International Journal of Systematic and Evolutionary Microbiology (2001), 51, 413–418 Printed in Great Britain

Mycoplasma agassizii sp. nov., isolated from NOTE the upper respiratory tract of the (Gopherus agassizii) and the (Gopherus polyphemus)

M. B. Brown,1 D. R. Brown,1 P. A. Klein,3 G. S. McLaughlin,1 I. M. Schumacher,3 E. R. Jacobson,2 H. P. Adams4 and J. G. Tully5†

Author for correspondence: M. B. Brown. Tel: j1 352 392 4700 ext. 3970. Fax: j1 352 846 2781. e-mail: mbbrown!nersp.nerdc.ufl.edu

1,2,3 Departments of Biochemical, serological and molecular genetic studies were performed on 1 Pathobiology and seven isolates that were recovered from the upper respiratory Small Animal Clinical Sciences2 , College of tract of clinically ill desert tortoises. The isolates were serologically related to Veterinary Medicine, each other but serologically distinct from previously described . Unique and Department of mycoplasma species-specific 16S rRNA nucleotide sequences were found in the Pathology and Laboratory Medicine, proposed type strain. The name Mycoplasma agassizii is proposed for these College of Medicine3 , isolates. The type strain is PS6T (l ATCC 700616T) which caused upper University of Florida, respiratory tract disease (URTD) in experimentally infected tortoises. Gainesville, FL 32610, USA 4 Department of Cell Biology, Baylor College Keywords: mycoplasma, tortoise, upper respiratory tract disease of Medicine, Houston, TX 77030, USA 5 Mycoplasma Section, National Institute of Allergy and Infectious Diseases, Frederick, MD 21702, USA

Recent dramatic population declines in the desert have also been isolated from the trachea tortoise (Gopherus agassizii) in California have resulted and lung. Antibody to the mycoplasma species was in the listing by the United States Federal Government found in 15 of 16 tortoises with respiratory lesions of this animal as a threatened species in areas north (Schumacher et al., 1993). In a serological survey of and west of the Colorado River (Department of desert tortoises in Nevada, 84% of animals with Interior, 1990). In addition to habitat degradation, clinical signs of URTD tested positive by ELISA drought and predation, an upper respiratory tract (Schumacher et al., 1997). In experimental infection disease (URTD) was identified as a major contributory studies (Brown et al., 1994, 1999) clinical isolates of the factor in population declines. We initially demon- new mycoplasma caused URTD in both the desert and strated by electron microscopy the presence of classical gopher tortoise (Gopherus polyphemus), fulfilling the mycoplasma-like organisms attached to the respir- majority of the Koch–Henle–Evans postulates and atory surface, often between the cilia (Jacobson et al., demonstrating the pathogenic potential of the species 1991). These observations led to attempts to culture (Evans, 1976). The isolates from the desert tortoise did and isolate the mycoplasmal species. In conjunction not react with antisera against other known myco- with a detailed pathology study, we have isolated plasmal species. In this study we describe the charac- mycoplasmas from 5 of 9 (56%) ill tortoises and 4 of teristics of the isolates and propose that these strains 12 (33%) apparently healthy tortoises, some of which should be classified as a new species in the had microscopic lesions. The most common sites of Mycoplasma, Mycoplasma agassizii sp. nov. The type isolation were the choana and nasal passages, but strain of this species is PS6T which was used in the

...... experiments to demonstrate an aetiologic role for the † Present address: 16400 Black Rock Road, Germantown, MD 20874, USA. mycoplasma in URTD. Abbreviation: URTD, upper respiratory tract disease. Desert tortoises used in this study were obtained under The GenBank accession number for the 16S rRNA gene sequence of special permit from the US Fish and Wildlife Service. Mycoplasma agassizii is U09786. Tortoises had been collected in the Las Vegas Valley,

01633 # 2001 IUMS 413 M. B. Brown and others

Table 1 Chelonian species from which Mycoplasma agassizii has been recovered by culture or detected by PCR

Scientific name Common name Status*

Geochelone chilensis Captive Geochelone pardalis Captive Geochelone elegans Captive Geochelone forstenii Travancore tortoise Captive Geochelone sulcata African spurred tortoise Captive Gopherus agassizii Desert tortoise Wild and captive Gopherus polyphemus Gopher tortoise Wild and captive sp. Unnamed tortoise Captive Terrapene carolina bauri Wild and captive Testudo graeca graeca Spur-thighed tortoise Captive Testudo graeca ibera Spur-thighed tortoise Captive

* Wild status refers to tortoises which were captured and sampled in the field. Captive status refers to tortoises which were part of a private collection or housed under captive conditions.

