Quick viewing(Text Mode)

Mycobacterium Avium, and Description of Mycobacterium Avium Subsp

Mycobacterium Avium, and Description of Mycobacterium Avium Subsp

INTERNATIONALJOURNAL OF SYSTEMATICBACTERIOLOGY, July 1990, p. 254-260 Vol. 40, No. 3 0020-7713/90/030254-07$02. oo/o Copyright 0 1990, International Union of Microbiological Societies

Numerical of Mycobactin-Dependent Mycobacteria, Emended Description of Mycobacterium avium, and Description of Mycobacterium avium subsp. avium subsp. nov. , Mycobacterium avium subsp. subsp. nov. , and Mycobacterium avium subsp. silvaticum subsp. nov.

MARIE-FRANCOISE THOREL,l MICAH KRICHEVSKY ,2 AND VERONIQUE VINCENT L&VY-FREBAULT3* Centre National d'Etudes Ve'te'rinaires et Alimentaires, Laboratoire Central de Recherches Ve'te'rinaires, 94703 Maisons Arfort Cedex, France'; National Institutes of Health, Department of Health and Human Services, Bethesda, Maryland 208922; and Unite' de la Tuberculose et des Mycobacte'ries, Institut Pasteur, 75724 Paris Cedex 15, France3

We performed a numerical taxonomy analysis of 38 Mycobacterium paratuberculosis and related mycobac- terial strains, including wood pigeon mycobacteria; this analysis was based on 22 tests, which were selected for their potential discriminative value from a total of 51 tests studied and produced four well-defined clusters. Cluster 1 contained the M. paratuberculosis strains, including two strains isolated from Crohn's disease patients; cluster 2 contained Mycobacterium avium and Mycobacterium intracellulare reference strains; cluster 3 consisted of the wood pigeon mycobacteria; and the only strain in cluster 4 was M. paratuberculosis 316F, which is used for antigen and vaccine production. Strains in cluster 1 were mycobactin dependent even when they were subcultured, whereas strains in cluster 3 were unable to grow on egg medium and their growth was stimulated by pH 5.5. Growth stimulation by pyruvate, resistance to D-cycloserine (50 pg/ml), and alkaline phosphatase activity also were characteristics that were useful for discriminating between clusters 1 and 3. The results of previous DNA-DNA hybridization studies have demonstrated that M. avium Chester 1901, M. paratuberculosis Bergey et al. 1923, and the wood pigeon mycobacteria belong to a single genomic , and we propose that the name of this species should be M. avium. On the basis of the results of previous genomic analyses based on restriction fragment length, the results of polymorphism studies, and DNA patterns determined by field inversion gel electrophoresis as well as the results of our phenotypic study, we propose that the species should be divided into subspecies which correspond to pathogenicity and host range characteristics. An emended description of M. avium Chester 1901 and descriptions of M. avium subsp. avium subsp. nov., M. avium subsp. paratuberculosis subsp. nov., and M. avium subsp. silvaticum subsp. nov. are presented; strains ATCC 25291, ATCC 19698, and CIP 103317 are the type strains of the three new subspecies, respectively.

Mycobacterium avium and Mycobacterium paratubercu- relatedness of serovars of M. avium and M. intracellulare (1, losis have been shown to belong to a single genomic species 2); this finding was recently confirmed by workers who used by DNA-DNA hybridization studies (23, 40, 52) and deter- DNA probes specific for either M. avium or M. intracellu- minations of the thermal stabilities of hybrids (52). The close lare (38). relationship between M. avium and M. paratuberculosis was However, other genomic approaches have revealed some demonstrated by the results of a comparison of the se- differences among M. avium, M. paratuberculosis, and the quences of a 383-base-pair segment of the gene encoding the wood pigeon mycobacteria. The restriction profiles of a large 65-kilodalton mycobacterial antigen. The sequences of the set of strains revealed stringent conservation of the genetic entire fragment have been found to be identical for M. avium composition of M. paratuberculosis and high levels of het- and M. paratuberculosis, whereas the oligonucleotides erogeneity within M. avium (9, 45, 49). Restriction patterns within this segment consist of probes that are useful for may be difficult to interpret because usually they rely on the differentiating M. avium from tubercle bacilli and Mycobac- detection and comparison of more than 50 closely spaced terium fortuitum (22). The results of DNA-DNA hybridiza- bands. In order to simplify such analyses the following two tion studies have supported the identification of strains techniques have been developed: restriction fragment length isolated from patients suffering from Crohn's disease as M. polymorphism (RFLP) analysis, which results in simplified paratuberculosis but have failed to differentiate these strains patterns as determined by hybridization of endonuclease from M. avium (31, 53). Strains designated wood pigeon mycobacteria also have been shown to belong to the cleavage products with specific DNA probes, and pulsed- genomic species containing the M. avium and M. paratuber- field gel electrophoresis (especially field inversion gel elec- culosis type strains on the basis of determinations of trophoresis [FIGE]), which allows the separation of large percentages of DNA-DNA homology in which a spec- DNA fragments and thus an analysis of patterns composed trophotometric method was used (39) and on the basis of of few bands after chromosomal digestion by endonucleases hybridization with radiolabeling in solution (23, 52). Myco- having low-frequency cleavage sites (6). The two methods bacterium intracellulare is a genetically distinct species, as have comparable sensitivities (27). demonstrated by Baess in her pioneering studies of the DNA RFLP analysis of the ribosomal or other DNA genes has shown that mycobacterial strains isolated from patients with Crohn's disease (namely, strains Ben, Dominic, and Linda) * Corresponding author. are identical to M. paratuberculosis (8, 30) and that M.

