INTERNATIONALJOURNAL OF SYSTEMATICBACTERIOLOGY, July 1977, p. 241-246 Vol. 27, No. 3 Copyright 0 1977 International Association of Microbiological Societies Printed in U.S.A.

Mycobacterium malmoense sp. nov.

K. H. SCHRODER AND I. JUHLIN Research Institute Borstel, Institute for Experimental Biology and Medicine, 2061, Borstel, Germany; and Department of Clinical Bacteriology, University of Lund, General Hospital, Malmo, Sweden

Strains of a new type of group 111 nonphotochromogenic mycobacteria have been repeatedly isolated from four patients with clinical and roentgenological signs of lung mycobacteriosis. These strains split nicotinamide and pyrazinam- ide as do those of Mycobacterium avium, but they show no esterase activity. Furthermore, the strains produce thermolabile catalase and hydrolyze Tween 80. They also have a unique lipid pattern and special sensitins and agglutinins. These strains, therefore, are considered as belonging to a new species of patho- genic, nonphotochromogenic mycobacteria, for which we propose the name Mycobacterium malmoense. Strain M-o 816 is the type strain, and it has been deposited in the American Type Culture Collection under the number 29571.

During the years 1968 through 1970, seven tol, rifampin, p-aminosalicylic acid, or thiosemicar- unusual mycobacterial strains were isolated bazone were inoculated with mg of mycobacte- from four patients from Malmo and Lund, Swe- ria per tube; those with ethionamide, capreomycin, den. All of the strains grew extremely slowly viomycin, kanamycin, or cycloserine were inocu- and poorly as tiny colonies. Their properties, as lated with mg per tube. Two control tubes were determined at laboratories in Malmo, Sweden, included in each test series, as were two additional controls, with an inoculum of lo+ mg of bacteria per and in Borstel, Germany, indicated that they tube. Cultures were incubated for 4 weeks at 37°C. belonged to a new species. This paper reports Strains were reported as resistant if they grew in the results of a taxonomic study of these the tubes containing the critical concentrations of strains. agent used (see Table 1) and showed at least 1% (inoculum, mg of mycobacteria per ml) or 10% MATERIALS AND METHODS (inoculum, lop4 mg of mycobacteria per ml) of the growth appearing in the control tubes; they were Bacterial strains. The seven strains of atypical reported as susceptible if they did not grow at these mycobacteria studied (referred to herein as the concentrations. The range between no growth and Malmo strains) and their sources are cited below. growth of 1% of that in the control tubes is called Strain M-6 648 was isolated from a needle aspiration “doubtful.” of a process in the right lung, and strains M-6 649 The influence of oxygen on growth was investi- and M-o 650 were isolated from gastric lavage and gated in the oxygen gradient of a semisolid agar bronchial secretion, respectively, of the patient. column (Lebek test, modification from Schmiedel Strains M-o 709 and M-o 710 were isolated from and Gerloff [14]), and the production of acid from gastric lavage and sputum, respectively, from the glycerol was determined in egg medium with bromo- second patient, who had two more sputum samples cresol purple as an indicator (6, 14). which were culture positive. Strain M-o 716 was Other properties studied included amidase pat- isolated from the sputum of the third patient, who tern (l), production of nicotinic acid (121, reduction had four more positive samples. Strain M-o 816 was of nitrak (2), hydrolysis of Tween 80 (161, production isolated from resected lung tissue in the fourth pa- of phosphatase (41, esterases (51, arylsulfatase (91, tient, who also had positive sputum cultures. and catalase, and the thermolability of catalase (3). The results of earlier studies (11) of strains of For pathogenicity tests (lo), 4-week-old cultures Mycobacterium avium, Mycobacteriurn intrace&- were used. Hens and rabbits were infected intrave- lare, and Mycobacterium xenopi were compared with nously and guinea pigs were infected subcutane- the properties of these new strains. ously with 1ml of a mycobacterial suspension of lop2 Strain characteristics. Colony morphology and mg of mycobacteria per ml of each strain. After 8 the ability of the strains to grow at various tempera- weeks, the animals were sacrificed if they had not tures (22, 31, 37, 39, 43, and 45°C) were observed already died. Reisolations were made from liver, during 6 weeks of incubation of cultures on Lowen- spleen, and lymph nodes. stein-Jensen medium inoculated with mg of Serological studies were performed by W. Anz mycobacteria per tube. and G. Meissner (Borstel, Germany) and by W. B. Drug susceptibility testing was performed with Schaefer (Denver, Colo.) by the method described by Lowenstein-Jensen medium in tubes closed with Schaefer (131, ‘lipid patterns were analyzed by P. stoppers which allowed exchange of air. Chemother- Jenkins (Cardiff, England) (81, and an investigation apeutic agents were added before inspissation. of sensitins was made by M. Magnusson (Copen- Tubes containing isoniazid, streptomycin, ethambu- hagen, Denmark) (7). 241 242 SCHRODER AND JUHLIN J. INT. SYST.BACTERIOL. RESULTS bonate with raised, compact centers sur- rounded by thin granular borders with entire or Microscopic observation of cells of the seven slightly irregular edges; some low, convex, cir- strains grown on Lowenstein-Jensen medium cular, coarsely granular colonies, 0.9 to 1.7 mm revealed coccoid to short rods, but, when grown in diameter, with sharp, entire edges were also on Middlebrook and Cohn 7H10 agar or corn- produced (Fig. 2 and 3). Although the strains meal agar, the cells were short to moderately have been subcultivated over a period of several long rods, acid alcohol-fast, and gram positive years, growth is still very slight upon subcul- but without such characteristics as cords, cross- turing. bands, or interwoven networks. No spores, cap- As shown in Table 1, the strains were suscep- sules, or aerial hyphae could be seen. tible to 1 pg of ethambutol, 16 pg of ethionam- On Lowenstein-Jensen medium, all of the ide, 8 pg of kanamycin, and 16 pg of cycloserine strains grew as dysgonic, smooth, colorless col- and were resistant to the remaining antibacte- onies when incubated from 22 to 37°C. Growth rial agents used. The susceptibilities of the at 22°C was observed only after 6 weeks of strains to viomycin were not uniform. incubation. The colonies remained colorless In the oxygen gra.dient of the Lebek test, the after exposure to light. Dilute inocula on Mid- Malmo strains grew at the same depth as Myco- dlebrook and Cohn 7H10 agar yielded smooth, bacterium bouis, which means that they needed glistening, grayish-white, opaque, domed, cir- less oxygen than other mycobacteria. As with cular colonies, 0.5 to 1.5 mm in diameter with M. bouis, they grew without color change in the entire margins, and some umbonate colonies, bromocresol purple medium. 0.6 to 2.7 mm in diameter, with compact, raised None of the strains produced nicotinic acid, centers and flattened, irregular edges (Fig. 1). and none of them reduced nitrate; however, all On cornmeal agar, the colonies were smaller, of them split the amides nicotinamide and pyr- 0.6 to 1.7 mm in diameter, grayish-white, um- azinamide. Tween 80 was hydrolyzed, and all of

