INTERNATIONAL JOURNALOF SYSTEMATIC BACTERIOLOGY, July 1990, p. 254-260 0020-7713/90/030254-07$02.oo/o Copyright 0 1990, International Union of Microbiological Societies
Vol. 40, No. 3
Numerical Taxonomy of Mycobactin-Dependent Mycobacteria, Emended Description of Mycobacterium avium, and Description of Mycobacterium avium subsp. avium subsp. nov. , Mycobacterium avium subsp. paratuberculosis 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'rinaireset 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 mycobacterial 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 species, 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 electrophoresisas 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 paratuberculosis have been shown to belong to a single genomic species by DNA-DNA hybridization studies (23, 40,52) and determinations of the thermal stabilities of hybrids (52). The close relationship between M . avium and M. paratuberculosis was demonstrated by the results of a comparison of the sequences of a 383-base-pair segment of the gene encoding the 65-kilodalton mycobacterial antigen. The sequences of the entire fragment have been found to be identical for M. avium and M. paratuberculosis, whereas the oligonucleotides within this segment consist of probes that are useful for differentiating M . avium from tubercle bacilli and Mycobacterium fortuitum (22). The results of DNA-DNA hybridization studies have supported the identification of strains isolated from patients suffering from Crohn's disease as M. paratuberculosis but have failed to differentiate these strains from M. avium (31, 53). Strains designated wood pigeon mycobacteria also have been shown to belong to the genomic species containing the M . avium and M . paratuberculosis type strains on the basis of determinations of percentages of DNA-DNA homology in which a spectrophotometric method was used (39) and on the basis of hybridization with radiolabeling in solution (23, 52). Mycobacterium intracellulare is a genetically distinct species, as demonstrated by Baess in her pioneering studies of the DNA
relatedness of serovars of M . avium and M . intracellulare (1, 2); this finding was recently confirmed by workers who used DNA probes specific for either M . avium or M. intracellulare (38). However, other genomic approaches have revealed some differences among M . avium, M . paratuberculosis, and the wood pigeon mycobacteria. The restriction profiles of a large set of strains revealed stringent conservation of the genetic composition of M. paratuberculosis and high levels of heterogeneity within M . avium (9, 45, 49). Restriction patterns may be difficult to interpret because usually they rely on the detection and comparison of more than 50 closely spaced bands. In order to simplify such analyses the following two techniques have been developed: restriction fragment length polymorphism (RFLP) analysis, which results in simplified patterns as determined by hybridization of endonuclease cleavage products with specific DNA probes, and pulsedfield gel electrophoresis (especially field inversion gel electrophoresis [FIGE]), which allows the separation of large DNA fragments and thus an analysis of patterns composed of few bands after chromosomal digestion by endonucleases having low-frequency cleavage sites (6). The two methods have comparable sensitivities (27). 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) are identical to M . paratuberculosis ( 8 , 30) and that M.
* Corresponding author. 254
VOL.40, 1990
MYCOBACTIN-DEPENDENT MYCOBACTERIA
paratuberculosis is distinguishable from the other mycobacteria examined, including different M . avium serotypes (10) and one wood pigeon bacillus (32). The differences between RFLP patterns could be related to the presence of a repetitive DNA sequence present in M . paratuberculosis but absent from all M . avium serotypes and from 18 other mycobacterial species (10, 19). RFLP analysis has enhanced the detection of heterogeneity among the M . avium serotypes, as suggested by previous studies of restriction patterns (21, 45). FIGE patterns obtained with restriction enzyme DraI confirmed the high level of homogeneity of M . paratuberculosis strains, including strains isolated from patients with Crohn’s disease, and the differentiation of this species from M . avium. FIGE analysis also demonstrated the homogeneity of the wood pigeon mycobacteria. All of the strains tested exhibited the same unique pattern, which differed from the patterns of M . avium and M . paratuberculosis (27). RFLP and FIGE analyses have established that M . avium, M . paratuberculosis, and the wood pigeon mycobacteria can be recognized and differentiated despite the fact they belong to a single species. The RFLP or FIGE patterns correlate well with differentiation based on pathogenicity and host range. M . paratuberculosis is the agent of paratuberculosis in ruminants (35) and may be involved in Crohn’s disease in humans (7), whereas wood pigeon mycobacteria may cause paratuberculosis in bovine calves and tuberculosis in birds (11, 28, 29, 43). Both organisms are isolated only from infected tissues, unlike M . avium, which is widely encountered in the environment and is the agent of tuberculosis in birds and various infections in other animals and in humans, especially in disseminated infections of patients suffering from acquired immunodeficiency syndrome (36, 48, 50). Thus, it would be worthwhile to establish phenotypic tests that agree with the genomic classification. The differential characteristics proposed previously, based on biochemical tests, fatty acid composition, peptidoglycolipid content, and immunodiffusion analysis results, did not allow reliable identification of the different taxa (4,5 , 13, 33, 40, 42). The purpose of this study was to evaluate biochemical tests that are easily applied to this set of difficult-to-grow mycobacterial strains for consistent identification of M . paratuberculosis and wood pigeon mycobacteria. We propose a change in nomenclature in order to recognize the taxonomic implications of the numerous recent genomic and phenotypic studies of these closely related mycobacteria. MATERIALS AND METHODS Mycobacterial strains. Table 1lists the 38 strains which we studied. All of the strains isolated from organs or feces were mycobactin dependent, at least when they were first isolated. Isolation of the strains. Initial isolation of all of the strains was performed on Herrold medium, Middlebrook 7H10 medium, or Middlebrook 7 H l l medium devoid of or supplemented with 2 mg of mycobactin per liter (16), and an adequate medium was selected for subsequent subcultures. The M. avium and M . intracellulare type strains were subcultured on Lowenstein-Jensen medium. Identification tests. The tests shown in Table 2 have been described previously (14-16); when necessary, some of these tests were performed on media supplemented with mycobactin. Stimulation by pyruvate (4.1 g/liter) and resistance to NaCl were tested on Middlebrook 7H10 medium; the ability to grow at pH 5.5 was determined on Middlebrook 7Hll
255
medium. Drug susceptibility was studied on Middlebrook 7 H l l medium, and resistance was determined by using the proportion method (16). Mycolic acid determinations were performed by using the method of Daff6 et al. (12). Numerical taxonomy analysis. The numerical taxonomy analysis was based on the simple matching coefficient, and similarities were calculated by using the unweighted average linkage method (44). Tests were designated according to their RKC code numbers (37). RESULTS AND DISCUSSION The 38 strains were subjected to 51 tests, including determinations of cellular and colonial morphology, optimal temperature for growth, growth on different media, production of pigment, tolerance to inhibitors, resistance to antibiotics, enzymatic activities, and mycolic acid content. Tests which gave positive or negative results for all strains were deleted from the numerical analysis. After we deleted these tests (which are reported below in the emended description of M . avium), the numerical analysis was performed by using the 22 remaining characteristics (Table 2) and generated four well-defined clusters (Fig. 1). Cluster 1, containing 18 strains, included the type strain of M . paratuberculosis and the two strains isolated from patients suffering from Crohn’s disease. The other strains in this cluster were isolated from cattle, goats, and sheep; diagnosis of Johne’s disease was confirmed for these animals by clinical evidence. The mean internal matching score for this cluster was 86%. The strains were characterized by an absolute requirement for mycobactin for growth (Table 2). A FIGE analysis of seven strains in cluster 1 (strains 1,4, 6 , 9 , 11, 13, and 18 in the numerical taxonomy matrix [Fig. 11) gave identical patterns, which were characteristic of M . paratuberculosis (27). Cluster 2, which had a mean internal matching score of 88%, consisted of six strains and included the type strains of M . avium and M . intracellulare. This demonstrates that differentiation of these two species cannot be achieved by using characteristics that are currently considered for numerical analysis (47) but requires other methods, such as hybridization to a specific probe (38) or a study of Tcatalases (46). All of the strains in cluster 2 except the M. avium type strain have been found to be NaCl resistant, although Mycobacterium triviale is usually considered to be the only slowly growing mycobacterial species that is able to grow on 5% NaCl (18). However, it should be noted that the test in this study was performed on Middlebrook 7H10 medium. This may explain the discrepancy with the results of previous studies of NaCl resistance, in which Lowenstein-Jensen medium was used. Strain St18 was previously identified as an M . paratuberculosis strain and was selected as a working strain until Merkal reported the unusual cultural properties of this strain, its mycobactin independence, and the absence of pathogenicity for calves (34). Further studies of peptidoglycolipid content, RFLP data, and conventional and pulsedfield electrophoresis data established that this strain is in fact an M. avium serotype 2 strain (5, 10, 27, 49). Strain 2103 was also identified as an M . paratuberculosis strain at first since this strain was mycobactin dependent and was isolated from a goat mesenteric lymph node. However, strain 2103 differed from M. paratuberculosis by the following characteristics: mycobactin independence after subculturing, the ability to grow in the presence of NaC1, suscep-
INT. J. SYST.BACTERIOL.
THOREL ET AL.
256
TABLE 1. Strains used in this study Serial no.
