A COIIPARISON OF AERGG-‘NIC I'ARIAN‘I‘S OF Pf ISTEURELLA mm P". AIULTOEIBA mm P ‘IELIIICTRGfiLA 73:03:: {:09 {Am chma. of‘ Mg 5. m BTW VII N? Iv“: 52/18.in a LiCLGga 3in «L 37 I 1quImurinmmrumrmmmu 1' 1/ 3 1293 01009 1068 m” & LIB RA R Y * Michigan State University *’ 052$ I92 “’33 ABSTRACT A COMPARISON OF AEROGENIC VARIANTS OF PASTEURELLA WITH P.1MULT0CIDA AND P. PNEUMUTROPICA By Mary T. Horan An attempt was made to group and compare 37 strains of PasteureZZa of various species and origin by their biochemical properties and other characteristics, including dye bacteriostasis. Of special interest in this study were the strains of Pasteurella which produced gas during carbohydrate fermentation. The data which were collected by these tests were analyzed by calculating the percent similarity and arranging the strains by cluster analysis. A similarity matrix was constructed to examine the relationship between strains. All the strains of P. multocida of both animal and human origin formed one group. Seven of the aerogenic strains were found to resemble P. pneumotropica most closely, the other aerogenic strain was most like P. multocida. A COMPARISON OF AEROGENIC VARIANTS OF PASTEURELLA WITH P. MULTOCIDA AND P. PNEUMOTROPICA BY Mary Theres a Horan A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Microbiology and delic Health 1971 ACKNOWLEDGEMENTS The author wishes to express her sincere appreciation to Dr. G. R. Carter for his continued guidance and assistance. Sincere appreciation is also expressed to the other members of the staff of the Clinical Microbiology Laboratory for their technical assistance and encouragement. I also wish to acknowledge Dr. Weaver, Dr. Frederickson,-and the Michigan State Department of Health for the cultures which were obtained from them. 11 TABLE OF CONTENTS INTRODUCTION. . . . . . . .'. . . .‘. REVIEW OF THE LITERATURE. . . . . . . Pasteurella multocida. . . . .- Nomenclature . . . .‘. . . . . Cellular Morphology. . . . .'. Colonial Morphology. . . . .~. Biochemical Reactions. . . . . Serological Studies of-Pasteurella Animal Infection . . . . . .'. Human Infection. . . . . . . .' Isolation of P. multocidh from PasteureZZa pneumotropica. . .v Biochemical Reactions. . . . . Serological Studies. . .‘. . . Pathogenicity. . .'. . . .*. . MATERIALS AND METHODS . . . . . . . . Cultures . . . . .‘. . . .‘. . Colonial Iridescence . . . . Biochemical Reactions. . . .'. we sensitiv1ty. O O O O O O O I cultures 0 O O . O O O O I 0 Preparation of Dye Discs Numerical Analysis . . .-. . . iii Page 10 10 13 14 14 15 - 16. 18 18 18 19 20 '20 20- ,20- RESULTS . . . DISCUSSION. . BIBLIOGRAPHY. iv Table LIST OF TABLES List of characteristics for numerical analysis Description of Pasteurella strains . . . . . . Summary of biochemical tests . . . . . . . . . Gas production in aerogenic strains. . . . . . Dye bacteriostasis - variant strains . . . . . Page 22 24 28 32 32 LIST OF FIGURES Figure Page 1 Similarity matrix. .‘. . . . . . . . . . . . . . . . . . . . 34 vi INTRODUCTION PasteureZZa multocidb has long been recognized as a veterinary pathogen of considerable economic importance (12). It is now realized that this species is,a human pathogen as well. Pasteurella multocida may cause local infections at the site of animal bites and may be involved in systemic infections, most often involving the respiratory and central nervous systems (24,25). The PasteureZZa genus is a very complex one taxonomically. There have been many.studies to examine the interrelationships of these species and to improve the classification (39,40,50,54). These studies have led to a simplification of the genus and the reclassification of some species to another genus (50). In this study, the strains which.were examined were P. multocidh of animal and human origin, the closely related genus of P. pneumotropica, and the aerogenic variants of Pasteurella which were tentatively identi— fied as P. "gas" (multocida?). Test strains of P. areas and P. hemo- Zytica were also included for comparative purposes. These strains were compared by means of biochemical and other tests. The biochemical tests were those recommended for Gram negative rods-(28,56). In addition, dye sensitivity tests were carried out on these strains in order to see if a repeatable specific pattern of susceptibility for each species or variety could be established. This had been tried previously without success (46). The data which were collected through these methods were analyzed by the calculation of the percent similarity and were arranged into groups by cluster analysis (2,52). The groups-were depicted diagrams matically by a similarity matrix which provided a means of comparing each strain individually with the other strains. REVIEW OF THE LITERATURE Pasteurella, which was created by Trevisan (54) to include the agent of fowl cholera and hemorrhagic septicemia bacilli, is a genus in the family, BruceZZaceae. This genus is defined as small Gramenegative rods, which generally occur in pairs or short chains and which are characterized by bipolar staining, pleomorphism, absence of gelatin liquification, and the fermentation of carbohydrates with slight acid production (4). In the 1957 edition of Bergey's Manual (4), there are the following 9 species listed for this genugg. P. multocida, P. septicaemiae, P. hemolytica, P. anatipestifer, P. pestis, P. pfthI, P. pseudbtuberculosis, P. tularen- sis, and P. novicida. These 9 species form a very heterogeneous group of organisms. Taxonomic studies have been attempted to understand the interrelationships of these species and the others which have been added to the genus and to improve the classification (39,40,50,54). ' Since the 1957 publication of Bergey’s Manual, P. tularensis (40) and P. novocida (39) were reclassified as belonging to the genus, FranciseZZa. P. pfhffi, P. septicaemiae, and P. anatipestifér were considered question- able members of this genus. P. pfhffi is not readily distinguishable from P. pseudbtuberculosis, and P. septicaemiae and P. anatipestifer are gelatin liquifiers which contradicts the definition of the genus (50). However, even with these simplifications, the remaining species still provide some difficulty in classification. WOrk by Talbot and Sneath (54) showed that P. multocida is only distantly related to 3 4 P. pestis and Pu'pseudbtubercuzosis, but the latter_species are closely related to each other. Smith and Thal (50), in 1965, confirmed these results when they employed the Adansonian method to classify 25 strains of PasteureZZa of various species. They found that the strains fell into 2 broad groups on the basis of the oxidase reaction. Group I which-is oxidase positive includedz‘ P. multocida,.P. pneumotropica, P. hemo- Zytica, and P. hemolytica var. ureae. Gtoup II which is oxidase negative consisted of: P. pseudbtuberculosis, P. pestis and P. "X"- Group II has been reclassified as belonging to the genus, Yarsinia (50). In Group I, the 2 most dissimilar species are P. multocida and P. hemolytica. They were recognized early in experimental studies as 2 distinct groups within the PasteureZZa genus. The atypical hemolytic, non-indole forming group was named P. hemolytica (37). Later, the non- hemolytic, indole positive organism was called P. multocidb (46). Pasteurella multocida PasteureZZa multocidh is a heterogeneous species. using computer taxonomic technique, strains of this species have shown an 85% similarity with each other (50). Therefore, there have been repeated attempts to subdivide the species. Nomenclature The name, P. multocida, is a recent one having been proposed by‘ Rosenbusch.snd Merchant in 1939 (46). It has its origin in the first name, Bacterium bipolare-multocidum, given to hemorrhagic septicemia organisms isolated from cattle, swine, deer, horses and goats by Kitt in 1885 (46). The large number of species names which appeared in the early literature is a direct result of the wide variety of animals infected and clinical diseases produced. 