.o .3; Emma Emmox .mimm THE PATHOGENESIS 0F COLIBACILLARY DIARRHEA OF THE NEWBORN An Electron Microscopic and. Histochemicai Study of the Disease in Neonatal ‘ Gnotohiotic Pigs Thesis for the Degree 'of Ph. D. MICHIGAN STATE UNIVERSITY BRUCE ROBERT CHRISTIE 1969 V: H {1518 Michigan 1 ' ..' .‘ ,;.. Uiu‘fi'hrfitt‘v J 21" J This is to certify that the thesis entitled THE PATHOGEIIESIS OF COLIBACILLARY DIARRHEA OF THE MMBORN An Electron Microscopic and Histochemical Study of the Disease in Neonatal Gnotobiotic Pigs presented by Bruce Robert Christie has been accepted towards fulfillment I of the requirements for Ph. D 6H“, 1 71 ' degree in _P_.““ ‘O‘OL'J' , ,7 j » z/ézz/4c4kc92fz/ z??? (6:? Major professor Date@?% 517% /7‘(“ 0-169 Imomo IV HMS G SOIS' 800K BINDERY INC LIBRARY IINDIRS S'IIIIPOIT Incum- ... A {ix/H1 3/ 3.4!. .v' ,y'p/(u (0/ g. ,5; I c / <1.» ' J ‘ . _ ABSTRACT THE PATHOGENESIS 0F COLIBACILLARY DIARRHEA OF THE NEWBORN An Electron Microscopic and Histochemical Study of the Disease in Neonatal Gnotobiotic Pigs by Bruce Robert Christie A major world problem in human and animal health is colibacil- lary diarrhea of the newborn. Little is known of the pathogenesis of this disease in neonatal domestic animals, and there have been few attempts to determine the pathogenesis of infantile colibacil- lary diarrhea. Pigs are considered suitable experimental animals for certain of the diseases of man, including that of infantile diarrhea and neo- natal diarrheal disease of domestic animals. Forty-eight gnotobiotic pigs were used in experiments in which 34 animals were fed between 1.8 and 3.0 x 106 Escherichia coli 0138:K81:NM organisms. Observations were made of the clinical effects of the infection. Pigs were killed from 4 to 96 hours after monocon- tamination. Observations were made of the macroscOpic signs of the disease. Tissues were taken at 5 levels of the small intestine and submitted to optical microsc0pic and electron microscopic examina- tion, and histochemical staining for cell inclusions and enzymes. Diarrhea occurred between 8 and 12 hours after monocontamina- tion with E. coli. Raised hair coat, projectile watery diarrhea and dehydration were also recorded within 18 hours of monocontamination. Bruce Robert Christie Optical microscopic lesions ranged from those of an acute enteri- t18~t0 a histologic picture, in persistently scouring pigs, similar to that seen in clinically normal germfree control animals. Vacuoli- zation and hydrOpic degeneration, microscopically indistinguishable from the vacuolization associated with absorption, were seen in both monocontaminated and germfree neonates. Electron microsc0pic examination revealed the basic difference between these two types of vacuolization. Absorption vacuoles were membrane enclosed, and were associated with normal function, normal cytologic enzyme activity and absence of diarrhea in germfree animals. Vacuolization of the mucosa of infected neonates was at first similar to that seen in the control animals, but within 12 to 20 hours of monocontamination there were diminution in form and histochemical function of the glycocalyx, loss of alkaline phosphatase and leucine amin0peptidase activity from the brush border, and degeneration of the microvilli. There were vacuolization of the cristae mito- chondriales leading to swelling and rupture of the mitochondria, vacuolization then lysis of the endOplasmic reticulum and Golgi complex, and eventual complete disruption of the internal cytOplasmic structure. Associated with these changes was a loss of activity of alkaline phosphatase, acid phosphatase, adenosine triphOSphatase, leucine aminoPeptidase, succinate dehydrogenase and lactate dehydrogenase. Effete cells generally were retained on the villal stalk by the persistence of the external cell membrane and the preservation of the mural interdigitations with neighboring cells. It is speculated Bruce Robert Christie that this retention could explain the histologic difference between colibacillary diarrhea and enteric diseases involving villal atrOphy in man and animals. In the monocontaminated neonates evidence of functional inactivity of the vasculature was associated with edema and infiltration of the interstices of the lamina propria with collagen fibrils. It is con- sidered that the endotoxic fraction of the E. coli enterotoxin is responsible for the inflammatory re3ponses observed in the villi. It is speculated (1) that the enterotoxin of E. coli 0138:K81:NM is readily absorbed by villal epithelial cells and is toxic for membrane-bound enzymes, (2) that it exerts its effects primarily on those energy producing systems involving adenosine triphosphatase -— the mitochondria, the endOplasmic reticulum and the Golgi complex, (3) that it is primarily toxic for the glycocalyx and secondarily toxic for this entity by its effect on the Golgi complex, (4) and that it is secondarily disruptive to the laminar organelles of the intestinal absorptive cell. It is considered that neonatal colibacillary diarrhea is pri- marily a disease of malabsorption. THE PATHOGENESIS 0F COLIBACILLARY DIARRHEA OF THE NEWBORN An Electron Microsc0pic and Histochemical Study of the Disease in Neonatal Gnotobiotic Pigs By Bruce Robert Christie A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Pathology 1969 Q @0297 f /;y_/[?—C]O For h h A; kl M. Ma h B A f a rt a us 8 ton ”cf/7’72” _ - , . M72. '7 .K‘ ACKNOWLEDGEMENTS I wish to express my appreciation to my major professor, Dr. Glenn L. Waxler, for his guidance, consideration, encouragement, understanding and practical help throughout the course of these studies. His energy and skill in helping to secure and raise the experimental animals was invaluable; his ready assistance at any hour during the course of the experiments was very much appreciated; and his patience with my enthusiasms and weaknesses, I will remember with gratitude. My sincere thanks to my academic committee: Dr. R. A. Fennell for his friendly help and advice in the histochemical aspects of this work; Dr. C. K. Whitehair, who first helped me to enter research involving gnotobiotic animals; and Dr. C. C. Morrill, Chairman of the Department of Pathology, Michigan State University, for providing facilities and supplies for this research. Dr. Robert F. Langham with his teaching of histOpathology opened many new fields to me. His dedication as a teacher, his skill as a pathologist, and his great warmth as a human being will long be remembered. The preserved and embedded tissues for electron microscopy were sectioned and stained by Mrs. June P. Mack, who with much patience and forbearance operated the EM-lOO whilst I hesitatingly entered the amazing world of electron microsc0py. I wish to record here my ii a indebtedness and thanks to her, and to Dr. Gordon C. Spink of the Electron Microsc0py Laboratory, Biology Research Center, Michigan State University. I am extremely appreciative of the secretarial assistance given by Mrs. Gretchen King throughout the five years of my graduate studies; of the technical help given by Mesdames Sunderlin, Miller, and Whipple,and Miss Patricia Lamb, of the histopathology laboratory; and particularly the conscientious and careful work of Mr. James Southern, gnotobiotic animal caretaker at the Veterinary Research Farm, Hagadorn Road, East Lansing. The kindness and help of Dr. E. Murray Pullar, Dr. A. K. Sutherland, Dr. Rodger Watson, and Dr. William A. Snowdon in sup- porting my application for graduate study in the United States of America will always be remembered by me. Again, the personal efforts on my behalf, of Dr. D. S. Wishart and Dr. D. M. Flynn of the Depart- ment of Agriculture, Victoria, when I sought study leave from the department is gratefully acknowledged. iii TABLE OF CONTENTS Page INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . 1 LITERATURE REVIEW. . . . . . . . . . . . . . . . . . . . . 5 9 Colibacillosis in Human Medicine. . . . . . . . . . . . Experimental Animal Models for Diarrheal Disease. . . . 12 Colibacillosis in Veterinary Medicine . . . . . . . . . l3 Colibacillosis in Swine. . . . . . . . . . . . . l4 Colibacillosis in "Specific Pathogen Free" 17 and Gnotobiotic Pigs. . . . . . . . . . . . . Extrinsic Factors Complicating Colibacillary Diarrhea . 21 The Mucosa of the Small Intestine . . . . . . . . . . . 22 The Villus . . . . . . . . . . . . . . . . . . . 29 Permeability of the Intestinal Mucosa. . . . . . 32 Macromolecular Absorption . . . . . . . . 32 Entry of Microorganisms to the Villal Mucosa . . . . . . . . . . . . . . . . 34 Absorption and Malabsorption. . . . . . . . . . . . . 37 Absorption of Water and Salts. . . . . . . . . . 38 Absorption of Carbohydrates, Proteins and Fats . 43 Histochemistry of the Intestinal Mucosa . . . . . . . . 47 Hydrolytic Enzymes . . . . . . . . . . . . . . . 48 The Phosphatases. . . . . . . . . . . . . 48 Alkaline Phosphatase . . . . . . . 48 Acid Phosphatase . . . . . . . . . 49 Adenosine Triphosphatase . . . . . 49 iv Page Proteolytic Enzymes . . . . . . . . . . . 50 Leucine Aminopeptidase . . . . . . 50 Oxidative Enzymes. . . . . . . . . . . . . . . . 51 The Dehydrogenases. . . . . . . . . . . . 51 Succinate Dehydrogenase. . . . . . 51 Lactate Dehydrogenase. . . . . . . 52 Cytochemical Stains for Cell Vacuoles and In- clusions and the Brush Border . . . . . . . . 52 The Pathology of Enteric Colibacillosis . . . . . . . . 53 In Human Infants and Calves. . . . . . . . . . . 53 In Conventional Pigs . . . . . . . . . . . . . . 55 In Gnotobiotic Pigs. . . . . . . . . . . . . . . 56 The Pathogenesis of Enteric Colibacillosis. . . . . . . 57 MATERIALS AND METHODS. . . . . . . . . . . . . . . . . . . . . 64 General Plan. . . . . . . . . . . . . . . . . . . . . . 64 Animals . . . . . . . . . . . . ... . . . . . . . . . . 65 Age of Experimental Animals. . . . . . . . . . . 66 Determination of Initial Sterility of Test Animals 0 O O O O O O O O O O O O O 0 O O O O 68 Control Animals. . . . . . . . . . . . . . . . . 69 Infective Agent . . . . . . . . . . . . . . . . . . . . 69 Infective Dose of Organisms. . . . . . . . . . . 71 72 NecrOpsy and Laboratory Procedures. . . . . . . . . . . Histopathologic Technique. . . . Histochemical Methods. . . . . . . . . . . . . . 76 Alkaline Phosphatase. . . . . . . . . . 76 Acid Phosphatase. . . . . . . . . . . . 78 78 Adenosine Triphosphatase. . . . . . . . . V Leucine AminOpeptidase. . . . . . . Succinate Dehydrogenase . . . . . . . Lactate Dehydrogenase . . . . . . Electron Microscopy Methods. . . . . . . . . Reclamation of the Test Serotype . . . . . RESULTS. . . . . . . . . . . . . . . . . . . . . . . . . . The Experimental Plan . . . . . . . . . . . . . . . Clinical and Gross Pathologic Findings. . . . . . . Recovery and Serotyping of the Organism . . . . . . Histopathologic Findings. . . . . . . . . . . . . . The Small Intestine. . . . . . . . . . . . . The Mucosa and Lamina Propria . . . . The Villal Epithelial Cell of the Small Intestine O O O O O O O O O O O O O The Submucosa and Other Structures of the Small Intestine. . . . . . . Enzyme H18 110011811118th 0 O O O O I O O O 0 O C 0 0 . Histochemistry of the Villus and Villal Epi- thelial cells 0 O O C O O O O O O O O O O The Phosphatases. . . . . . . . . . . Alkaline Phosphatase . . . . . Acid PhOSphatase . . . . . . . Adenosine Triphosphatase . . . Proteolytic Enzymes . . . . . . . . . Leucine Aminopeptidase . . . The Dehydrogenases. . . . . . . . . . Succinate Dehydrogenase. . . . Lactate Dehydrogenase. . . . . vi Page 79 79 79 80 81 82 82 82 87 87 87 90 105 117 117 117 117 117 124 136 141 141 145 145 154 Page Special Stain Techniques for IntracytOplasmic Con- tents and the Brush Border . . . . . . . . . . . . . 163 Oil Red 0 Stain. . . . . . . . . . . . . . . . . 163 Periodic Acid-Schiff Reaction. . . . . . . . . . 165 Best's Carmine Stain . . . . . . . . . . . . . . 165 Mayer's Mucicarmine-Stain. . . . . . . . . . . . 165 Alcian Blue Stain. . . . . . . . . . . . . . . . 165 Brown and Brenn Stain. . . . . . . . . . . . . . 166 Electron Microscopy . . . . . . . . . . . . . . . . . . 166 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . 255 Clinical Findings . . . . . . . . . . . . . . . . . . . 256 Histopathologic Findings. . . . . . . . . . . . . . . . 258 Histopathologic Lesions in Tissues Other Than the Alimentary Tract. . . . . . . . . . . . . 258 HistOpathologic Lesions in the Alimentary Tract O O I O I I O O O O O I O O O O O O O O 258 Villal Morphologic Change . . . . . . . . 266 Vacuolization of the Villal Epithelial cell 0 O O O I C I C C C O O O I O O O 269 Microbiologic Aspects . . . . . . . . . . . . . . . . . 270 Ligated Intestinal—100p Preparations. . . . . . . . . . 279 HistOCMistry. O O O O I O O O C I O O O O O O O O I O 281 Alkaline Phosphatase . . . . . . . . . . . . . . 282 Acid Phosphatase . . . . . . . . . . . . . . . . 286 Adenosine Triphosphatase . . . . . . . . . . . . 288 Leucine Aminopeptidase . . . . . . . . . . . . . 289 Succinate Dehydrogenase. . . . . . . . . . . . . 290 292 Lactate Dehydrogenase. . . . . . . . . . . . . . vii Electron Microsc0py . . . . . . . . . . . The Brush Border . . . . . The Glycocalyx. . . . . . . . . The Apical Trilaminar Membrane. The Terminal Web. . The Subapical Zone . . . . The Apical Tubule System. . . EndOplasmic Reticulum and the Apical The Nucleus. . . . . . . The Mitochondria . . . . . The Golgi Complex. . . . . Maturation and Necrobiosis thelial Cells . . . . . Mural Interdigitations . . The Lamina Prepria . . . . Absorption and Malabsorption. . . CONCLUSION . SUMMARY. LIST OF REFERENCES 0 o e o e o o o o o 0 VITA . . viii Vacuoles. Page 295 295 296 300 302 303 303 305 310 312 314 315 319 322 325 330 333 337 365 Table LIST OF TABLES Details of experimental animals . . . . . . . . . Time of exposure and necropsy of pigs . . . . . Preservation and fixation of tissues. . . . . . . Clinical appearance of pigs monocontaminated with E. 0011: 0138:K81:NM O O I O O C O O O O I O O O 0 ix Page 67 73 75 84 LIST OF FIGURES Figure Page 1 Pig 2258. Age 3 days. Terminal ileum. E. coli was introduced per 08 18 hours before necropsy C . O O C I O O C O U I O O O 0 O O 88 2 Pig 2257. Age 3 days. Duodenum. E. coli was introduced per 08 18 hours before necropsy. . . . . . . . . . . . . . . . . . . . . . 88 3 ‘ Pig 2257. Age 3 days. Duodenum. E. coli was introduced per 08 18 hours before necropsy O O O I O U 0 O O O O O O C O O 0 I O O O O 89 4 Pig 9047. Age 2-1/2 days. Duodenum. Germfree control. . . . . . . . . . . . . . . . . . 91 5 Pig 2253. Age 2 days. Terminal ileum. Germfree control. . . . . . . . . . . . . . . . . . 92 6 Pig 9046. Age 2 days. Terminal jejunum. Germfree control C I I I O O O O I O O O O O O O O O 93 7 Pig 9047. Age 2-1/2 days. Terminal ileum. Germfree control 0 C C . O O . O O O O O C O O O O O 94 8 Pig 1006. Age 4 days. Terminal ileum. E. coli was introduced per os 32 hours before necropsy . . . . . . . . . . . . . . . . . . .95 9 Pig 9047. Age 2-1/2 days. Terminal ileum. Germfree control. . . . . . . . . . . . . . . . . . 96 10 Pig 1010. Age 6 days. Terminal ileum. Germfree control. . . . . . . . . . . . . . . . . . 96 11 Pig 9046. Age 2 days. Terminal ileum. Gemfree control C O C O O O C l O C C C O O I O O O 97 12 Pig 1005. Age 3 days. Midjejunum. Germfree control C O C . . O . O C . C C C . O . O O . C C O 98 13 Pig 1306. Age 4-1/2 days. Terminal jejunum. E. 001i was introduced per 08 8 hours before necropsy. . . . . . . . . . . . . . . . . . . . . . 98 Figure 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 Page Pig 1004. Age 3-1/2 days. Anterior jejunum. E. coli was introduced per 03 28 hours before necropsy I I I I I I I I I I I I I I I I I I I I I I 100 Pig 0657. Age 30 days. Anterior jejunum. E. coli was introduced per os 48 hours before necropsy I I I I I I I I I I I I I I I I I I I I I I 101 Pig 1306. Age 4-1/2 days. Anterior jejunum. E. coli was introduced per as 8 hours before necroPSYI I I I I I I I I I I I I I I I I I I I I I 102 Pig 2255. Age 2-1/2 days. Terminal ileum. E. coli was introduced per os 16 hours before neCIopsy I I I I I I I I I I I I I I I I I I I I I I 103 Pig 2255. From the same field as Figure 17 . 104 Pig 9043. Age 3 days. Terminal jejunum. E. coli was introduced per os 18 hours before necropsy. . . . . . . . . . . . . . . . . . . . . . 106 Pig 9043. Age 3 days. Midjejunum. E. coli was introduced per 08 18 hours before necropsy. . . 107 Pig 9038. Age 2-1/2 days. Duodenum. E. coli was introduced per as 8 hours before necrOpsy . . . 108 Pig 0657. Age 30 days. Terminal jejunum. E. coli was introduced per 03 48 hours before necropsyI I I I I I I I I I I I I I I I I I I I I I 109 Pig 9046. Age 2 days. Duodenum. Germfree control I I I I I I I I I I I I I I I I I I I I I I 111 Pig 2262. Age 4 days. Duodenum. Germfree control . . . . . . . . . . . . . . . . . . . . . . 112 Pig 1305. Age 4-1/2 days. Duodenum. E. coli was introduced per 08 8 hours before necropsy . . . 113 Pig 9047. Age 2-1/2 days. Midjejunum. Germ- free control. . . . . . . . . . . . . . . . . . . . 114 Pig 9043. Age 3 days. Terminal ileum. E. coli was introduced per 08 18 hours before necrOpsy. . . 115 Pig 1306. Age 4-1/2 days. Anterior jejunum. E. coli was introduced per 08 8 hours before necrOpsy. . . . . . . . . . . . . . . . . . . . . . 118 Pig 0661. Age 29 days. Terminal ileum. Germ- free control. . . . . . . . . . . . . . . . . . . . 118 xi Figure 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Page Pig 1002. Age 3 days. Midjejunum. Germ- free contrOJ-o I I I I I I I I I I I I I I I I I I I 120 Pig 9046. Age 2 days. Duodenum. Germ- free control. . . . . . . . . . . . . . . . . . . . 121 Pig 9046. Age 2 days. Midjejunum. Germ- free control. . . . . . . . . . . . . . . . . . . . 122 Pig 9046. Age 2 days. Duodenum. Germfree contra]: I I I I I I I I I I I I I I I I I I I I I I 123 Pig 1005. Age 3 days. Anterior jejunum. Germfree control. . . . . . . . . . . . . . . . . . 125 Pig 1303. Age 4 days. Terminal jejunum. E. coli was introduced per 03 4 hours before necropsyI I I I I I I I I I I I I I I I I I I I I I 126 Pig 9041. Age 2—1/2 days. Duodenum. E. coli was introduced per os 14 hours before necropsy. . . 127 Pig 9041. Age 2-1/2 days. Terminal ileum. E. colt was introduced per 03 14 hours before necropsy. . . . . . . . . . . . . . . . . . . . . . 128 Pig 1009. Age 6-1/2 days. Terminal ileum. E. coli was introduced per os 96 hours before necropsyI I I I I I I I I I I I I I I I I I I I I I 129 Pig 9042. Age 2-1/2 days. Midjejunum. E, coli was introduced per 08 16 hours before necropsy. . . 130 Pig 9046. Age 2 days. Terminal jejunum. Germfree contrOI I I I I I I I I I I I I I I I I I I 132 Pig 9047. Age 2-1/2 days. Terminal ileum. Germfree control. . . . . . . . . . . . . . . . . . 133 Pig 9040. Age 2-1/2 days. Duodenum. .E. coli was introduced per os 12 hours before necropsy. . . 134 Pig 9044. Age 3 days. Terminal ileum. E. 001i was introduced per 08 20 hours before necrOpsy. . . 135 Pig 9045. Age 3 days. Terminal ileum. E. coli was introduced per 08 24 hours before necropsy. . . 137 Pig 9042. Age 2-1/2 days. Terminal ileum. E- 0011 was introduced per 03 16 hours before necropsy. . . . . . . . . . . . . . . . . . . . . . 138 xii Figure Page 46 Pig 9047. Age 2-1/2 days. Terminal ileum. Germfree control I I I I I I I I I I I I I I I I I I 139 47 Pig 9040. Age 2-1/2 days. Terminal ileum. E. coli was introduced per os 12 hours before necropsy. . . . . . . . . . . . . . . . . . . . . . 140 48 Pig 9043. Age 3 days. Terminal ileum. E. coli was introduced per os 18 hours before necropsy. . . 142 49 Pig 9046. Age 2 days. Terminal ileum. Germ— free contr01o I I I I I I I I I I I I I I I I I I I 143 50 Pig 9040. Age 2-1/2 days. Terminal ileum. E. coli was introduced per os 12 hours before necropsy. . . . . . . . . . . . . . . . . . . . . . 144 51 Pig 9042. Age 2—1/2 days. Duodenum. E. coli was introduced per 08 16 hours before necropsy. . . 146 52 Pig 9042. From the same field as Figure 51 . . . . 147 53 Pig 9044. Age 3 days. Terminal ileum. E. coli was introduced per os 20 hours before necropsy. . . 148 54 Pig 9046. Age 2 days. Duodenum. Germfree control I I I I I I I I I I I I I I I I I I I I I I 149 55 Pig 9046. Age 2 days. Duodenum. Germfree control . . . . . . . . . . . . . . . . . . . . . . 150 56 Pig 9046. Age 2 days. Terminal jejunum. Germ- free control. . . . . . . . . . . . . . . . . . . . 151 57 Pig 9047. Age 2-1/2 days. Terminal jejunum. Germfree controlI I I I I I I I I I I I I I I I I I 152 53 Pig 9038. Age 2—1/2 days. Terminal ileum. E. coli was introduced per 08 8 hours before necrOpsy . . . 153 59 Pig 9039. Age 2-1/2 days. Terminal ileum. E. coli was introduced per 08 10 hours before necrOpsy. . . 155 50 Pig 9041. Age 2-1/2 days. Terminal jejunum. E. 0017: was introduced per 08 14 hours before necr0psyo I I I I I I I I I I I I I I I I I I 156 61 Pig 9042. Age 2-1/2 days. Terminal jejunum. E. coli was introduced per os 16 hours before necropsy. I I I I I I I I I I I I I I I I I I I I 157 62 Pig 1005. Age 3 days. Duodenum. Germfree 158 control I I I I I I I I I I I I I I I I I I xiii Figure 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78. Pig 1002. Age 3 days. Duodenum. Germfree control I I I I I I I I I I I I I I I I I I I Pig 1005. Age 3 days. Terminal ileum. Germ- free contrOII I I I I I I I I I I I I I I I I Pig 1000. Age 3-1/2 days. Terminal ileum. E. 00113 was introduced per 03 24 hours before neCropsyI I I I I I I I I I I I I I I I I I Pig 1004. Age 3-1/2 days. Terminal jejunum. E. coli was introduced per 03 28 hours before necropsy I I I I I I I I I I I I I I I I I I Pig 1008. Age 4 days. Terminal jejunum. Midvillus. Germfree control. . . . . . . Pig 1008. Age 4 days. Terminal jejunum. Midvillus. Germfree control. . . . . . . . Pig 1008. Age 4 days. Terminal jejunum. Midvillus. Germfree control. . . . . . . . . Pig 1002. Age 3 days. Terminal ileum. Basal third of villus. Germfree control . . . . Pig 1002. Age 3 days. Terminal jejunum. Germfree control. . . . . . . . . . . . . Pig 1310. Age 4-1/2 days. Terminal jejunum. Midvillus. Germfree control. . . . . . . . . Pig 1306. Age 4-1/2 days. Terminal jejunum. Midvillus. E. coli was introduced per as 8 hours before necropsy . . . . . . . . . . . . . . Pig 2251. Age 2-1/2 days. Terminal ileum. Midvillus. E. coli was introduced per os 14 hours before necropsy . . . . . . . . . . . . Pig 1005. Age 3 days. Terminal ileum. Basal third of villus. Germfree control. . . . . . Pig 1310. Age 4-1/2 days. Terminal jejunum. Midvillus. Germfree control. . . . . . . . Pig 2253. Age 2 days. Terminal jejunum. Basal third of villus. Germfree control. . . . . . Pig 1310. Age 4-1/2 days. Terminal jejunum. Midvillus. Germfree control. . . . . . . . . xiv Page 160 161 162 164 168 170 172 174 177 179 181 183 186 188 191 193 Figure 79 80 81 82 83 84 85 86 87 88 89 90 91 92 Pig 1001. Age 3-1/2 days. Terminal jejunum. Midvillus. E. 00111 was introduced per os 24 hours before necropsy . . . . . . . . . . . . Pig 1001. Age 3-1/2 days. Terminal jejunum. Midvillus. E. coli was introduced per os 24 hours before necropsy . . . . . . . . . Pig 2251. Age 2-1/2 days. Terminal ileum. Basal third of villus. E2 coli was introduced per 08 14 hours before necrOpsy . . . . . . . . . . Pig 2251. Age 2-1/2 days. Terminal ileum. Basal third of villus. E. coli was introduced per os 14 hours before necropsy . . . . . . . . . Pig 2251. Age 2—1/2 days. Terminal ileum. Midvillus. E. coli was introduced per 08 14 hours before necropsy . . . . . . . . . . . . . . . Pig 2251. Age 2-1/2 days. Terminal ileum. Midvillus. E. coli was introduced per 03 14 hours before necropsy . . . . . . . . . . . . Pig 2251. Age 2-1/2 days. Terminal ileum. Midvillus- E. coli was introduced per os 14 hours before necrOpsy . . . . . . . . . . . . . Pig 2252. Age 2-1/2 days. Terminal jejunum. Midvillus. E. coli was introduced per os 14 hours before necrOpsy . . . . . . . . . . . . . Pig 2251. Age 2-1/2 days. Terminal jejunum. Midvillus. E. coli was introduced per 08 14 hours before necropsy . . . . . . . . . . . . Pig 2251. Age 2—1/2 days. Terminal jejunum. Midvillus. E. coli was introduced per os 14 hours before necropsy . . . . . . . . . . . . . . . Pig 1001. Age 3-1/2 days. Terminal jejunum. Midvillus. E. coli was introduced per 08 24 hours before necropsy . . . . . . . . . . . . . . Pig 1001. From the same field as Figure 89 . . . . Pig 1001. Age 3—1/2 days. Terminal jejunum. Midvillus. E. coli was introduced per os 24 hours before necropsy . . . . . . . . . . . . . Pig 2258. Age 8 days. Terminal jejunum. ‘Mid- villus. E. 0011 was introduced.per 08 18 hours before necropsy . . . . . . . . . . . . . . . . X? Page 195 197 199 203 205 207 209 211 214 216 218 220 222 224 Figure Page 93 Pig 1001. Age 3-1/2 days. Terminal jejunum. Midvillus. E. coli was introduced per os 24 hours before necropsy . . . . . . . . . . . . . . . 228 94 Pig 2258. Age 3 days. Terminal jejunum. Mid- villus. E. 00171 was introduced per 08 18 hours before necropsy . . . . . . . . . . . . . . . . . . 230 95 Pig 2251. Age 2-1/2 days. Terminal jejunum. Midvillus. E. coli‘was introduced per os 14 hours before necrOpsy . . . . . . . . . . . . . . . 232 96 Pig 1001. Age 3-1/2 days. Terminal ileum. Basal third of the villus. E. coli was introduced per 08 24 hours before necrOpsy . . . . . . . . . . . . 234 97 Pig 1001. Age 3-1/2 days. Terminal ileum. Basal third of the villus. E. coli was intro- duced per 08 24 hours before necrOpsy . . . . . . . 236 98 Pig 1001. Age 3—1/2 days. Terminal ileum. Basal third of the villus. E. coli was intro- duced per 08 24 hours before necrOpsy . . . . . . . 238 99 Pig 1001. Age 3—1/2 days. Terminal ileum. Basal third of the villus. E. coli was intro- duced per os 24 hours before necropsy . . . . . . . 240 100 Pig 1001. Age 3—1/2 days. Terminal ileum. Mid- villus. E. 001% was introduced per 08 24 hours before necropsy . . . . . . . . . . . . . . . . . . 243 101 Pig 1001. Age 3-1/2 days. Terminal ileum. Mid- villus. E. coli was introduced per os 24 hours before necropsy . . . . . . . . . . . . . . . . . . 245 102 Pig 1006. Age 4 days. Terminal ileum. Mid- villus. E- 001i was introduced per os 32 hours before necrOpsy . . . . . . . . . . . . . . . . . . 247 103 Pig 1006. Age 4 days. Terminal jejunum. Mid- villus. E. 001$ was introduced per os 32 hours before necropsy . . . . . . . . . . . . . . . . . . 250 104 Pig 1006. Age 4 days. Terminal ileum. Mid— villus. E. colt was introduced per 03 32 hours before necropsy . . . . . . . . . . . . . . . . . . 252 105 Pig 1009. Age 6-1/2 days. Terminal jejunum. Midvillus. E'- coli was introduced per os 96 hours before necrOpsy . . . . . . . . . . . . . . . 254 xvi PREFACE "I can't believe that!" said Alice. "Can't you?" the Queen said in a pitying tone. "Try again: draw a long breath, and shut your eyes." Alice laughed. "There's no use trying," she said: "one can't believe impossible things." "I daresay you haven't had much practice," said the Queen. "When I was your age, I always did it for half-an-hour a day. Why, sometimes I've believed as many as six impossible things before breakfast." Through the Looking-Glass Lewis Carroll (Charles Lutwidge Dodgson) 1832-1898 5"!‘1‘15‘ I INTRODUCTION There is a growing body of opinion in favor of the use of the domestic pig (She scrofh) as an experimental animal model for enteric diseases of man. Infantile colibacillary diarrhea is one of the major diseases of childhood in the underdeveloped countries of the world, and it is occasionally associated with high mortality in intrahospital epidemics in the United States of America and other of the so-called deve10ped countries, yet the pathogenesis of this disease remains undetermined. Colibacillosis is a term applied to a group of diseases'in man and the higher animals caused by infection with bacteria classified as Escherichia coli. This definition has been widened by some biologists to include infections of the mammary gland and urinary tract of these Species and coliform infections of birds. In this study, the term colibacillosis is used to refer to postnatal and weanling diarrheal disease of man and higher animals specifically associated with growth of E. coli in the gastrointestinal tract of the host. Pigs derived by hysterectomy or hysterotomy and maintained free of all detectable microorganisms, or specifically contaminated with known organisms, are called "gnotobiotic" pigs. The increased availa- bility of germfree and gnotobiotic pigs as experimental animals for investigation of infectious diseases has facilitated research into II. 4| 1 'lllll'l-l" II basic pathology and pathogenesis of diseases, uncomplicated by secon- dary infections and microflora in general and, most important, in the absence of nonpathogenic microbial flora of the same genus and species as the pathogen under investigation. Until.recent1y, pathologists studying colibacillosis have been confused somewhat by the presence of the microflora of the "normal" intestine and particularly by the presence of serotypes of E. coli of unknown pathogenicity. Very little significant work has been done to clarify the patho- genesis of colibacillosis of human infants despite the enormity of this public health problem. In veterinary research, an increasing interest has been shown in this field over the past five years. Since it is likely that the basic problem of pathogenesis can most successfully be investigated in experimental animal models, elucida- tion of the pathogenesis of the diarrhea of colibacillosis in an eXperimental animal such as the gnotobiotic pig should have applica- tion in the solution of the wider problem of colibacillary diarrhea 0f man and the higher animals. Recent studies of colibacillosis were concentrated on bacterio- logic and immunologic aspects of the disease. Some preliminary studies of the pathology and pathogenesis of colibacillosis in gno- tobiotes have been reported within the last two years and the sub- stance of this thesis is the direct development of one of these studies. In this laboratory, using E. coli 0138:K81:NM, the means whereby the orSanism established itself as an infection in the host following 0. both oral and subcutaneous contamination were established (Christie, 1967). Concurrent hematologic studies and observations of the clini- cal syndrome were made and recorded, relating these to the deve10pment of the infection. In addition, the gross and microsc0pic lesions of colibacillosis in gnotobiotes were recorded. From these observa- tions it was hypothesized that in coliform diarrheal disease -- the most commonly seen manifestation of colibacillosis in.neonata1 pigs -- the pathogenesis of the condition was most likely one of malabsorp- tion due to alteration of the absorptive component of villal epi- thelial cell activity. Fordtran (1967) defined diarrhea in terms of water and electro- lyte malabsorption as: interruption of the transport processes of the normal mucosal cell; deranged permeability of the intestine; the presence of nonabsorbable, osmotically active substances in the lumen of the intestine; and abnormal intestinal motility. There is, in addition, some evidence that the diseased villus loses excess fluid into the intestinal lumen, thereby contributing directly to the increased volume of chyme. The clinical state of diarrhea will depend on the region of the intestine involved in the disease process and on which of the above factors has initiated and is contributing to the disease process. In this thesis I propose to describe the villal changes contribu- ting to alteration of the complex dynamic process of fluid and electrolyte movement and, using as experimental models the gnoto- biotic pig and a specific serotype of E. coli, to speculate on the pathogenesis of diarrhea in enteric colibacillosis of human and ani- mal neonates. LITERATURE REVIEW "For out of olde feldes, as men seith, Cometh al this newe corn fro yeer to yere And out of old bokes, in good feith, Cometh al this newe science that men lere." Geoffrey Chaucer 1340—1400 Escherichia coli is an organism of many races, of great sero- logical variety and continually varying in virulence. It was Theobald Escherich, in 1885, who first found and described "Bacterium colt commune" from.the feces of breast-fed babies. His detailed description leaves no doubt that this was the organism we now know as E. coli (Sojka, 1965). Escherich did, however, consider this organism to be a harmless saprOphyte, and his published Opinions might have had considerable bearing on the fact that it has been only in recent years that a renewed interest has been taken in the potential of E. coli as an important pathogen of both man and animals. A natural diarrheal disorder of newborn calves, commonly called scours, was studied by Jensen in 1892. It was this man who pioneered the concept that there were enteropathogenic strains of a common intestinal organism (Jensen, 1892). Jensen distinguished between pathogenic and nonpathogenic strains of E. coli by oral exposure of neonatal calves with suspect strains (Barnum et al., 1967). Von Bahr in 1912 suggested that there was an etiological rela- tionship between E. coli scours in animals and human diarrheal disease and this theory was endorsed by Adam (1927). Theobald Smith and Orcutt (1925) studied the condition described previously by Jensen, and attributed it to special strains of colon bacilli which were subsequently identified bacteriologically and immunologically (Smith, 1927, 1928). Theobald Smith and others, from 1927 on, attempted to distinguish between pathogenic strains by bio- chemical means but were unsuccessful (Barnum at al., 1967). Certain serotypes of E. coli have been designated as "entero- pathogenic", this designation referring to those strains isolated from animals and man with diarrhea associated with the presence of these strains in the alimentary tract (Kohler and Bohl, 1966). Adam (1927) characterized these strains by their behavior in fermen- tation tests, but the first serious epidemiological study of human infantile gastroenteritis due to E. coli is attributed to Goldschmidt (1933), who identified enteropathogenic serotypes by serological methods. It was in 1927 that Adam isolated one or two strains of coli— form bacilli from human patients with infantile gastroenteritis (Barnum et a1., 1967). Further experiments with animals were carried out by Lovell in the United Kingdom, whose reports strengthened the concept of an etiological relationship of E. coli to diarrheal disease in man (Lovell, 1937; Lovell, 1951). Lovell (1937) employed precipi- tin tests and was able to distinguish serologically between patho- genic and nonpathogenic bovine strains of E. coli. 0f more recent import, in that it gave impetus to modern interest in human colibacillosis, is the work of Bray (1945) and Bray and Beavan (1948) in the United Kingdom, who demonstrated enteropatho- genic E. coli in 88% of cases of infantile diarrhea, but in only 3% of controls. Material gains in the understanding of the clinical and epidemi- ological significance of such infections were made possible by the publication of the serological classification of E. coli by Kauffman (1944, 1947) and by Kauffman and Dupont (1950). Their discovery of the "0" agglutination masking effect of the "L" antigen and subse- quent classification of the antigenic structure of E. coli according to the serological investigations of Kauffman - Knipschildt - Vahlne, was subsequently expanded by Ewing in the United States of America, and by ¢rskov in EurOpe (Sojka, 1965). This schema was based on the fact that three kinds of antigen occur in E. coli, and that these antigens can be selectively activated or inactivated by specific heat treatments. Since 1950 the classification of the organism according to sero- type has been widely accepted, and the literature pertaining to serotyping of E. coli antigens has been reviewed and tabulated by Edwards and Ewing (1962)- Recent excellent and succinct accounts of the bacteriology and serology of E. coli are available in the work of Sojka (1965) and Barnum.et a1. (1967)- Some serotypes of E. coli are pathogenic in their own right. Most serotypes, however, are of low virulence and cause disease because of either their overwhelming numbers or because the host has been weakened by other factors. Certain strains of E. coli produce powerful endotoxins capable of evoking tissue reaction when the organism invades tissues outside those of its normal habitat (Ribi at aZ., 1964), but one assumes that the power to invade tissues is not necessarily associated with the pathogenicity of the endotoxin of the organism. It is not yet known what factors are associated with virulence in a particular strain of E. coli, and there is no in vitro test for this property available at this time. Gay (1965) has recorded that, in a particular species of host animal, there is variation and over— lapping in the lethal dose between serotypes of E. coli considered to be pathogens and nonpathogenic strains from healthy animals. Studies on the pathogenicity of these strains, using laboratory animals or chick embryos, are of little value. In the experience of some workers (Saunders et aZ., 1960), a specific serotype may or may not be hemolytic, the conditions of incubation (particularly the carbon dioxide potential) being a determinant in this phenomenon. In general there is no confirmed relationship between virulence and O and K antigens. With serotypes isolated from "gut edema" and coliform gastroenteritis of swine, however, Gay (1965) recorded a relationship between these two antigens and pathogenicity. The most recent review of the identification and classification of the serotypes of the E. coli group of importance in veterinary medicine is well documented, brief and quite comprehensive (Gay, 1965). Colibacillosis in Human Medicine In human medicine the epidemiologic entity, infantile diarrhea or "weanling diarrhea" (Gordon at al., 1963),is generally observed with the transition of babies from a breast-fed existence to that of a mixed diet. This clinical syndrome is still considered one of the major health problems of the world, and in the underdeveloped nations is considered to be the major cause of death in infants less than two years of age (W.H.O. Report, International List of’Causes of Death, 1957; Gordon et aZ., 1963; Pearse, personal communication, 1968). In a number of authoritative studies the diarrheas and the dysenteries are recognizable as outstanding among the health problems of children (Wasti, 1959; Silhar and Maru, 1960; Carter, 1961). No more than an eighth of infants and young children of the world are free from appreciable risk of a lethal diarrhea (Cruick- shank, 1961), and despite the lesser freqency of this condition in school children and adults, a survey conducted by the Pan American Sanitary Bureau found this condition to be the leading cause of death in 8 of 17 countries studied (Anon., 1956). The weanling diarrhea syndrome includes several etiologic enti- ties. In most cases in babies no accepted pathogenic agent can be identified, and significant numbers carry agents of questionable disease-producing capacity or variously involved in multiple disease processes (Gordon et al., 1963). Although the clinical manifesta- tions of weanling diarrhea differ widely according to the environmental sanitation and the state of nutrition of the infant, the infectious 10 agents broadly encountered are much the same, suggesting that clini- cal variations relate mainly to host resistance and infecting dose. Not all acute diarrheas are of infectious origin, but most are (Public Health Dept., London County Council, 1953). Shigellosis or bacillary dysentery is the most common specific enteric infection among diarrheas of children (Hardy and Watt, 1945; Goodwin et al., 1960; Gordon, 1960), although ShigeZZa.is less common in the diarrheas of inhabitants of preindustrial countries where diarrheal disease is more prevalent (Gordon et aZ., 1963). Within recent years it has become increasingly apparent that a considerable porportion of the acute diarrheal disease of infants during their first year is due to particular enteroPathogenic strains of E. coli (Hodes, 1956; Shanks and Studzinski, 1952). Eacherichia coli is the causative agent of epidemic diarrhea of newborn infants in the United States of America. It kills 25 to 50% of affected babies in this age group. Escherichia coli also causes diarrhea in two- and three—year—old infants in the United States of America and is the major cause of infantile diarrhea in this country, constituting from 5 to 40% of all cases of diarrhea in the age group, according to the region in the U.S.A. that is sampled (Wheeler, 1967). In human colibacillosis enterOpathogenic serotypes are usually present only during the period of the illness (Kauffmann and Perch, 1943; Wallick and Stuart, 1943; Perch, 1944; Sears et a1., 1950). A causative relationship between implicated strains of E. coli and epidemic and sporadic occurrences of colibacillosis in humans was established by experiments in which adult volunteers and infants 11 were fed different strains of E2 coZi (Taylor et aZ., 1949; Neter and Shumway, 1950; Taylor and Charter, 1952; June at al., 1953; wright and Boden, 1953; Ross and Dawes, 1954; Ferguson and June, 1955; Kessner at aZ., 1962). These observations were made in coun- tries where general nutritional standards were adequate (Europe and U.S.A.), but there is confirmatory evidence of the role of colibacil- losis in diarrheal disease in other countries such as India (Anon., 1962), Indonesia (Joe et al., 1960), Guatemala (Gordon at al., 1962) and Chile (Rodriguez-Leiva, 1960). Intrahospital epidemics of diarrheal disease associated with EL 001i enteric infection are reported (Barnum et al., 1967) and representative series of sporadic cases among infants of a general pOpulation regularly include infections of this nature (Yekutiel, 1959). The ability to serotype enteropathogenic strains of E. coli enabled investigators to determine that they were relatively rare in the feces of normal infants (Giles et aZ., 1949; Ewing at al., 1955) but that there existed nonclinical carriers of pathogenic E. coli among the contact group of cases of infantile diarrhea due to organisms of the same serotype (Payne and Cook, 1950). Infections with E. coli can occur at any age but are particu- larly severe in infancy and in advanced life. The age distribution is probably due to lack of immunity in the young or general debility and increased susceptibility of the aged (Jubb and Kennedy, 1963). Ebcherichia coli sets up a symbiotic residence in the intestinal tract almost at birth, and it is thought that the organisms can 12 enter the tissues of the host from this source via the lymphatic route or by contamination of the body surfaces or through breaks in these surfaces (Hopps, 1966). In human medicine certain serotypes of this organism are impor- tant etiologic agents of focal pyogenic skin infections, omphalo- phlebitis, polyarthritis, meningitis, urinary tract infections, peritonitis, acute appendicitis, cholecystitis, cholangitis, and infectious biliary cirrhosis. In all of these sites E. coli evokes a nonspecific suppurative reaction with a degree of abscess forma- tion that is often indistinguishable from reaction of tissues to invasion with Staphylococcus spp. (Robbins, 1962). Experimental Animal Models for Diarrheal Disease The pathogenicity of strains of E. coli has been tested in embryonated eggs, white mice, guinea pigs, neonatal and weaned calves and "conventional" pigs and gnotobiotic pigs. Studies thus far have led to the conclusion that differences in the susceptibility of the test species, the source of the organism, the enhancement of viru— lence by host passage of the organism, the route of eXposure, the host defense mechanism (including intrinsic as well as extrinsic factors such as nutrition) and the serologic characteristics of the E3 coli strain, all lead to variation in the clinical and pathologic manifestations of infection. In 1966, Bustad emphasized the physiologic relationship of pigs to man and the value of these animals as models for human disease investigation (Bustad, 1966): 13 In earlier papers Dubos (1962) and Gordon et a2. (1963) specu- lated that the young germfree animal offered possibilities for explor- ing the result of the first invasion of organisms, ordinarily non- pathogenic for man, into the malnourished subject. Bustad and McClellan (1966) have reported a much closer physio— logic similarity between the domestic pig and man than that between the carnivores and man. Haelterman and HOOper (1967) consider that the baby pig is a particularly suitable model for infantile diarrhea studies, whereas Bustad has a broader opinion and considers pigs to be an ideal experimental model for many diseases of man (Bustad, 1966). Maronpot and Whitehair (1967) record their enthusiasm for the baby pig as a "useful eXperimental model for obtaining additional basic information on enteric disturbances" and go on to point out the par- ticular advantages of these animals especially in the investigation of enteric disturbances related to infection and nutrition. The unique qualities of the germfree neonatal pig as detailed by the above authorities, and in my own experience, make it an excel- lent tool for research into some or all of the abovementioned factors associated with infantile and weanling diarrhea. Colibacillosis in Veterinary Medicine In the veterinary field, Jubb and Kennedy (1963) summarized the prevailing concepts when they listed the three basic situations in which E. coli behaves as a pathogen. Certain strains act as primary pathogens; others of lower virulence act as pathogens only when they are present in numbers sufficient to overwhelm the host tissue de- fenses; and other strains act as secondary or opportunist invaders 14 which enter and produce, or combine with other organisms to produce, disease in a host tissue weakened by a previous invader or stress situ- ation. These authors (Jubb and Kennedy, 1963) list the predisposing factors in young non-primates: (1) inadequate housing with exposure to extremes of weather; (2) mistakes in dietary management (often overfeeding); (3) insufficiency of colostrum; (4) congenital weakness; and (5) specific vitamin deficiencies. In a study of "calf scours" Smith and Crabb (1956) could not find evidence of a specific strain of E. coli associated with an individual outbreak of the disease. They found, in the healthy con- trol group, 22 of the 23 phage types of E. coli that had been iso- lated from the diseased group. They also observed a change in the predominant phage type during the course of the disease. Colibacillosis in Swine Much confusion has resulted from the consideration by many workers that diarrhea of baby pigs is an entity rather than a clini- cal sign. The relationship between E. coli and porcine neonatal diarrhea has been accepted for a number of years. It was Jensen (1899) who first recorded that an infection by ER coZi could produce diarrhea in baby pigs. In 1934, there was described an outbreak of diarrhea in 3-day- old pigs that was considered to be a manifestation of E. coli infec— t1°n (McBryde, 1934). In his account, McBryde described duodenitis and jejunitis with no obvious lesions in other organs. He isolated Eh coZi from internal organs and mesenteric lymph nodes of these pigs. n-w 15 Glasser, Hupka and Wetzel (1950) described an acute outbreak of swine enteritis associated with E. coli infection as being characterized by diarrhea, anorexia and, at necropsy, catarrhal gastritis and enteritis with reddening and swelling of the intestinal mucous membrane. In their investigation of an acute diarrhea affecting pigs within 6 hours and up to a few days after farrowing, Gordon and Luke (1958) found that the pathogen involved was E. coli. The disease was char- acterized by high mortality in a litter that was apparently normal at the time of farrowing. They observed that, shortly after birth, the pigs became dull and were disinclined to suckle. A profuse diar- rhea (with yellow to orange feces) resulted in dehydration and was followed at the end of the first or second day by coma and death. Less acute cases were seen to have "scalded" hindquarters and, in some pigs, necrosis and sloughing of the tail. Gastroenteritis was a consistent lesion at necrospy. Almost pure cultures of E. coli were grown from large intestine and stomach samples, and occasional isolations of the same organism.were made from liver and spleen. Saunders et al. (1960), while investigating 58 outbreaks of disease in pigs in which E. coli was considered significant, came to recognize a clinically well defined syndrome affecting piglets within an age range of 48 hours to 2 weeks. The syndrome that these workers attributed to infection with E. coli was characterized by Purity, abundance and consistency of the serotype; absence of these features from other cases examined routinely during the same period was noteworthy. The findings of Saunders et a1. (1960) that E. coli 16 was not often recovered from viscera other than the intestines, even when there was a delay of several days between death and postmortem examination, concurs with the finding of Stevens (1961). In a recent review, Sojka (1965) listed 37 papers published since 1928 in which E. coli was considered to be involved in gastroenteritis and septicemia of young pigs. In a study of 100 pigs submitted to a diagnostic laboratory, Gossling and Rhoades (1966) obtained isolates of E. coli 0138:K81:NM from 6 pigs with "enteritis". These authors stated that "two sero- types, 08:K85:Hl9 and 0138:K81:NM, were identified as possible speci- fic etiologic agents of enteritis in baby pigs". In a comprehensive analysis of outbreaks of enteritis and diar— rhea over a 5-year period, Stevens (1963) was able to discern 3 groups in which a characteristic syndrome was associated with a Specific age group. These syndromes seemed to be associated with certain sero- types of E. coZi. These syndromes and the age groups were: (1) pig- let enteritis, 1 to 4 days of age; (2) enteritis of unweaned piglets, 3 weeks of age; and (3) postweaning enteritis of pigs, 10 to 12 weeks of age or shortly after weaning, manifested chiefly as "catarrhal enteritis". Stevens included hemorrhagic enteritis, edema disease and necrotic enteritis in his "post-weaning enteritis" group. The experimental reproduction of piglet enteritis with 2 sero— types frequently associated with field cases is recorded by Saunders at al. (1963). The disease resembled natural infections associated with these serotypes, diarrhea following the experimental exposure within 12 to 24 hours and some deaths occurring within 3 days. At 17 necropsy the same serotypes were obtained in profuse pure culture from the intestines and, at times, from other viscera. Varying degrees of ileitis and colitis, commonly associated with slight edema of the colonic mesentery, were found in the dead pigs, and mild gastritis was seen in approximately 1/3 of these animals. Colibacillosis in "Specific Pathogen Free" and Gnotobiotic Pigs Experimental reproduction of colibacillosis in "specific patho- gen free", "germfree" or gnotobiotic pigs is generally not well documented. Saunders et a2. (1963) used hysterectomy-derived, colostrum-deprived and artificially reared pigs in a series of experiments. In one eXperiment, using a known pathogenic serotype of E. coli (E68 type I) and oral doses ranging from 1000 to 100 x 106 organisms, they were able to produce diarrhea within 15 to 36 hours and recovery or, in some instances, death within 30 to 33 hours. In those pigs which died, the test organism was isolated from some or all of the tissues -- intestines, liver, brain and lung. Those pigs which recovered were killed 6 days after infection, and the organism was cultured at that time from some or all of 5 sites -- intestines, brain, heart blood and liver. In a second experiment, using a "non—pathogenic" serotype (P50) in 3 pigs, some diarrhea was observed within 20 hours of oral inoculation with l x 106, 100 x 106 and 100 x 106 organisms, reapectively. These pigs recovered quickly and were soon seen to be healthy with soft to normal feces. Two of these pigs were subse- quently inoculated orally with serotype E68, type 1, and died within \onv 18 2 and 5 days, respectively. The test organism was recovered in pro- fuse pure culture from the intestinal tract and in sparse but pure culture from the brain, heart blood and liver. In the final experiment of the series Saunders at al. (1963) found that the pathogenic serotype used in the previous eXperiments in neonatal pigs was equally pathogenic when used in 6-day—old pigs. They considered that there was no correlation between the severity of effects and the number of organisms administered and that the use of "pathogen-free" pigs, which were shown to be free of cytOpatho- genic viruses, indicated that their pathogenic serotype of E. coli was a primary pathogen. Using E. coli 08:K?:H21 in gnotobiotic pigs infected at 4 to 6 days of age, Kohler and Bohl (1966) detected a bacteremia as a constant finding for 3 days after infection. They attributed this bacteremia to the fact that the experimental animals were colostrum- deprived, since bacteremia was not detected in the clinically ill, conventionally reared pigs from which the strain of E. coli was originally isolated. These workers did not observe enteritis in gross examination of pigs which died in the acute phase of the disease. In a comprehensive study of the pathogenicity of E. coli, sero- type 083, Britt and Waxler (1964) used 7 litters of gnotobiotic pigs and recorded characteristic lesions in 78% of the 64 pigs exposed at ages ranging from 2 to 12 days. The outstanding lesion reported by these workers was a serofibrinous or fibrinopurulent polyserositis and polyarthritis. l9 Kenworthy and Allen (1966), in studies on germfree and gnoto- biotic pigs, demonstrated a variation in intestinal villus and crypt structure associated with the degree of bacterial contamination of the intestinal tract. The villi of the intestines of pigs monocon- taminated with E. coli were generally similar to those in the germfree control animals. Pigs contaminated with two serotypes of E. coli had branched and moderately foreshortened villi on the intestinal mucosa, and slight edema of the lamina propria. In pigs maintained in a "normal" environment, leaf-shaped villi were observed, with evidence of stunting, clubbing and fusion of villi. Those villi most severely affected were at the crests of the plicae. Cellular infiltration into the lamina prOpria was heavy, and reticular tissue appeared coarser and more abundant than in the monocontaminated pigs. These workers concluded that a biochemical interaction between the intestinal flora and the diet might result in the appearance of an altered metabolic pathway which, in turn, depended on the symbiosis or antagonism of the bacterial species present and the various sub- strates present in the intestinal tract. The morphology of the mucosa of the small intestine is dependent, in some measure, on these factors (Kenworthy and Allen, 1966), and the basic dynamic process of cell necrobiosis and replacement (Creamer, 1967). In three experiments, using 62 gnotobiotic pigs, Christie (1967) determined the clinical effects, and gross and microscOpic lesions of colibacillosis produced by 2 strains of E. coli -- 0138:K81:NM isolated in Michigan and 0138:K81 isolated in Minnesota. Precluding oral exposure of some test litters, he determined that both serotypes 20 readily colonized the intestinal tract from experimental infection simulating neonatal contamination of the umbilical stump. It was observed that bacteremia was produced within 1-1/2 hours of injec— tion, and by 24 hours the infection was generally established in the mucosa of the gastrointestinal tract. By 48 hours after injection the bacteremia had subsided so that only an occasional isolation from organs other than the gastrointestinal tract was made. The clinical signs of profuse watery diarrhea, dehydration and erect hair coat characteristic of colibacillosis in the neonatal pig were not appar- ent until the organism became well established in the gastrointestinal tract. The clinical signs of infection with E. coli 0138:K81 (Minne- sota origin) were significantly milder but edema of the subcutis was a constant feature of infection with this serotype. Oral exposure of 1 germfree litter to E. coli 0138:K81:NM (Michigan origin) pro- duced heavy colonization of the entire intestinal tract and diarrhea within 4 hours. Intermittent bacteremia was evident in this litter. No definite trends in body temperature or hematologic findings were observed, although in some animals there was a marked reduction in leukocyte counts (Christie, 1967). Christie (1967) reported that the gross necrOpsy findings, in general, agreed with those reported in the literature from field cases of colibacillosis in neonatal pigs. Histologically the lesions were predominantly in the gastrointestinal tract and ranged from an acute hemorrhagonecrotic enteritis to a histological picture, in persistently scouring pigs, microsc0pically indistinguishable from that seen in clinically normal germfree animals. He observed the 21 most common lesion to be in the villal epithelial cell and described it as a marked degenerative vacuolation of the cytoplasm that he termed "hydrOpic degeneration". He observed and described evidence of basic changes in apparent function of the villi, and hypothesized that these changes were responsible in part for malabsorption asso- ciated with colibacillary diarrhea (Christie, 1967). Extrinsic Factors Complicating Colibacillary Diarrhea Diarrhea is less likely to be caused by nutrient excess than it is by nutrient deficiency, as exemplified in pellagra and beri-beri. Protein deficiency is the outstanding factor in the etiology of kwashiorkor, and "kwashiorkor without diarrhea" is likely not to be kwashiorkor (Gordon et al., 1963). Recent attempts were made in Ontario to determine the role played by high and low energy diets in the development of enteric colibacillosis in experimental pigs (Barnum, D. A., personal communication). Although some changes in intestinal coliform p0pulations were observed, little of clinical or pathologic significance was recorded. In addition to colibacillosis and malnourishment, factors singly or severally associated with diarrhea in infants include Shigella spp., salmonella spp., many enteroviruses (including poliomyelitis virus, Coxsackie Group A, Coxsackie Group B, and many echoviruses), many adenoviruses, general infections, bacterial toxins, herbal medicines (used in tropical and underdeveloped countries), saline Purgatives and indiscriminate use of antibiotics. 