INTESTINAL BACTERIAL POPULATION IN RELATION TO ANTIBIOTIC GROWTH STIMULATION IN POULTRY by William Kent Warden An abstract Submitted to the School of Advanced Graduate Studies at Michigan State University of Agriculture and Applied Science in Partial fulfillment of the Degree of Doctor of Philosophy Department of Poultry Science 1959 Approved ABSTRACT Battery experiments were conducted with one-day-old broiler chicks and poults to study the importance of intestinal bacterial populations in explaining the mechanism(s) by which antibiotics stimulate growth. To practical rations containing high levels of broad and narrow spectrum antibiotics, the following compounds and bacterial preparations were added: (a) ristocetin, (b) polymixin, (c) sulfa drugs, (d) sialic acid, (e) yeast or yeast extract, (f) water-soaked barley, and (g) fresh, dried and autoclaved feces. In addition, birds in some experiments were given weekly or bi-weekly crop inoculations of an £• coli broth or its fractions. Significant responses ( P < .01) occurred from feeding growthstimulating- type antibiotics in most instances. Numbers of intestinal micrococci, enterococci, and lactobacilli were generally depressed in the presence of dietary antibiotics, whereas, the E. coli population varied. Among the E. coli treatments, only lysed cells were effective in significantly (P < .01) improving growth, and growth improvement was similar and non-additive to the response obtained from broad or narrow spectrum antibiotics. Sulfasuxadine had no effect on weight of chicks at four weeks, whereas the significant (P r .01) growth depression which occurred from feeding sulfaguanadine was significantly improved from feeding ristocetin in combination with either zinc bacitracin or terramycin. In three of five experiments, in which zinc bacitracin promoted a highly significant growth response (P <\01) over the basal, the addi­ tion of polymixin resulted in a further significant response (P <.05) over that obtained from zinc bacitracin fed singly. Enzymes of yeast, yeast extract, or those from water-soaked barley were ineffective in mediating antibiotic responses in chicks. Daily feeding of fresh feces from hens receiving an antibioticfree ration significantly depressed (P<(.01) growth. However, growth was restored equivalent to the basal-fed birds by addition of sialic acid and bacitracin, but not by addition of either compound alone. Fresh feces which were dried at 100° F. for 72 hours, or dried and autoclaved under 1^ pounds pressure for *JO minutes, had no effect on chick growth. INTESTINAL BACTERIAL POPULATION IN RELATION TO ANTIBIOTIC GROWTH STIMULATION IN POULTRY By William Kent Warden THESIS Submitted to the School for Advanced Graduate Studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Poultry Science 1959 ProQuest Number: 10008626 All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. uest ProQuest 10008626 Published by ProQuest LLC (2016). Copyright of the Dissertation is held by the Author. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code Microform Edition © ProQuest LLC. ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106- 1346 ACKNOWLEDGEMENTS The author wishes to express his appreciation to Dr. Philip J. Schaible for his personal interest, motivation and constant guidance throughout the experimental investigation. He is also indebted to Dr. W. L. Mallmann for the use of laboratory facilities and for his assist­ ance in contributing much useful information concerning microbiological techniques which were employed throughout the work. The writer is indebted to Dr. Stephen T. Dexter and Dr. Lester F. Wolterink for many helpful suggestions and their constructive review of this thesis. Appreciation is expressed to Dr. Howard C. Zindel and Dr. Theo H. Coleman for their constructive review of this thesis. In addition, sincere appreciation is expressed to his fellow graduate students Charles W. Pope, Robert H. Roberson, Simon T. L. Tsang, James B. Ward and Jerome D. Yates for their willing assistance. Gratitude is also expressed to Commercial Solvents Corporation for a financial grant which helped to support this work, and for their excellent cooperation in providing research materials and for making many constructive suggestions. Finally, the author is indebted, above all, to his wife Ruth for her patience and encouragement during this strenuous period of study and research. TABLE OF CONTENTS Page INTRODUCTION .................................................... 1 REVIEW OF LITERA T U R E ........................................... 5 ......................... Antibiotic mechanism of action Germ-free studies. ............................... Identification of intestinal organisms * • • • • • . ........ Effects of antibiotics on intestinal microflora. • • • • • • • Non-bacterial growth effects from antibiotics. . . . ........ The effect of various microorganism inhibiting agents. • • • • Gut physiology studies • • ............ . . . . . . . . . . . 5 5 6 8 10 11 12 GENERAL EXPERIMENTAL PROCEDURES................................. 15 CROP INOCULATION OF BROTH CULTURES OF E. COLI OR ITS FRACTION. . 18 Experiment I - Experimental procedure Results and Discussion • • • • • • • • • • • 18 19 Experiment II - Experimental procedure.......... Results and Discussion • • • • • • . • • • • 25 26 SULFAGUANADINE AND RISTOCETIN AS INTESTINAL BACTERIOCIDAL' A G E N T S ........................................................ Experiment III 29 - Experimental pro c e d u r e.............. Results and Discussion • • • • • • • • • • • 29 29 VERIFICATION OF ANTIBIOTIC RESPONSE IN SEXES REARED SEPAR A T E L Y ................................................... 55 Experiment IV- Experimental procedure.... . ......... . • • • • Results and Discussion • • • • . • • • • . . THE EFFECT OF ENZYMES FROM WATER-SOAKED BARLEY ................ Experiment V - Experimental procedure • • • • • • ......... Results and Discussion • • • » • • • • • • • POLYMIXIN AND SULFA3UXADINE AS INTESTINAL BACTERIOCIDAL AGENTS . Experiment VI - Experimental procedure .......... Results and Discussion..............• • • • • SYNERGISTIC EFFECTS OF POLYMIXIN ON ANTIBIOTIC RESPONSE........ Experiment VII - Experimental procedure. Results and Discussion. • • • • . . • • • • • • • • • 55 55 59 59 40 45 45 45 47 47 48 Page YEAST AND YEAST EXTRACTS OF SACCAROMYCES CEREVISIAE ............ 54 Experiment VIII - Experimental procedure......................... 5^ Results and Discussion. ...................55 THE EFFECT OF ANTIBIOTICS ON POPULATION OF SEVERAL INTESTINAL MICROORGANISMS........................................59 Experiment IX - Experimental procedure......................... 59 Results and Discussion. . . . . . . 59 Experiment X - Experimental procedure......................... 65 Results and Discussion.• • • • • • • • • • • 65 THE EFFECT OF FEEDING FECAL MATERIAL. Experiment Experiment XI ....................... 70 - Experimental procedure............... • • • • • Results and Discussion..................... 70 71 XII - Experimental procedure......................... ~j6 .................... J6 Results and Discussion. FEED TRANSIT TIME S T U D Y ........................................... 81 Experiment XIII - Experimental procedure............. • • • • • • Results and Discussion. • • • • • • • • • • • 81 82 THE EFFECT OF FEEDING CELLULAR CONTENTS OF S. COLI................. 88 Experiment XIV - Experimental procedure................ Results and Discussion • • • • • • • 88 88 GENERAL DISCUSSION..................................................95 CONCLUSIONS....................................................... 107 REFERENCES CITED. . ........................................... 114 LIST OF TABLES Table 1 Page THE EFFECT OF BI-WEEKLY CROP INOCULATIONS OF E. COLI IN THE PRESENCE OF ANTIBIOTICS IN THE FEED ON POULT GROWTH TO FOUR WEEKS OF A G E ............................. 21 RELATIVE CONCENTRATIONS OF E. COLI IN DIGESTIVE TRACT OF TURKEY POULTS................................... 22 5 FRESH CARCASS ANALYSIS OF FOUR-WEEK-OLD TURKEY POULTS 25 4 ANALYSIS OF VARIANCE OF WEIGHTS OF POULTS AT FOUR WEEKS OF AGE (Experiment I ) ............................. 24 THE EFFECT OF WEEKLY CROP INOCULATIONS OF FRACTIONS OF E. COLI CULTURES IN THE PRESENCE OF ANTIBIOTICS ON FOUR-WEEK CHICK GROWTH AND INTESTINAL E. COLI POPULATION............................................... 27 ANALYSIS OF VARIANCE OF FOUR-WEEK WEIGHTS (Experiment II) • • • • • • . . . . . . . . . 26 THE EFFECT OF SULFAGUANADINE IN THE PRESENCE OF CERTAIN ANTIBIOTICS ON CHICK GROWTH AND INTESTINAL E. COLI POPULATION....................................... 51 2 5 6 7 8 9 10 11 12 15 14 . . ANALYSIS OF VARIANCE OF FOUR-WEEK-OLD CHICKS (Experiment III).......... 52 THE EFFECT OF NARROW-SPECTRUM AND BROAD-SPHCTRUM ANTIBIOTICS ON INTESTINAL E. COLI POPULATION AND WEIGHT OF FOUR-WEEK-OLD COCKERELS AND PULLETS REARED SEPARATELY....................................... 55 ANALYSIS OF VARIANCE OF FOUR-WEEK-OLD COCKEREL CHICKS (Experiment IV). ............................... 56 ANALYSIS OF VARIANCE OF FOUR-WEEK PULLET CHICKS (Experiment IV) • • • • • . . . . . 57 EFFECT OF SPECTRUM OF ANTIBIOTIC ON FOUR-WEEK-OLD CHICK WEIGHT RESPONSE AND INTESTINAL E. COLI POPULATION............................................... 58 THE EFFECT OF WATER-SOAKED BARLEY IN THE PRESENCE OF ZINC BACITRACIN ON FOUR-WEEK-OLD CHICK WEIGHTS AND INTESTINAL POPULATION............................... 41 ANALYSIS OF VARIANCE OF FOUR-WEEK-OLD CHICK WEIGHTS (Experiment V)................ 42 Table 15 the effect of sulfasuxadine in the presence of ANTIBIOTICS ON CHICK GROWTH AND INTESTINAL E. COLIPO P U L A T I O N.......................................... 45 16 ANALYSIS OF VARIANCE OF CHICK WEIGHTS AT 27 DAYS(Experiment VI)...................................... 46 17 THE EFFECT OF A LOW AND HIGH LEVEL OF GROWTH STIMULATING ANTIBIOTIC IN THE PRESENCE OF POLYMIXIN ON CHICK GROWTH AND INTESTINAL S. COLI POPULATION AT FOURW E E K S ............................. 50 18 ANALYSIS OF VARIANCE OF FOUR-WEEK-OLD CHICK WEIGHTS (Experiment VII) • • • • • ............. • • • • • • • • • 51 19 THE EFFECT OF A LOW AND HIGH LEVEL OF GROWTH STIMULATING ANTIBIOTIC IN THE PRESENCE OF POLYMIXIN ON CHICK GROWTH AND INTESTINAL E. COLI POPULATION AT 60 DAYS................................. 52 20 ANALYSIS OF VARIANCE OF EIGHT-WEEK BROILER WEIGHTS (Experiment V I I ) . . . . . . . . . . . ........ • • • • • 55 21 THE EFFECT OF ANTIBIOTIC IN THE PRESENCE OF LIVE AND KILLED YEAST ON CHICK GROWTH AT 26 DAYS................. 56 22 ANALYSIS OF VARIANCE OF 26-DAY CHICK WEIGHTS (Experiment VIII).......................... 57 THE EFFECT OF ANTIBIOTIC TREATMENT ON RELATIVE POPULATION OF INTESTINAL BACTERIA IN FOUR-WEEK CHICKS. . . 58 25 24 THE EFFECT OF POLYMIXIN IN THE PRESENCE OF NARROW AND BROAD SPECTRUM ANTIBIOTICS ON CHICK GROWTH AT FOUR W E E K S .................................................. 61 25 ANALYSIS OF VARIANCE OF FOUR-WEEK CHICK WEIGHTS (Experiment IX)........................ 62 26 THE EFFECT OF POLYMIXIN IN THE PRESENCE OF GROWTH STIMULATING ANTIBIOTICS ON RELATIVE POPULATION OF INTESTINAL BACTERIA IN FOUR-WEEK CHICKS..................... 65 27 RESULTS OF CONFIRMING TESTS TO DETERMINE PURITY OF INTESTINAL S. COLI POPULATION CULTURED BY THE MALLMANN-PEABODY DROP PLATE METHOD ....................... 64 28 THE EFFECT OF POLYMIXIN IN THE PRESENCE OF GROWTH STIMULATING ANTIBIOTICS ON CHICKS AT FOUR WEEKS............. 67 29 ANALYSIS OF VARIANCE OF FOUR-WEEK-OLD CHICK WEIGHTS (Experiment X ) ............................ 68 Table 50 THE EFFECT OF POLYMIXIN IN THE PRESENCE OF GROWTH STIMULATING ANTIBIOTICS ON RELATIVE POPULATION OF SELECTED INTESTINAL BACTERIA IN FOUR-WEEK-OLD C H I C K S ................ 69 51 THE EFFECT OF FRESH HEN FECES ON ANTIBIOTIC RESPONSE IN CHICKS AT TWENTY-SEVEN DAYS..................... 72 52 ANALYSIS OF VARIANCE OF 27-DAY CHICK WEIGHTS (Experiment XI).......... • ...................... 75 ANALYSIS OF VARIANCE OF 27-DAY CHICK FEED EFFICIENCIES (Experiment XI) ............................. 74 THE EFFECT OF DIETARY FECES IN THE PRESENCE OF GROWTH STIMULATING ANTIBIOTICS ON RELATIVE POPULATION OF SELECTED INTESTINAL BACTERIA IN FOUR-WEEK CHICKS ........ 75 55 54 55 THE EFFECT OF FEEDING FECES IN THE PRESENCE OF BROAD AND NARROW SPECTRUM ANTIBIOTICS ON FOUR-WEEK CHICK G R O W T H ................................................78 56 ANALYSIS OF VARIANCE OF FOUR-WEEK CHICK WEIGHTS (Experiment X I I ) ............ 79 57 THE EFFECT OF FRESH, DRIED OR AUTOCLAVED FECES ON RELATIVE POPULATION OF CERTAIN INTESTINAL BACTERIA IN FOUR-WEEK CHICKS......................................... 80 58 THE EFFECT OF DIETARY FECAL CONTAMINATION ON ANTIBIOTIC RESPONSE IN THE PRESENCE OF SIALIC ACID IN FOUR-WEEK-OLD C H I C K S ............................... 84 59 ANALYSIS OF VARIANCE OF FOUR-WESK-OLD CHICK WEIGHTS (Experiment XIII).................................. 85 THE EFFECT OF ZINC BACITRACIN ON RELATIVE POPULATION OF CERTAIN INTESTINAL BACTERIA IN FOUR-WEEK-OLD CHICKS . • 86 40 41 MEASUREMENT OF FEED TRANSIT TIME AND GUT CAPACITY OF ANTIBIOTIC-FED CHICKS AT 50 DAYS OF A G E ................. 87 42 THE EFFECT OF FEEDING LYSED E. COLI AND FRESH FECAL MATERIAL ON ANTIBIOTIC RESPONSE IN FOUR-WEEK-OLD CHICKS* . 90 ANALYSIS OF VARIANCE OF 27-DAY-OLD CHICK WEIGHTS (Experiment XIV) .......................... 91 THE EFFECT OF ANTIBIOTICS, FECAL MATERIAL AND LYSED E. COLI ON RELATIVE POPULATIONS OF CERTAIN INTESTINAL BACTERIA IN FOUR-WEEK-OLD CHICKS ......................... 92 45 44- Table Page 4^atb ASSOCIATION OF ANTIBIOTIC FEEDING AND CERTAIN GRAM-POSITIVE INTESTINAL MICROORGANISM POPULATIONS . . . .102-105 45 EFFECT OF COMBINATIONS OF BACITRACIN AND POLYMIXIN ON FOUR-WEEK CHICK W E I G H T S ....................... 104 47 EFFECT OF VARIOUS ANTIBIOTIC TREATMENTS ON INTESTINAL E. COLI POPULATION AND GROWTH OF FOUR-WEEK-OLDCHICKS. . . 105 43 SUMMARY OF DRUG EFFECT ON BIRD WEIGHT........................ 106 1a EXPERIMENTAL R A T I O N S .........................................109 1b CALCULATED ANALYSIS OF STARTER R A T I O N S ................ . 110 1c COMPOSITION OF M E D I A ................................ 1d LYSED E. COLI CELL PREPARATION.............................. 