254 VOL. 40, 1990 MYCOBACTIN-DEPENDENT MYCOBACTERIA 255 paratuberculosis is distinguishable from the other mycobac- medium. Drug susceptibility was studied on Middlebrook teria examined, including different M. avium serotypes (10) 7Hll medium, and resistance was determined by using the and one wood pigeon (32). The differences between proportion method (16). determinations were RFLP patterns could be related to the presence of a repeti- performed by using the method of Daff6 et al. (12). tive DNA sequence present in M. paratuberculosis but Numerical taxonomy analysis. The numerical taxonomy absent from all M. avium serotypes and from 18 other analysis was based on the simple matching coefficient, and mycobacterial species (10, 19). RFLP analysis has enhanced similarities were calculated by using the unweighted average the detection of heterogeneity among the M. avium sero- linkage method (44). Tests were designated according to types, as suggested by previous studies of restriction pat- their RKC code numbers (37). terns (21, 45). FIGE patterns obtained with restriction enzyme DraI RESULTS AND DISCUSSION confirmed the high level of homogeneity of M. paratubercu- losis strains, including strains isolated from patients with The 38 strains were subjected to 51 tests, including deter- Crohn’s disease, and the differentiation of this species from minations of cellular and colonial morphology, optimal tem- M. avium. FIGE analysis also demonstrated the homogene- perature for growth, growth on different media, production ity of the wood pigeon mycobacteria. All of the strains tested of pigment, tolerance to inhibitors, resistance to , exhibited the same unique pattern, which differed from the enzymatic activities, and mycolic acid content. Tests which patterns of M. avium and M. paratuberculosis (27). gave positive or negative results for all strains were deleted RFLP and FIGE analyses have established that M. avium, from the numerical analysis. After we deleted these tests M. paratuberculosis, and the wood pigeon mycobacteria can (which are reported below in the emended description of M. be recognized and differentiated despite the fact they belong avium), the numerical analysis was performed by using the to a single species. The RFLP or FIGE patterns correlate 22 remaining characteristics (Table 2) and generated four well with differentiation based on pathogenicity and host well-defined clusters (Fig. 1). range. M. paratuberculosis is the agent of paratuberculosis Cluster 1, containing 18 strains, included the type strain of in ruminants (35) and may be involved in Crohn’s disease in M. paratuberculosis and the two strains isolated from pa- humans (7), whereas wood pigeon mycobacteria may cause tients suffering from Crohn’s disease. The other strains in paratuberculosis in bovine calves and in birds this cluster were isolated from , goats, and ; (11, 28, 29, 43). Both organisms are isolated only from diagnosis of Johne’s disease was confirmed for these animals infected tissues, unlike M. avium, which is widely encoun- by clinical evidence. The mean internal matching score for tered in the environment and is the agent of tuberculosis in this cluster was 86%. The strains were characterized by an birds and various in other animals and in humans, absolute requirement for mycobactin for growth (Table 2). A especially in disseminated infections of patients suffering FIGE analysis of seven strains in cluster 1 (strains 1,4, 6,9, from acquired syndrome (36, 48, 50). 11, 13, and 18 in the numerical taxonomy matrix [Fig. 11) Thus, it would be worthwhile to establish phenotypic tests gave identical patterns, which were characteristic of M. that agree with the genomic classification. The differential paratuberculosis (27). characteristics proposed previously, based on biochemical Cluster 2, which had a mean internal matching score of tests, fatty acid composition, peptidoglycolipid content, and 88%, consisted of six strains and included the type strains of immunodiffusion analysis results, did not allow reliable M. avium and M. intracellulare. This demonstrates that identification of the different taxa (4, 5, 13, 33, 40, 42). The differentiation of these two species cannot be achieved by purpose of this study was to evaluate biochemical tests that using characteristics that are currently considered for nu- are easily applied to this set of difficult-to-grow mycobacte- merical analysis (47) but requires other methods, such as rial strains for consistent identification of M. paratuberculo- hybridization to a specific probe (38) or a study of T- sis and wood pigeon mycobacteria. We propose a change in catalases (46). nomenclature in order to recognize the taxonomic implica- All of the strains in cluster 2 except the M. avium type tions of the numerous recent genomic and phenotypic stud- strain have been found to be NaCl resistant, although ies of these closely related mycobacteria. Mycobacterium triviale is usually considered to be the only slowly growing mycobacterial species that is able to grow on MATERIALS AND METHODS 5% NaCl (18). However, it should be noted that the test in this study was performed on Middlebrook 7H10 medium. Mycobacterial strains. Table 1lists the 38 strains which we This may explain the discrepancy with the results of previ- studied. All of the strains isolated from organs or feces were ous studies of NaCl resistance, in which Lowenstein-Jensen mycobactin dependent, at least when they were first iso- medium was used. lated. Strain St18 was previously identified as an M. paratuber- of the strains. Initial isolation of all of the strains culosis strain and was selected as a working strain until was performed on Herrold medium, Middlebrook 7H10 Merkal reported the unusual cultural properties of this medium, or Middlebrook 7Hll medium devoid of or supple- strain, its mycobactin independence, and the absence of mented with 2 mg of mycobactin per liter (16), and an pathogenicity for calves (34). Further studies of peptidogly- adequate medium was selected for subsequent subcultures. colipid content, RFLP data, and conventional and pulsed- The M. avium and M. intracellulare type strains were field electrophoresis data established that this strain is in fact subcultured on Lowenstein-Jensen medium. an M. avium serotype 2 strain (5, 10, 27, 49). Identification tests. The tests shown in Table 2 have been Strain 2103 was also identified as an M. paratuberculosis described previously (14-16); when necessary, some of these strain at first since this strain was mycobactin dependent and tests were performed on media supplemented with mycobac- was isolated from a goat mesenteric lymph node. However, tin. Stimulation by pyruvate (4.1 g/liter) and resistance to strain 2103 differed from M. paratuberculosis by the follow- NaCl were tested on Middlebrook 7H10 medium; the ability ing characteristics: mycobactin independence after subcul- to grow at pH 5.5 was determined on Middlebrook 7Hll turing, the ability to grow in the presence of NaC1, suscep- 256 THOREL ET AL. INT. J. SYST.BACTERIOL.