Laboratory designation“
Source
Originb
3418 1277 12998.5 ATCC 19698T 1528 CD Lyon 12998.12 12998.2 147.89 1466 1077 1394 7912 8642 12998.9 12998.4 1464 2569
Bovine ileocecal valve Bovine feces Bovine feces M . paratuberculosis bovine feces Bovine feces Human intestine biopsy Bovine feces Bovine feces Goat mesenteric lymph node Bovine feces Sheep Bovine feces Bovine rectum Goat mesenteric lymph node Bovine feces Bovine feces Bovine feces Human surgical specimen
E.N.V. Alfort, France SaBne et Loire, France Corrkze , France ATCC SaBne et Loire, France RhBne, France Corr2ze, France Corrkze, France Deux Skvres, France SaBne et Loire, France Iowa SaBne et Loire, France Haute SaBne, France SaBne et Loire, France Correze, France Corr&ze,France SaBne et Loire, France Lelystad, The Netherlands
19 20 21 22 23 24
ATCC 25291T M21 ATCC 13950T
Deer mesenteric lymph node Mycobactin preparation reference strain Mesenteric lymph node M. avium hen liver Wood pigeon M. intracellulare
Aube, France Iowa Deux Skvres, France ATCC Compton, United Kingdon ATCC
25 26 27 28 29 30 31 32 33 34 35 36 37
V172 2828 10059 853 4776 5329 3135 6409 6861 495 4043 7362 3585
Deer Wood pigeon liver Wood pigeon liver Wood pigeon liver and Wood pigeon liver Wood pigeon spleen Wood pigeon liver Wood pigeon liver and Wood pigeon liver and Wood pigeon liver and Crane liver Wood pigeon spleen Wood pigeon liver
Copenhagen, Denmark Yvelines, France Gers, France Yvelines, France Gers, France Yvelines, France Gers, France Val d’Oise, France Val d’Oise, France Loiret, France Haut Rhin, France Loiret, France Gers, France
316.F
Antigen and vaccine preparation reference strain
1 2 3 4 5 6
7 8 9 10 11 12 13 14 15 16 17 18
38
spleen
spleen spleen spleen
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 phosphatase. Pulsed-field electrophoresis patterns confirmed the differentiation of this organism from M . paratuberculosis (27). In addition, experimental infections of calves revealed pathogenicity similar to that of M. avium (20) since strain 2103 induced acute infections and killed the animals within 36 to 60 days after inoculation (43). Strain 12547 was isolated from a mesenteric lymph node of a deer. This strain did not exhibit the characteristic DNA pattern of M . paratuberculosis or wood pigeon mycobacteria as determined by pulsed-field electrophoresis (data not shown). Strain M21, which was isolated from a wood pigeon, fell into cluster 2 because of cultural characteristics that were similar to those of M. avium. Growth occurred on egg medium and was not prevented by p-nitrobenzoate or by sodium chloride on Middlebrook 7H10 medium. Growth was not stimulated at pH 5.5. However, strain M21 exhibited a pulsed-field pattern that was characteristic of the wood pigeon mycobacteria (data not shown). Experimental infections demonstrated the ability of this strain to produce
paratuberculosis in cattle and tuberculosis in hens (11, 28, 43). Cluster 3 had a mean internal matching score of 91%; this cluster was composed of 13 strains (12 strains isolated from birds and 1 strain isolated from a mammal). All of the strains isolated from birds produced lesions in livers and spleens similar to those observed in birds infected with M. avium. Concomitant pathogenicity for calves was confirmed by experimental infections with two strains (strains 6861 and 6409) (43). Similarly, strain VI72, which was isolated originally from a wild ruminant suffering from Johne’s disease, has been shown to cause avian tuberculosis in hens and paratuberculosis in calves (25, 43). Strains in this cluster are characterized by the inability to grow on egg medium and stimulation of growth by pH 5.5 (Table 2). Four strains belonging to this cluster were examined by FIGE and produced identical patterns, which were different from the profiles observed for M. paratuberculosis and M . avium strains (27). Cluster 4 consisted of a single strain, M . paratuberculosis 316F. This strain has been extensively subcultured since it
MYCOBACTIN-DEPENDENT MYCOBACTERIA
VOL.40, 1990
257
TABLE 2. Variable and differential characteristics of M. paratuberculosis and related mycobacteria ~
Cluster 1 Characteristic
Rods Coccobacillary Colonies smooth Colonies rough Growth at 42°C Growth stimulation by pyruvate' Growth stimulated in medium at pH 5.5' Mycobactin (2 mglliter) is required for growth' Growth on egg medium' 5% NaCl tolerance on Middlebrook 7H10 medium Enzymatic activities Catalase after heating at 68°C Acid phosphatase (2 h) Alkaline phosphatase (2 h)' Trehalase (5 h) Growth in the presence of para-Nitrobenzoate (500 pglml) D-Cycloserine (30 pglml) D-Cycloserine (50 pg/ml)' Ethambutol (7.5 pglml) Pyrazinamide (50 kg/ml) Rifampin (1 pglml) Ansamycin (1 pg/ml) Streptomycin (2 pg/ml) a
Cluster 3
RKC
Cluster 4 13
-
003008 003026 016030 016031 017035 098193 099031 099038 099030 018006
+ + + + + +
024425 034136 034137 034142
-
016264 098183 099033 099034 099035 040164 099036 019510
5/18' 13/18 0118 18/18 16/18 18/18 0118 18/18 18/18 0118
13/13 0113 0113 13/13 13/13 0113 13/13 0113 2/13 0113
-
18/18 0118 1/18 6/18
13/13 1/13 13/13 3/13
+ + + + + + +
4/13 18/18 17/18 18/18 14/18 13/18 15/18 13/18
0113 11/12 0/13 12/13 13/13 1/13 1/13 7/13
-
+
I1 E t 4 q )
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 Kingdom, laboratory (17) and is used as a live vaccine (17, 24). It is characterized by negative responses to all of the tests that are useful for discriminating among clusters 1, 2, and 3 (Table 2) and differs from other M . paratuberculosis strains by its inability to grow on egg medium, its mycobactin independence, its lack of stimulation of growth by pyruvate, and its susceptibility to cycloserine. Numerical analysis split the 38 strains which we studied into four well-defined clusters (Fig. 1). Cluster 1 corresponds to M . paratuberculosis; cluster 2 contains six strains, including the type strains of M . avium and M . intracellulare; cluster 3 contains the wood pigeon mycobacteria; and cluster 4 contains M . paratuberculosis 316F, which is used for vaccination. This separation based on cultural and biochemical tests is highly correlated with pulsed-field patterns and with pathogenicity, as confirmed by experimental infections in hens and calves (11, 27, 28, 43). The mean internal matching scores range from 85 to 91% for clusters 1 through 3. These percentages must be compared with circumspection with values reported in other numerical analyses which generated clusters corresponding to different mycobacterial species. In this numerical analysis we used a set of highly related strains and relied on selected tests. Our selection resulted in the high clustering levels observed in this study. Of the 38 strains studied, 2 were not correctly identified by the tests which we used. The M . intraceflufaretype strain and wood pigeon mycobacterium strain M21 both fell into cluster 2 along with M . avium strains. This misidentification may be explained by the nature of the tests. M . avium and M . intracellulare cannot be separated by cultural or biochemical tests but instead require hybridization to specific DNA probes or determination of the types of T-catalases
(38, 46). Determinations of the serotypes by detecting peptidoglycolipids (mycoside C) are valuable but have to be interpreted carefully because DNA-DNA hybridization results lead to a classification of the serotypes between M . avium and M . intracellulare that is different from the Schaefer classification (2, 38, 51). The misidentification of wood pigeon mycobacterium strain M21 was also due to the use of primarily cultural tests since this strain had peculiar growth characteristics which were more like those of M . avium than those of the other wood pigeon mycobacteria, especially growth on Lowenstein-Jensen medium, tolerance to inhibitors (such as p nitrobenzoate), and absence of stimulation by pH 5.5. M . paratuberculosis 316F, the only member of cluster 4, has been subcultured extensively since 1940 for antigen and vaccine production. This may explain the phenotypic and genomic differences now observed between this strain and the other M . paratuberculosis strains, as shown by our numerical analysis as well as by the peculiar pulsed-field pattern of this organism (27). M . paratuberculosis and wood pigeon mycobacteria can be easily separated on the basis of their responses to six tests. M . paratuberculosis is characterized by a mycobactin requirement, the ability to grow on egg medium, tolerance to cycloserine (50 pg/ml), growth stimulation by pyruvate but not by pH 5.5, and an absence of alkaline phosphatase production. Wood pigeon mycobacteria have the opposite characteristics in these six tests. This set of tests includes five cultural tests and only one biochemical test. Lack of balance between cultural and biochemical discriminative criteria may contribute to the misidentification of frequently subcultured strains which become adapted to a particular medium (e.g., strain M21 as discussed above). Additional
258
13
3418 1277 12998 5 ATCC19698T 1528 CDLYON 12998 1 2 12998 2 147 8 9 1466 1077 1394 7912
14
8642
1 2
3 4
5
6 7
8 g 10 11 12
INT. J . SYST.BACTERIOL.