Las Vegas, NV, USA, and were held at the Desert arginine, 10B, A8 and SP-4 (ATCC medium 988) Tortoise Conservation Center. Clinically ill and ap- (Freundt, 1983; Tully et al., 1979). Both horse serum parently clinically healthy animals were sent to the and fetal bovine serum were used as supplements. University of Florida for intensive pathology studies. Initial incubation temperatures included 20, 25, 30 and Isolates designated PS were cultured from these 35 mC. Cultures were incubated in both ambient air animals at necropsy. Additional samples, designated and 5% CO#. The only successful initial isolations as numerals only, were submitted as nasal washes from were obtained with SP-4 containing 20% fetal bovine tortoises in a relocation study. Nasal washes were serum. Optimal primary isolation rates were obtained obtained in the field by gently flushing the nares of the at 30 mC. Primary isolations required 3–6 weeks for tortoise with 1 ml sterile tryptose broth via a catheter growth in broth or on agar. Single colonies were attached to a 1 ml syringe. The broth was placed in expanded in broth, replated and the procedure re- cryotubes, frozen in liquid nitrogen and sent to the peated two times. Strains recovered under these University of Florida on dry ice. conditions were then purified by conventional filter- cloning procedures through three consecutive passages All strains described in this study were isolated from T (Tully, 1983). Subsequent broth passages could be tortoises with URTD. Strains PS6 , PS10, 553 and 111 adapted to grow at 37 mC. were isolated from the nares of desert tortoises with clinical signs of URTD. Strain PS13T was isolated Morphology was examined by conventional darkfield from the trachea of a tortoise with URTD; strain and electron microscopy. Ultrathin sections of micro- PS13C was isolated from the choana of that same organisms from broth cultures were prepared by tortoise. The micro-organism has been demonstrated techniques described by Deines & Bullivant (1968). in the upper respiratory tract of a number of other The morphology of strain PS6T cultivated in SP-4 captive chelonians (Table 1), but the pathogenic broth at 30 mC for 48 h was examined under darkfield potential in these species has not been confirmed microscopy at a magnification of i1250. Agar col- although clinical signs compatible with URTD have onies of strain PS6T were obtained on SP-4 agar after been observed in some of these species. Reference incubation at 30 mC under anaerobic conditions in the strains and antisera used in the serological cross- Gaspak system (BBL Microbiology Systems). reaction tests were from the Mycoplasma Reference Laboratory Collection, NIAID, Frederick, MD, USA All strains recovered from the tortoises that showed a (Table 2). Most, if not all, of these reference reagents colour change in broth were subjected to six passages are available from the Culture Collection, in broth without bacterial inhibitors. At each passage, University of Florida, Gainesville, FL, USA. Myco- subcultures were made on blood agar, brain heart plasma testudinis, the only other described myco- infusion agar and in brain heart infusion broth to determine if reversion to bacterial forms had occurred. plasmal species of chelonian origin (Hill, 1985), was T obtained from the American Type Culture Collection The ability of strain PS6 and other tortoise isolates to (ATCC), Manassas, VA, USA. Rabbit antiserum was pass through filters was determined by passage through T 0n45 and 0n22 µm porosity membrane filters (Tully, prepared to Mycoplasma agassizii PS6 and Myco- T plasma testudinis. 1983). The ability of strain PS6 and other tortoise isolates to ferment glucose, hydrolyse arginine and Initial isolation attempts were made using a wide urea, and produce phosphatase was tested using variety of media, including Frey’s, modified Hayflick’s, standard methods (Aloutto et al., 1970). Haemag-

