A DNA probe for detecting Mycoplasma genitalium in clinical specimens

Molecular and Cellular Probes

(1988) 2, 327-335

A DNA probe for detecting Mycoplasma genitalium in clinical specimens

George F. Risi Jr,* David H . Martin,* Jeffrey A. Silbermant and J . Craig Cohent *Department of Medicine, Section of Infectious Diseases and tSection of Genetics, Louisiana State University School of Medicine, New Orleans, Lousiana 70112, USA (Received 4 May 1987, Accepted 7 June 1987)

Despite decades of careful study, the etiologies of all cases of pelvic inflammatory disease (PID) and non-gonococcal urethritis (NGU) have yet to be described . Mycoplasma genitalium is a newly described organism which has been implicated as a cause of both PID and NGU . Because of fastidious growth requirements, prolonged incubation time and frequent overgrowth in clinical specimens by Mycoplasma hominis, non-culture methods need to be developed for its detection . We have cloned M . genitalium DNA by transfection into Escherichia coli using M13 as the vector . Using these segments as templates, we synthesized radiolabelled cDNAs that were tested for specific hybridization with M . genitalium, and clinically isolated genital mycoplasmas presumptively identified as M . hominis, and Ureaplasma urealyticum . A 256 base-pair segment was found to hybridize with M. genitalium with a sensitivity of 10 2 colour-changing units (CCUs) . No cross-hybridization was observed with M. hominis, and cross-hybridization was observed only with large concentrations (>106 CCUs) of U. urealyticum . Because of our choice of M13 as the vector, which contains the Lac Operon of E. coli, slight hybridization occurred with E . coli as well . This cDNA can be used against clinical specimens to determine the ecologic niche and spectrum of disease caused by M . genitalium .

In 1981, Tully et al. reported isolating a new species of mycoplasma from two men with urethritis in whom no other cause for the condition could be found ." The organism was given the name Mycoplasma genitalium . Indirect evidence has accumulated to implicate this organism as the causative agent in some cases of previously unexplained pelvic inflammatory disease (PID) and non-gonococcal urethritis (NGU) . Taylor-Robinson & Furr inoculated six female marmosets intravaginally with M . genitalium . The resulting infection in four animals was accompanied by vaginal

polymorphonuclear leukocyte (PMN) response and serum antibody rise .' Those investigators were also able to infect two of four male chimpanzees by intra-urethral Address correspondence to : George F . Risi, Department of Medicine, Section of Infectious Diseases, 1542 Tulane Avenue, New Orleans, LA 70112-2865, USA .

0890-8508/87/040327+09 $03 .00/0

327

© 1987

Academic Press Limited



328

G . F. Risi et al .

inoculation of the organism . 10 In a follow-up study, six out of six male chimps became infected as determined by four-fold elevations in antibody levels ." Urethral inflammation was confirmed in five chimps, and the organism was cultured from blood samples of two . After intravaginal inoculation, vaginitis and antibody rise developed in four out of four female chimps . Moreover, upper genital tract disease has been produced by direct oviduct inoculation of two female Grivet monkeys and three female marmosets .' All developed a moderate to severe endosalpingitis characterized by the infiltration of acute inflammatory cells into the tubal epithelium as well as lumenal exudate and adhesions between the mucosal folds . In addition to the work done in animals, other indirect evidence implicates M. genitalium as a pathogen in humans . Moller et al .' studied 31 women who had acute PID, but none of the pathogens currently known to be associated with this disease . Twelve showed a four-fold or greater change in antibody titre to M. genitalium during a 4-week follow-up period . Additionally, presumptive isolation of

M. genitalium

was made from the urethras of seven of 22 men having urethritis with no other apparent cause .' Several difficulties must be overcome before conclusive proof of pathogenicity of this organism in humans is established . The organism grows slowly, and 6-8 weeks may be required for primary isolation . Media may deteriorate during that time, lowering the sensitivity of the culture system . Finally, overgrowth by M . hominis is frequently encountered . For these reasons, techniques other than isolation by culture are needed to further study M. genitalium as a putative cause of human disease . We have developed a specific nucleic acid probe that can detect relatively small numbers of organisms . Here, we describe the development of the probe and its characteristics .