THOREL ET AL.
15 1 2 9 9 8 9 16 1 2 9 9 8 4 17 1 4 6 4 18 CD2569 19 1 2 5 4 7 2 0 S t 18 21 2103 22 A T C C 2 5 2 9 1 " 23 M21 24 A T C C l 3 9 5 0 ' 25 V 1 7 2
# ## ### #### ##### ###### &&&#### &&& # # # & #
8&&###&&#
&&&###&& # # //a&& # # #/ /
XXX///&&&&/# XXX///3€3€3€X/ ///XXX/=== x3€3€===3€3€
cluster 1 M.paratu berculosis
g
= #
x ## -= &## X3€3€///3€X3€X/ WX=# ===xxx=;/=== = # / # / #
---
--
# ## &##
w23 3
#ax# =xx #
26 2828 27 1 0 0 5 9 28 8 5 3 29 4 7 7 6 30 5329 31 3135 32 33 34 35 36 37 38
cluster 2
6409 6861 495 4043 7362 3585
316F
#
x# X& P
a## 3a###
X&####
cluster 3
=/##### wood pigeon mycobacteria =I # # # # # # =/####### %&&&#### XI///&&&/# =&&&&###&/# ==////&&&//## #
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 identification tests used for such strains. In order to take into account the results of genomic and phenotypic studies, we propose a change in nomenclature. M . avium, M . paratuberculosis, and wood pigeon mycobacteria have been shown to belong to a single genomic species. M . avium was described first, and thus the species name should be M . avium. The following designations are proposed: M . aviurn subsp. avium and M . avium subsp. paratuberculosis (as previously mentioned by several authors [18, 391). The name M . avium subsp. silvaticum is proposed to replace the trivial designation wood pigeon mycobacteria and the " M . avium subsp. name columbae," which was suggested by Saxegaard and Baess in their recent study on relationships among M . avium, M . paratuberculosis, and wood pigeon mycobacteria (39). The subspecific epithet of M . avium subsp. silvaticum reflects the various sources of these peculiar mycobacteria, which have been isolated from
wild animals, mammals (especially deer), and birds (especially wood pigeons and cranes). Emended description of M. avium Chester 1901. Cells are short to long rods with no cord formation. Colonies may be smooth or rough. The colony type may change after first isolation from pathological material through subsequent subcultures. Most strains are nonphotochromogenic and may become yellow on aging. However, some strains are scotochromogenic and have bright yellow pigment. Growth occurs after more than 7 days at 37"C, which is the optimal temperature , and may require mycobactin and/or special media other than egg-based media. Growth may be stimulated by pH 5.5 or pyruvate. Strains do not produce niacin, peroxidase, nitrate reductase, urease, arylsulfatase, penicillinase, or P-glucosidaseand do not hydrolyze Tween 80 in 10 days. Strains produce small amounts of thermoresistant catalase. All strains have the same mycolic acid pattern, namely, mycolate I (a-mycolate), mycolate IV (keto-myco-
VOL.40, 1990
MY COBACTIN-DEPENDENT MYCOBACTERIA
late), and mycolate VI (dicarboxymycolate) (12, 26). All strains are resistant to thiophene-2-carboxylic acid hydrazide and isoniazid. Strains may be susceptible to p nitrobenzoate, ethambutol, pyrazinamide, rifampin, and streptomycin. Most strains are susceptible to ansamycin. The validity of the species has been demonstrated by the results of numerous DNA-DNA hybridization studies and by determinations of the thermal stabilities of the hybrids (1, 2, 23, 31, 52, 53). The type strain is strain ATCC 25291. The guanine-plus-cytosine content is 70 mol% (3). The species includes both potential and obligate parasites and is divided into subspecies on the basis of pathogenicity and host range. Description of M . avium subsp. avium. M . avium subsp. avium is frequently encountered in the environment and is the agent of tuberculosis in birds. It may be involved in infections of a very large set of animals (cattle, sheep, goats, pigs, cats, kangaroos) (20, 41). The main infections in humans are pulmonary infections in adults, submandibular adenopathies in children, and disseminated infections in patients suffering from acquired immunodeficiency syndrome (36, 48, 50). The type strain is strain ATCC 25291. Description of M . avium subsp. paratuberculosis. M . avium subsp. paratuberculosis strains possess the properties of M . avium described above and the additional features indicated in Table 2. A mycobactin requirement for growth is characteristic of the subspecies. Strains have not been isolated from the environment, are obligate pathogens for ruminants, and cause paratuberculosis, a chronic enteric disease (35). This organism may be involved in Crohn’s disease in humans (7). The type strain is strain ATCC 19698. Description of M . avium subsp. silvaticum. M . avium subsp. silvaticum strains possess the properties of M . avium described above and the additional features indicated in Table 2. Inability to grow on egg medium and stimulation of growth at pH 5.5 are characteristic of the subspecies. Strains have not been isolated from the environment, are obligate pathogens for animals, and cause tuberculosis in birds and paratuberculosis in mammals (11, 28, 29, 43). The type strain is strain 6409 (= CIP 103317 [Collection Nationale de Cultures de Microorganismes, Paris, France]). ACKNOWLEDGMENTS
We gratefully thank H. L. David for continuous help and support during ths investigation and P. A. D. Grimont for helpful advice. LITERATURE CITED 1. Baess, I. 1979. Deoxyribonucleic acid relatedness among species of slowly-growing mycobacteria. Acta Pathol. Microbiol. Scand. Sect. B 87:221-226. 2. Baess, I. 1983. Deoxyribonucleic acid relationships between different serovars of Mycobacterium avium, Mycobacterium intracellulare, and Mycobacterium scrofulaceum. Acta Pathol. Microbiol. Scand. Sect. B 91:201-203. 3. Baess, I., and B. Mansa. 1978. Determination of genome size and base ratio on deoxyribonucleic acid from mycobacteria. Acta Pathol. Microbiol. Scand. Sect. B 86:309-312. 4. Barclay, R., D, F. Ewing, and C. Ratledge. 1985. Isolation, identification, and structural analysis of the mycobactins of Mycobacterium avium, Mycobacterium intracellulare, Mycobacterium scrofulaceum, and Mycobacterium paratuberculosis. J. Bacteriol. 164:89&903. 5 . Camphausen, R. T., R. L. Jones, and P. J. Brennan. 1988. Antigenic relationship between Mycobacterium paratuberculosis and Mycobacterium avium. Am. J. Vet. Res. 49:1307-1310. 6. Carle, G. F., M. Frank, and M. V. Olson. 1986. Electrophoretic separations of large DNA molecules by periodic inversion of the electric field. Science 232:65-68. 7. Chiodini, R. J. 1989. Mycobacterium paratuberculosis: an
259
emerging pathogen? Acta Leprol. 7:Sl&S17. 8. Chiodini, R. J., H. J. Van Kruiningen, W. R. Thayer, and J. A. Coutu. 1986. Spheroplastic phase of mycobacteria isolated from patients with Crohn’s disease. J. Clin. Microbiol. 24:357-363. 9. Collins, D. M., and G. W. De Lisle. 1986. Restriction endonuclease analysis of various strains of Mycobacterium paratuberculosis isolated from cattle. Am. J. Vet. Res. 47:222&2229. 10. Collins, D. M., D. M. Gabric, and G. W. DeLisle. 1989. Identification of a repetitive sequence specific to Mycobacterium paratuberculosis. FEMS Microbiol. Lett. 60:175-178. 11. Collins, P., A. McDiarmid, L. H. Thomas, and P. R. J. Matthews. 1985. Comparison of the pathogenicity of Mycobacterium paratuberculosis and Mycobacterium spp. isolated from the wood pigeon (Columba palumbus L.). J. Comp. Pathol. 95591-597. 12. DaffC, M., M. A. Laneelle, C. Asselineau, V. Levy-FrCbault, and H. David. 1983. IntCret taxonomique des acides gras des mycobactkries: proposition d’une mkthode d’analyse. Ann. Microbiol. (Paris) 134B:241-256. 13. Damato, J. J., C. Knisley, and M. T. Collins. 1987. Characterization of Mycobacterium paratuberculosis by gas-liquid and thin-layer chromatography and rapid demonstration of mycobactin dependence using radiometric methods. J. Clin. Microbiol. 252380-2383. 14. David, H. L. 1977. Alkaline phosphatases from Mycobacteriurn smegmatis. J. Gen. Microbiol. 101:99-102. 15. David, H. L., M. T. Jahan, A. Jumin, J. Grandry, and E. H. Lehman. 1978. Numerical taxonomy analysis of Mycobacterium africanum. Int. J. Syst. Bacteriol. 28:467472. 16. David, H. L., V. Levy-FrCbault, and M. F. Thorel. 1989. Mkthodes de laboratoire pour Mycobactkriologie clinique. Institut Pasteur, Paris. 17. Doyle, T. M. 1964-1965. Strains of Mycobacteriumjohnei used for the preparation of vaccine. State Vet. J. 19-20:154155. 18. Grange, J. M. 1983. The mycobacteria, p. 60-93. I n M. Parker (ed.), Topley and Wilson’s principles of bacteriology, virology and immunity, 7th ed., vol. 2. Edward Arnold, London. 19. Green, E. P., M. T. Moss, J. Hermon-Taylor, and J. J. McFadden. 1989. Insertion elements in mycobacteria. Acta Leprol. 7: S239-S242. 20. Griffith, F. 1911. Final report of the Royal Commission on Tuberculosis, part 2. Darling and Son, London. 21. Hampson, S. J., F. Portaels, J. Thompson, E. P. Green, M. T. Moss, J. Hermon-Taylor, and J. J. McFadden. 1989. DNA probes demonstrate a single highly conserved strain of Mycobacterium avium infecting AIDS patients. Lancet ii:65-68. 22. Hance, A. J., B. Grandchamp, V. LCvy-Frdbault, D. Lecossier, J. Rauzier, D. Bocart, and B. Gicquel. 1989. Detection and identification of mycobacteria by amplification of mycobacterial DNA. Mol. Microbiol. 3:843-849. 23. Hurley, S. S., G. A. Splitter, and R. A. Welch. 1988. Deoxyribonucleic acid relatedness of Mycobacterium paratuberculosis to other members of the family Mycobacteriaceae. Int. J. Syst. Bacteriol. 38:143-146. 24. Jorgensen, J. B. 1984. The effect of vaccination on the excretion of Mycobacterium paratuberculosis, p. 131-136. In J. B. Jorgensen and 0. Aalund (ed.), Paratuberculosis. Diagnostic methods, their practical application and experience with vaccination. Commission of European Communities, Brussels. 25. Jorgensen, J. B., and B. Clausen. 1976. Mycobacteriosis in a roe-deer caused by wood pigeon mycobacteria. Nord. Veterinaermed . 28539-546. 26. LCvy-Frebault, V., K. S. Goh, and H. L. David. 1986. Mycolic acid analysis for clinical identification of Mycobacterium avium and related mycobacteria. J . Clin. Microbiol. 24:835-839. 27. LCvy-Frebault, V. V., M.-F. Thorel, A. Varnerot, and B. Gicquel. 1989. DNA polymorphism in Mycobacterium paratuberculosis, “wood pigeon mycobacteria,” and related mycobacteria analyzed by field inversion gel electrophoresis. J. Clin. Microbiol. 27:2823-2826. 28. Matthews, P. R. J., and A. McDiarmid. 1979. The production in bovine calves of a disease resembling paratuberculosis with a Mycobacterium sp. isolated from a woodpigeon (Columba pa-
260
THOREL ET AL.