a

FIG. 1. A smooth, glistening colony (right) and an umbonate colony (left)with a raised compact center and with flattening at the periphery. Middlebrook and Cohn 7H10 agar. Strain M-o 816. VOL. 27, 1977 MYCOBACTERIUM MALMOENSE SP. NOV. 243

FIG. 2. A smooth, glistening colony (right) and an umbonate colony (left) with a raised, compact center and with flattening at the periphery. Cornmeal agar. Strain M-o 816. the strains produced catalase, which was inac- differs from those previously known. A descrip- tivated by heating at 68°C. Phosphatase and a- tion of this new serotype will be published and /I-esterases were not detectable, and the later. Furthermore, all of the strains showed arylsulfatase test was negative. In Table 2, the same unique lipid pattern (determined by some of the properties of the Malmo strains are P. A. Jenkins, Cardiff, England). Their sensitins compared with those of closely related mycobac- were similar, too, and they deviated from those teria. of M. avium, M. intracellulare, M. gastri, M. In guinea pigs, all of the strains produced a kansasii, and M. tuberculosis (with which they strong local reaction but no generalization. In were compared (by M. Magnusson, Copen- hens, four of the seven strains caused macro- hagen, Denmark). scopic lesions in the liver and spleen, and two of the hens died before the end of the test. A heavy DISCUSSION growth of mycobacteria appeared when cul- The Malmo strains are similar to those of M. tures were made from the organs of these hens; avium, M. intracellulare, and M. xenopi with no growth was obtained from specimens from regard to hen pathogenicity, amidase pattern, the surviving hens. The rabbits showed few or and inability to reduce nitrate. They are differ- no lesions; specimens from liver and spleen entiated from strains of the M. avium complex gave either no growth or only single colonies. by the phosphatase-esterase pattern, which is The serological investigation of the aggluti- specific for M. avium and related species, and nins (performed by W. Anz and G. Meissner, * by the thermolabile catalase. The ability to Borstel, Germany) revealed that all of the hydrolyze Tween 80 is an unusual property for strains are of the same serotype, one which pathogenic strains because this property ordi- 244 SCHRODER AND JUHLIN INT. J. SYST.BACTERIOL.