22 The Mucosa of the Small Intestine The transfer of the products of digestion from the intestinal lumen to the blood or lymph is the process of intestinal absorption. In a normal animal the major factors involved in absorption are gastrointestinal motility, integrity of the intestinal wall and blood flow. Circular contractions of the muscularis mucosa have the effect of moving ingesta through the intestinal tube, but they also squeeze the villi together, diminishing the surface available for chyme-mucosal contact and decreasing the opportunity for absorption. The small intestine is the main organ for absorption of dietary nutrients; the villus, and the epithelial cells clothing the villus, constitute the functional absorptive unit (Kenworthy, 1967). The mucosa of the small intestine consists of that of the tubular crypts in continuity with that of the villi, the long finger-like processes which project into the lumen. The integrity of the anatomy and physiology of the villus is essential to normal function of the intestinal tract. Movements of the villus are considered to be both undulating and vertically contractile. These movements are the result of pas- sive movement with the flow of ingesta, active secondary movement as a result of movements of the intestine muscular coats in peristalsis, and active primary movement due to contraction of the Breucke fibers. Finger-shaped villi are said to "pump" actively (Verzar and McDougall, 1936). whereas leaf—shaped villi never show pumping movements (Ken- worthy, 1967). 23 Unlike the brain and kidney, organs of the body composed of almost permanent cell populations, the gastrointestinal tract epi- thelium works on the basis of rapid cell replacement. The most dynamic part of the gastrointestinal tract epithelium is the small- intestinal mucosa. The intestinal epithelial cell has, as D. H. Smyth (1967) says, "a short life, but a merry one, if merriment is the equivalent in cellular terms of a remarkable range of activi- ties possessed by few other cells in the body". The life of the mucosal cell is brief, for the whole epithelial lining is renewed every few days. The Specialized function of the crypt epithelium is that of cell division, so that as cells are lost from the terminal portion of the villus, they are replaced by a continual streaming of villal epithelial cells from the crypts. The villal epithelial cells arise from the repeated division of relatively undifferentiated cells in the crypts of Lieberkflhn, and once division has taken place dif- ferentiation is complete relative to eventual function (Kenworthy, 1967). The concept that the crypts are almost solely for epithelial cell production whilst the villi provide a large surface area almost wholly for absorption was first positively prepounded by Bizzozero (l888),was amply confirmed by the exhaustive studies of Leblond and various co-workers (1948, 1955, 1958) and has been recently summarized in an excellent review (Creamer, 1967). The cells lost from the tips of the villi disintegrate rapidly so that little is ever seen as whole cells, or even recognizable residue, in the lumen of the small intestine (Leblond and Messier, 24 1958). Indeed, this disintegration in the small intestine of the normal human is so rapid that cells labelled with [3H] thymidine did not survive long enough, once shed into the lumen, for eXperi— mental biopsy in the cell migration studies reported by Deschner at al. (1963) . The very fast turnover of intestinal epithelial cells in general is illustrated by the following examples of the shortest and longest recorded turnover times, the shortest time being that of the villal epithelial cells of the ileum of the mouse -- 24 hours (Creamer, 1967) -- and the longest being that of the villal epithelial cells of the duodenum of man -- 5 to 6 days (MacDonald, Trier and Everett, 1964). Lipkin (1965) has estimated that the time taken for complete replacement of the epithelial lining of the gastrointestinal tract in man is 3 to 6 days. A more graphic expression of this dynamism is the determination of the flow-rate of villal epithelial cells in "positions per hour". Creamer (1967) has recorded a flow-rate for the human ileum at 0.8 cell positions per hour and for the human jejunum at 1.0 cell position per hour. For man and for mammals the turnover time is "remarkably constant" (Creamer, 1967). Germfree animals have an "abnormal" small intestine compared to that of conventional animals, that is, animals not reared in germfree conditions. Germfree animals have shorter crypts, much longer villi and a paucity of lymphoid and inflammatory tissue in the epithelium and the laminae prOpriae(Dubos, 1966). The villal epithelial cell turnover time is approximately twice that of the conventional animal, and the generation cycle of crypt cells in the germfree animal is 25 slower than that of the conventional animal (Abrams at al., 1963; Lesher et al., 1964). Kenworthy and Allen (1966), in studies on germfree and gnoto- biotic pigs, demonstrated a variation in intestinal villus and crypt structure associated with the degree of bacterial contamination of the intestinal tract. The villi of pig intestine monocontaminated witth. coli 0141:K8Sa,c(B):H4 were generally similar to those in the germfree control animal. Pigs contaminated with E. coli sero- types 0141:K85a,c(B):H4 and 08:H had intestinal villi with con- siderable variation in shape, including some branching forms. Kenworthy and Allen (1966) introduced a third group of axenic pigs into a piggery and reported that these animals developed a large, mixed, intestinal flora and a striking change in the mucosa with long crypts, increased mitoses and short villi, the picture, accord- ing to Creamer (1967), characteristic of very rapid cell turnover. Smyth (1967) has raised the question of mode of desquamation of cells in the zone of extrusion. Alternative suggestions are that either cells at the tap of the villus are pushed off by the continuum of new cells streaming from the crypts, or they are produced in response to a need to replace those sloughing at the apex. Smyth (1967) suggested that histochemical and microscopic studies of cells at the very tip of the villus, should they reveal basic pathologic change, would suggest an answer to this problem. Intestinal absorption is complex in a number of ways -- iron and Vitamin 812 are absorbed into the epithelial cells and kept there for a variable time, then later released from the cell, through the 26 basement membrane into the lamina propria. Dipeptides, disaccharides and glycerides are absorbed into the villal epithelial cell and hydro~ lyzed intracellularly. Conversion of fructose to glucose, glucose to lactic acid, and glutamic acid and eSpartic acid to alanine are well documented examples of the detailed metabolic digestion proces- ses occurring within the villal epithelial cells as an integral part of the serosal transfer of nutrients (Levin, 1967). Apart from the absorption of protein, carbohydrates and lipids the villal epithelial cell, being provided with membranes permeable to electrolytes and water, exchanges fluid and electrolytes with its environment in order to facilitate absorption of digestion products and as a means of stabilizing cell volume. Intestinal epithelial cells, in normal function, move substances , against their electrochemical potential and for this purpose use the energy of metabolism mediated through adenosine triphosphate synthe- sis and breakdown. Since all cells are required to regulate indi- vidual cell composition and volume, they utilize the energy of metabo— lism for osmotic work, a process aptly called "chemi-osmotic coupling" (Mitchell, 1961). The mechanism of membrane transport in relation to intestinal absorption is one of great complexity and to this date, investigation has been primarily directed to the elucidation of the individual steps of membrane transport in model systems such as the erythrocyte, nerve axon, and cholecyst epithelium (Diamond and Tormey, 1966; Smyth and Whittam, 1967). 27 Molecules and molecular aggregates such as soluble enzymes, intermediates of metabolism, and organelles carrying a surface charge function as electrolytes to which the villal epithelial cell membrane is impermeable. As a consequence there is an osmotic stress which tends to favor the imbibition of fluid by the villal epithelial cell. In a normal cell there is a stabilization of volume dependent upon the regulation of the concentration of diffusible ions within the cell in two ways; by active transport systems -- or "pumps" -- for ions, particularly of sodium and potassium, and loss of ions through leaks in the cell membranes. The Operation of the pumps requires a continual supply of free energy. Interference with, or destruction of such energy-producing systems, leads to a cessation of pump activity, a loss of cell potassium and increased intracytOplasmic sodium, chloride and water (Robinson, 1960; Parsons, 1967). In 1957, Lifson and Parsons demonstrated in the small intestine of experimental rats that sodium transport depended on aerobic glycolysis, and that this glycolysis was stimulated by glucose or mannose but not by acetate, pyruvate or lower fatty acids. They observed, however, that in the ileum and colon, fluid absorption was stimulated, under aerobic conditions, by the addition of glucose and acetate, pyruvate and the lower fatty acids (Lifson and Parsons, 1957). A direct coupling exists between inward movement of sodium at the brush border of villal epithelial cells and accumulation and transport of sugars and amino acids (Crane, R. K., personal communi- cation). Parsons (1967) considered that the absorption of either 28 sodium, or sugars and amino acids is stimulated when active absorp- tion of the other is proceeding, and he concluded that the energy for sugar and amino acid transport is derived from the movement of sodium and potassium ions down an electrochemical gradient that is maintained by cellular metabolism involving the hydrolysis of energy- rich substances such as adenosine triphosphate. Membrane adenosine triphOSphatases implicated in sodium and potassium tranSport have been described in the brush border and other bordering membranes of the villal epithelial cells and Parsons (1967) postulated that there is an outward oriented sodium pump in the brush border of the anterior jejunum. The obvious practical need for such a pump is that more sodium, than that provided by diet and exocrine digestive systems, is required in the chyme to drive the absorption of sugars and amino acids. Within the villal epithelial cell, alterations in the energy- producing systems of the mitochondria could lead to inability of these organelles to excrete metabolic water. Such distention of the mitochondria has been reported (Rouiller, 1960) and is considered to be the cause of the finely granular cytoplasm of the cell under— going cloudy swelling. Still further distention of the mitochondria will continue to that point at which the organellar membrane rup- tures releasing its store of metabolic water to the cytoplasm, and eventually resulting in the phenomenon of hydrOpic degeneration of the cell (Robinson, 1960)~ 29 The Villus Under the optical microscope the structure of the villus is deceptively simple. The villus is clad with a simple columnar epi- thelium bearing a faintly discernible brush border of microvilli. In the neonatal pig the brush border is more apparent in the duodenal and jejunal epithelial cells than it is on those of the ileum. InterSpersed amongst the villal epithelial cells are goblet cells, increasing in relative number caudally. The epithelium rests on a clearly demonstrable basement membrane. The lamina propria of the villus consists of connective tissue cells and noncellular stroma, capillaries, lymphatic vessels, a central chyle vessel termed a "lacteal" (it can be single or multiple), contractile fibers superficially resembling smooth muscle fibers (termed Breucke fibers), and variable numbers of reticuloendothelial cells and lymphocytes. It is not uncommon to find marked vacuolation of the villal epi— thelial cells in germfree and monocontaminated neonatal pigs. This vacuolation is more marked in infected pigs and is more common in sections taken from more caudal sites in the small intestine. These vacuoles have been described as the result of either normal intestinal absorption or of necrobiosis in the germfree pig and the more marked vacuolization seen in pigs monocontaminated with E. coli, as absorp- tion vacuolization, evidence of necrobiosis, or hydroPic degeneration associated with the toxic activities of the enteric organisms (Christie, 1967). 30 Other than the basic cellular structures -— nucleus, cytOplasm, cell wall, brush border, various vacuoles including fat droplets, hyaline bodies, glycogen and other substances there is little else in the villal epithelial cell that is visible to the Optical microsco- pist. The advent of the electron microsc0pe has enabled cytologists to observe the absorptive cell in much greater detail. As would be expected from the knowledge of the highly specialized functional demands of this cell, the ultrastructure is quite complex. The brush border consists of microvilli about 1 u long and 0.1 u in diameter (Granger and Baker, 1949).‘ In the jejunal epithelial cell of the 3-week—old pig, the individual microvilli are enclosed within a hirsute layer, the "glycocalyx". The glycocalyx is not present in neonatal pig villal epithelial cells, and Staley et al. (1968a) suggested that it was synthesized after whole protein absorp— tion ceased since it was present in the absorptive cell of the 3- week-old pig. A trilaminar membrane constituting what is virtually the basement membrane for the glycocalyx is continuous over the microvilli. The filamentous central core of each microvillus is termed the "cytoskeleton" and runs from the tip, through the shaft, to the base and into the apical cytOplasm, condensing in that region into well-defined "rootlets" (Staley at al., 1968a). Microvilli occasionally have a common base and in the crypts of Lieberkfihn and at the tips of the villi, the microvilli are shorter and some— what club-shaped. At the level of the rootlets, in the extreme apical cytOplasm of the cell, there is a mass of interlacing fibers around the rootlets, constituting what is called the terminal web *5— _,A..-.. .. w 31 (Palay and Karlin, 1959). The terminal web is not clearly apparent at the apex of the jejunal villal epithelial cell of the neonatal pig (Staley et al., 1968a). The subapical zone of the jejunal villal epithelial cell was observed by Staley et al. (1968a) to be free of organelles but with a faintly granular appearance. Apical tubules were seen, particu- larly in cells near the tip of the villus. These tubules were usually parallel, but did branch and anastomose, and penetrate the cell a distance of 2 or 3 u. The tubules were slender; near the luminal plasmalemma they were 30-60 mu wide but were sometimes dilated by their contents to form vacuoles. It was observed that these tubules had a trilaminar membrane 95 to 100 A thick with fine ex- tensions projecting into the lumen of the tubules. The apical tubules were absent from the absorptive epithelial cells in the 3- week-old pigs. It is clear that the apical tubule system described here corresponds to the "smooth-surfaced cisternae" described by Kenworthy et al. (1967). Staley et al. (1968a) described vacuoles in the jejunal villal epithelial cells as being prominent in the apical two thirds of the villus and, within the individual cell, generally small in the supranuclear position and larger below the nucleus. These workers suggested that the apical vacuoles originated either by pinocytosis or by dilatation of the apical tubular system; that these large vacuoles delaminated into the lumen forming "prominent myelin-like whorls"; and that they contained "little material except free mem— branes". The so-called "basal vacuoles" were of a somewhat 32 different nature in that they contained a flocculent material that was "either dense and homogeneous or clumped in bizarre arrangementsfl' Staley et al. (1968a) described the Golgi apparatus of the vil- lal epithelial cell of the neonate as being subnuclear and associated with the basal vacuoles. In less mature villal epithelial cells, the Golgi apparatus was in a paranuclear position. In the jejunal absorptive cells of pigs at 21 days of age, the Golgi apparatus was peripheral to the nucleus in the apical cytoplasm. In neonates, the mitochondria and granular endoplasmic reticulum were present at the base of the cell, whereas in pigs at 3 weeks of age, these organelles were generally diSpersed throughout the cell (Staley et al., 1968a). The other ultrastructural features of villal epithelial cells, namely desmosomes, mural invaginations, intramembranous space, smooth endOplasmic reticulum, lysosomes, micropinocytotic vesicles, multi- vesicular bodies, ribosomes, and the various intranuclear structures, have been described elsewhere in consummate detail in a number of standard references (Rhodin, 1963; Ham, 1965; Fawcett, 1966). Eggmeability of the Intestinal Mucosa Macromolecular Absorption. Since there is no transplacental movement of antibody in swine, cattle, or sheep, consumption of colostrum by the neonates of these species precedes a rise in serum gamma globulin in the first day of life to levels usually found in adult animals. These globulins are apparently absorbed intact into the lacteals through the intestinal 33 epithelium during the first 36 hours of life. The ability of the neonate to absorb intact proteins generally is lost between 8 and 16 hours of life (Brambell, 1958; Clark, 1959). This phenomenon of "closure" (Lecce, 1966) was considered due to decrease or loss of pinocytotic activity. According to Barnum et al. (1967), the pro- teins (other than immunoglobulins) absorbed from the gastrointestinal tract of newborn herbivora prior to closure are cleared from the renal circulation, resulting in proteinuria during the second and third day after birth. Comprehensive studies of the intestinal absorption of protein by neonatal (but not gnotobiotic) pigs have been reported by Mattisson and Karlsson (1964, 1965, 1966) and by Pierce and Smith (1967), who demonstrated the predominance of the jejunum over the duodenum and ileum in the intestinal absorption of immune lactoglobulins. Mattisson and Karlsson (1966) described a nongranular membrane system of tubules and vacuoles in the apical cytOplasm of jejunal epithelial cells. They further observed that, after colostrum feed- ing, lipid and granular electron dense material, suggestive of pro— tein, was present in pinocytotic tubules and vacuoles in the apical cytoplasm of the cells. The concept of absorption of proteins by the intracellular tubu- lar system.is not new. It was described in the epithelium of guinea pig yolk sac by Dempsey in 1953, in the neonatal jejunal cell of rats and mice by Clark (1959), in the absorption of ferritin by the intestinal cells of rats and rabbits (Graney, 1965; Kraehenbuhl et al., 1967), in the proximal convoluted tubules of the kidney 34 (Maunsbach, 1966) and presumptively in the intestine of 1—day—old puppies (Anderson, 1965). In neonatal pigs fed colostrum 3 to 6 hours earlier, Staley at al. (1968b) observed in duodenal villal epithelial cells protein- containing vacuoles, up to 10 u in diameter and "almost entirely filling the cytOplasm". These authors demonstrated, however, that duodenal villal cells no longer contain visible intracytOplasmic colostral vacuoles 48 hours after exposure to colostrum, and that at this time epithelial cells cannot be distinguished from mature cells. The jejunal absorptive cell of the neonatal pig is charac- terized by a poorly developed hirsute layer over the microvilli, an abundance of pinocytotic vacuoles and apical tubules, and a sub- nuclear location of the Golgi apparatus. The apical tubules are lined with Spinous processes and are retained only until the pig is first fed, or if unfed, for approximately 42 hours (Staley et al., 1968b). The disappearance of the tubular system after feeding sug- gests that this is the system involved with the absorption of pro- tein macromolecules in the newborn pig. The decreased ability to absorb proteins was suggested by Payne and Marsh (1962) to be a sign of maturation of the villal epithelial cell, as does the development of normal enzyme (particularly alkaline phosphatase) content in the cell during its migration from the crypt and maturation en route along the shaft of the villus (Wilson, 1962). Entry of Microorganisms to the Villal Mucosa The neonatal intestinal epithelial barrier to intraluminal bac- teria has not been defined, and the ease with which macromolecules w..- v Q... 'iuu ... . D... *u. ... 35 enter the neonatal villal epithelial cells indicates the possibility of a similar route of entry for microorganisms. Dixon (1960) was the first to clearly demonstrate the ability of viable bacteria to pass intact through the apparently normal intestinal epithelium. Using guinea pigs preconditioned by starvation and the admini- stration of opium, Takeuchi and Sprinz (1967) gave salmonella typhi- flmrium organisms per as and observed subsequent changes in the mucosa of the ileum. The enteritis was characterized histologically by blunting, swelling, and foreshortening of villi; elongation of crypts and emptying of gOblet cells; and increased and abnormal extrusion of epithelial cells, compensated in part by increased mitoses. Although the continuity of the epithelial lining was retained, the individual epithelial cells became flattened and their brush border attenuated concomitant with loss of cytoplasmic components. Inflammatory changes of the lamina propria were also reported by these authors (Takeuchi and Sprinz, 1967). Takeuchi and Sprinz (1967) described the process whereby salmonella typhimurium adhere to the brush border, disrupt the microvilli, and enter the cytoplasm of the apical portion of the cell as membrane-enclosed bodies. Within 14 hours the number 0f bacilli within the villal epithelial cell had decreased, but many were seen in the lamina propria. By 48 hours after challenge, there were reduced numbers of bacteria in the lamina propria; in some instances degeneration of the phagocytized bacilli was noted; in others the organisms persisted in the lamina propria and even proliferated Within macrophages. Associated with these bacterial migrations there _-_._ w W- 5‘ 36 were a number of significant morphologic changes in the villal epi- thelial cells. At first, the brush border was reduced in height; later the microvilli of affected cells were completely denuded. The short tubules of the granular endOplasmic reticulum were transformed into long curvilinear patterns or concentric whorls with intermittent vacuolar dilatations, and there were fewer free ribosomes present. Coincident with the epithelial changes, progressive intravascular clotting of blood in the microvasculature of the lamina prOpria was observed, associated with fragmentation of cellular membranes, slough- ing of endothelial cells and the extension of plasma fluid into the extravascular Space (Takeuchi and Sprinz, 1967). Staley et al. (1968c, 1968d) observed the migration of E. coli 055:BS:H7 organisms through the mucosa of the small intestine of the neonatal pig. It was seen that the organisms were aligned along the surface of a villal epithelial cell in close proximity to a goblet cell and became attached to the brush border in contact with the apices of the microvilli. In 20-hour-exposed ileum, all bacteria (nonfimbriated prior to the 20—hour interval) both in the lumen and attached to the villal epithelial cell, had a well developed fimbri- ated coating. The microvilli nearest the bacterium exfoliated to form a microerosion -- the authors termed it a "microulcer" -- which developed into an invagination of the plasma membrane containing the organism. The organism became fully enclosed by the "pinching off" of the invagination, to form an intracytoplasmic vacuole. The tri- laminar structure of the vacuole wall was lost and appeared just as a single thickened membrane. The microvilli of the nearby cells were 37 Shortened and swollen and appeared as though "budding". The organisms within vacuoles then could move throughout the cytOplasm down to the basement membrane, but were not seen within the basement membrane, the intercellular spaces or the lamina propria. The enclosed organisms apparently did not come into association with lysosomes, nor was there evidence of formation of secondary or compound lysosomes. There was no evidence of intracellular digestion of E. coli observed by these authors (Staley et al., 1968c, 1968d). Absorption and Malabsorption Absorption implies movement of fluid substance across the apical plasmalemma through the cytOplasm (possibly involving passage through the compartmental boundaries of organelles) and out of the cell at the serosal pole, which can be defined as those boundaries of the cell not in contact with the intestinal lumen or adjacent cells. The substances abSorbed by the intestine are water, hydrophilic substances in solution, and lipids in various forms. Movement of molecules and ions through the villal epithelial cell (so-called "translocation") implies a functional asymmetry of the cell, with transport systems distributed asymmetrically between the two faces bordering the two fluid compartments between which transport is occurring. It has been shown by Newey (1967) and by Fisher (1967) that absorption of sugars and amino acids required active pumping systems in the limiting membranes of the villal epi— thelial cells, and that for translocation of these substances it was necessary that they be maintained in the mucosal cell cytoplasm at 38 a concentration greater than the extracellular environment so that they would leak out at the Opposite face of the cell. Parsons (1967) recorded that with sodium, translocation required that the cytOplasmic sodium ion concentration be lower than that of the environment, so that sodium would leak into the cell at one face, then outward from the cell through a pump at the Opposite face. In normal circumstances, the duodenum and jejunum differ from the ileum in that the former will add or subtract sodium Chloride from the chyme until isotonicity with blood is established, then net absorption of the isotonic solution occurs. In the ileum, net absorption of sodium chloride will occur, even fronlhypotonic solu- tions (Vischer at al., 1944; McHardy and Parsons, 1957; Code, 1965; Parsons, 1967). Parsons (1967) has summarized the proof of active transcellu- 1ar transport of sodium and the evidence for membrane function in translocation of sodium, potassium, sugars and amino acids. Absorption Of Water and Salts Intestinal epithelial cell membranes are basically lipoidal in nature but allow non-lipids to penetrate by two pathways: through aqueous channels, or pores, in the membranes; or for some electro- lYtes (and non—electrolytes such as glucose) tranSport via a membrane carriers. There are no carriers for water, so its absorption is ex— clusively through these aqueous channels (Fordtran, 1967). The forces controlling passive movement of ions and water through the nenbrane pores are: electrochemical gradients, osmotic pressure gradients and "solvent drag" (Fordtran, 1967). ., v' ....- v... 39 It is of significance that, in man, pore size decreases with increasing distance from the ligament Of Treitz. In the duodenum and anterior jejunum, active tranSport of non-electrolytes such as glu- cose and amino acids constitutes the major driving force for sodium absorption and, in turn, water flow with its particular tendency to further absorb sodium ions by solvent drag. Sodium transport in— volves the active processes of bicarbonate absorption or hydrogen ion secretion. An active transport of sodium per se has been pro- posed for the duodenum and jejunum, but the ease of diffusion of the sodium ions back through the relatively large pores into the intestinal lumen makes this a very inefficient process (Fordtran, 1967). In the ileum and colon, however, the Opposite is true. The effective aqueous pore size is approximately 3A, and actively trans- ported sodium molecules are effectively restrained. Passive move- ment and solvent drag are relatively unimportant in sodium ion absorption in the ileum. However, the rate and direction of sodium movement affects net anion movement. In the anterior jejunum, bicar- bonate is absorbed in preference to chloride and against concentra- tion gradients whereas, in the ileum and colon, chloride is absorbed against steep concentration gradients and is in part coupled to bicarbonate secretion. Fordtran (1967) maintains that this recipro- cal anion movement is brought about by an exchange carrier in the membrane. In the proximal jejunum, potassium is passively absorbed whereas, in the ileum and colon, potassium is secreted (Fordtran and Locklear, 1966). 40 The data of Watten at al. (1959) and of Cooke (1957) indicated that as diarrhea increased in severity there was a progressive rise in fecal sodium and a decrease in fecal potassium. They showed that the degree of sodium loss varied according to the cause of the diar— rhea, whereas potassium loss relative to water loss seems more or less fixed regardless of the cause of the diarrhea. In a number of diarrheal diseases of man, including laxative diarrhea and "infantile gastroenteritis", the sodium and potassium concentrations in feces, together, exceed the total chloride content (Evanson and Stanbury, 1965). The type of food ingested is a major determinant of the direc- tion and rate of water and electrolyte movement in the intestine. Those characteristics of chyme most important to absorption being osmolality, the rate of addition of osmotically active solutes (the products of intraluminal digestion) and the relative concentrations of absorbable and nonabsorbable solutes in the chyme (Fordtran, 1967). In infantile diarrhea, fecal osmolality has been-recorded to be half the osmalality Of plasma (Teree et al., 1965). The source of fecal osmolality is considered to be malabsorption of hydrolyzed dietary products and sodium and potassium ions (Fordtran, 1967). Fordtran explained this phenomenon as possibly being due to a de- crease in size Of the aqueous channels in the jejunal absorptive cells and a marked loss of jejunal surface area coupled with oral inges- tion of hypotonic fluid. An increased rate of movement of ingesta through the intestinal tract is an almost invariable consequence of enteritis. Under this 41 circumstance sodium is less able to be reabsorbed and, as a result, sodium loss exceeds potassium loss during the course of infantile diarrhea (Darrow, 1946). It has long been known that diarrhea, even of slight duration, will result in depletion of sodium that is by far the major extracellular base in the body (Butler et al., 1933; Black, 1953). Relatively more of an infant's body weight is "highly labile" extracellular space. Sodium loss from this component leads to water loss, reduced extracellular fluid volume, disturbed renal function and eventual shift of water from an intracellular to an extracellular disposition (Darrow, 1945). In consequence, cellular dehydration involves loss of cell potassium to tissue fluids and serum (Elkinton et al., 1948). Shock and cellular anoxia have been Observed at 30 to 50% falls in levels of extracellular electrolyte (Darrow, 1946). At a particular but undetermined cell potassium concentration, sodium and hydrogen ions emigrate from serum to cell according to an approximate but not quantitative inverse correlation in the osmotic activity of these cations (Elkinton et al., 1948). There occurs, further, a metabolic acidosis as serum sodium levels fall. Potassium moves from the cells to the serum as a con— sequence of this (Smith and Etteldorf, 1961). Serum potassium so derived is lost in renal excretion (Darrow, 1945). Pitts (1963) has summarized the sequence as follows: a loss of extracellular sodium leads to acidosis, a decrease in extracellular volume, then a decrease in cellular volume with loss of cellular potassium. Intracellular migration of sodium occurs and bicarbonate follows sodium into the 42 cell, producing an intracellular alkalosis and an extracellular acidosis. Falling serum bicarbonate levels stimulate renal mechanisms to increase serum chloride levels in an attempt to maintain electri- cal balance, but in addition this accentuates extracellular acidosis. The end result, of grave consequence to the patient, is depletion of cellular potassium. It is Obvious that abnormalities of the villal epithelial cells, and change in the intestinal absorptive surface area, will alter the rate of absorption of water, electrolytes and other products of digestion. The enteric diseases of man and animals now recognized to be malabsorptive in nature are celiac disease, idiOpathic steator- rhea, nontrOpical sprue, tropical sprue (Wiseman, 1964), transmissible gastroenteritis of swine (Trapp at al., 1966; Maronpot and Whitehair, 1967) and sprue—like syndrome in dogs (Vernon, 1962; Kaneko at al., 1965). In these diseases the actual loss of cells and foreshortening and "clubbing" of the villi, in addition to the ultrastructural changes in the microvilli and the intracellular organelles, provide readily acceptable reasons for the malfunction Of the absorptive mechanism. In colibacillosis, villal changes under Optical microscopy are slight, so slight in fact that Moon at al. (1966) stated that "diarrhea occurs in E. coli infection without morphologic evidence Of enteri- tis...." This Opinion coincides with that of Stowens (1966), who reported that the intestines of babies dead from uncomplicated coli- bacillary diarrhea were without Signs of inflammation. It is only in recent years that some authors (Kenworthy and Allen, 1966; Christie, 1967; Barnum at al., 1967) have described changes in the villi or the lamina prOpria associated with E. coli infection. 43 Although there has been much evidence, particularly with ligated intestinal loop preparations, of the diarrhea of colibacillosis being associated with a massive outflow of fluid from the circulation to the intestinal lumen, it has been recently hypothesized that another possible major factor in the pathogenesis of colibacillary diarrhea has been malabsorption due to hydropic degeneration of the villal epithelial cells (Christie, 1967). Also Fordtran at al. (1967) reported that, although cell loss does Significantly reduce the absorptive surface area Of the jejunal mucosa in Sprue, the permea- bility of the remaining mucosal cells is markedly decreased to levels lower than that seen in the normal ileum. The resulting malabsorp- tion of water, sodium chloride, and other small, water soluble solutes contributes to the diarrhea associated with the abnormal osmotic pressure and ionic content of the chyme (Fordtran at al., 1967). Absorption of Carbohydrates, Proteins and Fats In vitro intestinal techniques have made the intestinal epithelial cell available for detailed study of a number of systems such as amino acid and hexose tranSport, which systems also are found in kidney and other tissues. It is apparent from the work of Auricchio at al. (1963) and Dahlqvist at al. (1963) that a major component of maldigestion and malabsorption is deficiency in the carbohydrases of the intestinal absorptive cell. Basically, the transport of carbohydrates from the lumen of the intestine through the brush border into the villal epi— thelial cell depends in part on the presence and function of the carbohydrases: alpha-dextrinase (isomaltase), maltase, lactase, and out . 44 sucrase (Gray, 1967). .Gray (1967) presented evidence that complete digestion of disaccharides is accomplished by mucosal enzymes at the luminal surface. Specific carbohydrases are located at the luminal portion of the columnar absorptive cell (Miller and Crane, 1961). Oligosac- charides -- alpha-dextrins, maltotriose, maltose, and others -— resulting from intraluminal digestion must be further hydrolyzed during physical contact with the intestinal mucosal surface, Other- wise carbohydrate malabsorption will occur. Any mechanism prevent— ing contact of the substrate with the brush border, or alternatively, disruption of the luminal border enzymes either spatially or chemi- cally, will interfere with carbohydrate absorption. Glucose comprises the bulk of monosaccharide that must be ab— sorbed,and the villal epithelial cell appears to have a Specific carrier mechanism.with which glucose or galactose can combine, and by which monosaccharide can enter the otherwise impermeable lipo- protein membrane of the cell (Gray, 1967). The specificity of intestinal absorption is such that the slight Structural difference between glucose and mannose is sufficient to determine that the intestinal epithelium will absorb the former and reject the latter. Such specificity requires special mechanisms, and one hypothesis is that there are present at the cell membrane carrier sites to which particularly structured chemical groups can attach. The attachment of large hydrophilic substances to lipid carriers may partly explain the movement of these complex particles, such as vitamin B into the cell. It is of interest to note that 12’ 45 the large vitamin B12 molecule cannot enter the epithelial cell until it is attached to its carrier molecule which is itself a larger molecule (Matthews, 1967). There are well documented specific acquired, secondary and con- genital carbohydrate malabsorption states associated with enzyme inadequacy, disaccharidase deficit, lactase deficiency and sucrase- isomaltase deficiency, and these have been tabulated by Gray (1967) in their associations with 14 of the described human malabsorption syndromes, including "infant diarrhea". In infantile diarrhea, and its associated disaccharidase de- ficiency, sugar malabsorption leads to bacterial conversion of unab- sorbed carbohydrates to organic acids which accumulate and cause metabolic acidosis (Weijers at al., 1961; Torres-Pinedo at al., 1966). After "closure" mammals, including man, can absorb no more than traces of intact protein or peptide (Dent and Schilling, 1949). Pro- teolytic enzymes hydrolyze dietary protein to peptides and amino acids. Peptides are then bound to or taken up by the microvilli where, with peptidases, they are further hydrolyzed to amino acids (Holt and Miller, 1961). Amino acids are absorbed by the mucosal cells -- most of them by an active system -- and transported to the capillary blood (Wiseman, 1953). Pancreatic lipase and bile salts interact with dietary trigly- cerides in the intestinal lumen, thus hydrolyzing them to fatty acids and monoglycerides. Bile salts are essential in the formation of micelles which serve to effectively solubilize fatty acids and mono- glycerides and lead to a clear solution rather than an emulsion 46 (Hoffman and Borgstrom, 1962; Isselbacher, 1967). Fatty acids and monoglycerides in the micellar phase are then brought to the surface of the villal epithelial cell, attach thereto, and the micelle dis- charges its lipid contents such that they enter the cell in a molecu— lar manner rather than in a micellar form (Isselbacher, 1967). This is a nonenergy-dependent process (Strauss, 1966). The monoglycerides and fatty acids are reesterified to triglycerides in reactions medi- ated by enzymes in clearly defined reactions (Senior, 1964) within the agranular reticulum of the apical cytoplasm of the villal epi- thelial cell (Porter, 1969). All of the glyceride synthesizing enzymes are located in the microsomes, and more specifically, in the smooth end0p1asmic reticu- lum, and it is significant that as soon as fat is seen beneath the terminal web it is almost always surrounded by membranes which appear to be components of the end0p1asmic reticulum (Isselbacher, 1967). Lipid passes through the basal intercellular wall of the villal epi- thelial cells in the form of a chylomicron, a central triglyceride core bound by an outer coating of protein, cholesterol and phospho- lipid (Zilversmit, 1965). Lipid dr0p1ets accumulate in the region of the Golgi zone prior to their exit from the cell (Dobbins, 1966) and in this area each droplet is enclosed by a membrane most representative of the rough endoplasmic reticulum. Isselbacher (1967) speculated that the "entire Packing process" of the chylomicron takes place while the triglyceride is surrounded by endoplasmic reticulum -- "the smooth providing the triglyceride synthetase mechanism, the rough the protein synthetic machinery". 47 Histochemistry of the Intestinal Mucosa As recently as 1966, Kohler and Bohl have stated, "the illumination Of the mechanism by which diarrhea and subsequent dehydration are caused without concurrent enteritis will undoubtedly require further extensive study. In fact, histochemical studies may well be required to detect changes other than morphological changes in the affected cells" (Kohler and Bohl, 1966a). The relationship between functional activity Of the villal epi— thelial cell and its changing cytochemistry has not been studied to any significant extent. According to Pearse and Riecken (1967), "...the evidence presented is restricted to a number of cytochemical phenomena associated with or related to absorption in the normal or diseased mucosa, and the significance of these is often obscure". Many cytochemical phenomena observed in studies of the small intestine still lack biochemical interpretation (Pearse and Riecken, 1967). Much of the histochemistry of the intestinal tract has been con- cerned with the establishment of normal distribution patterns of enzymes —— the hydrolytic, oxidative and proteolytic enzymes -- and cell inclusions such as lipids, mucopolysaccharides, amyloid and others. The distribution of enzymic activities and reactive groups bound to membranes has enabled workers such as Pearse and Riecken (1967) to map, histochemically, the functional organellar pattern of villal epithelial cells. The accuracy of such methods compared with electron micrOSCOpy is at present being studied. The relationship Of these results to organellar function is suggested by some histo- Chemists, but not yet established (Pearse and Riecken, 1967). There is no scientific literature, known to this author, concern- ing the enzyme histochemistry of the neonatal gnotobiotic pig. .... ....— In. 48 Hydro lgtic Enzymes The Phosphatases. The hydrolysis of phOSphoric acid esters such as glycerol-1— phosphate to glycerol and inorganic phosphate is catalyzed by the phosphatases. They are in fact esterases, but this term is retained for those enzymes that hydrolyze esters of the carboxylic acids. Alkaline Phosphatase.--Relatively few of the known enzymes at the present time can be conclusively demonstrated in animal tissues by histochemical means. Two of the alkaline phosphatase methods that are available are for phosphomonoesterase I, and for nonspecific alka— line phosphatase. Alkaline phosphatases hydrolyze most orthOphOSpho- monoesters at an optimum pH Of 9.0 to 9.6. PhOSphate esters, namely glycerophosphate, glucose-l-phOSphate, creatinine phosphate (and other phosphoamide linked esters), diphosphate esters such as hexose diphosphate, and the nucleotides (adenylic acid and adenosine tri— phosphate) are Split by alkaline phosphatase activity (Gomori, 1949; Feigin and Wolfe, 1955; Friedenwald and Gryder, 1958). It is possible that alkaline phosphatase is a number of closely related enzymes, all of which split phosphate at an alkaline pH (Emmel, 1950; Burgos et al., 1954; Burgos at al., 1955). The probable function of both acid and alkaline phosphatase is the dephosphoryla- tion necessary for absorption, tranSport and control of metabolism, possibly in the synthesis of some esters, and the maintenance of levels of intracellular inorganic phOSphate sufficient for osteo— genesis (Morton, 1961). 49 Alkaline phosphatase has been described as being present in the villal epithelial cells; at the apical border in tissues from man (Shnitka, 1960; Padykula at al., 1961), Specifically localized in the brush border in cells from man (Pearse and Riecken, 1967), in the brush border and apical cytOplasm of 3- to 5—week-old conventional pigs (Maronpot and Whitehair, 1967) and at the apical border in 1— to 2-week-Old conventional pigs (Thake, 1968). Acid Phosphatase.—-PhOSphomonoesterase II and nonspecific acid phosphatase hydrolyze a variety Of phOSphoric esters, all being hydrolyzed at an Optimum pH of 5.0. There are several enzymes that constitute this group (Barks, 1960; Barka, 1961a; 1961b; 1961c). Acid phosphatase has been described as being present in the villal epithelial cells in a number of locations: in the lysosomes confined to a narrow zone beneath the terminal web in man (Fric and Lojda, 1964; Riecken et al., 1966), as a loose scattering of granules in the apical cytOplasm in the rat and the mouse (Pearse and Riecken, 1967), continuous from the "villous apex to the crypt base except for goblet cells" of 3- to 5-week-old conventional pigs (Maronpot and Whitehair, 1967), and predominantly in the apical cytoplasm in l- to Z‘WEER-Old conventional pigs (Thake, 1968). Adenosine TriphOSphatase.