lumbus L.). Vet. Rec. 104:286. 29. McDiarmid, A. 1948. The occurrence of tuberculosis in the wild wood pigeon. J. Comp. Pathol. 58:128-133. 30. McFadden, J. J., P. D. Butcher, R. Chiodini, and J. HermonTaylor. 1987. Crohn’s disease-isolated mycobacteria are identical to Mycobacterium paratuberculosis, as determined by DNA probes that distinguish between mycobacterial species. J. Clin. Microbiol. 25796-801. 31. McFadden, J. J., P. D. Butcher, R. J. Chiodini, and J. HermonTaylor. 1987. Determination of genome size and DNA homology between an unclassified Mycobacterium species isolated from patients with Crohn’s disease and other mycobacteria. J. Gen. Microbiol. 133:211-214. 32. McFadden, J. J., J. Thompson, E. Hull, S. Hampson, J. Stanford, and J. Hermon-Taylor. 1988. The use of cloned DNA probes to examine organisms isolated from Crohn’s disease tissue, p. 515-520. In R. P. MacDermott (ed.), Inflammatory bowel disease: current status and future approach. Elsevier, Amsterdam. 33. McIntyre, G., and J. L. Stanford. 1986. ImmunodifFusion analysis shows that Mycobacterium paratuberculosis and other mycobactin-dependent mycobacteria are variants of Mycobacterium avium. J. Appl. Bacteriol. 61:295-298. 34. Merkal, R. S. 1979. Proposal of ATCC 19698 as the neotype strain of Mycobacterium paratuberculosis Bergey et al. 1923. Int. J. Syst. Bacteriol. 29:263-264. 35. Merkal, R. S. 1984. Paratuberculosis, p. 1237-1249. I n G. P. Kubica and L. G. Wayne (ed.), The mycobacteria. A sourcebook. Marcel Dekker, New York. 36. Portaels, F. 1987. Le SIDA et les mycobactCries atypiques. Ann. SOC.Belge Med. Trop. 67:93-116. 37. Rogosa, M., M. I. Krichevsky, andR. R. Colwell. 1971. Method for coding data on microbial strains for computers, edition AB. Int. J. Syst. Bacteriol. 21:1A-l75A. 38. Saito, H., H. Tomioka, K. Sato, H. Tasaka, M. Tsukamura, F. Kuze, and K. Asano. 1989. Identification and partial characterization of Mycobacterium avium and Mycobacterium intracelMare by using DNA probes. J. Clin. Microbiol. 27:994-997. 39. Saxegaard, F., and I. Baess. 1988. Relationship between Mycobacterium avium, Mycobacterium paratuberculosis and “wood pigeon mycobacteria.” Acta Pathol. Microbiol. Scand. 96: 37-42. 40. Saxegaard, F., I. Baess, and E. Jantzen. 1988. Characterization of clinical isolates of Mycobacterium paratuberculosis by DNADNA hybridization and cellular fatty acid analysis. Acta Pathol. Microbiol. Scand. 96:497-502. 41. Thorel, M. F. 1980. Note sur 1’Ctude comparative des skrotypes des souches de Mycobacterium avium isol6es de l’homme et de l’animal. Ann. Microbiol. (Paris) 131A:71-76.
INT.J. SYST.BACTERIOL. 42. Thorel, M. F., and P. Desmettre. 1982. Etude comparative de souches de mycobacteries mycobactine-dkpendantes, isolkes de pigeon rarnier, avec Mycobacterium avium et M . paratuberculosis: etude des caractkres biologiques et antigdniques. Ann. Microbiol. (Paris) 133B:291-302. 43. Thorel, M. F., P. Pardon, K. Irgens, J. Marly, and P. Lechopier. 1984. Paratuberculose experimentale: pouvoir pathogene chez le veau de souches de mycobacteries mycobactine-dependantes. Ann. Rech. Vet. 15365-374. 44. Walczak, C. A., and M. I, Krichevsky. 1980. Computer methods for describing groups from binary phenetic data: preliminary summary and editing of data. Int. J. Syst. Bacteriol. 30:615-621. 45. Wards, B. J., D. M. Collins, and G. W. De Lisle. 1987. Restriction endonuclease analysis of members of the Mycobacterium avium-M. intracellulare-M. scrofulaceum serocomplex. J. Clin. Microbiol. 252309-23 13. 46. Wayne, L. G., and G. A. Diaz. 1986. Differentiation between T-catalases derived from Mycobacterium avium and M . intracellulare by a solid-phase immunosorbent assay. Int. J. Syst. Bacteriol. 36:363-367. 47. Wayne, L. G., R. C. Good, M. I. Krichevsky, 2.Blacklock, H. L. David, D. Dawson, W. Gross, J. Hawkins, P. A. Jenkins, I. Juhlin, W. Kappler, H. H. Kleeberg, V. LCvy-Frebault, C. McDurmont, E. E. Nel, F. Portaels, S. Rusch-Gerdes, K. H. Schroder, V. A. Silcox, I. Szabo, M. Tsukamura, L. Van den Breen, B. Vergmann, and M. A. Yakrus. 1989. Third report of the cooperative, open-ended study of slowly growing mycobacteria by the International Working Group on Mycobacterial Taxonomy. Int. J. Syst. Bacteriol. 39:267-278. 48. Wendt, S. L., K. L. George, B. C. Parker, H. Gruft, and J. 0. Falkinham III. 1980. Epidemiology of infection by nontuberculous mycobacteria. 111. Isolation of potentially pathogenic mycobacteria from aerosols. Am. Rev. Respir. Dis. 122:259-263. 49. Whipple, D. L., R. B. LeFebvre, R. E. Andrews, and A. B. Thiermann. 1987. Isolation and analysis of restriction endonuclease digestive patterns of chromosomal DNA from Mycobacterium paratuberculosis and other Mycobacterium species. J. Clin. Microbiol. 251511-1515. 50. Wolinsky, E. 1979. Nontuberculous mycobacteria and associated diseases. Am. Rev. Respir. Dis. 119:107-159. 51. Wolinsky, E., and W. B. Schaefer. 1973. Proposed numbering scheme for mycobacterial serotypes by agglutination. Int. J. Syst. Bacteriol. 23:182-183. 52. Yoshimura, H. H., and D. Y. Graham. 1988. Nucleic acid hybridization studies of mycobactin-dependent mycobacteria. J. Clin. Microbiol. 26:1309-1312. 53. Yoshimura, H. H., D. Y. Graham, M. K. Estes, and R. S. Merkal. 1987. Investigation of association of mycobacteria with inflammatory bowel disease by nucleic acid hybridization. J. Clin. Microbiol. 2545-51.
