Detection of Urogenital Mycoplasmal Infections in Primates by Use of Polymerase Chain Reaction

Trenton R. Schoeb,1 Kevin Dybvig,2 Karen F. Keisling,1 Maureen K. Davidson,1 and Jerry K. Davis1

Abstract Urogenital mycoplasmal infections could affect use of primates as models for reproductive system studies and could affect reproduction in captive primates, but could be useful as animal models of similar human infections. We conducted a pilot study to assess detection of urogenital mycoplasmal infections in primates by use of polymerase chain reaction (PCR). Healthy animals were anesthetized, and vaginal, cervical, or endometrial and urethral swab specimens were collected from females and males, respectively. Speci- mens were tested by PCR supplemented with dot blotting and nonradiolabeled oligonucleotide probing for 16S rRNA sequences conserved among mollicutes. Specimens with positive results were tested by species- specific PCRs with primers for 16S rRNA sequences of Ureaplasma urealyticum and Mycoplasma hominis and for adhesin gene sequences of Mycoplasma genitalium. Spiked duplicate reactions were included as internal controls for each reaction. Results for 232 specimens from 166 animals indicate that naturally acquired urogenital infections are readily detected and suggest that urogenital mycoplasmal infections are common in laboratory primates (48/166 [29%] overall). M. hominis and U. urealyticum appeared to be common among the studied primates overall and especially in chimpanzees. Mycoplasmas other than M. genitalium, M. hominis, and U. urealyticum appeared to be at least as common as these three, with specimens from 18 of 48 animals (38%) having positive “generic” PCR results, but no positive results in species-specific PCRs.

Mycoplasmal infections may have roles in human uro- primates and could complicate use of primates for repro- genital and reproductive disease. Ureaplasma urealyticum ductive system research. They also could provide useful is a urogenital tract commensal that is sometimes associ- animal models of human infections. Rapid, sensitive de- ated with chorioamnionitis; increased risk of preterm detection methods would aid diagnosis of mycoplasmal infec- livery, low birth weight, and fetal death; and infection, res- tions in nonhuman primates with naturally acquired uro- piratory insufficiency, and bronchopulmonary dysplasia in genital and reproductive disease, and would facilitate se- premature infants (1–8). However, its causal role in these lection and development of primate models of these infec- conditions is controversial (2, 9–12). Similar associations tions by allowing identification of infected and uninfected may exist for Mycoplasma hominis, which occasionally may animals and by determining species in which natural in- be a fetal pathogen (13), but it appears to be less commonly fections with these organisms develop. We conducted a pi- associated with infections of the urogenital tract and in lot study to evaluate detection of urogenital mycoplasmal premature infants than is U. urealyticum. Mycoplasma infections in primates by use of polymerase chain reaction genitalium infection probably causes some cases of (PCR). The results suggest that such infections are com- nongonococcal urethritis (14), but whether it is a potential mon in laboratory primates. Although many animals har- cause of chorioamnionitis or fetal effects is not known. Most bored U. urealyticum or M. hominis, or both, many others mycoplasmas for which humans are considered the natu- probably carried mycoplasmas of other, unidentified, spe- ral hosts, including those typically found in the urogenital cies. tract, also have been isolated from various species of nonhuman primates, and Mycoplasma primatum has been iso- Materials and Methods lated from humans (15–20), suggesting that designation of “Generic” PCR: All specimens were tested by PCR, us- mycoplasmal species as “human” or “primate” is artificial ing the primers GPO-1 (5'-ACTCC TACGG GAGGC AGCAG and probably inaccurate. The importance of urogenital my- TA-3') and MGSO (5'-TGCAC CATCT GTCAC TCTGT coplasma infections in nonhuman primates is unknown, TAACC TC-3'), which recognize 16S rRNA sequences con- but in view of the suspected role of such infections in hu- served among most members of the Mollicutes (20). man urogenital and reproductive disease, it is reasonable Amplicons were visualized by electrophoresis of agarose to suspect that they could affect reproduction in captive gels and ethidium bromide staining. After optimization for Mg2+ concentration and annealing temperature, sensitivity was about 1,000 mycoplasma cells as assessed by PCR Division of Comparative Medicine, University of Florida, Gainesville, Florida1 and Department of Comparative Medicine, University of Alabama with serial dilutions of DNA from U. urealyticum, M. at Birmingham, Birmingham, Alabama2 hominis, M. genitalium, M. pneumoniae, M. penetrans, M.