112 111 1 INTRODUCTION Since the isolation of antibiotics and the recognition that these compounds stimulate early growth of chicks, research investigators have made many attempts to determine the exact mechanism(s) by which they exert their beneficial effect. Among the several theories which have been advanced to explain this action of antibiotics, the concept of quantitative or qualitative shifts in microbial population within the intestinal tract of the animal has received widespread attention. Several investigators including Schottelius (1899)> Balzam (1957)> and Reyniers et al, (19^9) have attempted to raise birds under germ-free conditions in order to study possible relationships between intestinal microflora and growth. While the results of the germ-free approach have not been particularly rewarding, they have made researchers aware of the importance of the part microorganisms may play in eliciting the growth effect from antibiotics. Characteristic chemical and spectral differences between "broadspectrum" (tetracyclines) and "narrow-spectrum" (penicillin and bacitracin) antibiotics may possibly cause differences in their growth stimulating mechanism(s). Since bacitracin is a complex polypeptide molecule that is not absorbed into the blood stream, the primary effect in growth stimulation must result from its action in the digestive tract. Tetracycline- type antibiotics, on the other hand, are absorbed into the blood to some extent and so could act systemically as well as enterically. In addi­ tion, the "broad-spectrum" antibiotics, as represented by the tetracyclines, 2 inhibit both gram-positive and gram-negative organisms whereas, "narrowspectrum” antibiotics, such as bacitracin and penicillin, act more specifically against the general disease-producing gram-positive forms# The research described in this thesis is concerned with the study of the effects of feeding both types of antibiotics, singly, or in com­ bination with various microbial inhibitory agents, bacterial preparations, or enzymes. Particular emphasis has been placed on the changes in Sscherichia coli in relation to observed growth differences among the various treatments# 5 REVIEW OF LITERATURE Antibiotic mechanism of action Despite the well established knowledge, Moore et al, (19^6), Stokstad _et al* (19^9), and Reed and Couch (1950) that antibiotics in­ crease growth rate and enhance feed utilization in poultry, the exact mechanism(s) by which these compounds function to improve performance has not been discovered since their initial use in feeds in 19^6. Gen­ eral concepts which have developed to explain this mechanism include: 1* An effect on the physiology of the animal. 2. Inhibition of harmful organisms in the are competing with the host animal for intestinal tract which nutrients. 5* An increase in beneficial intestinal organisms which aid in increasing assimilation of nutrients. A. An effect on sub-clinical diseases. The concept of meaningful bacteriological changes within the gut from antibiotic feeding has received extensive attention since it was demonstrated early that antibiotics are more growth stimulating when fed to birds in contaminated quarters than those fed in new poultry houses. Coates et al. (1951, 1952), Bird et al. (1952), Hill et al. (1955), Jacobs et al. (1955), and Waibel jet al* (195^)* lation of the gut has been foundto Also the bacterial popu­ change significantly (Price andZolli, 1959) when antibiotics are fed at levels as low as four grams per ton of feed. Germ-free studies Several investigators have studied the effect of microflora on animal performance in an attempt to correlate possible interrelationships between intestinal microbial population and rate of growth. 4 Schottelius (1899 j 1902, 1908, 1915) raised both normal and germfree chicks, but due to lack of essential nutrients in the ration, growth was abnormally slow in the germ-free birds. He concluded erroneously, therefore, that chicks could not be raised in a germ-free environment. Cohendy (1912) also attempted to raise chicks in a germ-free en­ vironment. Lack of knowledge of the complete nutrient requirements of starting chicks resulted in his inability to keep the birds alive for more than five weeks. Balzam (1957 )9 in a study on the effect of intestinal microflora on the vitamin requirement of birds, raised five germ-free chicks for a period of two months. Since growth and freedom from vitamin deficiencies were similar on germ-free and conventionally reared birds, he concluded that the intestinal flora of chickens exerted no measurable effect on feed utilization. Moore _et al. (19^6) reported a growth response from feeding sulfasuxadine, streptothricin, and streptomycin. Since streptomycin appeared to depress the total number of intestinal coliform and micrococci bacteria, including Escherichia coli and enterococci, while at the same time increas­ ing lactobacilli, this raised the question of enteric changes as an ex­ planation for antibiotic action. Experimental evidence presented by Coates et al. (1952) in England, lent strong support to the idea that an "infectious agent" must be pre­ sent in order to obtain growth stimulation in chicks from dietary anti­ biotics. This work encouraged others to initiate studies with antibiotics. While the initial germ-free chick studies conducted by Reyniers et al. (1949) were beset with mechanical and other difficulties, these researchers were able to successfully rear bacteria-free chicks for a 5 l6o-day period on an autoclaved commercial chick starting mash* Un­ fortunately, these workers did not attempt to determine the relative microorganism populations of the gut of the normal test birds, so only limited information was obtained. Subsequent studies by Luckey _et al. (1955) using limited numbers (4 or 5 birds per treatment) of day-old, New Hampshire chicks and Beltsville White poults showed inconsistent results. A feeding level of 46 mg. per kg. of procaine penicillin produced a significant growth response in four-week-old germ-free poults, while lower levels of oxytetracycline and procaine penicillin promoted only slight non-significant growth increases. In further studies along the same lines with germ-free chicks, Gordon ^t al. (1957-5®) attempted to correlate 54 to 57 day cecal bacterial counts in penicillin-fed birds with those in controls. No differences in the microflora population of the cecum were observed, with the possible exception of a decrease in streptococci count in the antibiotic-fed chicks. A comparison between untreated germ-free chicks and antibiotic-fed con­ ventional birds showed similar characteristics; such as reduced weight of small intestine and ileocecal tonsil, unchanged weight of spleen and adrenals, and increased lymphocyte concentrations of the thymus. Thirty- five day weights of germ-free birds were also similar to those of the conventionally reared antibiotic-fed Leghorn chicks. Forbes at al. (195®) recognizing the disadvantages of the limited numbers of birds employed in the Reynier and Luckey studies, re-examined the effect of antibiotics in two separate lots of approximately 25 germfree poults at 14 days of age. Penicillin fed at 45 mg. per kgn. and 6 oleandomycin fed at ^0 mg/kgm. produced significant growth responses in conventionally reared birds. G-ernnfree poults grew as rapidly as the conventionally reared antibiotic-fed birds but did not respond to anti­ biotic treatment. These workers concluded that their experiments sup­ ported the theory that growth response from antibiotics is due to their action on the microflora of the gut. At the First International Conference on Antibiotics in Agricul­ ture, Freerksen (1955) reviewed the numerous speculations concerning mechanisms of antibiotic growth stimulation in domestic animals and poultry up to that time. His a m work in Germany supported the concept that micro­ flora changes in the gut were involved in antibiotic growth stimulation, though not necessarily through increases in total numbers of organisms. Rather, he favored the theory that antibiotics act on certain mutant strains of organisms which produce toxins in the host. He also pointed out exactly contrary relationships where decomposed E. coli organisms are used therapeutically in human medicine. Identification of intestinal organisms Since it is known that bacteria, or their metabolic products, play essential roles in many biochemical reactions, much study has been directed toward the types of bacteria predominating in the normal intestinal tract of animals and poultry. Kern (1897) identified 88 species of intestinal organisms in poultry and found $2 different motile bacilli, 28 various micrococci, 8 sarcinae, as well as 20 species of other bacteria. In an early study of microflora of the chicken*s digestive tract, King (1905) established that 3. coli was the predominating microorganism 7 in the gut, but that it rarely occurred in the duodenal region, Clostridium welchii (or perfringens) was found to occur infrequently. The findings of Gage (1911) suggested that about sixty percent of the intestinal bacteria were of the gram-negative type; however, some difficulty was encountered in interpretation because much of the fiber and other non-bacterial debris stained gram positive. In these studies, which involved birds in dirt floor pens, the predominating organisms were E. coli with a few diplococci (enterococci) present. Gage noted the E. coli always predominated, but appeared to vary with age and changes in environmental conditions. It was noted by Menes and Rochen ( 19^) that the same species of microorganisms flourished throughout the intestinal tract, but that quantitative differences occurred depending upon the site. Their expla­ nation for this apparent uniformity of population was based on the likeli­ hood of lactic acid production, with its associated reduction of putre­ factive organisms. They considered the lactic acid flora to be Streptococcus faecalis, E. coli, and Lactobacillus piantarum. Emmel (1930) confirmed King1s observation that the coliform type of organism (E. coli, Aerobacter aerogenes) constituted about 65 percent of the bacterial population of the intestinal tract of chickens. Twenty of the chickens which he had under observation showed a decrease in in­ testinal E. coli while suffering from an enteritis outbreak. More recently Schumacher and Hauser (19^1) and Yacowitz and Bird (1933) Lave verified the predominance of E. coli as the principal organism in the intestinal tract of poultry. In addition, evidence has been pre­ sented by Couch et al. (19^*8), Driesens (1931)* and Sieburth et al. (1952) 8 that coliform bacteria have the ability to synthesize folic acid, ribo­ flavin, biotin and niacin as well as other nutritional factors. The chick's requirement for protein was reported by Machlin _et al. (1952) to be 19 percent with a diet containing chlortetracycline and 21 percent if the antibiotic were omitted. Weakley et al. (1955) for chick growth. Further studies along these lines led conclude that bacitracin also spared protein Groschke and Evans (1950) observed a sparing effect on water-soluble vitamins in chicks fed chlortetracycline while a vitaminsparing effect from antibiotics was also shown for riboflavin, niacin and folic acid by Biely and March (1951)* Insofar as fat-soluble vitamins are concerned, Burgess et al. (1951) noted improved utilization of vita­ min A and carotene due to penicillin feeding and investigations by Ross and Yacowitz (195^0 showed a decreased vitamin D requirement for normal bone deposition in the presence of dietary penicillin. Effects of antibiotics on intestinal microflora Since growth responses appeared to be related to the presence of particular organisms, Johansson £t al. (19^8 ) and Johansson and Sarles (1949) initiated studies of microflora changes in the presence of baci­ tracin and penicillin. They showed that the ceca are an important site for the synthesis of B vitamins and probably other unknown nutritional entities. Rhodes _et al. (195^0 showed that coliform organisms accumulated in the ceca of antibiotic-fed birds. However, Dixon and Thayer (1951) demonstrated that growth responses could occur with penicillin and aureomycin in chicks whose ceca had been removed. The early work of Groschke and Evans (loc. cit.) suggested that antibiotic growth responses were mediated by changes in intestinal 9 microflora which reduced competition from unfavorable bacteria. Sieburth et al. ( 1951) observed a reduction in cecal Clostridia in turkey poults from the feeding of 100 grams per ton of penicillin. The total cecal enterococci, however, appeared to be quite variable. March and Biely (1952) and Slam e-t al. (1955) suggested that anti­ biotic growth responses may be due to decreasing the incidence of enterotoxemia in birds — primarily by reducing the number of Clostridia and lactobacillus-type organisms in the gut. Romoser e*t _al. (1952a) found that A. aerogenes increased sharply upon penicillin feeding. This led to further work (1952b, 1955) i^ which lyophilized E. coli and A. aerogenes were fed to chicks in the presence and absence of penicillin. Addition of these cultures increased the re­ sponse to the antibiotic 64- to 84 percent; however, no measurable response occurred from the feeding of the bacterial cultures in the absence of penicillin. Work along the same lines was conducted by Anderson et al. ( 1952a,b; 1955^,b) in which normal- and abnormal-appearing strains of E. coli were isolated from cecal contents and fed to chicks. The killed coliform cells were without effect, but the live cells and their broth filtrate gave growth responses similar to penicillin in chicks. Poults, on the other hand, responded to penicillin and live cells, but not to the fil­ trate. Addition of micrococci depressed growth while anaerobic bacteria did not. Further studies with chicks by Anderson et al. (1958) related growth responses from chlortetracycline to its effect in reducing organisms of the enterococci-type. Feeding of ten percent (volume/weight) of a high 10 concent.rat.ion (2 8 x 10^ cells) of a mixed enterococcus culture caused a highly significant depression in growth. Addition of 4o grams per ton of chlortetracycline reduced the numbers of enterococci and lactobacilli and tended to maintain the coliform count at approximately that of the control group, while weight was restored equivalent to the control birds* Bogdonoff et al. (1957) obtained an added growth effect in broilers frombi-weekly crop inoculations of mixed cultures of S. coli and A. aerogenes in the presence of the 200 gram per ton level of several antibiotics. se had Addition of approximately 6 .8 x 10 12 coliform cells per ml. per no effect on growth or feed efficiency. Using pooled samples of the entire digestive tract of chickens, Price and Zolli (1959) were able to show significant increases in coliform populations in the presence of four grams per ton of oleandomycin. These microorganisms counts correlated positively with increased growth at nine weeks, but not at five weeks. Pooled samples of proteus, micrococcus, lactobacilli, enterococci, yeasts, molds or clostridia, as well as total aerobic and anaerobic populations from all segments of the tract were not significantly changed due to antibiotic feeding. Non-bacterial growth effects from antibiotics Whereas, the antimicrobial concept as a mechanism of antibiotic action has received considerable support, some evidence points to addi­ tional modes of action. For example, Stokstad and Jukes (1950) observed that alkali-treated chlortetracycline residues contained growth-stimulat­ ing properties but were inactive from a microbiological standpoint. Ad­ ditional work by Elam _et ai. (1951) lent further proof that more than one mechanism of growth stimulation by antibiotics might be operating. 