Vol. 40, No. 3
Numerical Taxonomy of Mycobactin-Dependent Mycobacteria, Emended Description of Mycobacterium avium, and Description of Mycobacterium avium subsp. avium subsp. nov. , Mycobacterium avium subsp. paratuberculosis 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'rinaireset 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 mycobacterial 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 species, 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 electrophoresisas 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 paratuberculosis have been shown to belong to a single genomic species by DNA-DNA hybridization studies (23, 40,52) and determinations of the thermal stabilities of hybrids (52). The close relationship between M . avium and M. paratuberculosis was demonstrated by the results of a comparison of the sequences of a 383-base-pair segment of the gene encoding the 65-kilodalton mycobacterial antigen. The sequences of the entire fragment have been found to be identical for M. avium and M. paratuberculosis, whereas the oligonucleotides within this segment consist of probes that are useful for differentiating M . avium from tubercle bacilli and Mycobacterium fortuitum (22). The results of DNA-DNA hybridization studies have supported the identification of strains isolated from patients suffering from Crohn's disease as M. paratuberculosis but have failed to differentiate these strains from M. avium (31, 53). Strains designated wood pigeon mycobacteria also have been shown to belong to the genomic species containing the M . avium and M . paratuberculosis type strains on the basis of determinations of percentages of DNA-DNA homology in which a spectrophotometric method was used (39) and on the basis of hybridization with radiolabeling in solution (23, 52). Mycobacterium intracellulare is a genetically distinct species, as demonstrated by Baess in her pioneering studies of the DNA
relatedness of serovars of M . avium and M . intracellulare (1, 2); this finding was recently confirmed by workers who used DNA probes specific for either M . avium or M. intracellulare (38). However, other genomic approaches have revealed some differences among M . avium, M . paratuberculosis, and the wood pigeon mycobacteria. The restriction profiles of a large set of strains revealed stringent conservation of the genetic composition of M. paratuberculosis and high levels of heterogeneity within M . avium (9, 45, 49). Restriction patterns may be difficult to interpret because usually they rely on the detection and comparison of more than 50 closely spaced bands. In order to simplify such analyses the following two techniques have been developed: restriction fragment length polymorphism (RFLP) analysis, which results in simplified patterns as determined by hybridization of endonuclease cleavage products with specific DNA probes, and pulsedfield gel electrophoresis (especially field inversion gel electrophoresis [FIGE]), which allows the separation of large DNA fragments and thus an analysis of patterns composed of few bands after chromosomal digestion by endonucleases having low-frequency cleavage sites (6). The two methods have comparable sensitivities (27). 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) are identical to M . paratuberculosis ( 8 , 30) and that M.
* Corresponding author. 254
VOL.40, 1990
MYCOBACTIN-DEPENDENT MYCOBACTERIA
paratuberculosis is distinguishable from the other mycobacteria examined, including different M . avium serotypes (10) and one wood pigeon bacillus (32). The differences between RFLP patterns could be related to the presence of a repetitive DNA sequence present in M . paratuberculosis but absent from all M . avium serotypes and from 18 other mycobacterial species (10, 19). RFLP analysis has enhanced the detection of heterogeneity among the M . avium serotypes, as suggested by previous studies of restriction patterns (21, 45). FIGE patterns obtained with restriction enzyme DraI confirmed the high level of homogeneity of M . paratuberculosis strains, including strains isolated from patients with Crohn’s disease, and the differentiation of this species from M . avium. FIGE analysis also demonstrated the homogeneity of the wood pigeon mycobacteria. All of the strains tested exhibited the same unique pattern, which differed from the patterns of M . avium and M . paratuberculosis (27). RFLP and FIGE analyses have established that M . avium, M . paratuberculosis, and the wood pigeon mycobacteria can be recognized and differentiated despite the fact they belong to a single species. The RFLP or FIGE patterns correlate well with differentiation based on pathogenicity and host range. M . paratuberculosis is the agent of paratuberculosis in ruminants (35) and may be involved in Crohn’s disease in humans (7), whereas wood pigeon mycobacteria may cause paratuberculosis in bovine calves and tuberculosis in birds (11, 28, 29, 43). Both organisms are isolated only from infected tissues, unlike M . avium, which is widely encountered in the environment and is the agent of tuberculosis in birds and various infections in other animals and in humans, especially in disseminated infections of patients suffering from acquired immunodeficiency syndrome (36, 48, 50). Thus, it would be worthwhile to establish phenotypic tests that agree with the genomic classification. The differential characteristics proposed previously, based on biochemical tests, fatty acid composition, peptidoglycolipid content, and immunodiffusion analysis results, did not allow reliable identification of the different taxa (4,5 , 13, 33, 40, 42). The purpose of this study was to evaluate biochemical tests that are easily applied to this set of difficult-to-grow mycobacterial strains for consistent identification of M . paratuberculosis and wood pigeon mycobacteria. We propose a change in nomenclature in order to recognize the taxonomic implications of the numerous recent genomic and phenotypic studies of these closely related mycobacteria. MATERIALS AND METHODS Mycobacterial strains. Table 1lists the 38 strains which we studied. All of the strains isolated from organs or feces were mycobactin dependent, at least when they were first isolated. Isolation of the strains. Initial isolation of all of the strains was performed on Herrold medium, Middlebrook 7H10 medium, or Middlebrook 7 H l l medium devoid of or supplemented with 2 mg of mycobactin per liter (16), and an adequate medium was selected for subsequent subcultures. The M. avium and M . intracellulare type strains were subcultured on Lowenstein-Jensen medium. Identification tests. The tests shown in Table 2 have been described previously (14-16); when necessary, some of these tests were performed on media supplemented with mycobactin. Stimulation by pyruvate (4.1 g/liter) and resistance to NaCl were tested on Middlebrook 7H10 medium; the ability to grow at pH 5.5 was determined on Middlebrook 7Hll
255
medium. Drug susceptibility was studied on Middlebrook 7 H l l medium, and resistance was determined by using the proportion method (16). Mycolic acid determinations were performed by using the method of Daff6 et al. (12). Numerical taxonomy analysis. The numerical taxonomy analysis was based on the simple matching coefficient, and similarities were calculated by using the unweighted average linkage method (44). Tests were designated according to their RKC code numbers (37). RESULTS AND DISCUSSION The 38 strains were subjected to 51 tests, including determinations of cellular and colonial morphology, optimal temperature for growth, growth on different media, production of pigment, tolerance to inhibitors, resistance to antibiotics, enzymatic activities, and mycolic acid content. Tests which gave positive or negative results for all strains were deleted from the numerical analysis. After we deleted these tests (which are reported below in the emended description of M . avium), the numerical analysis was performed by using the 22 remaining characteristics (Table 2) and generated four well-defined clusters (Fig. 1). Cluster 1, containing 18 strains, included the type strain of M . paratuberculosis and the two strains isolated from patients suffering from Crohn’s disease. The other strains in this cluster were isolated from cattle, goats, and sheep; diagnosis of Johne’s disease was confirmed for these animals by clinical evidence. The mean internal matching score for this cluster was 86%. The strains were characterized by an absolute requirement for mycobactin for growth (Table 2). A FIGE analysis of seven strains in cluster 1 (strains 1,4, 6 , 9 , 11, 13, and 18 in the numerical taxonomy matrix [Fig. 11) gave identical patterns, which were characteristic of M . paratuberculosis (27). Cluster 2, which had a mean internal matching score of 88%, consisted of six strains and included the type strains of M . avium and M . intracellulare. This demonstrates that differentiation of these two species cannot be achieved by using characteristics that are currently considered for numerical analysis (47) but requires other methods, such as hybridization to a specific probe (38) or a study of Tcatalases (46). All of the strains in cluster 2 except the M. avium type strain have been found to be NaCl resistant, although Mycobacterium triviale is usually considered to be the only slowly growing mycobacterial species that is able to grow on 5% NaCl (18). However, it should be noted that the test in this study was performed on Middlebrook 7H10 medium. This may explain the discrepancy with the results of previous studies of NaCl resistance, in which Lowenstein-Jensen medium was used. Strain St18 was previously identified as an M . paratuberculosis strain and was selected as a working strain until Merkal reported the unusual cultural properties of this strain, its mycobactin independence, and the absence of pathogenicity for calves (34). Further studies of peptidoglycolipid content, RFLP data, and conventional and pulsedfield electrophoresis data established that this strain is in fact an M. avium serotype 2 strain (5, 10, 27, 49). Strain 2103 was also identified as an M . paratuberculosis strain at first since this strain was mycobactin dependent and was isolated from a goat mesenteric lymph node. However, strain 2103 differed from M. paratuberculosis by the following characteristics: mycobactin independence after subculturing, the ability to grow in the presence of NaC1, suscep-
INT. J. SYST.BACTERIOL.
THOREL ET AL.
256
TABLE 1. Strains used in this study Serial no.