TABLE 1. Strains used in this study Serial no. Laboratory designation“ Source Originb 1 3418 Bovine ileocecal valve E.N.V. Alfort, France 2 1277 Bovine feces SaBne et Loire, France 3 12998.5 Bovine feces Corrkze , France 4 ATCC 19698T M. paratuberculosis bovine feces ATCC 5 1528 Bovine feces SaBne et Loire, France 6 CD Lyon Human intestine biopsy RhBne, France 7 12998.12 Bovine feces Corr2ze, France 8 12998.2 Bovine feces Corrkze, France 9 147.89 Goat mesenteric lymph node Deux Skvres, France 10 1466 Bovine feces SaBne et Loire, France 11 1077 Sheep Iowa 12 1394 Bovine feces SaBne et Loire, France 13 7912 Bovine rectum Haute SaBne, France 14 8642 Goat mesenteric lymph node SaBne et Loire, France 15 12998.9 Bovine feces Correze, France 16 12998.4 Bovine feces Corr&ze,France 17 1464 Bovine feces SaBne et Loire, France 18 2569 Human surgical specimen Lelystad, The Netherlands

19 Deer mesenteric lymph node Aube, France 20 Mycobactin preparation reference strain Iowa 21 Mesenteric lymph node Deux Skvres, France 22 ATCC 25291T M. avium hen liver ATCC 23 M21 Wood pigeon Compton, United Kingdon 24 ATCC 13950T M. intracellulare ATCC

25 V172 Deer Copenhagen, Denmark 26 2828 Wood pigeon liver Yvelines, France 27 10059 Wood pigeon liver Gers, France 28 853 Wood pigeon liver and spleen Y velines, France 29 4776 Wood pigeon liver Gers, France 30 5329 Wood pigeon spleen Yvelines, France 31 3135 Wood pigeon liver Gers, France 32 6409 Wood pigeon liver and spleen Val d’Oise, France 33 6861 Wood pigeon liver and spleen Val d’Oise, France 34 495 Wood pigeon liver and spleen Loiret, France 35 4043 Crane liver Haut Rhin, France 36 7362 Wood pigeon spleen Loiret, France 37 3585 Wood pigeon liver Gers, France

38 316.F Antigen and vaccine preparation reference strain Weybridge, United Kingdon