414 International Journal of Systematic and Evolutionary Microbiology 51 Mycoplasma agassizii sp. nov.

Table 2 Mycoplasma species tested by growth inhibition and epi-immunofluorescence for cross-reactivity with Mycoplasma agassizii

Mycoplasma adleri G145T Mycoplasma edwardii PG24T Mycoplasma meleagridis 17529T Mycoplasma agalactiae PG2T Mycoplasma elephantis E42T Mycoplasma moatsii MK405T Mycoplasma alkalescens D12T Mycoplasma equigenitalium T37T Mycoplasma mobile 163KT Mycoplasma alvi IlsleyT Mycoplasma equirhinis M432\72T Mycoplasma molare H542T Mycoplasma anatis 1340T Mycoplasma falconis H\T1T Mycoplasma muris RIII4T Mycoplasma anseris 1219T Mycoplasma fastidiosum 4822T Mycoplasma mustelae MX9T Mycoplasma arginini G230T Mycoplasma faucium DC333T Mycoplasma mycoides subsp. mycoides PG1T Mycoplasma arthritidis PG6T Mycoplasma felifaucium PUT Mycoplasma mycoides subsp. capri PG3T Mycoplasma auris UIAT Mycoplasma feliminutum BenT Mycoplasma neurolyticum Type AT Mycoplasma bovigenitalium PG11T Mycoplasma felis COT Mycoplasma opalescens MH5408T Mycoplasma bovirhinis PG43T Mycoplasma fermentans PG18T Mycoplasma orale CH19299T Mycoplasma bovis DonettaT Mycoplasma flocculare Ms42T Mycoplasma ovipneumoniae Y98T Mycoplasma bovoculi M165\69T Mycoplasma gallinaceum DDT Mycoplasma oxoniensis 128T Mycoplasma buccale CH20247T Mycoplasma gallinarum PG16T Mycoplasma penetrans GTU54T Mycoplasma buteonis Bb\T2gT Mycoplasma gallisepticum PG31T Mycoplasma phocacerebrale 1049T Mycoplasma californicum ST-6T Mycoplasma gallopavonis WR1T Mycoplasma phocarhinis 852T Mycoplasma canadense 275CT Mycoplasma gateae CST Mycoplasma phocidae 105T Mycoplasma canis PG14T Mycoplasma genitalium G37T Mycoplasma pirum 70-159T Mycoplasma capricolum subsp. Mycoplasma glycophilum 486T Mycoplasma pneumoniae FHT capricolum Calif.kidT Mycoplasma capricolum subsp. Mycoplasma gypis B1\T1T Mycoplasma primatum HRC292T capripneumoniae F38T Mycoplasma caviae G122T Mycoplasma hominis PG21T Mycoplasma pullorum CKKT Mycoplasma cavipharyngis 117CT Mycoplasma hyopharyngis H3-6BFT Mycoplasma pulmonis PG34T Mycoplasma citelli RG-2CT Mycoplasma hyopneumoniae JT Mycoplasma putrefaciens KS-1T Mycoplasma cloacale 383T Mycoplasma hyorhinis BTS7T Mycoplasma salivarium PG20T Mycoplasma collis 58BT Mycoplasma hyosynoviae S16T Mycoplasma simbae LXT Mycoplasma columbinasale 694T Mycoplasma imitans 4229T Mycoplasma spermatophilum AH159T Mycoplasma columbinum MMP-1T Mycoplasma iners PG30T Mycoplasma spumans PG13T Mycoplasma columborale MMP-4T Mycoplasma iowae 695T Mycoplasma sturni UCMFT Mycoplasma conjunctivae HRC581T Mycoplasma lagogenitalium 12MST Mycoplasma sualvi Mayfield BT Mycoplasma corogypsi BV1T Mycoplasma leonicaptivi 3L2T Mycoplasma subdolum TBT Mycoplasma cottewii VIST Mycoplasma leopharyngis LL2T Mycoplasma synoviae WVU 1853T Mycoplasma cricetuli CHT Mycoplasma lipofaciens R171T Mycoplasma testudinis 01008T Mycoplasma crocodyli MP145T Mycoplasma lipophilum MaByT Mycoplasma verecundum 107T Mycoplasma cynos H831T Mycoplasma maculosum PG15T Mycoplasma yeatsii GIHT Mycoplasma dispar 462\2T