Laboratory designation“
Source
Originb
3418 1277 12998.5 ATCC 19698T 1528 CD Lyon 12998.12 12998.2 147.89 1466 1077 1394 7912 8642 12998.9 12998.4 1464 2569
Bovine ileocecal valve Bovine feces Bovine feces M . paratuberculosis bovine feces Bovine feces Human intestine biopsy Bovine feces Bovine feces Goat mesenteric lymph node Bovine feces Sheep Bovine feces Bovine rectum Goat mesenteric lymph node Bovine feces Bovine feces Bovine feces Human surgical specimen
E.N.V. Alfort, France SaBne et Loire, France Corrkze , France ATCC SaBne et Loire, France RhBne, France Corr2ze, France Corrkze, France Deux Skvres, France SaBne et Loire, France Iowa SaBne et Loire, France Haute SaBne, France SaBne et Loire, France Correze, France Corr&ze,France SaBne et Loire, France Lelystad, The Netherlands
19 20 21 22 23 24
ATCC 25291T M21 ATCC 13950T
Deer mesenteric lymph node Mycobactin preparation reference strain Mesenteric lymph node M. avium hen liver Wood pigeon M. intracellulare
Aube, France Iowa Deux Skvres, France ATCC Compton, United Kingdon ATCC
25 26 27 28 29 30 31 32 33 34 35 36 37
V172 2828 10059 853 4776 5329 3135 6409 6861 495 4043 7362 3585
Deer Wood pigeon liver Wood pigeon liver Wood pigeon liver and Wood pigeon liver Wood pigeon spleen Wood pigeon liver Wood pigeon liver and Wood pigeon liver and Wood pigeon liver and Crane liver Wood pigeon spleen Wood pigeon liver
Copenhagen, Denmark Yvelines, France Gers, France Yvelines, France Gers, France Yvelines, France Gers, France Val d’Oise, France Val d’Oise, France Loiret, France Haut Rhin, France Loiret, France Gers, France
316.F
Antigen and vaccine preparation reference strain
1 2 3 4 5 6
7 8 9 10 11 12 13 14 15 16 17 18
38
spleen
spleen spleen spleen
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 phosphatase. Pulsed-field electrophoresis patterns confirmed the differentiation of this organism from M . paratuberculosis (27). In addition, experimental infections of calves revealed pathogenicity similar to that of M. avium (20) since strain 2103 induced acute infections and killed the animals within 36 to 60 days after inoculation (43). Strain 12547 was isolated from a mesenteric lymph node of a deer. This strain did not exhibit the characteristic DNA pattern of M . paratuberculosis or wood pigeon mycobacteria as determined by pulsed-field electrophoresis (data not shown). Strain M21, which was isolated from a wood pigeon, fell into cluster 2 because of cultural characteristics that were similar to those of M. avium. Growth occurred on egg medium and was not prevented by p-nitrobenzoate or by sodium chloride on Middlebrook 7H10 medium. Growth was not stimulated at pH 5.5. However, strain M21 exhibited a pulsed-field pattern that was characteristic of the wood pigeon mycobacteria (data not shown). Experimental infections demonstrated the ability of this strain to produce
paratuberculosis in cattle and tuberculosis in hens (11, 28, 43). Cluster 3 had a mean internal matching score of 91%; this cluster was composed of 13 strains (12 strains isolated from birds and 1 strain isolated from a mammal). All of the strains isolated from birds produced lesions in livers and spleens similar to those observed in birds infected with M. avium. Concomitant pathogenicity for calves was confirmed by experimental infections with two strains (strains 6861 and 6409) (43). Similarly, strain VI72, which was isolated originally from a wild ruminant suffering from Johne’s disease, has been shown to cause avian tuberculosis in hens and paratuberculosis in calves (25, 43). Strains in this cluster are characterized by the inability to grow on egg medium and stimulation of growth by pH 5.5 (Table 2). Four strains belonging to this cluster were examined by FIGE and produced identical patterns, which were different from the profiles observed for M. paratuberculosis and M . avium strains (27). Cluster 4 consisted of a single strain, M . paratuberculosis 316F. This strain has been extensively subcultured since it
MYCOBACTIN-DEPENDENT MYCOBACTERIA
VOL.40, 1990
257
TABLE 2. Variable and differential characteristics of M. paratuberculosis and related mycobacteria ~
Cluster 1 Characteristic
Rods Coccobacillary Colonies smooth Colonies rough Growth at 42°C Growth stimulation by pyruvate' Growth stimulated in medium at pH 5.5' Mycobactin (2 mglliter) is required for growth' Growth on egg medium' 5% NaCl tolerance on Middlebrook 7H10 medium Enzymatic activities Catalase after heating at 68°C Acid phosphatase (2 h) Alkaline phosphatase (2 h)' Trehalase (5 h) Growth in the presence of para-Nitrobenzoate (500 pglml) D-Cycloserine (30 pglml) D-Cycloserine (50 pg/ml)' Ethambutol (7.5 pglml) Pyrazinamide (50 kg/ml) Rifampin (1 pglml) Ansamycin (1 pg/ml) Streptomycin (2 pg/ml) a
Cluster 3
RKC
Cluster 4 13
-
003008 003026 016030 016031 017035 098193 099031 099038 099030 018006
+ + + + + +
024425 034136 034137 034142
-
016264 098183 099033 099034 099035 040164 099036 019510
5/18' 13/18 0118 18/18 16/18 18/18 0118 18/18 18/18 0118
13/13 0113 0113 13/13 13/13 0113 13/13 0113 2/13 0113
-
18/18 0118 1/18 6/18
13/13 1/13 13/13 3/13
+ + + + + + +
4/13 18/18 17/18 18/18 14/18 13/18 15/18 13/18
0113 11/12 0/13 12/13 13/13 1/13 1/13 7/13
-
+
I1 E t 4 q )
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 Kingdom, laboratory (17) and is used as a live vaccine (17, 24). It is characterized by negative responses to all of the tests that are useful for discriminating among clusters 1, 2, and 3 (Table 2) and differs from other M . paratuberculosis strains by its inability to grow on egg medium, its mycobactin independence, its lack of stimulation of growth by pyruvate, and its susceptibility to cycloserine. Numerical analysis split the 38 strains which we studied into four well-defined clusters (Fig. 1). Cluster 1 corresponds to M . paratuberculosis; cluster 2 contains six strains, including the type strains of M . avium and M . intracellulare; cluster 3 contains the wood pigeon mycobacteria; and cluster 4 contains M . paratuberculosis 316F, which is used for vaccination. This separation based on cultural and biochemical tests is highly correlated with pulsed-field patterns and with pathogenicity, as confirmed by experimental infections in hens and calves (11, 27, 28, 43). The mean internal matching scores range from 85 to 91% for clusters 1 through 3. These percentages must be compared with circumspection with values reported in other numerical analyses which generated clusters corresponding to different mycobacterial species. In this numerical analysis we used a set of highly related strains and relied on selected tests. Our selection resulted in the high clustering levels observed in this study. Of the 38 strains studied, 2 were not correctly identified by the tests which we used. The M . intraceflufaretype strain and wood pigeon mycobacterium strain M21 both fell into cluster 2 along with M . avium strains. This misidentification may be explained by the nature of the tests. M . avium and M . intracellulare cannot be separated by cultural or biochemical tests but instead require hybridization to specific DNA probes or determination of the types of T-catalases
(38, 46). Determinations of the serotypes by detecting peptidoglycolipids (mycoside C) are valuable but have to be interpreted carefully because DNA-DNA hybridization results lead to a classification of the serotypes between M . avium and M . intracellulare that is different from the Schaefer classification (2, 38, 51). The misidentification of wood pigeon mycobacterium strain M21 was also due to the use of primarily cultural tests since this strain had peculiar growth characteristics which were more like those of M . avium than those of the other wood pigeon mycobacteria, especially growth on Lowenstein-Jensen medium, tolerance to inhibitors (such as p nitrobenzoate), and absence of stimulation by pH 5.5. M . paratuberculosis 316F, the only member of cluster 4, has been subcultured extensively since 1940 for antigen and vaccine production. This may explain the phenotypic and genomic differences now observed between this strain and the other M . paratuberculosis strains, as shown by our numerical analysis as well as by the peculiar pulsed-field pattern of this organism (27). M . paratuberculosis and wood pigeon mycobacteria can be easily separated on the basis of their responses to six tests. M . paratuberculosis is characterized by a mycobactin requirement, the ability to grow on egg medium, tolerance to cycloserine (50 pg/ml), growth stimulation by pyruvate but not by pH 5.5, and an absence of alkaline phosphatase production. Wood pigeon mycobacteria have the opposite characteristics in these six tests. This set of tests includes five cultural tests and only one biochemical test. Lack of balance between cultural and biochemical discriminative criteria may contribute to the misidentification of frequently subcultured strains which become adapted to a particular medium (e.g., strain M21 as discussed above). Additional
258
13
3418 1277 12998 5 ATCC19698T 1528 CDLYON 12998 1 2 12998 2 147 8 9 1466 1077 1394 7912
14
8642
1 2
3 4
5
6 7
8 g 10 11 12
INT. J . SYST.BACTERIOL.
THOREL ET AL.