~

a T = type strain. ATCC, American Type Culture Collection, Rockville, Md. ; E.N.V., Ecole Nationale VktCrinaire. tibility to cycloserine (50 pg/ml), and production of alkaline paratuberculosis in cattle and tuberculosis in hens (11, 28, phosphatase. Pulsed-field electrophoresis patterns con- 43). firmed the differentiation of this organism from M. paratu- Cluster 3 had a mean internal matching score of 91%; this berculosis (27). In addition, experimental infections of cluster was composed of 13 strains (12 strains isolated from calves revealed pathogenicity similar to that of M. avium birds and 1 strain isolated from a mammal). All of the strains (20) since strain 2103 induced acute infections and killed the isolated from birds produced lesions in livers and spleens animals within 36 to 60 days after inoculation (43). similar to those observed in birds infected with M. avium. Strain 12547 was isolated from a mesenteric lymph node of Concomitant pathogenicity for calves was confirmed by a deer. This strain did not exhibit the characteristic DNA experimental infections with two strains (strains 6861 and pattern of M. paratuberculosis or wood pigeon mycobacteria 6409) (43). Similarly, strain VI72, which was isolated orig- as determined by pulsed-field electrophoresis (data not inally from a wild ruminant suffering from Johne’s disease, shown). has been shown to cause avian tuberculosis in hens and Strain M21, which was isolated from a wood pigeon, fell paratuberculosis in calves (25, 43). Strains in this cluster are into cluster 2 because of cultural characteristics that were characterized by the inability to grow on egg medium and similar to those of M. avium. Growth occurred on egg stimulation of growth by pH 5.5 (Table 2). Four strains medium and was not prevented by p-nitrobenzoate or by belonging to this cluster were examined by FIGE and sodium chloride on Middlebrook 7H10 medium. Growth was produced identical patterns, which were different from the not stimulated at pH 5.5. However, strain M21 exhibited a profiles observed for M. paratuberculosis and M. avium pulsed-field pattern that was characteristic of the wood strains (27). pigeon mycobacteria (data not shown). Experimental infec- Cluster 4 consisted of a single strain, M. paratuberculosis tions demonstrated the ability of this strain to produce 316F. This strain has been extensively subcultured since it VOL.40, 1990 MYCOBACTIN-DEPENDENT MYCOBACTERIA 257

TABLE 2. Variable and differential characteristics of M. paratuberculosis and related mycobacteria

~ Cluster 1 Cluster 3 RKC Cluster 4 Characteristic 13 I1 Et4q)

Rods 003008 - 5/18' 13/13 Coccobacillary 003026 + 13/18 0113 Colonies smooth 016030 - 0118 0113 Colonies rough 016031 + 18/18 13/13 Growth at 42°C 017035 + 16/18 13/13 Growth stimulation by pyruvate' 098193 + 18/18 0113 Growth stimulated in medium at pH 5.5' 099031 - 0118 13/13 Mycobactin (2 mglliter) is required for growth' 099038 + 18/18 0113 Growth on egg medium' 099030 + 18/18 2/13 5% NaCl tolerance on Middlebrook 7H10 medium 018006 - 0118 0113 Enzymatic activities Catalase after heating at 68°C 024425 + 18/18 13/13 Acid phosphatase (2 h) 034136 - 0118 1/13 Alkaline phosphatase (2 h)' 034137 - 1/18 13/13 Trehalase (5 h) 034142 - 6/18 3/13 Growth in the presence of para-Nitrobenzoate (500 pglml) 016264 4/13 0113 D-Cycloserine (30 pglml) 098183 + 18/18 11/12 D-Cycloserine (50 pg/ml)' 099033 + 17/18 0/13 Ethambutol (7.5 pglml) 099034 + 18/18 12/13 (50 kg/ml) 099035 + 14/18 13/13 Rifampin (1 pglml) 040164 + 13/18 1/13 Ansamycin (1 pg/ml) 099036 + 15/18 1/13 Streptomycin (2 pg/ml) 019510 + 13/18 7/13