5 The diversity of infection prompted an effort to subdivide these organisms according to a zoological classification (30). This system proved disadvantageous because organisms which caused a single type of disease were given a wide variety of names. The zoological classifica- tion was later abandoned as a result of many cross-pathogenicity and cross-immunity studies which proved that this was not a valid means of subgrouping these organisms (42). Even at the present time, the species name is not uniform. Topley and Wilson introduced the name P. septica in the.PrincipZes of’BacterioZogy and IMmunoZogy, lst edition, 1939, p. 48. This name appears widely throughout the literature of Great Britain and other Commonwealth nations. Cellular Morphology_ Pasteurella multocida is a Gram-negative organism which varies from a long rod to a coccobacillus (12). Smith (48) recorded small differences in the cellular morphology of cultures from different animal sources. He found the bovine strains to be pleomorphic measuring 0.5 by 1.2 u, while porcine strains were non-pleomorphic about 0.8 by 1.0 u. Dog strains varied from small coccobacilli to long S-shaped filaments. These differences have not been found to be adequate or precise distinc- tions. Cultures from all animal sources, especially those maintained on laboratory stocR culture media for a long period of time, tend to form long filamentous forms (12). Colonial Morpholggx Several colonial variants of P. multocida have been demonstrated. The terminology which has been used to describe these variants often dif- fers, but the 3 main variants have been described as: iridescent (smooth), blue (rough), and mucoid (14,16). 6 To distinguish these variants a common method is to observe the colonies through obliquely transmitted light. The smooth colonies will appear iri- descent with a greenish tint, while mucoid colonies are reddish iridescent, and the rough colonies appear blue (16). The 3 colonial variants also differ in physical appearance on a blood plate. The mucoid colonies are the largest with irregular margins, while the smooth and rough colonies are smaller with entire edges (12). The iridescent and mucoid variants are most often obtained from animals with acute diseases, and the blue variants are isolated frequently from animals with chronic infections (7,14). Attempts (18,13) have.been made to correlate colonial variation with other properties of the organism. It has been experimentally determined that there is a relationship between the colonial morphology and virulence. Iridescent colonies are the most virulent of the colonial variants for the mouse, and rough colonies vary in their virulence for mice (13). Although colonial variation does affect virulence and serological behavior, it seems to have no other significant effect. Biochemical characteristics including carbohydrate fermentation and indole produc- tion are not affected by colonial variation (18). BiochemicaliReactions In a biochemical description of a species, the fermentation pattern of carbohydrates is significant. A "typical" pattern for P. muZtocida (46,48,54), using 12 peptone water and Andrade's indicator, is acid from glucose and sucrose and no acid production, with rare exceptions, from lactose, arabinose, dulcitol and inositol. Approximately 50% of the strains tested produced acid in xylose, mannitol and sorbitol, while maltose is only rarely fermented. PasteureZZa multocido does not produce 7 gas routinely in carbohydrate.fermentation; however, there are reports in the literature of a few aerogenic strains (45,54). Talbot and Sneath (48) described a strain of P. multocida, isolated from a human dog bite infection, which produced gas in small quantities from sucrose, maltose, mannose and trehalose. There was also a report of an aerogenic strain from a purulent leptomeningitis in a dog (45). In all other respects these strains were typical of P. multocidh. In addition, Dr. Weaver (56) in his classification chart for Gramrnegative organisms included a group of PasteureZZa which produced small quantities of gas from.g1ucose.‘ These strains did possess other deviant reactions and were tenatively identified as P- "gas" (multocidh7). This phenomenon of gas production has not been studied by taxonomists in any further detail. The fermentation pattern of P. multocida is notable in its varia- bility. Some of the variation in certain carbohydrates may result from the comparison of results obtained by the use of different growth media and indicator systems. It has been found in a closely related species, P. pneumotropica, that with certain sugars which are slowly fermented varying the indicator will alter the outcome of the test (20). Certain variation from.the norm has been observed to predominate in one host species over another. Smith (48) found that strains from cattle and pigs. generally ferment xylose, sorbitol and mannitol, while dog strains usually ferment maltose, trehalose, and dextrin. In considering the biochemical reactions of a bacterial species the fermentation pattern is studied because it is thought of as a stable property. In P. multocida, there is some evidence that it is a dynamic. property. Rosenbusch and Merchant (46) observed 4 strains which showed changes in their fermentation of xylose, arabinose and dulcitol over a 8 2dyear period. In the more recent work of Higgins (22), he noticed a change in the fermentation pattern of a strain which was causing an out- break of fowl cholera in.a chicken flock. This has led to some specula- tion of an interrelationship of virulence, antigenic characteristics, and carbohydrate fermentation (22).‘ In addition to carbohydrate fermentation, certain other biochemical reactions may be used to characterize the species (12,28,46,50,54). Catalase, oxidase, and indole are produced and nitrates are reduced to nitrites. The methyl red and Voges-Proskauer tests are negative. Hydrogen sulfide is produced in small amounts which is detectable by suspending a lead acetate strip over a triple sugar iron agar slant. Urease production is generally negative, but it can be positive in a few rare cases. Citrate and malonate are not utilized, and gelatin is not liquefied. 0f the decarboxylase reactions, only ornithine decarboxylase is positive. Serological Studies of PasteureZZa muZtocida Many workers have.attempted to classify P. multocida through sero- logical techniques. Some of the early workers in this field included Cornelius, Ochi, Yusef, and Khalifa (12). One of the most important of the early studies was the work of Rosenbusch and Merchant (46). They divided strains of P. muZtoeido into 3 groups according to its fermenta- tion of xylose, arabinose, and dulcitol and agglutination tests. In general, the early studies underestimated the complexity of the species and used many different standards to evaluate and classify the organisms (12). The next attempt atva serological classification of the species was the-work of Little and Lyon (31). They divided 30 strains of 9 non-hemolytic PasteureZZa by slide agglutination test into 3 types. However, it is now appreciated that this typing procedure was inadequate (12). Another method to subgroup Pasteurella was with the serum pro- tection test in mice. In 1947, Roberts (44) used this method to separate 37 strains into 4 types which were numbered with Roman numerals I through IV. This classification was found to loosely parallel some host and biochemical differences. Types I and 111 did not ferment arabinose but was xylose positive. Type II fermented arabinose but generally did not ferment xylose. Type IV fermented arabinose and dulcitol, but there was an uncertain reaction pattern with xylose. At the present time a serological classification of P. multocida'has been established which uses the combined capsular and somatic antigenic make-up to categorize the species. The capsule of P. muZtocida consists of protein and carbohydrate components (1,43). Both the proteins associated capsular substance (1) and the capsular polysaccharides (6, 29,42) have been demonstrated to have serological activity, but the poly- saccharide component alone is not immunogenic. Through an indirect hemagglutinatiOn test (8) in which the capsular extracts were adsorbed to the erythrocytes, 5 serotypes were named. The serotypes were given the letters A through E. It was later suggested that the group C be- dropped because it did not adequately represent a capsular type (10). In addition to the capsular system, Namioka and Murata (36) further classified strains on the basis of somatic or 0 antigens. This and the work of Namioka and Bruner (35) divided P. muZtocida into 11 0 groups. Each 0 group is designated by an arabic number. If the 0 groups and the capsular types are correlated, 14 combined serotypes result (12). Although this dual serological classification promises to add to the present knowledge of P. multocida, it is acknowledged that the multiple 10 cross reactions, especially in determining the 0 serotype, can cause difficulty (35). Animal Infection PasteureZZa multocidhLis known as an animal pathogen of considerable economic importance.