.-‘,.-. ~ . » -o.. ’ ..-m.-:..:-- ;..>--MOCO;MMWflOW-Q*d o!” —- - . .‘f""' «WQOI'CO"WQFO~OO-MOOMOOC«“-o‘oc-‘oo—v...» .. -..-.9. o .9». “4....-- -n-- ......a-.... r . ACTENOMYCES VESCOSUS. A POTENTEAL PATHOGEN. {N DOGS Fhesis. for the Degree of M. S. :MECHIGAN STATE UNWERSETY ALLISON AM DAVENPORT 1974 ¢‘\‘I.I. ‘2', .‘g ‘1' a 3 . . . -.| 0 ~ . . .7 .- _ -: r ‘ . . . . . - ."-‘ K o v - O- . 9" ‘ .fid‘ ‘ rt- - ‘- -% —‘ o ‘ .‘ as - -.;"§ .',‘_' - 'V .‘ My 1 u. .- o .' .r“~ a? ”r:- g1, _ , . _ _' b g _. r; {Afié- _ _. ”$5.: _. :— - r»- - '7" ' ....... y .......... " ‘ r _ '. .r‘ ‘ v i . . Linda- . ‘-. .. .' ' -l.‘ .. ' ‘ ..x. u. ,4? . '- .- - . . .I a ‘ . - . -.I 0 on ‘ 0 ~~. - Q vac-o. or“. 94 .0 coo o..h¢~.6.4flo.c|3~.“ ......... nm_cj.9t.0.‘on--omdnsdoo A . . ‘v— Tml' -—-vm"r‘r--rr_ THESIS 3 1.4 3, Umvcrmty LI BRAR Y Michigan State J J 'Mm cwwvw - ~. A j“ ’ amoms‘w 3' MM & SUNS? 380K BVNUERY ‘N A i1 LlBRA av Bmoms M snzxarcat. mama ABSTRACT ACTINQMYCES VISCOSUS, A POTENTIAL PATHOGEN IN DOGS BY Allison Ann Davenport In one and one-half years, five cases of Actinomyces viscosus infections in dogs have been identified at the Michigan State University Veterinary Clinic. Infections were similar to those diagnosed of nocardiosis and acéinomycosis during a period of nine years at the Clinic. The possibility of mistaken identification of the etiologic agent in many of these infections should be considered. It seems likely that many were actually A. viscosus. Actinomyces viscosus is a member of the normal flora in periodontal plaque and has been recently reported in occasional suppurative, granulomatous infections in dogs and humans. A survey of the history and characteristics of infections with special reference to granule morphology, cellular morphology, and biochemical properties of the strains from these cases is discussed. It was concluded that seven biochemical tests and an acid-fast stain should speciate A. viscosus from other diphtheroidal or filamentous microorganisms. The tests are catalase production, urease production, gelatin hydrolysis, Voges-Proskauer, lactose fermentation, aesculin hydrolysis, and nitrate reduction. ACTINOMYCES VISCOSUS, A POTENTIAL PATHOGEN IN DOGS BY Allison Ann Davenport A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Microbiology and Public Health 1974 {{{z’ ACKNOWLEDGEMENTS I wish to express my sincere appreciation to the following people: Dr. G. R. Carter for his generous assistance and use of his laboratory facilities; Dr. E. S. Beneke for his initiation and planning of this project and donation of equipment; Dr. R. G. Schirmer for his patience and cooperation in examination of clinical histories; Dr. R. W. Cook for his interest in canine Nocardia infections which led to identification of the first A. viscosus isolates at the Clinic; and Mr. Harold McAllister, an extremely knowledgeable veterinary micro- biologist who provided the isolates for this thesis. I also want to thank Dr. A. W. Dade, Dr. C. A. Reddy, Dr. J. A. Bresnak, Dr. M. J. Patterson, Dr. A. L. Rogers, Ms. M. G. Shue, Ms. D. Boettger and the staff at the Veterinary Clinic. The Michigan State University Depart- ment of Microbiology and Public Health provided financial assistance for this project. My gratitude is also extended to Mr. and Mrs. T. I. Davenport for their support and encouragement which allowed me to pursue this degree. 11 INTRODUCTION . TABLE OF CONTENTS LITERATURE REVIEW. . . . . . . . . Taxonomical Methods . . . . . Numerical Taxonomy . . . General Characteristics. . Cell Wall Composition. . . Serological Properties . Fermentation Studies. . . . . . . Ultrastructure. . . . . . . . . . Isolation of A. viscosus. . . Pathology of Actinomycetales. . . Confusion of A. viscosus with Other Treatment of Canine Actinomycosis and Nocardiosis CLINICAL HISTORIES OF CANINE ACTINOMYCOSIS SEEN AT THE MICHIGAN STATE UNIVERSITY VETERINARY CLINIC JUNE 1972 To NOVMER 1973 O C C O O O O O O O O 0 Clinic Clinic Clinic Clinic Clinic Case 129820 (CF-470-72). Case 138631 (CF-99-72) . . Case 141564 (CF-295-73). Case 145833 (CF-1078-73, CF-1097-73) MATERIALS AND METHODS. . . . . . . . . . Organisms . . . . . . . . . . . . Culture Media . . . . . . . . . Organisms . Page N NU‘J-‘N 11 ll 21 28 37 39 39 4O 42 44 46 50 50 50 Isolation Methods . . . . . . . Preservation of Stock Cultures. Morphologic Studies . . . . . . Biochemical Tests . . . . . Carbon Dioxide Requirements . RESULTS. . . . . . . . . . . . . . . . Morphological Characteristics . Biochemical Tests . . . . . DISCUSSION . . . . . . . . . . . . . . ADDENDUM . . . . . . . . . . . . . Clinic Case 147495 (CF-194-74). BIBLIOGRAPHY . . . . . . . . . . . . APPENDIX 0 O O O O O O O O O O I O O O A. TECHNIQUES FOR ISOLATION AND IDENTIFICATION OF A. VISCOSUS IN THE CLINICAL LABORATORY. iv Page 51 51 52 52 54 55 55 58 62 69 69 71 77 77 Table LIST OF TABLES Page Canine actinomycosis: reported characteristics in 8 dogs by Swerczek et aZ., 1968 . . . . . . . . . . . . . l6 Canine and feline cases from which filamentous micro- organisms were isolated in the Clinical Microbiology Laboratory, Veterinary Clinic, Michigan State University, from 1965 to 1974 . . . . . . . . . . . . . . . . . . . . 29 Comparison of biochemical studies of A. viscosus strains isolated at MSU with reported reactions of A. viscosus and A. naeslundii cited in literature . . . . . . . . . . 59 Expected reactions of various Gramrpositive filamentous or diphtheroidal microorganisms . . . . . . . . . . . . . 79 Figure 1 2 3 4a 4b 6a,b LIST OF FIGURES A. bovis sulfur granules with clubbed border. . . . Higher magnification of A. bovis sulfur granules. . CF-lO97-73. Histopathology of A. viscosus granules from lung tissue. . . . . . . . . . . . . . . . . . CF-99-72. Reversion from rough to smooth colonial morphOIOgy O O C C O C O O C C C O O O O O O O O O O CF-99-72. Reversion from rough to smooth colonial morphOIOgy O O O O I O O O O C O O O O O O O O O C O CF-194-74. Diphtheroid form of A. viscosus associ- ated with smooth colonial morphology. . . . . . . . CF-470-72. Filamentous form of A. viscosus associ- ated with rough colonial morphology . . . . . . . . CF-194-74. Histopathology of A. viscosus granules from cervical abscess . . . . . . . . . . . . . . . vi Page 23 23 48 56 56 57 57 7O INTRODUCTION The seventh edition of Bergey's manual (Breed et aZ., 1957) places Gramrpositive pleomorphic organisms into the families Actinomycetaceae, Corynebacteriaceae, Streptomycetaceae, and Mycobacteriaceae. Many of these organisms, when isolated from human and animal sources, are considered to be normal flora or contaminants. Since these are classified as nonpathogens, cultural identification of these slow-growing, fastidious organisms is not attempted in most clinical laboratories. The organisms are ignored or reported as diphtheroids (Gerencser et aZ., 1969). Recently, certain diphtheroids have been isolated in material obtained from cases of periodontal disease, suppurative granulomatous abscesses and empyemas (Coleman et al., 1969; Johnson et al., 1970; Georg et aZ., 1972; Adeniyi-Jones et aZ., 1973). The possible patho— logical significance of these organisms has necessitated better methods of cultural isolation and identification. During a one and one-half year period, Actinomyces viscosus has been isolated from.suppurative lesions in five dogs admitted to the Michigan State University Veterinary Clinic. The case histories and cultural studies reported in this thesis should aid the clinician and the microbiologist in the diagnosis of A. viscosus infections. LITERATURE REVIEW Taxonomical Methods Numerical Taxonomy There are two different systems of taxonomy in the analysis used for microorganisms: the Adansonian and the Darwinian. Concern with the validity of the results, time involved, and financial practicality of a computer in the microbiology laboratory has eliminated the Adansonian method from most clinical laboratories. Routine analysis presently relies on monothetic Darwinian principles of taxonomy. No one perfect method for classification of the kingdom Procaryotae has been found, and strict adherence to subclassification levels, similar to those used in the plant and animal kingdoms, is almost impossible. The eighth edition of Bergey's Manual of Determinative Bacteriology (to be printed in May 1974) proposes the taxonomic structure of 19 parts, many of these lacking class, order, or family (Langlykke, 1973). Harrington (1966) used numerical taxonomy to study diphtheroids included in the genera Mycobacteria, Cbrynebacterium, and NOcardia. Morphological, metabolic and cell wall studies led to the conclusion that consistent data were lacking to justify use of three separate genera. The seventh edition of Bergey's manual uses acid-fastness and cellular morphology as criteria for genera separation; however, some 3 species of Nacardia have been shown to be non-acid-fast, and some Cbrynebacterium sp. show varying degrees of acid-fastness (Harrington, 1966). Bousfield et al. (1969) also attempted classification of coryneform bacteria with numerical taxonomy. The group of coryneforms included Microbacterium, CeZZuZumonas, Listeria, Corynebacterium, Brevibacterium, Nacardia, Mycobacterium, Erysipelothrix, and Arthro- bacter. He also had little success in genus separation. Holmberg and Hallander (1972b) successfully used a computer in their numerical taxonomic methods to classify 123 diphtheroid isolates from human oral, pathological, and typed reference material. The 77 characteristics examined formed phenon clusters of the genera Corynebacterium, Bacterionema, Rothia, Actinomyces, and NOcardia with a similarity level of 87.5%. Actinomyces viscosus and A. naeslundii agreed on a level of 92.5% and thus did not warrant separate species names. Analysis of data involved use of the similarity coefficient SJ (Sneath, 1957a), simple matching coefficient Ss (Sokal et aZ., 1958) and single cluster analysis dendogram (Sneath, 1957b). About 50% of the characteristics examined were morphological or colonial. Diphtheroids can demonstrate varia- tions in physical and colonial morphology which leaves much doubt as to the significance of taxonomic data based on such characteristics (Gordon, 1966). Even though A. viscosus and A. naesZundii may be found in similar specimens and have many characteristics in common, other taxonomic methods have demonstrated definite differences between these organisms. Cell wall composition, serology, and a limited number of metabolic 4 properties demonstrate consistent differences sufficient to justify the present use of two separate species names. The eighth edition of Bergey's manual (Langlykke, 1973) separates coryneform organisms into one part, viz., 17: "Actinomycetes and Related Organisms." Erysipelothrix, Listeria and Lactobacillaceae are not con- sidered in this part. Corynebacterium is grouped without an order or family. The order Actinomycetales includes the families Actinomycetaceae, Mycobacteriaceae, Actinoplanaceae, Frankiaceae, Dermatophilaceae, Nocardiaceae, Streptomycetaceae, and Micromonosporaceae. The family Actinomycetaceae includes the genera.Aetin0myces, Arachnia, Bifidb- bacterium, Bothia, and Bacterionema. The criteria on which this classi- fication is based are not available at this time, as the manual has not yet been published (Langlykke, 1973). General Characteristics The family Actinomycetaceae defined by cell wall composition and cellular morphology cannot be placed in the present classes of Schizo- mycetes or Mycota (Sykes et al., 1973; Prevot, 1966). Organisms are non-acid-fast, Gramrpositive, nonmotile, and possess a transitory branching mycelial phase. No true spores or aerial mycelia are present. The cells are about 1.0 u in diameter and occasionally break into coccoidal or rod-like elements. Cell membranes lack sterols and cell walls lack cellulose, chitin, and diaminopimelic acid (except Araehnia). Antibacterial agents are toxic to cells. The absence of sterols in cell membranes makes the organisms insensitive to polyene antifungal agents (Davis et aZ., 1973). Phage infection and lysozyme suscepti- bility have been reported in some species (Gledhill et al., 1969). 