METHODS

Sample procurement

and

isolation of DNA

Lyophilized M. genitalium, strain G-37, was propagated in SP-4 media which was prepared as described by Tully et al. 14 The pH of the medium was adjusted to 8 .0, and growth of the organism was detected by a colour change from red to yellow, indicating acid production by the organism . After three passages of the organism in our laboratory, a pool of organisms was prepared, aliquotted and frozen at -70 ° C in SP-4 for use in the studies described below . M. genitalium adheres to the glass or plastic surfaces of the culture container in which it grows . To produce large quantities of the organism, we used tissue culture roller bottles which provided the high surface to liquid ratio important for quantity cultivation . An inoculum of about 105 colour-changing units (CCUs) of M. genitalium was added to 100 ml of SP-4 medium and incubated in an 850 cm' roller bottle until colour change was seen (about 48 h) . After discarding the used medium, cells were lysed by the addition of proteinase K (0 . 1 mg/ml) and sodium dodecyl sulphate (SDS) 0 %) in DNA extraction buffer [20 mm Tris-HCI, pH 8 . 0, 10 mm ethylene diamine-



DNA probe to detect Mycoplasma genitalium

329

tetraacetate (EDTA) and 100 mm NaCI] . DNA was extracted with buffered phenol and buffered phenol :chloroform (1 :1) and spooled with two volumes of cold 95% ETOH . The DNA was resuspended in TLE (10 mm Tris-HCI, pH 8 . 0, 0 . 1 mm EDTA), treated with RNAse T1 (125 units/ml) and RNAse A (0 .05 mg/ml) at 37 ° C for 4 h, and then incubated with proteinase K (0. 05 mg/ml) overnight . Another DNA extraction was performed with buffered phenol and phenol :chloroform (1 :1), then spooled and dissolved in TLE .

Preparation of DNA fragments An aliquot of the DNA was subjected to ultrasonication to produce small, random segments . The sample was then subjected to electrophoresis through a 20% polyacrylamide gel . The gel slice with fragments ranging in size from 200-300 basepairs in length was cut out, and the fragments were eluted and passed over an S&S ELUTIP-D column . The fragment ends were blunted by T-4 DNA polymerase in the presence of all four nucleoside triphosphates at 37°C for 15 min . Fine end blunting was done after precipitation and resuspension in 10 mm Tris-HCI, pH 8 . 0, 5 mm MgCl, with the Klenow fragment of Escherichia coli DNA polymerase 1 in the presence of all four nucleoside triphosphates at 41°C for 15 min . The enzyme was heat-inactivated at 65°C for 15 min and the DNA was extracted in chloroform : isoamyl alcohol (24 :1) .

M13 cloning Bacteriophage M-13, strain MP18 (BRL), was prepared with Hinc II in the presence of calf alkaline phosphatase to prevent rejoining after the addition of DNA ligase . The M . genitalium fragments were ligated into this vector at 4°C for 48 h . Competent E . coli strain JM101 was transfected with 10 pg of the ligation mixture by the technique of Messing . 4 Transfection was done in the presence of CaCI Z on ice for 1 h . The bacteria were then heat-shocked at 42°C for 2 min to stimulate uptake of the phage DNA by the bacteria . The bacteria were then mixed with 0 . 33 mm IPTG and 0 . 02% XGAL, and then plated on YT plates and grown overnight at 37 ° C . Ten Lac (-) plaques were picked at random and grown separately in liquid PBS medium and centrifuged, with retention of the supernatant that contained free M-13 . An aliquot of this supernatant was combined with 20% PEG in 2 . 5 M NaCl . The aliquot was vortexed and centrifuged, and the pellet was resuspended in Tris-EDTA and then extracted with phenol and precipitated with Na acetate and 95% ETOH .

cDNA synthesis The M-13 clone was incubated at 55 ° C for 45 min with the universal pentadecamer sequencing primer in 10 mm Tris-HCI, 5 mm MgCl,, to allow annealing of the primer to the M-13 . Elongation of the primer over the inserted fragment was accomplished by incubation of the primed M-13 with dGTP, dATP and dTTP, alpha-32P-CTP and the Klenow fragment of E. coli DNA polymerase 1 . Polymerization was allowed to



330

G . F. Risi et al.

proceed for 60 min at 41°C . Unincorporated nucleotides were separated by passage over a Sephadex G-50 column, and the DNA was ethanol-precipitated and resuspended in 3 mm EDTA .