‘ The pathogenicity of the organism is variable. It is involved in infections which vary.from the highly severe virulent forms to non-clinical commensalism in.which organisms are found in the. nasopharynx (12). The organism also acts as a secondary invader in a number of viral and mycoplasmal infections (12). The mechanism of infection of this organism is not completely under- stood. Stresses of various kinds are important, especially in-the periodic outbreaks of shipping fever of pasteurellosis and fowl cholera (11). It seems most probable that the process of endotoxin poisoning is responsible for the pathologic effects in acute septicemias and death (12). The organism is transmissible by contact, contaminated food and water and in some cases by droplet infection (33). The principal acute infections in animals are fowl cholera, hemor- rhagic septicemia in cattle and water buffalos, enzootic pneumonia in swine, snuffles in rabbits, and periodic infections of a wide range of animals (12). Severe acute septicemic infection results in such clinical signs as blood vascular congestion, submucous and subserous hemorrhage and enteritis (11). The chronic type of infection in a localized area is characterized by abscessation and by anemia and diarrhea leading to a general debilitation (11). Human Infection It is becoming apparent that P. muZtocida must be considered a human as well as an animal pathogen. The first infection of man with 11 P. multocddhwwas reported by Brugnatelli in 1913 (15). It was isolated from a puerperal pyrexia. In 1947, Schipper summarized the literature Iof Pasteurella infections from 1930-1947 reporting approximately 40 cases (15). Meyer in 1955 updated this review counting 95 cases to that year (15). An additional 12 cases were counted by Swartz and Runs (58) from 1955-1959. There has been a considerable increase in reporting of isolations from.humans in the last few years due to increased familiarity with the. species (9). The relatively.small number of cases in earlier literature is probably the result of errors in diagnosis. This is dramatically shown by an 8-month study at the University of Colorado, in.which 10 cases were reported in this short time period (17). The largest number and the most thoroughly studied infections in. humans are in the local infections which result from bites of dogs, cats, and other mammals (15). It-has been proven that P. multocido is a common commensal of the respiratory tract of domestic animals (48). These organisms may be injected into the wound at the time of the bite. An epidemiological study (24) has been made of local P.1multocido infections to see if any trends in infection do exist. According to this~ study, these infections are most frequently located on the upper extremi- ties, and 732 of the people infected were 40 years of age or older. The infection rate resulting from cat bites was about 10 times greater than dog bites. The lower frequency of dog bite infections may be accounted for by certain differences between dog and cat strains. Cultures isolated from the throats of dogs were usually rough (49).' The appearance of clinical symptoms in infections occurs very quickly. In about 24 hours, the patient feels severe pain and swelling and has a gray odorous discharge at the site (3). If the patient does not seek 12 medical treatment, recovery may be slow and there may be underlying necrosis of the bone (49). The underlying bone involvement is a more frequent problem if the organism.has been injected into a poorly vascular- ized region. The result may beya chronic osteomyelitis, cellulitis, and adenitis with abscess formation (3,22). For these local infections, the treatment is usually antibiotic therapy (32,58). PasteureZZa multocidh is a Gramdnegative organism which is highly sensitive to penicillin. In some cases surgery may be needed (19). This is especially true if the bone becomes extensively involved. In analyzing the biochemical reactions of cultures isolated from human infections, it has been reported by Smith (49) that the cultures isolated from dog bite wounds resemble strains isolated from the throat of-healthy dogs. The same relationship appeared from strains isolated from cat infections. However, there is some contradictory data in.the literature (17,32) which show the isolation of mannitol positive strains from local human infections. This is not characteristic of dog strains isolated.from.hea1thy dogs. A second type of human infection with PasteureZZa is the systemic infection which is unrelated to animal bites. The most frequently diag- nosed conditions in these infections are meningitis, empyema and bronchiectasis (25). The largest group of internal infections are the infections of the respiratory system (47). The classic patient who develops a pulmonary infection with P. multocida is an individual who has had many years of contact with animals and has some underlying chronic respiratory problems (47). Systemic infections may also involve systems other than the respira- tory system. The non-respiratory systemic infections often involve the l3 appendix and the central nervous system (25). The 2 most common clinical conditions are_meningitis and chronic otitis media (25). A mechanism of pathogenicity has been suggested for the non-respiratory infections. The organism may reside as a commensal in a healthy respira- tory tree until some trauma to the host system causes it to invade the other host tissue (15). This is supported by the fact that there are reports of humans carrying this organism symptomless in their respiratory tract (49). A site such as a.carcinoma, appendiceal abscess or any perforated abscess may also provide P. multocidb with an opportunity to act as a secondary invader (15). In the systemic infections with P. multocida, septicemia may or may not occur, but if it does occur chances for recovery are poor-(15). The general treatment for these infections as with the local infections is antibiotic therapy (15). There has been only one report of a penicil- lin resistant strain (47). In P. multocida infection, an animal source is always.presumed. However, there have been reported cases where such a source could not be found (25). It has been postulated that in addition to animal to human infection, there might also be a human to human route of infection (25). Isolation of P. multocido from Human Clinical Specimens In the human diagnostic laboratory, P. multocida can be confused with certain other Gramrnegative rods. In direct smears from certain specimens, pleomorphic forms can be found which resemble Hemophilus (3)- Pasteurella multocida can be differentiated from.HemophiZus by the fact that P. multocida grows in the absence of X and V factors and is sensi- tive to penicillin (3). 