5 Membranous organelles are not present. Products of glucose metabolism include acetic, formic, succinic and lactic acids, and some have traces of propionic acid (Li et aZ., 1968). Catalase production varies and addition of carbon dioxide necessary for anaerobic growth usually enhances aerobic growth in facultative species. Oxygen requirements vary (Gledhill et aZ., 1969). No gas is produced from sugar fermenta- tion. Hydrolysis of gelatin and production of indol are not seen. Cells may form long branched filaments or pleomorphic diphtheroidal forms, both with round swollen primitive spores or clubbed ends. Cell Wall Composition Cell wall composition has become an important tool in taxonomy of the diphtheroids. Actinomycetes lack diaminopimelic acid (DAP) which is characteristically found in its meso form in Neeardia, Mycobacteria, Baeterionema, and some Corynebacterium species. Cell walls of fungi rarely contain DAP or amino acids (Cummins et aZ., 1958; Cummins, 1961). Boone and Pine (1968) partially identified many species of the Actino— mycetales by cell wall composition. Complete speciation would have involved quantitative measurements of the cell wall components. Diphtheroids containing LL-DAP are Corynebacterium acnes, Streptomyces sp. and Arachnia. Bifidbbacterium, Actinomyees, Rothia, LactobaeiZZus, and some corynebacteria lack DAP. Lysine is present in the above species of corynebacteria, Rothia, LaetobaeiZZus and Actinomyces. Actinomycetes also contain mannose and ornithine while one or both of these are missing in the cell walls of Lactobacillus and Rothia. Glycine is found in some corynebacteria but not in actinomycetes. 6 Only C. pyogenes cannot be differentiated from.Actin0myces sp. by these techniques (Pine, 1970; Cummins, 1961). Reed (1972) examined cell wall components of Strain T-6, A. viseosus, with the gas chromatograph. The only amino sugars isolated were N-acetyl—glucosamine, and N—acetyl muramic acid, both found in a peptidoglycan. Rhamnose was the major neutral sugar. No sugar alcohol or techoic acids were present. Phosphorus and phospholipids were in small amounts and amino acids detected included alanine, lysine, and glutamate. Other studies (Pine et aZ., 1965; Pine, 1970; Gerencser et aZ., 1969) found galactose, glucose, mannose, and ornithine. Boone and Pine (1968) detected trace amounts of phenylalanine, methionine, leucine, and isoleucine. Cell walls of A. naesZundii contain fucose and deoxytalose which have not been found in A. viscosus (Pine, 1970). Two colonial morphologies of actinomycetes are commonly noticed: a spidery, granular, irregular form, and a smooth, entire, convex form (Ajello et aZ., 1963). A difference in cell wall composition may be responsible for the morphological variations (Sykes et aZ., 1973; Pine et aZ., 1967). Moderate quantities of hexosamines, and little or no aspartate, are present in spidery colonies, while low amounts of hexos— amines and larger quantities of aspartate are in the smooth colonies. Pine (1970) admits that morphological changes may be associated with variation in cell wall composition; however, changes in the site of cell wall synthesis or enzymatic differences may be responsible for morphological variation. Neither smooth nor rough forms of A. viscosus contain aspartate in the cell walls. Serological Properties Serological reactions of A. viscosus and other diphtheroids have been examined with agglutination, immunodiffusion, and fluorescent antibody techniques. Agglutination studies produced many cross reac- tions. Snyder et al. (1967a) found strains of A. viscosus agglutinated with antisera prepared from A. isranii, A. naesZundii, A. odbnto- Zyticus, R. dentoeariosa, and B. matruchotii. The antisera from C. acnes and A. odbntolyticus reacted with A. viscosus in studies by Slack et al. (1971). Immunodiffusion gel studies (Snyder et al.. 1967a) with A. viscosus cells showed minor cross reactions with A, bovis, A. odbntolyticus, and A. naeslundii antisera. An excellent serological test for A. viseosus is the fluorescent antibody technique. Hamster strains of A, viseosus do not react with antisera from human A. viscosus strains or with A. naeslundii anti- sera (Holmberg et aZ., 1972b; Gerencser et al., 1969). Human strains did react with A. naesZundii antisera and with hamster A. viseosus antisera. Reciprocal tests of A. naeslundii isolates with A. viscosus antisera from either hamster or human strains reacted (Gerencser et al., 1969). When A. viscosus hamster antisera was sorbed with A. naesZundii, all cross reacting antibodies were removed; thus no reac- tion with human strains of A. viscosus or with A. naeslundii was possible. Antisera prepared from A. viscosus did not fluoresce at diagnostic titers with cultures of A. isranii. A. bovis, A. odbnto- Zyticus, A. eriksonii, A. propionica. R. dentocariosa. B. matruchotii. C. pyogenes, C. acnes, or Ramibacterium pleuritieum (Gerencser et al., 1969). Studies of fluorescent antibody reactions with.A. viscosus of human, hamster, and canine strains showed cross reactions of both human 8 and hamster strains with canine strain antisera (Georg et aZ., 1972). Types of cross reacting antigens varied with canine strain examined since absorption with A. naeslundii had variable effects. Fermentation Studies Gas-liquid chromatography of fermentation end-products is not only a useful technique for speciation of anaerobic bacteria (Henis et aZ., 1966), but it can also provide useful information for gross separation of some anaerobic diphtheroidal organisms (Pine, 1973; Li et aZ., 1968). Three types of detection apparatuses have been used for determina- tion of volatile and nonvolatile acids: flame ionization, electron capture, and thermal conductivity. The electron capture device is extremely sensitive for identification of organic compounds with halogenated or polar functional groups (Cherry et al., 1969). Flame ionization is very sensitive to formic acid and less sensitive to water than the thermal conductivity device, but acetic acid carryover may occur when multiple samples are being tested (Moore et al., 1966). Li and Georg (1968) successfully used the flame ionization detector to separate Ar. propionica from other actinomycetes. Qualitative and quantitative analysis with this detector differentiated Actinomyces from Corynebacterium, Nocardia, and other diphtheroids (Pine, 1973). Howell (1963) noted that A. viscosus requires carbon dioxide for aerobic growth but A. naeslundii does not. Buchanan et al. (1963) suggested that actinomycetes lack aspartate permease and thus must synthesize aspartic acid from C02 and small molecules. Large amounts 9 of carbon dioxide are fixed into the carbonyl group of oxaloacetate by the Wood-werkman reaction: ++ ATP, Mg L C02 + pyruvate phosphopyruvateT oxaloacetate carboxylase (Wood et aZ., 1941; Buchanan et al., 1966). Other C02 fixation mechanisms are operative but these are not major reactions. The addition of carbon dioxide to a growing culture of an actinomycete will rapidly increase growth and production of several end—products. Proposed equations of glucose fermentation by A. isranii, A. bovis, A. viscosus, and A. naeslundii had evolved through extensive chemical extractions and titra- tions performed before the gas chromatograph became popular for such research (Buchanan et aZ., 1966; Buchanan, 1967; Pine, 1970). Anaerobic glucose -———+ 2 lactate (homolactic) 3.5 glucose + 3 C02-—-¥ 3 formate + 3 acetate + 3 succinate + 1 lactate (heterolactic) Aerobic glucose + 02 ———4-2 acetate + 2 C02 + 2 H20 Malate can substitute for carbon dioxide under anaerobic conditions to produce: low glucose concentration glucose + 2 malate -——9-2 formate + 2 acetate + 2 succinate 10 high glucose concentration 2 glucose + 2 malate-—-+ 2 formate + 2 acetate + 2 succinate + 2 lactate Other diphtheroids do not require large amounts of carbon dioxide for growth. Arachnia propionica, Bifidiobacterium bifidum, Propionibac- terium, and Corynebacterium acnes have the following fermentation balances (Pine, 1970): B. bifidum 10 glucose + CO2 --*-l formate + 13 acetate + 10 lactate + 1 succinate 10 glucose-—-—+ 15 acetate + 10 lactate _ I Ar. propionica 18 glucose-————+ 10 CO2 + 2 formate + 10 acetate + 23 propionate + l succinate + 1 lactate P. pentoaceum, P. arabinosum 10 glucose -——+-6 CO2 + 2 acetate + 14 propionate + 2 succinate C. acnes 10 glucose -—-+ 5 C02 + 4 acetate + 12 propionate + 2 succinate + 1 lactate Actinomyces sp. ferment glucose predominantly by the Embden-Meyerhof pathway; however, the organisms also use the hexose monophosphate shunt, phophoketolase pathway, and Entner-Doudoroff pathway to a slight extent (Pine, 1970). When carbon dioxide is added to anaerobic cultures of Actinomyces sp., fermentation changes from.homolactic to heterolactic. Carbon dioxide acts as a hydrogen acceptor to oxidize lactate, thus 11 forming formate and acetate from a pyruvate intermediate. Carbon dioxide is also used to form oxaloacetate from which aspartate can be synthesized. Energy yielding studies by Payne (1970) showed the addi- 2, and 02 + CO2 to anaerobic cultures increase the energy yield of A. naeslundii by 22%, 150%, and 200%, respectively. tion of malate, CO Presently no enzymatic differences between A. viseosus and A. naeslundii have been suggested to explain why A. naeslundii shows no preference for CO2 when growing in oxygen but A. viscosus does (Howell, 1963). Ultrastructure Two studies on ultrastructure in the Actinomycetales have shown some possible differences in species. Overman et al. (1963) found complex or coiled membranes in Ar. propionica, simple coils in A. naeslundii and A. isranii, and no coils in A. bovis or Bifidbbac— terium bifidum. Duda et al. (1972) and Overman et a1. (1963) noted variations in cell wall thickness in different species; however, these findings showed no relationship to atmospheric requirements of the species. Isolation of A. viscosus The generic name, Actinomyces, was first used by Harz in 1877 (Gerencser et aZ., 1969), who was describing a "concentrically radiat- ing fungus" obtained from tongues of cows suffering from a disease spreading around the jaw bones. Dr. Bollinger, who originally sent the specimens to Harz, lectured and published papers describing this Actinomyees bovis, which Harz identified but could not culture. Israel 12 and Ponfick identified the disease in humans in 1878, but Actinomyces was not actually isolated from human material and injected into animals until 1891 by Wolff and Israel (Conant et aZ., 1971). An aerobic filamentous actinomycete resembling Actinomyees viscosus was first described by Gruter in Argentina in 1932 (Negroni et al., 1932). He cultured Actinomyces discofbliatus from.pus present in infected lacrimal concretions and suppurative gum lesions of a patient with an infection aggravated by an ill-fitting artificial palate. The cellular and colonial morphology resembled that of A. viscosus but the organism described could not hydrolyze starch or reduce nitrates. It did hydrolyze gelatin and produce indol from tryptophan. Catalase production had not been examined. These biochemical tests led the author to the conclusion that this organism was probably not A. viscosus. Brion (1939, 1942) isolated an organism from dog and cat lesions which also resembled A. viscosus. Georg et a1. (1972) noted that this organism, A. baudetii, was not preserved for future examina- tion, and Brion gave a description of metabolic properties of the organism too incomplete to adequately exclude the possibility that A. baudetii might have been an organism other than A. viscosus. Gerencser and Slack (1969) obtained cultures of A. baudetii from the Pasteur Institute in Paris and A. discofbliatus from P. Negroni in Argentina. Biochemical, morphological, and serological tests of these organisms demonstrated they were both.A. viscosus. Further investiga- tion showed these were not the original strains described by Gruter and Brion. Cultures which resembled only the strains' morphological 13 characteristics had been sent to Gerencser under these names. Filamentous microorganisms became an important topic for research in the field of dental medicine. Aerobic, non—acid-fast filaments isolated from human specimens were identified as Nocardia salivae2 in 1962 (Roth et aZ., 1962). Numerous other diphtheroids were also found in the mouth, and the possibility that these might cause disease was suspected. Hamsters orally inoculated with a suspension of Leptotrichia sp.,3 a filamentous organism found in dental caries, produced infections (Jordan et aZ., 1964a). Syrian hamsters fed high carbohydrate diets of confectioners sugar acquired periodontal infections (Jordan et aZ., 1964b; Keyes et al., 1964). The organism isolated from subgingival plaque in the infections was identified as Strain T-6. This strain was orally inoculated into uninfected hamsters and produced carious lesions from which T—6 could again be isolated. Howell and Jordan (Howell, 1963; Howell and Jordan, 1963) examined the morphological, colonial, and physiological characteristics of Strain T-6 and other diphtheroids isolated from hamster dental plaque. Their observations of an aerobic, catalase positive, filamentous "hamster organism" comprised the first detailed description of A. viscosus. The hamster organism was compared with R. dentocariosus, another catalase positive, aerobic, non-acid-fast actinomycete present in dental plaque. Only the "hamster organism" could be shown to induce periodontal disease. Rothia was beaded and 1Personal communication from M. A. Gerencser. 