Hybridization Two mycoplasma species were isolated from vaginal secretions obtained from pregnant women . Growth on solid A7B plates demonstrated the typical 'fried egg' appearance of Mycoplasma s p . b y low-power microscopy for one of the isolates . The other produced the typical, tiny, dark brown, granular colonies of Ureaplasma sp . On this basis and from the source of the isolates, the former was presumptively identified as M. hominis and the latter as U . urealyticum. These organisms were grown in SP-4 media that was modified for their growth requirements . For growth of M. hominis, the pH of the media was adjusted down to 7 . 0 . Growth of U. urealyticum was accomplished by adjustment of pH down to 7-0 by adding 10% urea, 4 ml/I . Plating on solid phase media was done to establish the purity of the pool of organisms used in subsequent work . Whole DNA was extracted from these organisms using the method described above for M . genitalium, and the DNA concentrations of these extracts and of an aliquot of M . genitalium were determined by measurement of u .v . absorption . The DNA from all three species was denatured by boiling for 10 min followed by rapid cooling on ice . The solutions were placed onto nitrocellulose filter paper via the S&S slot-blot apparatus using 5 µg of DNA per slot . The nitrocellulose was first saturated in 20 x SSC, prior to addition of DNA to each slot . After placement onto the filters, DNA was fixed on the filter by heating for 2 h at 80 ° C . The filters were then prehybridized for 2 h at 41 ° C in annealing buffer (50% formamide, 3 x SSC, 0-2 gg/ml yeast RNA, 0-05 mg/ml single-stranded salmon sperm DNA, 1 x Denhart's solution, 0-05 M Hepes) . Filters were hybridized in fresh annealing buffer with 32 P-labelled probe at 10 6 counts per sample . Slot blots of intact M. genitalium were done by passing 0. 5 ml volumes of freshly cultured organisms through nitrocellulose . Serial 1 :2 and 1 :10 dilutions were performed in fresh SP-4 medium . After passage onto the filter, cell lysis and DNA denaturation was performed by overlaying the filter onto Whatman 3 mm paper saturated with denaturation buffer (0 . 5 M NaOH, 1 . 5 M NaCl) . The neutralization step was done twice . The filters were then baked, prehybridized and hybridized as described above . An aliquot of the serial dilutions was placed at 35°C and checked daily for colour changes indicative of growth . The last dilution to show colour change was assumed to have contained one CCU . An aliquot of each sample was further plated onto SP-4 agar, and colonies were counted to correlate their numbers with CCUs .

RESULTS Ten randomly selected, radiolabelled cDNAs from M. genitalium DNA clones were hybridized with DNA from M. genitalium, M. hominis and U . urealyticum . Of these, three seemed to hybridize specifically with M . genitalium . One was noted to be darker than the others, indicating a greater degree of binding, and was chosen for



DNA probe to detect Mycoplasma genitalium

I

2

331

3

Fig. 1 .

Application of M. genitalium DNA probe from (1) Ureaplasma urealyticum, (2) Mycoplasma hominis and (3) Mycoplasma genitalium . Slot blot onto nitrocellulose with overnight exposure . 11 P label .

the subsequent studies (Fig . 1) . This fragment showed a faint band against an unquantified aliquot of U . urealyticum . When serial dilutions of known quantities of intact U . urealyticum were used in the assay, no hybridization was seen against the highest concentration, which corresponded to between 10 5 and 106 CCUs . The sensitivity of the cDNA fragment was assessed by applying it first against serial dilutions of intact organisms . A band of hybridization was seen at 10 CCUs . This corresponded to 100 organisms by colony count . Further determination of sensitivity was made by hybridization against serial dilutions of extracted DNA from M . genitalium . In the first experiment, a band was detected at 780 pg of DNA, and in the second at 100 pg (Fig . 2) . Given a genome mass of 5 X 108 Da,' or 5 X 10 8 g/mole of the intact organism, we calculated that the fragment used in the slot-blot system is capable of detecting about 1 . 2 X 104 organisms . Specificity and clinical utility were addressed subsequently . Hybridization was attempted with a variety of micro-organisms likely to be found in the genital tract .

100 ng 1 50 ng -> 25ng

30

12 . 5 ng -~ 6 .25 ng 3 . 13 ng 1 .56 ng --> 0 .78 ng - . 0 . 39 ng -~

Fig. 2. Hybridization against serial dilutions of DNA from Mycoplasma genitalium with a band detected at 0 . 78 X 10 - 'g .

332

G . F . Risi et al .

F- Mycop/osma genitalium E- Ureop/osma E- Trichomons vaginals

EEEEE-

Lactobacilus sp . Veillonella parvu/a Bacteroides dis/ens Bacteroides /ntermedius Bacteroides bivius

E- Bacteroides fragilis

E- Neisseria gonorrhoea E- Gardnerella vaginalls E- Staphylococcus aureus E- Gandido alb/cans E- Corynebacfer/um

EEE-E--

Proteus mirabillls Streptococcus faeco//s Escherichia co/i K/ebsiella pneumoniae

E- Group B streptococcus

Fig . 3 . Hybridization against organisms likely to be encountered in the genital tract of humans, with bands against Mycoplasma genitalium, Ureaplasma urealyticum and Eschicheria coli.