--At an alkaline pH, adenosine triphosphate is hydrolyzed to adenosine diphosphate and phosphoric acid in the presence of adenosine triphosphatase. This reaction is notable for the release of free energy with the breaking of the phos- phate bond. More than one enzyme can break this bond, including 50 nonspecific alkaline phosphatase, but each active enzyme essentially requires calcium or magnesium ions for activation (Hunt, 1966). Adenosine triphosphatase has been described as being present in these locations in the villal epithelial cells: in the brush border and mitochondria in man (Pearse and Riecken, 1967), from the base to the apex of the villus, and specifically in the brush border and throughout the cytOplasm, in 3- to 5-week-old conventional pigs (Maronpot and Whitehair, 1967), and uniformly throughout the cells in 1- to 2-week—old conventional pigs (Thake, 1968). Proteolytic Enzymes. The group of proteolytic enzymes is divided into 2 major sub- groups, the proteases (proteinases and endopeptidases) and the pep- tidases (exopeptidases) which hydrolyze the —CO-NH- linkages of proteins and peptides. Peptidases generally do not attack proteins but act on peptides or peptide derivatives which contain free carboxyl or amino groups. Leucine amingpeptidase.-—This enzyme is a proteolytic enzyme found in high concentrations in the liver, pancreas, kidneys and small intestine (Ticktin and Trujillo, 1966). Sites of leucine aminOpepti- dase activity have been reported to occur in the columnar cells of the rat, human and monkey small intestine (Nachlas at al., 1957; Nachlas et al., 1960). It is specifically localized at the level of the microvilli in man and does not appear until after birth (Pearse and Riecken, 1967). Maronpot and Whitehair (1967) reported leucine aminopeptidase to be localized in the villal epithelium, specifically 51 in the brush border, with strong activity in the supranuclear cyto- plasm of the villal epithelial cell. They did not detect enzyme activity in the crypts of LieberkUhn in 3- to 5-week-old conventional pigs. Oxidative Enzymes The Dehydrogenases. The oxidative process involves either the addition of oxygen to a substance or the loss Of hydrogen (and/or the loss of electrons) from that substance. The enzymes concerned with biological oxida- tion through hydrogen removal are termed dehydrogenases, the hydrogen generally being passed through the cytochrome system and combined with oxygen to form water. Succinate dehydrogenase.--This enzyme catalyzes the trans- formation of succinic acid to fumaric acid in the tricarboxylic acid cycle (Krebs cycle) with oxidation of the succinate fraction and the transfer of the hydrogen ions to the electron transport substance flavine adenine dinucleotide. Nachlas et al. (1957) have reported high levels of succinic dehydrogenase activity in the epi- thelium of the duodenum, jejunum and ileum, moderate activity in Brunnefs glands and the colonic epithelium, and slight activity in Peyer's patches. These reactions were observed in tissues from both rats and dogs. All of the positive reactions were confined to intra- cytoplasmic structures resembling mitochondria. 52 Pearse and Riecken (1967) found succinate dehydrogenase at three levels within the cytOplasm of the intestinal absorptive cells of man. These three zones corresponded to the distribution pattern of the bulk of the mitochondria, separated by the Golgi region and nucleus. Succinate dehydrogenase has been described as being present in these locations in the villal epithelial cells: in the subnuclear epithelial cytOplasm, both on the villi and in the crypts, of 3- to 5-week-Old conventional pigs (Maronpot and Whitehair, 1967), and throughout the cytOplasm, but more concentrated toward the base, in l- to Zdweek—Old conventional pigs (Thake, 1968). Lactate dehydrogenase.-~Lactate dehydrogenase is a glyco- lytic enzyme that catalyzes the reversible reaction of pyruvate to lactate in the presence of diphosphopyridine nucleotide. Although lactate dehydrogenase activity has been demonstrated in urine and other body fluids, the pathways of degradation and excretion are still Obscure (Ticktin and Trujillo, 1966). Lactate dehydrogenase is very widely distributed in a high pro- portion of the mitochondria of most tissues, including mucosal cells of the gastrointestinal tract of rats (Nachlas et al., 1958; Hess et al., 1958). There are no other reports of lactate dehydrogenase activity's being used as a measure Of mitochondrial activity in the villal epithelial cells in health or disease. cytochemical Stains for Cell Vacuoles and Inclusions and the Brush Border The literature relating to the demonstration of lipoidal inclusions and hydrOphilic substances within cells and for specialized structures, 53 such as the brush border, is voluminous and beyond the scOpe of this review. Complete histologic and biochemical details are available from many sources. The most encyclopedic recent review is available in the text of Thompson and Hunt (1966). The Pathology Of Enteric Colibacillosis In Human Infants and Calves The difficulties inherent in Obtaining suitable tissues from human patients with colibacillary diarrhea whose condition is uncom- plicated by malnutrition, secondary infection, multiple microbial contamination or medical treatment is readily understandable. The literature of the pathology of enteric colibacillosis in man is inadequate as exemplified in the papers of McKay and Wahle (1955), Handforth and Sorger (1961) and Stowens (1966). Stowens (1966), in describing the pathology of colibacillary diarrhea (epidemic diarrhea of the newborn) due to E. coli 0111, stated "... not only is there complete absence of signs Of inflamma- tion either in the bowel or elsewhere, but there appears to be a depression of the lymph nodes." He defined lymph node depression as depopulation of lymphocytes with disappearance of fOllicles and germinal centers. In 2 of 3 unrelated epidemics, extensive virologic studies of these dead infants were pursued with negative results. Stowens considered lymph node depression to be significant in epi- demic diarrhea of the newborn in that he has not observed this phe— nomenon "in any other disease in infants" (Stowens, 1966). "I? 54 In a recent experiment, Osborne (1967) studied the lesions pro- duced in neonatal calves following a composite inoculum of E. coli serotypes O78:K-:NM, 015:54:NM, 09:35:10, 026:60:11 and 0119:69z4. He suggested that the pathogenesis was primarily a disseminated intra- vascular coagulation with thromboembolism being wideSpread in the microvasculature of the lungs, liver, kidneys and the gastrointestinal tract. In his experimentally infected calves a severe gastroenteri- tis developed, and 18 of 33 calves died. Osborne has speculated that bacterial endotoxin moves across the intestinal membrane, pri- marily by passive cliffusion, that the direct action of the E. coli endotoxin activates the clotting mechanism by the release of a plate- let factor (Factor III),or that the endotoxin activates Hagemann factor (Factor XII) which acts as a trigger mechanism, in the presence of other clotting factors, for intravascular coagulation. Osborne also reported that E. coli endotoxin has been shown to effect directly selective vasoconstriction, and anticipating that this phenomenon would most likely favor thrombus formation where small vessels branch, considered that vasoconstriction at these sites was contributory to the thromboembolism that he observed in the pulmonary, renal and Splanchnic microvasculature of experimental calves. His conclusion was that blockage of the microvasculature caused local tissue and vascular endothelial hypoxia, and in turn necrosis; or relaxation of vascular tissue cells, vasodilatation, endothelial cell separation, edema and hemorrhage. Such "overwhelming pathOphySiological changes apparently stabilize in an irreversible state and the internal cellu- lar environment becomes incompatible with life" (Osborne, 1967). 55 There are important similarities between Osborne's findings and those of McKay and Wahle (1955), who described the pathology Of colibacil- losis as observed in 9 infants who had died during an epidemic of infantile diarrhea associated with infection by E. coli 0111:B4. In Conventional Pig§_ Barnum at al. (1967) stated, "The histological appearance of the intestinal tract of piglets with colibacillosis is usually normal", but they qualified the statement somewhat by claiming that there is occasionally a catarrhal enteritis and that after prolonged diarrhea there can be atrophy of intestinal villi. They described the villi as being short and blunt, the mucosa approaching that normally found in the colon, the normal eosinophilic columnar epithelium of these atrophic villi being replaced by a basophilic, cuboidal epithelium. Their description concluded with, "The atrophic change is similar to that described for 'sprue' in man and transmissible gastroenteritis (TGE) in piglets" (Barnum.et al., 1967). This similarity was also drawn by Maronpot and Whitehair (1967). Barnum at al. (1967) de- scribed similar atrOphic changes of the villi of the small intestine in their report Of lesions in weaned pigs with persistent coliform diarrhea. In addition they referred to hyperemia and sloughing of the mucosa of the gastric fundus associated with infarction by granu- lar thrombi lodged in the mucosal veins and capillaries. Their description of weanling colibacillosis of swine included that of an acute catarrhal enteritis with an accumulation of epithelial and inflammatory cells mixed with masses Of bacteria, fibrin and mucus ., s. ca .- 56 on the mucosal surface and accumulation of inflammatory cells in the lamina propria. Kenworthy at al. (1967) have reported their observations of ultra— structural changes in jejunal, apical-zone villal epithelial cells of conventional pigs with postweaning diarrhea associated with E. coli 0147:K89(B),K88a,c(L):Hl9 and other unnamed serotypes. They described absence of microvilli or microvilli "enormously variable in length" and saw no evidence of rootlets or terminal web, or else "a disorganized zone of Shadow immediately below the microvilli". In animals with mild diarrhea the terminal web remained organized, but the distance between the upper surface of the web and the cell membrane was increased. There was distortion of the free surface of the cell membrane and obliteration of microvilli, mitochondrial bodies were elongated and few in number and they described degenera- tion of the cell membranes whereas the desmosomal attachments remained in situ. There was loss of organelles from the cytoplasm reminiscent of the lesions of Shigella spp. infection in the guinea pig as described by Takeuchi at al. (1965). lgfiGnotobiotic Pig§_ A mild neutrOphilic infiltration of the laminae prOpriae of the villi of gnotobiotic pigs in response to specific infection with E. coli 0138:K81:NM was reported by Christie (1967). At times aggregates of neutrophils in the lamina propria formed microabscesses that dis- tended the shaft of the villus. This observation has since been con- firmed in gnotobiotic pigs infected with E. coli 08:K.:H2l in that a .\. I 1 ems. ne- ‘ I 1" 57 mild neutrophil infiltration of the villi of the small intestine was a consistent finding (Kohler, 1967). Both authors observed that the infiltration was very mild and was only apparent when compared with tissue sections from germfree control animals (Christie, 1967; Kohler, 1967). It is apparent, as would be expected, that different serotypes of E. coli would produce different lesions and of different degrees of severity. Christie (1967) described a range of lesions extending from a mild generalized hydropic degeneration to an acute hemorrhago- necrotic enteritis. Kohler (1967) reported that enteritis was not detected in his experimental animals and described only mild neutro- phil infiltration of the cecum and colon and the presence of small numbers of the organisms, usually in the capillaries. The Pathogenesis of Enteric Colibacillosis There have been no definitive studies of the pathogenesis of colibacillary enteritis in the human subject. It is only very recently that an attempt has been made to differentiate, in terms of pathology, the diarrheal diseases of man, and those studies of any Significance are few. Experimental colibacillosis in domestic and laboratory animals provides a significant basis for speculation, and confirmation of such speculation, on the pathogenesis of colibacillary enteritis and diarrheal disease associated with E. coli infections. In studies of experimental colibacillosis using a particular sero- type, Saunders at al. (1960) found the organism to be present in all can out 'o is la. n- a ‘ to. (II (a 58 the viscera of some baby pigs but only in the intestines of others in the same group. This phenomenon was observed when 2 other sero- types were used, and these workers concluded that bacteremia was a stage in the development of the infection. Jubb and Kennedy (1963) listed the favored Site of localization of E. coli as the intestinal tract and described the infection as septicemic since it is possible to isolate the organism from all tissues. Localization of E2 coli in the meninges, joints and lungs was described by these authors, who pointed out that, although these infections accompany the enteric phase, their manifestations appear later than does diarrhea. Moreover, umbilical infections might never Show themselves as enteric infections but remain as tissue infections. The favored Sites of localization of E. coli in the tissues are the meninges, joints, the lungs -- especially if infectious porcine pneumonia lesions are present -- and probably the kidney, in which site it produces the interstitial nephritis known as "white spotted kidneys". The renal infection may produce a descending pyelonephritis from which the organism is recoverable. It is evident from bacteriological examination of cases of coli- septicemia that systemic invasion can occur from the umbilical stump, nasal or pharyngeal mucosae, or from the intestinal tract (Fey and Margadant, 1961; Fey, 1962; Jubb and Kennedy, 1963). Fey at al. (1962) produced an E. coli septicemia by oral or intranasal infection of 3 colostrumedeprived neonatal calves despite ligation and section Of the esophagus. A fourth calf, with esophagus left intact and eXposed intranasally to the same serotype, had an 59 Eh coli bacteremia 20 hours after challenge, but the organism was not obtained from the contents of the small intestine, collected by laparotomy, nor from bile. These workers concluded that, in calves, infection of the intestine was by the circulatory system rather than by the digestive system. A number of workers have reported that strains of E. coli from cases of infant epidemic diarrhea have the ability to cause diarrhea in adults as well as in infants when the organism is fed in sufficient numbers (Neter and Shumway, 1950; Ferguson, 1956). Stevens (1963) has classified the syndrome Of porcine colibacil- lary diarrhea into 3 groups, namely, "piglet enteritis", "enteritis of unweaned pigs -- three-week enteritis", and "post-weaning enteri- tis". The so—called "three-week enteritis" of Stevens' classifica- tion is considered to be due to an increase in the number of E. coli organisms already existing in the alimentary tract and not to the introduction Of another serotype. Of the third of Stevens' clinical groups (Stevens, 1963), the only specific manifestation that need be mentioned here is of hemorrhagic enteritis. This acute disease is characterized by sudden death and is often associated with dietetic or environmental change. At necrOpsy the intestinal tract is partly filled with ingesta and blood, and one finds gross and microscopic lesions of severe hemorrhagic enteritis (Jennings, 1959; Philip and Shone, 1960; Thomlinson, 1963; Kinnaird, 1964; Bennett, 1964). Buxton and Thomlinson (1961) and Thomlinson and Buxton (1962) proposed that the hemorrhagic enteritis syndrome was an anaphylactic reaction. Stevens (1963) supported this view. 9. 60 The most comprehensive of the recent reviewsof the role of E. coli in disease (Sojka, 1965) does not include an adequate statement of the pathogenesis of colibacillosis, despite the author's statement that "...a theory of how enteritis (associated with E. coli) is produced in pigs has been recently advanced by Stevens (1963 a,b). It seems to be the most logical explanation of the sequence of events leading to development of various forms of enteritis and will be described in some detail." The concluding statement common to most of this author's chapters on the various diarrheal syndromes of colibacillosis are generally listed as "Diarrhoea and/or death due to toxic effect of E. coli" (Sojka, 1965). In a more succinct review, Barnum at al. (1967) have incorporated a number of hypotheses in their chapter, ”Pathogenesis", and have attempted to establish alternative possible sequential processes in the development of diarrheal disease associated with E. coli infection. Eacherichia coli organisms normally enter neonates through skin wounds, localized infections such as omphalophlebitis or, more usually, by ingestion (Jubb and Kennedy, 1963). In neonatal pigs certain serotypes of E. coli can move directly along the intestine to their multiplication sites in the terminal small intestine (Christie, 1967) and in certain instances can enter the circulation and persist for a few hours in the bacteremic state in gnotobiotic P188 (Christie, 1967) and conventional pigs (Barnum at al., 1967). 61 It has also been shown in calves (Fey et al., 1962) and in gnotobiotic eSOphagectomized pigs (Christie, unpublished data) that EL coli can leave the alimentary tract (presumably at the tonsils, through epithelial defects in the oral cavity, through the intact ore-pharyngeal epithelium, or via the respiratory tract), enter the circulation and become established in the gastrointestinal tract. Speculation, as reported by Jubb and Kennedy (1963) and Barnum at al. (1967), that intestinal infection may be preceded by5and result from bacteremia, was confirmed in one recent limited experi- ment with E. coli 0138:K81:NM in gnotobiotic pigs (Christie, 1967). A recent published account of pathogenesis (Barnum at al., 1967) emphasizes the bacteriologic aspects and factors concerned with the proliferation of the organism in the lumen of the intestine. These authors Stated that, "The mechanisms by which certain strains of E. coli produce diarrhea when present in large numbers in the intestine are unknown". Barnum et al. (1967) stated that gastric acid destroys the coliforms in the stomach, and the acid that floods the upper small intestine partially inhibits the growth of E. coli in this area. The high susceptibility of neonates to colibacillosis may be, in part, due to low gastric acidity during their first one or two days of life. Intestinal secretions and motility move bacteria caudally, constantly cleaning the tract of organisms and their meta- bolic products. Interference with secretion or motility would theo- retically allow for some build-up of potentially pathogenic populations of organisms. The presence of fimbriae on some serotypes of E. coli enables the organisms to adhere to the thin layer of seromucin on the brush border of the villal epithelial cells. 62 Strains of E. coli in man are known to be both resident and transient. The transient nature of the sojourn of E. coli strains in anhmals is well documented (Smith, 1960; Smith and Jones, 1963; Smith, 1965). The intestinal flora changes with age, change in diet, and interactions among intestinal bacteria and their products (such as colicines) (Barnum at al., 1967). Barnum at al. (1967) reviewed the concept that the absorption of endotoxins produced in the intestine is responsible for the physio- logical disturbances resulting in diarrhea. They speculated further that E. coli can produce toxins other than classical endotoxin and that these toxins might, as is the toxin recently isolated from Vibrio cholerae, be involved in the pathogenesis of diarrhea. Independently, Christie (1967), from his observations of the lesions of colibacillosis in gnotobiotic neonatal pigs, postulated that some of the pathologic changes that he saw were less character- istic of endotoxic phenomena than they would be if exotoxin were present and suggested that colonies of E. coli Ol38:K81:NM in the lumen of the pig intestine produced both endo- and exotoxin. The sequence of events suggested by Barnum at al. (1967) in the clinical syndrome of diarrhea due to E. coli is as follows: diarrhea; dehydration; hemoconcentration, leading to increased blood viscosity and resistance to blood flow; electrolyte losses (namely water, sodium, bicarbonate, chloride and at times potassium) and consequent electro- lyte imbalance. If untreated, severe uncompensated acidosis ensues, with detectable fall in blood pH, interference with cardiac impulse conduction, and death apparently due to cardiac failure. Terminal 63 \uemia has been observed in some instances, probably due to increased utilization of body protein and decreased renal circulation as a result of cardiac insufficiency. MATERIALS AND METHODS "The difficulty with which enteric colibacil- losis is produced under controlled experimental conditions remains the single greatest argu- ment against a primary causative role for E. coli in the disease". D. A. Barnum, P. J. Glantz and H. W. Moon, 1967 General Plan In a study that involves following the progress of a specific infection, the develOpment of a lesion or syndrome in body tissues or systems, and the collection Of tissues from the test animals so that the tissue changes can be correlated with the develOpment of the disease, it is necessary to make some assumptions regarding the behavior of organisms in host tissues. The validity of these assump- tions can be increased by standardizing genetic and environmental factors in the experiment. It was expected that litter mates held under the same environ- mental conditions and submitted to identical treatments, provided they were of similar physiological activity, would generally respond or react in a similar way. It was not possible to carry out the tissue sampling called for in this experimental method without killing the individual pig; hence, it was assumed that observations made on each individual animal at a particular time interval following exposure to pathogens would constitute a generally reliable picture of the development of the disease in any particular pig of that litter when 64 a. y.. .0 I‘ p '5‘- . :5. .— '0‘ 5.. a O o 65 held under the same circumstances and when the disease was permitted to run its course. The gnotobiotic pigs were obtained according to the modified hysterotomy method described by Waxler et al. (1966). This technique has been further modified by Dr. A. L. Britt* and Dr. G. L. Waxler,* in that a standard surgical skin preparation of detergent, alcohol, then chloroform was used prior to the sealing of the skin of the sow to the plastic sheet that formed the floor of the fiberglass ring set in the base of the surgical isolator (waxler et al., 1966). Other modifications have included the use of Allis tissue forceps to clamp the edge of the incised plastic sheet to the apposite edge of the incised skin of the sow at approximately 2-cm. intervals. This modification has helped to maintain the adhesion between these 2 sur- faces and to prevent movement of fluid from the surgical field out between the skin of the saw and the wall of the isolator. The use of umbilical cord clamps** to occlude the umbilical cord vessels prior to separation of the pig from the placenta has decreased the Operation time and considerably facilitated manipulations within the surgical isolator. Animals The experimental animals consisted of 5 litters of gnotobiotic pigs, using some as control (uninfected) animals and some as test (infected) animals in an accepted experimental design. Between-litter —__ * Department of Pathology, Michigan State University. ** "Double Grip" Diaposable Cord—Clamp, Hollister, Inc., Chicago, Ill. 66 variation was minimized so far as possible. The pigs used in this series of experiments were from related sows and were the progeny of one of two boars; they were fed a standard volume of the same diet;* they were maintained in sterile plastic-film isolators at an environ- umntal temperature of 85 to 90 F. Details of litter size, sex dis- tribution, bacterial serotype, number of organisms, and age of pigs at exposure are given in Table l. Agegof Experimental Animals Clark (1959) had described in mice and rats an apical tubular system in the neonatal jejunal cell which transported undigested pro- teins into the cell. The ability of neonatal pigs to absorb unaltered, colostral protein during the first 36 hours of life (Brambell, 1958) was possibly a function of a similar apical tubular system. It was considered in this study that the neonatal germfree pigs should be at least 48 hours old before being exposed to the test organism. By this time the phenomenon of "closure" (Lecce, 1966) should have occurred and one complicating factor in the pathogenesis of colibacil- losis, that of the direct absorption by the apical tubular system of endotoxin, exotoxin, or in the parlance of Barnum (1968) "enterotoxin", should be eliminated. Two other factors were also considered. It takes approximately 24 hours for the experimental pigs in isolators to learn to prOperly drink out of the feed pans, and by 48 hours after birth, the pigs are drinking effectively and have an established feeding regime. It is * SPF Lac, Borden Co., New York, N.Y. 67 «\HIN 1‘] nqoo 2 H m m N oeom 2 H N a.~ z m m a mean m H m amNN m H m omum m H m q a mmmm 2 H N m.H z a e «\Hue oHMH a H a e N\qu aomH S H a q.~ z N m o QHOH a H e QOOH m H m «OOH m H m m m nOOH 2.H N\HIN o.m z a N on Nome m H m N am Homo 2 H mm e.~ 27m H Amhmvv umnasz mmHm mo Amhmvv mOH x wme umos mo sonaaz mmeouooz um mad sOHumUHwHusovH now was nonssz munmoaxm um mamHsmwuo New was nonssz HouuHH mem we ow< mo Honssz mem Houuaoo omumshou mMHm vomoexm mHmsHsm HausoaHuomxo mo mHHmumn .H oHan 68 also more likely that the presence of material other than meconium in the gastrointestinal tract, prior to the introduction of the infective organisms, more closely approximates the situation that might obtain in natural infections of man or animals. In all litters in this experiment, there was a time lapse of at least 48 hours from delivery until the infective dose of organisms was administered. In Litter 1 it was observed that the clinical response to the organisms was unsatisfactory in terms of the produc- tion of a diarrheal syndrome, and in subsequent experiments all litters were infected within 5 days of delivery. Determination of Initial Sterility of Test Animals Immediately prior to the introduction of the infective material into the pigs in the sterile isolators, composite fecal, oral, nasal, and fecal tray samples were taken with sterile swabs and inoculated into 3 tubes of thioglycollate medium.* These tubes were incubated at 37 C, 50 C, and at room temperature. (In Litters 3, 4 and 5 the tubes were incubated only at 37 C and at room temperature.) These tubes were examined daily for 14 days. If there was no growth after 14 days, the isolators were considered to have been free of bacteria. Since these were very short-term experiments, the sterility of the isolators was determined immediately prior to the exposure of the pigs to the test organism, and generally this was done within 48 hours of the delivery of the pigs into the isolators. * Bacto Fluid Thioglycollate Medium (Dehydrated), Difco Labora- tories, Detroit 1, Michigan, U.S.A. ‘9 i u. 'ob». n-- -.s. . 0.. a u“. tu - 0 a. 69 Control Animals Some pigs from each litter were kept as control animals in order to check the sterility of the intrauterine environment during gestation and the sterility of the hysterotomy and delivery procedures. Only 2 or 3 pigs from each litter were retained for this purpose, since within a total of 5 litters this would give more than the necessary number of tissue controls for the experiment (see Table 1). It was planned that control animals from each litter be killed on a time schedule so that, together, they would adequately cover the actual-age scale of the test animals in the litters constituting the whole experiment. Since between-litter variation was negligible in terms of diet and environment, and genetic variation was minimal, it was considered that control animals from any 1 litter in the series would provide adequate tissue controls for test animals of equal age from other litters in the series. Infective Agent The organism used in all experimental animals (in this series) was E. coli Ol38:K81:NM and was isolated in August, 1959, from experi- mental pigs at Michigan State University. These organisms had been cultured, lyophilized in glass ampules, and stored at -20 C until required. Seventeen hours before the pigs were due to be infected, an ampule 0f the culture was broken open, and its contents were inoculated into liquid brain-heart infusion medium* and streaked onto a human blood * Bacto Brain Heart Infusion (B 37); Difco Laboratories, Detroit. 7O agar plate and a MacConkey Agar plate.* Only MacConkey Agar plates were inoculated prior to infecting Bitters 4 and 5. These culture plates were incubated at 37 C. After 16 hours' incubation, immediately before the experimental animals were due to be inoculated and when it was apparent that a pure culture was present, the tube culture was compared with a series of McFarland nephelometer tubes, from which comparison was derived an estimate of the number of bacteria present. The tube culture was then aseptically diluted with sterile saline (0.85% NaCl) until the desired number of organisms was contained in an appropriate volume of inoculum (see Table 1). This saline suspen- sion of the organisms was the material used in the isolators to inocu- late the test animals. The screw-capped tubes containing the inoculum were thoroughly sprayed with 22 peracetic acid (with approximately 0.12 wetting agent added).** The inoculum tubes (2 per isolator) were then introduced into the port of the sterile isolator, and the port was resprayed and closed. After 30 minutes the tubes were introduced into the isolator through the internal cap, and the inoculum was then ready for use. A standard dose volume of 1 ml. per pig was used throughout the experiment. The inoculum.was measured into a calibrated syringe, and a l-ml. ‘ dose was squirted onto the back of the tongue of each pig in the isolator. After each pig had been inoculated, 2 fluid ounces of the * Bacto MacConkey Agar (B 75); Difco Laboratories, Detroit. ** Nacconal, N.R.S.F., National Aniline Division, Allied Chemi- cal Corporation, New York, N.Y., U.S.A. 0. ! bu ' e 71 standard milk formula diet were offered. Infective Dose of Organisms Almost invariably, research workers have used infective doses in excess of l x 107 E. coli organisms. Saunders et al. (1963b) used graded doses of organisms from 1 x 103 to l x 108 and established infections with numbers of organisms as low as l x 103. Otherwise, much larger numbers of organisms have been used; 5 x 108 (Lecce and Keep, 1962), 1 x 108 up to 1 x 1012 (Saunders et al. , 1963a), 1 x 107 (Kohler and Bohl, 1966), 4.5 x 1010 daily for up to 10 days (Osborne, 1967), 0.42 x 106 up to 135 x 106 viable organisms (Christie, 1967). In a recent study, Staley et a2. (1968d) gave 1- to 2-hour-old pigs 100 x 109 E. coli 055:B5,H7 organisms. These organisms had been Packed by centrifugation and resuspended in saline prior to being administered to the pigs through a stomach tube. A number of theoretical objections to such large doses are immediately apparent, in that large numbers of organisms, submitted to extraordinary procedures such as centrifugation or prolonged incu- bation time, would lead to the introduction, together with the viable organisms, of dead organisms and preformed endotoxin and/or enterotoxin. In this experiment the infective dose was made from high dilu- tions of a young (16-hour) culture of organisms, and the diluted cul- ture was given directly, thereby minimizing the possibility of intro- d“91ng organic material other than the viable organism itself. The hiShest infective dose used in this experiment was 3 x 106 organisms. t ingested It was haped that this dose would more nearly approximate tha under natural conditions of exposure. 72 Necropsy and Laboratory Procedures Immediately prior to necrOpsy, oral and rectal contents of pigs were sampled with swabs, which were then placed in sterile tubes and removed from the isolator with the particular animal to be killed. These swabs were streaked onto MacConkey Agar plates and human blood agar plates and incubated at 37 C for 16 to 24 hours. Immediately after the pigs were removed from the isolators, they were killed by cerebral concussion and exsanguination (by severing the axillary vessels). A terminal blood sample of approximately 10 ml. was taken at this time. The identification numbers and the time intervals between inoculation of test animals and necropsy are given in Table 2. The body cavities were opened within a minute by excising the ventral thoracic and abdominal walls. The entire small intestine was freed from its attachments at the pylorus, the root of the mesentery and the ileocecal valve. The intestine was then quickly stretched out on a clean surface, and four l-cm.-long sections were taken from each of five sites: the duodenum, jejunum (approximately 50 cm. from the pylorus), jejunum (approximately 1 m. from the pylorus), jejunum (approximately 2 m. from the pylorus), and the terminal ileum at the level at which it bears two diametrically Opposed serosal reflections, one to the mesentery and the other to the cecum. Those sections of intestine intended for electron microscopic study were taken first and immediately opened along the mesenteric border. With fine scis- sors, groups of between 5 and 20 villi were snipped from the mucosal surface and immersed in cold (4 C) osmium tetroxide fixative (Palade» 1952). 73 ~\Hlo mooH om m meom m mnNN mm mmoo mm m “mum wH ~\H|~ Nqom Hm coco «\HIN mmmm Hm ammo so N\HI~ «mam oH ~\HI~ Hcoa om mnoo «\HIN mmmm om nmmo we «\HIN Hnmu «H N\HI~ oqom e mQQH «\qu momH s @00H mm N\an BOMH NH «\Hlm «OOH «\Hnm mo0H mm «\HIN mmom 0H m mQOm ~\H1N mmom ~\Hlm HOOH «\qu QOMH «\Hlm OOOH «N N\H1< mOMH m m seem m HcNN q womH m comm cm a momH q Ammmvv .oz me Amusonv mmeouomz Amzmvv .oz me Amusonv hmaouomz hemouooz um mw¢ was musmomxm smm3umm Hm>umucH hmeouomz um mm< mam unamoaxm aflmspu 0m HQKYHU UGH mmwm mo humouomc use ouamomxo mo mmEHH .N oHcmH .5. iii .0 'c 74 These samples were the first to be taken from the five levels of the intestine so that all tissues for electron microsc0py were in cold fixative within no more than 3 minutes of the concussion of the experimental animal . Preservation and fixation procedures for all experimental tissues are tabulated according to litter number in Table 3. Histopatholggic Technique Tissues from the five selected intestinal sites, taken for rou- tine histologic examination, were sliced approximately 0.5 cm. thick, identified, fixed in 10% neutral formalin buffered with sodium acetate, and processed for examination by procedures described in the Manual of'HistoZogic and Special Staining Technics, Second Edition, of the Armed Forces Institute of Pathology, washington, D.C. (1960). Sec- tions for the following stains: Movat's pentachrome (Thompson and Hunt, 1966), Alcian Blue (Pearse, 1961), Mayer's mucicarmine, Brown and Brenn stain for bacteria, hematoxylin and eosin, and the periodic acid-Schiff reaction (Armed Forces Institute of Pathology, 1960) were cut from Paraplast*-embedded tissues at a thickness of 6 u. Formalin-fixed tissues were frozen, 12 u sections were cut on an Ames Labtech Cryostat, and the sections were stained with Oil Red 0 Fat stain (Armed Forces Institute of Pathology, 1960). From the pigs in Litters 4 and 5, l—cm. lengths from each of the previously selected sites of the small intestine were fixed in * Paraplast Embedding Paraffin, Scientific Products, 1210 Leon Place, Evanston, 111., U.S.A. u do. 75 Table 3. Preservation and fixation of tissues r‘ J t Litter Number Methods 1 2 3 4 5 Carnoy's fluid fixative + + Fresh, unfixed tissue frozen over dry ice and stored in sealed containers at approx. -70 C + + Tissue frozen in a mixture of acetone and dry ice, then stored in sealed containers at approx. -70 C + + + 10% neutral formalin buffered with sodium acetate, and kept at approx. 40 + + + + 10% neutral formalin buffered with sodium acetate kept at room temp. + + + + + Osmium tetroxide fixative (Palade, 1952) + + + + + r— A- '- I .‘v 76 Carnoy's fluid and were subsequently stained by Best's method for the demonstration of glycogen (Mallory, 1942). Histochemical Methods All tissues for histochemical study, save for the Paraplast— embedded tissues, were cut at 12 u, except where stated otherwise, on either an International Harris Cryostat (cold acetone fixed tis- sues), an Ames Labtech Cryostat (unfixed frozen tissues) or a Pearse— Slee Cryostat (unfixed frozen tissues and cold formalin fixed tissues) and mounted on clean glass slides or coverslips and stored at -20 C prior to histochemical staining in litter batches. Adequate controls for each enzyme reaction were prepared by: (l) incubating duplicate tissue sections with and without substrate, (2) processing tissues from similar organ sites from germfree litter mates of the same age and kept under the same envirOnmental and die- tary conditions as the monocontaminated gnotobiotes, (3) processing all tissues of a particular litter and treatment at the same time, in one batch of incubating medium (except where specifically stated otherwise), (4) processing tissues known to be positive for the test material at the same time as the experimental tissues were being processed. Alkaline phosphatase. Pearse (1961) described the disadvantages of using unfixed frozen tissue in the histochemical visualization of alkaline phospha- tase, and stated that during incubation the enzyme, together with otherfprotein and nonprotein materials, is lost into the incubating tour ;-0 . but 'I 77 medium, leading to false localization of the enzyme by adsorption at other sites and deposition on the section of products of enzymic activity in the medium. Mbwry (1949) recommended fixation in cold acetone and storage at —70 C prior to processing for demonstration of alkaline phosphatase. Pearse (1961) has suggested preservation of tissues in cold formalin and preparation of frozen sections as being "particularly suitable" for the demonstration of alkaline phosphatase. The parti- cular advantages of this technique were claimed to be sharpness of localization of the enzyme and excellence of the morphologic picture (Pearse, 1961). In view of these conflicting reports, tissues were preserved in acetone at -70 C, and duplicate specimens were stored in 10% neutral formalin buffered with sodium acetate at 4 C. In addition, fresh tissues from Litter 5 were frozen over dry ice and stored in indivi- dual screw-cap aluminum vials prior to processing. Tissues from each of the four preservation methods were cut and mounted on clean glass slides. Alkaline phosphatase activity at pH 9.2 was determined, using the calcium-cobalt method for alkaline phos- phatase according to Gomori (1952). Tissues were incubated at 37 C for approximately two hours. One batch of fresh frozen unfixed tis- sues (Litter 5) was incubated at room temperature for 30 minutes in the hope that there would be less nonspecific artifactual staining of sections. l") .- 78 Acid phOSphatase. Two methods for the demonstration of acid phosphatase were used. Unfixed frozen tissue sections were cut at 12 u and mounted on clean glass coverslips. The enzyme was localized using the "Standard Coupling Azo Dye Technique" (Pearse, 1961). The acid phosphatase activity was determined at pH 5.0. Sodium alpha naphthyl phosphate in veronal acetate buffer Servedjas the incubating medium; the dia- zonium salt employed was the stable diazotate of O-amino azotoluene. The sections were incubated for 60 minutes at room temperature. In the second method, formalin-fixed, Paraplast-embedded tissues were cut at 6 u and mounted on clean glass slides. Acid phospha- tase was demonstrated according to the post-coupling technique de- vised by Rutenberg and Seligman (1955). Acid phosphatase activity was determined at pH 5.2. Sodium 6 benzoyl-Z- naphthyl phosphate in veronal acetate buffer served as the incubating medium, and the released insoluble naphthol was coupled to diazo blue B in phosphate buffer at pH 7.6. Tissue sections were incubated for 4 hours at room temperature . Adenosine triphosphatase. Tissues to be tested for the presence and distribution of adeno- sine triphosphatase were immediately frozen at -70 C and maintained at that temperature until processed to completion. The calcium method for adenosine triphosphatase of Padykula and Herman (l955a,b) was employed. Tissue sections were cut at 12 u and mounted on clean glass coverslips. A sodium barbiturate buffer at pH 9.4 was used, and the 79 tissue sections were incubated in the substrate solution for three hours at room temperature. Leucine aminOpeptidase. The method used for demonstration of leucine aminopeptidase activity was that devised by Nachlas, Crawford and Seligman (1957). Tissues were:incubated for three hours at room temperature in L-leucyl-4-methoxy-beta-naphthylamide solution containing sodium chloride, potassium cyanide and fast blue B salt. Acetate buffer at pH 6.5 was used. Succinate dehydrogenase. To localize succinate dehydrogenase, 12 u sections, cut from unfixed frozen tissues, were incubated for 30 minutes at 37 C in Tris-buffered sodium succinate with 3-(4,5-dimethyl—thiazolyl-2)- 2,5-diphenyl tetrazolium bromide (MTT) according to the method of Pearse (l96l). Lactate dehydrogenase. For the intramitochondrial demonstration of lactate dehydrogen- ase it is necessary that cold formalin-fixed tissues be used. When fresh, unfixed frozen sections are incubated with nitro—blue tetra- zolium, a crystalline artifact outside the mitochondria is produced (Novikoff and Masek, 1958). Pearse (1961) condemned the use of formalin fixation because some of the enzyme activity is destroyed by the fixative. Because of these conflicting opinions, both unfixed frozen sections and formalin-fixed tissues were cut and processed for lactate dehydrogenase demonstration. 80 Lactate dehydrogenase activity was determined according to the method of Hess et a1. (1958), using sodium DL-lactate as the substrate with diphosphopyridine nucleotide, sodium cyanide adjusted to pH 7.2, and MIT added. Electron Microscopy Methods After initial fixation of villi for 30 minutes to 2 hours in cold osmium tetroxide fixative (Palade, 1952) the tissue samples were immersed in fresh osmium tetroxide fixative at 4 C for 6 hours, then washed in 4 changes of Palade's sodium barbital buffer solution at pH 7.4 (Palade, 1952). The tissues were then dehydrated in a graded series of cold ethanol-water mixtures and embedded in Beem* capsules in Epon 812* according to the method of Luft (1961). The ratio of A:B mixture of the Luft (1961) resins was 5:5. The tissue' capsules were cured at 37 C for 16 hours and at 58 C for 24 hours. Sections were cut with glass and diamond DuPont knives on a Sorvall Porter—Blum MT-2 Ultramicrotome. The Epon blocks were trimmed so that cells at the base of the villi were first to be sectioned, then as these tissues were examined, the blocks were retrimmed so that sections of cells at the midvillus were taken. Sections were cut at a thickness between 700 and 900 A. Naked 400-mesh and Parlodion*-carboned grids of 100- and ZOO-mesh were used. Sections were stained with Reynolds' lead citrate (Reynolds, 1963), uranyl acetate (Pease, 1964), or a combination of both (Pease, 1964). Grids were observed on a Philips EM-lOO B electron microscOpe and * Ladd Research Industries, Inc., Burlington, Vermont, U.S.A. ‘u A“ ‘I DA 81 photographed on 35mm. Kodak Fine Grain film. Reclamation of the Test Serotype After the blood agar and MacConkeyéAgar plate cultures of the swabs removed from the isolators had been examined and read, a single sterile 100p was used to sample a typical colony from each of the cultures grown from samples from a particular isolator. This 100p of organisms, selected from different plates but all from the one isolator, was inoculated onto a nutrient agar slant, incubated for 24 hours at 37 C, sealed with a foam plug, and stored at room temperature until all samples were ready to be serotyped. In this way it was possible not only to check the validity of the monocontamination but also to ensure that the serotype isolated from the test animals was, in fact, the serotype that was used in the initial infecting exposure and that a mutant of altered serotype had not appeared during the course of the eXperiment. At the conclusion of all experiments, cultures harvested from each of the nutrient agar slants were serotyped according to the methods for o and K serotypi‘ng recorded by Sojka (1965). RESULTS "The fact that diarrhea occurs in E. coli in- fection without morphologic evidence of enteritis should be considered in any attempt to explain the pathogenesis of the diarrhea". H. W. Moon, D. K. Sorenson, J. H. Sautter and J. M. Higbee, 1966 The Experimental Plan In this experiment animals from 5 litters of germfree pigs were orally contaminated with an approximately equal number of E. coli 0138:K81:NM organisms and maintained under gnotobiotic conditions for a variable number of hours until necrOpsy. Clinical observations were made at this time. In this way it was h0ped to simulate natural infection of neonatal pigs and to study the pathogenesis of the associated disease in colostrum—deprived animals. Clinical and Gross Pathologic Findings The gross pathologic findings of infection of neonatal gnoto- biotic pigs with E. coli 0138:K81:NM (Mich.) have been reported in detail elsewhere (Christie, 1967). In this experiment the observations of gross lesions were con- fined to the intestinal tract in an attempt to relate the observed clinical manifestation of diarrhea with the gross appearance of the intestinal tract at necropsy. 82 ‘. 4v 83 The first gross lesion of disease was seen in Pig 1305 when it was killed 8 hours after inoculation. The mesentery of the entire length of the small intestine at the point of attachment to the in- testine was markedly edematous. The intestinal tube was relatively~ flaccid and moderately filled with homogeneous, very fluid contents. The clinical appearance of the pigs monocontaminated with E. coli Ol38:K81:NM in this experiment is tabulated (see Table 4). The first clinical sign was seen in Pig 1306 when it was killed 8 hours after inoculation. This pig did not have apparent diarrhea, but the rectal contents evacuated by the animal after concussion were a mixture of thin watery fluid, mucus and suspended curds. At necropsy the intestine was flaccid and distended with fluid contain- ing suspended milk curd and mucus. Hyperemia of the intestine was not grossly apparent. One pig (1307) died during the course of the eXperiment 2 hours before it was due to be killed. At necropsy there was a flaccid, fluid—filled small intestine, hyperemia of the entire gastrointestinal tract, particularly of the gastric mucosa, and a marked ascites. Twelve hours after inoculation the fluid ingesta persisted throughout the intestinal tract and was sometimes the only gross abnormality that was seen at necrOpsy. From 12 hours after inocula— tion, most inoculated pigs scoured profusely. At first, the peri- neum and hindquarters were pasted with the soft yellow feces charac— teristic of postmeconial feces of germfree neonatal pigs. As the diarrhea continued, the feces washed the perineum leaving it rela- tively clean, until in animals that had scoured for many hours (Pig 74.2 A D nix”... vs’ Av n-O‘I ..v'.‘ 1N I‘ll u use: 84 .3609 + soHumuvhnoQ + + some mo wanosm «monnumHv humans oHHuoonoum + + + + seas and: B33 + + + + + + mmsuumHv assume unamoum + + + mmoow aH mvuoo mooosa pom vHsHm humane voxfiz + + + + + + new? HeoHaHHo oz wH mH eH oH 0H «H «H eH NH NH NH 0H m m w a a soHumHsuosH nouns mason m m m m m m m m m.~ m.e m.e m.~ m.~ m.s n.e e s “eases eeeouede he use .6 .6 .b .6 .6 ,0 .6 .c ,0 +1 .L .0 ,0 TL T. T. T. 7s 7s nu 7d 7d no 7c 7d no r. to no no cc r. to r. a: m “assessmmmmemmmwm use: zzunxHNMHo mNou .m.nuH3 moussHEMusooosoa mmHa mo ousmsmoamm HmoHsHHu .e oHan 85 hmaouoma ou uoHuq hHmumHvosEH mason HmoHsHHo n + suede + + + + GOHumuvanon + + + + + some mo wsHsuus “monuumHv humans oHHuoonoum + + + + + + + + umoo uHmn vmmHmm + + .+ + + + + monuume humans mmomoum + + mooom sH mouse mucosa was pHus mucosa vmtz + + + . mome HmoHsHHo oz as me as so we as Nm .Nm ma mu em em «H cm on om mH aoHueHeuoeH house endow m.e mm Hm Hm on on a e .m m.m H m.m m.m m m m m Aeeeev eeeouuee ed use 001.... T.T.6 TTOZZG -33 mEOGBLILngSI vessHusooth oHAmH 86 1009) "scalding" of the perineum (erythema venenatum) was apparent. Diarrhea was observed in every pig exposed to 3 x 106 E. coli 0138:K81:NM at least 16 hours beforehand. The fecal material from those pigs killed early in this series was of watery consistency, pale yellow, and with suspended fine whitish curds. As time passed the color of the watery material varied among pigs, and the suspended curds were less commonly seen, even though the pigs continued to drink well throughout the experi- ment. Occasionally mucus was observed in the intestinal contents. lHowever, there was no gross evidence that excess mucus production contributed to the diarrhea in any way. Anorexia was not observed in any of these experimental pigs, but in few was the disease permitted to run its course. However, those animals left for a relatively long time before euthanasia did not lose appetite at any stage. I Loss of "bloom” in the hair coat, erection of the hair coat, and degrees of arching of the back and straining in an apparent attempt to defecate were observed variously from 16 through 24 hours after inoculation of the gnotobiotes with the organism. The first signs of dehydration were seen in pigs 20 hours after inoculation. In a previous experiment (Christie, 1967) significant tempera- ture changes were not observed in neonatal pigs infected with this serotype, except in terminal stages of the infection when the body temperature fell. Rectal temperatures were not recorded in either control or infected pigs in this experiment. 