a Strain 6409 (= CIP 103317) is proposed as the type strain of M. avium subsp. silvaticum. Number of strains positive/total number of strains in the cluster. Key test for differentiation of clusters 1 through 4. was first isolated in 1940 in a Weybridge, United , (38, 46). Determinations of the serotypes by detecting pep- laboratory (17) and is used as a live vaccine (17, 24). It is tidoglycolipids (mycoside C) are valuable but have to be characterized by negative responses to all of the tests that interpreted carefully because DNA-DNA hybridization re- are useful for discriminating among clusters 1, 2, and 3 sults lead to a classification of the serotypes between M. (Table 2) and differs from other M. paratuberculosis strains avium and M. intracellulare that is different from the by its inability to grow on egg medium, its mycobactin Schaefer classification (2, 38, 51). independence, its lack of stimulation of growth by pyruvate, The misidentification of wood pigeon mycobacterium and its susceptibility to cycloserine. strain M21 was also due to the use of primarily cultural tests Numerical analysis split the 38 strains which we studied since this strain had peculiar growth characteristics which into four well-defined clusters (Fig. 1). Cluster 1 corresponds were more like those of M. avium than those of the other to M.paratuberculosis; cluster 2 contains six strains, includ- wood pigeon mycobacteria, especially growth on Lowen- ing the type strains of M. avium and M. intracellulare; stein-Jensen medium, tolerance to inhibitors (such as p- cluster 3 contains the wood pigeon mycobacteria; and clus- nitrobenzoate), and absence of stimulation by pH 5.5. ter 4 contains M. paratuberculosis 316F, which is used for vaccination. This separation based on cultural and biochem- M. paratuberculosis 316F, the only member of cluster 4, ical tests is highly correlated with pulsed-field patterns and has been subcultured extensively since 1940 for antigen and with pathogenicity, as confirmed by experimental infections vaccine production. This may explain the phenotypic and in hens and calves (11, 27, 28, 43). genomic differences now observed between this strain and The mean internal matching scores range from 85 to 91% the other M. paratuberculosis strains, as shown by our for clusters 1 through 3. These percentages must be com- numerical analysis as well as by the peculiar pulsed-field pared with circumspection with values reported in other pattern of this organism (27). numerical analyses which generated clusters corresponding M. paratuberculosis and wood pigeon mycobacteria can to different mycobacterial species. In this numerical analysis be easily separated on the basis of their responses to six we used a set of highly related strains and relied on selected tests. M. paratuberculosis is characterized by a mycobactin tests. Our selection resulted in the high clustering levels requirement, the ability to grow on egg medium, tolerance to observed in this study. cycloserine (50 pg/ml), growth stimulation by pyruvate but Of the 38 strains studied, 2 were not correctly identified by not by pH 5.5, and an absence of alkaline phosphatase the tests which we used. The M. intraceflufaretype strain production. Wood pigeon mycobacteria have the opposite and wood pigeon mycobacterium strain M21 both fell into characteristics in these six tests. This set of tests includes cluster 2 along with M. avium strains. This misidentification five cultural tests and only one biochemical test. Lack of may be explained by the nature of the tests. M. avium and balance between cultural and biochemical discriminative M. intracellulare cannot be separated by cultural or bio- criteria may contribute to the misidentification of frequently chemical tests but instead require hybridization to specific subcultured strains which become adapted to a particular DNA probes or determination of the types of T-catalases medium (e.g., strain M21 as discussed above). Additional 258 THOREL ET AL. INT. J. SYST.BACTERIOL.