468 Urogenital Mycoplasmas in Primates

Table 1. Results of polymerase chain reaction analysis of primate urogenital tract specimens Mycoplasma Mycoplasma Ureaplasma Species “Generic” PCRa genitaliuma hominisa urealyticuma Unidentifiedb Macaca fascicularis 7/12(58%) 0/12(0%) 0/12(0%) 1/12(8.3%) 6/12(50%) M. mulatta 2/23(8.7%) 0/23(0%) 0/23(0%) 1/23(4.3%) 1/23(4.3%) M. nemestrina 21/70(30%) 0/70(0%) 7/70(10%) 2/70(2.9%) 12/70(17%) Papio sp. 8/49(16%) 4/49(8.2%) 1/49(2.0%) 5/49(10%) 3/49(6.1%) Pan troglodytes 10/12(83%) 0/12(0%) 7/12(58%) 10/12(83%) 0/12(0%) Totals 48/166(29%) 4/166(2.4%) 15/166(9.0%) 19/166(11%) 22/166(13%) aNo. animals with >1 positive specimen/total no. animals tested bNo. animals with positive “generic” PCR results but no positive specific PCRs/total no. animals tested pirum, M. fermentans, M. pulmonis, M. arthritidis, and M. 15,000 X g in a microcentrifuge. The supernatant was dis- neurolyticum, all of which produced amplicon samples of carded, and sediment samples were treated with 10 to 50 the expected size (715 bp). For each sample, a duplicate l of lysis buffer containing 1% Tween 20, 1% Triton X- reaction spiked with about 104 copies of M. fermentans or 100, and 0.5 mg of proteinase K/ml. Tissue specimens were M. pulmonis DNA served as an internal control. Samples homogenized in lysis buffer; aliquots of the resulting sus- not producing a visible product were dot-blotted and probed pensions were used in PCR reactions after clarification by with the oligonucleotide GPO-2 (5'-CTTAA AGGAA TTGAC centrifugation. GGGAA CCCG-3'), which recognizes a conserved sequence Statistical analysis: Results were analyzed by chi- within the PCR product (20). The probe was labeled with square analysis and Fisher’s test, with Bonferroni’s cor- digoxigenin (Oligonucleotide Tailing Kit; Boehringer- rection used for multiple comparisons where appropriate Mannheim, Indianapolis, Ind.) and visualized, using an al- (22). Calculations were done by use of Statistix for Win- kaline phosphatase-conjugated digoxigenin antibody and dows 1.0 (Analytical Software, Tallahassee, Fla.). Probabil- nitroblue tetrazolium (Genius 3 Nucleic Acid Detection Kit; ity estimates of < 0.05 were considered significant. Boehringer-Mannheim). Sensitivity after dot blotting and probing was about 100 mycoplasma cells. Results Specific PCRs: Samples giving positive “generic” PCR Results for 232 specimens from 166 animals are sum- results were tested by species-specific PCRs, using prim- marized in Table 1, and an example of a gel is shown in ers designed with MacVector software (Oxford Molecular, Figure 1. At least one specimen from 48 of 166 animals Oxford, England) for 16S rRNA sequences of U. urealyticum (29%) had positive results by use of the “generic” PCR; 15 (U.u.F1: 5' GCTAA TACCG AATAA TAACA TC 3'; U.u.R1: of 166 (9.0%) were M. hominis positive, and 19 of 166 (11%) 5' ATGGT ACAGT CAAAC TAAAA TC 3'; 333 bp