2N. saZviae was later called N. dentoeariosus and is presently Rothia dentoeariosa. 3Leptotrichia is now Baeterionema matrichotii. 14 did not produce long filaments seen in T-6. Howell (1963) mentioned that A. naesZundii had similar morphologic and metabolic properties as T-6; however, A. naeslundii was catalase negative and did not need additional carbon dioxide for aerobic growth, which T-6 did require. Nocardia and Rothia also grow well aerobically without C02. End- products from glucose fermentation by T-6 included acetic, formic, lactic, and succinic acids. Howell et a2. (1965) officially named the hamster organism Odbntomyees viscosus. At the time the basis for exclusion of T—6 from a new proposed family "Nocardiaceae" included lack of aerial hyphae, different cell wall components, inability to produce true spores, different atmospheric requirements, and fermentative ability. Howell maintained that production of catalase was a signifi— cant enough difference to justify the separate genus in the Actinomy— cetaceae family. Fluorescent antibody tests were performed and no cross reaction with A. isranii, but weak reactions with A. naesZundii, were seen. Supragingival and subgingival plaque in infected hamsters showed apical migration of epithelium, erosion of cementum, and inflammation of periodontal pockets, later progressing to erosion of the bone (Jordan et al., 1965b). Hamsters could become infected from direct plaque inoculation, close contact with infected rats, or from consuming large amounts of carbohydrates. Odbntomyces viscosus showed plaque formation (Jordan et al., 1965a). No evidence of soft tissue invasion in any of these infections could be shown. Some hamsters were pre- sented with enlarged cecums after several weeks, but no filamentous organisms were cultured from the cecums. 15 Melville (1965) isolated pathogenic plaque producing diphtheroids from the human oral cavity and identified these as A. naeslundii. He maintained that some of these strains of A. naeslundii were capable of producing catalase. This was the first reference of human infection with a possible strain of A. viscosus; however, lack of a full descrip- tion of the organism prevents verification. An indirect fluorescent antibody technique to demonstrate reac- tions to 0. viscosus showed reactions in one-half of the 14 patients examined by Snyder et al. (1967b). The ability of this test to identify 0. viscosus in patients is questionable since cross reactions occurred with four other organisms: A. isranii, 14/14; A. naesZundii, 8/14; E. dentocariosa, 6/14; and B. matrichotii, 14/14. No cultures grew 0. viscosus. Swerczek et al. (1968) examined eight cases of canine actinomycosis (Table 1). Seven of these involved large breeds. It was suggested that actinomycosis occurs frequently in dogs older than one year, especially hunting breeds, since they have a great probability of exposure to soil-contaminated trauma. Actinomyces sp. do not occur in soil; however, bacteria from.soil may be necessary to contaminate the lesion and debilitate the animal so an endogenous actinomycete can colonize in the wound. Lesions of all dogs contained granules that had non-acid-fast filaments in their centers. Gram stains revealed Gram- positive bacteria and filaments. A distinct border without any club formation differentiated the canine granules from loose, irregular granules seen in nocardiosis and clubbed granules of other actinomycoses. Although no cultural studies were performed to speciate the actinomycete, the author feels there is a strong possibility that the etiologic agent 16 .mnmofim pocket 60 unamlvwowlaoc .o>aufimoalamuu mums HH< « HHHHomn o>HuwmomIEmuu voNHa mHmoo oHeoomouowz nonuo auHB hHHmoHnouomam nacho Imaoafluom pouwamuosou mqmfim wasnmnomo u» n mamookfi oaeoomouoflz consuaso uoz Ionauom cOHmoH Hmoa>uoo hwwmm pumauom.um uh H mfiwooza oaaoomouoaz IoaHuom mmoomnm xamam emwoaw umxom um o mamouaaoaauom ofieoomouofiz cousuaoo uoz mmwomnm Hmunouoo mmwoam can“: H% N mfimoohaooauom uw>ofiuuom oamoowouoaz cousuaso uoz macheao owomuosa Namoam nowaou uaovm meoo oweoomouofiz cannuasu uoz luaocfiuom mmoomnm wand wmowm uwxom uh N mamoohfi oamoowouoez wounuaso uoz Iosfiuom um%o hum>fiamm «mmmn umnom H% m Hofinmmm owumq manuaso no: wasoo mHmoohaoaHuom maohaam Hmmmma umwafiuem uh m moasamuu «mamficmwuoouofiz mwmoawmflm ammo woman om< mo oufim mo mofiumfiuouomumnu mama ..No no xenouo3m an mmow w :H moHumHuouumumno wouuoemu umwmoomaonwuow mafinmo .H mHan 17 was A. viscosus in some of these cases. This is based on granule type, acid-fast reaction, and incidence of infections of dogs with various Actinomyces species. In 1967 the Subgroup on Taxonomy of’the Microaerophilic Actinomycetes amended the general description of the genus Actinomyces to then include catalase positive organisms (Georg et aZ., 1969b). It was suggested that 0. viscosus be changed to Actinomyces viscosus. Isolation of A. viscosus in human specimens of dental plaque was characterized and found to possess the same morphological and physiological properties as the hamster strains. Short rods in palisade arrangements and long dividing filaments were found in the same human isolates, and several strains which originally formed rough, heaped colonies later reverted to smooth, entire, colonial formations. Georg et al. (1969b) also used serological and cell wall analysis to identify these strains. The hamster organisms had the same cell wall components as the human strains but serological differences were apparent. Further studies on periodontal disease in hamsters and humans resulted in an interesting theory on colonization and growth require- ments. Gramrpositive organisms appear to be the predominant cultivable flora in the human oral cavity (Holmberg et al., 1972a). There must be some bacteria present in the mouth before colonization and the induc- tion of caries and periodontal disease can be effected. It appears that Streptococcus sanguis is present in early plaque formation in humans. As plaque continues to be deposited, Gramrpositive rods and filaments colonize, and St. sanguis disappears. The new flora consists of Corynebacterium, Actinomyces, and LactobaciZZus species. Cultures 18 of St. sanguis can inhibit growth of Actinomyces, Bacterionema, Cbrynebacterium, Nocardia and LactobaciZZus on the same culture plates (Holmberg et al., 1972a), which suggests a possibility that inhibition of diphtheroidal colonization by St. sanguis may occur in vitro in the oral cavity. Large amounts of sucrose catabolites which occur in organisms fed a high carbohydrate diet inhibit the growth of St. sanguis, and enhance the growth of A. viScosus., Actinomyces viscosus readily hydrolyzes sucrose and uses the resulting carbon and energy for other synthetic reactions. T—6, A. viscosus, when fed in large amounts to gnotobiotic hamsters and rats, will rarely produce disease symptoms other than carious lesions (Reed et aZ., 1972). The lesions found in gnotobiotic rats orally infected with A. viscosus or A. naesZundii are similar (Socransky et aZ., 1970; Frank et aZ., 1972). Actinomyces viscosus may be more pathogenic than A. naeslundii since orally infected rats fed high carbohydrate diets produced more dental lesions from A. viscosus than from.A. naeslundii. Actinomyces viscosus and A. naesZundii are not as pathogenic as the anaerobic actinomycetes. Few cases of noncarious type infections have been reported (Georg et al., 1972; Lewis et al., 1972; Adeniyi- Jones et aZ., 1973; Holmberg et aZ., 1972b). Experiments to Compare the pathogenicity of A. viscosus and A. naesZundii by intraperitoneal inoculation of laboratory animals showed slight differences between the two species (Coleman et aZ., 1969; Georg et aZ., 1968). Coleman et a1. (1969) described three experiments of hamster and mouse infections with A. naeslundii. Although A. naeslundii could be isolated from l9 lesions in many of the experimental animals, it was apparent that intraperitoneal inoculations with the obligate anaerobic actinomycetes were significantly more severe and more frequent. The severity of lesions in mice produced from single inoculations of several different Actinomyces sp. was examined by Georg et a2. (1968). Actinomyces naesZundii induced lesion formation in 104/119 (87%) mice and A. viscosus induced lesions in 23/24 (95%) mice. Actinomyces viscosus is known to be the causal agent of perio- dontal disease in hamsters, and it also has been isolated from the human oral cavity on numerous occasions (Georg et al., 1972). Georg et al. (1972) examined five strains of A. viscosus isolated from animal infections and experimentally inoculated these intraperitoneally into mice and newborn puppies. These strains were highly pathogenic to all of the mice producing abscesses on parietal and peritoneal surfaces which were restricted to the abdominal cavity. Most of the lesions involved the liver. Actinomyces viscosus produced lesions in only one of the six newborn puppies. The lesions were in the hepatic parenchyma and resembled those found in one of the dogs presented with the original infection. This is the only published case of a successful experimental infection in a large animal. Holmberg and Hollander (1972b) felt that the differences between A. viscosus and A. naeslundii were not significant enough to justify use of two separate species names. A calculated median organism was suggested that would include both species. Otherwise, Holmberg et al. recommended that A. viscosus be considered a variant of A. naeslundii. Their data for evaluation by numerical taxonomy methods did not include 20 cell wall analysis or serology, both of which are presently used to separate the species. Two cases of pathological involvement of A. viscosus in humans have been described. One was a 35—year-old male who suffered with a bilateral pleural effusion (Lewis et aZ., 1972). No organism was seen in stains of direct smears of aspirated material; however, Mycobacterium tuberculosis and A. viscosus were cultured from the specimen. The patient had a positive tuberculin skin test and, because he had tuberculosis, the amount of pathology due to A. viscosus could not be adequately described. The patient showed no signs of improvement when only antituberculosis drugs were administered, but did improve when bacterial antibiotics were added to the treatment. The other 19-year- old male was presented with a branchial cyst below the left ear (Adeniyi-Jones et aZ., 1973). The cyst contained hemorrhagic tissue resembling that seen in chronic inflammations and was filled with a yellow, nonodorous pus which was aspirated and stained. The Gram stain showed clumps of Gram-positive pleomorphic diphtheroid-like rods and some short filaments; some organisms had swollen ends or were filamentous. Only A. viscosus was grown in culture, and colonies on agar were of the smooth form: entire, glistening, white, convex. This, however, formed granular floccules when cultured in broth. Biochemical tests confirmed the identification of A. viscosus, and a positive reaction with the fluorescent antibody test at the Center for Disease Control (CDC) in Georgia was reported later. Holmberg and Hollander (1972b) stated that "no clear relationship of A. viscosus to human infections has been shown." Personal 21 communication with Dr. L. K. Georg has provided the author with two strains of A. viscosus isolated from a human lung biopsy and a lung abscess. Attempts to fulfill Koch's postulates with these strains from human specimens have not been made. It is thus impossible to unequivo— cally state A. viscosus can cause disease in man. Pathology of Actinomycetales The term "actinomycosis" usually refers only to Actinomyces infec- tions and nocardiosis is restricted to NOcardia