15 1 2 9 9 8 9 16 1 2 9 9 8 4 17 1 4 6 4 18 CD2569 19 1 2 5 4 7 2 0 S t 18 21 2103 22 A T C C 2 5 2 9 1 " 23 M21 24 A T C C l 3 9 5 0 ' 25 V 1 7 2
# ## ### #### ##### ###### &&&#### &&& # # # & #
8&&###&&#
&&&###&& # # //a&& # # #/ /
XXX///&&&&/# XXX///3€3€3€X/ ///XXX/=== x3€3€===3€3€
cluster 1 M.paratu berculosis
g
= #
x ## -= &## X3€3€///3€X3€X/ WX=# ===xxx=;/=== = # / # / #
---
--
# ## &##
w23 3
#ax# =xx #
26 2828 27 1 0 0 5 9 28 8 5 3 29 4 7 7 6 30 5329 31 3135 32 33 34 35 36 37 38
cluster 2
6409 6861 495 4043 7362 3585
316F
#
x# X& P
a## 3a###
X&####
cluster 3
=/##### wood pigeon mycobacteria =I # # # # # # =/####### %&&&#### XI///&&&/# =&&&&###&/# ==////&&&//## #
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 identification tests used for such strains. In order to take into account the results of genomic and phenotypic studies, we propose a change in nomenclature. M . avium, M . paratuberculosis, and wood pigeon mycobacteria have been shown to belong to a single genomic species. M . avium was described first, and thus the species name should be M . avium. The following designations are proposed: M . aviurn subsp. avium and M . avium subsp. paratuberculosis (as previously mentioned by several authors [18, 391). The name M . avium subsp. silvaticum is proposed to replace the trivial designation wood pigeon mycobacteria and the " M . avium subsp. name columbae," which was suggested by Saxegaard and Baess in their recent study on relationships among M . avium, M . paratuberculosis, and wood pigeon mycobacteria (39). The subspecific epithet of M . avium subsp. silvaticum reflects the various sources of these peculiar mycobacteria, which have been isolated from
wild animals, mammals (especially deer), and birds (especially wood pigeons and cranes). Emended description of M. avium Chester 1901. Cells are short to long rods with no cord formation. Colonies may be smooth or rough. The colony type may change after first isolation from pathological material through subsequent subcultures. Most strains are nonphotochromogenic and may become yellow on aging. However, some strains are scotochromogenic and have bright yellow pigment. Growth occurs after more than 7 days at 37"C, which is the optimal temperature , and may require mycobactin and/or special media other than egg-based media. Growth may be stimulated by pH 5.5 or pyruvate. Strains do not produce niacin, peroxidase, nitrate reductase, urease, arylsulfatase, penicillinase, or P-glucosidaseand do not hydrolyze Tween 80 in 10 days. Strains produce small amounts of thermoresistant catalase. All strains have the same mycolic acid pattern, namely, mycolate I (a-mycolate), mycolate IV (keto-myco-
VOL.40, 1990
MY COBACTIN-DEPENDENT MYCOBACTERIA
late), and mycolate VI (dicarboxymycolate) (12, 26). All strains are resistant to thiophene-2-carboxylic acid hydrazide and isoniazid. Strains may be susceptible to p nitrobenzoate, ethambutol, pyrazinamide, rifampin, and streptomycin. Most strains are susceptible to ansamycin. The validity of the species has been demonstrated by the results of numerous DNA-DNA hybridization studies and by determinations of the thermal stabilities of the hybrids (1, 2, 23, 31, 52, 53). The type strain is strain ATCC 25291. The guanine-plus-cytosine content is 70 mol% (3). The species includes both potential and obligate parasites and is divided into subspecies on the basis of pathogenicity and host range. Description of M . avium subsp. avium. M . avium subsp. avium is frequently encountered in the environment and is the agent of tuberculosis in birds. It may be involved in infections of a very large set of animals (cattle, sheep, goats, pigs, cats, kangaroos) (20, 41). The main infections in humans are pulmonary infections in adults, submandibular adenopathies in children, and disseminated infections in patients suffering from acquired immunodeficiency syndrome (36, 48, 50). The type strain is strain ATCC 25291. Description of M . avium subsp. paratuberculosis. M . avium subsp. paratuberculosis strains possess the properties of M . avium described above and the additional features indicated in Table 2. A mycobactin requirement for growth is characteristic of the subspecies. Strains have not been isolated from the environment, are obligate pathogens for ruminants, and cause paratuberculosis, a chronic enteric disease (35). This organism may be involved in Crohn’s disease in humans (7). The type strain is strain ATCC 19698. Description of M . avium subsp. silvaticum. M . avium subsp. silvaticum strains possess the properties of M . avium described above and the additional features indicated in Table 2. Inability to grow on egg medium and stimulation of growth at pH 5.5 are characteristic of the subspecies. Strains have not been isolated from the environment, are obligate pathogens for animals, and cause tuberculosis in birds and paratuberculosis in mammals (11, 28, 29, 43). The type strain is strain 6409 (= CIP 103317 [Collection Nationale de Cultures de Microorganismes, Paris, France]). ACKNOWLEDGMENTS
We gratefully thank H. L. David for continuous help and support during ths investigation and P. A. D. Grimont for helpful advice. LITERATURE CITED 1. Baess, I. 1979. Deoxyribonucleic acid relatedness among species of slowly-growing mycobacteria. Acta Pathol. Microbiol. Scand. Sect. B 87:221-226. 2. Baess, I. 1983. Deoxyribonucleic acid relationships between different serovars of Mycobacterium avium, Mycobacterium intracellulare, and Mycobacterium scrofulaceum. Acta Pathol. Microbiol. Scand. Sect. B 91:201-203. 3. Baess, I., and B. Mansa. 1978. Determination of genome size and base ratio on deoxyribonucleic acid from mycobacteria. Acta Pathol. Microbiol. Scand. Sect. B 86:309-312. 4. Barclay, R., D, F. Ewing, and C. Ratledge. 1985. Isolation, identification, and structural analysis of the mycobactins of Mycobacterium avium, Mycobacterium intracellulare, Mycobacterium scrofulaceum, and Mycobacterium paratuberculosis. J. Bacteriol. 164:89&903. 5 . Camphausen, R. T., R. L. Jones, and P. J. Brennan. 1988. Antigenic relationship between Mycobacterium paratuberculosis and Mycobacterium avium. Am. J. Vet. Res. 49:1307-1310. 6. Carle, G. F., M. Frank, and M. V. Olson. 1986. Electrophoretic separations of large DNA molecules by periodic inversion of the electric field. Science 232:65-68. 7. Chiodini, R. J. 1989. Mycobacterium paratuberculosis: an
259
emerging pathogen? Acta Leprol. 7:Sl&S17. 8. Chiodini, R. J., H. J. Van Kruiningen, W. R. Thayer, and J. A. Coutu. 1986. Spheroplastic phase of mycobacteria isolated from patients with Crohn’s disease. J. Clin. Microbiol. 24:357-363. 9. Collins, D. M., and G. W. De Lisle. 1986. Restriction endonuclease analysis of various strains of Mycobacterium paratuberculosis isolated from cattle. Am. J. Vet. Res. 47:222&2229. 10. Collins, D. M., D. M. Gabric, and G. W. DeLisle. 1989. Identification of a repetitive sequence specific to Mycobacterium paratuberculosis. FEMS Microbiol. Lett. 60:175-178. 11. Collins, P., A. McDiarmid, L. H. Thomas, and P. R. J. Matthews. 1985. Comparison of the pathogenicity of Mycobacterium paratuberculosis and Mycobacterium spp. isolated from the wood pigeon (Columba palumbus L.). J. Comp. Pathol. 95591-597. 12. DaffC, M., M. A. Laneelle, C. Asselineau, V. Levy-FrCbault, and H. David. 1983. IntCret taxonomique des acides gras des mycobactkries: proposition d’une mkthode d’analyse. Ann. Microbiol. (Paris) 134B:241-256. 13. Damato, J. J., C. Knisley, and M. T. Collins. 1987. Characterization of Mycobacterium paratuberculosis by gas-liquid and thin-layer chromatography and rapid demonstration of mycobactin dependence using radiometric methods. J. Clin. Microbiol. 252380-2383. 14. David, H. L. 1977. Alkaline phosphatases from Mycobacteriurn smegmatis. J. Gen. Microbiol. 101:99-102. 15. David, H. L., M. T. Jahan, A. Jumin, J. Grandry, and E. H. Lehman. 1978. Numerical taxonomy analysis of Mycobacterium africanum. Int. J. Syst. Bacteriol. 28:467472. 16. David, H. L., V. Levy-FrCbault, and M. F. Thorel. 1989. Mkthodes de laboratoire pour Mycobactkriologie clinique. Institut Pasteur, Paris. 17. Doyle, T. M. 1964-1965. Strains of Mycobacteriumjohnei used for the preparation of vaccine. State Vet. J. 19-20:154155. 18. Grange, J. M. 1983. The mycobacteria, p. 60-93. I n M. Parker (ed.), Topley and Wilson’s principles of bacteriology, virology and immunity, 7th ed., vol. 2. Edward Arnold, London. 19. Green, E. P., M. T. Moss, J. Hermon-Taylor, and J. J. McFadden. 1989. Insertion elements in mycobacteria. Acta Leprol. 7: S239-S242. 20. Griffith, F. 1911. Final report of the Royal Commission on Tuberculosis, part 2. Darling and Son, London. 21. Hampson, S. J., F. Portaels, J. Thompson, E. P. Green, M. T. Moss, J. Hermon-Taylor, and J. J. McFadden. 1989. DNA probes demonstrate a single highly conserved strain of Mycobacterium avium infecting AIDS patients. Lancet ii:65-68. 22. Hance, A. J., B. Grandchamp, V. LCvy-Frdbault, D. Lecossier, J. Rauzier, D. Bocart, and B. Gicquel. 1989. Detection and identification of mycobacteria by amplification of mycobacterial DNA. Mol. Microbiol. 3:843-849. 23. Hurley, S. S., G. A. Splitter, and R. A. Welch. 1988. Deoxyribonucleic acid relatedness of Mycobacterium paratuberculosis to other members of the family Mycobacteriaceae. Int. J. Syst. Bacteriol. 38:143-146. 24. Jorgensen, J. B. 1984. The effect of vaccination on the excretion of Mycobacterium paratuberculosis, p. 131-136. In J. B. Jorgensen and 0. Aalund (ed.), Paratuberculosis. Diagnostic methods, their practical application and experience with vaccination. Commission of European Communities, Brussels. 25. Jorgensen, J. B., and B. Clausen. 1976. Mycobacteriosis in a roe-deer caused by wood pigeon mycobacteria. Nord. Veterinaermed . 28539-546. 26. LCvy-Frebault, V., K. S. Goh, and H. L. David. 1986. Mycolic acid analysis for clinical identification of Mycobacterium avium and related mycobacteria. J . Clin. Microbiol. 24:835-839. 27. LCvy-Frebault, V. V., M.-F. Thorel, A. Varnerot, and B. Gicquel. 1989. DNA polymorphism in Mycobacterium paratuberculosis, “wood pigeon mycobacteria,” and related mycobacteria analyzed by field inversion gel electrophoresis. J. Clin. Microbiol. 27:2823-2826. 28. Matthews, P. R. J., and A. McDiarmid. 1979. The production in bovine calves of a disease resembling paratuberculosis with a Mycobacterium sp. isolated from a woodpigeon (Columba pa-
260
THOREL ET AL.
