Feb 27, 1991 - AND MICHAEL F. BARILE. Laboratory ofMycoplasma ..... Olson, L. D., S. W. Shane, A. A. Karpas, T. M. Cunningham,. P. S. Probst, and M. F. ...
INFECTION AND IMMUNITY, Sept. 1991, p. 3327-3329 0019-9567/91/093327-03$02.00/0 Copyright C 1991, American Society for Microbiology
Vol. 59, No. 9
Successive Synovial Mycoplasma hominis Isolates Exhibit Apparent Antigenic Variation LYN D. OLSON,* CAROLYN A. RENSHAW, STEPHANIE W. SHANE, AND MICHAEL F. BARILE Laboratory of Mycoplasma, Center for Biologics Evaluation and Research, Food and Drug Administration, Bethesda, Maryland 20892 Received 27 February 1991/Accepted 27 June 1991
The expression of surface proteins by 14 successive Mycoplasma hominis isolates obtained from the synovial fluid of a chronically infected septic arthritis patient was examined. Marked differences in the expression of surface proteins, as determined by monoclonal antibodies raised against the first isolate, were observed. However, identical restriction patterns and virtually identical hybridization patterns with probes containing the conserved genes of the Mycoplasma capricolum rRNA operon and the Escherichia coli elongation factor Tu suggest that the protein differences might reflect antigenic variation by M. hominis during infection. Sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) of whole-cell lysates was performed by the method of Laemmli (7) with 10% gels. Each well was loaded with 10 pug of protein. Immunoblotting was performed as previously described (13). DNA was isolated by using standard procedures except that proteolysis was performed first and RNase treatment followed the SDS precipitation with potassium acetate. The Coomassie-stained SDS-polyacrylamide gels of the 14 patient isolates are shown in Fig. 1. Although the overall protein profiles of the M. hominis isolates are very similar, distinct differences from lane to lane can be seen. Some protein bands appear to be more expressed in some strains than in others, while other protein bands are present in only some of the strains. Immunoblots of these proteins reacted with MAb-G5, MAb-H3, and MAb-A8 (Fig. 2) revealed avid binding of all three MAbs to similarly migrating proteins from the early isolates, followed by diminished expression or loss of reactive epitopes, and then progressive expression again of reactive protein bands for all three MAbs. It is of particular interest that following the administration of intravenous
Mycoplasma hominis is a prokaryotic human parasite that has been associated with a variety of clinical conditions. The mycoplasma has been isolated from both healthy tissues and from patients with clinically diverse diseases, including pelvic inflammatory disease, polynephritis, septicemia, and arthritis, as well as meningitis in infants less than two weeks old (12). Strains of M. hominis have been shown to vary genomically (2, 4-6) and serologically (8-11). No common genomic or antigenic feature that correlates with the pathogenicity of strains or the tissues that are colonized has been found. In this report, we describe the initiation of the molecular characterization of 14 M. hominis isolates from a chronic septic arthritis patient (14). These isolates, which were obtained from the synovial fluid surrounding a knee prosthesis during each exacerbation of the disease over a 6-year period (1984 to 1989), appear to vary antigenically. Analysis of M. hominis strains obtained from a single patient might provide information regarding the mechanism(s) used by this pathogen to vary antigenically within a host over time. An understanding of such a mechanism(s) might also provide information on how the mycoplasma is capable of evading the host immune system and establishing colonization, thereby facilitating the development of more effective treatment or prophylaxis. Table 1 lists the strain designations of the patient isolates, dates of isolation, and antibiotic therapy employed at the time of isolation. All strains (passage 2 to 5) were grown in modified arginine-containing Edward's media as previously described (3), harvested by centrifugation, washed twice with phosphate-buffered saline (PBS), and stored at -40°C. We have previously characterized three monoclonal antibodies (MAbs), MAb-A8, MAb-G5, and MAb-H3, which were raised against the first patient isolate, M. hominis 1620 (13). These MAbs bind to many surface-exposed integral membrane proteins; the number of bound proteins which are expressed varies from strain to strain from different patients, regardless of the site of isolation. The MAbs were used in this study to probe immunoblots for comparison of surface epitope expression of the M. hominis isolates from the septic arthritis patient. *
TABLE 1. M. hominis strains isolated from a single chronic septic arthritis patient Strain'
a, b, c, d, e,
1620 1621 1623 1628 1630
Date of isolation
(mo/day/yr) 2/24/84 3/2/84 3/6/84 3/13/84 3/22/84
f, 1635
4/4/85
g, 1656 h, 2101
9/10/86 8/4/88
i, 2111 j, 2116 k, 2119 1, 2125 m, 2126 n, 2127
11/17/88 11/30/88 12/8/88 4/20/89 5/18/89 7/6/89
a The letter designations
Corresponding author.