infections (Carter, 1973). However, earlier studies often referred to both Actinomyces and Nocardia infections as "actinomycosis" (Swerczak et al.. 1968); thus, review of the literature on this topic is quite difficult. Actinomycosis and nocardiosis are chronic, suppurative granulomatous diseases. Nocardia sp. produces subcutaneous lesions and soft tissue abscesses, but pulmonary involvement is the most frequent manifestation in humans. Sinus tract formation with dissemination is rare, but a tendency for hematogenous spread especially to the brain is occasionally reported (Murray et aZ., 1960). Actinomyces sp. are frequently involved in subcutaneous, cervico—facial, thoracic, and abdominal infections (Conant et aZ., 1971). Sulfur granules frequently found in Actinomyces infections are not present in NOcardia infections. The granules in anaerobic Actinomyces infections are hard and gritty in texture, while those seen in Nocardia and A. viscosus infections are soft. The granules vary in size from a microscopic mass to a diameter of several millimeters. The cheese-like, yellow-gray granules of A. bovis and A. isranii contain colonies of filaments embedded in a polysaccharide-protein matrix with approximately 22 50% calcium phosphate. Sulfur granules are also found in maduromycosis, staphylococcal actinophytosis, and actinobacillosis. In hematoxylin and eosin stained histological sections, a mass of acidophilic, club- like projections radiate from the centers of the granules in a rosette pattern (Figures 1 and 2). This "clu " formation is possibly a mani- festation of an interaction of a cellular response of the host with the organism, not a product of the microorganism (Swerczak et aZ., 1968). The center of the granule contains colony—like masses of rods and filaments which stain Gram positive. Sulfur granules are usually sur- rounded by neutrophils and then enveloped by lymphocytes, plasma cells, macrophages, and fibroblasts in fibroproliferative tissue (Swerczak et al., 1968). The granules found in nocardiosis are unlike the granules seen in bovine actinomycosis. These Necardia granules lack club formations and appear to be a mass of organisms held together in a loose, unstruc- tured, acidophilic matrix. The surrounding suppurative reaction is similar to that of actinomycosis (Kalnins, 1963). Swerczek et a1. (1968) has described granules found in canine actinomycosis which he maintains are unlike Nocardia or bovine Actinomyces granules, but have a well defined border without clubbing. Anaerobic bacilli and various types of cocci were frequently found with Actinomyces sp. in the granules. Georg et a1. (1968) also noted the absence of clubbed borders on granules found in mice which were inoculated intraperitoneally with A. viscosus. Many Actinomyces sp. are saprophytes of the oral cavity of man and animals (Davis et al., 1973). Only A. humiferus has been isolated from 23 Figure 1. A. bovis sulfur granules with clubbed border. Hematoxylin and eosin stain; X 335. Figure 2. Higher magnification of A. bovis sulfur granules. Hematoxylin and eosin stain; X 1350. 24 the soil (Gledhill et aZ., 1969). Actinomyces eriksonii, A. isranii. A. bovis, and A. naesZundii are commonly encountered in infections while A. odbntolyticus and A. viscosus are considered nonpathogenic. Although Nocardia Species are 8011 inhabitantS. Necardia asteroides, N. brasiZiensis, N. caviae, N. farcinica, and N. dassonvillei have been isolated from purulogranulomatous nodular lesions (Thompson et al., 1951). Although no cases of transmission of actinomycosis from man to man or from animal to man have been reported, caution should be observed by individuals handling infectious specimens. Suspected modes of infection include trauma, passage through carious lesions in the gums, direct aspiration, or ingestion of the organisms in saliva. Lesions have been found in bones, thorax, abdomen, cecum, neck, skin, brain, tongue, soft tissues, and several organs (Conant et al., 1971)- Cervico-facial actinomycosis can involve abscessation of soft tissue or bony hard tissue, including the maxilla or mandible. The soft granulomatous tissue contains necrotizing abscesses which coalesce to form draining sinuses or fistulous tracts (Georg et al.. 1968). Hard, dense tumorlike nodules simultaneously form in cutaneous tissue. The purulent exudate has a characteristic yellow-green color. Eventually fibrous scar tissue replaces regions of the granuloma. Osteitis from proliferation of the microorganism.in the bone may cause a honeycombing effect due to rarefaction and proliferation of the bone. This may cause blockage of nasal passages and induce dyspnea. Dissemination of the organism to other tissues is effected by sinus tract formation. Sulfur granules are frequently found in cervico-facial actinomycosis (Smith et al., 1972). 25 Aspiration is probably the usual mode of infection in thoracic actinomycosis. Symptoms may parallel those of pneumonia: irregular mild fever, cough, emaciation, anorexia, anemia, with subsequent production of bloody mucopurulent sputum, pleural pain, dyspnea, or abnormal respiratory sounds. Acute thoracic infections may cause pleural effusions with draining sinuses and fibrosis. Radiographs may reveal the spread of the infection from the base of the lungs (Hurst, 1950). Occasionally apical lung infections may resemble tuberculosis; however, no discrete circumscribed tubercule-like lesions will be apparent. Pleural adhesions and fluid accumulation may also be seen on x-rays. In pleural actinomycosis, mediastinal or other lymph nodes may become swollen and necrotic (Swerczak et aZ., 1968). As fluid accumulates in the thoracic cavity, lung expansion is impeded. Pulmonary congestion, edema, and atelectasis gradually increase, and cardiac output may be reduced sufficiently to cause fluid accumulation in other body cavities which can debilitate the host and abet spread of the infection. Hyperplastic, fimbrionated villous proliferation of the thoracic wall causing pleural thickening is the result of a chronic tissue response to irritated pleural lining (Swerczak et al., 1968). Abdominal actinomycosis may result from ingestion of masses of contaminated saliva or from metastatic spread from cervico-facial or thoracic actinomycosis. Chronic abdominal dissemination to bones, liver, kidney, spinal column, lungs, and other regions has been reported. weight loss, vomiting, fever, and abdominal pain are nonspecific symp- toms which can also suggest appendicitis or abdominal carcinoma. Hepatomegaly, splenomegaly, and pyelonephritis may be apparent without 26 actual presence of infectious organisms in these organs. Any region of the gastrointestinal system may be infected with Actinomyces organisms. Suppurative burrowing abscesses and draining sinuses become embedded in fibrotic scar tissue, or are surrounded by yellowish lipid-filled macrophages. These regions should be examined for actino- mycotic granules. Presence of granules excludes a diagnosis of syphilis, neoplasms, tularemia, amebiasis, disseminated tuberculosis, typhoid fever, granuloma inguinale, or certain deep-seated mycoses (Conant et aZ., 1971). Lesions due to different Actinomyces sp. cannot be differentiated. "Lumpy jaw" in cattle caused by A. bovis resembles a similar infection in humans due to A. isranii. Experimental infection of large animals with different species of actinomycetes has been rather unsuccessful. Even the most pathogenic actinomycetes, A. bovis and A. isranii, have induced infections in less than one-half of the inoculated test animals (Georg et aZ., 1968). Reported human infections with A. naeslundii include one abdominal abscess, one gallbladder empyema, three septicemias (one associated with leukemia and pyelonephritis), several sinus infections related to tooth extractions, one leg abscess, and one post-thyroidectomy wound abscess (Georg et al., 1969a). Human infections with A. viscosus involved a branchial cyst abscess (Adeniyi-Jones et al., 1973) and a case of pleural effusion complexed with tuberculosis. Georg et al. (1972) isolated A. viscosus from one case of enzootic pneumonia in pigs, a lymph node abscess in a goat, and two cases of abdominal infections in dogs. Canine strain 1231 was isolated from an abdominal mass attached 27 to the pancreas, stomach, spleen, intestines, and kidneys of a six— month-old terrier. The dog was presented with signs of emaciation, vomiting and general malaise. Granules found in a biopsy had irregular margins but no clubbed edges. Granules yielded Eubacterium Zentum, Ehterobacterceloacae and A. viscosus in culture. The latter two organisms are considered normal intestinal flora. A six-month-old mixed breed dog was presented with signs suggestive of abdominal effusion: lassitude, anemia, emaciation, abdominal distention, pale oral mucosa, hepatomegaly, and a mild leukocytosis with a shift to the left. No sulfur granules were seen; however, numerous abscesses were observed on the liver. The latter were sur- rounded by yellow zones of lipid accumulation, fibrotic tissue, and hemorrhagic material. On histologic examination the center of the abscesses contained mostly dying hepatocytes and colonies of A, viscosus filaments. Nocardiosis is a disease found in cattle, man, dogs, cats, marsupials, horses and goats, and tropical fish (Bruner et aZ., 1973). Subcutaneous, abdominal and thoracic forms have been reported in varying frequencies from different animal species. Subcutaneous abscesses, empyema, pyopneumothorax and infections of the stomach, brain, kidneys, bones and heart have been reported in dogs (Bruner et aZ.. 1973). Definitive diagnosis requires cultural identification of the organisms since lesions closely resemble those seen in actinomycosis. There is frequently extensive vascularization of lesions particularly in villous prolifera- tion of serosal surfaces. Non-clubbed flake-like granules containing filaments may aid in a tentative diagnosis if the organism is partially acid-fast. 28 Confusion of A. viscosus with Other Organisms Reported infections with A. naeslundfii, A. viscosus and N. asteroides have similar pathological and morphological characteristics. Nocardia sp. appeared as the cause of most Gram-positive filamentous infections in dogs seen at the Michigan State University Veterinary Clinic during the past eight years (Table 2). These statistics, however, are somewhat weighted since all Gramrpositive, aerobic, fila- mentous organisms were thought to be N. asteroides until 1973, when knowledge of more precise techniques for identification of Actinomy- cetales were applied. Failure to differentiate A. viscosus from N. asteroides probably accounts for the low incidence of reported cases of canine actinomycosis. Not listing an acid-fast stain, improper decolorization of the stain, and knowledge that some NOcardia species are non-acid-fast may result in some microbiologists reporting all Gram-positive, aerobic, catalase-positive filamentous organisms as N. asteroides (Georg et al., 1969b). The absence of clubs in granules of A. viscosus has not been reported in most literature on actinomycosis; however, many references are found which cite the presence of non—clubbed flaky granules in nocardiosis (Langham et al., 1959; Murray et aZ., 1960), thus causing further confusion in identification of these two potential pathogens (Swerczak et al., 1968). Diphtheroidal bacteria may be a contaminant or, on occasion, a pathogen (Johnson et al., 1970). A diphtheroidal non-filamentous form of A. viscosus also exists which physiologically and morphologically resembles these bacteria (Adeniyi-Jones et aZ., 1973). 