glutination and haemadsorption were tested using conjugated antisera to each of the type strains listed in guinea pig erythrocytes and tortoise erythrocytes as Table 2. described by Gardella & Del Giudice (1983). T T The 16S rRNA gene of PS6 was amplified using A requirement for sterol by strain PS6 was established the upstream primer (5h-AGAGTTTGATCCTGGC- by testing the susceptibility of the strain to 1n5% TCAGGA-3h) specific for Gram-positive digitonin and by growth responses in medium that did (Robertson et al., 1993) and the downstream primer not contain serum or other supplements (Freundt et (5h-TGCACCATCTGTCACTCTGTTAACCTC-3h) al., 1973; Rose et al., 1993). specific for the genus Mycoplasma (van Kuppeveld et Antiserum to PS6T was prepared in rabbits using a al., 1992) as previously described (Brown et al., 1995). standard technique (Senterfit, 1983). Standard disc For PCR, 50 pg mycoplasmal DNA was used as growth inhibition tests (Clyde, 1983) were performed template. Reaction conditions were 10 mM Tris, pH T on agar plates with strain PS6 and antisera to the 8n3, 50 mM KCl, 0n1% Triton X-100, 2n0 mM MgCl#, Mycoplasma species listed in Table 2, including anti- 200 µM deoxynucleoside triphosphates, 0n5 µM each sera to both fermentative and non-fermentative primer and 2n5UTaq polymerase. The template was T species. In addition, agar colonies of PS6 were also denatured for 45 s at 94 mC followed by primer tested in a direct epi-immunofluorescence test annealing at 59 mC for 1 min and a 2 min poly- (Gardella et al., 1983) against individual fluorescein- merization at 72 mC for 35 cycles in a Coy Thermo-

International Journal of Systematic and Evolutionary Microbiology 51 415 M. B. Brown and others