14 Meningeal smears may show Gramenegative diplococci which can be mis- taken for Neisseria and tha (15). The latter are, like P. multocida, oxidase positive. Pasteurella multocida can be distinguished by.the fact that it produces indole and reduces nitrate while Neisseria and Mime do not (3). Pasteurella multocida is sometimes overlooked in isolations from animal bites or other places from which enteric bacteria are expected. One reason that P- muZtocida is frequently overlooked is because it reduces nitrate and may resemble enteric bacilli in broth. However, it is distinguished as a separate species, by the fact that it does not grow in the presence of bile salts, is oxidase.positive and is sensitive to penicillin (2).' The main characteristics which should indicate that PasteureZZa might be involved in the infection is: the source of the swab, e.g., cat or dog bite, penicillin sensitivity, morphology and Gram-negativity (51). PasteureZZa(pneumopropica PasteureZZa pneumotropica is the species of PasteureZZa which is the most closely related in its biochemical characteristics to P. multocida. It was first isolated and described by Jawetz in 1950 (27). Biochemical Reactions PasteureZZa pneumotropica appears as a convex, round, entire, semi- translucent colony about 1 mm. in diameter with a smooth surface which varies in color from.white to grayish yellow (5,27,57). There is no hemolysis on horse, sheep or rabbit blood, but there is occasionally greening under the colonies (5,27,57). Similar to P. multocida, 15 P. pneumotropica is indole, nitrate, catalase and oxidase positive, and MacConkey, methyl red, and voges-Proskauer negative (20,23,57). PasteureZZa pneumotropica differs from P. multocido chiefly in that it is urease positive and ornithine decarboxylase negative (57). In addi- tion, P. pneumotropica is more susceptible to penicillin than P. multocida. The carbohydrate fermentation pattern is also slightly different for the 2 species. using the same test system which was used to test for the fermentation pattern of P. multocida, Andrade's indicator with 12 peptone water, P. pneumotropica produces acid but no gas from glucose, xylose, lactose, sucrose, maltose and arabinose (21). It does not ferment mannitol, sorbitol, dulcitol, and inositol (21). Serological Studies In serological studies of P. pneumotropiea, normal mice who carried this organism in the latent form in their lungs had no agglutination or complement fixation titer (23). After experimental inoculations, all the mice had complement fixation titers of 1:10 but no agglutination titers (23). Cross-absorption tests indicated that there were some. antigenic differences between strains isolated from different animal colonies which suggested an antigenic mosaic.similar to P. muZtocida (27). In further studies, P. pneumotropica cells were not agglutinated by antisera of P. pestis, P. pseudbtuberculosis, P. multocida, B. bronchi- septica, H. influenza, type B, and H. pertussis (27). However, interaction of P. multocida and P. pneumotropica cannot be ruled out completely because antisera for all the serotypes of P. multocida have not been tried. 16 Pathggenicity In its initial isolation P. pneumotropica was characterized as a latent organism which is an inhabitant of the lungs of mice (27). After 14 serial passages in mice, the infectivity was still very low. This- species caused lesions only in the lung and only.when injected intra- nasally. The only other injection route which caused any disease process was intracerebral injections which caused brain abscesses (27). The virulence of the organism obtained by serial passage was unstable (27). Repeated transfer on artificial media soon resulted in a partial loss of pathOgenicity (27). Jawetz speculated that P. multocida and P. pneumotropica might.share a common ancestor and represent 2 ends of a bacterial spectrum. Pus- teureZZa multocido at one end of the spectrum is an-invasive organism of potentially high virulence and wide host specificity, and P. pneumotropica is at the other end with tissue specificity and low virulence. He sug- gested that the different biochemical features might be the result of the gradual adaptive changes between the host and the parasite (27). Additional experimentation has shown that P. pneumotropicafihas a. wider spectrum of pathogenicity and greater virulence than first thought. In the mouse, it was found in the latent form in the brain, uterus, liver as well as the lung (23). PasteureZZa pneumotropica has 8180 been isolated from a number of other animals including rats, kangaroo rats and hamsters (5). PasteureZZa pneumotropica can also-be found associated with natural disease processes. It was found in the lungs of rats and mice with pneumonia (5) and from the eyes of mice with conjunctivitis (5). In dogs, it has been associated with otitis, peritonitis, renal abscess, pneu- monia, kidney abscess and septicemia (5). There have also been records 17 of human respiratory tract isolations (20) and isolations from local inflammation at the site of a dog bite (57) and cat bite (58). METHODS AND MATERIALS Cultures, Thirty-seven strains of Pasteurella were studied. The strains which were isolated from.cases of human pasteurellosis were obtained from 2 sources. Thirteen strains were provided by the Michigan Department of Public Health, Bureau of Laboratories, Lansing, Mich. Six additional cultures of human pasteurellosis isolates were supplied by Dr. Robert E. Weaver of the Center for Disease Control, Atlanta, Georgia. The 12 strains of Pasteurella multocida of animal origin studied were part of the stock culture collection of Dr. G. R. Carter, Michigan State University. Four cultures of PasteureZZa of species other than Pasteurella multocida were also obtained from this collection. Two strains of Pasteurella pneumotropica were.received from.the laboratory of Dr. Frederickson, Copenhagen, S. Denmark. All the stock cultures were maintained on Difco stock culture medium. The cultures were transferred to fresh stock culture medium at 3-month~ intervals. Colonial Iridescence Each strain.was grown on BBL trypticase soy agar for 24 hours. If the colonies were iridescent, it could be seen clearly when the plates were examined in a darkened room with a-strong oblique light falling on the plate from behind. 18 19 Biochemical Reactions Each strain.was.streaked for isolated colonies on blood agar plates which consisted of BBL trypticase soy.agar base and approximately 72 defibrinated bovine blood. 'Colonial morphology at 24 hours was noted. Growth from these plates was Gramestained and was used to inoculate the other biochemical media. -The biochemical tests, in general, were incu- bated at 37° and were read at 24 hours, 48 hours, and 7 days. Carbohydrate utilization was studied with 3 types of media. Tests were carried out in semi-solid phenol red broth containing Difco phenol red base, 0.152 agar, and 12 of the particular carbohydarte. Oxidation- fermentation test.was performed using the method of Hugh and Leifson (26) with various carbohydartes as substrates. The growth on 10% glucose and lactose slants was also recorded. Difco Simmon citrate slants were used to test for citrate utiliza- tion, while Christensen urea slants were used for urease production. H25 production was detected by suspending lead acetate strips over triple sugar iron agar slants. Esculin hydrolysis was studied by using the modified esculin agar recommended by Sneath (51). Decarboxylase tests for lysine and ornithine and the arginine hydrolase test were performed using Moeller decarboxylase broth (34). Oxidase production was detected with a few drops of a 0.5% solution of tetramethyl-p-phenylene-diamine dihydrochloride on 24 hours growth on a blood plate. For the catalase test, a few drops of 3% hydrogen peroxide on 24 hours growth on TGY slants were used. Nitrate reduction, indole production and the MR/VP tests were per- formed after 48 hours' incubation. Indole production was examined using Kovac's method. Nitrate reduction was tested by adding 3 drops of 0.8% sulfanilic acid and 5 N acetic acid mixture and 3 drops of a 0.5% alphanaphtholamine and 5 N acetic acid mixture. 