29 manuaso owuumm z I o uoa lawnmooo oHomuosa mmmqoa uovmunmq me\m snouaoo 4m monommlo mwmoficumooa . mo m0 assess .Am 2 I o mwuaamlo vmuaamumoow oowqoa mHHHou mo\¢ mfimofiwumoos Headmam unauaao HaD mumum cmwwnoflz .owafiao humnfluouo> .huoumuonmg thHOHn .N oHnMH 30 EswwwAAmoNb . mmmomnm mam uo>ofiuuom «assess .hm I I o ncImoqu Iofivumoos manwmmom mmomHH uovmunma no\m Ammwv + omouuxow + omouosm + Hows“ u z mommmros nolcmNIU mamofivumooa mxmnm .smuo INcm «wNoo .m z z o noImNNIo mwuwsouwuom mmNNHH omwuonwm no\m mHmOHvumoon o o o mmmumnm Hosa>uoo mowmoa mavoom oo\0H musuaso 4m ownoummsw no mo mmoomnm .mmspm oanouownm mwmofioumvoa «mmmwosmuuom z o o colmaqlo umasnficnmaasm omqmoa noxom oo\m BEES Um quoo .m qofiwuopNSE ooImeIo .m «mxuooooauwx_ ooImemIo mHmOfio quuoouopmmsem m oolmqmlu Immoos whoa .em unannonm .maoooootamAAm » I I I oeIeGSIo was mesa ..uaoa:m maoooa assume eo\a ousuasu uoo oownaa .um mo\m mumo owNeNH ouaamamz om\¢ mumo mmmmma amaawam oA\~ ousuaso nos fiasco manoeuumooa vase: mwumnowav mo we z I z molmoeIu mwufiaoaaooumo OHmHNH uoxamz Om\a mfimowvumooa ummeuHom mmmomnm Icon we musuaso z z I moIAHHHIu sumaaaxmansm mmnsma «Hammm ae\ma .amuw ousuaso uoa moImwon mfimowwumooa Houswom soHHow fiasco sneeze .cm I z I moImmeo Mononuo%m ooomma emfiawsm mo\m mamoauumuon maoasnwuw venom o I o “who sum>aamm «mamma “means mo\e vwsam osmmHu mo me u z summosmnosm HmlnooIo mamoawumoos wuonnonm .aswoosazs..m z o z aoINHMIo xenonuoam HHHmHH amaumo ao\a muoo HumoNH .um HN\~ mHmOHuumooo wmoomnm uo>mwuumm ousuaso «m mo mu 2 I I onImmHHIU humaawxmansm mnomma “command HB\H mamoavumooa z mHuHHmnmooaoowsHaoa o manfimmom o msoumaoassmuo Howmma meadow om\~H casuaso ea mwmowvumoos scum “on sauce I + z oAIAmoHIo ammo sum>aamm Hmumma «agenda: OA\HH mamowvumoos vumchom scum uon fiasco I o z OnImooan mmoomnm Hmow>uou HNNmNH .um ON\0H ummMIvHUMIsos m «masses .nm mHmOkumooa rumoumm assesseucememsob I o z anIoOAIU unwound unasnavnmz wmomma .um em\o m «unease .em «.mmsem m «masses NnIomeo Iosow2_qumeo&m NnImeIo mamaaoumoos wuonnom «newessuwwmm .pm I I I NnImmnIo mmoumnm unasnawnmz. Nmmmma .um ~m\HH I How .I won: .muoo onmNH Omaha -\o .amuw msouoooemmsnm NBIHmNIo mwmofivumoon msoa mama wonpaau > quuuouosowz I z o Nelnmmlo Inseam umasnfiwnnz NHNHMH umouu NN\¢ mwmoavumoos mama susouw on I o z HeloomIQ mmoomnm HwUH>HoU meema awoke N~\N mHmOHwHMUOd Hofiomem .smuw I I z HmImmnIU maoasnmuw oquma noxooo HN\NH mamoapumooa auaouw on I o z HaImooIo mmmomnm Hmuoaama memma anon ewe HA\¢ uqu ausouw o: I o z ANIHmHIo «scanner» Hmoapuoo Hosoma gaseous HA\¢ moaonmuw sumo thHOHnouon uuomom hwoao “swam .oz ammo mHmonmmHQ .oz ammo vomum mama mwo Isume hwo hMOHOfiA Hmoflafiao ofidfiao IHoan IoumHm IHoumo Iouoflz IOHOHE Aemaaauaoov N magma 35 mxoooown .n I momma o>amaouxo I Nae mwwamooz mmwwaooon I oasu memos ao mamoohaonauoo Iaso nn no mo .noaauon «UNNmaxoaoom Iowa .w I z MmIamIo mmoomno Hoaoeaoe Hmnwma aowoanon mm\a oamoawaooos oaaaon z z o QISIE 23283 $2.2 3338 2: manuaso 4n no oaooawaooon wannaon 8 «808832 2 z z fiImomIo 83:55 1328 332 .3 ~23 moHsnoau ouon hwoaoanoaoaz uaomon hwoao aooam .oz omou oaoonwoan .oz mono woman ouon hwo Inuoe hmo awoaoan Hooasaao oasaao IaoE Ioumam $8.8 Ioauaz Ioaofiz Awmsaaunoov N oHAoH .moasnoaw u .soaw ”ammo m.w:oasonom u nauowmd n I momwadooz u 2 an woaaomaoe wumou mooaoaos memos moons mamooaaosauuo wnsonxan I manage»; .w Inca .w Ioecsoe I saIsman amoa>umu mmsaaa amawmzuoz «A\a wmaaonaom mumou msoaosba «o. I 23839 a. 288 QIRSIU 380E838 38m ocwcoszoxm 4M Iowa .n z z mmInnoHIo oaoansn mmmmea aoxao: m~\aa mmoomno aouumm nusoaw on I I z mnlqooan nooam anwfln maqmqa awaaH mm\HH mnnaosmnon mswz «oxuoomoamfin mamoawaoooa noeomrwaxaon .NQ I I z mnlmonIo oasuman man Nomqea woxaz mn\m moaoaoaw ouon awoaoanoaoaz uaomon hmoao aooam .oz omou oamonwoan .oz moon woman ouon hwo Inuoa mwo hwoaoan Hosanaao oafianu IHoan Iouman IHouau Ioaoaz Ioaoaz Awosnauaoov N manoh 37 Treatment of Canine Actinomycosis and Nocardiosis Because there have been so few cases of canine actinomycosis diagnosed, there is very little information on the efficacy of treat- ment. Actinomyces viscosus is sensitive to most antibiotics. Penicillin is the drug of choice. Chloramphenicol, oxytetracycline, erythromycin, clindamycin, ampicillin, and chlortetracycline are also very effective (Lerner, 1974). Streptomycin, rifampin, and novobiocin are not as potent against these infections as many broad-spectrum antibiotics (Adeniyi-Jones et al., 1973; Murray et aZ., 1960; Davis et aZ., 1973). The latter drugs are usually given intravenously or orally for two or more weeks until no signs of active infection persist. Surgical excision of infected tissues and drainage of wounds cannot adequately eliminate infection from fibrotic scar tissue unless antibiotics are also used. Topical application or flushes with iodine solution are recommended in treatment of subcutaneous abscesses. Antibiotic therapy has not been too successful in progressive cases of canine actinomy- cosis when there is extensive damage to organs and development of fistulous tracts. Cure depends upon early diagnosis and treatment of the infection. In vivo Nocardia infections are resistant to penicillin and other antibiotics. Treatment usually includes sulfonamide drugs, but broad- spectrum antibiotics, streptomycin, novobiocin, or erythromycin are often prescribed along with the sulfonamides (Murray et al., 1968; Kalnins, 1963. Treatment of nocardiosis may have to be carried out for several weeks. Advanced infections respond poorly to antibiotic 38 therapy, so veterinary clinicians frequently recommend euthanasia. Thus proper diagnosis of the etiologic agent is very important. CLINICAL HISTORIES OF CANINE ACTINOMYCOSIS SEEN AT THE MICHIGAN STATE UNIVERSITY VETERINARY CLINIC JUNE 1972 TO NOVEMBER 1973 Clinic Case 129820 (CF-470-72) The referring veterinarian had been unsuccessful in three weeks of treatment of an abscess on a two and one-half-year-old male Great Dane. Eighteen months earlier the dog had been diagnosed as having bilateral osteochondritis dissecans of the humoral heads. No treatment for osteochondritis had been initiated. Inflamed tonsils and a cervical swelling were noted during the preliminary examination at the Veteri- nary Clinic on June 5, 1972. Tentative diagnosis suggested a salivary cyst. The enlarged mass was below the right cervical region, posterior to the mandible. The pathologist examined large, soft, mucoid granules present in the infected tissue and noted the presence of Gramrpositive, filamentous, "Nocardia-like" microorganisms. The microbiologist cul- tured a "probable Nocardia species" which was non-acid-fast, urease negative, gelatinase negative, and did not grow on Sabouraud's agar. A diagnosis of nocardiosis was made. The mass was excised, drained and flushed daily with Betadine solution,1 and Bactro-Vet2 was administered orally. Radiographs of the chest did not indicate thoracic infection. The University Clinic and 1Providone-Iodine, Purdue—Frederick. 2Sulfadimethoxine, Pitman—Moore. 39 40 the referring veterinarian have had no further admissions with this animal. Suggested route of infection through the mouth, presence of mucoid granules, and atypical cultural reactions left some doubt in the final diagnosis of nocardiosis. The strain was sent to Dr. L. K. Georg at CDC in Georgia for positive identification. The strain was dead when it arrived at CDC, but fluorescent antibody studies and catalase reaction were examined. Georg said this strain was catalase- negative and did not react with A. viscosus fluorescent antibody. At that time, information on identification of A. viscosus was brought to the attention of the microbiologist and pathologist at the Veterinary Clinic. Clinic Case 138631 (CF-99-72) This seven and one-half-year—old female Labrador Retriever had a history of false pregnancy, ovariohysterectomy, epilepsy, and a single episode of dyspnea due to inhalation of phlegm during one of the seizures. Slightly elevated blood sugar, 116 mg/100 ml, and a blood urea nitrogen of 22-24 mg% were also noted. Medication included Mebroin1 and Mlyelepsin2 prescribed by the family veterinarian for seizures, later substituted with Dilantin3 and Primidone4 prescribed by the MSU Veterinary Clinic. The family veterinarian had administered lDiphenylhydantoin, Winthrop. 2Primidone, Fort Dodge. 3Diphenylhydantoin, Parke, Davis & Co. 4Primidone, Ayerst. 41 broad-spectrum antibiotics as treatment for a temporal abscess which appeared on the animal during hunting season. The abscess seemed to have started on the nose and moved upwards towards the left eye causing a swelling on the left side of the head. Initially the swelling was firm but, after several weeks of intermittent treatment, it softened. Purulent material was aspirated from the masseter muscle region. The abscess was present for two months when first examined at the MSU Veterinary Clinic on January 1, 1973. This involved the left temporal and periorbital areas, causing the left eye to be nearly swollen closed. Part of the mass was fluctuant, part firm, and the entire mass was very painful and quite warm to touch. Poor pupillary response in the right eye, enlarged cervical lymph nodes, and dandruff were also noted during the initial examination. Marked eosinophilia, l738/mm3, and a slight increase in blood urea nitrogen, 22 mg%, were among the laboratory findings. Pus from the abscess contained numerous macrophages and polymorphonuclear leukocytes, in which abundant Gramrnegative bacilli, Gram-positive bacilli, and Gram- positive pleomorphic beaded filamentous microorganisms were seen. An acid-fast stain was negative; however, the pathologist remarked that the filaments were suggestive of Nocardia species. Heavy growth of Actinomyces viscosus and a lighter growth of AanZigenes fecaZis, Mima polymorpha var. oxidans, and a PasteureZZa species were cultured from the exudate. Chloramphenicol succinate administered both subcutaneously and orally caused a marked reduction in the size of the temporal mass within approximately two weeks, despite the presence of a Chloramphenicol E“ 42 resistant strain of Mima polymorpha. The abscess was lanced and drained, and the dog was discharged from the clinic on January 26. Chloramphenicol and Primidone therapy were continued after discharge. Follow-up examination one week later revealed the mass had shrunken to about one inch. Surgical removal of fibrotic tissue and a fistulous tract which extended to the skull bones and curved towards the maxilla was undertaken. Chloramphenicol was continued for another week. Except for drainage of surgical seromas, no further complica- tions or recurrence of infection have been reported. Clinic Case 141564 (CF-295-73) A three-year—old male English Setter with a history of hip problems resulting from being struck by a car was admitted to the Clinic on March 6, 1973. The dog had been emaciated and dyspneic for the preceding two months. The symptoms seemed to appear at the end of an active hunting season. Hookworms, mild fever (104 F.) and problems with urination had been reported by the referring veterinarian, who had x-rayed the chest region and had been treating the dog with tetra- cycline. Radiographs revealed an obscured mediastinal area which suggested fibrotic attachments forming adhesions. Mild anemia with a hemoglobin value of 10 mg% and a packed cell volume of 30% was evident. Preliminary examination at the Clinic revealed abnormal lung sounds, deformed hips, pale mucous membranes, and anemia. A normal white blood cell differential count with a slight terminal increase in monocytes was reported. Thoracic fluid was aspirated and examined microscopically. This contained a moderate number of degenerative 43 polymorphonuclear leukocytes and macrophages. Microorganisms seen in the fluid included coccobacilli, cocci, and Gramrpositive branching filaments "morphologically identical to Nocardia asteroides." Actino- myces viscosus and Clostridium perfringens were grown in culture from the exudate. A series of radiographs were taken. The first showed compressed lungs floating on top of serosanguineous exudate. When the dog was in a prone position the lungs remained in the dorsal region, apparently attached there by adhesions. There was a moderate thickening of the pleura probably due to villous proliferation and fibrosis. Gunshot— like objects were visible in the dorsal area near the spine. The lungs remained clear. After large amounts of fluid were aspirated from the thorax, radiographs again were taken which demonstrated slight clearing of the thoracic cavity; however, some fluid was still evident. One week later, radiographs showed that the condition had stabilized, but pleural thickening and adhesions were still prominent. The expansion ability of the lungs was restricted to about 60% because of these adhesions. A diagnosis of pyothorax due to A. viscosus was made. Treatment included oral tetracycline during the first and fourth weeks of therapy. Daily injection of penicillin-streptomycin intra- thoracically and Combiotic1 intramuscularly was maintained for about three and one-half weeks. Response to antibiotic therapy was 1Penicillin, dihydrostreptomycin, Pfizer. 