87 At necropsy of those animals killed 14 hours or longer after inoculation, the small intestine was seen to be flaccid and distended with fluid, with or without suspended matter such as milk curd, mucus and gas. Hyperemia of the tract was generally not grossly apparent. Throughout the experiment the control animals remained free of demonstrable bacteria, continued to pass soft or typically pasty feces, and maintained apparent good health until the time of euthanasia. Recovery and Serotyping_of the Orggnism Fourteen cultures were grown on nutrient agarslants. Each individual culture represented a composite of organisms harvested from the positive cultures of the experimental pigs of a particular litter that were housed in the same plastic isolator. The cultures harvested from the slants were serotyped according to the methods recorded by Sojka (1965) and all cultures were found to belong to the E. coli 0 group 138 and had the K 81 antigen. HistoEathologic Findings Histopathologic observations in this study were confined to the intestinal tract. The lesions observed ranged from those of a mild acute enteritis (Figures 1, 2 and 3) to a histologic picture, in persistently scouring pigs, microscopically indistinguishable from that seen in clinically normal germfree control animals. Ike Small Intestine Sections were routinely made from duodenum, jejunum at approxi- mately the 50-cm., lOO-cm., and ZOO-cm. levels, and the ileum close Figure l. Pig 2258. Age 3 days. Terminal ileum. E. coli was introduced per os 18 hours before necropsy. Mild acute enteritis. Neutrophils are invading the lamina propria. Hematoxylin and eosin. x 750. Figure 2. Pig 2257. Age 3 days. Duodenum. E. coli was introduced per 03 18 hours before necropsy. Neutrophils are present in the tissue spaces and have formed a microabscess in the lamina propria. Hema- toxylin and eosin. x 750. 89 Figure 3. Pig 2257. Age 3 days. Duodenum. E. coli was introduced per os 18 hours before necropsy. There is marked vacuolization of epithelial cells with cell breakdown and mucosal disarray. Neutrophils abound in a vessel within the lamina propria, and are free in the interstitium. Movat's pentachrome. x 750. 90 to the ileocecal valve. It is apparent that marked variations from the normal can occur as a result of infection with this serotype of E. coli. It is also apparent that much variation from the normal to abnormal can occur as one examines tissues taken from neighboring sites in an intestinal tract colonized with E. coli 0138:K81:NM. The mucosa and lamina propria. In some infected pigs and some germfree control animals the sur- face epithelium was ordered and regular with little or no vacuoliza- tion (Figures 4 and 5) and with occasional active goblet cells filled with mucin and becoming more frequent in more caudal sites (Figures 6, 7 and 8). According to the level of the intestine from.which the section was taken, lymphocytes were present singly or in aggregates in the submucosa and in increasing numbers caudad (Figures 6 and 9). Some slight but distinct changes from the above were also seen in the control and monocontaminated pigs. The volume of the cytOplasm of the individual columnar cells of the mucous epithelium was uni- formly greater and less intensely stained. In some instances the epithelial nuclei were centrifugally placed in relation to the cell cytOplasm and the axis of the villus (Figure 10); in other sections the epithelial cell nuclei were uniformly axially placed, close to the basement membrane of the epithelium (Figure 11). The cytoplasm of these cells was negative to Mayer's mucicarmine stain. A.more advanced degenerative change frequently observed was that in both control and infected animals columnar epithelial cells of the villi were markedly distended or vacuolate either singly or en masse (Figures 12 and 13). 91 Figure 4. Pig 9047. Age 2—1/2 days. Duodenum. Germfree control. The duodenal epithelium is ordered and regular with little vacuolization. Movat's penta- chrome. x 750. 92 ‘ , Figure 5. Pig 2253. Age 2 days. Terminal ileum. Germfree control. The epithelium is ordered and regu- lar, with moderate vacuolization in some areas. Inter- cellular fluid is indicative of active fluid absorption. Moderate subepithelial edema of the basal lamina is present. The lacteal canal is distended. Hematoxylin and eosin. x 470.’ ' 93 Figure 6. Pig 9046. Age 2 days. Terminal jejunum. Germfree control. The epithelium is or- dered with consistent moderate vacuolization char- acteristic of this region of the intestine. Goblet cells are more frequent here than in more anterior sites. Aggregates of lymphocytes are present in the submucosa. Movat's pentachrome. x 75. 94 Figure 7. Pig 9047. Age 2-1/2 days. Terminal ileum. Germfree control. The surface epithelium is markedly vacuolate, with interspersed goblet cells filled with mucin granules. This degree of vacuoli- zation is not unusual in this region of the intestine of neonatal germfree pigs. Movat's pentachrome. x 750. 95 Figure 8. Pig 1006. Age 4 days. Terminal ileum. E. coli was introduced per os 32 hours before necropsy. Villi are long,and slender. Breucke fibers are not contracted. Although the epithelium is regular it is consistently moderately vacuolate. Active goblet cells filled with mucin are plentiful. The villi tend to be compressed together. Hematoxylin and eosin. x 75. Figure 9. Pig 9047. Age 2—1/2 days. Terminal ileum. Germfree control. The epithelial cells are markedly vacuolate; goblet cells are frequent. Aggre- gates of lymphocytes are present in the submucosa. Movat's pentachrome. x 75. Figure 10. Pig 1010. Age 6 days. Terminal ileum. Germfree control. Villal epithelial cells are moderately vacuolate with centrifugally placed nuclei. There is no evidence of rupture of individual cells. Hematoxylin and eosin. x 750. 97 Figure 11. Pig 9046. Age 2 days. Terminal ileum. Germfree controlu These markedly vacuolate cells represent the maximum distention seen in the epithelium of the ileum of germfree neonatal pigs. Epithelial nuclei are situated axially to the villus at the cell base. Movat's pentachrome. x 750. 98 Figure 12. Pig 1005. Age 3 days° Midjejunum. Germfree control. Marked vacuolization and distention of the epithelial cells. Mild subepithelial edema, with moderately distended lacteal vessel. Hematoxylin and eosin. x 470. Figure 13. Pig 1306. Age 4-1/2 days. Terminal jejunum. E. coli was introduced per 08 8 hours be- fore necropsy. There is marked vacuolization and distention of the villal epithelial cells. Hematoxy- lin and eosin. x 190. ' 99 Varying degrees of subepithelial edema of the villi were seen to be associated with 1 or more of the epithelial changes described above (Figures 5, 12 and 14). Commonly observed was edema of the tips of the villi, forming subepithelial bullae (Figures 5, 12 and 14). In some instances this edema was restricted to the tip of the villus,.but it was also seen to extend in some cases the entire length of the villus. In other cases the subepithelial edema was observed at the base or midvillus region leaving the tip relatively unchanged. In a few instances distinct separation of the basement membrane from the lamina propria indicated that the subepithelial Space was a fixation artefact. Generally, the condition of subepithelial edema was indicated by subepithelial vacuolization bounded by fibrous stroma (Figures 15 and 16). At times, edema of the axial tissues of the lamina prepria of the villus was seen. Other changes observed in the intestinal villi of infected but not control pigs were hyperemia with or without hemorrhage and/or necrosis of individual cells (Figures 17 and 18). In rare instances there was surface necrosis with sloughing of individual cells, and where sufficient cells had been lost from the tip of a villus there was an apparent replacement of villal epithelial cells with apparently normal juvenile cells. A further condition seen in the infected intestine, and generally not associated with many of the changes described above, was that in which the villi were markedly enlarged due to vacuolization of the epithelial surface or due to subepithelial edema, so that the villi 100 Figure 14. Pig 1004. Age 3-1/2 days. jejunum. E. coli was introduced per os 28 hours be- fore necropsy. Mild subepithelial edema with disten- tion of the vessels of the lamina prOpria. and eosin. x'470.’ ' ‘ ' Anterior Hematoxylin 101 Figure 15. Pig 0657. Age 30 days. Anterior jejunum. E. coli was introduced per os 48 hours be— fore necropsy. Subepithelial vacuolization bounded by fibrous stroma, constituting subepithelial edema and edema of the intramembranous Space of the base- ment membrane. Hematoxylin and eosin. x 750. 102 Figure 16. Pig 1306. Age 4—1/2 days. Anterior jejunum. E. coli was introduced per 0s 8 hours be- fore necropsy. Subepithelial vacuolization bounded by fibrous stroma, indicating edema of the basal lamina and the lamina propria. Hematoxylin and eosin. x 750. 103 :"‘x‘ : . flee.» « ~- ‘Ur . , ’mafl ...}... I“; H ‘0‘ t 9.4. H. ‘. . ‘j -. [’3- 1‘7"~ “.,. 43“,". hut "nuts‘ w‘y‘.‘ . .m Figure 17. Pig 2255. Age 2-1/2 days. Terminal ileum. E. coli was introduced per os 16 hours before necropsy. Some epithelial cells are distended to the point of rupture, others have ruptured and become confluent with neighboring cells. Effete cells at the tip of the villus are sloughing (arrow). There is moderate hyperemia of the lamina prOpria. Hema- toxylin and eosin. x 190. 104 Figure 18. Pig 2255. From the same field as Figure 17. Epithelial cell distention to the point of confluence with neighboring cells, or rupture with loss of cytoplasm into the lumen. There is moderate' hyperemia. Hematoxylin and eosin. x 470. YEA: In 17‘ 70.: 105 were in very close apposition and appeared adherent to each other, the microvilli forming a dense eosinOphilic boundary common to both villi (Figures 8, 19, 20 and 21). The gross appearance in this case was of a thicker-walled intestine that was firm to the touch compared to the flaccid intestine that was commonly seen in colibacillosis. At times in the infected intestine the laminae propriae of some of the villi were infiltrated with inflammatory cells, primarily neutrOphils (Figures 1 and 2), and in some instances these cellular aggregates formed microabscesses that distended the shaft of the villus. Other inflammatory cells commonly associated with a general edema of the villi were lymphocytes (Figure 21), eosinOphils, plasma cells and occasional macrOphages. In some instances, the villi were seen to be very long and slender. There was almost uniform, advanced vacuolization of the epi— thelium of the villi in these sections, there was little or no sub— epithelial vacuolization (Figure 22), and Breucke fibers were generally not contracted. Frequently the laminae propriae of these villi were infiltrated with foci of neutrophils. The villal epithelial cell of the small intestine. The outstanding feature of villal epithelial cells of the small intestine of both neonatal germfree pigs and neonatal pigs monocon- tamined with E. coli 0138:K81:NM was the degree of vacuolization of the cytoplasm of cells clothing the midshaft and luminal end of the intestinal villus. Although this phenomenon is almost always present, and is in fact used routinely by histologists as one of the signs 106 (am 0 5* VJ __ . I; Ix " i l , l w \ Figure 19. Pig 9043. Age 3 days. Terminal jejunum. E. coli was introduced per os 18 hours be- fore necropsy. Marked vacuolization and rupture of intercellular walls with confluence of mucosal cells, and distention of the villi to the point of apposi- tion of brush borders. Breucke fibers are not active. Movat's pentachrome. x 470. 107 Figure 20. Pig 9043. Age 3 days. Midjejunum. E. coli was introduced per as 18 hours before necropsy. Epithelial vacuolization with villal distention and apposition of brush borders. Movat's pentachrome. x 750. 108 Figure 21. Pig 9038. Age 2-1/2 days. Duodenum. E. coli was introduced per as 8 hours before necropsy. Distention at the tip of the villus is due to collec- tion of absorbed material or inflammatory edema. The clear space midshaft is probably artifact. There is apposition of the brush borders with those of neighbor- ing villi. The lamina propria is infiltrated with lymphocytes (arrow). Movat's pentachrome. x 470. 109 Figure 22. Pig 0657. Age 30 days. Terminal jejunum. E. coli was introduced per os 48 hours be— fore necropsy. Edema of the interstitium of the lamina propria. Some neutrophils are present. Breucke fibers are contracted in this villal shaft. Movat's pentachrome. x 750. t . 110 for identification of the site of origin of the tissue, rarely is even the slightest reference made to this feature in the literature of the pathology of colibacillosis. In germfree neonatal pigs it can be assumed that the villal epi- thelial cells from the more caudal sites in the small intestine will be more often vacuolate. In these germfree neonatal pigs, rarely were duodenal epithelial cells vacuolate to any marked degree; more often they were regularly columnar with evenly stained cytoplasm, centrally placed nuclei, detectable brush border, and well defined basement membrane (Figure 23). In those instances in which active absorption was taking place immediately prior to the collection and preservation of tissues, the accumulation in the intercellular spaces of fluids en route from the epithelial cells to the lacteal and capillaries was readily apparent (Figures 5 and 24). This phenomenon was also observed in infected animals in which cellular disruption had not yet occurred (Figure 25). In the germfree neonatal pigs the degree of vacuolization of the villal epithelial cells became more marked with distance along the intestinal tract (Figures 7 and 26). Although in some instances the vacuolization of individual cells was considerable in villal mucosa from germfree control pigs it was not associated with marked disruption of the mucosa (Figures 7, 10 and 11), whereas in villal mucosa from infected pigs, vacuolization of the same order of magni- tude seemed to be associated with a degree of individual cell break- down and mucosal disarray (Figures 3, l9 and 27). 111 Figure 23. Pig 9046. Age 2 days. Germfree control. Duodenum. Regularly columnar villal epi— thelial cells with evenly stained cytoplasm, centrally placed nuclei, detectable brush border, and well defined basal lamina. Active absorption is indicated by the small clear cytoplasmic spaces, and by the fluid present in the intercellular space. Movat's pentachrome. x 750. 112 Figure 24. Pig 2262. Germfree control. Age 4 days. Duodenum. Active fluid absorption is indi- cated by intercellular collection of fluid. Lack of edema or accumulation of fluid in the lamina propria is a sign of normal function. Hematoxylin and eosin. x 750. 113 Figure 25. Pig 1305. Age 4-1/2 days. Duodenum. E. coli was introduced per as 8 hours before necrOpsy. As in Figure 24, there is evidence of epithelial cell function and active absorption. There is some accumu- lation of fluid in the lamina propria, possibly an early sign of developing infection. Hematoxylin and eosin. x 750. 114 Figure 26. Pig 9047. Age 2-1/2 days. Midjejunum. Germfree control. Moderate vacuolization of villal epithelial cells typical of this region of the in- testine of germfree neonatal pigs. Movat's penta- chrome. x 750. 115 Figure 27. Pig 9043. Age 3 days. Terminal ileum. E. coli was introduced per os 18 hours before necropsy. Vacuolization of epithelial cells, although of the same magnitude as that seen in germfree neo- nates, is here seen to be associated with cell break— down (arrow) and mucosal disarray. Movat's penta— chrome. x 750. 1‘?’ A 116 Generally in the pigs monocontaminated with E. coli Ol38:K81:NM vacuolization was more pronounced and extensive: The effect of the organism on the mucosa of the more cranially situated portions of the small intestine was less apparent. In many instances, particularly in those animals inoculated only a few hours beforehand, there was no apparent difference between the duodenal mucosa of the monocon— taminated gnotobiotes and the axenic control group. Mucosal cells from more caudal sites in the intestine were more markedly vacuolate (Figures 8, l3, l4, l9 and 27) than were those from comparable sites in axenic control pigs (Figures 7, 9, 10, 11 and 12), and the vacuolization was much—more extensive. At times, in the mucosa of infected pigs, the cells were seen to be so vacuolate as to be disrupted by the breakdown of the cell membrane either at the luminal surface or at the intercellular wall (Figures 19 and 27). Despite this degree of disruption, at no stage was the lamina prOpria seen to be denuded of epithelial cells. Those epithelial cells which were excessively distended up to and beyond the point of rupture were apparently still retained as part of the covering epithelium of the villus. There was no evidence of exces- sive shedding of prematurely effete villal epithelial cells, nor was there evidence of denudation of the villus, exposure of uncovered lamina propria nor of foreshortening or clubbing of the villus. At no stage during the examination of tissues with the optical microscOpe were bacteria seen on the mucosal surface or within the villal epithelial cells, neither in hematoxylin-and—eosin stained sections, nor in those submitted to the Brown and Brenn stain for 117 bacteria nor the other special stains used on tissues in this . experiment. The submucosa and other structures of the small intestine. The changes observed in the submucosa and the muscular and serosal layers of the small intestine of the monocontaminated pigs were pri- marily inflammatory edema and infiltration with lymphocytes, neutro— phils, eosinOphils and macrOphages. Hyperemia (Figures 17 and 18), distended lymphatic vessels of the submucosa, subserosal edema (Figure 28) and occasionally intramuscular edema were observed in the infected pigs. Mature lymphocytes were present in the submucosa of the terminal jejunum and ileum of both the infected and the germ- free control pigs, but there were increased numbers of lymphocytes in these tissues in the pigs heavily contaminated with the test organisms. Enzyme Histochemistgy Histochemistry of the Villus and Villal Epithelial Cells The Phosphatases. Alkaline Phosphatase.--Although there was considerable loss of alkaline phOSphatase activity in the formalin-fixed tissues, the histochemical reaction was intense enough to be observed, and there was satisfactory sharpness of localization of the enzyme with excel- lent morphologic detail of tissues (Figure 29). More of the enzyme was retained in the tissues fixed in cold acetone and stored at -70 C, but morphologic detail was less 118 Figure 28. Pig 1306. Age 4—1/2 days. Anterior jejunum. E. coli was introduced per as 8 hours before necropsy. Subserosal and submucosal edema and disten- tion of submucosal lymphatic vessels. Hematoxylin and eosin. x 75. Figure 29. Pig 0661. Age 29 days. Terminal ileum. Germfree control. There is satisfactory localization of alkaline phosphatase in the brush border and excel- lent morphologic detail of the tissue. Cold neutral formalin fixation. Calcium-cobalt method for alkaline phosphatase. x 750. 119 satisfactorily, and the stain was less sharply localized in the tis- sues and within the cells (Figure 30). Least satisfactory in terms of specific localization of enzyme and of morphologic detail were those tissues fresh frozen over dry ice and stored at —70 C (Figures 31 and 32). In the germfree neonatal pig, alkaline phosphatase was restricted to the villal epithelial cells and was not seen in the epithelial cells of the crypts (Figure 31). Within the individual villal epithelial cell the enzyme was strongly localized in the brush border. There was at times a narrow clearer zone at the apex of the cell that would correspond to the region of the terminal web, in which there seemed to be a region of reduced alkaline phosphatase activity (Figures 30, 32, and 33). This phenomenon did not depend on the method of fixation, the period of storage prior to processing, or the level of intestine from which the section was cut (Figures 30 and 32). The remainder of the alkaline phosphatase activity was demon- strated only in the frozen unfixed and cold acetone fixed tissues- and was concentrated in the apical third of the villal epithelial cell and in a narrow zone axial to the cell nucleus close to the basement membrane (Figure 30). This pattern of tissue and cell distribution of alkaline phos- phatase was the same for each of the five levels of the small in— testine that were sampled. In control tissues, in which the villal epithelial cells were markedly vacuolated, the normal distribution 0f the enzyme was somewhat altered in that the positive staining 120 'Figure 30. Pig 1002. Age 3 days. Midjejunum. Germfree control. Much enzyme activity is retained, but morphologic detail and enzyme localization is poorly defined. Alkaline phosphatase is strongly localized in the brush border and there is some enzyme activity in the apical cytoplasm. There is reduced enzyme activity at the terminal web (arrow). Tissue frozen in acetone at -70 C. Calcium-cobalt method for alkaline phosphatase. x 750. 121 Figure 31. Pig 9046. Age 2 days. Duodenum. Germfree control. The enzyme is poorly localized and morphologic detail is minimal. Alkaline phos- phatase is restricted to the villal epithelial cells and is not apparent in the epithelium of the crypts (c). Unfixed frozen (-70 C) tissue. Calcium— cobalt method for alkaline phosphatase. x 75. Midjejunum. The enzyme is poorly localized Alkaline phospha- tase is localized in the brush border (arrow) and a prominent band of reaction product —— probably indicative of nonspecific phosphatase activity -- Figure 32. Pig 9046. Age 2 days. Germfree control. and morphologic detail is minimal. is present in the remainder of the cytoplasm. There is a band of reduced enzyme activity in the region of the terminal web. Unfixed frozen (-70 C) tissue. Calcium-cobalt method for alkaline phosphatase. x 470. 123 ." fis‘d v ' Figure 33. Pig 9046. Age 2 days. Duodenum. Germfree control. Alkaline phosphatase is localized in the brush border. There is a band of reduced enzyme activity in the region of the terminal web (arrow). Unfixed frozen (-70 C) tissue. Calcium- cobalt method for alkaline phosphatase. x 750. 124 reaction of the brush border was seen to be less even in intensity and there was a slight increase in intensity of the staining reaction at the margin of the vacuole (Figure 34). In those animals killed 4, 8 or 10 hours after monocontamination with E. coli Ol38:K81:NM there was no distinct change in the distri- bution or intensity of staining of the alkaline phosphatase (Figure 35). By 12 hours after inoculation there seemed to be some loss of intensity of staining, and by 14 hours the loss of intensity of the staining was more apparent in the epithelium more distally placed along the villi (Figure 36), as it was also less intense in sections taken further along the small intestine (Figure 37). Loss of this intense staining reaction and a lack of specificity of the distribu- tion of alkaline phosphatase was seen in those animals killed at subsequent time intervals although it was apparent that by 96 hours after inoculation, there was a renewed intensity of the staining reaction in the villal epithelial cells (Figure 38). The thin clear zone in the region of the terminal web continued to be apparent in the monocontaminated animals despite the changes in distribution of the enzyme in the remainder of the cell (Figure 39). The changes in staining intensity and intracellular distribu- tion of the enzyme associated with monocontamination with the organism were qualitatively similar at the five levels of the small intestine from which samples were taken. Acid Phosphatase.--In the germfree control pigs at all of the five selected levels of the small intestine, acid phosphatase was 125 Pig 1005. Age 3 days. Anterior Alkaline phosphatase is localized in the brush border of markedly vacuolate There is some concentration of enzyme activity Cold Calcium-cobalt method for Figure 34. jejunum. Germfree control. cells. at the margin of the intracytoplasmic vacuoles. neutral formalin fixation. alkaline phosphatase. x 750. 126 Terminal E. coli was introduced per as 4 hours before Although markedly vacuolate, the enzyme distribution in the villal epithelial cells is un— changed from that seen in the control animal tissues Figure 35. Pig 1303. Age 4 days. jejunum. necrOpsy. and is generally confined to the brush border. Cold neutral formalin fixation. Calcium-cobalt method for alkaline phosphatase. x 750. 127 Figure 36. Pig 9041. Age 2-1/2 days. Duodenum. E. coli was introduced per as 14 hours before necropsy. There is partial loss of alkaline phosphatase from the epithelium of the middle and distal third of the vil- lus. Note that the intensity of the reaction in this tissue is generally more marked than that observed in the terminal ileum (Figure 37). Unfixed frozen (—70 C) tissue. Calcium-cobalt method for alkaline phospha- tase. x 75. 128 . . :0" J _ [A 'V v) 1." a ' i ( d Figure 37. Pig 9041. Age 2-1/2 days. Terminal ileum. E. coli was introduced per as 14 hours before necrOpsy. Almost complete loss of alkaline phosphatase activity. Slight enzyme activity is apparent in the recently generated cells at the bases of the villi. Unfixed frozen (-70 C) tissue. Calcium-cobalt method for alkaline phosphatase. x 75. 129 Figure 38. Pig 1009. Age 6—1/2 days. Terminal ileum. E. coli was introduced per as 96 hours before necropsy. Alkaline phosphatase activity is localized in the brush border. There is renewed intensity of the staining reaction for enzyme activity in the brush border and the enzyme is well localized in that site. Cold neutral formalin fixation. Calcium-cobalt method for alkaline phosphatase. x 750. 130 Figure 39. Pig 9042. Age 2-1/2 days. Midje— junum. E. coli was introduced per as 16 hours before necropsy. Despite the diminution in alkaline phos- phatase activity in the region of the brush border in infected pigs, the thin clear zone of enzyme inactivity in the region of the terminal web is still apparent. Unfixed frozen (-70 C) tissue. Calcium- cobalt method for alkaline phosphatase. x 470. 131 seen to be evenly distributed throughout the villal epithelial cells and along the villi. The enzyme stained slightly less intensely in those villal epithelial cells at the tip of the villus. There was little or no acid phosphatase activity in the cells of the crypts (Figure 40). In sections of duodenum and anterior jejunum, the specific distribution of acid phosphatase within the cell was appar- ently more restricted than it was in those cells from tissues that were more caudally located in the small intestine. In the duodenal and anterior jejunal cells, enzyme activity was concentrated in the apical cytoplasm of the epithelial cells whereas, more caudally in the intestinal tract, the enzyme was more evenly distributed in the cell cytOplasm, although some concentration in the apical cytOplasm of the villal epithelial cell was still apparent (Figure 41). In some instances a perinuclear zone was observed in which there was no enzyme activity (Figure 41). There was no apparent change in cellular or tissue distribution of acid phosphatase in those villi exposed to E. coli 0138:K81:NM at either 8 or 10 hours prior to the time of necropsy, although the general intensity of the staining reaction was reduced. Twelve hours after contamination, at which time diarrhea was observed clinically, acid phosphatase was seen to be typically dis- tributed within the villal epithelial cells, but the staining reac- tion was generally less marked5(Figure 42). Loss of staining intensity of sites of acid phosphatase activity was associated with the number of hours elapsed after contamination, and this loss was more marked in tissues more distally positioned along the intestinal tract (Figure 43). 132 Figure 40. Pig 9046. Age'2 days. Terminal jejunum. Germfree control. Acid phosphatase is seen to be evenly distributed along the villi. There is little or no enzyme activity in the cells of the crypts (at bottom right). Acid phosphatase is dis- tributed evenly throughout the cytoplasm of the indi- vidual villal epithelial cells.' Unfixed frozen (-70 C) tissue. Standard coupling technique for acid phosphatase (Pearse, 1961). x 75. 133 Figure 41. Pig 9047. Age 2—1/2 days. Terminal ileum. Germfree control. In more caudal sites in the small intestine, acid phosphatase activity is evenly distributed in the cytoplasm of individual cells. A perinuclear zone in which there is no enzyme activity is apparent. Unfixed frozen (-70 C) tissue. Standard coupling technique for acid phos- phatase (Pearse, 1961). x 750. 134 Figure 42. Pig 9040. Age 2-1/2 days. Duodenum. E. coli was introduced per as 12 hours before necropsy. Acid phosphatase is typically distributed within the villal epithelial cells, but the intensity of the stain reaction is markedly reduced as a result of infection. Unfixed frozen (—70 C) tissue. Standard coupling technique for acid phosphatase (Pearse, 1961). x 750. 135 Pig 9044. Age 3 days. Terminal E. coli was introduced per as 20 hours before Marked loss of enzyme activity associated both with number of hours elapsed after monocontami- nation and the intestinal site of the tissue sampled. Compare this reaction with that depicted in Figures Figure 43. ileum. necropsy. 41 and 42. Unfixed frozen (—70 C) tissue. Standard coupling technique for acid phosphatase (Pearse, 1961). x 750. 136 The particulate or finely granular appearance of the stain reac- tion that is associated with histochemical reactions of membrane- bound enzymes was progressively supplanted by a smoother, more homo- geneous staining reaction in the villal epithelial cells of pigs monocontaminated with E. coli 0138:K81:NM for 20 hours and longer (Figure 43). Twenty-four hours after contamination, the villal epithelial cells of the ileum were uniformly stained, and there was clumping of stained material at the luminal surface in the region of the brush border (Figure 44). As with the histochemical reactions for other enzymes, marked vacuolization of the cell resulted in displacement of the enzyme- bearing organelles to the periphery of the cell (Figure 45). There was no acid phosphatase-positive material seen within the vacuoles. Adenosine Triphosphatase.-—In the neonatal germfree con- trol pigs, at each of the five levels of the small intestine, adeno- sine triphosphatase was observed to be particularly localized in the region of the brush border but with moderate activity in the apical half of the villal epithelial cells (Figure 46). In the monocontaminated animals, the first change was seen in pigs killed 10 or 12 hours after eXposure to the organism, and this change was a partial loss of specificity of distribution and intensity of staining of adenosine triphosphatase. This change was much more marked in the terminal portion of the small intestine (Figure 47) 137 Figure 44. Pig 9045. Age 3 days. Terminal ileum. E. coli was introduced per as 24 hours before necropsy. The loss of the finely particulate staining reaction characteristic of histochemical reactions of membrane- bound enzymes is apparent. There is clumping of acid phosphatase-active material at the luminal margin of the villal epithelial cells. Unfixed frozen (-70 C) tissue. Standard coupling technique for acid phospha- tase (Pearse, 1961). x 750. 138 Pig 9042. Age 2-1/2 days. Trans- Figure 45. E. coli verse section of villus from terminal ileum. was introduced per as 16 hours before necropsy. Marked vacuolization of the villal epithelial cells resulting in the displacement of enzyme bearing organelles to the periphery of the cell or to the margin of the vacuole (arrow). Unfixed frozen (—70 C) tissue. Standard coupling technique for acid phos- phatase (Pearse, 1961). x 750. 139 Age 2—1/2 days. Terminal Adenosine triphosphatase Figure 46. Pig 9047. ileum. Germfree control. is particularly localized in the brush border, and there is moderate enzyme activity in the apical cyto— plasm of the villal epithelial cells. Unfixed frozen (-70 C) tissue. Calcium method for adenosine tri— phosphatase. x 750. 140 .u ”new ~ 'Figure 47. Pig 9040. Age 2-1/2 days. Terminal ileum. E. coli was introduced per as 12 hours before necropsy. Loss of specific concentration of adenosine triphosphatase activity from the brush border. Unfixed frozen (-70 C) tissue. Calcium method for adenosine triphosphatase. x 750. 141 whereas, in sections from the duodenum and anterior jejunum, stain distribution and intensity more closely resembled that of the tissues from germfree control animals. In villal epithelial cells from animals exposed to Eh coli Ol38:K81:NM for 18, 20 and 24 hours, respectively, and in which there was marked vacuolization, the stained enzyme reaction product was seen to be pushed into the cell extremities by the intracytOplas- mic vacuole. There was no positively stained material within the vacuoles. In these tissues a slightly positive stain reaction was seen in the region of the brush border, but the reaction was less intense than that seen in the germfree controls (Figure 48). Proteolytic Enzymes. Leucine Aminongtidase.--Leucine amin0peptidase activity was observed to be localized at the brush border of the intestinal epithelium of neonatal germfree control pigs (Figure 49). Leucine aminopeptidase staining intensity in the microvilli of cells from the monocontaminates was sharply reduced, and 12 to 14 hours after infection only traces of the enzyme activity were detected at the apical surface of the villal epithelial cells (Figure 50). In sections from jejunum and ileum in which there was marked vacuolization, there was no detectable enzyme at the apex of the Some leucine aminopeptidase remained in the brush vacuolated cells. Sixteen border of the occasional interposed nonvacuolated cells. hours after contamination, slight activity in the brush border of cells from the duodenum and early jejunum could be detected (Figures 142 Pig 9043. Age 3 days. Terminal Figure 48. ileum. E. coli was introduced per as 18 hours before necrOpsy. Stained enzyme reaction products are dis— There is no placed by intracytoplasmic vacuoles. enzyme activity within the vacuoles. Some adenosine triphosphatase activity in the brush border and the apical cytoplasm is still apparent. Unfixed frozen (-70 C) tissue. Calcium method for adenosine triphos- phatase. x 750. 143 Figure 49. Pig 9046. Age 2 days. Terminal ileum. Germfree control. Leucine aminopeptidase activity is localized at the brush border. Unfixed frozen (—70 C) tissue. Method for leucine aminopeptidase (Nachlas at al., 1957). x 750. . - 4...“. D Figure 50. Pig 9040. Age 2-1/2 days. Terminal ileum. E. coli was introduced per os 12 hours before ‘necropsy. Twelve hours after monocontamination, en— zyme activity in the brush border is markedly reduced, and only traces of active enzyme are seen at the apex of the villal epithelial cells. Unfixed frozen (-70 C) tissue. Method for leucine aminopeptidase (Nachlas at al., 1957). x 750. 145 51 and 52), whereas no enzyme activity could be demonstrated in the villal epithelial cells of more distal intestinal tissues, nor in the intestinal villi from pigs sampled 18 or more hours after monocon— tamination with EH coli Ol38:K81:NM (Figure 53). The Dehydrogenases. Succinate Dehydrogenase.--The sites for succinate dehy- drogenase activity were clearly apparent in the small intestine of the germfree neonatal pig. Throughout the small intestine succinate dehydrogenase was seen in the epithelial cells of both villi and crypts, although the staining reaction seemed to be more intense at the base and in the midshaft of the villus (Figure 54). At a cellu- lar level, villal epithelial cells in the duodenum (Figure 55) and the anterior jejunum seemed to be more intensely stained than the villal epithelial cells in the midjejunum and the terminal ileum (Figure 56). The specific localization of the enzyme within the villal epi- thelial cell was consistent in tissues taken from the germfree con- trol pigs at all levels of the small intestine. There was a concen— tration of enzyme in the apical cytoplasm of the cell and a more intense, but more narrow, band of enzyme activity at the basement membrane (Figures 55 and 57). Eight hours after infection with E. coli Ol38:K81:NM, there was a distinct and intense reaction, with an even more clearly defined distribution of succinate dehydrogenase following the pattern of that seen in the control animals (Figure 58). Within 12 hours of 146 .r-. q p z I ." o 1 3‘ l ‘ o O f _ ‘ . 7" -fig}- «4&1 3‘72, “44v " L W ‘k ".1 ‘- r. '19 .VL' " 1 _ . I “ ' -' . A. .' \r ‘ " . | I % Figure 51. Pig 9042. Age 2—1/2 days. Duodenum. E. coli was introduced per 08 16 hours before necropsy. Slight enzyme activity is seen in the brush border of duodenal villal epithelial cells. The loss of leucine aminopeptidase activity is uniform along the length of the duodenal villi. Unfixed frozen (-70 C) tissue. Method for leucine aminopeptidase (Nachlas at al., 1957). x 190. 147 Figure 52. Pig 9042. From the same field as Figure 51. Slight leucine aminopeptidase activity is faintly discernible at the brush border of duo— denal villal epithelial cells. x 750. 148 Figure 53. Pig 9044. Age 3 days. Terminal ileum. E. coli was'introduced per os 20 hours be- fore necropsy. Complete loss of leucine aminopepti- dase activity from the villal epithelial cells. Unfixed frozen (-70 C) tissue. Method for leucine aminopeptidase (Nachlas et al., 1957). x 750. 149 Figure 54. Pig 9046. Age 2 days. Duodenum. Germfree control. Succinate dehydrogenase activity is clearly demonstrated in the epithelium of the villi and the crypts. Unfixed frozen (—70 C) tissue. MTT (Pearse, 1961). x 75. 150 .—‘t£‘l’g. Figure 55. Pig 9046. Age 2 days. Duodenum. Germfree control. An intense staining reaction for succinate dehydrogenase is apparent throughout the cytoplasm of duodenal villal epithelial cells. MTT (Pearse, 1961). x 750. 151 ,..r:".- Figure 56. Pig 9046. Age 2 days. Terminal jejunum. Germfree control. Succinate dehydrogenase activity in the villal epithelial cells of the ter- minal small intestine is less marked than that of the anterior small intestine. Compare this figure with Figure 55. Unfixed frozen (-70 C) tissue. MTT (Pearse, 1961). x 750. 152 :Figure 57. Pig 9047. Age 2—1/2 days. Terminal jejunum. Germfree control. Succinate dehydrogenase is localized in the apical cytoplasm and in a narrow band at the base of the villal epithelial cells. Un- fixed frozen (-70 C) tissue. MTT (Pearse, 1961). x 750. Age 2-1/2 days. Terminal E. coli was introduced per 08 8 hours before Eight hours after monocontamination there is a distinct, intense reaction, with clearly defined distribution of succinate dehydrogenase following the pattern of that seen in germfree control animal tissues. Figure 58. Pig 9038. ileum. necropsy. Unfixed frozen (-70 C) tissue. MTT (Pearse, 1961). x 750. 154 monocontamination there was marked loss of enzyme activity (Figure 59) and appreciable amounts of staining of material in the laminae propriae was observed. In each case the alteration, both in intensity of the staining reaction and in distribution of the stained material within the cell, was more marked in the more caudal sites in the small intestine. Fourteen hours after inoculation, the succinate dehydrogenase reaction product was less intensely stained and was more diffusely distributed throughout the cell, although the specific distribution of the enzyme at the extremities of the cell was still apparent (Figure 60). In those villal epithelial cells in which there was marked vacuolization, the displacement of the stained material by an intracytOplasmic mass was readily observed (Figures 60 and 61). In animals inoculated 24 hours previously, the reduction in staining intensity was apparent from the duodenum to the jejunum and terminal ileum, although the specific intracellular distribution of the enzyme in these tissues could still be seen. In some instances, by 28 or 32 hours after inoculation there was still an apparent diminution in the amount of enzyme present in the apical cytOplasm of the villal epithelial cells. Lactate Dehydrogenase.--In the duodenal epithelium of neo- natal germfree pigs, lactate dehydrogenase was concentrated at the apical third of the villal epithelial cell and in a thin line at the basement membrane (Figure 62). Using a shorter incubation time during tissue processing there was a less intense histochemical reaction, 155 Figure 59. Pig 9039. Age 2—1/2 days. Terminal ileum. E. coli was introduced per os 10 hours before necrOpsy. There is appreciable loss of succinate dehydrogenase activity from the villal epithelial cells. Unfixed frozen (-70 C) tissue. MTT (Pearse, 1961). x 750. 156 Figure 60. Pig 9041. Age 2-1/2 days. Terminal jejunum. E. coli was introduced per os 14 hours be— fore necropsy. Succinate dehydrogenase activity is markedly reduced, although specific intracellular dis— tribution of the enzyme is still apparent. Unfixed frozen (—70 C) tissue. MTT (Pearse, 1961). x 750. 157 Figure 61. Pig 9042. Age 2—1/2 days. Terminal. jejunum. E. coli was introduced per os 16 hours be— fore necrospy. In vacuolate villal epithelial cells from infected pigs, the succinate dehydrogenase is reduced in activity and the enzyme is displaced to the periphery of the cell. Unfixed frozen (-70 C) tissue. MTT (Pearse, 1961). x 750. 158 Figure 62. Pig 1005. Age 3 days: Duodenum. Germfree control. Lactate dehydrogenase is concen- trated at the apical third of the villal epithelial cell and in a thin line at the basement membrane. Cold neutral formalin fixation. MTT method for lac— tate dehydrogenase (Hess at al., 1958). x 750. 159 but a more elegant picture of the distribution of the enzyme was ob- tained (Figure 63). In control pigs in which the villal epithelial cells were vacuo— late, the lactate dehydrogenase was distributed around the vacuoles, probably due to the mechanical disruption of normal intracytOplasmic relationships, but there was still some of the enzyme concentrated at the basement membrane (Figure 64). Some staining of lactate de- hydrogenase reaction products was observed in the laminae propriae of the villi of the neonatal germfree pigs (Figure 64). There was no staining of lactate dehydrogenase reaction products in the laminae prOpriae of villi from control pigs in which the villal epithelial cells were apparently normal (Figures 62 and 63). It was observed in the germfree neonatal pigs that the most intense enzyme activity was in the duodenum and the anterior third of the jejunum. In the more caudal intestinal tissue sections the intensity of the staining reaction was diminished (Figure 64). In the monocontaminated gnotobiotes there was a noticeable dif- ference in the specific distribution of lactate dehydrogenase within the villal epithelial cell 24 and 28 hours, respectively, after exposure to the organisms. Concentration of the enzyme in the apical third of the cell could be seen in some instances, but the enzyme was more evenly distributed throughout the cell and extended centrally around the nucleus and to the basement membrane. This change was observed throughout the small intestine (Figure 65). In some instances a slight concentration of lactate dehydrogenase was seen in the immediate subapical cytoplasm of the cell. 160 Pig 1002. Age 3 days. Duodenum. Half normal incubation time was As a result, stain Figure 63. Germfree control. used in this tissue preparation. intensity is reduced, but a refined depiction of enzyme activity concentration in the apical cyto- plasm and at the base of villal epithelial cells is obtained. Cold neutral formalin fixation. MTT method for lactate dehydrogenase (Hess at al., 1958). x 750. 161 Figure 64. Pig 1005. Age 3 days. Terminal ileum. Germfree control. Villal epithelial cell vacuolization resulted in peripheral displacement of lactate dehydrogenase, but the concentration of the enzyme activity at.the base of the cells is MTT apparent. Cold neutral formalin fixation. method for lactate dehydrogenase (Hess et al., 1958). x 750. WI 162 Figure 65. Pig 1000. Age 3-1/2 days. Terminal ileum. E. coli was introduced per os 24 hours before There is more uniform distribution of active : 1 \ necropsy. enzyme throughout the cell and extending centrally around the nucleus and to the basement membrane. enzyme activity in the lamina propria is apparent. Cold neutral formalin fixation. MTT method for lactate dehydrogenase (Hess at al., 1958). x 750. Some 163 At 28 hours after inoculation, it was observed that there was loss of both specificity of distribution and intensity of staining of lactate dehydrogenase reaction products (Figure 66). In some instances staining of this material in the interstices of the lamina propria was observed, as it was observed occasionally in tissues from the germfree control pigs (Figure 64). Special Stain Techniques for Intracytoplasmic Contents and the Brush Border Oil Red 0 Stain At no stage was the accumulation of fat responsible for the vacuolization of the villal epithelial cells. 011 red 0 failed to stain any material in the vacuoles of either the control or the monocontaminated pigs. In those pigs fed within an hour prior to being killed, positively stained material was seen in the apical cytoplasm of the villal epithelial cells of the duodenum and of the ileum, and in some instances oil-red—O—positive material was observed in the lacteals. Oil-red-O—positive material was not seen in sec- tions from the terminal jejunum or ileum of clinically affected pigs (although some instances of fat absorption were seen in electron— micrographs of the occasional absorbing cells from comparable tissues). In no instance did the distribution of this material resemble that of the material in the vacuoles of the villal epithelial cells, and it was concluded that the vacuolization was not due to the collection of fat in the villal epithelial cells or the intramembranous space. AL. =‘ ‘ fibula—~— 164 Age 3—1/2 days. Terminal E. coli was introduced per os 28 hours be- There is loss of specificity of dis— tribution and intensity of staining of lactate dehy- Figure 66. Pig 1004. jejunum. fore necropsy. drogenase reaction products. Cold neutral formalin fixation. MTT method for lactate dehydrogenase (Hess at al., 1958). x 750. 165 Periodic Acid-Schiff Reaction The contents of vacuolate villal epithelial cells reacted nega- tively to the periodic acid-Schiff reaction. In some instances, in both control and monocontaminated pigs, the brush border reacted as strongly positive to this test as did all goblet cells and the Brunner's glands. \ Best's Carmine Stain The reaction of vacuolate cells to Best's carmine stain was negative. The only material positive for this stain was seen to be in the Brunner's glands, the goblet cells, and the brush border. In those villal epithelial cells undergoing marked vacuolization, there was no evidence that the material filling the vacuoles was Best's carmine-positive. In all instances neighboring goblet cells were positive for Best's carmine stain. Mayer's Mucicarmine Stain The reaction of vacuolate cells to Mayer's mucicarmine stain was negative. The extreme examples from the least to the most vacuo— late cells within the tissues from the germfree control pigs did not contain stainable mucin or seromucin. Similarly, in the mono- contaminated animals, none of the vacuolate cells contained stainable mucin or seromucin. In each instance the neighboring goblet cells were strongly positive for mucin. Alcian Blue Stain In both germfree and monocontaminated pigs, the only alcian- blue-positive material was seen to be in the Brunner's glands and 167 Figure 67. Pig 1008. Age 4 days. Terminal jejunum. Midvillus- Germfree control. Identifiable at this magnification are the apical tubule system, mitochondria, Golgi complex (Cc), and lysosomes (1y)- The brush border is moderately deep and electron dense. Mitochondria are plentiful in both the apical and basal cytoplasm. The Golgi c18- ternae are moderately vacuolate. The close association of villal epithelial cells with the vasculature of the lamina propria is apparent. Sodium barbital-buffered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. x 9,100. Figure 67 167 Figure 67. Pig 1008. Age 4 days. Terminal jejunum. Midvillus. Germfree control. Identifiable at this magnification are the apical tubule system, mitochondria, Golgi complex (Gc), and lysosomes (1y)- The brush border is moderately deep and electron dense. Mitochondria are plentiful in both the apical and basal cytOplasm. The Golgi cis- ternae are moderately vacuolate. The close association of villal epithelial cells with the vasculature of the lamina propria is apparent. Sodium barbital-buffered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. x 9,100. Figure 67 169 Figure 68. Pig 1008. Age 4 days. Terminal jejunum. MidV111us- Germfree control. Identifiable at this magnification are microvilli, mitochondria, apical tubule system (a), end0p1asmic reticulum (Er), Golgi cisternae, and mural plasmalemmal interdigitations (p). The microvilli are moderately short and do not appear to be rigid. The cYtoskeletal forms are indistinct and there is no evidence of a termi- nal web. Multiple supranuclear membrane enclosed vacuoles vary mar- kEdly in size, from obvious dilatations of the apical tubules (1). to more voluminous vacuoles probably formed within the endoplasmic reti- culum (2). The contents of the vacuoles are particulate, probably protein, detached ribosomes, and absorption products. Sodium barbital- bU§§ered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. x ,1000 a". - ‘_—" 6“? w ~ ".3 - .v-v... “ =15;st 7 i... '1! ."" alga, " "iv Figure 68 171 Germfiiguzg :9.1 P18 1008' A83 4 days. Terminal jejunum. Midvillus. dria and thniro - Identifiable at this magnification are mitochon- rosettes ( e r cristae mitochondriales, polyribosomes (r) and glycogen of the lami), 60131 cisternae, and endoplasmic reticulum. Features ial cells 2: EEOpria include the basal lamina (arrow) of the epithel- of a thrombo ut interstitium, capillary endothelium (ca), portion cells -- rosybe (p), and cYtOPlasmic processes of connective tissue tetroxid pf a y fibr0b188t8 (F). Sodium barbital-buffered osmium e ixation. Lead citrate-uranyl acetate stain. x 23,800. "K;"-IIIIIIIHMMZ*F‘V#Ffi5{55" ”I '11" "57.1““? map: 1"?" I .9 I?! ‘ 1:9" ”Win r.)- N; . (if: ,t'Vh_ u A “'“mmi; ’3‘ ‘3; Figure 69 ' 173 Figure 70. Pig 1002. Age 3 days. Terminal ileum. Basal third of villus. Germfree control. Granular reticulum (rough endoplasmic retiCU1um) (Gr), and lysosomes (1y) are distributed throughout the api- cal cytoplasm. Individual microvilli consist of a cylinder formed from the trilaminar membrane of the apical plasmalemma enclosing a structural complex, the cytoskeleton, terminating in a cytOplasmic con- densation, the terminal web (tw). The microvilli, though long and slender, appear to be rigid. There is no evidence of an apical tubule system in this cell. Cristae mdtochondriales are visible as parallel, nonvacuolated membranes (arrow). Sodium barbital-buffered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. x 30,600. .- ' ..., ;_ 3.1‘11 a , j ’. ' . up (hr!) 0'“, _;r\. w . TEAL ,sna. n. s; .; .1512. nae-ii? i Jar-:35) radf" Til-1 r-vv". 131,5 , a: 1’ .' Figure 70 ‘o‘clb? 175 cytoskeleton consisted of a central rod surrounded by six radially placed skeletal rods of an apparently similar nature. There was some evidence of a meagre hirsute layer, the glycdcalyx, on the outer surface of the microvilli (Figure 71). Generally it was observed that in germfree neonatal pigs, the microvilli of those cells in which there was an active apical tubule system were somewhat shorter and more intensely stained (Figures 67 and 68) than those of more recently generated cells in which there was no evidence of an apical tubule system (Figure 70). In these cells the microvilli were longer with a more densely staining cytoskeleton terminating in an indistinct, finely granular layer. corresponding to the terminal web (Figure 70). It was also appar- ent that the microvilli of these cells were more rigidly structured than those of cells containing an apical tubule system (Figures 68 and 70). V I In the older, more mature cells (generated in the neonatal pig before, at, or soon after birth) in which there was an apical tubule system, the microvilli were seen to be shorter and less rigid, the cytoskeleton less electron—dense and with no apparent foundation in a terminal web (Figure 68). ' In control and monocontaminated pigs 2 to 5 days old, the apical tubule system was clearly recognizable, extending 2 or 3 u into the apical cytOplasm. The formation of‘pinocytotic vacuoles and the direct continuity of the tubules at the bases of the microvilli with 'the apical tubule system was clearly apparent (Figures 67, 72, 73 and 74). 