P

*'

I

x

1.0 mm FIG. 3. A small, smooth, glistening colony (lower left) and several umbonate colonies with compact centers and flattened peripheries. Cornmeal agar. Strain M-o 816.

TABLE1. Susceptibilities of the Malmo strains and related mycobacteria to antimicrobial agents a I I I Isonia- Strepto- Etham- Rifam. D- Amino- Thiosem- Ethion- Viqmy- Kana- Taxa zid Pin salicylic icarba- amide cin mycin (0.25 1 (4.0mr pg/ b;t;l (32.0 acid (0.5 zone (1.0 (16.0 (16.0 (8.0 pg/ml) 1 pg/ml) pg/ml: pg/ml) pg/d pglml) pg/ml) pg/ml) Strains: M-O 648 R ~s R R S R S R S M-O 649 ~R R S R R S S S R S M-O 650 S R R S R S R S M-O 709 R S R R S S S R S M-O 710 R S R R S R S R S M-O 716 R S R R S S S R S M-O 816 R S R R S S S R S Species: M. avium R R Var R R R S M. intra- R R S R R R S cell u- lare M. xenopi R R S'S S S S

a Symbols: R, resistant; S, susceptible; susceptible, but resistance may occur; Var, results not uniform. Sx, VOL. 27, 1977 MYCOBACTERIUM MALMOENSE SP. NOV. 245