primary isolation before a colour change indicative of growth was observed. After a few passages in SP-4 broth, strains did appear to adapt and grow somewhat faster, frequently changing the pH of the medium from acid production within 24–48 h. Strains grew best at 30 mC. If strains were in expo- nential growth in broth culture, some, but not all, could be successfully transferred to 37 mC. None of the strains grew at 37 mC if inoculated directly from a stock culture stored at k70 mC. All but one strain (PS13T) remained viable after storage at 4 mC for 2 months. This could have biological significance since the host species is a reptile and undergoes periods of hiber- nation during winter months. The ability to survive during the dormant period would enhance survival of the microbe in its host. No reversion to walled bacterial forms was observed after numerous passages in -free broth or agar plates. The strains were " all resistant to 500 units ml− , since this was a component in the original isolation medium or had been used routinely in most culture media to grow the organisms. Preliminary tests to determine the serum requirement ...... for growth on all the tortoise isolates indicated a need Fig. 1. Electron micrograph of ultrathin section of Mycoplasma for serum or sterol. It was demonstrated that PS6T agassizii strain PS6T isolated from a desert tortoise with upper respiratory tract disease. Note the presence of a trilaminar could be maintained through 23 serial dilutions in membrane and absence of a (arrow). Bar, 0n5 µm. medium containing fetal bovine serum, but that growth did not occur in serum-free broth or in a serum-free medium supplemented with 0n04% poly- oxyethylene sorbitan (Tween 80) (Rose et al., 1993). cycler model 110P (Robertson et al., 1993; van Strain PS6T and all other tortoise strains isolated at Kuppeveld et al., 1992). The amplified product was this time catabolized glucose, but did not hydrolyse predicted to be 1047 bp. Mycoplasma felis DNA was arginine or urea. Phosphatase was not produced. included as a positive control template and Escheri- Haemagglutination was somewhat variable, with titres chia coli served as the negative control template. ranging from 4 to 16. Colonies of most mycoplasma Sequencing of the amplified DNA segment to confirm strains in the study haemadsorbed both guinea pig the presence of conserved mycoplasmal genus-specific and tortoise erythrocytes, but two strains (PS10 and sequences (Robertson et al., 1993; van Kuppeveld et PS13C) failed to demonstrate significant haema- al., 1992) was performed by the DNA sequencing core dsorptive properties. All strains tested were filterable laboratory at the Interdisciplinary Center for Bio- through 0n45 µm filters and two strains (PS13T and technology Research, University of Florida. 111) were filterable through 0n22 µm. Morphological examination by electron microscopy of T Growth inhibition tests performed with antiserum to strain PS6 revealed coccoid to pleomorphic cell forms all previously recognized species in the genus Myco- between 350 and 900 nm in size. The micro-organisms plasma (Table 2) revealed that strain PS6T was sero- lacked a cell wall and were surrounded by a trilaminar logically distinct. These findings were confirmed by the unit membrane (Fig. 1). Some strains seemed to have a results of direct agar plate immunofluorescence tests. rudimentary tip structure but this was not well defined, The two test procedures clearly indicated that PS6T nor has it been shown as yet to be involved in cellular and other tortoise isolates examined here were distinct attachment. In darkfield microscopy, exponential- T from Mycoplasma testudinis. All strains described in phase broth cultures of PS6 examined at i1250 this study had multiple proteins recognized by rabbit magnification demonstrated a rather homogeneous antiserum to PS6T in Western blot analysis (data not population, containing mostly small round cells that shown) and the lack of cross-reactions between Myco- were non-helical and non-motile. Colony morphology T plasma testudinis and antiserum or conjugated anti- was variable. On primary isolation, strain PS6 serum to PS6T was also confirmed by Western blotting. exhibited extremely small mulberry colonies on SP-4 agar. However, after broth passage and on some Both the disc growth inhibition test and the epi- primary isolations, a more classic fried-egg mor- immunofluorescence test, as well as a metabolism- phology was sometimes observed. The strains were inhibition test (Taylor-Robinson, 1983), were used to slow growers and required 3–6 weeks incubation upon establish the interrelationships between the various