44 excellent, and the dog was discharged from the Clinic on April 11, 1973. The local veterinarian prescribed sulfadimethoxine for three weeks. One year later the dog was reported to be in good condition. He participated in a successful active hunting season during the fall of 1973. Fibrotic tissue was most likely reduced and lung capacity increased. Clinic Case 131568 (CF—492—73) A five-year-old male English Pointer was presented with a small flank abscess in January 1971. The local veterinarian lanced and drained the abscess on January 17 and February 5, and penicillin was administered orally and by injection. The dog was presented with multiple abscesses in the flank area February 25. The abscesses were drained, penicillin was administered by injection, and tetracycline was given orally. The animal was admitted to the Veterinary Clinic on March 10, with multiple flank abscesses and a circumscribed swelling on the left thorax. The general condition of the animal was fair. Radiographs of the left flank region showed an increase in density of the soft tissue, suggestive of swelling; however, no evidence of fistulous tracts was apparent. Injection of sodium iodide through a polyethylene catheter into the abscesses revealed two fistulous tracts in the next series of radiographs, one anterior to the left femur and one anterior to this, extending to the thirteenth rib. These tracts were confined to subcutaneous and muscular layers of the abdominal wall, apparently ending retroperitoneally in the region of the kidneys. 1Personal communication, Dr. R. G. Schirmer. ‘F' 45 Chest radiographs showed a normal thoracic cavity with the exception of what appeared to be three small lead pellets embedded in subcutaneous tissue. A Laboratory reports included isolation of Staphylococcus aureus and Necardia sp. (the latter did not grow on Sabouraud's agar) from flocculent material in the serosanguineous aspirate and the presence of large numbers of hookworm and whipworm ova in the feces. Normal levels of hemoglobin, total protein, blood urea nitrogen, and packed cell volume were noted. Blood cell differential counts showed a slight increase in nonsegmented neutrophils, 420-704/mm3. The body tempera- ture was 102 F. The abscess was excised and drained. Histological examination of excised tissue showed increased vascularization, suppura- tion, scar tissue and fibrotic connective tissue in which foci of macrophages, plasma cells and lymphocytes were present. No organisms were seen in the smears, but the pathology report mentioned the presence of Nocardia sp. (beaded, Gramrpositive filaments) in white flakes found in the exudate. Following surgery, the wounds were flushed daily for several days with Betadine solution. Oral administration of tetracycline, chlor- amphenicol, and ampicillin was prescribed sequentially for two weeks, and the dog was then discharged with a lO-day prescription of Bactro-Vet. The animal was readmitted on April 23, 1971, for surgical debride— ment of a newly formed abscess in the same area. The fistulous tract was again opened and granulomatous material was excised. Pathological examination again revealed filamentous beaded organisms "suggestive of Nocardia." The blood differential appeared essentially normal. The E 46 animal was treated with oral tetracycline for one month, and the wound was flushed daily with Betadine solution. The last admission of this dog to the Veterinary Clinic was on May 8, 1973. Reported fluctuation under the left flank and right rib cage associated with severe pain suggested regression. Occasional mucoid stool production and a slight abnormal gurgling or crepitation in the left dorsal region of the lungs was reported. Radiographs revealed a subcutaneous swelling over the rib cage of the right side. No internal thoracic lesions were noted and the lungs remained clear. The cause of a fibrotic change at the eleventh, twelfth and thirteenth thoracic_vertebrae could not be explained. Urinary tract function remained normal. Marked eosinophilia, 1848/mm3, and basophilia, 420/mm3,_were apparent five days after surgical debridement. Wright's stains of infected tissue showed large numbers of macrophages, neutro- phils, and occasional eosinophils and lymphocytes. Cultural studies provided positive identification of Actinomyces viscosus. This led the clinician to the conclusion that earlier iso- lates were also A. viscosus since earlier cultural techniques were inadequate for definite identification. Methods for differentiation of Actinomyces and Nocardia were not made known to the laboratory until 1973. The dog was again treated with oral tetracycline and Betadine flushes for 10 days. No further remissions have been reported. Clinic Case 145833 (CF-1078-73, CF-1097-73) A six—year-old male walker Hound with signs of labored breathing and thoracic empyema was admitted to the Clinic on November 3, 1973. The animal had been recently treated for diarrhea with administration 47 of Chloramphenicol and Kaopectate.1 Radiographs revealed an increase in opacity of the thorax. In the same kennel a hound had died one month earlier. This hound suffered with an acute onset of dyspnea, malaise, and fluid in the thorax. Necropsy findings included the presence of red-tinged fluid in the pleural cavity. The lungs were consolidated, only 50% functional, and were covered with white patches suggestive of a possible fungal infection. No cultural studies were performed. The preliminary examination of this second walker Hound suggested a tentative diagnosis of pneumonia. The animal was emaciated and dyspneic, and moist rales were heard in the dorsal area of the thorax. No sounds were heard in the ventral thorax, suggesting fluid accumula- tion. Blood cell differential indicated an inflammatory reaction: leukocytes 14,100, monocytes 423/mm3, eosinophils 423/mm3, lymphocytes 2115/mm3, segmented neutrophils 1551/mm3, nonsegmented neutrophils 9588/mm3, 2 nucleated erythrocytes/100 leukocytes. Fluid was aspirated from the thorax and examined microscopically. It consisted of large numbers of neutrophils and fewer mononuclear cells which had phagocytized Gramrpositive beaded and branching filamentous organisms. The pathologist suggested the filaments resembled the Nocardia-like microorganism similar to ones seen in other cases of canine pleural effusions. Radiographs revealed fluid in the abdominal and thoracic cavities. Pleural thickening was seen and changes in interstitial areas in the lung were questioned. The possible 1Kaolin, Pectin, Upjohn. 48 interstitial changes could have been indicative of consolidated, patchy granulation tissue.1 The animal became more dyspneic and died five days later.. Necropsy reports suggested a diagnosis of a Nocardia infection. Thoracic and abdominal cavities were distended and filled with fluid. Extensive villous proliferation in the pleura and pericardial sac, suggestive of a chronic infection, showed progressive replacement with granulation tissue. Only a small dorsal portion of the lung had not collapsed. The foul-smelling, red-tinged exudate restrained the lungs thus impairing expansion and contraction. This induced gradual atelectasis. Abundant gray-white, soft, sulfur-like granules were found in the lung (Figure 3). These ranged from one millimeter to five millimeters in diameter. There was a conspicuous lack of clubbing around the border of these granules, and filamentous microorgan- isms were dispersed throughout the granules. No granules were found in the abdominal tissue or in the white-brown abdominal fluid. Hepato- Figure 3. CF-1097-73. Histopathology of A. viscosus granules from lung tissue. megaly was observed, Gram's stain; X 1350. but the cause of enlarge- ment was not investigated. 1Personal communication, Dr. R. G. Schirmer. 49 Microbiological examination of thoracic fluid, specimen 1078, and lung tissue, specimen 1097, revealed A. viscosus as the filamentous organism. Gramrstained granules indicated the presence of Gramr negative bacilli and Gramrpositive cocci which did not grow in culture. A diagnosis of pyothorax due to A. viscosus was made. MATERIALS AND METHODS Organisms Four cultures of Actinomyces viscosus Strains CF-295-73, CF-470-72, CF—492-73, CF-99-72 were obtained from Mr. H. McAllister, who cultured purulent material on trypticase soy agar (BBL) with 5% bovine blood. Strain CF-1097-73 was cultured from granules obtained from lung tissue of a dog with pyothorax and strain CF-1078-73 was cultured from thoracic fluid aspirated from the same animal. Culture Media Media were prepared anaerobically on the Virginia Polytechnic Institute (VPI) Anaerobic Culture Media Apparatus (Bellco, Inc.) using the procedure described in the VPI manual (Holdeman et al., 1972). The Hungate-roll tube (Hungate, 1950) was not used since colonial morphology is difficult to observe in these tubes. Instead, flat- bottomed 100 ml prescription bottles were filled with 15 ml of media, under an atmosphere of 90% nitrogen and 10% carbon dioxide, and a rubber stopper sealed the opening of the bottle. This anaerobic apparatus provided a constant flow of oxygen-free gas during preparation of media and transfer of cultures. Traces of oxygen were removed by passage of gas through a cold catalyst (Dexco). The Eh value of media prepared by this method is about 150 mv., Liquid media were prepared 50 51 anaerobically on the same apparatus by the techniques described by Holdeman et al. (1972), dispensed in 5 ml portions into 18 mm tubes and capped with rubber stoppers. Aerobic media were of the same composition as anaerobic media but lacked the reducing agent, cysteine hydrochloride, and the Eh indicator, resazurin. Studies of aerobic growth in an increased carbon dioxide atmosphere were performed by placing aerobic cultures in a torbal jar, then evacuating and flushing with carbon dioxide. All cultures were examined after incubation at 37 C. Isolation Methods Primary isolation from clinical material was made by crushing granules washed with saline solution and streaking the material onto anaerobic and aerobic supplemented BHI-blood agar,1 while the exudate was directly streaked onto similar media. Gram stains of the crushed granules and exudate were prepared and then examined microscopically. Specimens were then frozen at -40 C to preserve for future use. Preservation of Stock Cultures Seven-day stock cultures were preserved in 0.5 mi defibrinated blood, fast frozen with dry ice and stored at -40 C. Cultures were also suspended in mist dessicans2 after 10 days' inoculation on sup- plemented BHI-blood agar and then lyophilized. 1Five percent bovine ox blood in Difco BHI agar supplemented with vitamin Krhemin solution and yeast extract (Holdeman et al., 1972). 2One part 6% glucose solution, 3 parts horse serum, 1 part BHI broth. 52 Morphologic Studies Strains of A. viscosus were inoculated on supplemented BHI agar, supplemented BHI-blood agar, then incubated at 37 C under aerobic, aerobic with CO , and anaerobic with CO conditions. Colonial mor- 2 2 phology was examined at 24 hours, 48 hours, 7 days, and 14 days with a dissecting microscope at 40X. Microscopic examination of colonies from supplemented BHI-blood agar incubated anaerobically with CO2 and aerobically with CO2 was performed after growth at 48 hours and 7 days. Supplemented BHI broth1 cultures were examined at 24 hours', 48 hours' and 7 days' incubation under aerobic with CO2 and anaerobic with CO2 atmospheric conditions. Gram stains of 48-hour and 7—day broth cultures were examined microscopically. The amount of growth was estimated visually as trace to 4+. A Kinyoun modified acid-fast stain (Georg et al., 1961) was per- formed on 4—day cultures grown aerobically with CO on supplemented 2 BHI—blood agar. Control stains of Nocardia brasiliensis and Strepto- myces madurae were stained for comparison with the Actinomyces specimens. Biochemical Tests Biochemical tests were done by the procedures recommended by Holdeman et al. (1972) using a basal liquid medium of peptone yeast extract with 5% or 10% carbohydrate concentration for carbohydrate fermentation tests. Carbohydrate media were dispensed in 5 ml amounts 1Difco BHI broth supplemented with vitamin Krheme solution (Holdeman et al., 1972). 