11 They observed that autoclaved penicillin contained no measurable anti­ microbial effect and did not appear to alter intestinal microflora, yet stimulated growth when fed or injected. The effect of various microorganism inhibiting agents Since the establishment by Moore at al. (19^-6) of tolerance levels of sulfasuxadine and the observation that this drug reduced coliform population in the ceca of the chick, interest has developed in using bacterial inhibiting compounds to study microflora changes in the gut, or vitro inhibition of disease-producing microorganisms. Peppier et al. (1950) 1*0 and 0.5 percent polymixin D in a charcoal adsorbate to chicks and noted a marked depression in coliform population which persisted for several days following removal of the compound. Streptococcus faecalis appeared to be the most resistant type of bacteria to the drug while lactic-acid-producing bacteria were unaffected. Ristocetin, a new antibiotic possessing activity against grampositive bacteria and mycobacteria, but not against the gram-negative species, was studied extensively in vitro by Grundy at al. (195&“57)* Ristocetin B proved three to four times as active in vitro as ristocetin A against several disease-producing gram positive forms and resistance to the drug was not readily acquired. Since doses of greater than 100 ug. per ml. were required to inhibit El. coli, this organism remained relatively unaffected in therapeutic studies. Intravenous or intramus­ cular doses in rabbits gave measurable blood levels for at least eight hours, while infections of Streptococcus pyogenes, Staphylococcus aureus and Diplococcus pneumoniae in mice were readily controlled. 12 Goebel and Barry (1957)* in investigating the biochemical proper­ ties of E. coli (strain K-255) reported that in a culture medium this organism elaborated colominic acid which had properties similar in many respects to those of sialic acid, Goebel reported that sialic acid has interested medical scientists in recent years because of its effect on viruses. When it is combined in its native state with certain proteins and sugars, the sialic-acid-containing complex interferes with the adher­ ence of certain viruses, such as the influenza virus, to living cells. In a study of the effect of coprophagy and refection, Barnes et al, (1959) observed a sparing effect of several vitamins when rats were fed a slowly digested carbohydrate such as dextrin. Closely associated with this observation was the slight growth increase noted upon thiamin supplementation and a further marked increase in growth when penicillin was added to a dextrin-type ration. However, when coprophagy was pre­ vented neither thiamin nor penicillin promoted increased weight. This suggested possible involvement of fecal entities in sparing essential nutrients, Mameesh et slI. with the whole diet. (1959) 0 ,1 and 1 .0 percent raw hen feces mixed In the presence of both levels of fecal contamina­ tion he observed a consistent growth response in four-week chicks to fifty ppm. of terramycin. Penicillin at the same level, on the other hand, gave a growth response in only one of six trials in the presence of fecal contamination. Gut physiology studies Several investigators have attempted to correlate increased feed consumption, decreased thickness of gut wall and differences in feed transit 13 time with antibiotic responses in poultry as a possible mechanism of growth stimulation. Whereas, increased feed consumption does occur due to antibiotic feeding, the added percentage of feed consumed does not, in most instances, equal the magnitude of response observed during the early period of growth stimulation. Measurements of intestinal walls have shown reduced thickness and lower overall weight of the cleaned intestine in antibiotic fed birds as compared to birds receiving the control ration* In addition, tracer studies have indicated that anti­ biotics may have an effect on the amount of time required for passage of feed through the alimentary tract; however, the meaning of these changes has not been established. Gordon £t al. (1957-58) showed a highly significant decrease in small intestine weight of chicks at 57 days (P<0.01) from feeding fifty mg. per kgm. of penicillin. In the absence of antibiotics, germ-free chicks also showed a highly significant decreased weight of small in­ testine compared to conventionally reared birds. Several investigators (Busse, 1952; Busse and Spiess, 1952; Schaumann et al. 1952) have shown that many drugs can affect the motility of the intestinal tract — extent. in most cases inhibiting peristalisis to some However, Nakatsuka et al. (1952) found that penicillin or chlor- tetracycline in concentrations as low as 1:10,000 to 1:5*000 stimulated contractions of isolated frog intestine. WoIterink et al. (1958) ob­ served a reduction in transit time of radioactive P ^ from the crop to the intestine due to the feeding of reserpine plus dienestrol diacetate to White Rock cockerels. However, a comparison of treated with control birds showed that treatment with the drugs resulted in a 44 percent faster absorption of P^ from the duodenal area. 14 Hillerman et al. (1955)» in a series of tests involving young and mature turkeys and chickens, showed an increase in average transit time for feed in the presence of fifty ppm of dietary penicillin. A capsule containing two-tenths gram of ferric oxide per bird was found to be an acceptable tracer for providing good coloration without affecting the consistency of the feces. Under the conditions of these experiments, average passage time was two hours, 57 minutes for birds receiving the control ration and three hours, 15 minutes for the antibiotic-fed birds. On the other hand, Jukes _et al_. (195^) using carmine or chromic oxide in the feed as tracers found significant decreases in time of feed transit ranging from 10 to 55 minutes from feeding ten ppm of penicillin or aureomycin. In these cage experiments, both heavy and light breed chicks were tested with observations being made at five-minute inter­ vals for appearance of the marker in the feces. Further studies along the same lines led Jukes et al, (195^) 'to conclude that average weight of the intestinal tract was less for antibiotic-fed chicks than for controls. Microscopic examination of serial sections of the duodenum of birds receiving ten ppm of penicillin re­ vealed, on the average, a significantly smaller diameter, thinner tunica propria layer, and shorter villi than those of controls. 15 General Experimental Procedures Similar procedures were employed for each of the fourteen experi­ ments outlined in this section, except where discussed under an individual trial. Randomization of chicks to replicate pens was made as follows: one-day-old Cobb's strain of White Rock broiler cockerels, as well as White Leghorn De Kalb chicks or turkey poults hatched at Michigan State University were individually weighed and distributed into consecutive weight ranges, such as 59-41 grams, etc., with each weight range having a three-gram spread. Birds from highest and lowest weight ranges were discarded and an equal number of birds from all remaining weight ranges were allotted to each battery pen. This method was employed to minimize any possible effect that initial weight differences might have on performance* Each experiment involved 56 or 4o birds per treatment. Lots were randomly assigned to replicate pens, except that antibiotic-free treatments were placed in top pens to reduce likelihood of these birds receiving antibiotic. This was felt to be necessary, since the bacteriological tests conducted on intestinal microflora could be markedly influenced by relatively small quantities of these compounds sifting into pens below. One replicate of each treatment was assigned to a fluorescent-lighted deck level in each of four electrically-heated, starting battery brooders having raised wire floors. This procedure was followed in order to equalize between lots any possible effects due to position of batteries in the room. In experiments conducted beyond four weeks, birds were transferred to finishing batteries using the same method of placement to avoid antibiotic sifting into the basal feed. 16 Batteries were thoroughly cleaned between experiments and tempera­ ture regulated at 96~98° F. at least 24 hours before arrival of the birds. Temperature was reduced 5° day when practical. weekly and discontinued at the fourteenth Feed and water were supplied _ad libitum throughout the experiment and feed was also placed on paper on the floor of all pens for the first three days. Group weights were recorded at the end of the second week and in­ dividual weights were obtained for statistical analysis at the end of the fourth or eighth week. Analysis of variance and Duncan's (1955) multiple range and multiple F tests were employed to determine differences between lots. Feed consumption was recorded for each replicate of each treatment and used to calculate feed efficiency based on weight gains. Chicks on all experiments received a practical broiler-starter mash, or starter and then broiler-finishing mash from the sixth week to the eighth week. Turkeys received a practical turkey starting mash for the first four or seven weeks. Composition and calculated analysis of these all-mash rations are shown in tables 1a and 2a, respectively. Representative birds were sacrificed from each treatment for bacteriological examination 48 to 72 hours after final weights were ob­ tained. Contents of the digestive tracts were removed aseptically and placed in physiological buffer solution at 55° F. without shaking with­ in a few minutes after slaughter. The Mallmann-Peabody (1957) drop plate method for determining intestinal E. coli was employed in each test. Each area of the digestive tract in experiment 1 was studied for microorganism population and an area two inches above the cecum was studied in all other experiments. 17 When studied, total gram negative intestinal population was de­ termined by culturing on laurel sulfate agar. Azide agar was used to obtain total gram positive intestinal population. Dextrose azide broth and eosin violet azide (E.Y.A.) broth were used to determine total micro­ cocci and enterococci intestinal populations, respectively. Trypsin digest agar was used as media in determining intestinal lactobacilli population. Procedures for preparation of these various media and culturing techniques for the organisms are given in the Difco Manual.* Composi­ tion of the media used in culturing the various microflora are shown in table 1c. Differentiation tests, as found in Standard Methods (1955) were employed to verify the purity of E. coli as coliform organisms cultured by the Mallmann-Peabody method. A table of most probable numbers pro­ vided statistical basis for total micrococci and enterococci intestinal populations. *Difco Manual of dehydrated culture media and reagents for microbiological and clinical laboratory procedures. Difco Laboratories, Inc., Detroit 1, Michigan. 18 EXPERIMENT I The possibility of an association of intestinal E. coli population with bacitracin growth stimulation in poultry prompted an initial experi­ ment, Its purpose was to determine if crop inoculations with a broth culture of E. coli would affect performance in the presence of broad or narrow spectrum antibiotics. Sroad spectrum antibiotics are known to in­ hibit intestinal coliform population, whereas narrow spectrum antibiotics do not. The effects of oral inoculation with these organisms was studied with respect to growth, feed utilization, carcass analysis, distensibility of intestinal wall and digestive tract E* coli population. Four replicate lots of one-day-old Broad 3reasted Bronze turkey poults were fed the supplemented all-mash starter ration for a 28-day test period. (Table 1) Equal numbers of males and females were reared together and fed 200 grams of zinc bacitracin or terramycin per ton of feed continuously throughout the test period. In addition, trypticase soy broth or a culture of E. coli in trypticase soy broth was administered directly into the crop of the birds by automatic syringe at bi-weekly in­ tervals. The broth culture of E. coli was adjusted on the basis of light transmission as follows: (1) One loopful of a stock culture* of E. coli was transferred to ten ml. of sterile trypticase soy broth and incubated at 57° G. for 18 hours. *Stock cultures of E. coli provided through courtesy of Commercial Solvents Corp., Terre Haute, Indiana. 19 (2) By use of a photoelectric meter, the cultured medium was adjusted from approximately 10 to 25 percent light trans­ mission by addition of sterile broth, (5) 0.2 ml. of the 18-hour culture was added to each 100 ml. of soy broth and incubated, with shaking, for an additional 18 hours at 57° 0. to produce approximately 6.8 X 10 12 organisms per ml. The bi-weekly dosage of broth or E. coli in broth was increased from one ml. per bird the first week (starting at three days of age) to four ml. at the fourth week — by one ml. increments. The birds were maintained on the experimental rations and moved to finisher batteries at the end of the fourth week. A determination of differences in gut distensibility due to treatment was made at the end of the seventh week. Ten cm. sections of intestine immediately anterior to the ceca were suspended in a 57° C. water bath and perfused by manometer with 57° C. water containing a red vegetable dye. Results and Discussion The results are shown in tables 1, 2, 5> and An analysis of variance (table 4) revealed that significant growth responses occurred with the addition of either broad or narrow spectrum antibiotics. Crop inoculation of E. coli was ineffective in mediating or affecting growth response to either type of antibiotic. A preliminary test of the E. coli population of the various areas of the digestive tract revealed no marked differences between treatments, based on ex­ amination of a single representative poult per lot. While no specific pattern of E. coli population was evident, there was a general increase 20 Hi* nearest the cloaca for all treatments with very few E. coli noted in the crop or duodenum* Carcass analysis of homogenous samples of dressed eviscerated birds revealed no significant differences in protein, water, or ether extract due to the various treatments. Slight differences in distensibility of gut were noted, with antibiotic-fed birds having somewhat greater water-holding capacity at a given water pressure. These differences, while consistent, did not ap­ proach significance even at the five percent level of probability due to the limited numbers of birds involved. 