176 Figure 71. Pig 1002. Age 3 days. Terminal jejunum. Germfree control. Transverse section of the microvilli. Each microvillus is a cylindrical tube formed of the trilaminar membrane from.the apical plasma- lemma. This membrane encloses the cytoskeleton, six rod-like structures placed radially around a central skeletal rod. There is evidence of a meagre hirsute layer, the glycocalyx (arrow). Sodium barbital-buffered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. x 134,000. .4 "—r:rln 178 Figure 72. Pig 1310. Age 4-1/2 days. Terminal jejunum. Midvillus. Germfree control. The formation of PinocytOtic veSiC1es and the r1:0. lationship to the apical tubule system is apparent. The cristaesm omes chondriales are of uniform width and are not vacuolated. Polyrl OS are dispersed throughout the apical cytOplasm' An apical vasuo gédium probably associated with active absorption. 13 membrane boun ' barbital-buffered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. x 32,200. 179 5".» ‘4' 1 :" .I' a. kit's V‘ a was!” ., ”I: $31.0 :._0q:¢gd. 1...“, ‘5 yarn-W4 -- Figure 72 180 Figure 73. Pig 1306. Age 4-1/2 days. Terminal jejunum. Mid- villus. E. coli was introduced per 03 8 hours before necrOpsy. For- mation of pinocytotic vesicles at the bases of the microvilli and their relationship with the apical tubule system is apparent. The contents of the vacuoles are particulate, probably consisting of protein, some detached ribosomes and absorption products. These membrane lined vacuoles are indistinguishable from the absorption vacuoles seen in cells from germfree neonatal pigs. There is an indistinct, slight vacuolization of the cristae in the mitochondria of the apical cyto- plasm. Sodium barbital-buffered osmium.tetroxide fixation. Lead citrate-uranyl acetate stain. x 11,200. ...---A .‘ fl ~ ‘ ‘9" ..L I“! ’_M’Ixh'fl'.' link - Figure 73 182 Figure 74. Pig 2251. Age 2—1/2 days. Terminal ileum. Midvillus. E. coli was introduced par os 14 hours before necrOpsy. The formation of pinocytotic vesicles and their continuity with the apical tubule system is clearly visible (arrow). Early vacuolization of the cristae mitochondriales is apparent, as is breakdown of the laminae enclosing the absorption vacuoles. Sodium barbital-buffered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. x 11,200. .2; . ‘ 43%)},5 _, " ‘ V ‘13, Figure 74 Illnahunfl“ “an 1". 184 Early formation of membrane enclosed vacuoles, generally apical to the nucleus, was seen (Figure 67). All stages of intracytoplasmic vacuolization comparable to those described for germfree neonates under "HistOpathologic Findings" were observed with the electron microsc0pe. These vacuoles ranged from those that were apparently dilatations of tubules of the apical tubule system (Figure 68) to multiple supranuclear vacuoles (Figures 68 and 75) and vacuoles so large as to be incapable of being contained in the field of the lowest acceptable tap-meter setting of the elec— tron microscope (Figure 76). These vacuoles were seen to be membrane enclosed and, where adequately resolved by the electron microscope, this membrane was seen to be a monolaminar structure, probably derived from the endo- plasmic reticulum (Figures 68, 72, 75 and 76). Particulate matter of a flocculent nature, some of it probably coagulated protein and some unattached ribosomes, was floating free in the vacuoles. Free membranes or membrane fragments were not seen in the vacuoles of cells from germfree neonatal control pigs. In the cells examined from both jejunum and ileum of the germ- free control pigs aged 2 to 5 days, organelles comprising mitochon- dria, lysosomes, granular endoplasmic reticulum and polyribosomes were seen to be present throughout the apical and basal cytOplasm; and mitochondria, particularly, were apparent in the subapical zone immediately beneath the terminal web (Figures 67, 70, 72 and 76). The cristae mitochondriales of cells from germfree neonates were con— spicuous by their uniformity of width and lack of vacuolization (Figures 69, 70 and 72). 185 Figure 75. Pig 1005. Age 3 days. Terminal ileum. Basal third of villus. Germfree control. Multiple supranuclear membrane enclosed vacuoles vary markedly in size, from dilatations of the apical tubule system (3) to large vacuoles probably formed within the end0p1asmic reticulum (4). The vacuole contents are particulate and probably con- sist of protein, detached ribosomes, and products of absorption. Free membranes or membrane fragments are not apparent. Sodium barbital- bufgered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. x l ,800. . 2L, ; . > .r‘: " l V J V 4 ’1‘” w -. x , .. .§ 9" I J '..- v7.l'\ I saw. 3'2. -" 187 Figure 76. Pig 1310. Age 4-1/2 days. Terminal jejunum. Mid- villus. Germfree control. Multiple supranuclear vacuoles are membrane-enclosed and are probably derived from dilatations of the apical tubule system (5) and the end0p1asmic reticulum (6). Note that free membranes or membrane fragments are not seen in the vacuoles in germfree neonatal pigs. Sodium barbital-buffered osmium.tetroxide fixation. Lead citrate-uranyl acetate stain. x 9,100. 6 7 .... u g i F 189 In germfree pigs, 2 to 5 days old, the Golgi complex was generally paranuclear and sometimes the Golgi cisternae were moderately vacuolated (Figures 67, 69 and 77). Occasional abnormal situation of the Golgi complex was associated with marked vacuolization in which the nucleus was pushed into an unusual position within the cell. EndOplasmic reticulum, both granular and agranular, was recog- nizable in the apical and basal cytOplasm of the villal epithelial cells (Figures 70, 78 and 79). In some instances, end0p1asmic reticulum was compressed or displaced to the periphery of the cell by intracytOplasmic vacuoles (Figures 75 and 78). Readily identifiable lysosomes and microbodies were not common (Figures 67, 70 and 77), but numerous polyribosomes, glycogen rosettes (identifiable particularly in those sections stained with lead citrate) and absorbed material such as fat draplets, were come mon both in the cytoplasm and within intracytOplasmic membrane sys- tems (Figures 69, 70, 72, 77, 79, 80 and 81). The intricate interlocking of the mural invaginations was remarkable (Figures 68 and 82) in tissues from both germfree and monocontaminated animals, but as cells became vacuolate and swollen the complexity of the interdigitations was reduced until in some instances interdigitations were not apparent (Figures 75 and 76). Generally the villal epithelial cells were evenly placed upon the basal lamina, in close association with the endothelium of the vasculature of the lamina prOpria (Figures 67 and 69). 190 Figure 77. Pig 2253. Age 2 days. Terminal jejunum. Basal third of villus. Germfree control. Notable in the cytOplasm.are occasional lysosomes, absorbed free intracytOplasmic fat drOplets (f), numerous polyribosomes and glycogen rosettes. The interstitium of the lamina propria consists of a few fine collagen strands, chylomicrons and fat drOplets in a fluid matrix, a fibroblast, and the vascular structures. It is enclosed within the basal lamina of the epithelium. The endo- thelial cell of the lymphatic vessel is recognized by its intracyto- plasmic fat droplets, relative.lack of micropinocytotic vesicles and mitochondria, and delaminating internal membranes. The basal laminae of both endothelium and epithelium are marked (arrow). Sodium barbital- bugfered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. X , 00. .. "m“ r: ‘13; . . . x‘rd‘ra‘g 1r,- ‘ . ; so. leap. 2”")“31 7 .:..n$' .34.?“ 72"" Hire-um. 4:22.12 fz‘imip; wash... am im: ... 3 .n. r,¢JxJa 295.2%.V '15 hr“ g: ' "fi'fii', ‘ “If: Wk): 1“ :1 115 x. ‘ u ' flifi Max:333 56:31 W tau-193.5. :: .. ;' 3" . ’ ' ' " ‘mu lvtfifimmg1 .,9_ . Figure 89 219 Figure 90. Pig 1001. From the same field as Figure 89. Detail of disruption of the cytostructure by membrane-lined, and nonmembranous, vacuoles. In a neighboring cell there is vacuolization of the cristae mitochondriales. Sodium barbital-buffered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. x 17,500. Figure 90 221 Figure 91. Pig 1001. Age 3-1/2 days. Terminal jejunum. ‘Mid- villus. E. coli was introduced per os 24 hours before necropsy. Non- membranous vacuolization of almost the entire cytOplasmic space, in- cludes cell debris, the nucleus and other organelles, together with many membrane lined vacuoles and tubular fragments. Marked internal vacuolization and disarray of the cristae of many of the mitochondria (M) is clearly apparent. Sodium barbital-buffered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. x 14,400. - ln-n «.- f. .i Figure 91 223 Figure 92. Pig 2258. Age 3 days. Terminal jejunum. Midvillus. E. coli was introduced per os 18 hours before necropsy. The contents of the multiple membranous vacuoles in the apices of these villal epi- thelial cells become confluent as their separating membranes rupture. Eventually, most of the cytOplasmic space is occupied by the vacuoles. Sodium barbital-buffered osmium tetroxide fixation. Lead citrate- uranyl acetate stain. x 11,200. , ‘ Al . .1 1,." ,n 1 -';.*.,F “,3: . '. i‘.’ . ' f.» 9 ._ 'f w: i'; )1“, r '71 ,u ’. - No". ‘3..‘1i“',. 5| .1 r. . ' ,_ 7 . 1, {.119}: ' v 4 W 9 a A”? E '4 .. v the .1 '_1- ‘1! - . 1" -'-’ '” I. h t??? "(3:935'. 3.4) ‘43: _"-. i. Figure 92 225 mitochondriales. COpious amounts of absorbed material such as lipid were present in the cytOplasmic matrix, within the endoplasmic reticu- lum, and in the basal intercellular (intermural) space (Figures 79 and 80). In epithelial cells in which there were early signs of mitochon- drial degeneration, but not extensive vacuolization, the interdigi— tations of the cell membrane with those of neighboring cells was apparently unaffected (Figures 82 and 87). There was little change in the brush border that could be ascribed to the effects of exposure of this structure to E. coli for a period of 8 to 18 hours. In some cells with mitochondrial changes, the brush border appeared normal (Figures 81, 83 and 84), yet in some instances there was evidence (Figures 84 and 85) that the glycocalyx was very much reduced or missing. Variations in the height and'density of the brush border were inconsistent, and only in grossly disrupted cells was it apparent that the brush border was morphologically incapable of function (Figure 88). In other cells undergoing early degenerative changes the brush border was remarkable for its density, depth and integrity (Figure 87). It was observed that the microvilli were generally longer in the less mature cells and that, unless the degeneration associated with infection was very marked, the further along the villus that cells were sampled, the shorter were the microvilli (Figures 67 and 70). Between 8 and 24 hours after infection there was observed to be an increase in the amount of interstitial fluid present in the 226 intramembranous space of the lamina propria of the villi. Sometimes this was clear fluid, and in these instances the associated capillary or lymphatic endothelial cells were micrOpinocytotically active (Figures 86 and 93). Such edema was associated with villal epithelial cells in which early retrogressive changes were apparent (Figure 86). At times the interstitial fluid contained free floating cytoplasmic. membrane fragments, occasional chylomicrons and unidentified particu- late matter (Figure 94). The basal laminae of the capillaries were seen to be much thicker at this stage than the basal lamina of the neighboring epithelium. Again, in other instances, the interstitium of the basement membrane was seen to be edematous but with additional collagen fibers present. In these instances the capillaries were less micrOpinocytotically active (Figure 95). By 24 hours after inoculation, the contaminating organisms had become so plentiful in the intestinal tract that sections of the organisms were found during the electron microscopic examination. Ehoherichia coli 0138:K81:NM was seen to be nonfimbriated and of characteristic and readily identifiable structure (Figures 96, 97, 98 and 99). At no stage of this investigation were organisms seen within the villal epithelial cells, nor was there evidence of E. coli organisms having entered the apical plasmalemma, apical tubule sys- tem, or any of the cytoplasmic substance of the villal epithelial cells. Where the organisms were in close association with, and possibly attached to, the brush border, the microvilli were seen to be 227 Figure 93. Pig 1001. Age 3-1/2 days. Terminal jejunum. Mid- villus. E. coli was introduced per 03 24 hours before necrOpsy. There are increased amounts of interstitial fluid containing suspended chylo- microns, flocculent material, and fine collagen fibrils. Slight micro- pinocytotic activity and relative lack of laminar folding is charac- teristic of normal lymphatic endothelium. Sodium barbital-buffered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. x 11,200. -’ -‘-m—- —._—_\ m .ofi‘f'fi" {\t'f - , ' as?“ ago! \2 to «v: o. -M'N‘W 111452.. we #3311 In shun magi?! -.§..i':’"‘." and .U o ,4! t”: has '- a}. 3."b(c1 warm. 3c: 1341.! evnu- 50‘s ~(.Jf.-nt..; .. final" 'W'IMZ 4.2505". .5. (Lladiobfla .2. 'xrzf Lammr mm... man-ct. leagas-p 1.2.3135: ham: . is a} 9hr: Figure 93 229 Figure 94. Pig 2258. Age 3 days. Terminal jejunum. Midvillus. E. coli was introduced per os 18 hours before necropsy. The inter stitial tissue fluid of the lamina prOpria of this villus contains free, floating, cytoplasmic membrane fragments, occasional chylomicrons, collagen fibrils and unidentified flocculent material. Sodium barbital-buffered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. x 23,800. J." "- t :rr 1"! . 4!- . Figure 94 231 Figure 95. Pig 2251. Age 2-1/2 days. Terminal jejunum. Mid- villus. E. coli was introduced per os 14 hours before necrOpsy. There is marked interstitial thickening and edema of the lamina pro- pria of this villus. Compared with other sections, there is extens- ive collagen production, increased thickness of the capillary basal laminae, and reduced endothelial micropinocytosis. Sodium barbital- buffered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. x 11,200. 5 9 m u on .1 F 233 Figure 96. Pig 1001. Age 3-1/2 days. Terminal ileum. Basal third of the villus. E. coli was introduced per os 24 hours before necrOpsy. The monocontaminating organisms are of characteristic struc- ture and are nonfimbriated. They are closely associated with atrophic microvilli, whereas it can be seen that the E. colt organisms are separated from the unaffected microvilli by a relatively uniform, nonstaining layer, presumably a glycocalyx. There are some slight changes in the cristae mitochondriales of the affected cell; however, the Golgi complex and its cisternae appear to be normal. Sodium barbital-buffered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. x 28,000. ______ -_ -.-L-’ .m".:. - .i'afl’". ' W' a v-r' O ,fl' 0" ’m.’ T‘- Figure 96 235 Figure 97. Pig 1001. Age 3-1/2 days. Terminal ileum. Basal third of the villus. E. coli was introduced per os 24 hours before necropsy. The organisms are nonfimbriated and are closely associated with atrOphic microvilli of the apical plasmalemma of a mature affect- ed cell (above), its maturity characterized by the presence of an apical tubule system. The microvilli of the younger villal epithelial cell (it has no apical tubule system) have had less contact with the E. coli organisms and are apparently unaffected. Sodium barbital- buffered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. x 28,000. 3 236 .4 Figure 97 237 Figure 98. Pig 1001. Age 3-1/2 days. Terminal ileum; Basal third of villus. E. coli was introduced per os 24 hours belozeChar_ necropsy. Transverse sections of E. 002% 0138:K81:NM revea 118 or acteristic internal structure. There is no evidence of flage 1111 fimbriae in the capsule of these organisms. Although the microv1 of the least mature cell in this field appear unaffected (thereh :e some evidence of a meagre hirsute layer at the lumen surface) ; edria is slight vacuolization of the cristae within the apical mitoc on . Sodium barbital-buffered osmium tetroxide fixation. Uranyl acetate stain. x 32,200. ‘.J .1" ... ' . ,é’->s’..~rn_ ,_ 1‘14". ‘63 chi ‘ ”...-3’”: 1.91.1.3. tug: by! “1' «1.4: “W “...". .451 Figure 98 239 Figure 99. Pig 1001. Age 3-1/2 days. Terminal ileum. Basal third of villus. E. coli was introduced per 08 24 hours before necropsy. The plasma membrane of E. coli 0138:K81:NM (arrow) is a trilaminar structure. The microvilli in close association with the bacterial wall are markedly atrOphied, and cytoskeletal structure is not apparent. The trilaminar structure of the membrane of both the microvilli and the apical tubules can still be seen. There is pre- sumptive evidence of either a thin electron-transparent bacterial capsule, or the remnants of an electron-transparent glycotalyx, be- tween the tips of the microvilli and the external trilaminar membrane of the microorganism. There is no evidence of flagella or fimbriae. Sodium barbital-buffered osmium tetroxide fixation. Lead citrate- uranyl acetate stain. x 133,700. Figure 99 241 markedly foreshortened into stubby or globular, sometimes branched, forms in which the glycocalyx was reduced and in which neither the cytoskeleton nor its foundation in a terminal web could be detected (Figures 96, 97 and 99). Early vacuolar changes were detected in the cristae mitochondriales of nearby mitochondria (Figures 96 and 100) and at times these mitochondrial changes were observed in cells in which morphologic changes were not yet apparent in the microvilli (Figure 98). Transverse sections of microvilli from comparably affected cells were examined. The proximity of neighboring microvilli was sug- gestive that the glycocalyx was reduced in depth (Figure 101). The lack of cytoskeletal detail in these microvilli compared to transverse. sections of microvilli from villal epithelial cells of germfree con- trol animals (Figure 71) was apparent and coincided with the obser- vations made of longitudinal sections of microvilli from monocontami- nated animals. Twenty-four hours after inoculation, the vacuolar degeneration and disruption of organellar membranes was clearly apparent (Figures 89, 90 and 100). In more advanced degenerative states, the nuclei and membranous vacuoles were seen to be floating free in a matrix of fluid and suspended mitochondria, tubular fragments, membrane frag- ments, polyribosomes, and cytoplasmic debris (Figure 91). By 32 hours after inoculation, in affected cells the mitochondria were swollen, with disruption of the cristae mitochondriales, the vacuolar degenerative changes were maximal, and only traces of nor- mal cytostructure remained (Figure 102). Vacuolate cells became 242 Figure 100. Pig 1001. Age 3-1/2 days. Terminal ileum. Midvil- lus. E. coli was introduced per os 24 hours before necropsy. There is one markedly swollen mitochondrion, in others there is distinct vacuolization of the cristae mitochondriales; the microvilli appear foreshortened and their cytoskeletal structures are not apparent. Sodium barbital-buffered osmium tetroxide fixation. Lead citrate- uranyl acetate stain. x 28,000. _ .-rq -.. 0 ‘3 aim-.31 .132in hut/r51 fang-» Rte-C H.“ $33.; at": ~...-'.-1.r :1» turn? ~ var-:1. 9.103115 *tmd »" 1’. “sq 53:10.20: 'r'i' wrw " . \ :or'H‘elb e1 e~-...;- #111136 .1? 1‘" .5011 a .‘cft'otir reflux . iii-1 .' 9.23;.- H'.’ 7033.14" v.3 “1111.15.15.13” r. 2122’ *r ...13 3.1 mm; ' {quanta m“ a: '. ear! :31 “rag .Le‘_.."Haou.. '...~.~;..‘ 1m bush.- " ”5;",‘73 buy- .totzfru’ 115x13 : a... we Hiau‘iufi-“s u ”.3 :.~ -000 L1 x .t.:.: 9:3;131 t; Figure 100 244 Figure 101. Pig 1001. Age 3-1/2 days. Terminal ileum. Midvil- lus. E. coli was introduced per os 24 hours before necropsy. Trans- verse section of microvilli from cells nearby to those shown in Figure 100. The proximity of neighboring microvilli is suggestive that the glycocalyx is reduced in depth, although in some areas a hirsute layer is present where the preparation is adequate for the representation of fine detail of the trilaminar membrane (arrow). Cytoskeletal detail is lacking. Sodium barbital-buffered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. x 104,300. 1 V “ .”"&4I ~ \ .- ' fl 1 Q 1 ug- 246 Figure 102. Pig 1006. Age 4 days. Terminal ileum. Midvillus. E. coli was introduced per os 32 hours before necropsy. Little of the normal cytostructure is retained in this vacuolate cell. There is marked disarray of the apical plasmalemma and microvilli; the mito- chondria (M) are swollen and broken, and there is disruption and degeneration of the cristae mitochondriales. Sodium barbital—buffered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. x 13,500. - '33 “"3“. 7" fa . ' ,‘W'fl t" Figure. 102 248 confluent with similarly degenerated neighboring cells, and the con- fluent cytoplasm consisted of broken membranes and organellar frag- ments floating in a fluid matrix (Figure 103). Eventually the apical plasmalemma was seen to be broken and the disrupted cell contents released to the lumen of the intestine (Figure 104). By 96 hours after inoculation the only additional remarkable feature was that of the degree of collagen formation in the intra- membranous space of the midvillal lamina prOpria (Figure 105). Such collagen formation was not a feature of the midvillal lamina prOpria of jejunal or ileal villi of germfree control neonatal pigs. 249 Figure 103. Pig 1006. Age 4 days. Terminal jejunum. Midvillus. E. coli was introduced per os 32 hours before necropsy. The confluent cytOplasm of neighboring degenerate cells consists of broken membranes and organellar fragments floating in a fluid matrix. The nucleus appears relatively unaffected. Although now separated from their attachment to the parent cells, the interdigitated mural plasmalemmae remain adherent as a unit (arrow). Sodium barbital-buffered osmium tetroxide fixation. Uranyl acetate stain. x 22,500. 1- . m '11" .-_-'_ «(‘3 In “I 0‘ 31 5'. . tags; .05. ... r in and“ he; Amour 2-. a... 1. 'NW Ia runner a! .113 .BM‘Job {cl :0; 131.1: Lo :- ‘7 ah 2e- (can I 11‘ was.“ among”! 131.111.5215 2113141 .maenm tievnsir 5,-7.2 , , Jun; .11: c me- . 1"."1 I .3 31:1" ii: .12), Hbflu; Figure 103 251 Figure 104. Pig 1006. Age 4 days. Terminal ileum. Midvillus. E. coli was introduced per os 32 hours before necrOpsy. Vacuolization and disruption of the cytOplasmic structure, associated with morpho- logic change in the apical plasmalemma, have resulted in rupture of the brush border and extrusion of the cell contents toward the lumen of the intestine. Sodium barbital—buffered osmium tetroxide fixation. Lead citrate-uranyl acetate stain. x 18,900. .ld. .0 l e r u g .1 F 253 Figure 105. Pig 1009. Age 6—1/2 days. Terminal jejunum. Mid- villus. E. coli was introduced per 08 96 hours before necrOpsy. With prolonged monocontamination the intramembranous space of the midvillal lamina prOpria is markedly infiltrated with collagen. Such collagen formation is not a feature of comparable tissues from germfree pigs but may well be present in the laminae prOpriae of conventional, clini- cally normal, neonatal pigs. Sodium barbital-buffered osmium tetroxe ide fixation. Uranyl acetate stain. x 9,100. Figure 105 DISCUSSION "The other error...is that of thinking Science to consist in mere observation, and that the accumulation of records, without generaliza- tion, hypothesis, or the use of instruments which the understanding has forged, can ever be of the slightest advantage in the advance- ment of knowledge". F. G. Crookshank, 1922 The events leading to diarrhea in the neonate are interrelated and occur sequentially and often simultaneously. For the sake of clarity, and in relation to the voluminous literature that is availa— ble on the enteritides, malabsorption and diarrhea, these events will be discussed separately, somewhat along disciplinary lines. The difficulties encountered by Barnum at al. (1967) in produc— ing enteric colibacillosis under controlled experimental conditions and used by these authors as " ...the single greatest argument against a primary causative role for E. coli in the disease" (colibacillary diarrhea) have not been encountered in this laboratory. The fact that, under gnotobiotic conditions, the administration of known num— bers of organisms of a specific serotype of E. coli to neonatal gnotobiotic pigs is consistently followed some hours later by severe diarrhea in these animals; that consistently the same organism can be recovered in_pure culture from these animals and not from clini- cally unaffected, germfree litter mates kept under identical condi- tions; and that the recovered organism will reinfect 255 256 experimental neonatal pigs, is demonstration of the primary causative role for E. coli in the disease, colibacillary diarrhea. Clinical Findings The single litter of older pigs used in these experiments (Pig numbers 0657 through 0663, Tables 1 and 2) were not so adversely affected clinically (see Table 4) by infection with the test organism, as were the gnotobiotic neonates in the remaining 4 litters (Tables 1 and 4). It was clearly evident that neonatal, colostrum-deprived pigs are readily infected by oral contamination with 3 x 106 E. coli 0138:K81:NM organisms and that severe diarrhea results. It is ,apparent that even a brief period of agalactia in.a parturient sow, thus preventing pigs from.ingesting colostrum, would permit the gastrointestinal tract of neonetal pigs exposed to pathogenic sero- types of E. coli to become heavily contaminated, thereby constitut- ing a serious threat to the survival of the litter. It is likely that the profuse diarrhea characterizing colibacil- losis in gnotobiotic pigs due to E. coli Ol38:K81:NM is dependent on the presence either in the mucosa of the intestinal tract or the lumen of this organ, of sufficient numbers of the organism, and a remarkably consistent latent period of 8 to 16 hours (Table.4). It is apparent that the oral dose of approximately 3 x 106 organisms of this serotype is sufficient to produce diarrhea in a gnotobiotic pig within 12 hours of the exposure. In the gnotobiotic pigs systemically infected with this sero- type of E. coli in a previous experiment (Christie, 1967), it was 257 observed that sufficient time must elapse for the organism to colonize the gastrointestinal tract, and in sufficient numbers, before the physiological activity of the intestine was so altered as to produce diarrhea. Escherichia coli 0138:K81zNM consistently produced very pro- nounced diarrhea and dehydration and, under field conditions, one would anticipate infection of neonatal pigs with this particular serotype to result in a relatively high mortality rate. With this serotype, in those instances where the perineum was grossly and persistently contaminated with feces containing the organism, erythema venenatum was observed. In these circumstances it is not surprising that, in earlier experiments with the same sero- type, an occasional urocystitis, probably from an ascending infection of the female urethra, was recorded (Christie, 1967). The clinical signs of the disease produced by oral exposure of gnotobiotic neonatal pigs with E. coli Ol38:K81:NM closely resembled the clinical descriptions given elsewhere for other pathogenic sero- types of E. coli (Gordon and Luke, 1958; Saunders et al., 1960; Stevens, 1963; Dunne, 1964). Anorexia may or may not be a part of the clinical syndrome as it is described in the literature. With more pathogenic serotypes of E. coli, anorexia is more apparent as the pigs become dehydrated and lethargic, and it is likely that anorexia is a sign of a deteriorating clinical state rather than a specific manifestation of colibacillosis. Anorexia was not a sig- nificant clinical finding in the experiments reported here. 258 Other clinical signs of diarrhea and dehydration -— the physical nature of the fecal material, changes in the hair coat, erection of the hair shafts, and postural attitudes characteristic of scouring baby pigs, described by others in clinical reports of neonatal coli— bacillosis, were consistently present in those animals left alive for a sufficiently long period of time in these experiments. HistOpatholggic Finding§_ Histopathologic Lesions in Tissues Other Than the Alimentary Tract The histOpathologic lesions of E. coli 0138:K8lzNM infection of gnotobiotic neonatal pigs in tissues other than the alimentary tract have been described previously (Christie, 1967). Generally these changes were of a vascular nature, indicated by edema, and epithelial degenerations, such as hydropic degeneration, in the respiratory and urinary tracts and the aortic endothelium. The epithelial and endothelial changes were inconsistent and were possibly associated with incidental local concentrations of endotoxin or enterotoxic activity of the test organism. Histopathologic Lesions in the Alimentary Tract The changes in the epithelium and the lamina prOpria of the villi recorded as being common to both infected and germfree pigs are described in detail in "Histopathologic Findings" because the difference between the groups is distinctly one of degree. In the process of absorption of fluids from the lumen of the intestine, some of the histologic changes observed in the experi- mental pigs are likely to be found regardless of the presence or 259 absence of E. coli. The presence of enterotoxin of E. coli so modi- fies these changes that they occur in cells situated nearer to the base of the villus, and at an earlier stage in the short lifetime of the cell, than in cells of comparable tissue from germfree neonates. It is in this way that the histologic changes become significant in a pathologic sense. The Lamina PrOpria. It is suggested that the hyperemic and inflammatory changes found in the infected pigs are manifestations of a response to E. coli endotoxins, whereas the less marked histologic changes such as cytOplasmic vacuolization and breakdown which probably account for the major clinical sign of diarrhea are possibly due to other com- ponents of enterotoxin (possibly an exotoxin) and resemble to a degree some of the physiologic changes that one can see in the di- gestive tract of the germfree pig. Smith and Jones (1963) and Moon at al. (1966) have reported that there were no histologic indications of any inflammatory response in the intestinal tract of pigs with colibacillosis. From the histo- logic studies made with E. coli Ol38:K81:NM in these experiments and those previously reported (Christie, 1967), it is clear that there are inflammatory changes associated with colibacillosis but that these changes are insufficiently pronounced to be readily apparent in tissues taken from field cases of the disease. There are appar- ently no lesions pathognomonic of the infection at an optical- microscopic level. Occasional foci of neutrophils and other inflame matory exudates were observed in sections from the lamina propria and 260 submucosa of the intestinal tract of these experimentally infected pigs. Endotoxins have long been known to induce neutrophilia (Carpenter, 1965), and it may be that, as organisms multiply and they or their products "spill over" into the lamina prOpria from various sites of phagocytosis or localization in the crypts, they or their endotoxins can produce transient microabscesses in these tissues (Figures 1 and 2). Bacteria stimulate an inflammatory response in the "normal" intestine of the healthy conventional pig, and we have come to accept a constant state of acute inflammation (characterized by the presence of neutrOphils and lymphocytes) as being the normal. It is now apparent that there is a difference between the "normal" inflammatory state of the intestine in which normal function proceeds in tissues that have developed a balance between cell formation, cell function, and cell destruction (and are reacting to the con- stant threat of invasion by these organisms) and actual tissue damage by the products of their existence such as enterotoxins. The histOpathologic changes in the intestinal tissues of the infected pigs from each of the experimental litters were essentially the same. The ileum and terminal jejunum.were more severely affected than the duodenum and anterior jejunum as judged by villal epithelial changes and inflammatory changes in the lamina prOpria. Varying degrees of subepithelial edema, involving the inter- stices of the lamina prOpria, were associated with two possible states within the villus -- active absorption and colibacillary enteritis. There were several poorly defined conditions apparent 261 within the laminae propriae which enabled an arbitrary decision to be made concerning which condition, absorption or malabsorption, was present. Confirmation of these somewhat subjective judgments was obtained during electron microscopic examination of infected and control tissues. The optical microscopic features associated with active absorp- tion were considered to be: moderate hyperemia, moderate distention of the lacteals, some interstitial edema (confined generally to the immediate vicinity of the capillaries and lacteals), and activity of the Breucke fibers -- indicated by their appearance in sections as short, blunt, contracted cells. Those features considered indicative of absorptive inadequacy of the laminae propriae associated with colibacillosis were: relaxa- tion (or paralysis) and elongation of the Breucke fibers; sometimes marked hyperemia, at other times diminished capillary activity (according to the stage of the condition and level of enterotoxin in the environment); pooling of lymph in the lacteals; and particu- larly collection of fluid in the immediate vicinity of the basal lamina of the epithelium. Sometimes the collection of fluid became so marked that the fluid-filled spaces were interlaced with strands of fibrous stroma, and the villal stalk was noticeably distended (Figures 15, 16 and 25). There were occasions when one or more of these evidences of villal dysfunction were seen in villi from con- trol animals and hence could be misleading in defining colibacillary enteritis if not considered together with the other Optical and electron microscopic lesions of the disease. 262 In discussing the possible pathogenesis of these changes in the lamina prOpria, some consideration has already been given to the likely role of the endotoxic fraction of E. coli enterotoxin (aids supra), but some further amplification is necessary. Bacterial exotoxins are potent, relatively unstable compounds with highly specific toxic effects (often sufficiently characteristic to identify the toxin) and are generally highly antigenic. Endotoxins are confined within the bacterial cell and are pres- ent in both pathogenic and nonpathogenic gram-negative bacteria. Endotoxins derived from the Ehterobacteriaceae are weakly toxic, relatively stable products of bacterial destruction. They are weakly antigenic and cause well characterized clinical manifestations such as sweating, fever, prostration, diarrhea, and hyperglycemia followed by hypoglycemia. There is leukopenia, followed by leukocytosis, vascular malfunction (vasomotor shock, edema, and hemorrhage) and, sometimes, death. These so-called classical responses to endotoxins include altered reSponse to epinephrine and tissue necrosis at sites of endotoxin concentration (Ribi at al., 1964; Smith at al., 1964). There is no evidence that nonlethal levels of bacterial endotoxins have a generalized effect on the integrity of the capillary walls, but Zweifach et a1. (1956) and Zweifach (1958) have demonstrated vasoconstriction at the venular end of the capillary bed of labora— tory mammals following the application of endotoxin extracted from E. coli. "Nontoxic" doses of this endotoxin produced a slowing of blood flow through the microcirculation, but there was no change in the systemic blood pressure during this reaction. With increased 263 levels of endotoxin there was a brief hyper-reactive vascular phase as before, then a hyporeactive phase in which damage to the capillary and venular barrier per se was present. A continuing exposure to endotoxin led to protracted derangement of vascular reactivity, then capillary hemorrhage and stasis (Zweifach, 1964). In this series of experiments, and in previous studies (Christie, 1967), instances of these vascular changes were seen in the laminae prOpriae of the villi as hyperemia, edema and occasionally capillary hemorrhage and stasis. Edema of the intramembranous space and of the interstices of the lamina prOpria was so common in the villi of the monocontaminants that it could most likely be considered as a less severe manifestation of E. coli endotoxic activity on the vas— culature. Dunne (1964) has associated the edema of the mesocolon, seen in baby pigs, with colibacillosis. Christie (1967) considered that edema of the colon could in fact be an indication of the disease, but mild edema of the mesocolon was nonsignificant under experimental conditions in that it was commonly found in germfree neonatal pigs. A rapid increase in number of E. coli organisms in the intestinal tract of a previously germfree pig would be associated with a corre- sponding increase in the amount of endotoxin present. The entero- bacterial somatic 0 antigens are a portion of the endotoxic molecular complex that is a main constituent of the wall of the bacterial cell. The lipopolysaccharide portion of this endotoxic complex exerts endo- toxic as well as 0 antigenic properties. Exposure of a host to sub- lethal amounts of the endotoxin can lead to sensitization and so increase the lethal effect of toxin absorbed from the lumen of the 264 intestine at a later time. Coliforms present in sterilized feed, though nonviable, could still sensitize an animal, germfree or other— wise, so that a subsequent exposure to viable coliforms could well produce an intensified endotoxic reaction (Barnum at al., 1967). This could explain the variation in clinical response observed by Christie (1967) and the two recorded cases of hemorrhagic enteritis that he observed, but it is not likely to be a factor associated with the less overt clinical syndrome of persistent diarrhea seen in the infected animals in this experimental series. The presence of circulating or intestinal-lumen antibody, a complicating factor in the expression of colibacillary diarrhea in "conventional" neonates of animals and man, was circumvented to a large extent in the neonatal gnotobiotic pigs used in these eXperi- ments. Comparable animals used in previous experiments and held under identical environmental and dietary conditions (Christie, 1967) were independently found to be agammaglobulinemic (Brooks, 1966). Again, the short periods of time between birth, monocontamination with E. coli, and necropsy of the test animals in these experiments (Table 2) precludes their significant sensitization by nonviable coliforms which might possibly be introduced with the sterile diet. Endotoxic shock requires, for its full expression, endotoxin, endotoxin antibodies and a protein complex -- "complement". Accord- ing to Barnum at al. (1967), it is the relative prOportions of antibody and of endotoxin in an animal that determines whether the antibody will increase or decrease host resistance. Smooth strains of E. coli contain relatively more endotoxin than rough strains, but 265 it is not the endotoxin per se that determines virulence. The other two significant metabolic byproducts of E. coli metabolism are colicines (Heatley and Florey, 1946) and hemolysins. Although the prOperty of hemolytic activity is a useful marker to a bacteriolo- gist, hemolysis is no longer considered to be an indicator for pathogenicity or virulence (Barnum et al., 1967). It is accepted by some that the pathogenesis of the diarrhea of cholera is primarily that of altered vascular permeability of the lamina prOpria (Elliott at al., 1968). Elliott at al. (1968) have reported experimental infection of dogs with Vibrio sp., causing the canine cholera syndrome. Light microscOpic and electron microscopic studies were made. Biopsy of control dogs and those with the un- treated cholera syndrome led these workers to report minimal acute inflammation of the small intestine and slight capillary dilation in the lamina prOpria. Otherwise, no differences.between the groups were detected. They postulated that the Vibrio organisms probably released toxic material that altered mucosal function but did not cause visible tissue damage even at the ultrastructural level. The Vibrio cholerae toxin, choleragen, induced hyperemia and interstitial edema in the lamina propria of infant rabbit intestine with associated loss of fluid of low protein content through "histo- logically normal" epithelium into the lumen of the intestine (Finkelstein at al., 1964). Gordon at al. (1966) and Phillips (1966) found that the toxin of V. coli caused alteration of vascular and mucosal permeability in the intestine. Electron microscopic changes, probably associated 266 with increased permeability, were seen in the basal capillaries and numerous microvesicles were seen in the cytOplasm of their associated villal epithelial cells. It was hypothesized that, in cholera, increased capillary permeability would allow hydrodynamic movement of fluid from blood to the interstices of the lamina prOpria, through the basal lamina into the mucosa, and thence into the intestinal lumen. From his own studies of the intestinal capillary pressure and filtration coefficients of the human small intestine, and citing the findings of Gordon at al. (1966) relating to fecal serum albumin con- tent of human cholera patients, Fordtran (1967) presented strong arguments against basal capillary permeability's being a major fac- tor in the pathogenesis of cholera diarrhea. Despite this reinterpretation, it is possible that the patho— genetic mechanisms proposed by Gordon at al. (1966) might have some application in the study of neonatal colibacillary diarrhea, at least insofar as it could be a contributory factor to the outflow of fluid into the intestinal lumen. Fluid outflow will be considered briefly later in this discussion. Villal Morphologic Change. In the porcine viral disease transmissible gastroenteritis (TGE), Trapp et al. (1966), Hooper and Haelterman (1966) and Maronpot and Whitehair (1967) described shortening, blunting and at times fusing of the intestinal villi and morphologic alteration of the intestinal epithelium.from simple columnar to cuboidal type. A similar morpho- logic change in the villi has been described for E. coli infection 267 in infants (Handforth and Sorger, 1961), E. coli infection in baby pigs (Barnum at al., 1967), experimental Shigella fiexneri infection in germfree guinea pigs (Sprinz at al., 1961), celiac disease in man, tropical sprue in man, and untreated idiopathic steatorrhea in man (Shiner, 1967). As pointed out by Christie (1962) and Rubin and Dobbins (1965) the more highly Specialized an organ or tissue, the more limited is, it§,range of morphologic response to different noxious factors. This concept was reiterated by Maronpot and Whitehair (1967) and seems to have been borne out by their observations with TGE, and yet it is of interest to note that in no papers reporting histOpathology of coli- bacillosis in gnotobiotic pigs is shortening and blunting of the villi mentioned. It is likely that, at a cellular level, the morphologic response to different noxious factors -- such as the TGE virus, certain entero- toxins of E. coli, the toxic byproducts normal both to the existence of the flora of the intestinal tract and to mucosal necrobiosis -- is the same or similar, although in any given case the actual biochemi- cal site of toxin activity be different. At a tissue or structural level, however, the response may well be markedly different. In colibacillosis as seen in monocontaminated gnotobiotes, the effete villal epithelial cells are retained on the villal stalk for almost the entire period of time normally required for cell movement along the villus. With TGE, the diseased cells are lost from the villus, leaving a denuded lamina propria which retracts to a level close to the surface of the crypts. In the latter case, within a few hours, 268 cuboidal epithelial cells grow out over the stump, and the short blunt villi that result are considered the "lesion" of TGE. Hypo- thetical reasons for this main difference in manifestation of cell damage will be prOposed later in this discussion. It is reasonable to assume that, should colibacillosis be com- plicated by the presence of other elements of the intestinal flora and their metabolic byproducts, by malnutrition, or by any other of a number of factors, these additional factors could contribute to the sloughing of cells (made effete by the action of enterotoxin of E. coli), resulting in premature loss of epithelium and a resulting histologic picture of foreshortened, blunt, stubby villi, reminiscent of those seen in TGE, shigellosis, celiac disease and idiOpathic steatorrhea. So-called "conventional" hosts, that is, those bearing a bacterial flora in addition to E. coli, were described as being characterized by stunted, clubbed and occasionally fused villi (Kenworthy and Allen, 1966). Gnotobiotic pigs contaminated with two serotypes of E. coli in the same eXperimental series had intes- tinal villi with considerable variation in shape, including some branching forms, but no stunting or clubbing of the villi was reported (Kenworthy and Allen, 1966). In fatal cases of epidemic infantile gastroenteritis associated with E. coli Olll, Handforth and Sorger (1961) observed markedly foreshortened villi covered with flattened epithelial cells, and the histologic picture they describe does indeed closely resemble the lesions of TGE in pigs (Trapp at al., 1966; Maronpot and Whitehair, 1967). However, these five human cases of diarrheal disease attributed 269 to enteric colibacillosis (Handforth and Sorger, 1961) were of long duration -- between two and five weeks -- and, according to the authors, presented pathologic changes markedly different from those seen in acute epidemic infantile gastroenteritis. It would be a mis- representation of Handforth and Sorger (1961) to claim on their evidence, without regard to the severe malnutrition and other pre- disposing factors that were present, that the villal atrophy and abnormal thinning of the small-intestinal wall they described was characteristic of "...E. coli infection in infants" as is stated by Maronpot and Whitehair (1967). Vacuolization of‘the Villal Epithelial Cell. Preeminent in the small intestinal mucosa of clinically affected neonates are large intracytoplasmic vacuoles which are also present to a considerable extent in the intestinal mucosa of germfree neona- tal pigs. The determination by histochemical means of the nature of the contents of these vacuoles was relatively unsuccessful. The use of dilute solutions of Alcian blue 868 was introduced by Steedman (1950) as a rapid, easily applied specific stain for mucins. In practice, Alcian blue does not stain the majority of mucoproteins, but will Specifically stain acid mucopolysaccharides of both epithelial and connective tissue mucin (Pearse, 1961). None of the material contained within the vacuoles was Alcian blue-positive, thereby excluding the possibility of its being acid mucopolysaccharide. Similarly excluded by negative response of the vacuole contents to the periodic acid-Schiff reaction were all other types of known 270 epithelial mucins, connective tissue mucins, enterochromaffin (Thompson and Hunt, 1966), plasma and serum, the dystrophic sub- stances observed in the so-called glycogen and lipid storage diseases of man (Lillie, 1951), and the "serous coagula" -- presumably sero- fibrinous material, fibrin, neutral polysaccharides, glycogen and glchprotein (Thompson and Hunt, 1966). The vacuole contents were negative for Mayer's mucicarmine stain, Best's carmine stain, and oil red 0, thereby excluding epi- thelial mucin, glycogen, and fat, respectively. From these results, and from the appearance of the vacuoles on electron microsc0pic examination, it was concluded that the fluid contents of the vacuoles were of a serous nature. The fact that these contents were of low fat, protein, and carbohydrate content (significantly lower than serum levels of these substances) suggests that the material was similar to interstitial fluid. Microbiolggic Aspects Unlike Shigella app. and Vibrio coli, E. coli is considered a normal resident of the intestinal tract of man and animals -- hence its common name "the colon bacillus" -- and its adaptation to a commensal life in the intestinal tract, even of the pathogenic sero- types, will have a marked effect on its pathogenicity and pathogenesis. The ecology of the "normal" intestine is complex and has been the subject of study for many years. One criticism of the use of gnotobiotic animals in research into intestinal disease is that account cannot be made of the effects of interacting microorganisms and their roles in the establishment or maintenance of such disease 271 processes. Herein lie both the advantages and disadvantages of gno- tobiotic animal experimentation. Recent experiments concerned with colibacillosis in neonatal pigs have involved oral inoculation of experimental animals with pre- scribed doses of organisms and immunologic, histOpathologic and electron microscopic studies of the organism in its intestinal habitat. Some variation in results, probably associated with dif- ference in infective serotype, dose of organisms, and experimental animals, has been observed. Staley et a2. (1968d) examined tissues submitted to Gram's stain and saw bacteria in the lumen, attached to the distal half of the villi of both jejunum and ileum. They found bacteria occa- sionally in the "core" of the villi. In the experiments described in this thesis, sections of all small—intestinal tissues sampled were stained with the Brown and Brenn stain for bacteria. Although these were examined carefully, there were no organisms detected in the lumen,or within either the mucosa or lamina propria, or in the intracytOplasmic vacuoles. Staley at al. (1968d) used a very high dose of organisms to infect their experimental animals (100 x 109 organisms) and a fim- briated serotype of E. coli 055:BS:H7, which may account for this difference. According to Barnum et a1. (1967), in order to reproduce enteric colibacillosis it is necessary that a susceptible host be "infected" with large numbers of an enteropathogenic strain of E. coli. The ambiguity of this statement is illustrated by Saunders et al. (1963), 272 who observed diarrhea in a group of colostrum—deprived pigs after feeding as few as 1000 living E. coli organisms. It is obvious that with 100 x 109 organisms present in the intestine, as was used by Staley et a2. (1968d), there will be many organisms found in the tissues that will represent those experimentally administered and not necessarily organisms that have reproduced and colonized within the intestine. It was for this reason that an infective dose of approximately 3 x 106 organisms was used in each of the eXperiments reported here. It was hOped to introduce sufficient young organisms that an infec- tion would be established but insufficient that the disease syndrome would be caused by the organisms actually introduced. With a suf- ficiently low dose, it was considered more probable that the disease would be associated with multiplication and growth of the organisms within the host. The presence of fimbriae on E. coli 055:B5,H7 might be a major factor in the visualization of the organism with Gram's stain in infected tissues as described by Staley et a2. (1968d). The stain- ing of fimbriae with the Gram's counterstain would assist somewhat the visualization of fimbriated serotypes compared to nonfimbriated serotypes such as E. coli 0138:K81:NM and might eXplain in this experi- ment -- and other experiments with this serotype (Christie, 1967) -- this consistent inability to see organisms in tissues stained either with hematoxylin and eosin or the Brown and Brenn stain for bacteria. Staley at al. (1968d) examined sections of duodenum, jejunum and ileum with the electron microscope 2, 6 and 20 hours after dosing 273 hysterotomy-derived neonatal pigs with 100 x 109 E. coli 055:B5,H7 organisms. They saw neither intracellular organisms nor cellular damage in any of the duodenal or jejunal tissues, but they did detect bacterial invasion in the villal epithelial cells of the 20— hour-exposed ileum. The organisms were attached to the cell sur- face, and attachment sites were characterized by microulcers and increased density of the terminal web immediately beneath the attachment site. They observed that E. coli 055:B5,H7 entered the ileal villal epithelial cell by invagination of the plasma membrane where the terminal web was least dense. Once within the cell, the invaginated plasma membrane enveloped the bacteria in a membrane lined vacuole (Staley et al., 1968d). Although comparable tissues (jejunum and ileum) were examined during these experiments, at no stage were organisms seen either free or membrane-enclosed within the villal epithelial cells. There were many instances (for example Figures 96, 97, 98 and 99) where E. coli 0138:K81:NM was in close association with the brush border. Although it was apparent that, in the vicinity of the organisms, the micro— villi were foreshortened and somewhat atrophied, there was clear evi- dence that the trilaminar structure of the apical plasmalemma and the apical tubules nearby was retained (Figure 99). There was no evidence to suggest that microerosion had occurred, or was about to occur, nor was there any evidence of invagination of the apical plasmalemma or engulfment of the organism by the apical cytOplasm. One could explain these differences by taking into account the dif- ferent chemical and physical natures of the envelOping coat of the bacterial serotypes used in these various experiments. 