53 into test tubes and autoclaved at 110 C for 12 minutes. Carbohydrates in a concentration of 1.0% included: cellobiose, dulcitol, fructose, galactose, glucose, inositol, inulin, lactose, maltose, mannitol, mannose, raffinose, rhamnose, salicin, sorbitol, starch, sucrose, and xylose. Those diluted to a final concentration of 0.5% were: adonitol, arabinose, melibiose, ribose, and trehalose. Reagents and media for indol production, catalase, nitrate reduction, methyl red reaction, starch hydrolysis, gelatin hydrolysis, urease production, acetylmethyl- carbinol (Voges-Proskauer) production, and aesculin hydrolysis tests were prepared according to the techniques described in the Anaerobic Laboratory Manual (Holdeman et al., 1972). Bacto litmus milk was prepared according to directions provided by the manufacturer. Inoculum for all tests was made from 2—day-old cultures grown aerobically with CO2 on supplemented BHI—blood agar. A heavy suspension of A. viscosus was prepared by mixing the culture in physiological saline, and four drops were added from a Pasteur pipette to each tube of medium. The litmus milk tube received seven drops of inoculum to assure a strong visible reaction. All tubes were incubated in a candle jar for 10 days at 37 C. Carbohydrate fermentation was determined by measuring the final pH of the media on a pH meter (Model 265 Instrumentation Laboratory, Inc.). A pH of 6.2 or lower was considered positive, and 6.3 or above was a negative reaction. Ribose, xylose, and arabinose had to have a pH reading lower than 5.7 before a reaction was positive, since CO2 tends to lower the pH of this medium (Holdeman et al., 1972). Inter- pretation of results of other biochemical tests was by standard 54 procedures described in many manuals (Holdeman et al.. 1972). All tests were repeated once. Isolates which did not give results con- sistent with those reported in literature (Table 3) were again repeated using freshly made test media. Carbon Dioxide Requirements The basal peptone yeast extract was prepared under pure nitrogen gas which had been passed through sodium hydroxide pellets to remove trace amounts of carbon dioxide. Procedures for inoculation of media and preparation techniques were similar to those described in prepara- tion of anaerobic media with C02. Growth was examined after seven days and estimated visually as trace to 4+. Comparison of aerobic growth, with aerobic plus CO2 growth, was made on seven-day broth cul- tures, supplemented BHI agar and supplemented BHI agar with 5% bovine blood described in the section on morphological studies (page 52). Estimated value of carbon dioxide in the atmosphere is .03% (Spector et al. , 1956). However, presence of technicians and CO2 containing apparatus in the laboratory may have increased the concentration to 0.5% CO . Carbon dioxide added to the torbal jars was not quantitated, 2 but more than 10% was estimated to be delivered. RESULTS Morphological Characteristics All strains of Actinomyces viscosus were Gram-positive, diphther- oidal rods or filaments, and were non-acid-fast. A slight preference for supplemented BHI agar with blood over supplemented BHI-agar was seen. However, BHI agar supported much better growth of A. viscosus than trypticase blood agar. No strains grew on Sabouraud's agar. Two morphological types were seen: a smooth entire convex glistening mucoid form (Strains CF-295-73, CF-492-73, CF-99-72, CF-1078-73) and a dry granular heaped irregular colony type form (Strains CF-1097-73, CF-470—72, CF-492-73, CF-99-72, CF-1098-73). Complete reversion from smooth to rough morphology was seen with Strains CF-492-73 and CF-99-72 (Figures 4a and 4b). Other strains occasionally showed a partial change in morphology; however, this did not persist in subculture. Smooth colonies from broth and agar cultures microscopically demonstrated a predominance of diphtheroidal bacilli-type forms with rare or no filaments (Figure 5). Diphtheroids were arranged in X, Y, V and S shapes, occasionally forming knobs and clubs in older cultures. No microcolonies were seen in these forms. The broth and agar cultures of granular colonies formed spidery microcolonies which microscopically consisted of a mass of tightly interwoven filaments and diphtheroids (Figures 6a and 6b). Branches, X, Y, V, and S forms were common. 55 56 Figure 4a. CF-99-72. Reversion from rough to smooth colonial morphology. 2X. Figure 4b. CF—99-72. Reversion from rough to smooth colonial morphology. 2X. 57 Figure 5. CF-l94-74. Diphtheroid form of A. viscosus associated with smooth colonial morphology. Gram’s stain; X 350. Figure 6a‘ Figure 6b Figures 6a and 6b. CF-470-72. Filamentous form of A. viscosus associated with rough colonial morphology. Gram’s stain; X 350. 58 Knobs, clubs and highly irregularly-shaped filaments became more frequent in older cultures. Beading was often seen in Gramrstained and acid-fast-stained smears when long thin filaments were examined. Filaments were shorter, diphtheroidal and coccobacilliforms were more frequent, and pleomorphic shapes, knobs, and clubs were rare. Broth cultures associated with smooth colony formation showed ropy growth while broth cultures from rough colony forming strains demonstrated a loose flocculent, granular type of growth. Anaerobically prepared media with 10% carbon dioxide showed slight or no differences in amount of growth when compared with aerobic cultures; however, studies of broth cultures prepared with pure nitrogen demonstrated signifi- cantly less growth. All pure anaerobic cultures demonstrated very weak growth, but total inhibition was not evidenced. Biochemical Tests All organisms were catalase positive, indol negative, gelatinase negative, hydrolyzed aesculin, hydrolyzed starch, reduced nitrates to nitrites, urease negative, acetylmethylcarbinol negative, methyl red positive, and produced acid reaction in litmus milk (Table 3). The strains grew well on all carbohydrates. Carbohydrates fermented included fructose, galactose, glucose, lactose, starch and sucrose. Variable reactions were obtained from fermentation of glycerol, inulin, maltose mannitol, mannose melibiose, raffinose, ribose, salicin, sorbi- tol, trehalose, and xylose. Isolates did not ferment adonitol, arabinose, cellobiose, dulcitol, inositol, and rhamnose. 59 (‘1 I I E . I I I I I I 3383 I I I I I I I I I moosanoan “odoauouuoaamw ouoawhnonaoo Bow 33 30m 33 38m 38m 32. 38 3% ads 353 I I I I I I I I I oHowaH A+VI AIV+ + + + + + + + .PE 5983 + TV+ + + + + + + + .PE 3383 I I I I I I I I I moosauofloo I I I I I I I I I momma: + AIV+ + + + + + + + woa oumauaz I I I I I I I I I aosoxmoamIoowo> + + oz + + + + + + woa Hanson I + + + + + + + «+ omoaouou «evenness: «383.8 i 2 ma 2 ma 2 2 38.. .w .n Iqu Inmoa Inmoa lam Imma Ian Imam Iao Iao Iau Iau Iao Iao Iao , mnaoaum Hoasuoamuaa ea woaao wwwsxmomo: .n woo osmomowa .w no onoau Iooma wouaoaoa saws am: am wouoaooa onaoaum osmomowe .< no woawsum Hoofiaonooan no sooaaonaoo .m manoH 60 IHINHHHLW + + oz + + I + + + omoaaoamz + AIV+ oz I I + I + I omoaoaaoz + AIV+ oz + I I + I + omooamz I I I I I I I I + Houaoooz + + + + + + + + I omouaoz + + + + + + + + + omouuoo > > oz + + I I I I oaaooa + > oz I I I I I I Simona + + + + + + + + + omouoao I > oz + + I + I + aoaoozao I TV+ + + + + + + + 33038 + + + + + + + + + mmouooan I I I I I I I I I Houaoasn I A+VI I I I I I I I omoaooaaou wwwexNooo: monsoon: «a mm mm mm mm mm mm umoa .w .n Iqma Inaoa Innoa Imm Immq Iona Imam Inc Inc Inc Inu Inn Inc Ino moaoaum Awoooauooov m manoa 61 .xoos n 3 "manoaao> n > “woaaonaoo uoo ammo n nz mooauoooa o>auowoo n I mooauooma o>auamoo u.+« .Aooauooaosaaoo Homemaoo .oaooz .o .n .3 “mama ..Nm um ooamwaon "mnma ..No no waomn mawma ..Ns em ooassmou “memo ..Ns um HangamHmw "memo ..Ne no nomom “memo ..Nu em Houses “memo ..os um aaoaomv moawouo Hoao>om scan ouasooa no moafiooosoo mam wwwranmmo: .n woo osoooowo .n no mooauoooa wouaoooa may .wwwexnmom: .n woo mxoooovo .w on maoauuooa wmuoooxo waowooum onu woamoooo ea scamsmooo nose ou on o>on moaaouoaoooa uooaonmaw oa wooasoxm moaoaum woo moooaonuou Houoosaaooxm oa mooauoaao> a > > + + + I + + + oooaonoaa I I I I I + I I I omoazN + + + + + + + + + omoaoam > AIV+ + + + + + + + ouamum A+VI I I I + I I I I Houanaom A+VI AIV+ oz + + + + + I aaoaamm Aavl I + I + I I I I moonan I I I I I I I I I moooaonn seesxommu: ,mstuwea _ an an ma Na ma Na ma ammo .w .w Iqma lumen Innoa lam Immq IOna Imam Ioo Izo Iao Ioo Imo Iau Iou :woann Aemsaauaoov m manna DISCUSSION Three questions must be considered when evaluating the clinical importance of A. viscosus: 1. Is it possible to differentiate A. viscosus from other microorganisms? 2. Is accurate speciation of this microorganism of clinical importance to the microbiologist? 3. Does isolation of A. viscosus from clinical lesions signify infection in the host? Numerical taxonomy and monothetic Darwinian taxonomy have been analyzed in differentiation of A. viscosus from other filamentous and diphtheroidal microorganisms. Numerical taxonomy is a successful method for grouping distinctly different genera, but speciation and separation of biochemically and morphologically related genera require consideration of other factors. Corynebacterium, Rothia, Bacterionema, Actinomyces, Nocardia, Arachnia, and Bifidiobacteria morphologically resemble each other. Actinomyces viscosus produces a diphtheroidal form which cannot be distinguished from Corynebacterium and a filamentous form which resembles these other genera. Acid-fast stains are not of much use unless simultaneous controls are included. Inadequate decolori- zation.will produce false positive results, and excessive use of acid in decolorizing can give false negative reactions. Some species of 62 63 corynebacteria may be partially acid-fast when stained by the Kinyoun's modified acid-fast stain (Harrington et al., 1966). Certain species of Nocardia, and old cultures or excessively subcultured specimens of N. asteroides may stain non-acid-fast. Biochemical characterization will aid in speciation of most of the microorganisms, but certain problems in technique may be encountered. Review of literature concern- ing biochemical identification of A. viscosus has shown significant differences in results. Choice of media, size of inoculum, length of incubation, method of analysis, and number of strains examined are the probable causes of discrepancies. Slow-growing, fastidious micro- organisms require longer periods of incubation and enriched media, which are not used in many clinical laboratories. Biochemical tests suggested in Appendix A, performed by described procedures, should be adequate for differentiation of A. viscosus from similar microorganisms. Actinomyces viscosus strain CF-470-72 was sent to L. K. Georg at CDC; however, the organism was not viable when it arrived in Georgia. A catalase test and fluorescent antibody studies were still performed. Georg reported a negative catalase reaction, yet the strain was clearly catalase positive when it was examined at the'MSU Clinical Laboratory. The catalase enzyme probably had been degraded when the strain died, thus producing the negative result. The VPI method of examining the change in pH of carbohydrate test media is a sensitive technique for determination of fermentation of sugars. Holdeman et al. (1972) reported weak fermentation of xylose, and negative to weak fermentation of ribose by A. viscosus. Strains CF-lO78-73 and CF-l94-74 fermented ribose and Strain CF-99-72 fermented 64 xylose. Many carbohydrates are not fermented by all strains of A. viscosus; thus, only the recommended biochemical tests suggested in Appendix A should be used as criteria for rapid speciation. Cell wall composition is a good analytical technique for separa- tion of diphtheroids. This requires complicated apparatuses, gas chromatography or extensive extractions and chemical modifications which are not available to most clinical microbiology laboratories. Examination of cell walls for presence of diaminopimelic acid, lysine, mannose, ornithine, deoxytalose, and fucose is the minimum procedure for speciation of A. viscosus by this method. Serology is a popular taxonomic procedure used in many specialized large diagnostic laboratories. Agglutination studies are of limited value; however, fluorescent antibody and immunodiffusion tests are quite sensitive for differentiation of A. viscosus from most other pleomorphic Gram-positive microorganisms. Unadsorbed fluorescent antibody which will react with all strains of A. viscosus and shows no fluorescent reaction with A. naeslundii is not presently available. Three strains of A. viscosus isolated from canine infections at the Clinical Microbiology Laboratory at Michigan State University did not fluoresce with the fluorescent antisera prepared at CDC. M. A. Gerencser at West Virginia University has reported that many strains she isolated did not fluoresce with these antisera. Presently the MSU isolates are being examined by