amplicon) were U. urealyticum positive. Mycoplasma genitalium se- and M. hominis (M.hom.L1: 5' CAATG GCTAA TGCCG quences were found only in specimens from 4 of 49 baboons GATAC 3'; M.hom.R1: 5' GTACC GTCAG TCTGC AATC 3'; (8.2%), or 4 of 166 (2.4%) of all primates studied. Positive 333 bp product), and published primers for adhesin gene “generic” PCR results with no positive specific PCR results sequences of M. genitalium (21) (MG1: 5' AGTTG ATGAA were obtained for 22 of 166 (13%) of the studied primates. ACCTT AACCC CTTGG 3'; MG2: 5' CCGTT GAGGG GTTTT CCATT TTTGC 3'; 281 bp amplicon). The PCR prod- Discussion ucts were visualized by ethidium bromide staining of aga- Although the purpose of the study was to establish the rose gels that had undergone electrophoresis. Each PCR feasibility of detecting urogenital mycoplasmal infections was optimized for Mg2+ concentration and annealing tem- by PCR, rather than specifically to study prevalence, the perature, and sensitivity of each was about 1,000 myco- results indicate that, with an overall prevalence of 29% plasma cells as assessed by PCR with serial dilutions of (95% confidence interval: 22 to 36%), urogenital mycoplas- DNA from each target organism. Specificity was verified mal infections probably are common among laboratory pri- by testing each PCR with DNA of each of the other organ- mates. Prevalence of such infections was significantly isms described previously. For each sample, each PCR was higher in chimpanzees than in other species, with 10 of 12 conducted in duplicate, with the second reaction spiked with (83%) having at least one specimen with positive “generic” approximately 104 copies of DNA of the target organism as PCR results. Specific PCR results indicated that M. an internal control. genitalium was significantly less common than M. hominis Specimens and their preparation: Vaginal, cervical and U. urealyticum, with 7 of 70 (10%) pigtailed macaques or endometrial swab specimens (females) and urethral swab having M. hominis, 5 of 49 (10%) baboons having U. specimens (males) from anesthetized healthy macaques, urealyticum, and 7 of 12 (58%) chimpanzees having M. baboons, and chimpanzees were provided by colleagues. hominis and U. urealyticum. Organisms other than M. Tissue specimens from two aborted fetuses also were sub- genitalium, M. hominis, and U. urealyticum appeared to mitted. (This study was approved by the University of be at least as common as these three, with 22 of 166 ani- Florida Institutional Animal Care and Use Committee.) mals (13%) having positive “generic” PCR results but no Swabs were vortexed in 2 ml of nonenzymatic cell dissocia- positive specific PCR results. Actual prevalences may be tion buffer (Sigma Chemical Co., St. Louis, Mo.), after which even higher than those shown, inasmuch as no results were the swab was discarded and the buffer was centrifuged at obtained for 39 of 166 animals (23%) due to inhibition of