274 In a normal intestinal tract there are occasional disruptions in the integrity of the mucosal surface (microulcers) where an organism, although ill-equipped for epithelial invasion, can enter subepithelial tissues and enter the circulation via capillaries or lymphatic chan- nels. If this site, for example a tonsillar crypt, be anterior to the stomach, then this occasional contamination could effectively enable an organism such as E. coli to bypass the threat of destruc— tion in the acid contents of the stomach and re-enter the gastroin- testinal tract in the region of the jejunum or ileum where environ- mental circumstances are more favorable. This possible mechanism for infection has been established for E. coli (Fey at al., 1962) and for E. coli 0138:K81:NM in gnotobiotic pigs (Christie, 1967; Christie, unpublished data). There is no doubt that E. coli 0138:K8lzNM has the ability to enter the epithelial barrier and produce a bacteremia very early in an infection, but it may be that in most circumstances, with this organism at least, this phenomenon is entirely irrelevant to the establishment of a colibacillary diarrhea, provided sufficient num- bers of the organism can enter the intestinal tract. "Positive" results have been obtained in intestinal-loop studies both with viable organisms -- including E. coli Ol38:K81:NM -- and sterile preparations of enterotoxin from pathogenic serotypes of E. coli (Barnum, personal communication). The development of diar- rhea, associated with typical microscOpic lesions, in gnotobiotic neonatal pigs orally dosed with sterile samples of E. coli entero- toxin from pathogenic serotypes has also been observed (Christie, unpublished data). 275 With these results in mind, it is clear that an essential factor in the production of colibacillary diarrhea is not the ability of the organism to enter the villal epithelial cell or even invade the lamina propria (although these capabilities may directly or indirectly assist the organism in the colonization of the intestine) but to papulate the lumen, the intervillal spaces and perhaps the crypts, thereby maintaining a level of enterotoxin in the chyme bathing, and being absorbed by, the mucosal surface. Not only does the apical tubule system absorb colostrum, but probably also products of coliform colonization and metabolism, specifically enterotoxin. There was no evidence that the infectious agent itself was absorbed. It is no longer sufficient to say that E. coli is "readily absorbed by neonatal intestinal epithelium" (Staley, 1969) in that it is clear that some serotypes of E. coli are readily invasive of the epithelial mucosa whereas,with other serotypes, such as E. coli Ol38:K81:NM,this evidence was not found despite intensive search. Quite apart from the above considerations, it is likely that the ingestion of E. coli in toto by the villal epithelial cell is not an effective means of reducing infection nor of removing bac- teria. In view of the rapid turnover rate of the villal epithelial cells of the small intestine, the intracytoplasmic sojourn of E. coli is likely to be short, soon to be ended by the necrobiosis of the villal epithelial cell and the early release of the microorganism into the lumen. w.-- --..” 276 Alternatively, it may be that the encapsulated bacteria are brought into association with a lysosome to form a compound lysosome (deDuve, personal communication), and in this way the bacteria is perhaps lysed and digested, although Staley et a2. (1968d) did not see evidence of lysosomal destruction of the encapsulated organisms. As in these experiments, so also Staley at al. (1968d) did not observe in their electron microscOpic evidence of bacteria in the lamina prepria of the villus; neither were they observed in the lateral intercellular spaces, nor was there any conclusive evidence of intracellular digestion of the enveloped bacteria. In the case of an organism exerting its toxic effect by the production of enterotoxin, factors which assist in the maintenance of the organisms or growth of colonies, which maintain a suitable level of contact between the toxic material and its site of action, and which inhibit the production or distribution of antibiotic, antitoxic, and protective humoral or cellular factors will increase the pathogenicity of the organism. Losos (1964) speculated that the fimbriae of E. coli acted as a mechanism for attaching the microorganism to the epithelium of the small intestine. Smith and Halls (1967) considered that E. coli which cling to the mucosal surface, or those that papulate the crypts, would be more pathogenic than those that do not, in that they would resist the "washing out" effect of the chyme as it moved through the intestinal lumen. The absence of fimbriae from E. coli 0138:K81:NM might be a major factor in the apparent inability of this microorganism to 277 erode and enter the brush border of a normal epithelial cell such that it would be in a position favorable for pinocytosis by the apical plasmalemma. In these experiments, E. coli Ol38:K81:NM, although not fim- briated, did actively migrate within the lumen to the base of the villi, and possibly even further (although sections of the crypts were not examined under the electron microscope). From a study of the structure of the intestinal wall it is not difficult to imagine that, even in instances of diarrhea, particulate matter at the center of the lumen would be swept along rapidly, whereas particu- late matter near the base of the villi would move relatively more slowly, and possibly at no more rapid pace than in clinically normal hosts. It is likely that the rate of movement of particulate mat- ter from the bases of the villi would depend more on the compression and relaxation of individual villi on neighboring villi during "pumping" movements than on the axial flow of chyme associated with peristalsis. If the degree of vacuolization of the villal epithelial cells is marked, and the level of endotoxin present in the intestine is such that the Breucke fibers are relaxed (or paralyzed), yet the muscular coats are stimulated to hyperactivity (by endotoxic sensi- tization of smooth muscle to exogenous and endogenous stimuli) then, theoretically, conditions exist that would allow a rapid axial flow of chyme yet permit a relatively stable particulate papulation to exist at the level of the crypt Openings, trapped there by the villi themselves. The histologic picture seen many times during this series of experiments gave support to this speculation. _& 278 If, in addition to these hypothetical mechanical factors, this particular serotype had the ability to cling to the mucosal surface by either chemical or physical means, then as Smith and Halls (1967) suggest, pathogenicity would probably be increased. Although it has been demonstrated that serotypes of E. coli can enter the circulation of a host animal, whuh then undergoes intermittent bacteremic episodes (Christie, 1967), and that certain serotypes can regularly invade and establish infections in extra- enteric sites such as the pleural and peritoneal cavities (Britt, 1967), the most common manifestation of colibacillosis in neonates is its enteric form. It is more likely that E. coli 0138:K81:NM, and perhaps other serotypes, exert their pathologic effects pri- marily on the gastrointestinal tract because that environment is more favorable in terms of multiplication and survival and because the organism itself is basically noninvasive and tends to stay where it has successfully colonized. Indeed, there is no ecological need for E. coli 0138:K81:NM to invade the mucosa or lamina prOpria, and such invasion could be just as hazardous a process for the micro— organism as it would be for the host. Unlike enteric serotypes of E. coli, the pantrOpic serotype 083:K.:NM (Britt, 1967) is invasive for gnotobiotic pigs, rapidly colonizing serous surfaces and producing toxins which are retained and can only be absorbed by their environment. It is understandable that this infection kills rapidly, and in so doing limits the micro- organismis own continued existence and its potential for infecting other susceptible hosts. 279 Those more commonly known pathogens such as E. coli with the 0138 and the K81 antigens are less invasive, predominantly enteric in habitat,.and except for neonatal pigs (or older pigs in adverse environmental or physical circumstances) they produce a high mor- bitidy but low mortality, thereby promoting their own widespread distribution among susceptible hosts and their continued success- ful existence. Ligated Intestinal-loop Preparations Barnum et a2. (1967) described experiments with ligated intestinal loOps of susceptible pigs inoculated with E. cbZi in which they demon- strated that the net outflow of water into the "positive" loops took place through the intact epithelial membrane "as it apparently does in enteric colibacillosis". In ligated intestinal-100p preparations in 21-day-old conventional pigs it has been found that, in common with a number of other sero- types known to be pathogenic for swine, the organism used in these experiments is "strongly reactive" (D. A. Barnum, personal communica- tion). A "strongly reactive" or "positive" ligated intestinal-100p preparation is one in which ligated pig intestine has been injected with the enterOpathogen 12 to 14 hours previous and has become mar- kedly distended with fluid. "Positive loOps are a clear demonstra- tion of the net flow of water into the intestinal lumen which occurs in response to certain enteropathogenic bacteria" (Barnum at al., 1967). 280 This biological preparation, although made in the host animal, does constitute an artificial and abnormal situation. The intraluminal fluid is retained, distending and flattening the intestinal wall be- yond the limits found in pigs with clinical diarrhea and submits the vessels and mucosae to extraordinary pressures. Despite these limitations, it has been described as a pathogenicity test by a number of authors (Nielsen, 1953; Moon, 1965; Gyles and Barnum, 1965). Any lesion of the small intestine which interferes with orderly intestinal peristalsis and causes stasis of intestinal contents eventually results in intraluminal bacterial overgrowth (Donaldson, 1967). This fact confirms the importance of normal peristalsis in limiting bacterial growth in the small intestine. Those lesions associated with stasis and massive proliferation of enteric micro- organisms include surgical blind 100ps (Doig and Girdwood, 1960) and strictures (Seyderhelmen at al., 1924; Halsted at al., 1956). The real significance of a positive ligated intestinal-100p preparation is that the ligated segment of intestine distends with- out their being any significant amount of chyme in the lumen of the intestine prior to inoculation with the test organism. The fluid collecting in the 100p must therefore be derived from the circula- tion, and its production cannot be associated with the main patho- genetic theories of diarrhea such as inhibition of active sodium transport, impairment of chloride and bicarbonate absorption, hypo- kalemia, osmotic diarrhea and deranged intestinal motility. The intestinal-loop preparation does not, at present, contribute signi- ficantly to our understanding of the pathogenesis of neonatal colibacillary diarrhea. 281 Histochemistry The term "enzyme activity" has been used in this thesis to describe the observed tissue staining associated with a reaction, even though histochemists generally agree that this staining process might not represent the true functional capabilities of the particu- lar enzyme. The enzyme activity was qualitatively assessed indi- vidually and comparatively within litters and within tissue groups preserved the same way prior to processing. Melnyk at al. (1966) agree that qualitative tissue and cellular localization is the most reliable parameter for evaluating enzyme distribution and activity. Assuming proper technique and apprOpriate tissue and enzyme controls, decreased intensity of histochemical staining of enzyme catalyzed reaction products is an indication of the deficiency of the enzyme in a particular tissue or cell. "Alteration of the brush border and decrease of specific functional enzymes in intestinal epithelial cells bears a significant relation- ship to nutrient absorption". (Maronpot and Whitehair, 1967). It has been argued that enzymic complement is reduced where villal epithelial cell turnover rate is increased in that the cell is still acquiring its full enzymic load while it occupies the nor- mal position of an adult cell on the surface of the villus (Creamer, 1967). Although from these experiments and a previous report (Christie, 1967), it is clear that in colibacillosis due to E. coli 0138:K81:NM there is considerable damage to individual villal epithelial cells, there is no clear histopathologic evidence of a marked in- crease or decrease in cell turnover rate. 282 Creamer (1964b), with his own findings and those of Padykula at al. (1961), presented convincing evidence that, within a species, villal shape and length of crypt to villus ratio are relative indi- cations of cell turnover rate. Villal length and shape in tissues from monocontaminated pigs were not noticeably altered from those of the germfree control pigs in these experiments. It is apparent that any difference of turnover rate between the groups in these experiments was not marked. One can reasonably conclude that the alteration in intracellular enzyme concentration, activity, and dis- tribution in these experimental tissues is more likely to be due to a direct effect of the microorganism, or its enterotoxin, than secondarily through cell destruction and replacement with biochemi- cally and functionally immature epithelial cells. Apart from the functional and biochemical deductions to be made from histochemical studies, it was Suggested by Pearse and Riecken (1967), but not yet established, that the demonstration of enzyme groups bound to membranes would provide an indicator of the struc- tural integrity of the particular membranes to which they are bound. These experiments constituted an opportunity to study cells histochemically, predict functional and morphologic change, and then confirm or refute such predictions by electron microscopic examina- tion of comparable cells. Alkaline Phosphatase Biochemical studies of the brush border separated by microdis- section have revealed that alkaline phosphatase (as well as leucine aminOpeptidase and the disaccharidases) are associated with the 283 microvilli (Eichholz and Crane, 1965; Crane, 1966). Earlier, Miller and Crane (1961) found that the brush border contained all of the cellular activity of the terminal digestive enzymes including the alkaline phosphatases, and yet in a more recent publication (Maronpot and Whitehair, 1967) alkaline phosphatase activity was depicted as being characteristically present in the brush border and concentrated in the apical cytoplasm of villal epithelial cells in 3- to 5-week—old conventional pigs. In the germfree neonatal pig, alkaline phosphatase activity was restricted to the epithelial cells of the villi and, by the techniques used, was not demonstrated in the epithelial cells of the crypts. Within the individual cell, there was a high concentration in the brush border. At times there was a narrow reaction-free zone at the apex of the cell that would correspond to the terminal web and the apical tubule region. The rest of the alkaline phosphatase activity was concentrated in the apical third of the cell and in a narrow zone close to the basement membrane. Since the latter two sites for enzyme activity were not demonstrable in the tissues preserved in cold formalin, it is apparent that such staining was associated with nonspecific phosphatases and should be discounted as indicative of alkaline phosphatase activity. Localization of alkaline phOSphatase at the brush border of the villal epithelial cell in neonatal germfree pigs corresponds with localization of this enzyme at the apical border of mucosal cells in conventional baby pigs (Maronpot and Whitehair, 1967; Thake, 1968) and in other species (Shnitka, 1960; Padykula at al., 1961; Miller and Crane, 1961; Eichholz and Crane, 1965; Crane, 1966; Ito, 1969). 284 It is apparent that the variation in amount of nonspecific phos— phatase activity seen in the stained tissues in this experiment was due to the differing preservation procedures. The most satisfactory preparations were those preserved in cold formalin prior to staining, the method advocated by Pearse (1961). The very specific localization of the enzyme with a minimum of enzyme diffusion artifact and the excellent histologic detail was in sharp contrast with the lack of Specificity and histologic detail of tissues fixed in cold acetone or of frozen unfixed tissues. It was possible to increase the specific localization of alkaline phosphatase in frozen unfixed tissues by reducing the substrate incubation time markedly, but the result was less satisfactory than that obtained with formalin-fixed tissue. In the diagrammatic representations of enzyme distributions in the jejunum and ileum of their control group pigs, Maronpot and Whitehair (1967) have depicted alkaline phosphatase (and leucine aminOpeptidase) in both the brush border and the apical third of the villal epithelial cells. With improved preservation methods and localization techniques used in the processing of tissues from germ— free control pigs in these epxeriments, it was observed that these enzymes were in fact localized in the region of the brush border. These results are in agreement with those of Barrnett (1959), Shnitka (1960), Padykula et a2. (1961), Eichholz and Crane (1965), Crane (1966), Pearse and Riecken (1967) and Ito (1969). It is likely that the results of Maronpot and Whitehair (1967) were in part due to nonspecific phosphatase activity in the apical cytOplasm and partly to the diffusion of alkaline phosphatase (and leucine aminopeptidase) 285 from the brush border into the apical cytoplasm during cold preserva- tion and tissue processing. The glycocalyx is a hirsute layer comprising the filamentous structure closely associated with the unit membrane of the microvilli (Ito, 1965). Ito suggested that the glycocalyx had a protective function and Johnson (1966) described the localization of enzymatic activity in this structure. Staley et al. (1968b) suggested that the early appearance (by 48 hours) of the glycocalyx might reflect the precocious develOpment of alkaline phosphatase activity in the pig. Some confirmation of this hypothesis is borne out in alkaline phospha- tase studies in this experiment. In monocontaminated pigs, there was a loss of alkaline phospha- tase activity Starting 12 to 14 hours after eXposure, and this loss was more apparent in the epithelium more distally placed along the villi. The degree of staining was also less intense in sections taken further along the small intestine. By 93 hours after exposure there was renewed intensity of stain- ing for alkaline phosphatase activity throughout the small intestine. The loss of alkaline phosphatase activity associated with enteric disease and diarrhea in this experiment agrees with the find- ings of Maronpot and Whitehair (1967) and Thake (1968), who observed loss of alkaline phosphatase activity from the villal epithelial cells of baby pigs clinically affected with TGE. It is clear that, whether primarily or secondarily, alkaline phosphatase activity of the brush border of villal epithelial cells is reduced or lost for a period during enteric colibacillosis and 286 that such loss will impede terminal digestion and the process of absorption in the affected cell. Whether loss of alkaline phospha- tase activity is a Sign of a direct toxic effect of E. coli entero- toxin on the enzyme or its precursors, or whether it is a Sign of ultrastructural disorientation of the glycocalyx or brush border, will be discussed later in this thesis. It can be forecast, however, that the enterotoxin of E. coli 0138:K81:NM exerts an effect at the level of the brush border, which is significant in terms of the processes of digestion and absorption. Acid PhosPhatase As described in man (Fric and Lojda, 1964; Riecken at al., 1966), the rat and the mouse (Pearse and Riecken, 1967) and baby pigs (Maronpot and Whitehair, 1967; Thake, 1968), acid phosphatase was seen to be distributed throughout the cytOplasm of the intestinal absorptive cells of germfree neonatal pigs, particularly in the apical cytoplasm. This probably corresponds to the distribution of lysosomes within the cytOplasm of these cells. In germfree neonatal pigs the clear perinuclear zone seen in villal epithelial cells stained for acid phosphatase corresponds to the zone in the normal cell that is occupied by the Golgi complex and the end0p1asmic reticulum. Lysosomes are generally in the periphery of the cell, and the normal cell, when stained for acid phosphatase activity, would be expected to have a clear perinuclear zone of minimal acid phosphatase activity. Twelve or more hours after monocontamination with E. coli 0138:K8lzNM, there was histochemical evidence of a decrease in acid 287 phosphatase activity, and there was a replacement of the granular appearing Stain reaction with a smoother, more homogeneous staining of-the cytOplasm of villal epithelial cells. It is speculated that these changes are indicative of reduced lysosomal activity, possibly associated with disruption of lysosomes, either by direct action of enterotoxin or by indirect action of the enterotoxin on the meta- bolic activities or on the membrane-transport activities of the lysosomes.. It would be expected that disruption of a number of lysosomes within a particular cell would result in loss of Specific site location of the stain reaction products and release of formerly confined lytic enzymes (including acid phosphatase) so that the stain reaction was seen to be homogeneous throughout the cell. AS the cell contents were lysed by the released lytic enzymes (includ- ing acid phosphatase), general loss of intensity of the staining reaction would be experienced. Considering these speculations and the results obtained for acid phosphatase activity in this experiment, it could be reasonably anticipated that 12 to 14 hours after exposure of gnotobiotic neo- natal pigs to E. coli 0138:K81:NM, there would be disruption of lysosomes and beginning cytOplasmolysis of villal epithelial cells. subsequent dissolution of the cell, probably associated with rupture of the apical plasmalemma and discharge of the fluid contents of the cell into the intestinal lumen, could be expected. 288 Adenosine Triphosphatase As described in man (Pearse and Riecken, 1967) and in conventional baby pigs (Maronpot and Whitehair, 1967; Thake, 1968), adenosine triphosphatase was seen to be distributed throughout the cytOplasm but particularly localized in the apical cytoplasm and the brush border of the villal epithelial cells of germfree neonatal pigs. Thake (1968) considered that the slightly more intense reaction for adenosine triphosphatase activity seen in the apical cytoplasm may have been due to alkaline phosphatase since the reactions for demon- stration of both enzymes are carried out at pH 9.4. The most distinctive change in adenosine triphOSphatase distri— bution that was associated with colibacillosis was the partial loss of enzyme activity from the brush border of cells, particularly from the jejunum and ileum, within 12 hours of monocontamination with the microorganism. The second detectable change in the infected pigs was that of loss of intensity of the Staining for adenosine triphOSphatase activity within the cytOplasm of the villal epithelial cells. It is only in erythrocytes that a correlation has been shown between the rates of adenosine triphosphate hydrolysis and rates of membrane; transport of fluids (Whittam and Ager, 1965). Since adenosine tri- phOSphatase plays many roles in cell metabolism, apart from fluid tranSport, it is not yet possible to Specify those stages of the absorptive process blocked by loss or destruction of adenosine tri- phOSphatase in neonatal pigs with colibacillary diarrhea. 289 The losses of adenosine triphosphatase described above do, how- ever, constitute events of Significance because of the very nature of the role of this enzyme in biochemistry. Loss of adenosine triphosphatase from the brush border presents a significant block to the first stage of membrane transport at the apical plasmalemma, in that the source of energy for such a process is dependent on the proper function of adenosine triphosphatase. Within the mitochondrion, adenosine triphosphatase is an integral part of the essential functions of this organelle -- the oxidation of carbohydrates, fats and amino acids, the formation of high energy yielding adenosine triphosphate, active intracellular tranSport of ions, and intramitochondrial water metabolism (Lehninger, 1964). Disruption and loss of functional adenosine triphosphatase would most likely result first‘in the inability of the mitochondrion to excrete metabolic water. Collection of fluid within the mitochon- drion would be seen first as vacuolization of the mitochondrial matrix, then as distention of the mitochondrion to the point of eventual rupture. The disruption of mitochondria would be a terminal lesion in a cell of such diverse metabolic activity as the villal epithelial cell, and this lesion alone would be sufficient to explain all other changes observed within the intestinal cells of neonatal gnotobiotic pigs monocontaminated with E. coli 0138:K81:NM. Leucine AminOpeptidase As described in the columnar cells of the rat, human, and monkey small intestine (Nachlas at al., 1957, 1960) and within the brush 290 border and supranuclear cytOplasm of villal epithelial cells of 3- to S-week-old conventional pigs (Maronpot and Whitehair, 1967), leu- cine aminopeptidase was also found in the villal epithelial cells of neonatal germfree pigs. In these animals, as in man (Pearse and Riecken, 1967), the enzyme was apparently localized within the brush border, with only slight traces of enzyme activity in other sites. The enzyme was not concentrated in the supranuclear cytOplasm to any significant extent. The loss of leucine aminopeptidase activity from the brush border of villal epithelial cells of the jejunum and ileum 16 to 18 hours after monocontamination with the test serotype of E. coli was distinctive. Since leucine aminOpeptidase is a ”membrane associated" enzyme (Pearse and Ricken, 1967), this loss of cytochemical activity, when considered with the loss of alkaline phosphatase activity, pre— sents convincing evidence of a biochemical lesion in the brush border. It is reasonable to forecast that such a lesion might be asso- ciated with morphologic change in the microvilli and glycocalyx of villal epithelial cells from neonatal pigs with colibacillary diarrhea. Succinate Dehydroggnase As described in the mucosa throughout the small intestine of rats and dogs (Nachlas at al., 1957), man (Pearse and Riecken, 1967), 3- to Seweek-old conventional pigs (Maronpot and Whitehair, 1967) and l- to 2dweek-old conventional pigs (Thake, 1968), succinate dehy- drogenase activity was also observed in the cytOplasm of mucosal cells of both villi and crypts, at all levels of the small intestine, of neonatal germfree P185° 291 The distribution of the enzyme corresponded closely with the intracytoplasmic distribution of mitochondria, as seen in electron micrographs of duplicate tissues from these eXperimental animals. Considering the role of succinate dehydrogenase in mitochondrial metabolism, such distrubtion was to be expected. I The reaction for succinate dehydrogenase results in the intra- mitochondrial deposition of formazan as small dots with an average diameter of 0.2 to 0.3 micron (Scarpelli and Pearse, 1958). Pearse (1958, 1961) has emphasized the sensitivity of methods for, and the accuracy of localization of, succinate dehydrogenase, in which "it is probable that the amount of formazan deposited is directly related to enzyme acti- vity. We can now see‘ and record small altera- tions...in single cells and we can thus detect changes of cellular function such as that from normal growth to neOplasia". (Pearse, 1961). The particular value of the demonstration of succinate dehydro- genase in these experimental animals lies in the specificity of the stain reaction and its increased reaction velocity in swollen but metabolically active mitochondria. Succinate is oxidized much faster by swollen mitochondria (Raaflaub, 1953) and this observation was confirmed histochemically by Scarpelli and Pearse (1958), who found that the velocity of the succinate dehydrogenase reaction, but not of the DPNH-diaphorase reaction, was markedly increased in swollen mitochondria. The increased intensity and definition of the stain reaction within a few hours of monocontamination that was observed in some instances in these experiments (Figure 58) was probably due to 292 swelling of the mitochondria as a preliminary step in the eventual distention and rupture of these organelles. Within 12 hours of monocontamination, there was marked loss of succinate dehydrogenase activity from the mucosal cells, and this loss was more marked in more caudal Sites of the jejunum and ileum and was observed in tissues from the pigs with all clinical Signs of acute colibacillary diarrhea. The loss of enzyme activity was still apparent 32 hours after monocontamination. It is considered that this knowledge of the succinate dehydro- genase activity and distribution, when considered with that of adenosine triphosphatase and lactate dehydrogenase activity, pro- vides reasonable evidence of mitochondrial dysfunction associated with colibacillosis. Lactate Dehydrogenase As described in the mucosal epithelium of rats (Nachlas et al., 1958; Hess at al., 1958), lactate dehydrogenase was observed to be present in both the apical and basal cytoplasm of intestinal absorp- tive cells at all levels of the small intestine of the neonatal germfree pigs used in these experiments. The most marked enzyme activity was observed in tissues from the more anterior Sites, duodenum and anterior jejunum, and this activity might be related to the higher energy requirements of the more intense digestive-absorptive activity claimed for the anterior small intestine by some authors (Nielsen at al., 1968). 293 The binding of lactate dehydrogenase with mitochondrial mem- branes, and with those of the endoplasmic reticulum, provides an explanation for the described distribution of this enzyme in the intestinal absorptive cells of the neonatal germfree pig. The loss of lactate dehydrogenase activity observed to be associated with monocontamination of the experimental animals with E. coli 0138:K81:NM was much less marked than the observed loss of activity of the other enzymes (Dido supra), although such a compari— son is quite invalid in terms of assessment of function or meta- bolic activity. The distribution of lactate dehydrogenase in both mitochondria and end0p1asmic reticulum would explain the rather more persistent staining for activity of this enzyme if it was that either endoplasmic reticulum or mitochondria persisted relatively unaltered for a longer period of time in a cell eXperiencing entero— toxic stress. Twenty-eight hours after monocontamination, by which time diar— rhea is very well established in gnotobiotic neonatal pigs, there is a distinct loss of lactate dehydrogenase activity, which is indica- tive of dysfunction or disruption of either the mitochondria or endo— plasmic reticulum, or both. Lactate dehydrogenase is very widely distributed in a high pro- portion of the mitochondria of most tissues (Pearse, 1961), and it is tempting to speculate that the subjective assessment of increased staining for lactate dehydrogenase of the laminae propriae of villi from infected pigs is a Sign of the release of enzyme following dis- ruption of its organellar relationships and its escape with tissue 294 fluid into the laminae prOpriae. If this were the case, however, one would expect that one of the major sources of serum lactate dehydro- genase would be that amount of enzyme released from the continuing extensive necrobiosis of alimentary tract mucosa, known to be part of normal intestinal activity. Again, in diseases involving wide- Spread damage to the intestinal mucosa, such as TGE or colibacillary enteritis, one would expect to see elevated levels of serum lactate dehydrogenase. Such events have not been reported. Considering the lesions of colibacillosis already described in these experimental animals, it is more likely that enzymes released from organellar membranes are lost, first to the cytOplasm generally, and then through the ruptured apical plasmalemmae to the lumen of the intestine. A second consideration is that the enzyme itself is a large protein molecule and would not be absorbed, even by intact functional villal epithelial cells, without considerable proteolysis, which would in turn most likely destroy its enzymic function and render it undetectable by histochemical means. Function of intramitochondrial enzymes is most likely to be dependent on Specific spatial arrangement, in sequence, with other enzymes associated with a particular metabolic cycle (Lehninger, 1964), and disruption of this lattice would certainly remove or mar- kedly diminish their enzymic activity. The Spatial arrangement of the end0p1asmic reticulum in parallel arrays, separated by a remarkably uniform distance which varies according to cell type and function (Fawcett, 1966), is strongly suggestive of a similar functional structure. Presumably, distortion of the parallel array of these 295 membranes, or chemical breakdown of the membrane components, would effectively block the function of lactate dehydrogenase in cell metabolism, and the demonstration of lactate dehydrogenase activity in histochemical tests. In summary it is probable that the staining for lactate dehy— drogenase activity seen in the laminae prOpriae of the villi of both germfree and gnotobiotic neonatal pigs is associated with the activity of mitochondria and end0p1asmic reticulum of cells constituting the lamina propria, and not the consequence of enteric colibacillosis. Electron Microscopy The role of the intracellular membrane as a means of delineat- ing and defining cytoplasmic Spaces is apparent but of relatively little importance. It is no longer considered to be simply a bimo— lecular layer of phospholipid with nonlipid (usually protein) mole- cules bound to its polar surface by ionic bonds, thereby acting as an inert barrier to free diffusion of specific substances, with a degree of relative permissiveness to lipid—soluble solutes. It is increasingly clear that membranes are "specific arrangements of catalytic proteins" (Korn, 1969), and with this concept in mind, much of the behavior of neonatal porcine intestinal cells in absorption and malabsorption associated with experimental colibacillosis be- comes explicable. The Brush Border "With increasing knowledge regarding the spa- tial order and relationships within the brush border and the microvillus, it seems probable that close anatomical relationships within 296 this structure will carry with them impor— tant significance for the digestive and absorptive prOperties of this organelle". (Eichholz, 1969). Crane (1969) has stated that it was only in the last 8 years that the brush border was recognized to be a digestive— absorptive surface organelle, less complex perhaps than the mito— chondria, but a structurally integrated, subcellular organelle con- trolling and interacting with the internal environment of the villal epithelial cell as well as forming its luminal surface. Enzymes catalyzing digestive and absorptive functions seem to be spatially arranged in the brush border so as to offer a kinetic advantage for absorption (Crane, 1968; Crane, 1969) not only of digested foodstuffs, but of sodium and possibly other ions, by a mechanism of c00perative interaction between these and other absorbed molecules in the mobile carrier transport system (Crane, R.-K.,1' personal cemmunication). The Glycocalyx. Ito (1969) has summarized the role of the glycocalyx as being reaponsible for, or contributing to, "absorption, transport, electri- cal characteristics, immunological behavior, and adhesive nature" of the apical plasmalemma. The glycocalyx is most probably struc— turally layered, with a surface component of intrinsic hydrolases closely applied to the trilaminar membrane (Crane, 1969) and required primarily for the terminal digestion of carbohydrates and proteins. Ugolev (1965) has described this region of the plasma membrane as the most important site for terminal digestion. Adenosine 297 triphosphatase and alkaline phosphatase have been shown by many to be localized in or near the brush border, and this localization has been confirmed for neonatal germfree porcine villal epithelial cells in these experiments. Staley at al. (1968a) stated that the glycocalyx was not pres— ent in the neonatal pig villal epithelial cells and suggested that it was synthesized after whole protein absorption ceased. Yet in the present experiments, there was ample evidence of a glycocalyx, albeit meagre yet clearly present, in many animals from 3 days of age and older. It is likely that the staining methods used by Staley et al. (1968a) were less satisfactory for the demonstration of the glycocalyx than was the osmium tetroxide fixation and Reynold's lead citrate-uranyl acetate staining used for the experimental tissues in these experiments. Other evidence is also provided in these experiments for the presence of the glycocalyx in neonatal pigs. The strongly positive staining of the brush border of all control pig tissues with Best's carmine Stain and the strongly posi- tive periodic acid-Schiff reaction were indicative of the presence of the glycocalyx in pigs within, at most, 48 hours of surgical delivery. Further evidence for the presence of a functional glyco- calyx in germfree neonatal pigs was the demonstration of the presence of alkaline phosphatase and leucine aminopeptidase in the brush border, and the partial loss, or absence, of the activity of these enzymes from the brush border of the monocontaminated pigs. Ito (1969) has demonstrated, in the cat intestinal absorptive cell, that alkaline phosphatase activity is localized in the deeper 298 layers of the glycocalyx, immediately adjacent to the plasma membrane, and on the external and internal surfaces of the trilaminar membrane. Our present concepts of the biochemistry of carbohydrate digestion and absorption require an arrangement such as this for the success— ful passage of a molecule such as glucose through the trilaminar membrane. The exact chemical and structural nature of the mobile carrier transport system within the lipoidal membrane is not yet known, but there is strong evidence to suggest that the common car- rier for both sodium, and a limited number of specific carbohydrates (such as glucose and galactose), is most likely protein, obeys the kinetic laws characteristic of enzyme binding, and may be a "tranSport- enzyme" (Crane, R. K., personal communication). If such "transport enzymes" were in fact proteins, and integral parts of the membrane fabric with its tranSport "acceptor Sites" exposed to both inner and outer surfaces of the membrane, the only requirement necessary for movement of an "accepted" solute molecule or molecules from one side of the membrane to the other would be an-allosteric.transformation of the carrier molecule (Jardetsky, 1966). Sodium-potasSiumdmagnesium adenosine triphosphatase probably transports cations by this process (Skou, 1965). Ito (1969) found that the glycocalyx of the cat intestinal ab- sorptive cell is more labile when the underlying cell is moribund, and he concluded that this component might be a dynamic surface component "requiring an intact cell for its synthesis and mainte- nance". His autoradiographic investigations have clearly shown that the glycocalyx is continuouSly synthesized by the villal epithelial 299 cell and moved to the surface, the whole process in the cat requiring 30 to 60 minutes, their being a constant replacement of the surface coat with a minimum turnover time of about 4 hours and an upper limit of 8 to 10 hours. In the experiments desctlbed in this thesis, it has been shown histochemically and microsc0pically that there is some loss, both in form and function, of the glycocalyx of villal epithelial cells from neonatal gnotobiotic pigs monocontaminated with E. coli 0138:K81:NM. Whether this loss is a primary effect of the enterotoxin or secondary to enterotoxic damage of the intracytoplasmic structures responsible for the production of the glycocalyx, is equivocal. The early concentration of autoradiographic components of the glycocalyx in the Golgi complex, and their subsequent appearance in the glycocalyx, considered together with the accepted role of the Golgi complex as a major intracellular site for synthesis of cell products, is presumptive evidence of the role of the Golgi complex in the continuous formation of the glycocalyx. Assuming that the glycocalyx of the neonatal pig intestine is derived in a similar man- ner to that of the cat, it is clear that the internal cellular dis- ruption associated with monocontamination.with E. coli 0138:K81:NM would immediately lead to cessation of synthesis and transport of the materials required for the continuous replenishment of the glycocalyx. Although Ito (1969) has Speculated that the glycocalyx might be 300 "1.:characteristic for each Species and for ‘each cell type, and furthermore that it may vary in its structure and function in the same cell under different physiological con— ditions..." there is little doubt that the glycocalyx is an integral part of the complex sequence of events occurring in the brush border, and that alteration in, or loss of, the glycocalyx will result in loss of essential sequential steps in the digestive and absorptive process. Whether of a primary or secondary nature, the partial or complete loss of the glycocalyx from villal epithelial cells observed in neonatal pigs monocontaminated with E. coli 0138:K81:NM would be eXpected to contribute materially to inadequate digestion and malabsorption. The Apical Trilaminar Membrane. This membrane, to which the glycocalyx is closely applied, is continuous over the microvillus, the support of the whole Structure apparently depending on the integrity of the cytoskeleton. The cyto- skeleton, or core, of the microvillus consists of groups of rods, each rod composed of "...a single strand of dense material helically wound upon itself to form a compact rod" (Eichholz, 1969). The design of the arrangement of these rods as described in this thesis (see Results) will no doubt be found to be more complex as improved stained preparations are available for examination, and indeed this has already been found to be so (Dress, D. T., personal communication). The effect on microvilli of the presence of E. coli 0138:K81:NM, or its enterotoxin, is not immediately apparent. Staley et a1. 301 ‘! (1968c, 1968d) described "exfoliation" of the microvilli to form microerosions in the apical plasmalemma prior to entry of the micro- organisms into the villal epithelial cell. Although in the present experiments many instances were seen in which E. coli 0138:K81:NM organisms were closely approximated to the brush border, no discrete erosions of the microvilli directly associated with these microor- ganisms were seen. Compared with Ehe.intracyt0plasmic organelles, microvilli are relatively resistant to the toxic principle associ- ated with E. coli 0138:K81:NM, although in areas of high concentra- tion of the organism or its enterotoxin, such as close proximity to the viable bacteria, the microvilli are affected. It was observed that, where bacteria were adherent to the brush border, there was the formation of bizarre forms, foreshortening and stunting of the microvilli (Figures 96 and 99) extending over the entire apical surface of a particular cell and, at times, extending to neighboring cells. It was further observed, however, that in both germfree and monocontaminated pigs, as cells matured and moved along the basal lamina toward the tip of the villus, the microvilli be- came generally shorter and broader. One reason for this change can be prOposed. There are no ribosomes or granular endoplasmic reticu- lum associated with the membranes constituting the apical plasma- lemma, and there is no apparent ability of the apical trilaminar membrane to repair itself in case of injury or to replace used apical membrane. The regularity with which pinocytotic vesicles are foamed from the apical plasmalemma during the Short life of the villal epithelial cell is remarkable. It is reasonable to speculate 302 that, as pinocytotic Vesicles form and are pinched off at the bases of the microvilli, so the trilaminar membrane of the apical plasma- lemma is drawn from the walls of the microvilli to accommodate de- pletion of the trilaminar membrane. As the cell reaches the end of its life at the apex of the villus, so according to the amount of pinocytotic activity it has accomplished, it will have either long or foreshortened microvilli. In those cells altered by intracytOplasmic vacuolization and organellar disruption, the brush borders were seen to have under- gone marked degrees of degeneration. In summary, in these experiments, three basic forms of micro- villal change were observed: first, foreshortening and the forma- tion of abnormal forms associated with proximity to E. coli Ol38:K81:NM.and presumably due to high concentration of enterotoxin in the immediate vicinity; second, frank disruption of the microvilli associated with intracyt0plasmic disruption and disorganization of the cell with which the microvilli are associated; and third, a normal process of foreshortening apparently associated with matura— tion of the cell. The lack of cytoskeletal detail, seen in sections of microvilli from monocontaminated animals,was not a consistent finding, and too few examinations of this particular feature of diseased cells were made for valid conclusions to be drawn. The Terminal Web. McNabb and Sandborn (1964) considered that the terminal web provided a structural basis for the cytoskeletal rootlets of the 303 microvilli. This structure was not seen in tissues from baby pigs according to Staley et al. (1968a), who used glutaraldehyde fixation for their experimental animal tissues. It may be that this structure is more easily stained and demonstrated after osmium tetroxide fixa- tion since there was observed slightly more densely staining material in the region of the terminal web in villal epithelial cells of neonatal animals used in these experiments. It is understandable that, in those cells in which there is an extensive apical tubule system, the terminal web will be somewhat attenuated and perhaps made indistinct by the presence of the apical tubules. There were no apparent changes observed in the terminal web of cells from monocontaminated pigs that could be attributed to the primary effects of the enterotoxin. Changes that were observed were inconsistent and were associated secondarily with disruption of the apical plasmalemma in diseased cells. The Subapical Zone Unlike the situation reported by Staley et al. (1968a), organelles were commonly seen in the subapical zone of the intestinal absorptive cells of both germfree and monocontaminated neonatal pigs (Figures 70. 72, 74, 76, 80, 84, 85, 96, 100, and 102). The Apical Tubule System. The presence of the apical tubule system as described by Staley at al. (1968a) was consistently observed in the mature villal epi- thelial cells of germfree and monocontaminated neonatal pigs. Accord- ing to Staley at al. (1968a), the similarity of the membrane structure 304 of the apical tubules, the vacuoles, and the plasmalemma of the vil- lal epithelial cell leads one to conclude that the vacuoles might be derived from, or be extensions of, the plasmalemmal membrane struc- ture extending from the cell surface in the form of tubules into the apical cytOplasm. There were many instances in the experiments reported herein where it was clearly apparent that the laminae of the apical tubule system were continuous with (and possibly derived from) both the trilaminar membrane of the brush border and the mem- branes of the apical end0p1asmic reticulum. As was hypothesized in the discussion of the apical trilaminar membrane (vide supra), it might be that this membrane, in a process somewhat resembling pinocytosis and conjointly with the end0p1asmic reticulum, contribute laminar material to the apical tubule system. Some evidence for this speculation lies in the observation that the microvilli are relatively longer in those more recently produced villal epithelial cells, than they are in cells that have an extens- ive apical tubule system. Changes in the apical tubule system Specifically associated with monocontamination with E. coli 0138:K81:NM were not apparent in these eXperimentS. It was not possible to determine whether the disruption of the apical tubule system in the monocontaminated animals was secondary to general disorganization and disruption of the cells or if it was primarily affected by enterotoxin. The possible role of the apical tubule system in the absorption of enterotoxin is discussed elsewhere in this thesis. 