21 c *rf THE EFFECT OF BI-WEEKLY CROP INOCULATIONS OF E. COLI IN THE PRESENCE IN THE FEED ON POULT GROWTH TO FOUR WEEKS OF AGE OF ANTIBIOTICS cd O • CM CM O * CM o- -=J- CM mo O- INCO CO IS 00 N 15 t*cd O P co co co co co CO m • © +3 to X * -S. £ p 10 JDH4 E b < o -P fl a cd * ON VO VO VO CO CO NO VO VO VO CO N MO -H H ■ft, ^3 W O P p O O H Q* u ft • m w O o 3 (0 S © •H d© c 8 fl)»l-l 15 8 tf O tu m CM £ •H •d O © £> o C •H CQ o C •H O O CM CO O o CM *=! •rl £ s cd i •* H © C •HI & s cd £ © P o o P o o VO vO CM CM 22 © Sfl O O O«b ON O O- O O o «k o UN CN O o o • CO o rH Q 3 O J23 £ O 8 -P Ctf O s «fc UN VO o o CO * ON O O On «* CM CM U he Sh O O O SJ -ch o o * R rH s* CM ft CM O INOn CM Ov s s G O 523 rH * * rH O o ft 2 S3 6 0 O 00 123 ft un O § S3 ft UN S3 * co ft ft o S 0 O » W CM UN VO £ © ^ A is 8 0 G ft g S £8 to w U 0 0 0 -p e b.O G M O © O P CM 25 a O '-p u P-. ON 00 VO rv o \r\ o IN* O n ON rH NO O NO v O ON O Ov CM rH 00 UN unun CM CM -3* rH rH CM rH CM IN CN • • • • CM H • • © u p p p “" © ^ -H ^ O S ON ON O • • IN c n • • IN- IN • • © © o © §P •H u > © © * 24 TABLE 4 ANALYSIS OF VARIANCE OF WEIGHTS OF POULTS AT FOUR WEEKS CF AGE. Experiment I Source of variation Degrees of freedom Total F values Calculated P = 0.05 Mean square P = 0, 211 23 24,961 2.29 1 .9 1 Lot 5 83,894 7*71** 3.14 Replicate 3 12,639 1.16 2.67 15 7,780 0.72 2.10 165 10,877 Subclass L X R Error Comparison among lots at 1 percent level of probability Lot Wt. in grams 2 1 5 6 3 ^ 573 594 665 678 678 683 ** Significant (P ( .01} 25 EXPERIMENT II In view of “the negative findings with whole broth cultures in the previous experiment, a second test was conducted to determine if a frac­ tion of an E. coli broth culture might alter the growth pattern of poultry fed a broad or narrow spectrum antibiotic. A broth culture similar to that employed in the first trial was fractionated as follows: The ^>6^hour incubated culture was centrifuged at 36OO r.p.m. for 45 minutes and the supernatant saved. The centrifuged Ei. coli cells were then triple-washed by centrifugation in physiological saline and combined with a volume of saline equivalent to the original broth volume. Twelve lots of one-day-old Cobb's White Rock chicks were allotted as previously described into four replicates of ten chicks each. Equal numbers of cockerel and pullet chicks per pen were fed the supplemented chick starter ration (Table 1a) continuously for a 28-day period (Table 5)* Crop inoculations of trypticase soy broth, E. coli in broth, super­ natant of an E. coli culture, or centrifuged E. coli cells in saline were administered weekly. Dosage was increased from one ml* per bird the first week (starting at three days of age) week — to four ml. at the fourth in one ml. per week increments. One representative cockerel chick was selected from each lot for determination of the E. coli population. Limited numbers of E. coli occur in the anterior regions of the digestive tract, due probably to the high acidity. On the other hand, the excessively high numbers of E. coli ordinarily observed in the cecum make interpretation of this 26 population of questionable value. Therefore, E. coli population was determined in the area approximately two inches above the cecum, since it was felt that this location would represent an area which would give meaningful differences in population, if E. coli were involved in growth stimulation from antibiotics. Results and Discussion Effects of treatments on intestinal E. coli population and bird performance are shown in tables 5 and 6Weekly inoculations into the crop of chicks of the several frac­ tions of E. coli culture were ineffective in changing growth rate in the presence of either zinc bacitracin or terramycin. Only zinc bacitracin treatments promoted a significant growth response (Table 6) over the controls and no relationship between improved growth and E. coli popula­ tion was evident. In addition, none of the various coliform-crop inocu­ lation treatments tempered or increased antibiotic responses. The fact that the terramycin-fed birds did not grow significantly faster than the controls did not appear to be associated with intestinal E. coli, since microorganism population did not vary markedly between treatments. It is interesting to note that despite the lack of significant growth response from the broad spectrum type, feed utilization was im­ proved from feeding either antibiotic. Furthermore, crop inoculation of the live coliforms in broth or coliform cells was associated with maximum improvements in feed conversion in the presence of bacitracin, but not when terramycin was fed. This effect may be coincidental, or may indi­ cate associative effects between E. coli and zinc bacitracin. X O •rl 10 rH © ,£3 27 XI •H -P 0 O rH O g O C ft-H • M O ftC M © > o O « • o 1 • bO H O O U30 • * • H H O C O • O- 5 CM CM XTWO H H O On • • ♦ • CM CM CM H ft hO O O 3 25 VACM CO on 00 On rH rH rH rH H H rH Xj © Q © •h .5 ■£ X3 rH Q CO CJN Cb-CT C°v £>- On ON 00 • d - Q \ ^ t iH NrlOON C ^O ^k ON ON ON (*>C^kc*b © © © ©^ M -P JS S r , * ! ,s2 © e s a $ ■ * bO m > XI 0 -p W ^ G G *H •H O O © © 5Ht G P *H O o © © X> X> © © © © c S f o C o o ol o o o M O NO O 5 5 . 5 TJ G G G G G G © © © +J P - p © © £ © rH © © TJ rH 5-1 G © 43 O O O O O O O Q S25 !SJ S3 }2! CM CM CM CM CM CM CM CM H VPy\D O - CO ON O H CM H H H CM C n -3- . . G *H rH O O © G £>-nO X3 P * O O O O .5.5 « ta ts o Xl G *H O © G a ^ •H -H -p hJO -P XJ ■• &a CM © -p G rQ b° hOV-i -H G © ICO •rl * G G H © O O £ o O *0 *H © P © ft © © © o •H P a 8 j f f l ♦H G © G .- P rj C . C H G e-» * Or © © «aJOO * * * * * * * * * 28 TABLE 6 ANALYSIS OF VARIANCE OF FOUR WEEK WEIGHTS. Source Degrees of freedom variation >tal Mean square Experiment II F values Calculated P — 0.05 P = 0.01 465 Subclass 33 5*542 2.08** 1.74 Lot 11 22,058 3.98** 3.64 Replicate 3 9,363 Sex 1 333.273 S X L 11 1,866 S X R 3 2,093 33 2,317 370 2,653 S X R X L Error 1.69 6.70 125.62** Comparison among lots at 5 percent level of probability Lot A wk. wt. 4 3 2 12 1 9 11 10 411 414 419 421 425 441 449 452 7 8 6 5 460 469 471 477 at 1 percent level of probability Lot 4 wk. wt. 4 411 3 414 ** Significant (P <[0.0l) 2 12 1 419 421 425 9 1 441 1 449 10 7 452 460 8 469 6 5 471 477 29 EXPERIMENT III Since the findings of trials I and II did not support the concept • C° H involvement per se in mediating responses to broad or narrow spectrum antibiotics, a third trial was conducted. In this test an attempt was made to reduce the numbers of intestinal microorganisms to a minimum. It was theorized that, if an equivalent antibiotic response occurred in the presence and absence of high E. coli population, the likelihood of E. coli contributing to the antibiotic mechanism of action could be discounted. Accordingly, nine lots of Gobb*s V/hite Rock one-day-old cockerels were allotted, as previously described, into four replicate pens of eight chicks each. Antibiotic, sulfaguanadine, and/or ristocetin were added to the broiler basal diet (Table 1a) as shown on the experimental design (Table 7) and fed for a 28-day period. Sulfaguanadine is effective in reducing gram negative forms, such as E. coli, while in vitro and in vivo work with ristocetin by Grundy et al. (1956-57) showed this com­ pound to be highly effective in reducing the common disease-producing gram—positive organisms* One representative bird was selected from each experimental lot at the fourth week for determination of intestinal E. coli population. Results and Discussion Neither zinc bacitracin nor terramycin produced a significant improvement in weight gains (Table 8). While there was a significant growth-depressing effect due to the presence of sulfaguanadine in the ration (Table 7), addition of either antibiotic or ristocetin partially 50 compensated for this though it did not entirely restore performance equivalent to the basal group. Ristocetin plus sulfaguanadine promoted a significant growth in­ crease over nsulfa” alone, which was not further improved by addition of either type of growth-stimulating antibiotic. This suggested that ristocetin may have exerted a maximum depressing effect on harmful grampositive organisms which were competing with or inhibiting the host in some way, and that bacitracin or terramycin did not further reduce these detrimental microorganisms. It was noted that bacitracin restored the intestinal E. coli population more effectively than did terramycin in the presence of the E. coli-depreasing sulfa drug and that the increased population was as­ sociated with significantly greater growth. The observation that terramycin per se as compared to bacitracin effected a reduction in E. coli population without causing a growth de­ pression, suggested possible differences in mode of action of broad and narrow spectrum antibiotics in mediating growth. It was an interesting observation that feed utilization was markedly depressed by addition of the sulfa drug, and that under these conditions zinc bacitracin was more effective in improving efficiency of feed utili­ zation than was terramycin. Results, insofar as intestinal E. coli population was concerned, were not clear cut in this particular trial. However, it appeared that sulfaguanadine acted to inhibit intestinal coliforms and that ristocetin counteracted this depressing sulfa effect directly or indirectly, and thereby permitted increases in E. coli bacteria which were associated with increased growth. 51 to © © o o o o ft *rl • pq f t c\i © • • CO * A CA VO * A O• O• VA • 00 rH >A O RI ^ 6 CM 525 e-t CM P © •rl P CA 00 © O Sh rH OT -TO O -rH ft G I -P ■83 © © h£) 10 44 © CO O - O • • • rl rl rl O - CM CO O 00 O n • • • W H H h .g O n 00 • * rH rH E S © X> •rl TS <4 O P to © bfl © © CM © > ca p •H P f-i X H Vt «) O Ov co -5j\ rH O- Q> CA CM CM O— rH rH C\2 CA CA •H O © O O O •8 G rH CA CA CA © P 'i P to © Jg 44 St ©’44 © G 0-=J- E SP § rH -3- U') © *rl h£> > .EJ P W ■4 o ta CACM * A rH VO -=fC A C A CA CA O n VO O CA CA © •rl P O •rl •rl o •0 o o -P o o Sos -a5 ■a © s? © rl P © c •H 'O © G -3 © © 2 § o ,o 0 *H bp P <3 *h P 0 © O 00 A -P © © © e o so so S SB 55 »A iA »A • • • o o o VA A « • O » 3 P P S © o tH 'b -p ^ H c i —i *A O "© •H TJ S W as V PG§ § c C © © g 4> G •rl P © O G © as •rl u “3 W r~> © © c 13 -O o O p •rl © P ,£ © •H p E hO O P -4 © •rl P o •rl rO •rl P p G P •H © © © t>a ,© © § G o £ g •H © ISJ p © C O o o o o 3 CM CM rH CM CA © l>s § © G G G •rl © fr3 P G O O o o o S CM CM -d- VAVO G + G P •H G •H G P P *H P © © © © >> o P © © t © X> O g O © P Cu P o © © G © p G •rH g •rl © •rl G © G * •rl p G O OS p O o OV o o CM 0-00 CO p » S o G •rl © © G •rl 8 OV © © P G o p © o o 0 p © 1 © o ■8 8 p cO p p | o o E-* o ss; E-t * * 32 TABLE 8 ANALYSIS OF VARIANCE CF WEIGHTS OF FOUR-WEEK-OLD CHICKS Experiment III Source of variation Degrees of freedom Kean square Calculated 35 13,352 3 .55** 1.94 Lot 8 *4-5,808 1 2 .16** 4.91 Replicate 3 4,000 1 .0 6 8.54 24 3 ,7 0 2 0.98 2.27 197 3,766 'otal F values P = 0.05 267 Subclass L X R Error Comparison among lots at 5 percent level of probability Lot 4wk. wt. 4 6 5 7 8 313 346 362 363 379 1 9 4ol 3 *K)4 2 425 434 at 1 percent level of probability Lot 4 wk. wt. P = 0.01 4 313 6 346 ** Significant (P<[ 0.01) 5 362 7 363 8 1 379 9 **01 3 *K)4 2 425 434 55 EXPERIMENT IV Based on the findings in experiment II, in which zinc bacitracin caused a significant improvement in chick growth, whereas terramycin did not, it seemed advisable to verify these results and possibly learn if sex differences occurred with respect to growth performance or intestinal coli population* One-day-old, sexed, cockerels and pullets, hatched from a White Leghorn X De Kalb cross at Michigan State University, were separately allotted, as previously described, into six lots, each containing four replicates of ten chicks each. To the broiler basal diet shown in table 1a, 200 grams per ton of zinc bacitracin or terramycin were added. These rations were fed continuously for the 28-day test period. One cockerel and one pullet representative of each experimental lot were selected for intestinal E. coli determination at the fourth and sixth week* Results and Discussion Of the two antibiotics tested, only zinc bacitracin promoted a significant growth response (tables 10 and 11), and the improvement was noted in both sexes reared separately. No differences in intestinal E. coli counts were evident at the fourth week. However, bacitracin-fed birds were the only ones showing coliform population in the gut areas two inches above the cecum at six weeks. While evidence did not strongly support involvement of E. coli in zinc bacitracin-fed chicks, an associa­ tion may have existed. Apparently, the intestinal E. coli population 54 was variable and probably was affected by age and sex as well as other physiological factors (i.e. pH, location in the digestive tract, differences in diet). Associated with the growth response to bacitracin was a slight improvement in feed utilization in both sexes, which was not observed in the terramycin-fed lots. Insofar as chickens were concerned, bacitracin, at high levels contained a mechanism of growth-promoting action which was not consistently evident with terramycin throughout these particular experiments. In five experiments (II, III, IV, IX and XII) where the two types of anti­ biotics were compared directly on the same basal ration, zinc bacitracin produced significant responses four times, while terramycin promoted a response in only one instance. In each experiment, intestinal E. coli population was higher at four or six weeks in the bacitracin-fed birds than in the terramycin-fed birds (table 12). Furthermore, in the experi­ ment where terramycin gave significant growth increases, intestinal E. coli population was markedly increased over those experiments showing no responses. 55 o rH X 3 O O * * c o § O © o o !2« © © P X! o © O rH C © •H o VT*l CH NO d a & s O CM ■a Pi Xi C © *H CO On © © co 00 h£» O > .3* P •rl -H U jh r-1 «H cti O CM ON • si rH W 73 JX © O © O J55 C30 co C"\ o -d- CO O rN -d {>- 3 © ^ bfl a or j* i J5 © s F* O -d- tiO © »H > P Eh CM o © ao © p p •H © O ■« © © c P C P o O o p •rl -H © P P X3 w f iO *•£ hJ ^ o 5 S p O © C o 125 © p o rH 2 © i5 to •H O © O JO o •rl (X, •5 O © P P O P •H C3 O O CM o FS © 5 u © c o •H irt P © p © p O a o c3 o CM P T3 P E ©& 5 rH hH C o rH © © p; o o o rE>» •rl tt iH P c o © & © o o o © i —! Cf LJ , © 2 © rH SH -H °P © o P o •H P O CM r\ LT\ VO 3S £5 * * * 56 TABLE 10 ANALYSIS OF VARIANCE OF FOUR-WEEK-OLD COCKEREL CHICKS. Experiment IV Source of variation Total Degrees of freedom Mean square F values Calculated P = 0.05 P = 0.