469 Vol 47, No 5 Laboratory Animal Science October 1997

talapoins and patas monkeys from colonies with high rates of infertility and spontaneous abortion (15), and experimentally induced U. urealyticum, M. hominis, and M. genitalium infections can cause urethritis or sal- pingitis in primates of one or more species (23–25). Our results also document the applicability of the PCR methods used to test clinical specimens. These methods could be used to aid development of nonhuman primate models of human infections by providing a rapid and sensitive means to identify and monitor uninfected animals destined for breeding and to help evaluate experimentally induced infections. Inasmuch as natural urogenital mycoplasmal infections clearly occur, and because experimentally induced infections have been established, it is likely that practical models Figure 1. Ethidium bromide-stained agarose gel of PCR products. Lane 1, molecular weight markers; lane 2, M. hominis-positive specimen; could readily be developed for study of effects of such lane 3, M. hominis-negative specimen; lane 4, M. hominis spiked du- infections on fertility and pregnancy outcome. This plicate specimen; lane 5, M. genitalium-positive specimen; lane 6, M. would allow studies to assess the causal roles of these genitalium-negative specimen; lane 7, M. genitalium spiked duplicate organisms in urogenital and reproductive tract disease, specimen; lane 8, U. urealyticum-positive specimen; lane 9, U. to study pathogenetic mechanisms, and to develop pre- urealyticum-negative specimen; lane 10, U. urealyticum spiked duplicate specimen; lane 11, positive “generic” PCR; lane 12, negative “ge- ventive measures and treatments. The PCR methods neric” PCR; lane 13, spiked duplicate “generic” PCR. also could be used in clinical diagnosis to assess the role of urogenital mycoplasmal infections in spontane- ously developing reproductive and urogenital tract dis- the “generic” PCR; considering only those animals with at ease in laboratory primates. Although PCR is techni- least one noninhibiting specimen indicated overall preva- cally demanding, it offers advantages over mycoplas- lence of 48 of 127 (38%). Inhibition was especially common mal culture methods, which are less sensitive and sub- for specimens from baboons; 40 of 69 (58%) specimens, rep- ject to inhibition by antibiotics, and which may require resenting 23 baboons, failed to amplify. On the basis of weeks for positive results to be evident. noninhibiting specimens, the proportion of baboons with at least one positive “generic” PCR result was 8 of 26 (31%). However, comparisons among prevalences for specific mycoplasmas, or among prevalences of infections among host species, should be cautiously interpreted, inasmuch as Acknowledgements specimens from each primate species were obtained from We thank D. Anderson, K. Brasky, K. Mansfield, L. Reyes, and B. Swenson for collection and submission of samples. This work only one or two sources, and sample sizes for chimpanzees was supported by funds from the Division of Sponsored Research, and cynomolgus macaques were small. University of Florida, and by National Institutes of Health grant Our findings are consistent with those of previous stud- no. R01 AI33164. ies, in which urogenital mycoplasmal infections were identified by culture in chimpanzees, baboons, macaques, talapoin monkeys, patas monkeys, and marmosets (15–18). References Similar to our results, U. urealyticum infection was com- 1. Cassell, G. H., J. K. Davis, K. B. Waites, et al. 1987. Patho- mon in chimpanzees (17). In one study, a large proportion genesis and significance of urogenital mycoplasmal infections. Adv. Exp. Med. Biol. 224:93–115. of vaginal isolates could not be identified by use of specific 2. Eschenbach, D. A. 1993. Ureaplasma urealyticum and pre- growth-inhibiting antisera (18), suggesting that nonhuman mature birth. Clin. Infect. Dis. 17(Suppl. 1):S100–S106. primates may harbor unknown species of mycoplasmas or 3. Wang, E. E., G. H. Cassell, P. J. Sanchez, et al. 1993. ureaplasmas in the urogenital tract. Perhaps such organ- Ureaplasma urealyticum and chronic lung disease of prema- isms were present in some of the 22 primates we studied turity: critical appraisal of the literature on causation. Clin. Infect. Dis. 17(Suppl. 1):S112–S116. that had positive “generic” PCR results but no positive re- 4. Waites, K. B., D. T. Crouse, and G. H. Cassell. 1993. Sys- sults by use of U. urealyticum, M. hominis, or M. genitalium temic neonatal infection due to Ureaplasma urealyticum. Clin. PCRs, although many of these organisms may belong to Infect. Dis. 17(Suppl. 1):S131–S135. other species previously isolated from primates (19). In any 5. Stahelin-Massik, J., F. Levy, P. Friderich, et al. 1994. Meningitis caused by Ureaplasma urealyticum in a full term case, urogenital mycoplasmal infections appear to be suffi- neonate. Pediatr. Infect. Dis. J. 13:419–421. ciently prevalent among captive primates that there is ap- 6. Payne, N. R., S. S. Steinberg, P. Ackerman, et al. 1993. preciable potential for adverse health effects and for com- New prospective studies of the association of Ureaplasma plication of use of primates in reproductive research. urealyticum colonization and chronic lung disease. Clin. In- Whether these infections can be associated with adverse fect. Dis. 17:S117–S121. 7. Cassell, G. H., K. B. Waites, H. L. Watson, et al. 1993. effects, such as infertility or abortion, has not been estab- Ureaplasma urealyticum intrauterine infection: role in prema- lished, although ureaplasmal infections were found in turity and disease in newborns. Clin. Microbiol. Rev. 6:69–87.