305 Endoplasmic Reticulum and the Apical Vacuoles There was morphologic evidence that the membrane-enclosed vacuoles observed in the cytOplasm of both germfree and monocontaminated neo- natal pigs were probably distended elements of the end0p1asmic reticu- lum. In view of the widespread gross vacuolization seen in both germfree and monocontaminated animals it seems that the most likely explanation for the presence of these vacuoles is that they are associated with absorption. Electron microsc0pically it was apparent that the vacuolization associated with E. coli infection was dif- ferent from that of the germfree neonates in that, in infected ani- mals, the membranes enclosing the vacuoles eventually fragmented, allowing escape of the vacuolar fluid into the extramembranous cytoplasm. In sharp contrast, the vacuoles of the villal epithelial cells from germfree control pigs were seen to remain surrounded by intact membranes, despite instances of extreme distention with absorbed ' were present in the vil- fluids. So-called "nonmembranous vacuoles' lal epithelial cells of both the germfree and monocontaminated ani— mals, and these were considered to be localized areas of fluid con— taining little or no particulate matter which had been absorbed through the apical plasmalemma into the cytoplasmic matrix but had not yet entered the endoplasmic reticulum. Such vacuoles are apparent in Figure 72 and can be compared with membrane-enclosed vacuoles in the same electron micrograph. The spine—like structures observed in the membrane-enclosed tubules of the subapical cytOplasm of the intestinal absorptive cell 306 have been postulated by Staley et al. (1968a) to be possible recep- tor sites or possible sites for enzyme activity. The presence of phosphatases in the plasmalemma (Goldfischer at al., 1964) and their association with pinocytotic vesicle formation (Marchesi, 1965) has been cited by Staley et al. (1968a) as evidence of the probability of their postulate. The end0p1asmic reticulum has a number of metabolic activities which are characterized by the enzymes associated with its membrane component. Although an electron histochemical study of the changes in the end0p1asmic reticulum associated with colibacillosis was not ‘made in these experiments, the changes in vacuolar membrane struc- ture and function, when considered with the histochemical findings (vide supra) of a general reduction within the cytoplasm of both adenosine triphosphatase and lactate dehydrogenase activity, provides additional support for the concept of a biochemical lesion in the membranes of the end0p1asmic reticulum's being associated with enterotoxin activity. It is reasonable to expect that such lesions would cause a failure in membrane function, expressed either as dis- ruption of the membrane or as an inability of the membrane to trans- port fluids in one or both directions. If this be true, the disten— tion and rupture of the membrane-enclosed vacuoles observed in the infected cells could be expected to occur. The major electron microscOpic difference between the granular and agranular end0p1asmic reticulum is the presence of ribosomes in the limiting membranes of the former structure. Indeed, it is likely that the two major types of end0p1asmic reticulum are a continuous 307 system of planar membranes and tubules, their respective functions and enzyme content varying according to whether simple transport, synthesis of protein "packaging" materials (in which ribosomes are required), specialized transport enzymes, and/or synthetase mechanisms, etc., are present. In those markedly vacuolate cells from control animals it can be hypothesized that there is a lag in the transport of fluids from the vacuoles of villal epithelial cells of neonatal germfree pigs but that this removal is successfully accomplished with the passage of time. It may be that such vacuolate cells are incapable of fur- ther absorption until the fluids that have already been absorbed are successfully evacuated by the end0p1asmic reticulum. The fact that the membranes remain intact around these vacuoles, deSpite marked distention, is probably indicative of their health and is in very strong contrast with the extensive disruption of membranes within the cells of animals monocontaminated with E. coli 0138:K81zNM. The vacuolization so characteristic of villal epithelial cells from the more terminal Sites of the small intestine of neonatal germfree pigs is probably a Sign of either increased absorptive efficiency of the apical plasmalemma or alternatively (and more likely) relative deficiency or inefficiency of one or more enzymes of the endoplasmic reticulum that are required for the rapid and efficient transport of absorbed fluids within the cisternae of the end0p1asmic reticulum away from the cell and into the lamina prOpria. Whether this postulated enzyme deficiency is characteristic of germfree pigs and is not made up until a normal intestinal flora 308 is established is not known. Certainly, the feces of germfree pigs are more fluid than those of conventional animals for at least 4 weeks from birth, and the villal epithelial cells from germfree pigs retain a markedly vacuolate appearance throughout this time (Christie, unpublished data). It is apparent that, in germfree animals, despite vacuolization of these cells, absorption proceeds at an adequate level, whereas in neonatal pigs monocontaminated with E. coli 0138:K81zNM, the vacuolization and eventual disruption of the villal epithelial cell is such that the cell fails as an integral part of the absorptive mechanism. There was no evidence that the vacuolization of the villal epi- thelial cells of either the control animals or the monocontaminates was due to the collection of fat within the cytOplasm. The presence of stained fat in the apical cytOplasm of villal epithelial cells in germfree and some monocontaminates agrees with the concept of lipid absorption in micellar phase through the apical plasmalemma (Isselbacher, 1967), and by the apical pinocytotic sys— tem of tubules and vacuoles (Mattisson and Karlsson, 1964; 1965; 1966). Once inside the cell, the finely dispersed lipid aggregated just below the terminal web, where drOplets could be seen with no surrounding membrane. Deeper within the cell, the draplets were seen to be surrounded by both granular and agranular endoplasmic reticulum. According to Dawson (1967), the drOpletS leave through the base of the cell into the intercellular space, pass into the extra- cellular fluid and then proceed through pores into lymphatics. Some 309 evidence in confirmation of this chain of events was seen in electron micrographs of a villal epithelial cell from a 24-hour monocontami- nated pig which was actively absorbing fat. It is clear that impairment of digestion of fat within the lumen will occur if bile salts or'pancreatic enzymes are deficient or in some way made ineffective. There is no evidence to suggest that E. coli Ol38:K81:NM infection adversely affects either the pancreas or the liver in neonatal gnotobiotic pigs (Christie, 1967). However, in view of the marked derangement of membranous intracellular struc- tures of the villal epithelial cells that is associated with this infection, one might expect that the derangement would also lead to deranged function of membrane-bound enzymes. It was apparent that stainable fat had entered the apical cytOplasm of villal epithelial cells of the duodenum and early jejunum in monocontaminated and clinically affected pigs, whereas in tissues more distal in the small intestine, no oil—red-O—positive material was observed in the laminae prOpriae. Since fat in the form of triglycerides or chylomicrons could be demonstrated in electron micrographs of affected cells, and since the entry of the micellar lipid is a nonenergy process not mediated by enzyme, it may well be that fatty acid and monoglyceride in mi- cellar form.can enter a diseased villal epithelial cell through a patent brush border, yet not be further absorbed because of disrup— tion of the triglyceride synthetase mechanism. Enterotoxin of E. coli could disrupt this mechanism in many ways —- either directly or indirectly. This could occur directly 310 by chemical destruction of triglyceride synthetase in the agranular reticulum or by combination with one or more of the enzyme or sub- strate molecules of either the biochemical energy-producing cycles or protein—synthesis mechanism of the granular reticulum. This could occur indirectly by physical disruption of these or other inter- dependent systems. Electron microsc0pic evidence of specific injury to either the granular or agranular end0p1asmic reticulum could be considered presumptive evidence of this aSpect of pathogenesis of fat malabsorption. The villal epithelial cells from the jejunum and ileum of the clinically affected animals were generally markedly vacuolate, in some instances to the point of complete disruption. It is more likely that deficiency of the triglyceride synthetase mechanism, or failure of the protein synthesis required in the manufacture of chylomicrons, is a phenomenon secondary to more generalized organel- lar disruption that contributes to the overall malfunction of the absorption mechanism in colibacillosis. The Nucleus Staley at al. (1968a) described the position of the nucleus within the cell and related it to the position of the epithelial cell on the villus. At the base of the villus they described the nucleus as being closer to the basement membrane, whereas toward the tip of the villus the nucleus "...occupied the apical portion of the cytoplasm". 31l In contrast, it was seen in these experiments and in those previously reported (Christie, 1967) that the position of the nucleus within vacuolate villal epithelial cells was not consistent accord- ing to position of the cell on the villal stalk. A degree of con— sistency of nuclear position in neighboring villi was observed, in that on a single villus, and often in neighboring villi, cells would generally have basally situated nuclei, whereas sometimes in the same section of the tissue, but at a little distance from that site, villi and their immediate neighbors would generally have apically Situated nuclei. It is unlikely that this positioning of the nucleus indicates a basic difference in the nature of the end0p1asmic reticulum of the subapical and the basal cytOplasm. Whether the nucleus is pushed to the apex of the cell or to the base of the cell will depend on which of the sites of end0p1asmic reticulum activity is less efficient in transport of absorbed fluids and hence becomes vacuolate. It is tempting to accept the observation of Staley et al. (1968a) and postulate that the change is one of enzyme maturation, aging, or senility, according to the position of the cell on the villal stalk, but their observation was not confirmed by this author. An alternative hypothesis, that the vacuolization and therefore the nuclear position depends on the relative content of the chyme -- carbohydrate, fat, protein and water -— bathing the villi at that particular time, is proposed. It was of interest to observe that in electron micrographs of cells in which absorbed fat was readily detected, the nucleus was displaced into the apex of the cell (Figure 312 80), whereas in those vacuolate cells in which fat was not readily observed, the vacuolization was apical and tended to push the nuclei towards the base of the cells. No relationship was detected between infection with E. coli and the position of the nucleus within affected cells from the tissues examined in these experiments. At no stage in these tissues were the nuclei of the affected cells seen to undergo the classic changes described for necrosis -- pyknosis, karyoschisis, karyorrhexis or karyolysis. The Mitochondria The mitochondrion is composed of two discrete membranes, "...aggregates of controlled, directional metabolic processes" (Korn, 1969), which catalyze the biochemically integrated reaction sequences of fatty acid oxidation, the tricarboxylic acid cycle, oxidative phOSphorylation and electron transport. Its functions include selective permeability, organization of the electron trans— port and adenosinetriphOSphate—generating cycles, conversion and storage of chemically derived energy, and loci for specific anti- gens (Johnson, 1969). The enzymes of these processes are not soluble nor are they floating freely in the matrices of the cristae mito- chondriales but are most probably membrane-bound in a lattice or mosaic of integral subunits constituting the membrane itself and arranged in a spatial configuration that favors the sequential reac- tions of the particular metabolic cycle. 313 As was forecast Speculatively in the interpretation of the histochemical tests for adenosine triphosphatase and succinate dehy- drogenase and confirmed morphologically during electron microscopic examination of comparable tissues, vacuolization of the cristae mitochondriales, followed successively by swelling, internal dis- ruption, and eventually rupture of the mitochondria, was associated with monocontamination of the villal epithelial cells of neonatal pigs with E. coli 0138:K81:NM. The early changes, primarily moderate vacuolization of the cristae mitochondriales, were consistent and easily detected; more advanced degenerative changes were such that at times it was diffi- cult to recognize the disrupted mitochondria, and they were likely to be confused at this stage of degeneration with membrane-enclosed vacuoles containing membranous and particulate material. Again, it is not possible to state from these results whether the mitochondrial changes are primarily the result of enterotoxin activity on their constituent enzymes or secondarily the result of generalized disruption of the cytOplasmic constituents. However, considering the basic biochemical role of adenosine triphosphatase both in the mitochondria and other organelles, it is tempting to Speculate that E. coli 0138:K81:NM.enterotoxin has a specific toxic effect on adenosine triphosphatase and, through this effect, has the ability to produce the Spectrum of ultrastructural changes observed in the tissues from these experimental animals. 314 The Golgi_Complex In mature cells, lytic enzymes are transferred from the Golgi complex by coated vesicles to digestive vacuoles where degradation of some intracytOplasmic protein occurs (Friend and Farquhar, 1967), and it may be that this is one of the functions of the Golgi complex in the villal epithelial cells of the neonatal pig. The probable role of the Golgi complex in the production of the glycocalyx has already been discussed (vide supra). Some vacuolization of the arch0plasm.was observed in cells from germfree control pigs and in the cells from monocontaminated pigs. The general reduction of acid phOSphatase activity observed in the villal epithelial cells from monocontaminates could indicate loss of lysosomal activity or failure of the Golgi complex in its lytic enzyme-producing activity. Again, the reduction in amount of detectable glycocalyx associ- ated with monocontamination with E. coli could be the result of failing activity of the Golgi complex, either due to organellar disruption or direct toxic activity of the E. coli enterotoxin. Other reasons accounting for this observation suggest themselves -- inability of the cell transport system to carry the preformed elements of the glycocalyx from the Golgi complex to the brush border, dis- ruption of the transport route, changes in the brush border prevent- ing the deposition or replenishment of the glycocalyx, or direct lytic or toxic activity of the enterotoxin upon the glycocalyx. The precise effect of enterotoxin on the Golgin complex was not determined in these experiments. 315 Maturation and Necrobiosis of Villal Epithelial Cells The high turnover rate of villal epithelial cells theoretically necessitates a high level of reproduction of epithelial cells within the crypts and requires that the same number of cells be lost, or Shed, from the tips of the villi. Even if the villal epithelial cell turnover time for neonatal pigs is as long as that recorded for man, it is obvious that some hundreds of cells per day are lost from each villus. The exact nature of this necrobiosis has not yet been determined despite the work reported by Leblond and Messier (1958), Deschner at al. (1963), Abrams at al. (1963) and Lesher et a1. (1964). It is clear that, with this level of necrobiosis occurring, two major considerations need to be made in interpretation of epithelial lesions associated with colibacillosis in neonatal gnotobiotes. First, it is likely that epithelial cells near the tip of the villus are already undergoing changes which will soon lead to necro- biosis. If this be true, then cells at the tip Of the villus observed under both Optical and electron microscopes will have Structural changes not only associated with any specific or experimental condi- tion (such as colibacillosis) but will also exhibit changes associ- ated with incipient necrobiosis. Such changes may lead to confusion in interpretation of ultrastructural changes associated with the intended stress. Second, Since there is ultrastructural evidence of limited ability of the villal epithelial cell for biosynthesis of macro- molecules (except for the function Of the granular end0p1asmic 316 reticulum), the changes leading to necrobiosis may well be simply the complete utilization of the existing stores of enzymes and organellar components and an inability to repair or replace these deficiencies. Hence, after a Short life Span, the cell becomes effete and is lost from the villus. Since highly specialized cells, such as the villal epithelial cells, have a limited range of response to stress, it can be readily imagined that an effete villal epi- thelial cell, no longer replete with its normal enzyme complement and other components, will probably react morphologically with vacuolization and breakdown of organelles, hydropic degeneration, and rupture of the cell. Those cells not so expended are presumably pushed from the basal lamina at the tip of the villus by the stream of ascending cells following behind. It is postulated that the ultrastructural changes within the villal epithelial cells associated with colibacillosis due to E. coli 0138:K81:NM resemble, to a degree, the ultrastructural changes asso- ciated with necrobiosis. The most likely difference is that most of the villal epithelium is affected in enteric colibacillosis and that such epithelium is retained on the villus (see later), whereas in conventional or germfree neonates the cells undergoing necrobiosis are restricted to the tips Of the villi and are rapidly shed when no longer functional. In order to avoid the possibility of confusion between ultra- structural changes in villal epithelial cells due to E. coli entero- toxin and those due to necrobiosis in experiments reported in this thesis, only villal epithelial cells from the midvillus and basal third of the villus were examined. 317 Kenworthy at al. (1967) reported an electron microscopic study of the epithelium from the tips of the villi of pigs contaminated with Specific serotypes of E. coli. Their description of ultrastruc- tural changes in these cells closely resembled those changes one would expect to find in effete villal epithelial cells, and there was apparently no way these authors differentiated between the patho— logic effects of their test organisms and ultrastructural changes associated with necrobiosis. In tissues from pigs 42 hours Old, Staley et al. (1968a) recorded finding on the villi long columns of degenerating cells "...stacked on their tips". The figure purporting to display this phenomenon does not bear out their conclusion. It is apparent that the villus depicted had curved out of the plane of the tissue section and they had in fact photographed a sagittal section Of a portion Of the villus. The alleged change in cell type depicted in their photograph, from columnar to cuboidal, and the apparent loss of the brush border, is entirely compatible with change in the plane of the cells from longi- tudinal section, as they depict in midvillus, to transverse section as is depicted toward the tip of the villus. There is little doubt that some of the cells they Observed were degenerating. Their description of the intracellular changes of swelling and fragmenta- tion of the microvilli, swelling, vacuolization, and loss of the cristae mitochondriales, are what one would anticipate in effete villal epithelial cells; and their electron microscOpic confirmation of this concept constitutes a valuable contribution to the litera- ture. It is Of interest to note that an additional "degenerative" 318 change described was the presence of intracytOplasmic membrane- enclosed vacuoles possibly originating from degenerating apical tubules and mitochondria. There are alternative conclusions that can be drawn from their evidence (vide supra) which tend to support the hypothesis that these authors (Staley et al., 1968a) have possibly confused the intracellular signs of normal function of the jejunal villal epi- thelial cell with the Signs of cellular degeneration. In interpreting Vodovar's conclusion that the pig intestinal epithelial cell must elongate during maturation (Vodovar, 1964) Staley et al. (1968a) stated that the villi reach apparently mature length before 42 hours because, in their experimental pigs killed 42 hours after birth, degenerating cells were evident at the tip of the villus. Although the latter statement has already been ques- tioned (vide supra) it is also an unlikely assumption that the process of growth, maturation and necrobiosis Of villal epithelial cells does not begin and continue until after the birth Of the fetus. It is more reasonable to assume that from the moment of the embryological development of the Specialized epithelium of the small intestine these cells multiply, mature, die and are shed during the intrauterine life of the fetus. In this case one would expect to find mature cells undergoing necrobiosis in the villal epithelium of pigs at the moment of birth. In relation to the position of the Golgi complex in the mature cell, Staley et al. (1968a) intimated that in the neonatal pig the Golgi complex is in a subnuclear position whereas 319 "...the Golgi is located in a supranuclear position in the three week old piglet, indicat- ing a change in polarity of the cell as the cell matures". Arguing along similar lines as before, one could reasonably expect to find "mature" cells immediately prior to the stage of necrobiosis at the tip of the villus in the neonate, and if the hypothesis of Staley et al. (1968a) is correct, one would see cells in that posi— tion with supranuclear Golgi apparatus. This is generally not so, hence another explanation can be pro- posed, and that is a process of reorganization rather than maturation. Payne and Marsh (1962) reported that pigs maintained on a protein- free diet up to 106 hours after birth still actively absorbed immuno- globulins, whereas contact with colostrum and/or protein initiates the phenomenon of "closure" (Lecce at al., 1964). Lecce et al. (1964) were also able to "close" the epithelium to globulin absorp- tion with a heat-stable low-molecular—weight component dialyzed from colostrum or skim milk. It may be that the reorganization of the cell associated with "closure" not only involves the loss of the apical tubular system and decline in pinocytotic activity but also relocation of the Golgi complex and elongation of the entire villal epithelial cell. Mural Interdiggtations It is morphologically evident that the villal epithelial cells are attached laterally to each other by tight junctions called "junctional complexes" (Farquhar and Palade, 1963) or, more commonly, desmosomal attachments or terminal bars. These unions generally 320 extend for a short distance below the brush border. The remainder of the cell surface, laterally and in contact with the basal lamina should be considered the base of the cell through which water and solutes leave. There was, however, a remarkable degree of interdigi- tation of the neighboring cell walls, sometimes associated with desmosomes and sometimes not. Again, there were occasions when desmosomes were found low on the cell walls, close to the basal lamina. Kenworthy at al. (1967), in their paper describing ultrastruc- tural changes of villal epithelial cells associated with E. coli infection, depicted the alleged degeneration and disappearance of cell membranes on either side of desmosomal plaques. Reinterpret- ing their electron micrographs in the light of the findings of the experiments described herein, and considering the ultrastructural appearance and behavior of broken or ruptured membranes, it seems more likely that the phenomenon of intermittent disappearance of the cell membrane on either side of the desmosomal plaque that they depict is artefactual in that the membrane is present but is simply not resolved by the electron microscOpe at that particular point. Numerous instances of this phenomenon were observed during the study described herein and were readily distinguishable from those instances where membranes were in fact ruptured or disintegrating. Electron microscopically, it was seen that, even in cells in an advanced state of disruption, there were few, if any, of the "classical" nuclear signs of necrosis that are familiar to the opti— cal microscopist. It is not possible to say that the cells affected 321 by E. coli enterotoxin are necrotic in the classical sense. Although the villal epithelial cells are disrupted and effete, the mural interdigitations are retained. As a consequence, the affected cell is not lost or sloughed from the villus -- as it is in TGE -- hence the lamina prepria is not left denuded and does not retract, or shorten, as it does in TGE. In enteric colibacillosis the villus retains a covering of effete villal epithelial cells which, although relatively or completely ineffective in absorption (and possibly contributing to loss of tissue fluid from the edematous lamina prOpria) do serve to maintain the long villi that are characteristic of primary colibacillary enteritis. Staley at al. (1968a) stated that "...despite apparent degeneration and insta- bility of the cell membrane, the desmosomal attachments were intact and held the cells to gether. There was no evidence of nuclear de- generation as indicated by shrinking, serra- tion of the nuclear membrane, or abnormal clumping of the chromatin". After examination of their electron micrographs, and in consideration of the findings of the eXperiments described herein, it is likely that Staley et al. (1968a) had, in fact, been examining absorptive cells in a sagittal section of the villus. The lack of nuclear changes normally associated with degeneration and necrosis indicates that the cells they were describing were not, in fact, primarily degenerating cells but mature villal epithelial cells carrying on their functional activity. Such activity of necessity involves intracytOplasmic changes, some of which are found in a degenerating cell, and which in the final stages of necrobiosis will be associated 322 with the disruption of the organelles and extrusion of the nucleus and cytoplasmic contents, or else shedding of the intact cell from the extreme tip of the villus. The Lamina Propria In the germfree neonatal pig the lamina prOpria of the villus was seen to be relatively free of excess interstitial fluid (Figures 67, 69 and 77) and consisted primarily of lymphatic and vascular endothelium, Breucke fibers, some histiocytes, and occasional fibro— blasts in a fluid matrix in which some fine collagen strands, chylo— microns and fat drOplets were suspended (Figures 69, 77 and 78). The endothelial cells of capillaries were characterized by the presence of mitochondria and many micrOpinocytotic vesicles within the cytOplasm, marginal folding of the internal membrane, and the presence of erythrocytes in the lumen (Figure 78). Lymphatic endothelium was characterized by fewer mitochondria, fewer micropinocytotic vesicles and marginal folds, and the presence of particulate matter and delaminating membranes within the lumen (Figure 77). It was apparent that monocontamination with E. coli 0138:K81:NM was associated with increased amounts of interstitial fluid (edema) containing membranous fragments, collagen fibrils, and unidentified flocculent material, but these changes were not accompanied by dis- tinctive evidence of endothelial damage or malfunction of either capillaries or lymphatics. 323 It was observed in the laminae propriae of villi from neonates monocontaminated 14 or more hours beforehand that there were increased amounts of fluid of edema and collagenous fibrils. There was appar- ently reduced micropinocytotic activity of capillary endothelial cells and increased thickness of the basal laminae surrounding the capillaries (Figure 95). By 96 hours after monocontamination with E. coli 0138:K81:NM the lamina prepria was densely infiltrated with collagenous fibrils suspended in tissue fluid. Such collagen formation was not a fea- ture of comparable tissues from germfree pigs, but may well be pres- ent in the laminae prOpriae of conventional, clinically normal, neo- natal pigs. Zweifach (1964), reviewing E. coli endotoxin activity, recorded that it did not appear to have a direct cytotoxic action, but he did present evidence of a direct effect on the smooth muscle of the vasculature. The pattern of the effect of increasing levels of endotoxin is, first, at an increased rate, then at higher levels of endotoxin, at a decreased rate of sequential reaction of the vascula- ture due to endotoxic sensitization of the muscular vessels to exogen- ous and endogenous stimuli. There was no evidence of damage to the endothelium of the capillaries and venules. There was increased capillary permeability, and E. coli endotoxin -- a lipOpolysaccharide -- did produce stasis of the capillary bed by a mechanism of narrow- ing of the lumina of the draining vessels (Zweifach, 1964). Insufficient numbers of tissues from the laminae prOpriae were examined in these experiments to enable valid conclusions to be 324 drawn. In the tissues that were examined, however, the impression gained was not that of epithelial breakdown and leakage, but rather that of capillary contraction and pericyte contraction which agrees in part with Zweifach's findings. There was also an impression of impermeability and inactivity of the vessels of the lamina propria that would probably lead to inefficient drainage of fluids from the lamina prOpria, through the vessels, to the general circulation. This comment is somewhat contradictory to the observations made for the effects of E. coli endotoxin (Zweifach, 1964) and of Vibrio sp. toxin (Elliott at al., 1968) on the vasculature. A similarity between human intestinal lymphangiectasia, as described by Dobbins (1966), and certain aspects of colibacillosis of neonatal gnotobiotic swine can be seen in that there is apparent partial "closing down" of capillaries, retarded emptying of lacteals, and collection of fluid of edema in the intramembranous space and the interstices of the lamina propria. As it is for human intestinal lymphangeictasia in which there is closure of the interendothelial cell spaces and "...unusually prominent intracellular filaments", there are some detectable differences between the intestinal lymphatic endothelial cells of the germfree and the monocontaminated experi- ‘mental pigs. It is not likely that the mechanism of "exit block" that is responsible for fluid collection in the villal epithelial cells of lymphangiectasia patients is a major factor in the patho- genesis of fluid retention in the villi of the monocontaminated neo- natal pigs, although the role of this phenomenon as a contributory factor to colibacillary diarrhea should be kept in mind. 325 Absorption and Malabsorption In the state of absorption the net flow of fluid is from the intestinal lumen into the lamina propria and thence to the vessels of the villus. In the diarrheal state, the net flow is from the vessels to the intestinal lumen. In the latter case the net flow might be due either to increase in the amount of fluid excreted or lost from the villus to the lumen, the absorption component remain- ing unchanged; a diminished rate of absorption, the vessel to lumen flux remaining at a normal level; or else both diminished absorption and excessive fluid excretion or loss. This bidirectional flux holds for water and its electrolytes -- chloride, sodium, potassium, bicarbonate and others whose net flow varies markedly in diarrhea. The present state of knowledge is such that it is not yet possible to relate specific ultrastructural alterations of the villal epi- thelial cell to a particular biochemical defect. It would be fortui- tous if it were found that the enterotoxin of E. coli Ol38:K81:NM produced its effects through a single biochemical lesion, but it is clear from the clinical, histopathologic and ultrastructural manifes— tations that this is most unlikely. From the information abstracted in the Literature Review (vide supra) and the histochemical and ultrastructural changes reported in this thesis (at almost every step, from the terminal digestion of foodstuffs in the superficial layers of the glycocalyx to the eventual removal of metabolites by the vessels of the lamina propria), there are indications of disturbed function. Which of these absorp- tive systems, singly or in combination, is the specific target of 326 E. coli 0138:K813NM enterotoxin, and which are secondarily disrupted, could not be determined from the results of these experiments. It is clear that E. coli 0138:K81:NM adversely affects the brush border morphologically and enzymatically, so that the terminal digest— ive processes, particularly of proteins and sugars, essential for their initial absorption through the apical plasmalemma (Holt and Miller, 1961; Miller and Crane, 1961; Gray, 1967) will be markedly reduced or prevented. Those processes described for amino acid transport (Wiseman, 1953; Fisher, 1967; Newey, 1967), sodium and glucose transport (Lifson and Parsons, 1957; Parsons, 1967; Crane, R. K., personal communication), the transport of water (Fordtran, 1967), reciprocal- anion movement in ileum and colon (Fordtran, 1967), micellar fat absorption (Hoffman and Borgstrom, 1962; Isselbacher, 1967), and many others, less clearly defined, require an intact functional plasmalemma for Optimal function. Although much is yet to be discovered of the absorptive process at the level of the brush bor— der, it is apparent that the histochemical and morphologic changes in the brush border that are associated with colibacillary diarrhea will be significant factors in malabsorption. Again, energy dependent reactions involved in absorption, such as those for the sodium and potassium pumps (Robinson, 1960, Fisher, 1967; Newey, 1967; Parsons, 1967), absorption of basic carbohydrate molecules (Lifson and Parsons, 1957; Parsons, 1967; Crane, 1968), the postulated processes of "chemi-osmotic coupling" (Mitchell, 1961) —‘ _L_ 327 and the outward sodium pump at the brush border (Parsons, l967),and reesterification of triglyceride (Senior, 1964; Porter, 1969) will depend on functional mitochondria, end0p1asmic reticulum, and pos- sibly other cellular energy sources and utilization points. Mito- chondrial degeneration, and the diminution of succinate dehydrogen- ase and adenosine triphosphatase activity, are strongly suggestive of the likely failure of the energy-dependent essential absorptive processes in the intestinal mucosa of neonates monocontaminated with E. coli 0138:K81:NM. The intracellular processes concerned with absorption, carbo- hydrate metabolism (Aurrichio at al., 1963; Dahlqvist at al., 1963; Levin, 1967) and synthesis of chylomicrons from protein, cholesterol and phOSpholipid (Zilversmit, 1965) require the integrity of intra- cytOplasmic membranes for normal function, are enzyme mediated, and will be similarly disrupted in cells undergoing the degenerative changes observed in these experiments. Flow of water in absorption is so much a matter of controversy at this time that it is likely to be both an active and a passive process. Osmotic changes of the lumen contents have a marked effect on net water flow (Fordtran, 1967). Water absorption is known to depend on the active absorption of an ion such as sodium, but water also follows sodium down an osmotic gradient. On an osmotic basis, to postulate net flow of water from the vessels of the villus to the lumen requires that there be an increased number of osmotically active particles in the lumen. Theoretically, the source of these particles could be from vessels of the lamina prepria subsequent to 328 endothelial damage, allowing escape of large proteins and colloids to the lumen, from the action of bacteria in manufacture of osmoti- cally active metabolites, or from the failure of osmotically active particles, such as sodium, to be absorbed. From the results of these experiments, endothelial damage of vessels in the lamina prOpria, leading to escape of osmotically active substances, is not likely to be a factor in the enteritis associated with monocontamination with E. coli 0138:K81:NM, although further study of this aspect of pathogenesis for this and other serotypes of E. coli is necessary. Barnum at al. (1967) suggest that E. coli splits macromolecules normally present in the lumen into smaller molecules of increased osmotic activity. If this be true, should the presence of E. coli not alter the capacity of the intestinal epithelial membrane as a semipermeable membrane -— thereby permitting osmosis -- but so change the membrane as to prevent the proper absorption of the digested particles, then a net outflow of fluid from the intestinal wall to the lumen will result. Fordtran (1967) has referred to the ease of diffusion of sodium ions back into the lumen through the large aqueous pores of the apical plasmalemma of the villal epithelial cells in the duodenum and anterior jejunum. This same "leak back" is prevented in the terminal jejunum and ileum because of the relatively smaller size of the aqueous pores in that region of the intestine. Darrow (1946) considered that intestinal hypermotility was a major cause of inade- quate sodium absorption in infantile diarrhea. 329 Considering all of the above factors mediating against absorp— tion and the general ultrastructural disruption of the villal mucosa -- particularly in the terminal jejunum and ileum -- it is likely that sodium loss will be an important factor in neonatal colibacil- lary diarrhea. Factors external to the villus, and possibly of importance in the pathogenesis of colibacillary diarrhea have been discussed in detail elsewhere (Christie, 1967). CONCLUSION The need for future research will be at the ultrastructural and biochemical level if a more complete understanding of the patho— genesis of colibacillosis is to be achieved. The more detailed characterization of the ultrastructural and histochemical lesions, some of which are described in this thesis, awaits only the application (and in some instances the development) of apprOpriate electron histochemical procedures. The determina- tion of which lesions are primary and which are secondary should remove much of the speculation about pathogenesis that is presented in this thesis. Obviously, important lesions to be further investigated are those of the basic energy-producing processes involving adenosine triphosphatase and its related organelles -- the mitochondria, possibly the endoplasmic reticulum,and the brush border. The altera— tions of the glycocalyx are also obvious targets for intensive study and are likely to be important contributory factors in the patho- genesis of colibacillary diarrhea. Of equal importance are membrane transport studies, but these studies will require the techniques and attention of a number of scientific disciplines. It seems appropriate that future experimental work in neonatal colibacillary diarrhea should include analysis of intestinal contents, and ligated intestinal-loop contents, at all levels of the tract, 330 331 for sodium, potassium, bicarbonate, chloride, glucose and amino acid content. Suitable experimental treatment groups might well be con- ventional and germfree neonatal pigs, monocontaminated with E. coli 0138:K81:NM or inoculated with sterile preparations of enterotoxin from such an organism. The role of the colon in water absorption and ion exchange has long been known. Recently Levitan at al. (1962) and Fordtran (1967) have reported that in the normal animal only about 20% of the colon's capacity to absorb water and electrolytes is used. It is clear that future studies of colibacillary diarrhea should include consideration of the factors of secretion and absorption in the colon, and the motility of the large intestine, in addition to such changes that are known to exist in the small intestine. Although the experiments and observations described in this thesis are not comparable in some respects with the clinical syn- drome of infantile colibacillary diarrhea or with colibacillary diarrhea of conventional baby pigs, at least two of the unique char— acteristics of the neonatal gnotobiotic pig -- lack of placental transfer of immune globulins and relative freedom from bacterial antigens -- enables one to examine in much more elementary fashion the basic lesions of enteric colibacillosis. Such a study, uncomplicated by the manifestations of sensitiza- tion, immune phenomena, antigen-antibody responses, malnutrition, and the multiple histologic, physiologic and immunologic effects associated with the presence of other intestinal flora, is open to criticism from the viewpoint that infantile diarrhea and the neonatal 332 diarrheas of animals are never seen clinically in such an uncompli- cated state. If these factors are not forgotten, but put aside temporarily whilst the basic lesion is considered, then we might more easily understand the pathogenesis of colibacillary diarrhea. With a basic concept of the pathogenesis, it should then be possible, according to our knowledge of the modifying superimposed and complicating factors present in particular clinical situations, to develOp the means of prOphylaxis, alleviation or cure of the disease, in an individual, or as an entity. SUMMARY Forty-eight gnotobiotic pigs were used in experiments in which 34 animals were fed between 1.8 and 3.0 x 106 E. coli Ol38:K81:NM organisms. Observations were made of the clinical effects of the infection at varying times from 4 to 96 hours after monocontamination. The animals were killed and observations were made of the macro- scOpic lesions of disease. Tissues were taken from the mucosa at 5 levels of the small intestine and submitted to optical microscopic and electron microscOpic examination, and histochemical staining for alkaline phosphatase, acid phOSphatase, leucine aminopeptidase, succinate dehydrogenase, and lactate dehydrogenase activity. Diarrhea occurred between 8 and 12 hours after monocontamina- tion with E. coli. Raised hair coat, projectile watery diarrhea and dehydration were also recorded within 18 hours of monocontamination. At necropsy, the intestines were at times flaccid and fluid filled, and varying degrees of edema and hyperemia of the gastric and intestinal walls were observed. Occasionally ascites was pres- ent and in those animals scouring for longer periods, erythema venenatum was seen. The optical microscOpic lesions ranged from those of a mild acute enteritis to a histologic picture, in persistently scouring ‘lpigs, similar to that seen in clinically normal germfree control animals. The most common lesion observed in the villal epithelial cells from monocontaminated animals was marked vacuolization and 333 334 hydropic degeneration. This lesion was microsc0pically indistinguish— able from the vacuolization associated with absorption which was seen in both monocontaminated and germfree neonates. Electron microscopic examination revealed the basic difference between these two types of vacuolization. So-called absorption vacuoles were membrane-enclosed, probably within the end0p1asmic reticulum, and were associated with normal function, normal cyto— logic enzyme activity and distribution, and absence of diarrhea in host animals. Vacuolization of the mucosa of infected neonates was at first similar to that seen in the control animals, but within 12 to 20 hours of monocontamination with E. coli 0138:K81:NM, there was observed diminution in the form and histochemical function of the glycocalyx, loss of alkaline phOSphatase and leucine aminopep- tidase activity from the brush border, and degeneration of the microvilli. There was vacuolization of the cristae mitochondriales leading to swelling, gross enlargement and rupture of the mitochon- dria, vacuolization then lysis of the endoplasmic reticulum.and Golgi complex, and eventual complete disruption of the internal cytoplasmic structure. The cell then became a fluid-filled sac with its nucleus floating in a fluid matrix of membrane fragments and unidentified particulate matter. Associated with these changes were loss of activity of alkaline phosphatase, acid phosphatase, adenosine triphOSphatase, leucine aminopeptidase, succinate dehy- drogenase and lactate dehydrogenase. These effetecells generally were retained on the villal stalk by the persistence of the external cell membrane and the preservation 335 of the mural interdigitations with neighboring cells. At times rup- ture of the apical plasmalemma and extrusion of the cell contents into the intestinal lumen were seen. Edema of the lamina prOpria was observed in the monocontaminated neonates, followed by infiltration of the interstices with collagen fibrils. Some evidence of contraction and functional inactivity of the vasculature of the lamina prOpria was observed. It is speculated (1) that the enterotoxin of E. coli Ol38:K81:NM is readily absorbed by the villal epithelial cell and is toxic for membrane-bound enzymes, (2) that it exerts its toxic effects pri- marily on those energy-producing systems involving adenosine triphosphatase -- the mitochondria, the endoplasmic reticulum and the Golgi complex, (3) that it is primarily toxic for the glycocalyx and secondarily toxic for this entity by its effect on the Golgi complex, and (4) that it is secondarily disruptive to the laminar organelles of the intestinal absorptive cell. Persistence of mural interdigitations permitting continued intercellular adhesion with retention of the effete cells on the villal stalk is hypothesized as an explanation for the major dif- ference in the histologic appearance of the small intestine in colibacillary diarrhea and the enteric diseases involving villal atrOphy in man and animals. It is speculated that the endotoxic fraction of E. coli 0138:K81:NM.enterotoxin is responsible for the inflammatory re- aponses observed in the lamina propria and the basic changes in villal function. 336 It is considered that neonatal colibacillary diarrhea is pri- marily a disease of malabsorption. LIST OF REFERENCES LIST OF REFERENCES Abrams, G. D., Bauer, H., and Sprinz, H.: Influence of the normal flora on mucosal morphology and cellular renewal in the ileum. A comparison of germrfree and conventional mice. Lab. Invest., 12,(l963): 355-364. Adam, A.: Dyspepsiekoli. Zur Frage der bakterieZZen Ztiologie der sogen. alimentaren Intoxikation. Jb. Kinderheilk. phys. Erzieh., ll6,(l927): 8-40. Albersheim, P., and Killias, U.: The use of bismuth as an electron stain for nucleic acids. J. Cell Biol., l7,(1963): 93-103. Anderson, J. W.: Ultrastructural correlates of intestinal protein absorption. Report of the Ross Conferences on Pediatric Research, 1963—1965, Ross Laboratories, Columbus, Ohio, 50,(1965): 28-39. Anon.: Summary of reports on the health conditions in the Americas, 1950-1953. (Scientific Publication No. 25) Pan American Sanitary Bureau, Washington, D.C. (1956). Anon.: Proceedings of the National Serminar on Enterobacteriaceae, June 8/9, 1961. Kasauli, Punjab. Indian J. med. Res., 50,(1962): 640-659. Armed Forces Institute of Pathology: Manual of'Histologic and Special Staining Technics, 2nd edn, Washington, D.C. (1960). Auricchio, S. A., Rubino, A., Landolt, M., Semenza, G., and Prader, A.: Isolated intestinal lactase deficiency in the adult. Lancet, 2,(l963): 324-326. Badenoch, J.: The blind loop syndrome. In Mbdern Trends in Gastro- enterology second Series, F. Avery Jones, ed. Paul B. Hoeber, Inc., New York, N.Y. (1958): 231-242. Bahr, L. von,: Zentbl. Bakt. Parasitkde., 66,(l912): 335. cited by Adam (1927). Barka, T.: On the acid phosphatases of liver and RE-cells. J. Histochem. Cytochem., 8,(l960): 320-321. Barka, T.: Studies of acid phosphatase. I. ElectrOphoretic sepa- ration of acid phosphatases of rat liver on polyacrylamide gels. J. Histochem. Cytochem., 9,(l961): 542-547. 337 338 II. Chromatographic separation of acid phosphatases of rat liver. J. Histochem. Cytochem., 9,( 1961): 564-571. Barka, T.: Electrophoretic and chromatographic heterogeneity of acid Egosphatases of rat liver. J. Histochem. Cytochem., 9,(l961): 0-6310 Barnum, D.A.: Personal communication, Michigan State University, July 24, 1968. Barnum, D.A., Glantz, P.J., and Moon, H.W.: Colibacillosis. Ciba Pharmaceutical Co., Summit, New Jersey (1967). Barrnett, R.J.: The demonstration with the electron microscope of the end-products of histochemical reactions in relation to the fine structure of cells. Expl Cell Res., Suppl. No. 7, (1959): 65-89. Bennett, P.C.: Edema Disease. In Diseases of'Swine, 2nd edn, H.W. Dunne, ed., The Iowa State Univ. Press, Ames, Iowa (1964). Bessis, M.: Studies on cell agony and death: An attempt at classi- fication. In Cellular Injury (Ciba Foundation Symposium) Symposium on Cellular Injury. A.V.S. de Reuck and Julie Knight, eds., J. and A. Churchill, Ltd., London (1964): 299. Bizzozero, G.: Uber die Regeneration der Elemente der schlauch— fbrmigen Drusen und des Epithels des Magendarmkanals. Anat. Anz., 3,(1888): 781—784. Black, D.A.K.: Body-fluid depletion. Lancet, l,(1953): 353-360. Brambell, F.W.R.: The passive immunity of the young mammal. Biol. ReVQ, 33’(1958): 488-5310 Bray, J.: Isolation of antigenically homogeneous strains of Bact. coli neapolitanum from summer diarrhoea of infants. J. Path. Bact., 57,(l945): 239-247. ion of Bacterium coli var. Bra J. and Beavan T.: Slide agglutinat y, , ’ J. Path. Bact., 60,(l948): Neapolitanum in summer diarrhoea. 395-4010 Britt, A. L.: Pathologic effects of Escherichia coli O83:K.:NM in gnotobiotic pigs. Ph.D. Thesis, Michigan State University, East Lansing (1967). Eacherichia coli in gnotobiotic pigs. ium and Workshop, Michigan State June 7 (1964). Abstract. Britt, A. L., and Waxler, G.L.: Proceedings, Gnotobiotic Sympos University, East Lansing, Mich., 339 Brooks, R. A.: Serum proteins and antibodies in the germfree and gnotobiotic pig. M.S. Thesis, Michigan State University (1966). Burgos, M. A., Deane, H. W., and Barrnett, R. J.: Sites of reaction products in alkaline phosphatase preparations incubated with two classes of organic phosphate esters. J. Histochem. Cyto- chem., 2,(1954): 442—442. Burgos, M, A., Deane, H. W., and Karnovsky, M. L.: Histochemical and chemical evidence for more than one alkaline phOSphomono- esterase. J. Histochem. Cytochem., 3,(l955): 103-115. Bustad, L. K.: Pigs in the laboratory. Scient. Am., 214(6),(l966): 94-100. Bustad, L. K., and McClellan, R. 0.: Swine in biomedical research. Frayne Printing Co., Seattle, Washington, U.S.A. (1966). Butler, A. M., McKhann, C. F., Gamble, J. L., and March, P.: Intra- cellular fluid loss in diarrheal disease. J. Pediat., 3, (1933): 84-92. Caro, L.: Electron microsc0py radioautography: Golgi zone as a site of protein concentration in pancreatic acinar cells. Biochem. Cytol., 10,(196l): 37-43. Carpenter, P. L.: IMmunology and Serology, 2nd edn W. B. Saunders Co., Philadelphia, Pa. (1965). Carter, F. 8.: Paediatrics in Freetown, Sierra Leone. Archs Dis. Childh., 36,(l96l): 190-199. Christie, A. C.: A Study of the argentaffin cells with the electron microscope. Q. J1. microsc. Sci., 96,(1955): 295-301. Christie, B. R.: A muscular dystrOphy of Merino lambs in Western Victoria. Vict. vet. Proc., 20,(196l-62): 30-32. Christie, B. R.: Experimental colibacillosis in gnotobiotic pigs. M.S. Thesis, Michigan State University, East Lansing (1967). Clark, S. L.: The ingestion of proteins and colloidal materials by columnar absorptive cells of the small intestine in suckling rats and mice. J. biophys. biochem. Cytol., 5,(1959): 41-49. Code, C. F.: The semantics of the process of absorption. Perspect. Biol. Med., 3,(1960): 560-562. Code, C. P.: Proceedings of the cholera research symposium. U.S. Department of Health, Education and Welfare, Washington, D.C. (1965): 87. 