20 All the media and reagents used in the tests referred to above and the other biochemical reactions studied were obtained from the Difco Company or Baltimore Biological Laboratories. Dye Sensitivity Cultures. All the strains for dye sensitivity testing were grown in Difco beef heart infusion broth at 37° C. for 18 to 24 hours. Preparation of Dye Discs. Stock solutions of pyronin (1:500), methyl violet (1:100), basic fuschin (1:100), and thionin (1:200) were made up in the specified concentrations (w/v) in 50% ethyl alcohol. Corrections were made for the dye impurities. The stock solutions were stored in the dark for no longer than 2 months at 4° C. Taking care to maintain asepsis the dye discs were prepared by completely saturating Difco Bacto concentration discs sterile blanks (1/2" in diameter) with dye solution. The saturated discs were allowed to dry in sterile containers in the 37° incubator. Blood plates were used for the disc sensitivity testing. Sterile swabs were saturated with the broth culture. The plates were inoculated- by swabbing the blood plate in 3 directions. This procedure was shown in control plates to yield confluent growth. The dye discs were then placed on the inoculated plates with sterile forceps. The plates were read as millimeters of inhibition from the edge of the dye disc. A zone of inhibition of 4 mm. or more was recorded as a positive. Numerical Analysis The reaction of the strains were scored using method III of Beers et al. (56). In this procedure, the features to be examined are outlined, and each strain is scored as + if it possesses the characteristic, and 21 - if it does not possess it. The designation, 0, is used when the char- acteristic does not apply to the individual strain. Table 1 contains a list of characteristics tabulated in this study. Percent similarity coefficients (8) were calculated for each possible pair of strains by counting n which is the number of times + signs coin- cide and n' which is the number of times + and — signs coincided, and substituting in the formula: . 100 n , n+n' S 0 signs were ignored. The strains were sorted manually according to the methods of cluster analysis described by Sneath (52).: The groupings which resulted were_ depicted diagrammatically by a similarity matrix. Table 1. List of 22 characteristics for numerical analysis 9. 10. ll. 12. 13. 14. 15. l6. l7. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30 0' 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 24 hr. coloni a1 morphology: cleardwhite colony, 0.5-1.0 mm. round, raised, entire, buttery, shiny. 24 hr. coloni a1 morphology: cleardwhite colony, pinpoint- 0.5 mm., round, raised, entire, transparent. Gram stain: Gram stain: Gram stain: Fermentation - Fermentation Fermentation Fermentation Fermentation Fermentation Fermentation Fermentation Fermentation Fermentation Gas productio Fermentation Fermentation Catalase prod Gram negative coccobacillus. Gram negative rod. Gram.negative rod with pleomorphism. of 1% glucose. of-lz xylose. of 1% lactose. of 1% mannitol. of 1% sucrose. of 12 maltose. of 1% sorbitol. of 12 arabinose. (of 1% inositol. of 12 dulcitol. n from carbohydrates. of 10% glucose. of 102 lactose. uction. Oxidase production. Growth'in nut rient broth. Growth on MacConkey's agar. Growth on SS Growth on Get Utilization o agar. rimide agar. f citrate. Production of urease. Nitrate production.- Indole production. Acid production in 48 hrs. on TSI slant. Acid production in 48 hrs. on TSI butt. Darkening of lead acetate in 48 hrs. Methyl red test. Voges-Proskauer test. Gelatin liquification. Motility. Esculin hydro lysis. Growth on TGY slant at-25° C. Growth on TGY slant at 37° C. Growth on TGY slant at 42° C. Inhibition of Inhibition of growth by pyronin (1:500). growth by methyl violet (1:100). Inhibition of growth by basic fuschin (1:100). Inhibition of growth by thionin (1:200). bi t0 RESULTS Table 2 presents descriptions of strains used in this work. From this initial information, the cultures could be subdivided into the“ following 5 classes: (1) P. multocidozof human origin; (2) P. muZtocida of animal origin; (3) P. pneumotropica; (4) PasteureZZazof other species; and (5) P. "gas" (mltaaidafl. Table 3 is a tabulation of the results of biochemical and_other tests according to these 5 classes. The strains of P. hemolytica and P. ureae can be distinguished by their lack of production of indole and their fermentation reactions. Pasteurella pneumotrOpica can be separated from P. multocida by its production of urease and its carbohydrate fermen- tation pattern. Pasteurella pneumotropica, unlike P. multocida, consist- ently fails to ferment mannitol and sorbitol, but it does ferment maltose. The biochemical reactions of P. "gas" (multocida?) appear more closely related to P. pneumotropica than P. muZtocida. There is no sig- nificant difference among strains isolated from animal and human infections of P. multocida. In Table 3, dye sensitivity testing with discs does give.signifi- cantly consistent and different patterns for P3 multocida, P. pneumo- tropica, and P. "gas" (muZtocida?). This technique is one that has not been used to distinguish PasteureZZa species. The above 5 classes proved to be an inadequate way to represent the biochemical data obtained for these organisms. For example, in addition to the 6 aerogenic strains provided by Dr. Weaver, 2 other isolates, 23 24 Table 2. Description of PasteureZZa strains Strain Species Disease and/or Colonial No. no. origin Source specimen source variant 1 CDe90A. Human Mich. Dept.‘ Human septicemia Iridescent of Public (fatal) and blue Health 2 CD-623 Human Mich. Dept. Brain abscess Blue of Public Health 3 CE-154 Human Mich. Dept. Oida septicemia Blue of Public Health 4 CC-ll93 'Human Mich. Dept. Dog bite (face) Blue of Public. Health 5 CC-1213 Human Mich. Dept. Dog bite (leg) Blue and of Public iridescent Health 6 CE-28L Human Mich. Dept. Dog bite (leg) Blue of Public Health 7 CC—1183 Human Mich. Dept. WOund drainage Blue of Public Health 8 CD-1058 Human Mich. Dept. wound infection Iridescent of Public Health 9 CC-1043 Human Mich. Dept. Cat bite Blue and of Public iridescent Health 10 CC-97l' Human 'Mich. Dept. Cat scratch Blue and of Public iridescent Health 11 OL-l43 Human MSU Vet. Opossum bite (hand) Blue and Micro. Lab. iridescent 12 CE-49 Human Mich. Dept. Tiger bite Blue of Public Health Table 2 (cont'd.) Strain Species Disease and/or Colonial No. no. origin Source specimen source variant 13 CC-llSO Human Mich. Dept. Sputum Iridescent of Public Health 14 B-92 Human Dr. Weaver Dog bite Blue 15 B-767 Human Dr. Weaver Dog bite Blue 16 A-8708 Human Dr. Weaver Dog bite Blue 17 B-704 Human Dr. Weaver Foot.wound Blue 18 B-972 Human Dr. Weaver Ear Blue 19 B-1682 Human Dr. Weaver Leg wound Blue 20 Insein Bovine MSU Vet. Hemorrhagic Blue and B~ Micro. Lab. septicemia iridescent 21 P—1235 Bovine MSU Vet. Unknown Blue and Micro. Lab. iridescent 22 r-473 Bovine MSU Vet. Hemorrhagic Blue Micro. Lab. septicemia 23 19427 Unknown MSU Vet. Unknown Blue Micro. Lab. 24 932A Bovine MSU Vet. unknown Blue Micro. Lab. 25 3397 Porcine MSU Vet. Unknown Iridescent Micro. Lab. 26 M+17 Sheep MSU Vet. Pneumonia Blue and Micro. Lab. iridescent 27 C-566 Canine MSU-Vet. Unknown Blue Micro. Lab. 28 C-55 Canine MSU Vet. Unknown Blue Micro. Lab. Table 2 (cont'd.) 26 Strain Species Disease and/or Colonial No. no. origin Source specimen source variant 29 F—24 Feline MSU Vet. Unknown Blue Micro. Lab. 30 P-1059 Avian MSU Vet. Fowl cholera Blue and Micro. Lab. iridescent 31 M9355 Opossum. MSU Vet. Throat swab Blue Micro. Lab. 32 AP. hemo. Bovine MSU Vet.' Pneumonia Iridescent B-128-7l Micro. Lab. 33 IP. ureae Human MSU Vet. Bronchiectasis Iridescent Micro. Lab. 34 P. Rat MSU Vet. Unknown Blue and pneumo. Micro. Lab. iridescent M2199-70 35 P. Rat MSU Vet. Unknown Blue pneumo. Micro. Lab. 36 P-421 Unknown Dr. Unknown Blue Frederickson 37 P-309 Unknown Dr. Unknown Blue Frederickson Key: Alk. () (No.) Positive reaction Negative reaction Acid production Alkaline .Delayed reaction Number of deviant strains 27 28 UHt‘OMHgt-flm Albtflwciu Table 3. Summary of biochemical tests P. "Gas" P. muZto- P. muZto- (muZto- cide (13) cide (12) cide?) P. pneumo- P hemo- Test~ (human) (animal) (6) tropica (4) Zytica (l) P. ureae (l) Colo- clear clear- clear- white clear grayish nial white white white Morph. round round round round round round (24 0.5- pin- 0.25 