470 Urogenital Mycoplasmas in Primates

8. Ollikainen, J., H. Hiekkaniemi, M. Korppi, et al. 1993. 17. Khatamee, M. A. 1982. Mycoplasma colonization in primate Ureaplasma urealyticum infection associated with acute res- animal laboratories. Int. J. Gynaecol. Obstet. 20:245–249. piratory insufficiency and death in premature infants. 18. Hutchison, V. E., M. E. Pinkerton, and S. S. Kalter. 1970. J. Pediatr. 122:756–760. Incidence of mycoplasma in nonhuman primates. Lab. Anim. 9. Smyth, A. R., N. J. Shaw, B. C. Pratt, et al. 1993. Care 20:914–922. Ureaplasma urealyticum and chronic lung disease. Eur. J. 19. Somerson, N. L., and B. C. Cole. 1979. The mycoplasma Pediatr. 152:931–932. flora of human and nonhuman primates, p. 191–216. In J. G. 10. Saxen, H., K. Hakkarainen, M. Pohjavuori, et al. 1993. Tully and R. F. Whitcomb (ed.), The mycoplasmas, vol. 2. Hu- Chronic lung disease of preterm infants in Finland is not as- man and animal mycoplasmas. Academic Press, New York. sociated with Ureaplasma urealyticum colonization. Acta 20. van Kuppeveld, F. J., J. T. van der Logt, A. F. Angulo, et Paediatr. 82:198–201. al. 1992. Genus- and species-specific identification of myco- 11. Izraeli, S., Z. Samra, L. Sirota, et al. 1991. Genital mycoplasmas by 16S rRNA amplification. Appl. Environ. Microbiol. plasmas in preterm infants: prevalence and clinical signifi- 58:2606–2615. cance. Eur. J. Pediatr. 150:804–807. 21. Jensen, J. S., S. A. Uldum, J. Sondergard-Andersen, et 12. Heggie, A. D., M. R. Jacobs, V. T. Butler, et al. 1994. Fre- al. 1991. Polymerase chain reaction for detection of Myco- quency and significance of isolation of Ureaplasma plasma genitalium in clinical samples. J. Clin. Microbiol. urealyticum and Mycoplasma hominis from cerebrospinal fluid 29:46–50. and tracheal aspirate specimens from low birth weight in- 22. Snedecor, G. W., and W. G. Cochran. 1980. Statistical meth- fants. J. Pediatr. 124:956–961. ods, p. 115–117. Iowa State University Press, Ames, Iowa. 13. Meis, J. F., F. J. van Kuppeveld, J. A. Kremer, et al. 1992. 23. Moller, B. R., E. A. Freundt, F. T. Black, et al. 1978. Ex- Fatal intrauterine infection associated with Mycoplasma perimental infection of the genital tract of female grivet mon- hominis [letter]. Clin. Infect. Dis. 15:753–754. keys by Mycoplasma hominis. Infect. Immun. 20:248–257. 14. Horner, P. J., C. B. Gilroy, B. J. Thomas, et al. 1993. As- 24. Stipkovits, L., A. N. Marantidi, E. K. Dzikidze, et al. 1990. sociation of Mycoplasma genitalium with acute non-gonococ- Urethral infection of male monkeys by Mycoplasma hominis cal urethritis. Lancet 342:582–585. and Ureaplasma urealyticum. Zentralbl Veterinarmed [B] 15. Kundsin, R. B., T. Rowell, M. C. Shepard, et al. 1975. T- 37:125–134. strain mycoplasmas and reproductive failure in monkeys. Lab. 25. Taylor-Robinson, D., P. M. Furr, J. G. Tully, et al. 1987. Anim. Sci. 25:221–224. Animal models of Mycoplasma genitalium urogenital infec- 16. Furr, P. M., and D. Taylor-Robinson. 1976. The occurrence tion. Isr. J. Med. Sci. 23:561–564. of ureaplasmas in marmosets. Lab. Animals 10:393–398.

471