< 116 11 4,669 3 .38** 2.43 Lot 2 13.816 10.00** 4.82 Replicate 3 2,068 1.50 2.70 L X R 6 2,900 2*11 2.19 105 1.382 Subclass Error Comparison among lots at 5 percent level of probability Lot grains wt. 1 5 3 308 31? 344 1 5 3 308 317 at 1 percent level of probability Lot Grains wt. ** Significant (P < 0.01) 344 27 TABLE 11 ANALYSIS OF VARIANCE OF FOUR-WEEK PULLET CHICKS Experiment IV Source of variation Total Degrees of freedom Mean square F values Calculated P » 0.05 P =0.01 112 11 3478 3.58** 2.43 Lot 2 12,168 12.53** 4.82 Replicate 3 259 0.27 8.56 L X R 6 2,192 2.26 2.19* 101 971 Subclass Error Comparison among lots at 1 percent level of probability 2 6 4 273 275 305 Lot Grams wt* * S ignif ic ant (P { 0 .05 ) Significant (P^O.Ol) 2.99 53 o o CM CO vr\ o o a &q CM ON O cd CN- * O CM XI d g © *H © CO ft ho © Vi i-l & R 0o 1—l ON co 00 CS- On ON CO UN ON ON o ON NO W d © © © © H G •H cO p d G © •rl © G G © © J>a p © O ■S x o t o cO *H © -p © ON ON ON on ON on CO ON © iH Is ■d © p p p *d c o *© o © © X bp P CO 3 © © G JS © s > O -d* ho <0 *rH •w' P P G •rH © © G i>> P G hO © •rH & rH G cO P •H d rH G d_ © rH © X *H X © £ © O 1 © CO G G § o o a © d © © d I H G G •H 8 S O •rH P O •H 5 P © G cO o P p © w XI & p w o 5-8 p 3 © G O 53 C »H o © G p •H O © P © G ■H CQ O O CM © G O S3 G •rH o © G P •rl O © P O G •H CQ O O CM rH CM ON -3" © G O UN G •rH O © G P •H O © P O G •H CS) o o CM NO © § 0 *=> £ 2 rH P o 1 © © G w G © <8 "© co •H 8 X G H' © © S > © p XI W >» © 9 o © © © G P X P 3 * * * 42 TABLE 14 ANALYSIS OF VARIANCE OF FOUR-VJEEK-OID CHICK WEIGHTS. Experiment V. Source of variation Degrees of freedom Total Mean square F values______ ______ Calculated P = 0.05 P = 0*01 227 23 20*722 4.80** 1 .8 8 Lot 5 82*962 19.09** 3.11 Replicate 3 7,245 1.67 2.65 L X R 15 2*763 0.64 2 .1 0 Error 204 4,331 Subclass Comparison among lots at 5 percent level of probability Lot 4 wk. wts • 3 5 1 309 340 362 4 4ol 2 6 410 434 At 1 percent level of probability Lot 4 -wk. -wts. 3 5 1 4 2 6 309 340 362 401 410 434 ** Significant (P <(0.0l) EXPERIMENT VI Whereas, the evidence thus far did not strongly support the in­ volvement of coliforms in the mechanism of antibiotic action, increased intestinal E. coli population in the presence of bacitracin supplementa­ tion was generally observed in the previous five trials. The findings of Moore et al. (1946) indicate that sulfasuxadine is not growth depressing at the 0.5 percent level as is sulfaguanadine. It was, therefore, deemed advisable to re-investigate the effect of zinc bacitracin on growth in relation to intestinal E. coli in the presence of certain bacterial-inhibiting compounds. Eight lots of C o b b ^ White Rock, one-day-old cockerels were allotted, as previously described, into four replicate pens of ten chicks each and fed for a 27-day test period. To the basal chick starting diet shown in table 1a, 50 grams per ton of polymixin B sulfate and 0.5 percent sulfa­ suxadine were added and fed continuously. The polymixin compound was employed since it is effective against gram negative intestinal organisms, while sulfasuxadine is extremely effective in inhibiting gram-negative E. coli as well as some gram-positive forms (Moore _et al•, 1946). One representative bird from each experimental lot was selected for intestinal E. coli determination at the fourth week. Results and Discussion Sulfasuxadine, while not growth inhibitory in itself, did depress intestinal E. coli population when fed in combination with bacitracin (table 15)* This phenomenon suggested that reduction of intestinal E. coli is not, in itself, a factor in the mechanism of bacitracin action, but that large increases in the organisms may have been related to optimum bb performance. Certainly, the improved growth pattern and increased feed efficiency as well as coliform count were strongly suggestive of a close relationship between this organism and chick performance. Since polymixin alone did not stimulate growth or affect feed conversion, but appeared to increase intestinal S. coli, the possibility existed that bacitracin may have provided the "trigger mechanism" neces­ sary to obtain a growth response. It can be theorized that the anti­ biotic causes the release of intestinal enzymes, which act on the nutrients present to improve assimilation and subsequent utilization by the bird. There was an apparent synergism between polymixin and zinc baci­ tracin in supporting a large increase in intestinal E. coli organisms. The increased organism population was also associated with the maximum growth observed and was significantly greater, at the five percent level of probability than the highly significant (P = •' 0.01) response from antibiotic alone (table 16). 45 O M O •H •H -P o CO 0 © O H CO xl O H O 2 O o o IB • • G > ftH O • O X on CM p£J ft CM <0 S*« hD to © o 3 ■rl rl ^ O • * lACO CA £N H IN-VO ’B •rl JED CO 44 > o -rj o o o• o• o o CM Ov IN OV rH v/v • • O CO IN u~\ • • • • f—IrH H H H IN Ov Q 00 Ov rH rH H © -P G ■H CD tfl -P X © o > © © G G O O 3 S G *d o to to -P © VO G •O O S *H *3 © Vv G *O -P G *H CO vrvxrv • * o tsa © O G ft © t* o o E *H S ft G ft hD -P © G •H O <0 G •rl O cO * a o •H -P O •H -P G <4 $ cO G X> ©■ 3 10 S o -p ho a 8 © 3 m G O O o S CM •H © tO cO •H G X> *rl *H s X) to ■ G rH O © ft G O O !2f a 5 § J U i —IrH o o Si ft o CM o O * r \ VV G G -P •H O G .9 O (0 G G a -P cO X> o G .a O > 3 (—I cm n^a - trvvo >s « ft o CM I N CO W ft o ft o a t © 0 o £ •rH -P rCS o © -H TTf 3 -P -p O g S *ri O © ft •rH © •H rH O O o cO © § x> O G •H 46 TABLE 16 ANALYSIS OF VARIANCE OF CHICK WEIGHTS AT 27 DAIS. Source Degrees of freedom variation Total Mean square Experiment VI F values____________ Calculated P — 0.05 F = 0.01 310 Subclass 31 11.301 4.51** 1.79 Lot 7 33•893 13.51** 2.73 Replicate 3 2,613 1.04 21 5.013 2.00** 248 2,508 L X R Error 8.54 Comparison among lots at 5 percent level of probability Lot 27 day wt. 1 452 3 461 4 5 7 2 8 6 469 473 491 506 514 537 at 1 percent level of probability Lot 27 day wt. 1 452 3 4 461 469 ** Significant (P <^0.01) 5 473 7 2 491 8 506 6 514 537 1.97 47 EXPERIMENT VII The observation in experiment VI of possible synergistic action between bacitracin and polymixin in increasing intestinal E# coli, with the associated added growth above normal antibiotic response, prompted re-evaluation of this apparent effect# Oleandomycin, having a narrow spectrum of microbial activity similar to penicillin (Price et al# 1959) was consequently compared with two levels of zinc bacitracin# Low levels (1 or 5 grams per ton) were compared with high levels of antibiotic (200 grams per ton) for an eight and one-half week period to learn if associative effects existed between intestinal E# coli population and performance and, if so, did it persist beyond four weeks# Each growth-promoting antibiotic was fed singly, or in combination with polymixin, at approximately a four to one ratio throughout the experiment# However, due to the pronounced growth-depres­ sing effect of oleandomycin at the higher level, lots 8 and 9 (table 16) were discontinued at the end of the fourth week# Nine lots of one-day-old Cobb's White Rock cockerels were allotted, as previously described, into four replicate pens of ten chicks each and fed the supplemented broiler-starter ration (table 1a) through the fourth week# At that time, Lots 1 through 7 were transferred to finisher bat­ teries, as previously described, and were fed the supplemented broilerfinisher ration (table 1a) for the remaining four and one-half weeks. One representative bird from each experimental lot was selected for intestinal El# coli determination at the fourth week and one at the ninth week# 48 Results and Discussion Polymixin, in combination with the high level of bacitracin, again promoted a maximum growth response which was significantly greater at the five percent level of probability (table 18) than the highly significant response obtained over the basal with bacitracin alone. This improved performance was not, however, associated with an increase in intestinal E. coli count, based on the single sample cultured. Where­ as, increased weight gains occurred when polymixin was added to each growth-stimulating antibiotic, with the exception of the growth-depres­ sing higher level of oleandomycin, an analysis of variance (table 18) revealed that these differences were not significant. It was noted that except for an absence of organisms in the con­ trol lots no marked differences in intestinal E. coli occurred in any of the experimental lots, despite the highly significant early growth re­ sponses from the antibiotics. Since the higher level of oleandomycin showed a count of intestinal E. coli similar to the other lots, but a highly significant growth depression, mechanism of action in depressing growth was probably not due to reduction in E. coli alone. This suggests that high levels of this particular antibiotic may inhibit vital metabolic processes by blocking essential enzyme functions, or by increasing in­ testinal microorganisms which act in a toxic manner toward the host. At eight and one-half weeks, significant responses to all anti­ biotic treatments had disappeared (table 20). This was in accordance with the view held by many investigators that antibiotics exert their greatest growth stimulating effect during the early growing period and 49 tend to lose their effectiveness in late growth. No marked increases occurred in intestinal E. coli due to bacitracin supplementation at the ninth week, except in the presence of the low level (table 19)* This microflora difference could not be associated with growth, since lots showing equivalent growth had a lower intestinal E. coli population. 50 JO H EFFECT OF A LOW AMD H3>H LEVEL OF GROWTH STIMULATItG ANTIBIOTIC IN THE PRESENCE OF POLTMIXIN ON CHICK GROWTH AND INTESTINAL E. COLI POPULATION AT k WEEKS.. o rj § M oc ♦H W o 0 ■H P rH ® P O ^ hO hO CU*H O & cm o o 'S 5 8 fH -P 1o £ o su u o P c > H hD P P O H O ® O o • • O pH rH ^ pH O * • O CAP ^Vf\ O -H T j *H iH *M ifi r j q | O *P o on on VO -3" ON ON H ON ON un VO ON ON ON NO ON m cD J-4 © -H PU m * © P w hp £ Si ® .id £ to g O -3“ hD ON > P g *rl 5 G G p •H r*5 P E >4 O E rCJ O G *3 ^ G © cD rH O >rt G G •H cD « © rH rH CM rH Q rH rH O O o © o O O G •H a E5 5 •8 © © rH O O o H -P 8 p cD rH P P © P P o G o H CM CN-d- UN VO IN- 00 On 51 TABLE 18 ANALYSIS OF VARIANCE OF FCXJR-WEEK-OLD CHICK "WEIGHTS« Experiment VII Source Degrees of" freedom variation Total Mean square F values ____________ Calculated P = 0,05 P = 0.01 310 Subclass 35 43,176 7.54** 1.74 Lot 8 164,979 28*83** 2.60 Replicate 3 8,820 1*54 2.65 24 6,870 1.20 1*57 240 5,722 L X R Error Comparison among lots at 5 percent level of probability Lot 4 wk. wt. 8 9 1 2 6 3 * 7 5 371 375 469 512 521 530 531 538 552 7 5 at 1 percent level of probability Lot 4 wk. wt. 8 371 9 375 ** Significant (P<^0.01) 1 469 2 6 512- 3 521 ^ 530 531 538 552 P O * M op •H © I o P © - UN UN i— 1 iH © © © °• £ O O © O* So u a J3 G •H O cd G P ■H o cd P o G © -H G « o 13 UN G ■rl G o •H cd O G cd p G *rt P O •rH cd O P cd P o G O *H G « •H « o O UN CM iH CM CAP* UNVO ec; o A h VO M E~t R Q S 3 & M Ou O P^Ph Ph tuO © p o •H P O (T\ CM cn e -j M o la R m £ © • t>» W C? © P P «r o Jh Ti VO © G © S to P » -H P H 'O O rrt O P-, oi P G G O O {3 o < g •rH O > S o TJ G cd © H o . rH Eh P O R G •iH O r*» 0 P G cd © rH O H 55 TABLE 20 ANALYSIS OF VARIANCE OF EIGHT WEEK BROILER WEIGHTS. Source of variation Total Degrees of freedom Mean square Experiment VII F values Calculated P = 0.05 220 Subclass 27 25,317 Lot 6 27,990 1 .0 3 1.5*+ Replicate 3 18,83*+ 1 .1*+ 3.70 18 25,506 0.77 8.55 166 2*+,513 1 .0*+ 2 .0 6 L X R Error No significant differences among lots at the 5 percent level of probability. 54 EXPERIMENT VIII The possibility of synergistic action between bacitracin and polymixin (Experiments VI and VII) in affecting intestinal E. coli population and growth, suggested that other microorganisms, their en­ zymes, or their metabolic products which might be found in the intestine, could combine favorably to enhance antibiotic growth stimulation and shed light on possible mechanism(s) involved. In the following experiment, one percent of live yeast cells of Saccaromyces cerevisiae or 0.5 percent of an acid hydrolysate of killed cells of S. cerevisiae were fed singly and in combination with bacitracin or bacitracin plus polymixin. One-day-old Cobb*s White Rock cockerels were allotted, as pre­ viously described, into eight lots, each containing four replicate pens of ten chicks each. The basal broiler-starter (table 1a) was supplemented with the levels of bacitracin, polymixin, yeast, or yeast extract as shown in the experimental design (table 21) and fed continuously for the 26-day test period. Since intestinal E. coli involvement as a basis for explaining the growth-stimulating mechanism of bacitracin appeared uncertain, rela­ tive populations of other intestinal microorganisms were cultured aseptically, as described previously, from the area approximately two inches above the cecum. A representative bird was selected from only the con­ trol and bacitracin-fed groups for intestinal microorganism count, since yeast treatments were ineffective in stimulating growth. 55 Results and Discussion Neither live cells of S. cerevisiae nor an acid hydrolysate of killed cells were effective in altering the growth rate of the chicks, which indicated that yeast cells or their metabolic products are not directly involved in the mechanism by which bacitracin stimulates growth. Bacitracin alone produced highly significant increases in 26-day weights (table 22), but further increases were not obtained in this trial by addition of polymixin. No noticeable differences were observed in intestinal E. coli count from bacitracin supplementation in this test. However, based on the sample cultured, there were increased numbers of total gram-positive organisms in the antibiotic-fed birds. Total numbers of micrococci, enterococci-type organisms, lactobacilli and the total gram-negative population were decreased in the presence of bacitracin (table 25)* These findings are in agreement with Anderson et al. (195^) who showed a decrease in enterococci and lactobacilli-type organisms from feeding chlortetracycline. 