pin- 0.5 mm. 1 mm. hrs.) 1 mm. point- mm. point 1.5 mm. 1 mm. raised raised raised raised raised raised entire entire entire entire entire entire buttery buttery buttery buttery buttery buttery shiny shiny shiny. shiny shiny shiny Gram Grams, Gram- Gram- Gram- Gram-I Grams Stain cocco- coccie cocco- cocco- diplo- pleo- bacilli- pleo- bacilli- bacilli- cocci morphic pleo- morphic. pleo- pleo- rods morphic rods' morphic morphic~ rods rods. rods Action on I - - - — - - Blood 0F F F F F F F F Glucose + + + + xylose +(3) +(5) - +(2) + - Lactose -(1) - - - + - Mannitol +(2) +(2) - -(l) + + Sucrose +(1) + + + + + Maltose - -(2) + + + + , Sorbitol +(2) +(2) - - + - ;Arabin- ose - -(1) - —(l) + - Inositol - - - -(2) + - Dulcitol - - - - ' - - Glucose Al-(l) Al-(l) A/+(l) A/- A/- A/- Xylose A(3)/- A(4)/- (-/-) -/- A/- -l- Lactose -(1)/- -/- -/- -/- Al- -/- Mannitol A(2)/- A(3)/- -/- -(1)/- Al- Al- Sucrose A(l)/- Al-(l) A/+(3) A/- A/- A/- Maltose -/-' -(1)/-(1) Al+ A/- Al- Al- Sorbitol A(2)/-(1) A(3)/-‘- -/- -(2)/- A/- -/- Arabinose -/- -(l)/- -/- -(l)/- A/- -/- Inositol -/- -/- -/- -(2)/-— A/- -/- Dulcitol -/- -/— -/- -/- -/- -/- 29 Table 3 (cont'd.) P. "(hen P. muZto- P. muZto— (muZto- cido (l3) cide (12) side?) P. pneumo- P.-hemo- Test (human) (animal) (6) tropica (4) Zytica (l) P. ureae (1) 10zflucose A A A A A A actose Alk.(l) Alk. Alk Alk.(1) A Alk. Catalase + + + + + + Oxidase + + + + + + Nut. Br. - + + + + + + MacConkey. - - - - - - ss - - - - - - Cetrimide - - - - - - Citrate - - - - - - Urea - -(l) + + - + N03 + + + + + + Indole + +(l) + + - - Slant 24-48 hr.‘ 24-48 hr. 48 hr. 24 hr. 24 hr. 24 hr. A(l) A A A A A TSI 7 days 7 days 7 days 7 days 7 days 7 days A(l) A A A A A . Butt 24-48 hr. 24-48 hr. 48 hr. 24 hr.- 24 hr. 24 hr. A A A A A A 7 days 7 days 7 days 7 days 7 days 7 days A(l) A A A A A Paper 24-48 hr. 24-48 hr. 24-48 hr. 24 hr. 24 hr. 24 hr. - trace-+ +n + + +~ 48 hr. - HZS- 7 days 7 days 7 days 7 days 7 days 7 d. - + + + + +‘ Butt 7 days 7 days 7 days 7 days 7 days 7 days un/vr -/- -/- -/- -/- -/- -/- Gelatin - - - - - - Litmus Milk — - — - - - Matility -‘ - - - - ‘ - Esculin hydrolysis - - - - - - 259 c. +1 + + + + + TGY 37° c. + + + + + + 42° C. - - - - - - Arginine - - - - - - Lysine - - - - -. - Ornithine +(1) +(2) - + - - Flo Agar 6 ' - - - - - Tech Agar - - - - - Pyronin +(2) :41) -(1) -(1) + + 30 Table 3 (cont'd.) I1 P. "as" P. muZto- P. multo- (malto- cido,(l3) cide (12) cide?) P. pneumo- P. hemo- Test (human) (animal) (6) tropica (4) Zytica (l) P. ureae Methyl violet -(3) -(2) +(1) -(1) - + Basic Fncahin -(1) -(2) + - - + Thionin +(2) +(1) -(1) + + + 31 strain 12 and strain 27 were also gas producers. Table 4 tabulates the. gas production data observed using a Durham tube. In addition to the reported production of gas from glucose (56), it was also observed from other carbOhydrates which these strains fermented. Three of the strains showed a.more limited ability to produce gas. A control tube of phenol red base alone was used to eliminate the possibility of a non-specific breakdown of another substance in the medium being responsible for the gas production.> Among the aerogenic strains, it is possible to correlate the carbo- hydrate.fermentation pattern with urease production. The 7 strains which were urease positive had an identical fermentation pattern while strain 12, which is urease negative, had a pattern more closely resembling P. multocida. The use of phenol red base browth in the Durham tube and phenol red semi-solid in other fermentation reactions allowed for a comparison of the 2 media. The phenol red base semi-solid was found to be a better medium for the slowbgrowing Pasteurella. The broth in the Durham tube» often gave delayed-results especially in the fermentation of xylose. The broth tube gave a false positive for 24—48 hours, but in 4-7 days the tube reverted to negative. The same strain when tested on semi-sclid medium gave a negative reading for the 7 days. It was observed that the organism grew more slowly in broth than in the semi-solid medium. The dye bacteriostasis experiment also resulted in a significant pattern of variation which is not apparent in Table 3. Table 5 presents a list of the strains which differed from the dye sensitivity pattern (Table 3) in their respective classes. Some of these variations seem to be more significant than random experimental errors. Strains 4, 6, 12, and 26 are all methyl violet positive although they are P. multocida- 32 Table 4. Gas production in aerogenic strains No. 14 15 16 17 j 18 19 27 12 Strain No. B92 B767 A8708 B704 B972 31682 C566 CE49 Glucose'. +/+* +/+ +/+- +/- +/+ +/+ +/+ +/+ xylose <-/—) <-/-> (-/-> <-/-> (-/-> <-/-) <-/-> <-/-> Mannitol -/- - - - - - - - — —/- —/- -/- Lactose -/6 -/- -/-. -/— -/- -/- -/- -/- Sucrose, +/‘ +[- +/- +/- +/+ +/+ +/+ -/- Maltose +/+ +/+ +/+ +/+ +/+ +/+ w +/+ -/.. Sorbitol -/- —/- -/- -/- -/- -/- -/- +/+ Arabinose: -/- -/- -/- -/- —/— -/_ -/- -/- Inositol -/- -/- -/— —/- -/- -/- -/- -/- Dulcitol -/- -/- -/- -/_ _/- -/- -/- -/- Control -/- -/- -/- -/- -/- -/- -/- -/- Urea + + + + + + + - *Acid production/gas production. Table 5. Dye bacteriostasis - variant strains No. 2 4 6 10 12 13 17 18 26 27 30 36 Strain CD CC CE CC CE CC ' No. 623 1193 28L 971 49 1150 B704: B972 M17 C566 P1059 P309 PyroninI (1 :500) - + - + + + - + + - + + Methyl violet (1:100) - + - +1 + - — + 1+ + _ + Basic fuschin (1 :100) - - - - - + + + - + +. - Thionin (1:200) - + +1 +1 + + - + + - + + 33 Strain 27, urease positive gas producer of animal origin, was shown to have an identical pattern to other aerogenic strains of Dr. Weaver. Due to the variation in each class, all the strains were compared to each other individually by calculating the percent similarities. The calculated S values range from 100% to 44%. Figure l is a similarity matrix based on this data with the strain order rearranged to form clusters of similar strains. This matrix shows 4 groups. Group I is a cluster of 7 strains, including the 6 strains of P. "gas" (multocida?) and strain 27. This is a very homogeneous cluster with most S values 90% or over. The largest group, IV, which is composed of P. multocida of animal and human origin exhibits a great deal of heterogeneity. In this cluster, there is no separation of animal and human strains. In most cases, there is a high degree of similarity among them. However, there are some strains near the edge of the cluster which possess a low similarity. Strains 4 and 6 show a greater similarity to each other and strain 26 than the rest of the group. In general, the strains showing the lowest percent simi- larities are those from canine or feline origin or local infections of man. A simple relationship with this dissimilar group cannot be assumed because these-strains often differ markedly from each other. Between Groups I and IV are.the 2 intermediary groups. Group II is the 3 unre- lated strains of P. hemolytica, P. ureae, and strain 12. Group III is composed of the 4 strains of P. pneumotropica. The similarity matrix shows no cross relationship between Groups I and IV. A greater degree of similarity is seen between Groups I and III within a range of 70-79%. Group I exhibits little similarity to strain 12 which is the other aerogenic strain, but this group does show a simi- larity of 70-74% to P. ureae which is the other urease positive strain. It is unrelated to P. hemolytica. 