3327
Patient drug therapy
Keflex, Bactrim Bactrim Intravenous tetracycline Doxycycline
Ciprofloxacin Ciprofloxacin; irregular intake over several weeks Ciprofloxacin Ciprofloxacin Ciprofloxacin Ciprofloxacin Ciprofloxacin Ciprofloxacin
correspond to lanes in the figures.
NOTES
3328 a
b
c
d e f gh ij
I4 I
tww klIm n
r _!1 __ w_I
w
1z
INFECT. IMMUN.
a
abcdef
94-. 84-
ghi j k Imn
_
62-v
-..
1
I..
~ J_ FIG. 1. Coomassie blue-stained polyacrylamide gel of sequential M. hominis patient isolates shows occasional differences in expression of protein bands. Lanes a through n correspond to letter designations of isolates given in Table 1.
tetracycline and oral doxycycline there was marked suppression of all reactive epitopes, even though the patient's condition was clinically poor. In spite of the marked differences observed among the surface-exposed proteins of the successive isolates, we presume that the patient is chronically infected with M. hominis, rather than that the patient is repeatedly reinfected with different strains of M. hominis. We analyzed the DNAs of the 14 strains to examine this assumption. A variety of restriction enzymes for cleavage of the DNAs, including Ncol, BamHI, PstI, MspI, HpaII, and others, produced uniformly indistinguishable patterns from all 14 isolates. Band differences were observed in the DNA restriction patterns from the reference strain PG25. An example of this observation is presented in Fig. 3, which shows the patterns obtained with the enzyme Ncol. Since many different strains of M. hominis have been shown to exhibit similar but very distinct restriction patterns (2), illustrating the frequently extreme differences in DNA-DNA homology among members of this species, the identical restriction patterns among all the patient isolates support the hypothesis of chronic infection of the patient by a single strain. The cleaved DNAs were hybridized with the plasmid pMC5, which contains most of the rRNA operon from Mycoplasma capricolum (1), and plasmid pEF-Tu, which contains the Escherichia coli gene for elongation factor Tu (16). It has been shown that a finite number of different hybridization patterns with these probes are frequently observed among strains of a species (15). For example, the pMC5 hybridization patterns of three M. hominis strains cleaved with EcoRI have been shown to each produce a unique pattern (15). If the patient is being reinfected with various strains of M. hominis, one would therefore expect at least some of the hybridization patterns to differ from the others. Hybridization of pEF-Tu with EcoRI and HindlIl digests of the patient isolate DNAs produced identical patterns, as did the EcoRI digests hybridized with pMC5 (data not shown). These results argue in favor of the patient being persistently colonized by a single strain of M. hominis. Interestingly, however, pMC5 hybridized with HindIII digests of the patient isolate DNAs also produced identical
b
abcde f
gh i j k fmn
94--
8462-2
c
a
b
c
de
fg h i
j
k I
mn
94-84-
62--
116p0
FIG. 2. Immunoblots of sequential M. hominis patient isolates reacted with MAbs specific for surface-exposed integral membrane proteins. (a) MAb-H3; (b) MAb-G5; (c) MAb-A8. Lanes a through n correspond to letter designations of isolates given in Table 1. Numbers indicate molecular mass in kilodaltons.
patterns with all DNAs, except for two successive isolates,
strains 1630 and 1635 (Fig. 4, lanes e and f). The DNAs from these isolates were observed to possess an additional band that was not present in the M. hominis strains obtained before and after these two isolates. (DNA from the reference strain PG25 had a middle band that migrated faster than those of any of the patient isolates [data not shown].) This might suggest that some change at the DNA level, even as small as a single-base alteration, occurred and was subsequently reversed. Another possibility is that a cryptic plasmid is expressed in the presence of tetracycline and tetracycline derivatives. More precise studies of the DNA at the individual gene level are being performed. This report is the first analysis of serial M. hominis isolates, that is, multiple isolates from a single patient. The results demonstrate that antigenic variation of M. hominis in situ does occur, perhaps in response to the patient's immune system and/or drug therapy. The mechanism(s) governing this variability cannot be determined from these data. In summary, we have screened the surface-expressed
NOTES
VOL. 59, 1991
a
b
c
d
e
f
g
h i j k I
mn
o
3329
operon. The apparent antigenic variation is under further study.
kb
23.1 9.4 6.6 4.4
I.