56 £ *©8 •H 0 C^WO i>-VQ Cn.vO hD rl rl rH H On ON v r\ o~\ S3 O O- NO NO UN, cn NO CN• • rH rH P 0 ■e cO . § 0 £ O r«£H o o O O O £ rH rH H 0 o }/st o u n ,vr. 0 M -p * o 0 rH •H 3 -P cn, • * C*\ • o p2 p O •H 0 £ O 0 0 £ fl o o JE5 !S fl •H O cd £ P -rH O cd P 0' P P O P o »2! J25 £ £ 0 cO cO 0 p •rH O 0 P P £ •rl -H P P •H »rH $ & o £ *H N O H £ -H (SI 0 £ O O o P5 CM 0 fl O O O 3 CM O UN, C CO CO P P O O rH £ £ O ■H -H CU (SI CSI o o o vr, CM 8 B £ e hO £ 0 0 £ fl •H O o 3 0 ♦ 5 *H a 0 0• 0 P * • u •H *H £ ■° -H & tsJ o •H £ £ •H O O & O © o £5 0 O O 0 aa & £ 0 0 0 0 1 £ cO o o 0 CO H CM CrN-^' in VO JNwOO of Saccaromyces 18 cO 0 £ O S hydrolysate a 0 c o Acid CM 0 £ o ** ?> ■d 0 rH CO rH XI •H cerevisiae * 57 TABLE 22 ANALYSIS OF VARIANCE OF 26-DAY CHICK WEIGHTS. Source of variation Source of freedom Total Mean square Experiment VIII Calculated F values P = 0 .0 5 P = 0 .0 1 290 Subclass 31 11,812 2.88** 1.79 Lot 7 35,777 8.73** 2.70 Replicate 3 15 *958 3.89* 2.65 21 3,232 0.79 1 .6 2 228 4,096 L X R Error Comparison among lots at 1 percent level of probability Lot 26-day wt. 6 1 3 4 8 2 5 7 443 449 455 502 505 506 512 516 Same comparison at 5 percent level of probability * S ignif icant (P 0 #01) ** Significant (P<( 0*05) 58 \D •H'v, • r-W" M O © OH CM «V 12 O © *rl £ > ■£ bfl O © O © P rHdh cd 3 O S? I'd Eh JO CN ■a G PL, ■a TO 0 G CO CO Po •3* «"*> •H 6 4 O d rH *P •rl-J O O O cd cd rH P P ' g •rl CM CO -P G co £ o o vn P P I ocp s IH -H bO © O Eh +5 0 © G O ft M o^ cd g •rl o ON IN" © s o 525 p co © p CO Q © o u +5 0 © OOP, £h O"—' rH *H OP cv CO o R © § S3 •rl G cd B o XI 03 © G I— I eh 1 •H P rH © *-x ^ S Fj «rt to cd Eh •rfc}' O hD CO O Eh P •cfr cp ON cm * * O S Eh a d o •H P © G cd O P © to £; b/) o .3 3 rii, P 5 p P P 3 O cd Eh 3 § P s •H b) o o CM H1O •H •H P cd P, G O o o o •H o a •H •rl P P p cd *H cd cd r—I*H rH I— I G O a aP,sO sS,.♦rl a Eoh © o £ ■©£ 3P 0• «rl cd W © p •H bfl to © © © 0 !> •H p, 1 p P cd •rl CO 0 U bJ} PH 1 1—! * s cd P Eh Eh 2 U Eh o© © p HP EH I I II I cd P Eh Eh tip P cd cdco O Eh S’P S SPSP & P © o © © ©p p ** X © © cd cd a CQ bf) ft 3 iH Eh © G © G P P P W EH p CM H CM CP^t XT\vO 00 © p 6 4s X © P © rH G O 1—I © o TO P G O >rl O +3 G © rH § O_ bp -H © HX * & © © - P TO © X Eh p •rl G P G 0 © o 4J © o O +5 •H rH TO P, g © o Eh Eh © O E $ G «H O o o © 1 SP Eh ' j> 8 <4 EH * * * 59 EXPERIMENT IX The likelihood that, other intestinal microorganisms, in addition to E. coli, might be involved in the mechanism by which antibiotics stimulate growth prompted this experiment# In addition to evaluating certain intestinal gram-positive and gram-negative bacteria, a series of* several confirming tests for S. coli was also conducted. Accordingly, five lots of one-day-old Cobb*s White Rock cockerels were allotted, as previously described, into four replicate pens of ten chicks each. The starter basal (table 1a) was supplemented with anti­ biotic as shown in the experimental design (table 24) and fed continu­ ously for a 28-day period. A representative bird was selected from each experimental lot from which to culture intestinal microorganisms from an area approximately two inches above the cecum. These included total gram-positive bacteria, total micrococci, enterococci and lactobacilli, as well as total gramnegative population and the S. coli fraction of this group. In addition, five confirming tests were made on several representative E. coli colonies from each of the experimental lots which had been incubated by the MaiImann-Peabody (1957) drop plate method. Results and Discussion Significant growth response occurred from bacitracin supplementa­ tion alone or in combination with polymixin, but not from terramycin (table 25). No marked change in intestinal E. coli was observed from the feeding of either type of growth-stimulating antibiotic alone, based on the sample cultured. However, addition of polymixin in combination 6o with both broad and narrow spectrum compounds was associated with a marked increase in E. coli> but with no apparent effect on the growth pattern. While some of the previous experiments indicated that an optimum E. coli population may be associated with growth responses, this particular test (table 26) did not support their involvement in explaining the mechanism of antibiotic action. On the other hand, antibiotic supplementation re­ sulted generally in a depression of the total gram-positive forms, as well as enterococci and lactobacillus groups, though reduction in numbers was not associated with increased growth in the presence of terramycin. Results of the five confirming tests for the cultured intestinal E. coli (table 27) strongly indicated the presence of pure strains of coliforms and added support to the validity of this method of culturing. 61 "O C © hH 0) (0 VO K VO Cn cn Ov -d- VO 00 (N- o o- VO *n cn cn o S' hD 8 M Fm O M 0 ro(D o © t> p O *H fl .M •rl H £ O Vl (fl (H O

a*H p s hD rH P O P o EH © -p © fl o 25 2; fl •H O P a* W f l od OOP •H *H P Cd fl P •H rH o •H P © fl •H *H p © P C £ T) P <£ © O P s O cd P O P 6 hE © fl o 25 © © g C H O © fl P •H O © fl •H O P o fl •rl « O C *t-1 N O O fl fl © P O CM 8 CM CM cn o o tn o fl *r4 O >s § fl fl © P O O CM 62 TABLE 25 ANALYSIS OF VARIANCE OF FOUR-WEEK CHICK WEIGHTS. Experiment IX Source of variation Degrees of freedom Total Mean square F_values_________ _ Calculated P = 0*05 P = 0*01 181 Subclass 19 96,889 34.30** 2.00 Lot 4 37*993 13.40** 3-^5 Replicate 3 4,704 1.66 2*67 12 1.501 .53 2.3^ 143 2,825 L X R Error Comparison among lots at 1 percent level of probability Lot 4 wk. wt. 1 445 4 463 5 2 3 WO 507 524 Significance between lots at 5 percent level of probability did not differ from significance at 1 percent level of probability ** Significant (P<^0.01) B .^ O H OH o 0 0 rH VO CO CO os CM 0> iH 4 1« H rH 0 E-* ■35' ft o » ‘ ' m CM CO OSHf OS CM H CM § 1o ■a 65 OS ho S3 S<^—-* 4 O *P O dJ2> r—I — *rijO O 0 p G o o rH 0 *H E (J (fl rl ► H X "— G •H P I 0 H 0 0 0 0 -=J* 55 £3 S5 S5 O G G CM O O hC P *H 0 O P P O © ft p VO OS OvvO rH CO CM G G O O E O CO rH QSMS VO -3rH O H (§ O 0 0 p e n 0 -0 CM S 0 S 0 G O O CM C o •H P 3 •H X p 0 0 53 G •r) CtJ •Aou -P o © O O ft Eh O - o rH X 0 o o o S R S CM H CM 0 O G O O CM K CM 4 •8 gK rH •rl c 0 vO CM ©^ > E •H hfl H 0 3 P P p •H it O hi) 0 o O i—I Eh ft oB 3 g m. CM O H VO S o 0 O G o p ©' 53 CO P E hO o ft 03 -p © © S3 G O O O £3 {25 ' A o •> «k P o o o 0 G O O !aj vrs rH rH •* Jh rH 0 ft * o o P o XI G C •H *H O O 53 O iH 0 0 B o P P •H O 0 53 3 co 0 G *H 53 P S3 X o p O O p {3 •H •H 0 “ O 53 0 53 •H 4 P e p P. X O fajO 0 P ■rl P 0 rH G e S'S P O PI P P *H O 0 X O O G G •H IS] 0 G O O oo o {25 CM CM © •• .5.5 R R 5 H U OO P G O P P 0 iH •* P P o o P rH ft © * P O P P 0 0 o o CM CM 0 © 0 H <5H H CM OS us 6k H TO • TO © • © 53 « © © TO £ P Pi TO © •rl hD P Pi © P t s D £ O P O C P if s 53 rH rH O 3 (3 43 TO O O © P •rl © •H o © 3 S O C 43 •8 V. O O o O © P < P i o •rl o 43 • tf © Xl O Pi M TO B P M Pi © - p M Pi 43 © Pi »** g © o o O 43 o P Pi Pi Pi 43 © O *H Pi Pi i—1 O P P > P !> o P TO > P !> P P t> P > p -H © P © •ri •rl 53 P © P © •H © i—! P © p © o 53 -P O Pi P O (3 P © p O © P O •P O O p O p O O •r1 P © © •H © •rl TO © P © •rl © •H O p O © TO TO XS 43 TO © TO TO © X> T5 txO TO xs hO X5 £ O 53 P * 53 tf O Pi $ 53 O 13 © £3 • o 53 SP 13 • © Pi P. P O K P O CL, ftJSJ •H CL| •H 52J P M PL, •rH £3 P M o ft O 8 If 3 OXI a x'— * h TO 53 O P -P O c0 © 33 Pi M ft. ft 9 > U o 10 I (D C •H h O E J3 CO ■3 a O r pj CO © co Pt -©p «®H X>5? ,Q TO O 2 £ P o co CO « U © PH P P-, 1 CO *m ■(p0 81 0> P (0 Pi so P T5 O CO S 3 © Pt 5 © Pi Pi •rl r© *H © © n TO © r~I O ■ MH P % J! 5O ©E 05 «M PI -P 10 o 0) c Hr~l CM VTi Standard Methods for the Examination of Water, Sewage, and Industrial Wastes, Tenth Edition, 1955 TO P 65 EXPERIMENT X Despite the apparent variability in intestinal E. coli population, it was felt that additional work should be done to accumulate more in­ formation on the probability of their involvement in antibiotic action* Since interfering debris made counting of intestinal population of total gram—positive and total gram-negative forms difficult to interpret these bacteriological tests were not included. The work of Perdue et al. (195^) suggested possible differences in growth effect and tolerance to erythromycin compared with penicillin which could be reflected in differences in intestinal microflora popula­ tion of animals fed different narrow spectrum antibiotics. This prompted the following experiment in which performance and certain intestinal microorganism population were determined. Six lots of one-day-old Cobb,s White Rock cockerels were allotted, as previously described, into four replicate pens of ten birds each. The basal diet (table 1a) was supplemented with antibiotics as shown in the experimental design (table 28) and fed continuously for a 28-day period. One representative bird was selected from each experimental lot for intestinal bacteria determination at the fourth week. Among the gram^- positive forms, total micrococci and enterococci were determined, while the S. coli population among the gram-negative species was counted. Results and Discussion Highly significant growth responses were obtained from the addi­ tion of bacitracin or erythromycin to the ration (table 29)- The further addition of polymixin to the bacitracin diet results in a large increase 66 in intestinal S. coli, which was also associated with a slight non­ significant increase in growth. Whereas, erythromycin alone promoted a significant growth increase, equivalent to that from bacitracin, further supplementation with poly­ mixin resulted in a suppression of growth and a pattern of intestinal microorganism population resembling that from the bird fed the basal ration. This may have indicated a suppressing effect from the combina­ tion of these two compounds on certain gut microorganisms that were a prerequisite for optimum antibiotic response in the chick. Total intestinal micrococci and enterococci populations were depressed below the basal lot by feeding polymixin or zinc bacitracin (table 50), though it was noted that depression of these organisms which occurred when polymixin was fed singly was not associated with a growth increase. Erythromycin, on the other hand, appeared ineffective in altering these gram positive organisms, yet promoted a significant growth response in the chicks. 67 G ■8 •H P3 +> «a} O © -dUA 'A CM VO vr\ vo O vo *A CM vo vo v\ •pH 00 CO CM 5 C5 § .3 S'5 P ^ ra P> *H 43 g rH O 2 T? ^o W PL, © -P © § S3 o vn, © 0 £3 •5 0 © G +3 ■H O © * © © G X O •H +3 C P O O O +3 •H •H © ' ^ X -rH -P ,£> -P *H © E +> 73 _ bfl G 'O O »— <3j © -P *8 © § 35 © § S3 CM O G *rH N o G •H O © G +> •H O © § >rl CM © 8 sg * G •H O >» £ O G G © O O CM CM O O O r— 1 CA -=fr *A O *A CM S *ri I B © o o rH R. 68 TABLE 29 ANALYSIS OF VARIANCE OF FOUR-WEEK-OLD CHICK WEIGHTS. Experiment X Source of variation Degrees of freedom Total Mean square F values Calculated P = 0.05 P = 0.01 222 Subclass 23 10,985 2.75** 1»91 Lot 5 31.136 7.79** 3.1k Replicate 3 12,11k 3 .03* 2.67 15 k,0k2 1.01 1.76 176 3.996 L X R Error Comparisons among lots at 5 percent level of probability* Lot k wk. wt. 1 2 6 5 3 k 55k 562 569 60k 606 625 at 1 percent level of probability Lot 4 w k • wtr. 1 2 6 5 3 k 55k 562 569 60k 606 625 * Significant (P<'0*05) ** Significant (P ( 0.01) 69 CO •H 0 ft ft G •H O G C O ® W o ft *H 1 £ ^ t1 tfl 0 o CO o o ft o ft CM + o o o ft- o o CO CM CM + o o o + o o o«b * 15 -P (0 0 ft £ o G fCMt O O UN 0 0 O UN g 523 UN ca ft O rH G 0 Ph £5 CO G •rl O C6 G ft •rl *H *H 15 'g I er O ■a w as G 43 O +> JB to -p £ hO o ft 1-3 0 G 0 G ft ft O H O CM G -P •rl 5 t> O aj cfl ft ft O O G o om G •H O CTJ 2 ft .3 & 8 G ft •H G •H CSJ 1? 0 *& o O O O CM CM 9 o CN o ft • O UN 0 o ft G •H ft Q rH 0 •» > H •H ft CO O P o G o 0 A w H 0 •• G rH 0 G O o NO 0 rH •• rH 5 * 70 EXPERIMENT XI Th© findings of Barnes et al. (1959) that rats fed penicillin increase their growth if they have access to feces but not in the ab­ sence of coprophagy suggested possible relationships between the ingest­ ing of fresh droppings and the mechanism of response to broad and narrow spectrum antibiotics. In order to find what associative effects might occur, an experi­ ment was conducted in which fresh feces were fed to chicks daily in the presence and absence of antibiotics. Ten lots of Cobb*s White Rock one-day-old cockerels were allotted, as previously described, into four replicate pens of nine chicks each and fed for a 27-day test period. To the basal chick starting diet (table 1a) antibiotics were added as shown in table 51 and the chicks were given access to fresh fecal material in the following manner. Fresh feces were collected daily from caged-layer hens receiving an antibioticfree diet. The fecal material was diluted in an equal volume of physio­ logical saline solution and poured in a strip on top of the mash each morning. The moisture content of the feces was 82 percent, based on the 72-hour 90° 0. oven-dried sample tested which agrees with the work of Rosenberg and Palafox (195^)* On this basis of feeding each chick had the opportunity to receive a daily intake of approximately 0.5 to 0*75 grams of fecal material on a dry basis during the test period. One representative bird from each experimental lot was selected for intestinal bacteria determination at the fourth week. Total micro­ cocci, enterococci and E. coli were cultured, as previously described. 71 Results and Discussion In each instance, presence of fresh fecal material in the diet caused a highly significant growth depression (table 31)* This was not completely compensated for by the addition of antibiotics* Aureomycin restored growth more nearly equivalent to the basal than did zinc baci­ tracin (table 31) s^nd this improvement in growth was significant at the five percent level of probability (table 32)• This suggested that the fecal material may have contained growth-depressing organisms, which are more effectively inhibited by wide spectrum antibiotics, or combinations of compounds having wide spectrums of activity rather than by those having narrow spectrums of activity* The presence of fresh fecal material also caused a depression in feed utilization in each instance (table 33) despite the presence of anti­ biotic treatment. It should be pointed out that antibiotic supplementa­ tion resulted in an improvement in feed efficiency over non-antibioticfed birds. Apparently, the antibiotics exerted part of their action by inhibiting these depressing microorganisms, or substances produced by certain microbes that interfere with nutrient assimilation. Bacterial populations of micrococci, enterococci and E. coli were changed, some forms markedly so, in each experimental lot which had re­ ceived fecal material. A depression in the gram-positive population was particularly noted among the micrococci organisms (table 3^)« A signifi­ cant (P < .01) depression in growth occurred from feeding bacitracin in combination with fresh fecal material, compared to the bacitracin-fed lot receiving no feces. Since intestinal S. coli population was not de­ pressed from feeding feces, intestinal E. coli per se are probably not 72 an important consideration in the mechanism by which antibiotics stimu­ late growth* Rather it appeared that a balance between populations of beneficial intestinal microorganisms, including E* coli and an inhibi­ tion of certain detrimental forms, were an important mechanism that permitted increased growth and improved feed conversion* 72 G % ‘H CD TO ^ bD o H CO ON vo o rH CM i—I CM UN^t -3- on VO 0 0 rH rH CM CM -3- co i —Ii —I rH rH onp O *H G •rt n ^ flS O P VT\-3- ON ON VO VT\ c^ry UNvO o n on VO vo ON (TV G o TO o p CD CD G * CIS CD [S T t»0 ^ TO PI D— 10 ^ o w g 0 *H > p <4 O Q) VO 00 (TV UN un-3- CM UN ON ON u n c jn vo j$- UNrH ON ON UN UN CM ON Q ON vO UN p TO <0 &> 0 rH I+ O -H 0 TO I+ l+ I+ I+ fe Tf rH •H A O CD O O TO O § CD G CD P g G § bJO (0 g 3 TO 0) TO G l>* TO O «H rH rH O P Tj TO •H G *£« ,-Q O •rH "S •3.S % •H CD P w '£ TO 0 w 6? p unvo on on P 0 •rl CD 5 P G TO & 1 •i5N!