The organisms tested included E . coli, group B streptococcus, Klebsiella pneumoniae, Streptococcus faecalis, Proteus mirabilis, Corynebacterium sp ., Staphylococcus aureus, Gardnerella vaginalis, Neisseria gonorrhoeae, Bacteroides Ira gills, B . bivius, B . disiens, B. intermedius, Veillonella parvula, Lactobacillus sp ., Candida albicans and Trichomonas vaginalis . Organisms were grown in broth culture until turbidity was detected or were grown on solid medium and colonies resuspended in broth . A 0 . 2 ml aliquot was fixed to nitrocellulase paper by the slot-blot method . A minor band of hybridization was detected with E . coli but with no other organisms (Fig . 3) . To simulate testing of clinical specimens, we hybridized the fragment against known concentrations of M . genitalium that had been diluted in vaginal secretions . A pool of vaginal secretions was made by combining 2 ml aliquots of saline washes from 100 consecutive, asymptomatic, pregnant women attending a routine prenatal clinic . Serial dilutions of M . genitalium were made in SP-4 (1 :2 and 1 :10), and then the dilutions were mixed 1 :1 with the pooled vaginal secretions ; hybridization was done as described above . No cross-reactivity or diminution of sensitivity was observed (Fig . 4) .

DISCUSSION Despite decades of careful study, the cause(s) of two major sexually transmitted disease syndromes has not been completely resolved . Recognized pathogens for



DNA probe to detect Mycoplasma genitalium

3 33

12 .5 ng -~ 625 ng -a 3 , 13 ng 1 . 56 ng -~ 0 . 78 ng --

Fig . 4 . Hybridization against Mycoplasma genitalium added to pooled vaginal secretions from asymptomatic, pregnant women showing no diminution of probe sensitivity (for comparison see Fig . 2) .

.', is PID include N . gonorrhoeae, C. trachomatis, M . hominis and anaerobic bacteria A large proportion of women with clinical PID, however, do not show cultural or serological evidence of infection with any recognized pathogen . Estimates are that this category encompasses 20-30% of all cases . Similarly, in men with NGU, C . trachomatis and U. urealyticum are the recognized pathogens, but these account for only about 70% of cases .' Studies in our population in New Orleans suggest that the proportion of cases of NGU without either of these organisms may be even higher . These statistics have stimulated the continued search for new pathogens by the application of novel technologies and media to clinical material . Such was the case when SP-4 medium was used to culture the urethral discharge of men with NGU .12 This medium was developed in the course of attempts to cultivate a new tick14 derived, pathogenic spiroplasma, the suckling mouse cataract agent . In preliminary studies, Tully et al. found it supported the growth of a number of fastidious mycoplasmas of human and animal origin, such as M . pneumoniae, M . alvi and M . hyopneumoniae . 11 Of 13 men in the first series, two were found to be infected with a previously unrecognized mycoplasma . Serologically, the organism was similar to M . pneumoniae 2 ' 3 but distinct from genital tract strains . 12 Electron microscopy showed that its structure was similar to recognized pathogens such as M . pneumoniae . Because of the technical difficulties encountered in culturing this organism, we have developed a nucleic acid probe that should be sensitive enough to detect the organism in clinical specimens and specific enough to discriminate among the various strains of genital tract mycoplasmas . Some aspects of our results deserve further comment . The sensitivity of the probe differed substantially between intact organisms and extracted DNA, with a 100-fold greater sensitivity against the intact organism . Several possibilities exist to explain the sensitivity differences . Our probe may crosshybridize with RNA present in the intact cell . Because many more copies of RNA exist per cell, the probe would be more sensitive to the intact cell than to purified DNA