I I,. s
1.
~~~~~~~~~~~~~~101
Ll--i101-4-1
2.
J
]
2.3 2.0
3. 4.
FIG. 3. Restriction cleavage patterns of sequential M. hominis patient isolates cut with NcoI. Lanes a through n correspond to letter designations of isolates given in Table 1; lane o contains the reference strain PG25. Arrows mark some of the band differences between the restriction patterns of the reference strain and the patient isolates.
5.
6.
proteins and DNAs of successive M. hominis isolates from a chronically infected septic arthritis patient. The expression by the M. hominis isolates of surface-exposed proteins bound by the three MAbs varied significantly throughout the 6-year period. In contrast, the DNAs have apparently identical restriction patterns, although slight differences were noted in the HindIII hybridization patterns obtained with pMC5, which contains most of the highly conserved rRNA
7.
8.
9. 10.
a
b
c
d
e
f
g
h 11.
ww
.1
lowllw
12.
13.
40 14.
15. FIG. 4. Hybridization patterns of pMC5, which contains most of the rRNA operon from M. capricolum, with HindIlI-cleaved DNA from patient isolates. Lanes a through h correspond to letter designations of isolates given in Table 1.
16.
REFERENCES Amikam, D., S. Razin, and G. Glaser. 1982. Ribosomal RNA genes in mycoplasma. Nucleic Acids Res. 10:4215-4222. Barile, M. F., M. W. Grabowski, E. B. Stephens, S. J. O'Brien, J. M. Simonson, K. Izumikawa, D. K. F. Chandler, D. TaylorRobinson, and J. G. Tully. 1983. Mycoplasma hominis-tissue cell interactions: a review with new observations on phenotypic and genotypic properties. Sex. Transm. Dis. 10(Suppl.):345354. Chandler, D. K. F., S. Razin, E. B. Stephens, R. Harasawa, and M. F. Barile. 1982. Genomic and phenotypic analyses of Mycoplasma pneumoniae strains. Infect. Immun. 38:604-609. Christiansen, C., J. Frydenberg, G. Christiansen, H. Andersen, and L. Hedegaard. 1984. Analyses of the mycoplasma genome with recombinant DNA technology. Isr. J. Med. Sci. 20:781784. Christiansen, G., and H. Andersen. 1988. Heterogeneity among Mycoplasma hominis strains as detected by probes containing parts of ribosomal ribonucleic acid genes. Int. J. System. Bact. 38:108-115. Christiansen, G., H. Andersen, S. Birkelund, and E. A. Freundt. 1987. Genomic and gene variation in Mycoplasma hominis strains. Isr. J. Med. Sci. 23:595-602. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227:680-685. Lin, J. S. 1980. An antigenic analysis for membranes of Mycoplasma hominis by cross-absorption. J. Gen. Microbiol. 116: 187-193. Lin, J. S., S. Alpert, and K. M. Radney. 1975. Combined type-specific antisera in the identification of Mycoplasma hominis. J. Infect. Dis. 131:727-730. Lin, J. S., and E. H. Kass. 1974. Serological reactions of Mycoplasma hominis: differences among mycoplasmicidal, metabolic, inhibition, and growth agglutination tests. Infect. Immun. 10:535-540. Lin, J. S., and E. H. Kass. 1974. Serological typing of human genital mycoplasmas by a complement-dependent mycoplasmicidal test. J. Infect. Dis. 126:658-663. Mardh, P., B. R. MoUer, and W. M. McCormack (ed.). 1983. M. hominis: a human pathogen. Sex. Transm. Dis. 10(Suppl.):4. Olson, L. D., S. W. Shane, A. A. Karpas, T. M. Cunningham, P. S. Probst, and M. F. Barile. 1991. Monoclonal antibodies to surface antigens of a pathogenic Mycoplasma hominis strain. Infect. Immun. 59:1683-1689. Snell, M., F. Wellborne, M. F. Barile, and P. Plotz. 1986. Prosthetic joint infection with Mycoplasma hominis. J. Infect. Dis. 153:174-175. Yogev, D., D. Halachmi, G. E. Kenny, and S. Razin. 1988. Distinction of species and strains of mycoplasmas (Mollicutes) by genomic DNA fingerprints with an rRNA gene probe. J. Clin. Microbiol. 26:1198-1201. Yogev, D., S. Sela, H. Bercovier, and S. Razin. 1988. Elongation factor (EF-Tu) gene probe detects polymorphism in Mycoplasma strains. FEMS Microbiol. Lett. 50:145-149.