>T&J 0I 0 rH g o o UN UN £ TO 8 0 TO G jcj CD ft P O rH CD 0 H p G G ■3 TO r s 0 •rl P G TO •ft G o P O O •H •H 0-~-^ p ^ TO 5 •rf p e T> he 1 TOpO W P •S 0 0 G G o o P5 525 G G •rH .H O O TO TO G G P P •H *H O O TO TO •ft P •rl *H O O G G •H *rH N IS) O O G G •H *rl O O O O CM CM O O O TO G P •H O TO G P *H O TO TO •Q P O o q CM CM P TO -» * G G -H -H p p rN S g o o 0 0 G G P P TO TO o o o o CM CM G G on-3 " unvo in- co TO O 0 G 3 TO g O 0 G TO TO O O o O CM CM On O iH P b O 1U 1 0 TO TO CD bfi TO ’ d <2 33 0 TJ 8? O TO rH TJ *H P TO • O O ■—i 1 <83 • o TO X ft TO 0 >>P rH I •H 0 p rH CM s CD •H •H P P p-j •3 S u P & O TO U 0 P TO P <$ * * 13 TABLE 32 ANALYSIS OF VARIANCE OF 27-DAY CHICK 'WE3GHTS, Source of variation Degrees of freedom Total Mean square Experiment XI F values Calculated P = 0*05 P = 0.01 352 39 29,727 6.59** 1 .6 8 Lot 9 117,004 2 5 .93** 2.49 Replicate 3 2,655 0.59 8.54 27 3,643 0.81 1.70 27*+ *+,512 Subclass L X R Error Comparison among lots at 5 percent level of probability Lot 27 day wt. 2 4 6 8 1 10 7 9 5 3 458 495 499 531 536 539 595 602 615 632 536 539 595 602 615 632 at 1 percent level of probability 27 day wt. ^58 ^95 ** Significant ( P <0.01) ^99 531 74 TABLE 33 ANALYSIS OF VARIANCE OF 27-DAY CHICK FEED EFFICIENCIES. Experiment XI Source Degrees of freedom variation Total 39 Lot 9 Error 30 Mean square F values_____________ Calculated P = 0.05 P = 0.01 0.0449 0.1656 19.03** 3.06 0.0087 Comparison among lots at 5 percent level of probability Lot 27-day feed eff. 5 3 1.64 7 I .65 9 1.72 1 1.74 6 10 1.80 1 .8 8 8 1.09 4 1.92 2 2.Q4 2.31 at 1 percent level of probability Lot 25-day feed eff. 5 1.64 3 7 I .6 5 ** Significant (P^ 0.01) 1.72 9 1.74 I 1.80 10 6 1.88 1.89 8 1.92 4 2 2.04 2 .3 1 ____ 75 6 CM rH £ O 0 O ft to -p £ 0 £ £ o o is! s; o o o o o o CM CM - CM ONVO © Is u © "3 % © c\J !m -p CO Q> OX O CO O^C^N-CT on Cx-vo Ov ON ON ON IN-IN- O ON O N -5 © • t s O •H Cx-H on VP\_j- UN UN vo ox vo <4 43 -P unvO vo 4-\D n UNVO UN vo O ON o © If vr\ i —l R 3 © P=t CN g 3 TT u 43 TJ © § TJ *3 S c o *H -P P •>H rH © O O © © u © m 5 Si rH © © W © H « -P n © -p •8 © «s © U «H O I o Pi •rH O -P -P •oH •rH -P * * * TJ © > © © © 5 3SH © o TJ O © *H ho ho © o o I—I o •H & •a © © «H o © ■d E-i © © c © o 43 SH S3 S3 S3 •H •H •rl O O O © © © o *H a *©d. td S3 •P £ 4> O o o -P •H *H ©' 43 *P 43 © *H *rl -P s -P Tf WJ O $ 3 -P s © O © © tf o © U U u -p •P •rH *H O O © c5 © ,o 4> 43 o O O £3 © © © § g SS 125 •H N ■g . 95 csi O o o Q O p cvi •H cvj cm © •H © CM CM CM © £ R HH -P 3 r f CM (^ 4 - UNVO O- 00 ON O O each morning. Feces from above source dried at 100° F. passed through hammer mill and added to ration at the 1 percent level. Feces from above source dried at 100° F, passed through hammer mill, autoclaved for thirty minutes at 15 pounds pressure, and added to ration at the 1 percent level. ■8 0 'w* 3 A- H 3 v O Q v vO CA xa3 »Aio VAVO vo 3 vO CA vo O O n V A N O *A 3 £?P •H 0 CM 6 CO b f O ho *3 ' p 0) XA 00 CM CM On o o H * ON « S p ON CO O v O CO ON CArH 00 A * r H XA CAH •rl o O g rl o o o o P ■8 * * *T 2= * E“* C P P O O O •rl »rl *H 0 P aj as P P P P P •H -H *H O O O 0 0 0 45 43 43 O O O *3 CO as p P 4> o O •H *H 0 4> P 43 •H -rl p P TJ •§133 G 0 0 0 0 S O f Ol f Ol f Oi 3 3 3 3 P P •H * r l »H « « « o o o _ ~ Q CM ® 8 CM P •rl P ■rl tf £ P P p P P p 0 0 0 P P p O o o O o o CM CM CM CM H CM C A 3 xav O A- 00 ON O rH CA P 81 EXPERIMENT XIII Since G-oebel and Barry (1957) reported that sialic acid, effective against certain disease—producing viruses, closely resembled colominic acid, which is produced by E. coli, an experiment was conducted to evalu­ ate possible antibiotic-related effects this compound might have on chick performance. In order to test for possible effectiveness against the growth-depressing entity in fresh chicken fecal material, sialic acid was fed singly, or with zinc bacitracin, in the presence or absence of fresh feces. Eight lots of Cobb*s White Rock one-day-old cockerels were allotted, as previously described into four replicates of ten chicks each and fed the supplemented basal chick starter ration (table 1a) continuously for a 28-day period (table 5^)* Two representative birds each were selected at the fourth week from lots 1, 2, 5 6 for certain intestinal bacteria determinations, as previously described. The possibility that the mechanism of antibiotic action might be closely related to the amount of feed consumed, as well as feed transit time in the digestive tract, prompted a comparison of certain gut effects between control and antibiotic-fed birds. Thirty randomly selected chicks from each of Lots 1 and which were normally full-fed beforehand, were fed mash containing a ferric oxide tracer from 8:00 a.m. until 11:00 a.m. and then fed the normalcolored mash for the following three-hour period. approximately 0.1 gram of FeO in the mash. Each bird received This proved adequate as a 82 marker and did not affect the consistency of the droppings. Measurements were made of the rate of movement of normal colored feed through the tract in three hours following tracer feed. Additional measurements included live weight, length of tract from gizzard to cloaca, full and empty weights of the intestine and weight of intestinal contents on each treatment (table 41). Results and Discussion Fresh fecal material significantly depressed chick weights below the basal at four weeks when fed alone or in combination with zinc bacitracin, and/or sialic acid (table 59)* Th© addition of either the acid or the antibiotic in the presence of fecal material significantly improved chick growth at the five percent level of probability, but it required the ad­ dition of both compounds to produce growth equivalent to birds receiving the basal diet. Apparently, the sialic acid was partially effective in counteracting the deleterious effects of fresh fecal material, though the optimum level may have been higher than that employed in this study. The acid was ineffective in further enhancing the highly significant growth effect from bacitracin in the absence of fresh fecal material. Counts of intestinal populations of two birds per lot showed rela­ tively good agreement between replicates for total micrococci, enterococci, and lactobacillus organisms, with an indication that the numbers increased in the presence of fecal material and were decreased when bacitracin was added to the diet (table 40). On the other hand, although E. coli increased when fecal material was fed, wide variations between replicates made the counts of the bacitracin-fed lots difficult to in­ terpret. 85 Since the antibiotic-fed lot gained 65 grams more, on the aver­ age, and consumed only 76 grams more feed on the average than the controls (table 41), all the increased gain from bacitracin cannot be accounted for on the basis of increased feed consumption alone. This appeared true since the antibiotic prompted an eleven percent added gain from only eight percent additional feed intake. In this experiment, average weight of intestine, as well as feed transit time, was less for control birds than for antibiotic-fed chicks, though the individual variations among the limited number of samples determined revealed that these differences were not significant. Large numbers of birds would have been required to measure whether differences between gut capacity and feed transit rate would be meaningful in ex­ plaining mechanism of antibiotic growth stimulating action. 84 •G O S! *8 3 © © ft hO VO O- G VO rH £N-ON CM CM £>- CO O CM O- ON VO 0 0 p rH rH rH rH rl H rH rH *=«1 o G 3 rH © Q CO 33 CM 6 •H ft w o tc W p p •H • CO G o ^ t © »H G (0 © J3 44 bp $ © 10 s t0 -P ^ O N -3" O CM vm-3- CM ON CMVO \M^p VO CM VO OVM^t O ON vo CO *n-3- © P I © g S. TJ © © •J* •H t o •rHi—1«H © CO O P © **H H ? •rH £ G O •H P G © G G O © E ©P G o «H © I © O J? *H TO P © O © •H «H P •rH S fot § O © o P p G O S’ CO * * W {>» © rH O »H o fri © © I+ f t TJ I + t + i+ 00 3 CO G CM 9 rH ft * rH T) O tfi •rH -p © O © £ © G P O P O O .rH © •rl P .£ w rH *H -P S © T! G to S •H T j O © P O © © © © G O O K t£ © © o o CM CM § e S G3 o o CM CM O o P bf» § k O © G ■P •rH O © P O © rO o © .£ W P S £ TJ bfi •H 'd o N1 CD p s © © G G O O £5 £5 H CM © © G O G O £3 pg CM-3" O O O O CM CM vT\VO O O O O CM CM O'-CO rb rH © O rH W TJ © © C hfl »«H © rH 0 © CO 1 i—1 © £ •rH O © © O o r —I © o «H •rl © 8 Tt G Ti •H O © g p G P ft to r—v G O •H P -H © p *H {> O G G *H •H 'O > © •H 2 © T) O © O •H rH © •H CO w t © ft i w H © •H O © © ft «H * * 85 TABLE 39 ANALYSIS OF VARIANCE OF FOUR-WEEK-OLD CHICK WEIGHTS. Source Degrees of freedom Experiment XIII Mean square Calculated 31 22,465 4 .39** 1.76 Lot 7 82,398 16.10** 2.72 Replicate 3 4,555 0.89 8.54 21 5,046 0.99 1.84 262 5,118 variation Total F values____________ P = 0.05 P = 0.01 315 Subclass L X R Error Comparison among lots at 5 percent level of probability Lot 4 wk. wt. 2 4 6 8 1 3 7 5 434 469 472 489 509 523 560 566 at 1 percent level of probability Lot 4 wk. wt. 2 4 6 8 1 3 7 5 434 469 472 489 509 523 560 566 ** Significant (P^.Ol) 86 m p 5 ..Sif po ©? .4 ON O 4 VO VO i? CM Cv4 to 0 0 0 •H 04 O P 0 hC| w '4 or G cn ON • • o o I G 0 -H VO VO CM 00 00 • • H op 4 cp C"i—1 cL CM CM 00 00 u*\ to P S I fjSb O 3 *rl4 o o o « Ctf H iP P — g CM ON CPvO VO ON «k «k - 00 vo {>- O-H H«hCP A -P *h a -p o .3 ^ to P *rl 0 o to P O G SI w o op .© o o e s 1— 1 (0 g o S gc o o o o o o o -4 -4 CM CM + o o * * VT\ CM + 1,300 1,700 + •H 5 § O O ON O 4 -4 * CM -P G TJ cd 0 _ TJ G -O td cd ho © C 1 -H p >H •H O + o o o o o* o•» O -4 -4 CM CM l—1 VT\ O * O O •» m CM © > o 0 ip G p 0 G G O 0 *H © P td P -P _ P S rH o *H rH &, CM H CM H CM H CM 0 G «H O 0 0 0 G 0 .H OH 0 Cd «H 0 40 rH 4 0 O G •H bl) 0 p$ * to 0 o & o *8 5 0 •H 0 8P i>> •H TJ Hi. o -p 0 0 0 E H P G GP o 0 0 © p SP G 0 0 > o • 0o TJ O GO »rl • P O 1—1 0 •• > rH *H P * 0 O P o GO 0 » 0H 0 •• GH Pi 0 « GO O o rH ♦* e-t H * * • *8 p p © s rH TJ P G 0 -rH P O TJ G G 0 p 0 rH gSf o Q P. 0 p * 0 0 P 1 —1 G 0 Pi 0 0 0 £ e 0 G to P o— GO H o CM « CM x CD J ® «H H £ CHICKS AT 30 DAYS OF AGE MEASUREMENT OF FEED THAI'S IT TIME AND GUT CAPACITY OF ANTIBIOTIC-FED CO to o o 0 o 0 2 G •H P 0 0 P G ♦H s O P p 0 9* s X S ho G &a s-x cfl N N •H « & '—to* • o CM CM CM vrv VO • rH JP CO CO cts O, p to G CD H > oj • o (O 2 «—I * • o VO OH ct; o *H P s 03 G O 0 0 £ G rH r> VO cS X p 0 1OS H 0 G <•-» X ■Gg *H 0 0 G O G0 »o rH •• P rH rH CCS X G CCS H H rH s 6 O .H P G 0 G G tSl O Xl CCS ax o JQ cd p CO s to *H *H 'H hfl •rl G X G O £ g 0 >* 0 CO 0 P 0 G p G o csS O G -H 0) p o GPP p *H *H -H p XS 0 0 ca o cm * O 0 5O o G o •H O P 8 o S3 O o CM O 0 00 pG £ O •H G VPi CQ

-vo • • • • io © o Id -e o -H ^ G •rl H X! P Nr) w OR OR UR UR-3" NO CM OR OR U R UR CR CM CM CM UR UR <*} O © 1 K © g u © E-* * G O * © •H © O © I I I I I I+ 1 I+ © G O fjq ft G O O © •HI HI CM -3- O o| G © * ■d © © ,© •rf © p •H P ■t8 b X5 £O CO 'fcR O I+ I+ + I+ + I+ © G 3 G •H o rH G G G ■rt •H •H O O O © © © *H O O •rl •H c g •H •H •H O O O © © © © © G G G P P P o P> o •H , Q *H -P G © © ©• s o p © •H G O -P *H I —1 X> x> O o o G .5 •H © © o O « CQ g g a. o. O O O O rO c o -p © e hO G G G •H •H •H © © § G o £5 a ©ft O o CM CM CM o o o o CM CM © © 5.5O u © g G G © © -p p o o o o_ CM CM t § o ,G bD G O G X P TS © TS •H > B ft 3 H CM CR -d' URNO R- 0O ON O rH Method of preparation described in table Id* ** Daily feeding of fresh feces from hens receiving an antibiotic-free diet. One-half feces and one-half physiological saline solution by volume strip-fed on top of mash each morning © © 91 O o VO II CO -3- CM A- A- d va d VA VO CA VA 'A O (0 0 d d CM CO vO* 5 cO PU VA VA > P«4 00 ■8 co § •3 s CA * * rH ON • CA S o * * CA 00 • CA i—1 CM O • CM CM 00 • VA o (0 0 S VA VO CO CO g> <0 0 c 0r *• CM ON a CA CM CO rH *H JO C O JO O U A rH ON tv On O i—1 o CM H 0 > CM rH (0 o 0 CM CA VA rH CA VA VA CA CM VA On CM CM 'A O H d rH CM CA VA CA CM VA CM CM VA VA 0 d On CA ON » X O fH CM c 0 O © CO -P 5 •p 0 >* 0 A Xl I A(M -P a cu ■p o r’a O V/ »3 XI I A CM •H C % uy * -x- 92 0 •3 •H g O bo o vt"" o rH G 5^»H vo On CO CM VO CO rH VO CM ON VO 00 VO CO o o- 00 O CM O £>» rH 02 P C 0 . G O 0 2 e -P rH^h O -H O rH 01 G •H P «d o H *-3 cflw X § 6 •rH h£) 0 O G 2} 0 P o 0 a SJ t§ 8 g.H (0 0 -P G •H .5 ftSS1 ■'-S3 CO rH CO CM VO CM CM + + o o o O O Q o• o«k -2r ft -3- -3- vo CM CM O O O NO C O O CM CM O •fe § 8 H CM o o O p vo O« * CO-X CM •H 0 O O o s ■a § s O •H G oS o a E * <1 *0 0 o i£ ■g0 H + o o o o o-3r CM CO O CO c o oat X- X- vo Xh O O o CM CM O O CO OrH CM CM o o 9• >» S« CO VO I I+ I I+ I I I I I+ + I+ + I+ I+ 0 03 t-H *H ^ O + o o o G X O H •H 0 O P X O -H p rb ^ • •H O o ££j c5 -P G G •rl •H G G G •rl ■H •H O O O 03 0 0$ G G G P P P *0 03 G •H -P G Xi O P O O •H •H 0 X P X 0 *H *H -P e hD P Til S % $ *rl •rl •H O O o 05 05 03 X X X o P o o o o G G G •rl •H Q cq C2 C •H 0 0 0 G G O O S 3 rH CM CO O O o O O O rH rH •H •rl O O •rl •H G G © © a a 0 G © C G • r ) *H •H •H o3 03 & I § g G 0 P P o O o O a a O o o o G G © G G o o CM CM CM CM CM c3 CM -3- VO NO 0-00 ON O rH ** See footnote» table 37 OF ANTIBIOTICS, FECAL MATERIAL AND LEED E. COLI ON RELATIVE OF CERTAIN INTESTINAL BACTERIA IN FODR-WEEK-OID CHICKS POPULATIONS O oj o 93 GENERAL DISCUSSION The specific manner in which antibiotics stimulate growth has been a subject of extensive investigation during the past decade. From the numerous tests conducted several theories have been proposed to ex­ plain the action of antibiotics. According to Bird (1956), at least four mechanisms may be involved. 1. Direct effect on the metabolism or physiology of the animal. 2. Sparing of certain essential dietary nutrients* Reduction of toxic entities or sub-clinical disease in the bird or animal. 