1 3418 # 2 1277 ## 3 12998 5 ### 4 ATCC19698T #### 5 1528 ##### 6 CDLYON ###### 7 12998 12 &&&#### 8 12998 2 &&& ###& # g 147 89 8&&###&&# 10 1466 &&&###&& ## cluster 1 11 1077 //a&&//### M.paratu berculosis 12 1394 XXX///&&&&/# 13 7912 XXX///3€3€3€X/ g 14 8642 ///XXX/=== = # 15 12998 9 x3€3€===3€3€ x ## 16 12998 4 --- -- = &## 17 1464 X3€3€///3€X3€X/ WX=# 18 CD2569 ===xxx=;/====#/#/# 19 12547 # 20St 18 ## 21 2103 &## cluster 2 22 ATCC25291" - w23 3 23 M21 #ax# 24 ATCCl3950' =xx # 25 V1 72 # 26 2828 x# 27 10059 X& P 28 853 a## 29 4776 3a### 30 5329 X&#### cluster 3 31 3135 =/##### wood pigeon mycobacteria 32 6409 =I # # # # # # 33 6861 =/####### 34 495 %&&&#### 35 4043 XI///&&&/# 36 7362 =&&&&###&/# 37 3585 ==////&&&//## 38 316F # cluster 4 FIG. 1. Numerical taxonomy matrix of 38 strains of slow-growing mycobacteria. The following symbols are used to indicate internal matching scores: =, 75 to 80%; X , 80 to 85%; I,85 to 90%; &, 90 to 95%; #, 95 to 100%. biochemical tests might greatly improve the set of identifi- wild animals, mammals (especially deer), and birds (espe- cation tests used for such strains. cially wood pigeons and cranes). In order to take into account the results of genomic and Emended description of M. avium Chester 1901. Cells are phenotypic studies, we propose a change in nomenclature. short to long rods with no cord formation. Colonies may be M. avium, M. paratuberculosis, and wood pigeon mycobac- smooth or rough. The colony type may change after first teria have been shown to belong to a single genomic species. isolation from pathological material through subsequent sub- M. avium was described first, and thus the species name cultures. Most strains are nonphotochromogenic and may should be M. avium. The following designations are pro- become yellow on aging. However, some strains are sco- posed: M. aviurn subsp. avium and M. avium subsp. paratu- tochromogenic and have bright yellow pigment. Growth berculosis (as previously mentioned by several authors [18, occurs after more than 7 days at 37"C, which is the optimal 391). The name M. avium subsp. silvaticum is proposed to temperature , and may require mycobactin and/or special replace the trivial designation wood pigeon mycobacteria media other than egg-based media. Growth may be stimu- and the "M. avium subsp. name columbae," which was lated by pH 5.5 or pyruvate. Strains do not produce niacin, suggested by Saxegaard and Baess in their recent study on peroxidase, nitrate reductase, urease, arylsulfatase, penicil- relationships among M. avium, M. paratuberculosis, and linase, or P-glucosidase and do not hydrolyze Tween 80 in 10 wood pigeon mycobacteria (39). The subspecific epithet of days. Strains produce small amounts of thermoresistant M. avium subsp. silvaticum reflects the various sources of catalase. All strains have the same mycolic acid pattern, these peculiar mycobacteria, which have been isolated from namely, mycolate I (a-mycolate), mycolate IV (keto-myco- VOL.40, 1990 MY COBACTIN-DEPENDENT MYCOBACTERIA 259 late), and mycolate VI (dicarboxymycolate) (12, 26). All emerging ? Acta Leprol. 7:Sl&S17. strains are resistant to thiophene-2-carboxylic acid hy- 8. Chiodini, R. J., H. J. Van Kruiningen, W. R. Thayer, and J. A. drazide and . Strains may be susceptible to p- Coutu. 1986. Spheroplastic phase of mycobacteria isolated from nitrobenzoate, ethambutol, pyrazinamide, rifampin, and patients with Crohn’s disease. J. Clin. Microbiol. 24:357-363. streptomycin. Most strains are susceptible to ansamycin. 9. Collins, D. M., and G. W. De Lisle. 1986. Restriction endonu- clease analysis of various strains of Mycobacterium paratuber- The validity of the species has been demonstrated by the culosis isolated from cattle. Am. J. Vet. Res. 47:222&2229. results of numerous DNA-DNA hybridization studies and by 10. Collins, D. M., D. M. Gabric, and G. W. DeLisle. 1989. determinations of the thermal stabilities of the hybrids (1, 2, Identification of a repetitive sequence specific to Mycobacte- 23, 31, 52, 53). The type strain is strain ATCC 25291. The rium paratuberculosis. FEMS Microbiol. Lett. 60:175-178. guanine-plus-cytosine content is 70 mol% (3). The species 11. Collins, P., A. McDiarmid, L. H. Thomas, and P. R. J. Mat- includes both potential and obligate parasites and is divided thews. 1985. Comparison of the pathogenicity of Mycobacte- into subspecies on the basis of pathogenicity and host range. rium paratuberculosis and Mycobacterium spp. isolated from Description of M. avium subsp. avium. M. avium subsp. the wood pigeon (Columba palumbus L.). J. Comp. Pathol. avium is frequently encountered in the environment and is 95591-597. 12. DaffC, M., M. A. Laneelle, C. Asselineau, V. Levy-FrCbault, and the agent of tuberculosis in birds. It may be involved in H. David. 1983. IntCret taxonomique des acides gras des myco- infections of a very large set of animals (cattle, sheep, goats, bactkries: proposition d’une mkthode d’analyse. Ann. Micro- pigs, cats, kangaroos) (20, 41). The main infections in biol. (Paris) 134B:241-256. humans are pulmonary infections in adults, submandibular 13. Damato, J. J., C. Knisley, and M. T. Collins. 