34 H Hm n mm 5 mm «N NN 0H Hm mm m MH on m H ON w q mm o «m mm H mH 0H mH «H NH mm H HN m mm. mm em mm OH wfl Hm as. m >H MH msooo N on m 3. ON w e mm L em mm om HHH mm moose NH mm HH mm nsouo wH, mH 0H H 3 1:55 H H N; 35 Strain 12 provides an interesting similarity relationship. It shows its highest-correlation of 80-84% with strains 10 and 26 which are members of the group of P. muZtocido. DISCUSSION In this study, 37 strains of PasteureZZa which were initially grouped by species origin and disease process were analyzed and compared to each other by biochemical reactions and other tests. Of special interest in this study were the strains which were found to be aerogenic. Most of the cultures were in the rough or blue form. There were some iridescent variants, but no attempt was made to eliminate them because virulence testing was not being carried out. It has been observed that colonial variation does not significantly affect certain biochemical characteristics such as indole production or carbohydrate fermentation (18). Using the Durham tube to observe gas production, 8 of the 37 strains were shown to be aerogenic (Table 4). In all cases, the gas produced.was in small quantities. Six of these 8 strains had been reported in the literature as positive for gas production from glucose, but the presence or absence of gas production with other sugars was not noted (56). It was not possible to repeat this gas production from glucose in the case of 1 strain, B-704. However, this strain was observed to produce gas from maltose. The other 5 strains also produced gas from maltose and in 3 cases from sucrose. In addition to these 6 strains, 2 other strains were observed to be aerogenic. C-566, a canine isolate, produced gas from glucose, maltose, and sucrose, while CE-49, a human strain, produced gas from.g1ucose and sorbitol. The production of gas from different carbohydrates by a single strain of Pasteurella has been previously reported. Talbot and Sneath 36 37 (54) isolated a strain of P. muZtoeido from a dog which produced small quantities of gas from sucrose, maltose, mannose, and trehalose. The 8 aerogenic strains studied did not ferment the same carbohydrates. A direct correlation between carbohydrate fermentation and urease produce tion was observed. All the urease positive strains fermented glucose, sucrose and maltose and were negative for sorbitol, while the one urease negative strain fermented glucose and sorbitol but did not ferment sucrose and maltose. There were, therefore, 7 closely related aerogenic strains correspond- ing in biochemical reaction to P. "gas" (multocida?). The other aerogenic strain was unrelated. In addition to the biochemical reactions and other tests which are. used to classify PasteureZZa, an attempt was made to characterize this species by dye bacteriostasis. Rosenbusch and Merchant (46) tried unsuc— cessfully to do this in 1939. With a modified version of the dye sensi- tivity disc testing method of Pickett, Nelson, Hoyt and Eisenstein (41) a repeatable pattern of dye sensitivity was obtained. PasteureZZa mul- tocida, P. pneumotropica, and P. "gas" (multocido?) had unique patterns. The strain of P. hemolytica studied and P. muZtocida could not be dis- tinguished by this method. If the dye sensitivity testing method proved consistent with a large number of strains, it might become a useful technique to use with other tests to subgroup P. multocida into a series of biotypes similar to the- work done on BruceZZa. An indication of this possibility was given‘whenv 3 strains of P. multocida which varied in a consistent manner from.the "typical" dye sensitivity pattern of the other strains proved to be more. closely related to each other than the majority of other P. multocida strains when analyzed in the similarity matrix (Figure 1). 38 The aerogenic strains of Pasteurella could not be adequately com- pared to each other by using tabulated charts of biochemical data. An attempt was made to analyze the data according to the method used in Smith and Thal (50). They manually constructed a similarity matrix comparing 25 strains of Pasteurella of various species. They found this method to be very useful for comparing small numbers of strains. If more extensive studies are undertaken, computer analysis is necessary. With this method, the 8 aerogenic strains could be compared indi- vidually to each other and to the other strains tested. All the strains rearranged in the similarity matrix fell into 4 broad groups. Group I consisted of 6 aerogenic strains of Dr. Weaver and strain 27. Group II consisted of 3 unrelated strains-of P. hemolytica. P. ureae, and strain 12. Group III was composed of strains of P. pneumotropica. Group IV was made up of human and animal strains of P. muZtocida. The validity of the similarity matrix constructed was confirmed by a comparison of similarities between species established by this matrix and the one constructed by Smith and Thal (50). The percent similarity of P. multocida and P. pneumotmpica, P. muZtocida and P. hemolytica, P. ureae and P.~muZtocida are generally in agreement between the 2 matrices. When a slight variation occurred, the percent similarity of the experimental matrix is lower than the percent similarity contained in the matrix of Smith and Thal. When the Group I strains are compared to the other groups in.the experimental matrix, it was found to be the most closely related to P. pneumotropica and P. ureae. It is felt that the similarity is probably closer to P. pneumotropica strains than P. ureae. The limited number of AP. pneumotropica used in this study had some irregular carbohydrate fer— mentation reactions, and proved to be ornithine decarboxylase positive 39 even though this species is considered to be typically negative for this test. If a typical strain of P. pneumotropica (50) was compared to the P. "gas" (multocidd?) strains, the S values would be in the range of 902. PasteureZZa areas is also less likely to be the most closely related species because P. ureae and P. "gas" (multocida?) differ in indole reaction which is considered a key reaction of the PasteureZZa species in classical taxonomy. The Group I aerogenic.strains differed completely from strain 12, the urease negative aerogenic strain. Strain 12 showed its greatest similarity to some of the P. multocidb strains. It was not similar to all of the strains. Group IV varied considerably within the group. There have been repeated attempts to correlate the variations of P. multocidb‘with species differences of host or some other traceable variable. In this matrix, it was the canine, feline and local infections of man which were the most dissimilar and were found on the fringes of the group. Canine and feline strains have been observed to differ from the species as a whole (48). The subgrouping of P. multocidh is probably highly complex. Strains of Group IV dissimilar to the more common varieties of the group are often dissimilar in varying degrees to each other. In summary, the aerogenic strains of Dr. Weaver and strain-27 seem most closely related to P. pneumotropica. Strain 12 seems to be the most closely related to P. multocida. It would appear that the best way in which to taxonomically rank these strains are as varieties of the previously established species. To establish new species would probably only complicate an already complex situation. It must be remembered that any bacterial taxonomic classification system is at best a series of approximate points on a scale of natural progression which is constantly adapting and making biologic adjustment. BIBLIOGRAPHY 10. 11. 12. BIBLIOGRAPHY Bain, R. V. S. 1955. British Veterinary Journal 111: 492-498. Studies on haemorrhagic septicemia of cattle. IV. A preliminary examination of the antigens of Pasteurella multocida type I. ' Beers, R. C., J. Fischer, S. Megraw and W. R. Lockhart. 1962. Journal of’General Microbiology 28: 641-652. A comparison of methods . for computer taxonomy . Branson, D. and F. Bunkfeldt, Jr. 1967. The American Journal of Clinical Pathology 48: 552—555. Pasteurella multocida in animal bites of humans. Breed, R. S., E. D. C. Murray, N. R. Smith, et al. Bergey's Manual of’Determinative Bacteriology, 7th edition: 395-402. The Williams' and Wilkins Company. Brennan, P. C., T. E. Frita, and R. J. Flynn. 