4* Stimulation of beneficial microorganisms in the digestive tract. Much circumstantial evidence has been presented in favor of each theory, though no positive proof of a single major action has been shown. Therefore, the possibility still exists that multiple physiological phenomenon may be operating in effecting the increased weight gain and improved feed utilization usually observed from feeding antibiotics to animals. Certainly, the germ-free approach represents a realistic method of establishing whether the growth effect of antibiotics is a direct one of increasing uptake of certain essential nutrients, or of altering intestinal wall structure more favorably for nutrient assimilation. Luckey (19^6) pointed out that some differences may occur in the gross morphology, nutritional requirement, and chemical composition of germfree birds as compared with conventionally-reared chicks, but that these 9b differences do not detract from the general similarity noted* For ex­ ample, Luckey found that a mature Bantam chicken, which has been reared to an egg-laying state in a germ-free environment compared favorably in most physiological measurements with a normal Bantam bird. The major differences noted in the germ-free bird were smaller proventriculus and adrenals, thinner intestinal and ceca walls, less lymphatic tissue and a larger thymus than the conventionally reared bird. Gordon (1952) made further studies on the morphology of germ-free birds and found no effect from feeding antibiotics in the absence of microorganisms, though he did suggest that feeding penicillin to conventionally-reared chicks tended to change these birds morphologically so that they resembled germ-free chicks. Gordon also found that neither terramycin nor streptomycin caused physiological changes, although the terramycin-fed birds did show marked increases in biotin, folic acid and vitamin B ^ cecal contents com­ pared with the control chicks. It appears then, that antibiotics do exert an influence on the morphology of certain specific tissues in the bird. tion of these physiological alterations is difficult. However, interpreta­ From a strictly physiological standpoint, the effect of dietary antibiotics in thinning the wall of the intestine suggests the possibility that this may improve absorption of hydrolyzed amino acids, simple sugars, and fatty acids which are in direct contact with the villi of the duodenal mucosa. On this basis, improvement in growth rate from antibiotic feeding would be due to the physiological function leading to the transportion of more essential nutrients across the less resistant, thinner membrane. 95 On the other hand, certain intestinal microorganism populations raay increase the amount of several water-soluble vitamins within the digestive tract of the bird. But it is unlikely that added vitamins Per se explain the growth—stimulating effect of antibiotics since usually dietary vitamin levels have been adequate for maximum growth. Since total protein, balance of essential amino acids, as well as vitamin and mineral levels were all adequate with respect to known requirements, it was felt that likelihood of a nutrient— sparing effect in the experiments described was not great. The possibility of antibiotics reducing the incidence of biochemical degradation of nutrients within the digestive tract, on the other hand, may offer an explanation of a mechanism of their action. If, for example, organic, inorganic, or mixed organic- inorganic oxidation: reduction potentials for enzymes were changed by the presence of an antibiotic, the availability of the nutrients present in the gut might conceivably be altered due to differences in enzyme efficiency. Results of several tests reported in this thesis (table 45a,b) are generally in agreement with those shown in the literature insofar as an antibiotic-depressing effect on the gram-positive microorganisms is concerned. However, all tests (Exp. XII, XIII) did not show reduc­ tion in total gut micrococci or enterococci from antibiotic feeding. Nor was an antibiotic growth increase noted in each instance (Sxp. IX) where these bacterial populations were reduced. Since many of the micrococci forms of bacteria are disease pro­ ducers in animals, they may be indicative of adaptive forms capable of depressing optimum performance. Staphlococcus, streptoccoccus, gonococcus, 96 pneumococcus and diplocci-type organisms are examples of those which might logically be suspected as capable of reducing growth potential, feed utilization, or livability when present in the digestive tract* It is conceivable, too, that measurable populations of these organisms in the gut, even though they be present in sub-clinical numbers, could produce toxins or antimetabolites which could interfere with normal as­ similation of nutrients, or block vital enzymes. The growth-depressing effect observed from feeding fresh fecal material in trials XI, XII, XIII and XIV, demonstrated the presence of a fraction, which is effective in extremely small amounts and is sensi­ tive to heat and drying at 100° F. (approx. 3&° C.) or less. Jordan and Burrows (194-7) state that the thermal death point of bacteria is dependent on moisture and time, and that most vegetative forms are killed at 55° ° 5&° 0• in ten minutes under moist heat. Despite the fact that the temperature at which the fecal material was dried was below the thermal death point for most vegetative forms, the extended drying time (72 hours) probably proved lethal for most organisms. Since no growth depression resulted from dried fecal material, live microorganisms, rather than products of their metabolism are responsible for the observed growth retarding effect. The noticeable increase in intestinal micro­ cocci and enterococci in the lots receiving fresh fecal material (table 44) suggests that these may be the organisms responsible for depressing growth, though the birds receiving terramycin and fecal material showed a significant growth improvement, despite the marked increase in these particular gram-positive forms. 97 On the other hand, depression of lactobacillus population in the intestine noted from antibiotic feeding may reflect adjustments in gut pH to a more nearly ideal medium for maximum absorption of some hydrolyzed nutrients. Proteins, for example, are precipitated from solution at different isoelectric pH points and, therefore, will probably be attached by the appropriate proteolytic enzymes much more readily when in solution. In contrast to the concept of depressing harmful intestinal bac­ teria, many investigators have taken the view that certain beneficial microorganisms in the digestive tract are directly, or indirectly, in­ volved in the mechanism by which antibiotics stimulate growth. Numerous investigators (loc. cit.) have observed that the coliform bacteria com­ prise sixty percent or more of the bacterial population of the intestinal tract, and some have associated these types of organisms with antibiotic action in mediating growth patterns. The principal objective outlined in the experiments of this thesis has been to learn if E. coli is directly, or indirectly, involved in ex­ plaining the mechanism of antibiotic action. It is apparent from experi­ ment I and II that introduction of live E. coli organisms via crop inocu­ lation does not provide a satisfactory explanation. Furthermore, attempts to learn the effects of antibiotic action when reducing gut S. coli by drug inhibition were not particularly fruitful in elucidating a specific pathway of antibiotic function. Trials III and VI did indicate that zinc bacitracin was effective in restoring depleted intestinal E. coli population and growth due to sul­ fa drug inhibition. The depressing action of sulfonamides on S. coli is due to their antimetabolite-like action whereby the sulfonamide drug is 98 preferentially substituted for p-aminobenzoic acid in the enzyme system of E. coli, thereby depriving the organism of essential growth factors* Since zinc bacitracin effectively reduces the growth inhibition on S. coli caused by sulfaguanadine (Exp* III), part of the growth stimulating effect of the antibiotic in the bird may be due to its action in block­ ing certain unknown antimetabolites. In experiment VI, an apparent synergism on growth was observed from combining polymixin with zinc bacitracin and this was associated with a marked increase in intestinal coliform population (table 46). Additional experiments, however, indicated that no particular pattern could be ascribed for these organisms due to treatment, notwithstanding the improved growth noted from feeding the combination of drugs. Further, when intestinal E. coli from all experiments are considered (table 47) no pattern emerges that can be associated with antibiotic-stimulated growth responses. Any beneficial effect of microorganisms in improving chick growth must be an indirect one in which the organisms act to pro­ vide necessary enzymes that may be missing from the digestive tract, or are present only in limited amounts. If, for example, these digestive enzymes are sub-optimal in amount or concentration in the absence of a specific microorganism, antibiotic action might be explained on the basis of increased enzyme action on those nutrients not normally available. No completely satisfactory explanation exists concerning the ac­ tion of enzymes, however, the most probable postulate is that given by Michaelis and Menton (1915) and currently supported. According to their theory, action of enzymes is explained on the basis of "active centers" on the surface of the enzyme molecule. by the following equations. The specific effect is described 99 Enzyme + substrate — enzyme substrate complex Enzyme substrate complex .--=r=— ?*■ enzyme + products of enzyme action While the available evidence points favorably toward involvement of the cellular material from E. coli in explaining the mechanism of bacitracin action in stimulating chick growth (table 4-2), it appears that if the action is enzymatic there is a certain specificity of enzyme action* This is borne out by the fact that neither the cellular yeast material, with its multiplicity of enzymes, nor the enzymes liberated from soaked barley were effective in altering bacitracin-stimulated growth. The significant growth response noted from inclusion of cellular E. coli material, but not to crop inoculated live organisms, may possibly be explained by the concept that the cell membrane of E. coli represents an effective barrier against nutrient contact with certain enzyme sur­ faces. Since narrow spectrum antibiotics, such as zinc bacitracin or penicillin, do not depress intestinal E. coli, their numbers may be in­ creased in direct proportion to the microflora which these compounds inhibit. Increased numbers of live S. coli will then result in larger numbers of dead coliforms in the intestine. The cell membrane of these dead organisms may not be as much of a barrier against movement of specific enzymes which could be involved in antibiotic stimulated growth as is the cell membrane of live organisms. Fruton and Simmonds (1955) state that extracts of E. coli contain a multiplicity of enzymes, including phospho trans acetylase and aspartase. Aspartase is required in the Kreb cycle exergonic conversion of fumarate to aspartate. Conceivably then, more effective conversion of protein to amino acids with less loss of energy 100 from certain of these exergonic enzyme reactions offers a possible ex­ planation for the mechanism of antibiotic action* On the other hand, experiment XI, XII and XIII support the con­ cept that antibiotics and antiviral compounds reduce the growth-depres­ sing effects of fresh feces* This may explain their mechanism of action* Since birds normally have access to droppings, growth would in most in­ stances be depressed below a certain attainable level due to effect of fecal microorganisms on the host animal1s metabolism* However, when the organisms are weakened or destroyed by antibiotic action, the growth in­ hibition is removed. The possibility also exists that antibiotics may stimulate growth due to an effect on the appetite, since antibiotic-fed birds consumed more feed than did non-antibiotic-fed chicks in all experiments except II, III and VII. While it is conceivable that the increased feed intake observed in antibiotic-fed birds may provide more opportunity for build­ ing tissue than in control fed birds, both groups were consuming feed far in excess of maintenance requirements. Furthermore, the increased efficiency of added gains usually observed from antibiotic feeding cannot be explained on the basis of increased intake alone. It would, therefore, seem that factors in addition to increased feed intake are operating in effecting growth increases from antibiotic substances. From these studies, it appears that further work is indicated in the area of enzymes as the explanation for the major action of antibiotics in stimulating growth in poultry and farm animals. _In vitro experiments should be conducted to resolve the specific nature of the enzyme(s) present in cellular material from E. coli, which has the capacity to 101 influence chick growth. In addition, further tests involving feeding of fecal material should add to our knowledge of certain specific harm­ ful microorganisms that are probably depressing performance in the poultry house. Finally, the area of virus inhibition on bird performance requires further investigation. Titration of various levels of anti-viral agents, or those compounds showing promise, should prove fruitful in unlocking some of the many secrets of the digestive tract as yet unknown to us. 102 va -CM dr CO 1 CA CO I Hm rH o VO CM ON 1 m 3 CM 3 CM 1 m 3CM 3CM 1 O O 3CM 1 VA On VA CM CA VO O VO CA VA rH e x o 3CM -aO VO vO O vo -3T VA G 0 *H P VA 3 ca eh ♦ Ci5 • 23 • C3• ts XS • C3# !2J CA & 3 aj 'O © p m S' • o <5• 3 CM 15 0 « C5• 1U CN O VA !25 hO O 1 CO 15 o O 3 CM O tN CM 525 1— 1 VA O CM * H CQ O O ON IN- a• -31 o ON N* VA VA S w> Jh a o s o uo o „ U J-t © © P ft pHSw* © Pi a?*H O h &i % O •H -G O " to u u p 8 G •H M 1 1 c hO £ gS rH 0 O P O o g •H 5 O U U © P 1 CQ Eh CM CA O >a s O U JQ •t5 r* © i -d- 5 A - aureamycin (5 105 O o va CA VO ft o o o VA CA CM VA 5 rH CA O I—1 rH rH ch O VO CM CA O On o A- S -3 CM A- iA VO VO VA ca VA fe m ca o to cq -p C M o o o • -= t CM o A- -3 VA VO CA -3 CM o o O O VA H P © rH rH A - CA CO 0)M rO |JhW S' 0 o O O O p CM .5 CM v a O O ON CM C A CA Onvo H rH ON O VA CA Er* + O e-» o o•k -3 * -3 CM CM CQ o O O o o* o VO O o H vo CO H ft O o o •3 H G •rl vO VA CM •H o •rH G o t—I 0 o o o o o 00 3 o H CO (0 G *H -P CO ® •p G M G cti » O o / —£ > su (0 o •rl u 6 h0 U rH 0 0 -P s_/ ft O Er* © +5 o ft G » o x> o • CM c5 •» • ON i2 * vo O- CO CM ON *G- cm 3 NO o o CO CM VO OO^t CM CO vo o o iH CO 10 G 0 •H xi o £ o xi 5 h w S W -3O NO ff U o VO h£ -3 o «H c •H -3- CO vo •H O 3 •H a G *G 0 o CH O a CO CO t>- I—1 o vo S