416 International Journal of Systematic and Evolutionary Microbiology 51 Mycoplasma agassizii sp. nov.

ABCDE name Mycoplasma agassizii is proposed for strain T bp PS6 . 1353 1078 Description of Mycoplasma agassizii sp. nov. 872 Mycoplasma agassizii [agahssi.z.ii. M.L. gen. n. 603 agassizii of Agassiz, named after J. Louis R. Agassiz, a notable Swiss born American naturalist who first described the North American Testudinata (Agassiz, 310 1857) and after whom the desert tortoise was named 281/271 (Cooper, 1863)]. 234 194 Cells are coccoid to pleomorphic with no cell wall but with a clearly defined three-layer membrane. Non- 118 helical and non-motile. Some strains appear to possess 72 a rudimentary tip structure. Colonies on agar are variable, with both mulberry and fried-egg charac- ...... teristics. Mulberry colonies are common on primary Fig. 2. Mycoplasma genus-specific amplification of a segment isolation plates. Primary isolation requires 3–6 weeks of the 16S rRNA gene. Lanes: A, HaeIII-digested φX174 size of incubation. It does not grow well above 30 mC. Best standard; B, negative PCR control; C, Mycoplasma felis positive growth is achieved with SP-4 medium. Cells pass PCR control; D, 1047 bp Mycoplasma agassizii PCR product; E, Escherichia coli negative specificity control. PCR amplification through 0n45 µm filters and some, but not all strains products were separated on a 1% agarose gel and visualized by also pass through 0n22 µm filters. No reversion to staining with ethidium bromide. bacterial forms is seen. It is inhibited by digitonin and requires serum or sterol for growth. Glucose is fermented, but arginine and urea are not hydrolysed. tortoise isolates reported here. In all these com- Phosphatase is not produced. It haemagglutinates parisons, strain PS6T was clearly serologically related guinea pig and tortoise erythrocytes. Most, but not all to the other tortoise isolates examined in this study. strains haemadsorb guinea pig and tortoise erythro- cytes. It contains conserved mycoplasmal genus- Conserved mycoplasmal 16S rRNA sequences were T specific sequences in the 16S rRNA gene and is found in strain PS6 (Brown et al., 1995). A PCR band serologically distinct from other Mycoplasma species. of the predicted size was observed with Mycoplasma T The 16S rRNA gene sequence is unique. The habitat is agassizii PS6 and with the positive control, Myco- the respiratory tract of the desert tortoise and gopher plasma felis; no band was seen with the negative tortoise, with the most common sites of isolation being control, Escherichia coli (Fig. 2). Based on the 16S the nares and choana. It is virulent and has been rRNA sequence, Mycoplasma agassizii was placed in demonstrated in experimental infections to cause the Mycoplasma hyorhinis clade, with Mycoplasma upper respiratory tract disease in both the desert T mobile as its nearest neighbour (Brown et al., 1995). tortoise and gopher tortoise. Type strain is PS6 (l T The properties of strain PS6T described in this paper ATCC 700616 ) and is also available from the fulfil the criteria for species descriptions of members of Mollicutes Culture Collection (Curators Dr J. K. the class Mollicutes (International Committee on Davis and M. K. Davidson, University of Florida). Systematic Bacteriology Subcommittee on the Tax- onomy of Mollicutes, 1995). The organism does not Acknowledgements have a cell wall, is filterable, fails to revert to walled bacteria when grown in the absence of , is This work was funded by grants from The Nature Con- servancy and Walt Disney Development Company to resistant to penicillin and produces colony forms E.R.J., M.B.B. and P.A.K. typical of other mollicutes. Based on the growth requirement for serum or cholesterol, an optimum growth temperature of 30–37 mC, the presence of the References conserved sequences and the genus-specific ampli- Aloutto, B. B., Wittler, R. G., Williams, C. O. & Faber, J. E. (1970). fication of the 16S rRNA gene, the failure to hydrolyse Standardized bacteriologic techniques for the characterization urea and the requirement for sterol, we propose that of mycoplasma species. Int J Syst Bacteriol 20, 35–38. this micro-organism should be assigned to the family Agassiz, L. (1857). Contributions to the Natural History of the and the genus Mycoplasma. Be- United States of America. Vol. 1. North American Testudinata, cause the strains do not cross-react serologically with pp. 235–452. other known species in the genus Mycoplasma,we Brown, D. R., Crenshaw, B. C., McLaughlin, G. S., Schumacher, conclude that these strains represent a new species of I. M., McKenna, C. E., Klein, P. A., Jacobson, E. R. & Brown, M. B. pathogenic mycoplasma as defined by the Inter- (1995). Taxonomic analysis of the tortoise mycoplasmas Myco- national Committee on Systematic Bacteriology Sub- plasma agassizii and Mycoplasma testudinis by 16S rRNA gene committee on the of Mollicutes (1995). The sequence comparison. Int J Syst Bacteriol 45, 348–350.

International Journal of Systematic and Evolutionary Microbiology 51 417 M. B. Brown and others