1965. Laboratory Animal Care 15: 307-312. Pasteurella pneumotropica: Cultural and biochemical characteristics, and its association with disease in. laboratory animals. ’ Briefman, L., and K. E. Yaw. 1958. Journal of'Bacteriology 75: 236. Monosaccharide isolated from the capsule of Pasteurella multocida. Carter, G. R. 1952. Canadian Journal of’Comparative Medicine 16: 150-152. Some comments on pasteurellosis. Carter, G. R. 1955. American Journal of'Veterinary Research 18: 210-213. Studies of Pasteurella multocida. II. Identification of antigenic characteristics and colonial variants. Carter, G. R. 1962. Canadian Journal of'Public Health 53: 158-161. Animal serotypes of Pasteurella multocida from.human infections. Carter, G. R. 1963. The Veterinary Record 75: 1264-1265. Proposed modification of the serological classification of Pasteurella multocida. Carter, G. R. 1967. Diagnostic Procedures in Veterinary Bacteriology and.Mycology. Charles C. Thomas. Carter, G. R. 1967. .Advances in Veterinary Science 11: 321-379. Pasteurellosis:v Pasteurella multocida and Pasteurella hemolytica. 40 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 41 Carter, G. R. and R. V. S. Bain. 1960. Veterinary.Heviews and Annotations 6: 105—128. Pasteurellosis (Pasteurella multocida). A review stressing recent developments. Carter, G. R. and C. H. Bigland. 1953. Canadian Journal of’Compara- tive Medicine 17: 473-479. Dissociation and virulence in strains of Pasteurella multocida isolated from a variety of lesions. Controni, G. and R. S. Jones. 1967. American Journal of'Medical Technology 33: 379-387. Pasteurella meningitis. A review of the literature. Elberg, S. S. and C. Ho. 1950. Journal of Comparative Pathology and Therapeutics 60: 41-50. Studies on dissociation in Pasteurella multocida. Hawkins, L. G. 1963. Journal of Bone and Joint Surgery 51: 363-366. Local Pasteurella infections. Heddleston, K. L., L. P. Watko and P. A. Rebers. 1964. Avian Disease 8: 649-657. Dissociation of a fowl cholera strain of Pasteurella multocida. Henderson, A. 1963. A. van Leeuwenhoek Journal ofHMicrobiology and Serology 29: 359-367. Pasteurella multocida in man. A review of the literature. Henriksen, S. D. 1962. Acta Pathologica et Microbiologica Scandinavica 55: 355-356. Some Pasteurella strains from human respiratory tract. A correction and supplement. Heyl, J. G. 1963. A. van Leeuwenhoek Journal of.Microbioloay and Serology 29: 79-83. A study of Pasteurella strains from animal sources. Higgins, D. A. 1969. British Veterinary Journal 125: 622-634. Some studies of fowl cholera in Hong Kong. Hoag, W. C., P. W. Wetmore, J. Rodgers and H. Meier. 1962. Journal of’Infectious Disease 111: 135-140. A study of latent Pasteurella infection in a mouse colony. Hubbert, W. T. and M; N. Rosen II. 1970. American Journal of' Public Health and the Nation's Health 60: 1103-1108. Pasteurella multocida infection due to animal bite. Hubbert, W. T. and M. N. Rosen II. 1970. American Journal of Public Health and the Nation's Health 60: 1109-1117. Pasteurella multocida infection in man unrelated to animal bite. Hugh, R. and E. Leifson. 1953. Journal of'Bacteriology 66: 24-26. The taxonomic significance of fermentative versus oxidative metabol- ism of carbohydrates by various Gram-negative bacteria. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 42 Jawetz, E. 1950. Journal of’Infectious Disease 86: 172-183. A pneumotrOpic Pasteurella of laboratory animals. I. Bacteriology and serological characteristics of the organism. King, E. 0. 1964. The Identification of'Unusual Gram-negative Bacteria. 64th Annual Meeting of American Society of Microbiology. Published as training material by CDC Public Health Service, Atlanta, Georgia. antigens of Pasteurella multocida type 1. I. Capsular polysac- charides. ' Lignieres, J. 1901. Annales de l'institute Pasteur 15: 734-736. Contribution a 1'etude et la classification des septicemie hemorrhagiques les Pasteurelloses. Little, P. A. and B. M. Lyon. 1943. American Journal of Veterinary Research 4: 110-112. Demonstration of serological types within- the non-hemolytic Pasteurella. Maccabe, A. F. and N. Conn. 1968. Scottish Medical Journal 13: 242-244. The isolation of Pasteurella septica following dog and cat bites. A report of five cases. Merchant, I. A. and R. A. Packer. 1967. Veterinary Bacteriology and Virology. Iowa State University Press. Mbeller, V. 1955. Acta Pathologica et Microbiologica Scandinavica 34: 115-126. Simplified test for some amino acid decarboxylases and for arginine hydrolase. ‘ Namioka, S. and D. W. Bruner. 1963. Cornell Veterinarian 53: 41-53. Serological studies on Pasteurella multocida. IV. Type- distribution of the organism on the basis of their capsule and 0 groups. Namioka, S. and M. Murata. 1964. Cornell Veterinarian 51: 522- 528. Serological studies on Pasteurella multocida. III. 0 anti- genic analysis of cultures isolated from various animals. Newsom, I. E. and F. Cross. 1932. Journal of American Veterinary Association 80:n333: 711-719. Some bipolar organisms found in pneumonia in-sheep. Olson, J. R. and T. R. Meadows. 1969. The American.Journal of' Clinical Pathology 51: 709-710. Pasteurella pneumotropica infection resulting from a cat bite. Owen, C. R., E. 0. Bucker, W. L. Jellison, D. B. Lackman and J. F. Bell. 1963. Journal of’Bacteriology 87: 676-683. Comparative studies of Francisella tularensis and Francisella novicida. Philip, C. B. and C. R. Owen. 1961. Int. Bull. Bact. Nomenclature, Tamonomy 11: 67. Comment on the nomenclature of the causative agent of tularemia. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 43 Pickett, M. J., E. L. Nelson, R. E. Hoyt and B. E. Eisenstein. 1952. Journal of’Laboratory and Clinical Medicine 40: 200-205. , Speciation with Genus Brucella. I. Dye sensitivity of smooth Brucellae. Prince, G. H. 1969. Journal of'Comparative Pathology and Thera- peutics 79: 179. Production of antigens of Pasteurella multocida and detection of antibodies against them in bovine serum. Prince, G. H. and J. E. Smith. 1966. Journal of'Comparative Pathology and Therapeutics 76: 315—320. Antigenic studies on Pasteurella multocida using immunodiffusion techniques. 11. Relationships with other Gram-negative organisms. Roberts, R. S. 1947. Journal of Comparative Pathology and Thera- peutics 57: 261-278. An immunological study of Pasteurella septica. Rogers, R. J. and J. K. Eldes. 1967. Australian Veterinary Journal 43: 81-82. Purulent leptomeningitis in a dog associated with an aerogenic Pasteurella multocida. Rosenbusch, C. T. and I. A. Merchant. 1939. Journal of'Bacteri- ology 37: 69-89. A study of the hemorrhagic septicemic.Pasteurellae. Schmidt, E. C. ‘H., L. U. Truitt, and M. L. Koch. 1970. American Journal of’Clinical Pathology 54: 733-736. Pulmonary abscesses with empyema caused by Pasteurella multocida. Report of a fatal case. Smith, J. E. 1959. The Journal of’Comparative Pathology and Therapeutics 68: 315-328. Studies on Pasteurella septica. 11. Some cultural and biochemical properties of strains of different‘ host species. Smith, J. E. 1959. The Journal of’Comparative Pathology and Therapeutics 69: 231-235. Studies on Pasteurella septica. III. Strains from human beings. Smith, J. E. and E. Thal. 1965. Acta Pathologica et.Microbiologica Sbandinavica 64: 213-223. A taxonomic study of the genus Pasteurella using a numerical technique. Sneath, P. H. A. 1956. Journal of General Microbiology 15: 70-98. Cultural and biochemical characteristics of genus Chromobacterium. Sneath, P. H. A. The construction of taxonomic groups. Microbial classification. 12th Symposium of Society for General Microbiology, ed. Ainsworth, G. C. and P. H. A. Sneath, 1962. Swartz, M. N. and L. J. Kunz. 1959. New England.Journal ofVMedicine 261: 889-893. Pasteurella multocida infections in man. Talbot, J. M; and P. H. A. Sneath. 1960. Journal of'General hficrobiology 22: 303-311. A taxonomic study of Pasteurella septica, especially of strains isolated from human sources. 55. 56. 57. 58. 44 Topley, G. S. and A. A. Miles et al. 1964. Principles of’Bac- teriology and Immunity. Vol. II. The Williams and Wilkins Company. Weaver, R. E., Unclassified Groups of Aerobic Gram-negative Bacteria Isolated from Clinical Specimens. Published from the Meeting of the American Society of Microbiology. Winton, F. W. and N. S. Nair. 1969. Microbios 2: 155-162. Pasteurella pneumotropica isolated from a dog bite wound. Wong, P. and C. H. Chan-toeh. 1967. Journal of Clinical Pathology 17: 107-110. A report of three strains of Pasteurella septioa isolated in Hong Kong. “"' Au“ 4 HICHIGQN STQTE UNIV. LIBRQRIES 31293010091068