334

G . F. Risi et al ,

which organisms, although not growing, would still bind the probe . Finally, the SP-4 medium may not be able to support the growth of all organisms transferred to it . These possibilities are all currently under investigation . A faint band of cross-hybridization between our probe and U . urealyticum was seen when large concentrations of ureaplasma were fixed to nitrocellulose . When attempts were made to measure the amount required for detection, no hybridization was seen at 106 CCU, the highest concentration tested . Slight cross-hybridization with E . coli was also detected . This almost certainly results from our selection of M-13 as the vector . Polymerization past the probe onto M-13, which contains the E. toll Lac operon, would result in some hybridization with E. coli DNA . Common DNA sequences have been described between M . genitalium and M . pneumoniae, 16 so the possibility exists that our probe might cross-hybridize with M . pneumoniae . We are presently investigating this possibility, but even if a minor degree of cross-hybridization does occur this would not detract from the potential usefulness of our probe, since M . pneumoniae is not recognized as a genital tract pathogen . Sequencing of the probe has been accomplished, and it is 256 base-pairs in length . By the technique of synthetic oligonucleotide formation, smaller pieces of this probe can be made . The smaller the length, the less cross-reactivity with related organisms will be found, and the cross-reactivity with E. coil would be expected to disappear. Nevertheless, the probe we now have can be applied to begin to define the ecology of M . genitalium in our patient population .

ACKNOWLEDGEMENTS We are grateful to J . H . Tully of the National Institutes of Health, Bethesda, Maryland, for providing us with the G-37 strain of Mycoplasma genitalium .

REFERENCES 1 . Bowie, W . R. (1984) . Urethritis in males . In Sexually Transmitted Diseases . (Holmes, K . K . et al., eds) pp . 638-650 . New York: McGraw-Hill . 2 . Clyde, W . A . & Hu, P . C . (1986) . Antigenic determinants of the attachment protein of Mycoplasma pneumoniae shared by other pathogenic mycoplasma species . Infection and Immunity 51, 690 . 3 . Lind, K ., Lindhardt, B. O ., Schutten, H . J ., Bloom, J . & Christiansen, C . (1984) . Serological crossreactions between Mycoplasma genitalium and Mycoplasma pneumoniae . Journal of Clinical Microbiology 20, 1036 . 4 . Messing, J . (1983) . New M13 vectors for cloning . Methods in Enzymology 101, 20. 5 . Moller, B . R ., Taylor-Robinson, D . & Furr, P . (1984). Serological evidence implicating Mycoplasma genitalium in pelvic inflammatory disease . Lancet May 19, 1102 . 6 . Moller, B . R ., Taylor-Robinson, D ., Furr, P . M . & Freundt, E . A. (1985) . Acute upper genital tract disease in female monkeys provoked experimentally by Mycoplasma genitalium . British journal of Experimental Pathology 66, 417 . 7 . Razin, S . (1985) . Molecular biology and genetics of mycoplasmas (mollicultes) . Microbiological Reviews 49, 419 . 8. Taylor-Robinson, D . & Furr, P . M. (1982). The pathogenicity of a newly discovered human mycoplasma (strain G-37) for the genital tract of marmosets . Journal of Hygiene (Cambridge) 89, 449 .



DNA probe to detect Mycoplasma genitalium

335

9 . Taylor-Robinson, D ., Furr, P. & Hanna, N . F. (1985) . Microbiological and serological study of nongonococcal urethritis with special reference to Mycoplasma genitalium . Genitourin Medicine 61,319 . 10 . Taylor-Robinson, D ., Tully, J. G . & Barile, M . F . (1985) . Urethral infection in male chimpanzees produced experimentally by Mycoplasm a genitalium . British Journal of Experimental Pathology 66, 95 . 11 . Tully, J . G ., Rose, D . L ., Whitcomb, R . E . & Wenzel, R . P . (1979) . Enhanced isolation of Mycoplasma pneumoniae from throat washings with a newly modified culture medium . Journal of Infectious Diseases 139, 478. 12 . Tully, J . G., Taylor-Robinson, D ., Cole, R . M . & Rose, D . L . (1981) . A newly discovered mycoplasma in the human urogenital tract . Lancet June 13, 1288 . 13 . Tully, J . G ., Taylor-Robinson, D ., Rose, D . L ., Furr, P . M ., Graham, C . E . & Barile, M . F . (1986). Urogenital challenge of primate species with Mycoplasma genitalium and characteristics of infection induced in chimpanzees . Journal of Infectious Diseases 153, 1046 . 14 . Tully, J . G ., Whitcomb, R . E ., Clark, H . E . & Williamson, D . L . (1977). Pathogenic mycoplasmas : cultivation and vertebrate pathogenicity of a new spiroplasma . Science 195, 892 . 15 . Westrom, L . & Mardh, P . (1984) . Salpingitis . In Sexually Transmitted Diseases (Holmes, K . K . et al ., eds) pp . 615-32 . New York : McGraw-Hill . 16 . Hyman, H . C ., Yoger, D . & Razin, S . (1987). DNA probes for detection and identification of Mycoplasma pneumoniae and Mycoplasma genitalium . Journal of Clinical Microbiology 25,726-8 .