narily is seen only among the nonpathogenic Barbro Olsson (Lund), Lillemor Fredriksson (Malmo), and strains of group 111 (nonphotochromogenic) my- Kathe Hoff (Borstel) for their skillful technical assistance. cobacteria. REPRINT REQUESTS The susceptibility of the Malmo strains to ethambutol, ethionamide, kanamycin, and cy- Address reprint requests to: Assoc. Prof. Ingmar Juhlin, University of Lund, Department of Clinical Bacteriology, closerine in combination with resistance to the Malmo General Hospital, Malmo, Sweden. remaining drugs studied is remarkable. This S214 01 pattern is not introduced by treatment, and it LITERATURE CITED has not been previously reported. 1. Bonicke, R. 1961. Die Bedeutung der Acylamidasen Furthermore, the Malmo strains are distin- Fur die Identifizierung und Differenzierung der ver- schiendenen Arten der Gattung Mycobacterium. guished by a unique surface antigen, a new Jahresber. Borstel. lipid composition pattern, and a new type of 5:7-87. 2. Bonicke, R., E. Rohrscheidt, and E. Pascoe. 1962. Die sensitin. For these reasons, we regard these Verbreitung der Nitratreduktase innerhalb der Gat- strains as belonging to a new species, for which tung Mycobacterium. Naturwissenschafken 49:43-44. 3. Kubica, G. P., and G. L. Pool. 1960. Studies on the we propose the name Mycobacterium mal- catalase activity of acid-fast bacilli. I. An attempt to moense (M.L. adj. malmoensis belonging to subgroup these organisms on the basis of their cata- Malmo, Sweden, the source of the strains on lase activity at different temperatures and pH. Am. which the original description of this species is Rev. Respir. Dis. 81:387-391. 4. Kappler, W. 1965. Zur Differenzierung von Mykobak- based). Strain M-o 816 is the type strain of M. terien mit dem Phosphatase-Test. Beitr. Klin. Tub- malmoense; a culture of this strain has been erk. Spezifischen Tuberk. Forsch. 130:223-226. deposited in the American Type Culture Collec- 5. Kappler, W. 1965. Acetyl-Naphthylamin-Esterasen-Ak- tion under the number 29571. tivitat von Mykobakterien. Beitr. Klin. Tuberk. Spe- Up to the present, M. malmoense strains zifischen Tuberk. Forsch. 13O:l-4. 6. Lebek, G. 1958. Uber den Nachweis des unterschied- have been isolated only from human beings. lichen Sauerstoffoptimums des humanen und bovi- The isolates were pure cultures, and all of the nen Mykobacterium tuberculosis. Zentralbl. Bakte- patients had clinical and roentgenological signs riol. Parasitenkd. Infektionskr. Hyg. Abt. 1 Orig. of lung mycobacteriosis. No other mycobacteria Reihe A 173:581-587. 7. Magnusson, M. 1967. Identification of species of myco- or other pathogens were isolated from these bacterium on the basis of the specificity of the delayed patients. A report of the clinical data will be type reaction in guinea pigs. Z. Tuberk. 12755-56. published separately in the near future. 8. Marks, J., P. A. Jenkins, and W. B. Schaefer. 1971. Thinlayer-chromatography of mycobacterial lipids as an aid to classification: technical improvements: My- ACKNOWLEDGMENTS cobacterium avium, M. intracellulare (Battey bacilli). We thank W. Anz and G. Meissner, who performed the Tubercle 52:219-225. serological typing, the late W. B. Schaefer, who verified the 9. Marks, J., and D. R. Trollope. 1960. A study of the serological typing, P. A. Jenkins, who investigated the lipid “anonymous” mycobacteria. I. Introduction; colonial pattern, and M. Magnusson, who made the sensitin-typing characteristics and morphology; growth rates; bio- studies. We also want to express our gratitude to B. G. chemical tests. Tubercle 4151-62. Simonsson and B. Arvastsan (at the Department of Internal 10. Meissner, G. 1959. Untersuchungen an atypischen My- Chest Diseases, University of Lund), who made the case kobakterien. 11. Vergleichende tierexperimentelle reports of three of the patients available to us, and also to Untersuchungen zur Frage ihere Pathogenitat und 246 SCHRODER AND JUHLIN INT. J. SYST.BACTERIOL.

Virulenz. Beitr. Klin. Tuberk. Spezifischen Tuberk. 79~663-665. Forsch. 121:365-380. 13. Schaefer, W. B. 1965. Serologic identification and clas- 11. Meissner, G., K. H. Schroder, G. E. Amadio, W. Anz, sification of the atypical mycobacteria by their agglu- S. Chaparas, H. W. B. Engel, P. A. Jenkins, W. tination. Am. Rev. Respir. Dis. 92:85-92. Kappler, H. H. Kleeberg, E. Kubala, M. Kubin, D. 14. Schmiedel, A., and W. Gerloff. 1965. Dreifach-Differen- Lauterbach, A. Lind, M. Magnusson, ZD. Mikova, S. zierung von Mykobakterien in der Agar-Hohen- R. Pattyn, W. B. Schaefer, J. L. Stanford, M. Tsuka- Schicht-Kultur. Prax. Pneumol. 19:528-536. mura, L. G. Wayne, I. Willers, and E. Wolinsky. 15. Wagener, K., and E. Mitscherlich. 1951. Zwei neue 1974. A co-operative numerical analysis of nonscoto- Nahrboden zur kulturellen Differenzierung der Tub- and nonphotochromogenic slowly growing mycobac- erkelbakterientypen. Tuberkulosearzt 5:274-283. teria. J. Gen. Microbiol. 83:207-235. 16. Wayne, L. G., J. R. Doubek, and R. L. Russel. 1964. 12. Runyon, E. €I.,M. J. Selin, and H. W. Harris. 1959. Classification and identification of mycobakteria. I. Distinguishing mycobacteria by the niacin-test. A Tests employing Tween 80 as substrate. Am. Rev. modified procedure. Am. Rev. Tuberc. Pulm. Dis. Respir. Dis. 90:588-597.