her. Hopefully, the fluorescent antibody she has prepared will be more sensitive. Gas-liquid chromatography is an expensive but necessary and rapid technique for speciation of anaerobic bacteria (Holdeman et al , 1972) 65 and may be useful in separation of genera of diphtheroids. Analysis of end-products from fermentation of glucose has produced a separate group of the actinomycetes. Speciation of A. naeslundii and A. viscosus with this technique has not been reported, but quantitative measurements of the amount of acetic acid produced when cultures are grown aerobically without additional carbon dioxide may be of value. Actinomyces viscosus reportedly grows better in oxygen when carbon dioxide is added, while A. naeslundii apparently shows no requirement for CO2 in aerobic cultures. Quantitative analysis of end-products may be lower with A. viscosus than with A. naeslundii when grown in pure oxygen environment. Electron microscopy is of doubtful use in taxonomy of diphtheroids at the present time. It may be of some value in the future, but cost will probably not allow it to be a tool for bacteria or fungal specia- tion in a clinical laboratory. Correct speciation of diphtheroids is necessary when the micro- organism is isolated in pure culture from a clinical specimen. Actino- myces viscosus, when freshly isolated from suppurative material, may form either palisading pleomorphic Gram-positive rods similar to Corynebacterium sp., or filaments closely resembling N. asteroides and other members of the Actinomycetales. Colonial morphology may be smooth, entire, mucoid, glistening, and convex or rough, heaped, granular, irregular, and dry. Both forms may be pathogenic. Routine biochemical media with the addition of PPLO-serum (Difco) reduces laboratory cost, yet is adequate for speciation. Infections from actinomycetes, Nocardia and corynebacteria are not common; however, 66 they are being reported. Nocardia and certain Actinomyces infections cannot be differentiated without cultural identification since the lesions of both types of infections may be closely similar. It is essential that the cause of subcutaneous granulomas be correctly determined as soon as possible so that proper antibiotic treatment can be instituted. Once draining sinuses and fistulous tracts have formed or progressive empyema has developed, the prognosis is poor in spite of antibiotic therapy and surgical excision. Nocardia sp. are more virulent and the infections are treated with different drugs. Penicillin is of little use in Nocardia infections, yet most Actinomyces strains are extremely sensitive to this drug. A report of a "diphtheroid, probably contaminant" in cases of chronic granulomatous and suppurative infections may serve as a tremendous hindrance to effecting a cure for the patient. One must carefully evaluate reported references to diphtheroidal species in the literature. Carbon dioxide requirements for aerobic growth were not clearly evidenced with the canine strains of A, viscosus examined by Gerencser (Gerencser at al., 1969) and in this study, yet others use this as one of the definitive characteristics of A, viscosus (Georg et al., 1972). Reports that A. naeslundii does not ferment starch, and many erroneous generalizations about lactose fermentation and other biochemical reactions of corynebacterium have been made (Holmberg et al., 1972; Georg et al., 1969a). Morphological variation within the same strain of diphtheroidal microorganisms is common; presence of two or three different colonial types on an agar medium, or presence of diphtheroidal and filamentous microorganisms in a Gram stain, may not necessarily indicate contamination. 67 Actinomyces viscosus, other actinomycetes, and diphtheroids are part of the normal flora in periodontal plaque of man, hamsters and probably other animals. Isolation of these microorganisms from plaque or sputum does not usually indicate infection. Multiple isolation of a diphtheroid in pure culture, or presence of filaments in granules isolated from granulomatous or suppurative tissue, is necessary to confirm infection with A. viscosus. When filamentous microorganisms are observed in granules, even if smaller amounts of bacteria are present, the former microorganism is the probable cause of infection. Swerczek et al. (1968) has indicated that dogs under one year of age are more susceptible to Nocardia infections, while dogs older than one year are more susceptible to Actinomyces infections. Both cases of canine actinomycosis due to A. viscosus reported by Georg et al. (1972) occurred in dogs less than six months of age. Attempts to correlate size of dog and environment of the animal to A. viscosus infections have been made. Hunting dogs and other large breeds most likely are exposed to major and minor traumas such as puncture wounds, abrasions, and gunshot wounds which become contaminated with soil bacteria causing mild debilitation. This lowering of body defenses is apparently necessary for endogenous organisms to proliferate. Actino- myces viscosus is most likely a secondary invader superimposed on an underlying disorder. Four of the five cases described in this thesis were found in dogs with histories of previous disease which may have caused temporary debilitation (gunshot wounds, epilepsy, osteochondritis of the joints). These findings have not been examined in other reports, and the actual significance of minor debilitation is only a theoretical 68 surmise. Actinomyces viscosus is not as pathogenic for healthy labora- tory animals as N. asteroides or other actinomycetes. However, it occasionally infects dogs, mostly large breeds, and rarely infects man and cats. Signs of chronic granulomatous tumorlike swellings, draining abscesses leading from fistulous tracts, and dyspnea, emaciation and low grade fever seen with thoracic empyemas are frequent observations in actinomycosis and nocardiosis. Hard, well defined actinomycotic granules result from deposition of acidophilic material around filaments. Granules associated with nocardiosis and A. viscosus are soft or white, flaky, poorly defined masses of filaments which are not cemented with this hard calcium-phosphate matrix. Purulent material and granules have been occasionally submitted to the microbiology laboratory for culture and yet no growth on media was observed. Direct smears may have shown Gram-positive branching fila- ments, and specimens had been immediately cultured aerobically and anaerobically with little success. No explanation has satisfactorily accounted for all instances of failure to grOWIA. viscosus or other filaments. Occasionally inadequate specimens are submitted to the laboratory, or time between collection of specimen and culturing has been prolonged, which can kill fastidious pathogens. Bacteria present in granules may rapidly overgrow fastidious filaments on routine media. A final contributing cause for poor isolation of A. viscosus is recent antibiotic therapy, which inhibits growth or kills most of the micro- organisms. The best specimen for submission to the laboratory is immediate delivery of granules obtained from the infected area before antibiotic treatment has begun. ADDENDUM Clinic Case 147495 (CF—l94—74) After completion of most of the research for this thesis, another case of canine actinomycosis was diagnosed at the Clinic. Lacking sufficient data to be included in the body of this thesis, morphological and biochemical properties of the strain are presented in an abbreviated form. A 1.5 cm-sized swelling was noticed in the cervical region of a six-year-old female Norwegian Elkhound late in January 1974. The family veterinarian prescribed penicillin and streptomycin, which apparently caused temporary regression of the mass. No history of hunting accidents had been reported; however, the dog reportedly spent a lot of time outdoors and in wooded areas. The mass increased in size to about 6 cm in diameter by March 4, 1974, so the veterinarian referred the case to the Michigan State University Veterinary Clinic. The abscess was excised and drained, and portions of tissue and exudate were collected for examination. The animal was treated with penicillin and streptomycin and was discharged on March 7. A month later, the animal was reported in excellent condition, with no signs of regressionfi Examination of a smear from the aspirated exudate did not reveal granules or filamentous microorganisms. A large number of nucleated cells, 248,000/mm3, neutrophils, erythrocytes with occasional 69 70 eosinophils, histiocytes, and mononuclear phagocytes were observed. Histological examination of the tissue revealed a chronic suppurative granuloma consisting of highly fibrotic connective tissue and vascular proliferative tissue infiltrated with neutrophils, histiocytes, and plasma cells. The granuloma penetrated surrounding skeletal muscle causing some myodegeneration. Gram stains of this specimen (Figure 7) showed Gramrpositive branch? ing filamentous microorgan- isms in discrete masses without any club-shaped borders, which is suggest- ive of A. viscosus or Nocardia. Confirmation of the diagnosis was made L, from isolation and identi- Figure 7. CF-l94—74. Histopathology fication of the etiologic of A. viscosus granules from cervical abscess. Hematoxylin and eosin stain; agent, A. viscosus. X 1350. ' This strain grew slowly on blood agar, did not grow on Sabouraud's agar, but grew well on supplemented BHI—blood agar under anaerobic and aerobic conditions with increased carbon dioxide. Broth cultures revealed both rope-like and flocculent growth. The colonial morphology was of the smooth form with a slight tendency to form mucoid, heaped colonies. Gram stains of a 48-hour agar culture showed short filaments and diphtheroidal elements. ~Branching was infrequent, but a predominance of X and V forms was apparent. Biochemical reactions of this non-acid— fast, catalase-positive strain, as reported in Table 3, are typical of A . viscosus . BIBLIOGRAPHY BIBLIOGRAPHY Adeniyi—Jones, C., J. A. Minielly, and W. R. Matthews. 1973. Actino- myces viscosus in a branchial cyst. Am. J. Clin. Path. 69:711-713. Ajello, L., L. K. Georg, W. Kaplan, and L. Kaufman. 1963. C.D.C. Laboratory Manual for Medical Microbiology. U.S. Dept. of Health, Education, and Welfare, C.D.C., Atlanta, Georgia, pp. G39-G79. Boone, C. J., and L. Pine. 1968. Rapid method for classification of actinomycetes by cell wall composition. Appl. Microbiol. lg; 279-284. Breed, R. S., E. G. D. Murray, and N. R. Smith. 1957. Bergey's Manual for Determinative Bacteriology, 7th ed. The Williams and Wilkins Co., Baltimore, Maryland, pp. 713-744. Bruner, D. W., and J. H. Gillespie. 1973. Hogan's Infectious Diseases of‘Animals, 6th ed. Cornell University Press, Ithaca, New York, pp. 467-480. Bousfield, B. J. 1972. A taxonomic study of some coryneform bacteria. J. Gen. Microbiol. 11:441-455. Brion, A. 1939. L'Actinomycose (hi chien et du chat. Rev. Méd. Vet. 9_1: 121-159. Brion, A. 1942. L'Actinomycose

n > mooauoooa o>auamoo n + mooauoooa o>auowoo u I “wooaaaouow uos n ozN .mama .aouamo “memo ..Ns no Housemau “mesa ..Ne so nasmmo “mesa ..Ns so amameaom ”ohma .umumoaaaoz “mama ..Ns so mamaaaoo a + I + I AIV+ I + I I o>auouaoooo osooomwo .n oanoaoooo Anamooowo .n uoov + I + I > I I I I zaaoomo .moaao> mounsoewamn o>auouaouom + + I I oz I + I oz .0333. Sagan oz I I I > I I I oz moaao> mannwooooaoon oz oz + I I I I I oz oaooao< swanaoommwomnn Asmama I Imva A+VI > > > AIV+ Ava oz moaao> m sxwaoemoooemaob I I + I + I I I oz manoaooo< mwxnmoan o>aaouaooow + + I I + I + I oz .oanoao< osorowamaoon 92 s AIV+ I I I I I oz vapoumma< saoemsuseooeeaom oz I I > I > + zHHoHuaoo + oanoaon Bewaomoz mam amoox .saon moo coau moo mood umon aow< ouomamaaswon mamaoonao Izaoawzn Imoao omou Ioauo Ioowon Imao Iouoo Iwao< o.wooa=onom oowzxo oaasomo< Iwowo> Icon IHoo ouoauaz no cusoao Noumma Hmamaoowaooaoaa Howaoaoounoaw ao moouomaoaan m>auamoolaoau msoaao> no mooauoooa wouoooxn .Hln mHooH 80 taken to prevent contamination since the media are incubated for such long periods of time. Routine laboratory tests for determination of biochemical reactions are satisfactory. Actinomyces viscosus may or may not form filaments in culture, is non-acid-fast, catalase positive, hydrolyzes aesculin, ferments lactose and usually reduces nitrates. Urease production and Voges- Proskauer reactions are negative. Smooth or rough colonial morphology is evident on aerobic and anaerobic culture media. No growth occurs on Sabouraud's agar (Table A-l).