lumbus L.). Vet. Rec. 104:286. 29. McDiarmid, A. 1948. The occurrence of tuberculosis in the wild wood pigeon. J. Comp. Pathol. 58:128-133. 30. McFadden, J. J., P. D. Butcher, R. Chiodini, and J. HermonTaylor. 1987. Crohn’s disease-isolated mycobacteria are identical to Mycobacterium paratuberculosis, as determined by DNA probes that distinguish between mycobacterial species. J. Clin. Microbiol. 25796-801. 31. McFadden, J. J., P. D. Butcher, R. J. Chiodini, and J. HermonTaylor. 1987. Determination of genome size and DNA homology between an unclassified Mycobacterium species isolated from patients with Crohn’s disease and other mycobacteria. J. Gen. Microbiol. 133:211-214. 32. McFadden, J. J., J. Thompson, E. Hull, S. Hampson, J. Stanford, and J. Hermon-Taylor. 1988. The use of cloned DNA probes to examine organisms isolated from Crohn’s disease tissue, p. 515-520. In R. P. MacDermott (ed.), Inflammatory bowel disease: current status and future approach. Elsevier, Amsterdam. 33. McIntyre, G., and J. L. Stanford. 1986. ImmunodifFusion analysis shows that Mycobacterium paratuberculosis and other mycobactin-dependent mycobacteria are variants of Mycobacterium avium. J. Appl. Bacteriol. 61:295-298. 34. Merkal, R. S. 1979. Proposal of ATCC 19698 as the neotype strain of Mycobacterium paratuberculosis Bergey et al. 1923. Int. J. Syst. Bacteriol. 29:263-264. 35. Merkal, R. S. 1984. Paratuberculosis, p. 1237-1249. I n G. P. Kubica and L. G. Wayne (ed.), The mycobacteria. A sourcebook. Marcel Dekker, New York. 36. Portaels, F. 1987. Le SIDA et les mycobactCries atypiques. Ann. SOC.Belge Med. Trop. 67:93-116. 37. Rogosa, M., M. I. Krichevsky, andR. R. Colwell. 1971. Method for coding data on microbial strains for computers, edition AB. Int. J. Syst. Bacteriol. 21:1A-l75A. 38. Saito, H., H. Tomioka, K. Sato, H. Tasaka, M. Tsukamura, F. Kuze, and K. Asano. 1989. Identification and partial characterization of Mycobacterium avium and Mycobacterium intracelMare by using DNA probes. J. Clin. Microbiol. 27:994-997. 39. Saxegaard, F., and I. Baess. 1988. Relationship between Mycobacterium avium, Mycobacterium paratuberculosis and “wood pigeon mycobacteria.” Acta Pathol. Microbiol. Scand. 96: 37-42. 40. Saxegaard, F., I. Baess, and E. Jantzen. 1988. Characterization of clinical isolates of Mycobacterium paratuberculosis by DNADNA hybridization and cellular fatty acid analysis. Acta Pathol. Microbiol. Scand. 96:497-502. 41. Thorel, M. F. 1980. Note sur 1’Ctude comparative des skrotypes des souches de Mycobacterium avium isol6es de l’homme et de l’animal. Ann. Microbiol. (Paris) 131A:71-76.
INT.J. SYST.BACTERIOL. 42. Thorel, M. F., and P. Desmettre. 1982. Etude comparative de souches de mycobacteries mycobactine-dkpendantes, isolkes de pigeon rarnier, avec Mycobacterium avium et M . paratuberculosis: etude des caractkres biologiques et antigdniques. Ann. Microbiol. (Paris) 133B:291-302. 43. Thorel, M. F., P. Pardon, K. Irgens, J. Marly, and P. Lechopier. 1984. Paratuberculose experimentale: pouvoir pathogene chez le veau de souches de mycobacteries mycobactine-dependantes. Ann. Rech. Vet. 15365-374. 44. Walczak, C. A., and M. I, Krichevsky. 1980. Computer methods for describing groups from binary phenetic data: preliminary summary and editing of data. Int. J. Syst. Bacteriol. 30:615-621. 45. Wards, B. J., D. M. Collins, and G. W. De Lisle. 1987. Restriction endonuclease analysis of members of the Mycobacterium avium-M. intracellulare-M. scrofulaceum serocomplex. J. Clin. Microbiol. 252309-23 13. 46. Wayne, L. G., and G. A. Diaz. 1986. Differentiation between T-catalases derived from Mycobacterium avium and M . intracellulare by a solid-phase immunosorbent assay. Int. J. Syst. Bacteriol. 36:363-367. 47. Wayne, L. G., R. C. Good, M. I. Krichevsky, 2.Blacklock, H. L. David, D. Dawson, W. Gross, J. Hawkins, P. A. Jenkins, I. Juhlin, W. Kappler, H. H. Kleeberg, V. LCvy-Frebault, C. McDurmont, E. E. Nel, F. Portaels, S. Rusch-Gerdes, K. H. Schroder, V. A. Silcox, I. Szabo, M. Tsukamura, L. Van den Breen, B. Vergmann, and M. A. Yakrus. 1989. Third report of the cooperative, open-ended study of slowly growing mycobacteria by the International Working Group on Mycobacterial Taxonomy. Int. J. Syst. Bacteriol. 39:267-278. 48. Wendt, S. L., K. L. George, B. C. Parker, H. Gruft, and J. 0. Falkinham III. 1980. Epidemiology of infection by nontuberculous mycobacteria. 111. Isolation of potentially pathogenic mycobacteria from aerosols. Am. Rev. Respir. Dis. 122:259-263. 49. Whipple, D. L., R. B. LeFebvre, R. E. Andrews, and A. B. Thiermann. 1987. Isolation and analysis of restriction endonuclease digestive patterns of chromosomal DNA from Mycobacterium paratuberculosis and other Mycobacterium species. J. Clin. Microbiol. 251511-1515. 50. Wolinsky, E. 1979. Nontuberculous mycobacteria and associated diseases. Am. Rev. Respir. Dis. 119:107-159. 51. Wolinsky, E., and W. B. Schaefer. 1973. Proposed numbering scheme for mycobacterial serotypes by agglutination. Int. J. Syst. Bacteriol. 23:182-183. 52. Yoshimura, H. H., and D. Y. Graham. 1988. Nucleic acid hybridization studies of mycobactin-dependent mycobacteria. J. Clin. Microbiol. 26:1309-1312. 53. Yoshimura, H. H., D. Y. Graham, M. K. Estes, and R. S. Merkal. 1987. Investigation of association of mycobacteria with inflammatory bowel disease by nucleic acid hybridization. J. Clin. Microbiol. 2545-51.