340 Cooke, W. T.: Water and electrolyte upsets in the steatorrhea syn- drome. J. Mt. Sinai Hosp. N.Y., 24,(l957): 221-231. Cori, C.F.: The fate of sugar in the animal body. I. The rate of absorption of hexoses and pentoses from the intestinal tract. Crane, R. K.: Hypothesis for mechanism of intestinal active transport ‘of sugars. Fedn Proc. Fedn Am. Socs exp. Biol., 21,(1962): 891-895. Crane, R. K.: Enzymes and malabsorption: A concept of brush border membrane diseases. Gastroenterology, 50,(1966): 254-262. Crane, R. K.: A concept of the digestive absorptive surface of the small intestine. In Handbook of’Physiology. Alimentary Canal, C. F. Code ed., American Physiological Society, Washington, D.C. (1968). Crane, R. K.: Introductory remarks. Fedn Proc. Fedn Am. Socs exp. Biol., 28,(l969): 5-5. Crane, R. K.: Personal communication, Michigan State University, February 19, 1969. Creamer, B.: Variation in small-intestinal villous shape and mucosal dynamics. Br. med. J., 2,(l964a): 1371-1373. Creamer, B.: Small-intestinal mucosal dynamics and the environment. Br. med. J., 2,(l964b): 1373-1375. Creamer, B.: The turnover of the epithelium of the small intestine. Br. med. Bull., 23,(l967): 226-230. Crookshank, F. G.: (1922) cited by W. I. Card: Br. med. Bull., 23,(l967): 294. Cruickshank, R.: Diarrhoeal diseases in under-develOped countries. In The Theory and Practice of'Public Health, W. Hobson ed., Oxford University Press, London (1961): 139. Dahlqvist, A., Hammond, J. 3., Crane, R. Ki, Dunphy, J. V., and Littman, A.: Intestinal lactase deficiency and lactose in- tolerance in adults. Gastroenterology, 45,(l963): 488-491. Darrow, D. C.: Body-fluid physiology: the relation of tissue compo- sition to problems of water and electrolyte balance. New Engl. J. Med., 233,(l945): 91—97. 341 The retention of electrolyte during recovery from Darrow, D. C.: severe dehydration due to diarrhea. J. Pediat., 28,(1946): 515-540. Br. med. Bull., 23,(l967): 247-251. Dawson, A..M.: .Absorption of fats. De Duve, C.: The lysosome concept. In Lysosomes, Ciba Foundation Symposium, A.V.S. deReuck and M. P. Cameron, eds., J. and A. Churchill, London (1963). Electron microscOpy of the visceral yolk sac epi- Dempsey, E. W.: thelium of the guinea pig. Am. J. Anat., 93,(1953): 331-363. Dent, C. E., and Schilling, J.A.: Studies on the absorption of Bio- proteins: the amino-acid pattern in the portal blood. chem. J., 44,(1949): 318—333. Deschner, E., Lewis, C. M., and Lipkin, M.: In vitro study of human rectal epithelial cells. I. Atypical zone of H3 thymidine incorporation in mucosa of multiple polyposis. J. clin. Studies on the structural basis Diamond, J. M., and Tormey, J. M.: Fedn Proc. of water transport across epithelial membranes. Fedn Am. Socs exp. Biol., 25,(l966): 1458-1463. Passage of 1131 - albumin and 1125 - Dich, J., and Nielsen, K.: gamma globulin into the small intestine of the pig. Can. J. comp. Med., 28,(l964): 257-262. Dixon, J. M} S.: The fate of bacteria in the small intestine. Path. Bact., 79,(l960): 131-140. J. Dobbins, W. 0.: An ultrastructural study of the intestinal mucosa in congenital B-lipoprotein deficiency with particular empha- sis upon the intestinal absorptive cell. Gastroenterology, 50,(1966): 195-210. Electron microsc0pic study of the intestinal mucosa Dobbins, W. 0.: Gastroenterology, 51,(1966): in intestinal lymphangiectasia. 1004-1017. The absorption of folic acid and Q. J1. D013, A., and Girdwood, R. H.: labelled cyanocobalamin in intestinal malabsorption. Med., 29,(l960): 333-374. Donaldson, R..M}: Normal bacterial papulations of the intestine and their relation to intestinal function. New Eng. J1 'Med., 270,(1964): 938-945. 342 Studies on the pathogenesis of steatorrhea in the Donaldson, R. M.: J. clin. Invest., 44,(1965): 1815-1825. blind 100p syndrome. Donaldson, R. M.: Role of enteric microorganisms in malabsorption. Fedn Proc. Fedn Am. Socs exp. Biol., 26,(1967): 1426-1431. Doniach, I., and Shiner, M.: Duodenal and jejunal biopsies. II. Histology. Gastroenterology, 33,(1957): 71—86. The epithelium Donnellan, W. L.: The structure of the colonic mucosa. Gastroenterology, and subepithelial reticulohistiocytic complex. 49,(l965): 496-514. Drees, D. T.: Personal communication, Michigan State University, April 30, 1969. Dubos, R.: The microbiota of the gastrointestinal tract. enterology, 51,(1966): 868-874. Dubos, R., and Schaedler, R. W.: Some biological effects of the digestive flora. Am. J. med. Sci., 244,(l962): 35/265-41/271. Dunne, H. W.: Streptococcosis, colibacillosis and bordetellosis. In Diseases of’Swine, 2nd edn, H. W. Dunne, ed., The Iowa State University Press, Ames, Iowa (1964). Gastro- Duthie, H. L.: Methods for studies of intestinal absorption in man. Br. med. Bull., 23,(l967): 213-216. Edwards, P. R., and Ewing, N. H.: Identification of'Enterobacteriaceae, 3rd edn, Burgess Pub. Co., Minneapolis, Minn. (1962). Eichholz, A.: Fractions of the brush border. Fedn Proc. Fedn Am. Socs exp. Biol., 28,(l969): 30-34. Eichholz, A., and Crane, R. K.: Studies on the organization of the brush border in intestinal epithelial cells. I. Tris disrup- tion of isolated hamster brush borders and density gradient J. Cell. Biol., 26,(1965): 687-691. separation of fractions. Transfers of Elkinton, J. R., Winkler, A. W., and Danowski, T. 8.: cell sodium and potassium in experimental and clinical condi- tions. J. clin. Invest., 27,(1948): 74-81. Elliott, H. L., Carpenter, C. C. J., Sack, R. B., and Yardley, J. H.: Small bowel morphology in experimental canine cholera. A In Program and Abstracts light and electron microsc0pic study. The American Association of'the Sixty-fifth Annual Meeting. of Pathologists and Bacteriologists. Chicago, Ill. (1968). 343 Emmel, V. M.: Effects of HCl on alkaline phosphatase in kidney and intestine: Histochemical and quantitative study. Proc. Soc. exp. Biol. Med., 75,(l950): 114-117. Evanson, J. M., and Stanbury, S. W.: Congenital chloridorrhea or so- called congenital alkalosis with diarrhoea. Gut, 6,(l965): 29-38. Ewing, W. H., Tanner, K. E., and Tatum, H. W.: A new serotype of Escherichia coli associated with infantile diarrhea. Pub. Health Rep., 70,(1955): 107-114. Farquhar, M. G., and Palade, G. E.: Junctional complexes in various epithelia. J. Cell Biol., l7,(1963): 375-412. Fawcett, D. W.: An Atlas of'Fine Structure. The Cell. Its Organelles and Inclusions. W. B. Saunders Co., Philadelphia, Pa. (1966). Feigin, I., and WOIfe, A.: The phosphatase of the nervous system. J. Neuropath. exp. Neurol., l4,(1955): 11-22. Ferguson, W. W.: Experimental diarrheal disease of human volunteers due to Escherichia coli. Ann. N. Y. Acad. Sci., 66,(1956): 71-77. Ferguson, W. W., and June, R. C.: EXperiments on feeding adult volun- teers with Escherichia coli 111, B4, a coliform organism associated with infant diarrhea. Am. J. Hyg., 55,(l952): 155-169. Fey, H.: Neuere Untersuchungen fiber die Colisepsis des Kalbes. Schweiz. Arch. Tierheilk., 104,(l962): 1-12. Fey, H., and Margadant, A.: Zur Pathogenese der Kalber-Colisepsis. I. verteilung des Sepsistyps in den Organen. Zentbl. Bakt. Parasitkde., 182,(l96l): 71-79. Fey, H., Lanz, E., Margadant, A., and Nicolet, J.: Zur Pathogenese der Kalbercolisepsis. Dtsch. tierarztl. Wschr., 69,(1962): Finkelstein, R. A., Norris, H. T., and Dutta, N. K.: Pathogenesis of experimental cholera in infant rabbits. 1. Observations on the intra intestinal infection and experimental cholera produced with cell-free products. J. infect. Dis., ll4,(l964): 203-216. Fisher, R. B.: Absorption of proteins. Br. med. Bull., 23,(l967): 241-246. 344 Fordtran, J. S.: Speculations on the pathogenesis of diarrhea. Fedn Proc. Fedn Am. Socs exp. Biol., 26,(1967): 1405-1414. Fordtran, J. S., and Locklear, T. W.: Ionic constituents and osmo- lality of gastric and small-intestinal fluids after eating. Am. J. dig. D189, 11,(1966): 503’5210 Fordtran, J. S., Rector, F. C., Locklear, T. W., and Ewton, M. F.: Water and solute movement in the small intestine of patients with sprue. J. clin. Invest., 46,(1967): 287-298. Fric, P., and Lojda, Z.: Enzymes of the human jejunal mucosa. Acta gastro-enter. belg., 27,(l964): 526-530. Friedenwald, J. S., Gryder, R. M., and Gryder, J. W.: The substrate specificity of alkaline phosphatase of rat kidney. Archs Biochem. Biophys., 73,(1958): 262-272. Friend, D. S., and Farquhar, M. C.: Functions of coated vesicles during protein absorption in the rat vas deferens. J. Cell Biol., 35,(1967): 357-376. Fujimoto, Y., and Olson, C.: The fine structure of the bovine wart. Path. vet., 3,(1966): 659—684. Gay, G. C.: Eacherichia coli and neonatal disease of calves. Bact. Rev., 29,(l965): 75-101. Giles, C., Sangster, G., and Smith, J.: Epidemic gastro-enteritis of infants in Aberdeen during 1947. Archs Dis. Childh., 24,(1949): 45-53. Gitter, M.: The enteropathogenic effect of E. coli in rabbits and pigs. Br. vet. J., 123,(1967): 403-409. Glasser, R., Hupka, E., and Wetzel, R.: Die Krankheiten des Schweines, 5th edn, M. and H. Schaper, Hannover (1950). Goldfischer, 8., Essner, E.,;and Novikoff, A. B.: The localization of phosphatase activities at the level of ultrastructure. J. Histochem. Cytochem., 12,(l964): 72-95. Goldschmidt, R.: Untersuchungen zur Atiologie der Durchfallser- krankungen des Shuglings. Jb. Kinderheilk. phys. Erzieh., 139,(l933): 318-358. Goldstein, F., and Wirts, C. W.: Steatorrhea following gastrojejuno- tomy and vagotomy. Gastroenterology, 48,(1965): 819-819. .Gomori, C.: .Microscopic Histochemistry; Principles and.Practice. The University of Chicago Press, Chicago (1952). 345 Gomori, G.: Histochemical Specificity of phosphatases. Proc. Soc. exp. Biol. Med., 70,(1949): 7-11. Goodwin, M. H., Mackel, D. C., Ganelin, R. 8., Weaver, R. E., and Payne, F. J.: Observations on etiology of diarrheal diseases in Arizona. Am. J. trop. Med. Hyg., 9,(l960): 336-342. Gordon, J. B.: control of Communicable Diseases in Man. 9th edn, American Public Health Association, New York (1960). Gordon, J. E., Chitkara, I. D., and Wyon, J. B.: Weanling diarrhea. Am. J. med. Sci., 245,(l963): 129/345-161/377. Gordon, J. E., Pierce, V., Ascoli, W., and Scrimshaw, N. S.: Studies of diarrheal disease in Central America. II. Community prevalence of Shigella and salmonella infection in childhood. Am. J. tr0p. Med. Hyg., ll,(l962): 389-394. Gordon, R. S., Feeley, J. C., Greenough, W. B., Sprinz, H., and Oseasohn, R.: Cholera. Ann. intern. Med., 64,(l966): 1328- 1351. Gordon, W. A. M., and Luke, D.: Gastroenteritis in young piglets associated with Escherichia coli. Vet. Rec., 70,(l958): 542-543. Gossling, J., and Rhoades, H. B.: Serologic types of Escherichia coli isolated from certain pigs with enteric disorders. Cornell Vet., 56,(l956): 344-353. Govan, C. D., and Darrow, D. C.: The use of potassium chloride in the treatment of the dehydration of diarrhea in infants. J. Pediat., 28,(1946): 541-549. Graney, D. 0.: Uptake of ferritin by intestinal lining cells of suckling rats. Report of the Ross Conferences on Pediatric Research l963-l965,50,(l965): 18-27. Granger, B., and Baker, R. F.: Electron microscOpe investigation of the striated border of intestinal epithelium. Anat. Rec., 103,(l949): 459-459. Graumann, W.: Handbuch der Histochemie. Band II. Pblysaccharide. Gustav Fischer Verlag, Stuttgart (1964): pp. 13, 45, 329-332. Gray, G. M.: Malabsorption of carbohydrate. Fedn Proc. Fedn Am. Socs exp. Biol., 26,(1967): 1415-1419. Gray, G. M., and Santiago, N. A.: Disaccharide absorption in normal and diseased human intestine. Gastroenterology, 51,(1966): 489-498. 346 Gyles, G., and Barnum, D. A.: (1965) Unpublished data cited by Barnum, Glantz, and Moon (1967). Haelterman, E. 0., and H00per, B. E.: Transmissible gastroenteritis of swine as a model for the study of enteric disease. Gastro- enterology, 53,(l967): 109-113. Halsted, J. A., Lewis, P. M., and Gasster, M.: Absorption of radio- active vitamin B12 in the syndrome of megaloblastic anemia associated with intestinal stricture or anastomosis. Am. J. Med., 20,(l956): 42-52. Ham, A. W.: Histology, 5th edn, J. B. Lippincott Company, Philadelphia (1965). Handforth, C. P., and Sorger, K.: Failure of regeneration of small bowel mucosa following epidemic infantile gastroenteritis. Can. med. Ass. J., 84,(l96l): 425-427. Hardy, A. V., and Watt, J.: Studies of the acute diarrheal diseases. XII. Etiology. Pub. Health Rep., 60,(l945): 57-66. Hartman, R. S., Conrad, M. E., Hartman, R. E., Joy, R. J. T., and Crosby, W. H.: Ferritin—containing bodies in human small intestinal epithelium. Blood, 22,(l963): 397-405. Heatley, N. G., and Florey, H. W.: An antibiotic from Bacterium coli. Brit. J. exp. Path., 27,(1946): 378-390. Hess, R., Scarpelli, D. G., and Pearse, A. G. E.: The cytochemical localization of oxidative enzymes. II. Pyridine nucleotide- 1inked dehydrogenases. J. biophys. biochem. Cytol., 4,(l958): Hodes, H. L.: Etiology of infantile diarrhea. In Advances in Pediatrics, S. Z. Levine ed., Year Book Publishers, Chicago (1956). Hofman, A. F., and Borgstrom, B.: Physico-chemical state of lipids in intestinal content during their digestion and absorption. Fedn Proc. Fedn Am. Socs exp. Biol., 21,(1962): 43-50. Holt, J. H., and Miller, D.: The intestinal brush border as a digestive surface: Localization of aminopeptidase and phos- phatase in the mucosal epithelial cell. J. Lab. clin. Med., 58,(l96l): 827-827. Hooper, B. E., and Haelterman, E. 0.: Growth of transmissible gas- troenteritis virus in young pigs. Am. J. vet. Res., 27,(l966): 286-291. 347 Hopps, C. H.: Bacterial diseases. In Pathology, 5th edn, Vol 1, W. A. D. Anderson, ed. The C. V. Mosby Company, St. Louis, Mo. (1966). Hunt, R. D.: Microscopic histochemical methods for the demonstration of enzymes. In Selected Histochemical and Histopathological Methods, S. W. Thompson and R. D. Hunt. Charles C. Thomas, Springfield, Illinois (1966). Isselbacher, K. J.: Biochemical aspects of lipid malabsorption. Fedn Proc. Fedn Am. Socs exp. Biol., 26,(1967): 1420-1425. Ito, S.: The enteric surface coat on cat intestinal microvilli. Ito, S.: Structure and function of the glycocalyx. Fedn Proc. Fedn Am. Socs exp. Biol., 28,(l969): 12-25. Jackson, S. H.: The effect of food ingestion on intestinal and serum alkaline phosphatase in rats. J. biol. Chem., 198, (1952): 553-559. Jardetsky, 0.: Simple allosteric model for membrane pumps. Nature, Lond., 211,(l966): 969-970. Jelliffe, D. B., and Dean, R. F. A.: Protein-calorie malnutrition in early childhood. J. trop. Pediat., 5,(1959): 96-106. Jennings, R. A.: Gastroenteritis in the pig. Vet. Rec., 7l,(1959): 766-771. Jensen, C. 0.: (1889) C. 0. Jensen's Selected Papers 1886-1908. Ejnar Munskgaard, Copenhagen, l,(l948): 290, cited by W. J. Sojka, (1965). Jensen, C. 0.: Ueber die Kalberruhr und deren Aetiologie. Monatschr. prakt. Tierh., 4,(1892): 97-124. Joe, L. K., Sahab, K., Yauw, G. S., and Makaliwy, C.: Diarrhea among infants and children in Djakarta, Indonesia, with special reference to pathogenic Eacherichia coli. Am. J. tr0p. Med. Johnson, C. F.: Localization of hamster intestinal disaccharidase activity to a repeating unit associated with the microvillous plasma membrane. J. Lab. clin. Med., 68,(l966): 883-884. Johnson, C. F.: Hamster intestinal brush-border surface particles and their function. Fedn Proc. Fedn Am. Socs exp. Biol., 28,(l969): 26-29. 348 Jubb, K. V. F., and Kennedy, P. C.: Pathology of’Domestic Animals, VOZume 2. Academic Press, New York (1963). June, R. C., Ferguson, W. W., and Worfel, M. T.: Experiments in feeding adult volunteers with Escherichia coli 55, B5, a coliform organism associated with infant diarrhea. Am. J. Hyg., 57,(1953): 222-236. Kaneko, J. J., Moulton, J. E., Brodey, R. S., and Perryman, V. D.: Malabsorption syndrome resembling nontropical sprue in dogs. J. Am. Vet. med. Ass., l46,(l965): 463-473. Kassirer, J., Berkman, P. M., Lawrenz, D. R., and Schwartz, W. B.: The critical role of chloride in the correction of hypokalemic alkalosis in man. Am. J. Med., 38,(l965): 172-189. Kauffman, F.: Zur Serologie der Coli-Gruppe. Acta path. microbiol. scand., 21,(1944): 20-45. Kauffman, F.: The serology of the coli group. J. Immun., 57,(l947): Kauffman, F., and Dupont, A.: Escherichia strains from infantile epidemic gastroenteritis. Acta path. microbiol. scand., 27, (1950): 552-564. Kauffman, F., and Perch, B.: fiber die Coliflora des gesunden Menschen. Acta path. microbiol. scand., 20,(1943): 201—220. Kenworthy, R.: Influence of bacteria on absorption from the small intestine. Proc. Nutr. Soc., 26,(1967): 18-22. Kenworthy, R., and Allen, D.: Influence of diet and bacteria on small intestinal morphology with special reference to early weaning and Escherichia coli. Studies with germfree and gnotobiotic pigs. J. Comp. Path., 76,(l966): 291-296. Stubbs, J. M., and Syme, C.: Ultrastructure of small- d and unweaned pigs and pigs J. Path. Bact., 93,(1967): Kenworthy, R., intestinal epithelium in weane with post-weaning diarrhoea. 493-498. Kessner, D. M., Shaughnessy, H. J., Googins, J., Rasmussen, C. M., Rose, N. J., Marshall, A. L., Jr., Andelman, S. L., Hall, J. B., and Rosenbloom, P. J.: An extensive community out- break of diarrhea due to enteropathogenic Escherichia coli 0111:B4 I. Epidemiological studies. Am. J. Hyg., 76, (1962): 27-43. Kim, Y. S., Spritz, N., Blum, M., Tery, J., and Sherlock, P.: The role of altered bile acid metabolism in the steatorrhea of experimental blind loop. J. clin. Invest., 45,(l966): 956-962. 349 Kinnaird, P. J.: Progress of HyPAR pigs under commercial conditions at Dookie College. Vict. vet. Proc., 22,(l964): 37-38. Kohler, E. M.: Studies of Escherichia coli in gnotobiotic pigs. V. Evaluation of the effects of oral and parenteral admin- istration of immune serum. Can. J. comp. Med., 31,(1967): 283-289. Kohler, E. M.: Personal communication, University of Illinois, October 3, 1968. Kohler, E. M., and Bohl, E. H.: Studies of Eacherichia coli in gnotobiotic pigs. 1. Experimental reproduction of colibacil- losis. Can. J. comp. Med., 30,(l966): 199-203. Korn, E. D.: Current concepts of membrane structure and function. Fedn Proc. Fedn Am. Socs exp. Biol., 28,(l969): 6-11. Kowlesser, O. D., Haeffner, L. J., Riley, E. M., and Sleisenger, .M. H.: Comparative study of serum leucine amino-peptidase, 5-nucleotidase and non-specific alkaline phosphatase in diseases affecting the pancreas, hepatobiliary tree and bone. Am. J. Med., 31,(196l): 231-237. Kraehenbuhl, J. P., Gloor, E., and Blanc, B.: Resorption intestinale de la ferritine chez deux especes animales aux possibilites d"absorption proteique neonatale differentes. Z. Zellforsch., 76,(l967): 170-186. Leblond, C. P., and Carriere, R.: The effect of growth hormone and thyroxine on the mitotic rate of the intestinal mucosa of the rat. Endocrinology, 56,(l955): 261-266. Leblond, C. P., and Messier, B.: Renewal of chief cells and goblet cells in the small intestine as shown by radioautography after injection of thymidine-H3 into mice. Anat. Rec., l32,(l958): 247-259. Leblond, C. P., and Stevens, C. E.: The constant renewal of the intestinal epithelium in the albino rat. Anat. Rec., 100, (1948): 357-377. Lecce, J. 6.: Absorption of macromolecules by neonatal intestine. Biologia Neonat., 9,(l966): 50-61. Lecce, J. G., and Reep, B. R.: Escherichia coli associated with colostrum-free neonatal pigs raised in isolation. J. exp. 350 Lehninger, A. L.: The1Mitochondrion; Molecular Basis of'Structure and'Function. W. A. Benjamin, Incorporated, New York (1964). Lesher, S., walburg, H. E., Jr., and Sacher, G. A., Jr.: Generation cycle in the duodenal crypt cells of germ-free and conventional mice. Nature, Lond., 202,(1964): 884-886. Levin, R. J.: Techniques, terminology, and parameters in intestinal absorption. Br. med. Bull., 23,(l967): 209-212. Levitan, R., Fordtran, J. S., Burrows, B. A., and Ingelfinger, F. J.: water and salt absorption in the human colon. J. clin. Invest., 4l,(1962): 1754-1759. Lifson, N., and Parsons, D. S.: Support of water absorption by rat jejunum in vitro by glucose in serosal fluid. Proc. Soc. exp. Biol. Med., 95,(l957): 532-534. Lillie, R. D.: Histochemical comparison of the Casella, Bauer, and periodic acid oxidation-Schiff leucofuchsin technics. Stain Technol., 26,(1951): 123-136. Lipkin, M.: Cell proliferation in the gastrointestinal tract of man. Fedn Proc. Fedn Am. Socs exp. Biol., 24,(1965): 10-15. Losos, G. J.: Pathogenesis of enteric infection of swine. M.Sc; Thesis, University of Toronto (1964). Lovell, R.: Classification of Bacterium coli from diseased calves. J. Path. Bact., 44,(1937): 125-139. Lovell, R.: Infection and resistance in young animals. Lancet, 2, (1951): 1097-1100. Luft, J.H.: Improvements in epoxy resin embedding methods. J. biOphys. biochem..Cytol., 9,(l961): 409-414. MacDonald, W. C., Trier, J. S., and Everett, N. B.: Cell prolifera- tion and migration in the stomach, duodenum, and rectum of man. Radioautographic studies. Gastroenterology, 46,(1964): .Marchesi, V. T.: Histochemical and biochemical studies of adenosine triphosphatase activity bound to membranes of pinocytotic vesicles. J. Histochem. Cytochem., l3,(l965): 713-714. Maronpot, R. R., and Whitehair, C. K.: A aprue-like villous atrOphy in pigs. Fedn Proc. Fedn Am. Socs exp. Biol., 25,(l966): 752-752. 351 Maronpot, R. R., and Whitehair, C. K.: Experimental sprue-like small intestinal lesions in pigs. Can. J. comp. Med., 31,(1967): 309-316. Matthews, D. M.: Absorption of water-soluble vitamins. Br. med. Bull., 23,(l967): 258-262. Mattisson, A. G. M., and Karlsson, B. W.: Sub-light microsc0pical changes in the epithelial cells of the intestine of the pig- let reared with and without colostrum. Third European Regional Conference on Electron Microscopy, Prague, Czechoslovakia (1964). Mattisson, A. G. M., and Karlsson, B. W.: Observations on structure of intestinal epithelial cells in newborn piglets. J. Ultrastruct. Res., 12,(l965): 243-243. Mattisson, A. G. M., and Karlsson, B. W.: Electron microscopic and immunochemical studies on the small intestine of newborn piglets. Ark. 2001. Series 2, 18,(l966): 575-589. Maunsbach, A. B.: Albumin absorption by renal proximal tubule cells. Nature, Lond., 212,(l966): 546-547. .McBryde, C. N.: Acute enteritis in young pigs due to infection with the colon group. J. Am. vet. med. Ass., 84,(1934): 36-50. .McHardy, G. J. R., and Parsons, D. S.: The absorption of water and salt from the small intestine of the rat. Q. J. exp. Physiol., 42,(l957): 33-48. McKay, D. G., and wahle, G. H.: Epidemic gastroenteritis due to Escherichia coli 011134. II. Pathologic anatomy (with special reference to the presence of the local and generalized Schwartzman phenomenon). Archs Path., 60,(1955): 679-693. MCManus, J. F. A.: Histological and histochemical uses of periodic acid. Stain Technol., 23,(1948): 99-108. McNabb, J. D., and Sandborn, E.: Filaments in the microvillous border of intestinal cells. J. Cell Biol., 22,(l964): 701- 704. Melnyk, C. S., Brancher, R. E., and Kirsner,.J. B.: Colon mucosa response to injury. II. Histochemical study. Gastroenterology, 51,(1966): 50-58. Michel, M: C.: Activitélmetabolique de la flore totale isolee de Z’intestin db porc. Amino acides, Peptides, Proteines, 5, (1962): 157-173. 352 Miller, D., and Crane, R. K.: The digestive function of the epithelium of the small intestine. II. Localization of disaccharide hydrolysis in the isolated brush border portion of intestinal epithelial cells. Biochim. biophys. Acta, 52,(l96l): 293-298. Mitchell, P.: Coupling of phosphorylation to electron and hydrogen transfer by a chemi-osmotic type of mechanism. Nature, Lond., l9l,(l96l): 144-148. Moe, H., Rostgaard, J., and Behnke, 0.: 0n the morphology and origin of virgin lysosomes in the intestinal epithelium of the rat. J. Ultrastruct. Res., 12,(l965): 396-403. Moog, F.: DevelOpmental adaptations of alkaline phosphatase in the small intestine. Fedn Proc. Fedn Am. Socs exp. Biol., 21, (1962): 51-56. Moon, H. W.: The association of Escherichia coli with diarrheal disease of the newborn pig. Ph.D. Thesis, University of .Minnesota, Minneapolis, Minnesota (1965), cited by Moon, Sorenson, and Sautter, 1966. Moon, H. W., Sorenson, D. K., and Sautter, J. H.: Escherichia coli infection of the ligated intestinal 100p of the newborn pig. Am. J. vet. Res., 27,(l966): 1317-1325. Moon, H. W. Sorenson, D. K., Sautter, J. H., and Higbee, J. M.: Association of Escherichia coli with diarrheal disease of the newborn pig. Am. J. vet. Res., 27,(l966): 1007-1011. Merton, R. K.: Acid and alkaline phosphatases. In Biochemist's Handbook, C. Long, ed., D. Van Norstrand Co., Inc., Princeton, N.J., (1961): 249. Mbwry, R. W.: Comparison of technical procedures for histochemical preservation of alkaline phosphatase. Bull. int. Ass. med. Nachlas, M} M;, Crawford, D. T., and Seligman, A. M.: The histo- chemical demonstration of leucine aminOpeptidase. J. Histo- chun. Cytochem., 5,(1957): 264-278. Nachlas, M. M., Monis, B., Rosenblatt, D., and Seligman, A. M.: Improvement in the histochemical localization of leucine aminopeptidase with a new substrate, l-leucyl 4-methoxy-2- naphthylamide. J. biophys. biochem. Cytol., 7,(1960): 261- 264. 353 Nachlas, M. M. Tsou, K. G., Souza, E. D., Cheng, C. S., and Seligman, A. M.: Cytochemical demonstration of succinic dehydrogenase by the use of a new p-nitrOphenyl substituted ditetrazole. J. Histochem. Cytochem., 5,(1957): 420-436. Nachlas, M. M., Walker, D. G., and Seligman, A. M.: A histochemical method for the demonstration of diphOSphOpyridine nucleotide diaphorase. J. biophys. biochem. Cytol., 4,(l958): 29-36. Neter, E., and Shumway, C. N.: Escherichia coli serotype D443: Occurrence in intestinal and reapiratory tracts, cultural characteristics, pathogenicity, sensitivity to antibiotics. Proc. Soc. exp. Biol. Med., 75,(l950): 504-507. Newey, H.: Absorption of carbohydrates. Br. med. Bull., 23,(l967): Nielsen, N. 0.: Studies on edema disease of swine. Ph.D. Thesis, University of Minnesota, Minneapolis, Minnesota (1963), cited by Moon, Sorenson, and Sautter, 1966. Nielsen, N. 0., Moon, H. W., and Roe, W. E.: Enteric colibacillosis in swine. J. Am. vet. med. Ass., 153,(l968): 1590-1606. Novikoff, A. B., and Masek, B.: Survival of lactic dehydrogenase and DPNH-diaphorase activities after formol-calcium fixation. J. Histochem. Cytochem., 6,(l958): 217-217. Ordway, N. K.: Diarrhoeal disease and its control. Bull. World Health Organ., 23,(l960): 73-101. Osborne, J. C.: Escherichia coli serotypes orally. I. Induced gastroenteritis and pathological changes in neonatal calves. Cornell Vet., 57,(1967): 204-217. Osborne, J. C.: Escherichia coli serotypes intravenously. II. Pharmacological effects and tolerance. Cornell Vet., 57,(1967): 217-227. Osborne, J. C.: Escherichia coli serotypes intravenously. III. Pathological changes and pathogenesis. Cornell Vet., 57, (1967): 227-238. Padykula, H. A., and Herman, E.: Factors affecting the activity of adenosine triphosphatase and other phosphatases as measured by histochemical techniques. J. Histochem. Cytochem., 3, (1955a): 161-169. Padykula, H. A., and Herman, E.: The specificity of the histochemi- cal method for adenosine triphosphatase. J. Histochem. Cyto- chem., 3(1955b): 170-195. 354 Padykula, H. A., Strauss, E. W., Ladman, A. J., and Gardner, F. H.: A morphologic and histochemical analysis of the human jejunal epithelium in nontropical Sprue. Gastroenterology, 40,(l961): 735-765. Palade, G. E.: A study of fixation for electron microscOpy. J. exp. Med., 95,(l952): 285-298. Palay, S. L., and Karlin, L. J.: An electron microscopic study of the intestinal villus. II. The pathway of fat absorption. J. biophys. biochem. Cytol., 5,(1959): 373-383. Panish, J. F.: Experimental blind loop steatorrhea. Gastroenterology, 45,(l963): 394-399. Parsons, D. S.: Salt and water absorption by the intestinal tract. Br. med. Bull., 23,(l967): 252-257. Payne, A. M. M., and Cook, G. T.: A specific serological type of Bacterium coli found in infants' home in absence of epidemic diarrhoea. Br. med. J., 2,(1950): 192-195. Payne, L. C., and Marsh, C. L.: Absorption of gamma globulin by the small intestine. Fedn Proc. Fedn Am. Socs exp. Biol., 21, (1962): 909-912. Pearse, A. G. E.: Extension of the limits of cellular pathology: The role of enzyme histochemistry. J. clin. Path., ll,(1958): 520-534. Pearse, A. G. E.: Histochemistry Theoretical and Applied, 2nd edn, Little, Brown and Company, Boston, Mass. (1961a): 882. Pearse, A. G. E.: Histochemistry Theoretical and Applied, 2nd edn, Little, Brown and Company, Boston, Mass. (1961b): 910. Pearse, A. G. E.: Personal communication, Michigan State University, March 1, 1968. Pearse, A. G. E., and Riecken, E. 0.: Histology and cytochemistry of the cells of the small intestine, in relation to absorption. Br. med. Bull., 23,(l967): 217-230. Pease, D. C.: Histological Techniques fbr Electron MicroscOpy. Academic Press, New York (1964). Perch, B.: weitere Untersuchungen uber die Cbliflora des gesunden Menschen. Acta path. microbiol. scand., 21,(1944): 239-247. Peterson, M. R., and Leblond, C. P.: Uptake by the Golgi region of glucose labeled with tritium in the l or 6 position, as an indicator of synthesis of complex carbohydrates. Expl. Cell Res., 34,(l964): 420-423. 355 Phelps, P. C., Rubin, C. E., and Luft, J. H.: Electron microsc0pe techniques for studying absorption of fat in man-with some observations on pinocytosis. Gastroenterology, 46,(1964): Phillips, R. A.: Water and electrolyte losses in cholera. Fedn Proc. Fedn Am. Socs exp. Biol., 23,(l964): 705-712. Phillips, R. A.: Cholera in the perspective of 1966. Ann. intern. Med., 65,(l966): 922-930. Pierce, A. E., and Smith, M. W.: The in vitro transfer of bovine immune lactoglobulins across the intestine of new-born pigs. J. Physiol., 190,(l967): 19-34. Pink, 1. J., Curran, R. G., and Creamer, B.: The ultrastructure of the small intestine under normal and pathological condi- tions. In Methods and.Achievements in Eaperimental Pathology. Volume III, Ultrastructural, Histopathologic and Chemical Approaches. E. Bajusz and G. Jasmin, ed., Year Book Medical Publishers, Inc., Chicago, U.S.A. (1967). Pitman, F. E., and Pitman, J. C.: Electron microscopy of intestinal mucosa. Archs Path., 81,(l966): 398-401. Pitts, R. F.: Physiology of’the Kidney and Body Fluids. Year Book Medical Publishers, Inc., Chicago, U.S.A. (1963). Platt, B. 8., Heard, C. R. C., and Stewart, R. J. C.: The effects of protein-calorie deficiency on the gastrointestinal tract. In The Hole of’the Gastrointestinal Tract in Protein Metabolism: A symposium. H. N. Munro, ed., F. A. Davis Company, Philadelphia, Pa. (1964): 227-238. Porter, K. R.: Independence of fat absorption and pinocytosis. Fedn Proc. Fedn Am. Socs exp. Biol., 28,(l969): 35—40. Public Health Department, London County Council: Castro-enteritis in infancy. Brit. J. prev. soc. Med., 7,(l953): 194-204. Raaflaub, J.: Die schwellung isolieter Lebezellmitochondrien und ihre Physikalische-chemische beeinflusbarkeit. Helv. Physiol. Acta, ll,(l953): 142-156. Reynolds, E. S.: The use of lead citrate at high pH as an electron- opaque stain in electron microscOpy. J. Cell Biol., l7,(1963): 208-212. Rhodin, J. A. C.: An Atlas of’Ultrastructure. The W. B. Saunders Company, Philadelphia (1963). ”03-“ 356 Ribi, E., Anacker, R., Fukushi, K., Haskins, W. T., Landy, M., and Milner, K. C.: Relationship of chemical composition to bio- logical activity. In Bacterial Endotomins. M. Landy and W. Brown, ed., Institute of Microbiology, Rutgers, The State University, New Brunswick, N.J. (1964). Riecken, E. 0., Stewart, J. 5., Booth, C. G., and Pearse, A. G. E.: A histochemical study on the role of lysosomal enzymes in idiOpathic steatorrhoea before and during a gluten-free diet. Gut, 7,(1966): 317-332. Robbins, S. L.: Textbook of'Pathology with Clinical Applications, 2nd edn, W. B. Saunders Co., Philadelphia, Pa. (1962). Robertson, J. D.: The Unit Membrane. In Electron Microscopy in Anatomy. J. D. Boyd, F. R. Johnson, and J. D. Lever, ed., Edward Arnold, London (1961): 74. Robinson, J. R.: Metabolism of intracellular water. Physiol. Rev., 40,(l960): 112-149. Rodriguez-Leiva, M.: Escherichia coli studies. VII. Incidence of enteropathogenic Escherichia coli in the acute diarrhea of infants in Santiago, Chile.. Am. J. Hyg., 72,(1960): 162-168. Roe, W. E., Quamme, G., and Nielsen, N. 0.: University of Saskatchewan, Saskatoon, Sask., unpublished data, 1968, cited by Nielsen, Moon, and Roe (1968). Rosalki, S. B., and Wilkinson, J. H.: Diagnostic Enzymology. Dade Reagents, Miami, Florida (1966). Ross, C. A. G., and Dawes, E. A.: Resistance of the breast-fed infant to gastro-enteritis. Lancet, l,(l954): 994-998. Rubin, C. E., and Dobbins, W. O.: Peroral biOpsy of the small in- testine. A review of its diagnostic usefulness. Gastro- enterology, 49,(1965): 676-697. Rutenberg, A. Mi, Monis, B., and Banks, B. M.: Histochemical distri- bution of hydrolytic enzymes in the alimentary tract of man in health and disease. J. Histochem. Cytochem., 6,(l958): 102-102 0 Rutenburg, A. M., and Seligman, A. M.: The histochemical demonstra- tion of acid phosphatase by a post-incubation coupling technique. J. Histochem. Cytochem., 3,(l955): 455-469. ,g... 357 Samloff, I. M., Davis, J. 8., and Schenk, E. A.: A clinical and histochemical study of celiac disease before and during a gluten-free diet. Gastroenterology, 48,(1965): 155-172. Sanford, P. A.: The inhibition and mechanism of intestinal absorp- tion. Proc. Nutr. Soc., 26,(1967a): 12-17. Sanford, P. A.: Inhibition of intestinal absorption. Br. med. Bull., 23,(l967b): 270-274. Sandritter, W., and Lasch, H. G.: Pathologic aspects of shock. In Methods and.Achievements in Esperimental Pathology. vol. III. Ultrastructural, Histopathologic, and Chemical Approaches. E. Bajusz and G. Jasmin, ed., Year Book Medi- cal Publishers, Inc., Chicago, U.S.A. (1967). Saunders, C. N., Stevens, A. J., Spence, J. B., and Betts, A. 0.: Escherichia coli infection: Reproduction of the disease in "pathogen-free" piglets. Res. vet. Sci., 4,(1963): 347-357. Saunders, C. N., Stevens, J. A., Spence, J. B., and Sojka, W. J.: Escherichia coli infection in piglets. Res. vet. Sci., 1, (1960): 28-35. Saunders, C. N., Stevens, A. J., Spence, J. B., and Sojka, W. J.: Escherichia coli infection: Reproduction of the disease in naturally reared piglets. Res. vet. Sci., 4,(1964): 333- 346. Scarpelli, D. G., and Pearse, A. G. E.: Physical and chemical pro- tection of cell constituents and the precise localization of enzymes. J. Histochem. Cytochem., 6,(l958): 369-375. Sears, H. J., Brownlee, I., and Uchiyama, J. K.: Persistence of individual strains of Escherichia coli in the intestinal tract of man. J. Bact., 59,(l950): 293-301. Senecal, J.: Les diarrhées infantiles en Afrique-Occidentale francaise. Bull. World Health Organ., 21,(1959): 321-336. Senior, J. R.: Intestinal absorption of fats. J. Lipid Res., 5,(1964): 495-521. Seyderhelmen, R., Lehmann, W., and Wichels, P.: Emperimentelle intestinale pernizioze anamie beim Hand. Klin. Wochschr., 3,(1924): 1439-1440. Shanks, R. A., and Studzinski, L. P.: Bacterium coli in infantile diarrhoea. Brit. med. J., 2,(1952): 119-123. 358 Sheehy, T. W., and Floch, M. H.: The Small Intestine its Functions and Diseases. Hoeber Medical Division, Harper and Row, New York (1964). Sherratt, H. S. A., and HUbscher, C.: Properties of mitochondrial preparations from small-intestinal mucosa of the guinea-pig. Biochem. biophys. Acta, 69,(1963): 403-405. Shiner, M.: Feinstrukturelle Untersuchungen am Zottenepithel des menslichen Dunndarms a) Die normale Feinstruktur, b) Fein- strukturelle Veranderungen bei der Einheimischen Sprue (Coliakie, idiopathische steatorrhae). Internist, 7,(1966): Shiner, M.: Ultrastructure of jejunal surface epithelium in untreated idiOpathic steatorrhea. Br. med. Bull., 23,(l967a): 223-225. Shiner, Mt: Personal communication (l967b), cited by Sanford (1967b). Shiner, M.: The structure of the small intestine and some interest- ing relations to its function. Proc. Nutr. Soc., 26,(1967c): 1-4. Shnitka, T. K.: Enzymatic histochemistry of gastrointestinal mucous membrane. Fedn Proc. Fedn Am. Socs exp. Biol., 19,(1960): 897-904. Silhar, A. S., and Maru, A. M.: Morbidity in children. Indian J. Child Hlth., 9,(l960): 507-515. Skou, J. C.: Enzymatic basis for active transport of Na+ and K+ across cell membrane. Physiol. Rev., 45,(l965): 596-617. Smith, D. T., Conant, N. F., and Overman, J. R.: Microbiology. 13th edn, Appleton-Century-Crofts, New York (1964). Smith, H. L., and Etteldorf, J. N.: Parenteral fluid regimens in the treatment of severe diarrhea in infants. J. Pediat., 58, (1961): l-l6. Smith, H. W.: The ecology of the intestinal bacterial flora of the calf with particular reference to Escherichia coli. Vet. Rec., 72,(1960): 1178-1183. Smith, H. W.: Observations on the flora of the alimentary tract of animals and factors affecting its composition. J. Path. Bact., 89,(l965): 95-122. Smith, Smith, Smith, Smith, Smith, Smith, Smith, Smith, Smyth, Smyth, Sojka, 359 H. W., and Crabb, W. E.: The typing of Escherichia coli by bacteriophage, its application to the study of the E. coli pOpulation of the intestinal tract of healthy calves and of calves suffering from white scours. J. gen. Microbiol., 15,(l956): 556-574. H. W., and Halls, 8.: Observations by the ligated intestinal segment and oral inoculation methods on Escherichia coli in- fection in pigs, calves, lambs, and rabbits. J. Path. Bact., 93,(1967): 499-529. H. W., and Halls, S.: Studies on Escherichia coli entero- toxin. J. Path. Bact., 93,(1967): 531-543. H. W., and Jones, J. E. T.: Observations on the alimentary tract and its bacterial flora in healthy and diseased pigs. J. Path. Bact., 86,(l963): 387-412. T.: Studies on pathogenic B. coli from bovine sources. III. Normal and serologically induced resistance to B. coli and its mutant. J. exp. Med., 46,(1927): 141-154. T.: The relation of the capsular substance of B. coli to antibody production. J. exp. Med., 48,(1927): 351-361. T., and Little, R. B.: Studies on pathogenic B. coli from bovine sources. I. The pathogenic action of culture fil- trates. J. exp. Med., 46,(19 7): 123-131. T., and Orcutt, M. L.: The bacteriology of the intestinal tract of young calves with special reference to the early diarrhea ("scours"). J. eXp. Med., 4l,(1925): 89-106. D. H.: Intestinal absorption. Br. med. Bull., 23,(l967): D. H., and Whittam, R.: Membrane transport in relation to intestinal absorption. Br. med. Bull., 23,(l967): 231-235. W. J.: Escherichia coli in Domestic Animals and Poultry. Commonwealth Agriculture Bureaux, Farnham Royal, Bucks, England (1965). Sprinz, H., Kundel, D. W., Dammin, G. J., Horowitz, R. E., Schneider, H., and Formal, S. B.: The response of the germfree guinea pig to oral bacterial challenge with Escherichia coli and Shigella flexneri. Am. J. Path., 39,(196l): 681-695. 360 Staley, T. E.: Clinical implications of normal and abnormal intestinal epithelial maturation in the neonatal pig. Journal of the Oklahoma Veterinary Medical Association, 21,(1969): 10-14. Staley, T. E., Jones, E. W., and Marshall, A. E.: The jejunal absorp- tive cell of the newborn pig. An electron microsc0pic study. Anat. Rec. (In Press), (1968a). Staley, T. E., Jones, E. W., and Corley, L. D.: The fine structure of the duodenal absorptive cell in the newborn pig before and after feeding of colostrum. Am. J. vet. Res., (In Press), (1968b). Staley, T. E., Jones, L. D., and Corley, L. D.: Intestinal monocon- tamination in the neonatal pig -- microbiological and micro- scopic studies. Paper B-Z. Abstracts. symposium on Gnoto- biology: Eaperimental and Clinical Aspects. Statler Hilton Hotel, Buffalo, New York (1968c). Staley, T. E., Jones, E. W., and Corley, L. D.: Intestinal moncon— tamination in the neonatal pig: Microbiological and micro- sc0pic studies. Gnotobiotic Symposium, October 1968 (In Press), (1968d). Steedman, H. F.: Alcian blue 8G8: A new stain for mucin. Q. J1 Microsc. Sci., 91,(1950): 477-479. Stevens, A. J.: A Laboratory Handbook of'Veterinary Diagnostic .Procedures. Animal Health Branch Monograph No. 3, F.A.O., Rome (1961). Stevens, A. J.: Coliform infections in the young pig and a practical approach to the control of enteritis. Vet. Rec., 75,(1963): 1241-1246. Stowens, D.: .Pediatric.Pathology 2nd edn, The Williams and Wilkins Co., Baltimore (1966): 218. Strauss, E. W.: Electron microscopic study of intestinal fat absorp- tion in vitro from mixed micelles containing linolenic acid, monoolein, and bile salt. J. Lipid Res., 7,(1966): 307-323. Takeuchi, A.: Electron microsc0pe studies of experimental Salmonella infection. I. Penetration into the intestinal epithelium by Salmonella typhimurium. Am. J. Path., 50,(1966): 109-136. Takeuchi, A., and Sprinz, H.: Electron microscope studies of experi- mental Salmonella infection in the preconditioned guinea pig. II. Response of the intestinal mucosa to the invasion by Salmonella typhimurium. Am. J. Path., 51,(1967): 137—161. 361 Takeuchi, A., Sprinz, H., LaBrec, E. H., and Formal, S. B.: Experi- An electron microscopic study mental bacillary dysentery. of the response of the intestinal mucosa to bacterial invasion. Am. J. Path., 47,(l965): 1011-1044. The isolation and serological types Taylor, J., and Charter, R. E.: of Bacterium coli in two residential nurseries and their relation to infantile gastro-enteritis. J. Path. Bact., 64,(1952): 715-728. Taylor, J., Powell, B. W., and wright, J.: Infantile diarrhoea and vomiting. Brit. med. J., 2,(1949): 117-125. Teree, T. M;, Mirabal-Font, E., Ortiz, A., and wallace, W. M.: Stool losses and acidosis in diarrheal disease of infancy. Pediatrics, 36,(1965): 704-713. Jejunal epithelium in transmissible gastroenteritis Thompson, S. W., and Hunt, R. D.: selected Histochemical and Histo- (pathological Methods. Charles C. Thomas, Springfield, Illinois, U.S.A. (1966). Thake, D. C.: of swine. serum Enzymes in Diagnosis. Ticktin, H. E., and Trujillo, N. P.: Year Book Medical Publishers, Inc., Chicago (1966). Torres-Pinedo, R., Rivera, C. L., Fernandes, S., and Maldonado, M.: Studies on infant diarrhea. II. Absorption of glucose and net fluxes of water and sodium chloride in a segment of the J. clin. Invest., 45,(l966): 1916-1922. jejunum. Lesions of the small Trapp, A. L., Sanger, V. L., and Stalnaker, E.: intestinal mucosa in transmissible gastroenteritis-infected germfree pigs. Am. J. vet. Res., 27,(l966): 1695-1702. Changes in intestinal and TriantaphleOpoulos, E., and Tuba, J.: serum alkaline phosphatase levels during absorption of cer— tain amino acids. Can. J. Biochem. Physiol., 37,(1959): 711-719. Trier, J. S., Phelps, P. C., and Rubin, C. E.: Electron microscOpy of mucosa of small intestine. J. Am. med. Ass., 183,(1963): 768-774. 8., and Rubin, C. E.: Electron microsc0py of the small Gastroenterology, 49,(1965): 574-603. Trier, J. A review. intestine: 362 Tuba, J., and Dickie, N.: The role of alkaline phosphatase in in- testinal absorption. IV. The effects of various proteins on levels of the enzyme in intestinal mucosa. Can. J. Biochem. Physiol., 33,(1955): 89-92. The response of intestinal alkaline J. biol. Tuba, J., and Robinson, M. I. phOSphatase of fasted rats to forced feeding of fat. Chem., 203,(1953): 947-951. Membrane (contact) digestion. Physiol. Rev., 45, Ugolev, A. M.: (1965): 555-595. A simplified lead citrate stain Venable, J. H., and Coggeshall, R.: for use in electron microsc0py. J. Cell Biol., 25,(l965): J. Am. vet. med. Ass., Vernon, D. F.: Idiopathic sprue in a dog. l40,(l962): 1062-1067. Absorption from the Intestine. Verzar, F., and McDougall, E. J.: Longmans, Green, and Co., London (1936). Visscher, M} B., Varco, R. H., Carr, C. W., Dean, R. B., and Erickson, Sodium ion movement between the intestinal lumen and Am. J. Physiol., 141,(l944): 488-505. Structure histologique D.: the blood. Vodovar, N.: Intestine grele du porc. II. des parois et plus particulierement de la tunique muqueuse en fonction de l'age de l'animal. Annls Biol. anim. Biochem; Biophys., 4,(1964): 113-139. Untersuchungen zur Pathogenese der Voigt, A., and Abadjieff, W.: Kolienterotoxamie des Sehweines unter besonderer Berucksichti- gung der Bedeutung der Resistenz des Makroorganismus und Archs exp. vet. Med., 21,(1967): 155- der.Berdenprophylaxe. 164. Antigenic relationships of Escherichia J. Bact., 45,(1943): 121- wallick, H., and Stuart, C. A.: coli isolated from one individual. 126. Pakistan wasti, S. M.: Diarrhoea, the worst killer of child life. Methods for the use of indican J. biophys biochem. watson, M. L., and Adridge, W. G.: as an electron stain for nucleic acids. Cytol., ll,(l96l): 257-261. watten, R. H., Mbrgan, F. M., Songkhla, Y. N., Vanikiata, B., and Water and electrolyte studies in cholera. Phillips, R. A.: J. clin. Invest., 38,(1959): 1879-1889. 363 Waxler, G. L., Schmidt, D. A., and Whitehair, C. K.: Technique for rearing gnotobiotic pigs. Am. J. vet. Res., 27,(l966): 300- 307. weijers, H. A., Van de Kamer, J. H., Dicke, W. K., and Ijsseling, J.: Diarrhoea caused by deficiency of sugar splitting enzymes. Acta Paediat., 50,(1961): 55-71. Wheeler, W. E.: Fluorescent Antibody Detection of'Ehteropathogenic Escherichia coli. Script of film No. M-1456. U. S. Depart- ment of Health, Education and Welfare, Public Health Service, Chamblee, Georgia (1967). Whitehair, C. K., Grummer, R. H., Phillips, P. H., Bohstedt, G., and McNutt, S. H.: Gastroenteritis in pigs. Cornell Vet., 38,(1948): 23-39. The connexion between active cation Whittam, R., and Ager, M. E.: Biochem. J., 97, transport and metabolism in erythrocytes. (1965): 214-227. Wilson, T. H.: Intestinal Absorption. W. B. Saunders Co., Phila— delphia, Pa. (1962). Wirts, C. W., and Goldstein, F.: Studies of the mechanism of post- gastrectomy steatorrhea. Ann. intern. Med., 58,(1963): 25-36. Wiseman, 6.: Absorption of amino-acids using an in vitro technique. J. Physiol., 120,(1953): 63-72. Academic Press, London Wiseman, G.: .Absorption from the Intestine. (1964). world Health Organization Report: Manual of'the International Sta— tistical Ciassification of'Diseases, Injuries, and Causes of Death; 6th Revision of'the International Lists of'Diseases and'Causes of'Death based on the Recommendations of the 7th .Revision conference 2955: and adopted by the 9th Werd Health Assembly under WZE;0. nomenclature regulations. World Health Organization, Geneva (1957). Escherichia coli 05535 infection in Epidemiological applications of H wright, J., and Boden, A. T.: a gastro—enteritis ward. antigen type determinations. Epidemiological methods used in the study of infantile Yekutiel, P.: Bull. world Health Organ., 21,(1959): diarrhoea in Israel. 374-381. Zamcheck, N}: Dynamic interaction among body nutrition, gut mucosal metabolism and morphology and transport across the mucosa. Fedn Proc. Fedn Am. Socs exp. Biol., l9,(1960): 855-864. 364 Zilversmit, D. B.: The composition and structure of lymph chylomicrons in dog, rat, and man. J. clin. Invest., 44,(1965): 1610-1622. Zweifach, B. W.: Microcirculatory derangements as a basis for the lethal manifestations of experimental shock. Br. J. Anaesth., 30,(1958): 466-484. Zweifach, B. W.: Vascular effects of bacterial endotoxin. In Bac- terial Endotoxins. M. Landy and W. Brown, ed., Institute of Microbiology, Rutgers, The State University, New Brunswick, N.J. (1964). Zweifach, B. W., Nagler, A. L., and Thomas, L.: The role of epi- nephrine in the reactions produced by the endotoxins of gram negative bacteria. J. exp. Med., 104,(1956): 881-896. VITA Bruce Robert CHRISTIE .2255: 22nd. September 1932 at Colac, Victoria, Australia. Father: Richard William Christie Mother: Christina Christie (nee Lemon) Married: Greta Constance Straton 19th. May 1956 at Hobart,.Tasmania. Children: Richard James Iain Christie Bruce Angus Straton Christie Fiona Christine Christie Catriona Clark Christie Education: 1936-40: Meeniyan State School, Victoria 1941-42: Hamilton State School, Victoria 1943-47: Hamilton High School, Victoria 1948-50: Ballarat College, Victoria 1951: Melbourne University, Victoria 1952-56: University of Sydney, New South wales 1964-69: Michigan State UniverSity, Michigan, U.S.A. Academicgggrees: 1957: Bachelor of veterinary Science Faculty of Veterinary Science University of Sydney Sydney, New South wales, Australia 365 1967: 1969: 366 Master of Science Department of Pathology Michigan State University East Lansing, Michigan, U.S.A. Doctor of Philosophy Department of Pathology Michigan State University East Lansing, Michigan, U.S.A. Professional Certification: 1957: Professional Posts: 1957: 1957-61: 1957-61: 1961-64: 1962: 1964-66: 1965-69: Registered Veterinary Surgeon, State of Victoria, Australia Veterinary Officer Livestock Division, Department of Agriculture Melbourne, Victoria District Veterinary Officer Livestock Division, Department of Agriculture Horsham, Victoria Lecturer in Veterinary Science Longerenong Agricultural College Dooen, Victoria Veterinary Research Officer (Cattle Infertility) Cameron Laboratory State Research Farm Werribee, Victoria Representative Technical Committee on Cattle Infertility Launceston, Tasmania Assistant Instructor Department of Pathology College of Veterinary Medicine Michigan State University East Lansing, Michigan Post Graduate Research Fellow Australian Wool Board Michigan State University East Lansing, Michigan 367 1966-69: Senior Graduate Assistant Department of Pathology College of Veterinary Medicine Michigan State University East Lansing, Michigan Professional Societies: Australian Veterinary Association - Member Honorary Secretary Victorian Division 1963-64 Member Federal Council and State Representative 1963 Australian Institute of Agricultural Science - Member Australian and New Zealand Association for the Advancement of Science - Member Australian Society of Animal Production - Member Register of Scientists and Engineers for Commonwealth of Australia - Member Phi Zeta Sigma Xi Research and Publications: Christie, B. R.: Actinobacillosis of Sheep. Vict. vet. Proc., 17, (1958-1959): A22. Christie, B. R.: Trace Elements in Plant and Animal Nutrition - The Animal. Proceedings, Conference of the Grassland Society of Victoria, Melbourne, May, 1961. Christie, B. R.: A.Muscular DystrOphy of Merino Lambs in Western Victoria. Vict. vet. Proc., 20, (1961—62): 30-32. Christie, B. R.: Nutrition and Management of Beef Cattle as Applied to Reproductive Failure under Victorian Conditions. Vict. vet. Proc., 20, (1961-62): 16-16. Christie, B. R.: Ovine Muscular Dystrophy in Victoria. Aust. Christie, B. R.: Hysterectomy Produced Artificially Reared Pigs and Their Possible Use in the Control of Disease. Aust. vet. J., 39, (1963): 148-152. Christie, B. R.: Experimental Colibacillosis in Gnotobiotic Pigs. M. S. Thesis, MHchigan State University, East Lansing, Mich., 1967. Christie, B. R.: The Pathogenesis of Colibacillary Diarrhea of the Newborn. An Electron Microscopic and Histochemical Study of the Disease in Neonatal Gnotobiotic Pigs. Ph.D. Thesis, Michigan State University, East Lansing, Mich., 1969. GI. ”'ifil'fififlj‘lljlflifllfnlflfijflflfllflifljflflfflflllfl