1987. Characteri- adenopathies in children, and disseminated infections in zation of Mycobacterium paratuberculosis by gas-liquid and patients suffering from acquired immunodeficiency syn- thin-layer chromatography and rapid demonstration of myco- drome (36, 48, 50). The type strain is strain ATCC 25291. bactin dependence using radiometric methods. J. Clin. Micro- Description of M. avium subsp. paratuberculosis. M. avium biol. 252380-2383. 14. David, H. L. 1977. Alkaline phosphatases from Mycobacteriurn subsp. paratuberculosis strains possess the properties of M. smegmatis. J. Gen. Microbiol. 101:99-102. avium described above and the additional features indicated 15. David, H. L., M. T. Jahan, A. Jumin, J. Grandry, and E. H. in Table 2. A mycobactin requirement for growth is charac- Lehman. 1978. Numerical taxonomy analysis of Mycobacterium teristic of the subspecies. Strains have not been isolated africanum. Int. J. Syst. Bacteriol. 28:467472. from the environment, are obligate for ruminants, 16. David, H. L., V. Levy-FrCbault, and M. F. Thorel. 1989. and cause paratuberculosis, a chronic enteric disease (35). Mkthodes de laboratoire pour Mycobactkriologie clinique. In- This organism may be involved in Crohn’s disease in humans stitut Pasteur, Paris. (7). The type strain is strain ATCC 19698. 17. Doyle, T. M. 1964-1965. Strains of Mycobacteriumjohnei used Description of M. avium subsp. silvaticum. M. avium subsp. for the preparation of vaccine. State Vet. J. 19-20:154155. M. 18. Grange, J. M. 1983. The mycobacteria, p. 60-93. In M. Parker silvaticum strains possess the properties of avium de- (ed.), Topley and Wilson’s principles of bacteriology, virology scribed above and the additional features indicated in Table and immunity, 7th ed., vol. 2. Edward Arnold, London. 2. Inability to grow on egg medium and stimulation of growth 19. Green, E. P., M. T. Moss, J. Hermon-Taylor, and J. J. McFad- at pH 5.5 are characteristic of the subspecies. Strains have den. 1989. Insertion elements in mycobacteria. Acta Leprol. not been isolated from the environment, are obligate patho- 7: S239-S242. gens for animals, and cause tuberculosis in birds and paratu- 20. Griffith, F. 1911. Final report of the Royal Commission on berculosis in mammals (11, 28, 29, 43). The type strain is Tuberculosis, part 2. Darling and Son, London. strain 6409 (= CIP 103317 [Collection Nationale de Cultures 21. Hampson, S. J., F. Portaels, J. Thompson, E. P. Green, M. T. de Microorganismes, Paris, France]). Moss, J. Hermon-Taylor, and J. J. McFadden. 1989. DNA probes demonstrate a single highly conserved strain of Myco- bacterium avium infecting AIDS patients. Lancet ii:65-68. ACKNOWLEDGMENTS 22. Hance, A. J., B. Grandchamp, V. LCvy-Frdbault, D. Lecossier, J. We gratefully thank H. L. David for continuous help and support Rauzier, D. Bocart, and B. Gicquel. 1989. Detection and identi- during ths investigation and P. A. D. Grimont for helpful advice. fication of mycobacteria by amplification of mycobacterial DNA. Mol. Microbiol. 3:843-849. LITERATURE CITED 23. Hurley, S. S., G. A. Splitter, and R. A. Welch. 1988. Deoxyri- 1. Baess, I. 1979. Deoxyribonucleic acid relatedness among spe- bonucleic acid relatedness of Mycobacterium paratuberculosis cies of slowly-growing mycobacteria. Acta Pathol. Microbiol. to other members of the family . Int. J. Syst. Scand. Sect. B 87:221-226. Bacteriol. 38:143-146. 2. Baess, I. 1983. Deoxyribonucleic acid relationships between 24. Jorgensen, J. B. 1984. The effect of vaccination on the excretion different serovars of Mycobacterium avium, Mycobacterium of Mycobacterium paratuberculosis, p. 131-136. In J. B. Jor- intracellulare, and Mycobacterium scrofulaceum. Acta Pathol. gensen and 0. Aalund (ed.), Paratuberculosis. Diagnostic meth- Microbiol. Scand. Sect. B 91:201-203. ods, their practical application and experience with vaccination. 3. Baess, I., and B. Mansa. 1978. Determination of genome size Commission of European Communities, Brussels. and base ratio on deoxyribonucleic acid from mycobacteria. 25. Jorgensen, J. B., and B. Clausen. 1976. Mycobacteriosis in a Acta Pathol. Microbiol. Scand. Sect. B 86:309-312. roe-deer caused by wood pigeon mycobacteria. Nord. Veteri- 4. Barclay, R., D, F. Ewing, and C. Ratledge. 1985. Isolation, naermed . 28539-546. identification, and structural analysis of the mycobactins of 26. LCvy-Frebault, V., K. S. Goh, and H. L. David. 1986. Mycolic Mycobacterium avium, Mycobacterium intracellulare, Myco- acid analysis for clinical identification of Mycobacterium avium bacterium scrofulaceum, and Mycobacterium paratuberculosis. and related mycobacteria. J . Clin. Microbiol. 24:835-839. J. Bacteriol. 164:89&903. 27. LCvy-Frebault, V. V., M.-F. Thorel, A. Varnerot, and B. Gic- 5. Camphausen, R. T., R. L. Jones, and P. J. Brennan. 1988. quel. 1989. DNA polymorphism in Mycobacterium paratuber- Antigenic relationship between Mycobacterium paratuberculo- culosis, “wood pigeon mycobacteria,” and related mycobac- sis and Mycobacterium avium. Am. J. Vet. Res. 49:1307-1310. teria analyzed by field inversion gel electrophoresis. J. Clin. 6. Carle, G. F., M. Frank, and M. V. Olson. 1986. Electrophoretic Microbiol. 27:2823-2826. separations of large DNA molecules by periodic inversion of the 28. Matthews, P. R. J., and A. McDiarmid. 1979. The production in electric field. Science 232:65-68. bovine calves of a disease resembling paratuberculosis with a 7. Chiodini, R. J. 1989. Mycobacterium paratuberculosis: an Mycobacterium sp. isolated from a woodpigeon (Columba pa- 260 THOREL ET AL. INT.J. SYST.BACTERIOL.