Brown, M. B., Schumacher, I. M., Klein, P. A., Harris, K., Correll, T. Mollicutes (Division Tenericutes). Int J Syst Bacteriol 45, & Jacobson, E. R. (1994). Mycoplasma agassizii causes upper 605–612. respiratory tract disease in the desert tortoise. Infect Immun 62, Jacobson, E. R., Gaskin, J. M., Brown, M. B., Harris, R. K., 4580–4586. Gardiner, C. H., LaPointe, J. L., Adams, H. P. & Regiarrdo, C. Brown, M. B., McLaughlin, G. S., Klein, P. A., Crenshaw, B. C., (1991). Chronic upper respiratory tract disease of free-ranging Schumacher, I. M., Brown, D. R. & Jacobson, E. R. (1999). Upper desert tortoises, Xerobates agassizii. J Wildl Dis 27, 296–316. respiratory tract disease in the gopher tortoise is caused by van Kuppeveld, F. J., van der Logt, J. T., Angulo, A. F., van Zoest, Mycoplasma agassizii. J Clin Microbiol 37, 2262–2269. M. J., Quint, W. G., Niesters, H. G., Galama, J. M. & Melchers, Clyde, W. A. (1983). Growth inhibition tests. Methods Myco- W. J. (1992). Genus- and species-specific identification of myco- plasmol 1, 405–410. plasmas by 16S rRNA amplification. Appl Environ Microbiol Cooper, J. G. (1863). Description of Xerobates agassizii. Proc 58, 2606–2615. Calif Acad Sci 2, 118–123. Robertson, J. A., Verkis, A., Bebear, C. & Stemke, G. W. (1993). Deines, L. & Bullivant, S. (1968). Morphology and reproductive Polymerase chain reaction using 16S rRNA gene sequences processes of -forms of bacteria. II. Comparative study of - distinguishes the two biovars of Ureaplasma urealyticum. J Clin forms and Mycoplasma with the electron microscope. J Microbiol 31, 824–830. Bacteriol 95, 672–687. Rose, D. L., Tully, J. G., Bove, J. M. & Whitcomb, R. F. (1993). A Department of Interior (1990). Endangered and threatened test for measuring growth responses of mollicutes to serum and wildlife and plants: determination of threatened status for the polyoxyethylene sorbitan. Int J Syst Bacteriol 43, 527–532. Mojave population of the desert tortoise. Federal Register 55, Schumacher, I. M., Brown, M. B., Jacobson, E. R., Collins, B. R. & 12178–12191. Klein, P. A. (1993). Detection of antibodies to a pathogenic Evans, A. S. (1976). Causation and disease: the Henle–Koch mycoplasma in desert tortoises (Gopherus agassizii) with upper postulates revisited. Yale J Biol Med 49, 175–195. respiratory tract disease (URTD). J Clin Microbiol 31, Freundt, E. A. (1983). Culture media for classical mycoplasmas. 1454–1460. Methods Mycoplasmol 1, 127–136. Schumacher, I. M., Hardenbrook, D. B., Brown, M. B., Jacobson, Freundt, E. A., Andrews, B. E., Erno, H., Kunze, H. & Black, F. T. E. R. & Klein, P. A. (1997). Relationship between clinical signs of (1973). The sensitivity of Mycoplasmatales to sodium-poly- upper respiratory tract disease and antibodies to Mycoplasma anetholsulfonate and digitonin. Zentbl Bakteriol Parasitenkd agassizii in desert tortoises from Nevada. J Wildl Dis 33, Infektkrankh Hyg Abt 1 Orig Reihe A 225, 104–112. 261–266. Gardella, R. S. & Del Giudice, R. A. (1983). Hemagglutination, Senterfit, L. (1983). Preparation of antigens and antisera. hemadsorption, and hemolysis. Methods Mycoplasmol 1, Methods Mycoplasmol 1, 401–404. 379–384. Taylor-Robinson, D. (1983). Metabolism inhibition tests. Gardella, R. S., Del Giudice, R. A. & Tully, J. G. (1983). Immuno- Methods Mycoplasmol 1, 411–417. fluorescence. Methods Mycoplasmol 1, 411–439. Tully, J. G. (1983). Cloning and filtration techniques for myco- Hill, A. C. (1985). Mycoplasma testudinis, a new species isolated plasmas. Methods Mycoplasmol 1, 173–177. from a tortoise. Int J Syst Bacteriol 35, 489–492. Tully, J. G., Rose, D. L., Whitcomb, R. F. & Wenzel, R. P. (1979). International Committee on Systematic Bacteriology Subcom- Enhanced isolation of Mycoplasma pneumoniae from throat mittee on the Taxonomy of Mollicutes (1995). Revised minimum washings with a newly modified culture medium. J Infect Dis standards for the description of new species of the class 139, 478–482.

418 International Journal of Systematic and Evolutionary Microbiology 51