lumbus L.). Vet. Rec. 104:286. 42. Thorel, M. F., and P. Desmettre. 1982. Etude comparative de 29. McDiarmid, A. 1948. The occurrence of tuberculosis in the wild souches de mycobacteries mycobactine-dkpendantes, isolkes de wood pigeon. J. Comp. Pathol. 58:128-133. pigeon rarnier, avec Mycobacterium avium et M. paratubercu- 30. McFadden, J. J., P. D. Butcher, R. Chiodini, and J. Hermon- losis: etude des caractkres biologiques et antigdniques. Ann. Taylor. 1987. Crohn’s disease-isolated mycobacteria are identi- Microbiol. (Paris) 133B:291-302. cal to Mycobacterium paratuberculosis, as determined by DNA 43. Thorel, M. F., P. Pardon, K. Irgens, J. Marly, and P. Lechopier. probes that distinguish between mycobacterial species. J. Clin. 1984. Paratuberculose experimentale: pouvoir pathogene chez Microbiol. 25796-801. le veau de souches de mycobacteries mycobactine-dependa- 31. McFadden, J. J., P. D. Butcher, R. J. Chiodini, and J. Hermon- ntes. Ann. Rech. Vet. 15365-374. Taylor. 1987. Determination of genome size and DNA homology 44. Walczak, C. A., and M. I, Krichevsky. 1980. Computer methods between an unclassified Mycobacterium species isolated from for describing groups from binary phenetic data: preliminary patients with Crohn’s disease and other mycobacteria. J. Gen. summary and editing of data. Int. J. Syst. Bacteriol. 30:615-621. Microbiol. 133:211-214. 45. Wards, B. J., D. M. Collins, and G. W. De Lisle. 1987. 32. McFadden, J. J., J. Thompson, E. Hull, S. Hampson, J. Stan- Restriction endonuclease analysis of members of the Mycobac- ford, and J. Hermon-Taylor. 1988. The use of cloned DNA terium avium-M. intracellulare-M. scrofulaceum serocomplex. probes to examine organisms isolated from Crohn’s disease J. Clin. Microbiol. 252309-23 13. tissue, p. 515-520. In R. P. MacDermott (ed.), Inflammatory 46. Wayne, L. G., and G. A. Diaz. 1986. Differentiation between bowel disease: current status and future approach. Elsevier, T-catalases derived from Mycobacterium avium and M. intra- Amsterdam. cellulare by a solid-phase immunosorbent assay. Int. J. Syst. 33. McIntyre, G., and J. L. Stanford. 1986. ImmunodifFusion anal- Bacteriol. 36:363-367. ysis shows that Mycobacterium paratuberculosis and other 47. Wayne, L. G., R. C. Good, M. I. Krichevsky, 2. Blacklock, H. L. mycobactin-dependent mycobacteria are variants of Mycobac- David, D. Dawson, W. Gross, J. Hawkins, P. A. Jenkins, I. terium avium. J. Appl. Bacteriol. 61:295-298. Juhlin, W. Kappler, H. H. Kleeberg, V. LCvy-Frebault, C. McDurmont, E. Nel, F. Portaels, S. Rusch-Gerdes, K. H. 34. Merkal, R. S. 1979. Proposal of ATCC 19698 as the neotype E. strain of Mycobacterium paratuberculosis Bergey et al. 1923. Schroder, V. A. Silcox, I. Szabo, M. Tsukamura, L. Van den Int. J. Syst. Bacteriol. 29:263-264. Breen, B. Vergmann, and M. A. Yakrus. 1989. Third report of the cooperative, open-ended study of slowly growing mycobac- 35. Merkal, R. S. 1984. Paratuberculosis, p. 1237-1249. In G. P. teria by the International Working Group on Mycobacterial Kubica and L. Wayne (ed.), The mycobacteria. source- G. A Taxonomy. Int. J. Syst. Bacteriol. 39:267-278. book. Marcel Dekker, New York. 48. Wendt, S. L., K. L. George, B. C. Parker, H. Gruft, and J. 0. 36. Portaels, F. 1987. Le SIDA et les mycobactCries atypiques. Falkinham III. 1980. Epidemiology of by nontubercu- Ann. SOC.Belge Med. Trop. 67:93-116. lous mycobacteria. 111. Isolation of potentially pathogenic my- 37. Rogosa, M., M. I. Krichevsky, andR. R. Colwell. 1971. Method cobacteria from aerosols. Am. Rev. Respir. Dis. 122:259-263. for coding data on microbial strains for computers, edition AB. 49. Whipple, D. L., R. B. LeFebvre, R. E. Andrews, and A. B. Int. J. Syst. Bacteriol. 21:1A-l75A. Thiermann. 1987. Isolation and analysis of restriction endonu- 38. Saito, H., H. Tomioka, K. Sato, H. Tasaka, M. Tsukamura, F. clease digestive patterns of chromosomal DNA from Mycobac- Kuze, and K. Asano. 1989. Identification and partial character- terium paratuberculosis and other Mycobacterium species. J. ization of Mycobacterium avium and Mycobacterium intracel- Clin. Microbiol. 251511-1515. Mare by using DNA probes. J. Clin. Microbiol. 27:994-997. 50. Wolinsky, E. 1979. Nontuberculous mycobacteria and associ- 39. Saxegaard, F., and I. Baess. 1988. Relationship between Myco- ated diseases. Am. Rev. Respir. Dis. 119:107-159. bacterium avium, Mycobacterium paratuberculosis and “wood 51. Wolinsky, E., and W. B. Schaefer. 1973. Proposed numbering pigeon mycobacteria.” Acta Pathol. Microbiol. Scand. 96: scheme for mycobacterial serotypes by agglutination. Int. J. 37-42. Syst. Bacteriol. 23:182-183. 40. Saxegaard, F., I. Baess, and E. Jantzen. 1988. Characterization 52. Yoshimura, H. H., and D. Y. Graham. 1988. Nucleic acid of clinical isolates of Mycobacterium paratuberculosis by DNA- hybridization studies of mycobactin-dependent mycobacteria. DNA hybridization and cellular fatty acid analysis. Acta Pathol. J. Clin. Microbiol. 26:1309-1312. Microbiol. Scand. 96:497-502. 53. Yoshimura, H. H., D. Y. Graham, M. K. Estes, and R. S. 41. Thorel, M. F. 1980. Note sur 1’Ctude comparative des skrotypes Merkal. 1987. Investigation of association of mycobacteria with des souches de Mycobacterium avium isol6es de l’homme et de inflammatory bowel disease by nucleic acid hybridization. J. l’animal. Ann. Microbiol. (Paris) 131A:71-76. Clin. Microbiol. 2545-51.

Top View