Genotypic and Phenotypic Analysis of Mycoplasma fermentans Strains Isolated from Different Host Tissues

Abstract

A correlation was found between the expression of a specific Mycoplasma fermentans surface antigen (Pra, proteinase-resistant antigen) and the site of isolation of the organism from the infected host. Strains which expressed Pra were most frequently associated with cells of bone marrow origin, and strains which lacked expression of Pra were most commonly isolated from the respiratory tract, genital tract, and arthritic joints, i.e., epithelial cell surfaces. Pra was previously shown to be resistant to degradation by proteinases and was hypothesized to play a protective role at the organism surface and perhaps to influence which host tissue site was colonized by the organism. The methods used for this phenotyping scheme required isolation and growth of the mycoplasma in quantities sufficient for immunoblot analysis using monoclonal antibodies. We wanted to determine a more rapid and less cumbersome technique to supplement this method for determining the Pra phenotype directly in clinical specimens. Here we describe PCR studies to investigate the movement of a previously identified M. fermentans insertion sequence (IS)-like element. These data showed a correlation between a specific IS genotype and the Pra⁺ phenotype. Production of a 160-bp product using a single set of IS-based primers was associated with expression of Pra. The genomic IS location resulting in the 160-bp product was determined by using Southern blot analysis and was found to be a stable insertion site characteristic of genotype I strains. Additional analyses of sequences within and flanking the IS insertion sites revealed another pair of PCR primer sites which resulted in the consistent production of a 450-bp amplicon. The stability of this site was dependent on the absence of the IS-like element between the primer sites. The production of this 450-bp amplicon correlated with the Pra mutant phenotype and was characteristic of genotype II strains. The data showed that the sequence within the IS may be unstable and that reliable genotyping sequences are more easily found in the stable genomic sites which flank the IS element.

First mistakenly identified as a novel AIDS-associated virus (18), Mycoplasma fermentans incognitus, during the ensuing years, was considered to be a possible cofactor contributing to acceleration of the progression of this immune disorder (8, 16, 20–22, 29). Immediately following the first reports, several laboratories began probing into this question, but to date, the hypothesis of a mycoplasma-AIDS association remains unproved. However, these studies have added much to our basic knowledge of mycoplasmas. It has been documented that M. fermentans, as well as some other mycoplasmas, can occur intracellularly, which was only an occasionally reported and unproved observation prior to these studies. The ability of specific subpopulations of these organisms to survive within host cells could account, at least in part, for the characteristic chronicity of mycoplasmal disease, as well as for the frequent difficulty of isolation by culture. Additionally, an impressive volume of literature is accumulating which describes the induction of various cytokines by mycoplasma infection (3, 5, 15, 27, 28, 30, 37). The potential to alternately stimulate or suppress the immune system would impart a distinct advantage to any pathogen (or commensal organism) attempting to survive in the hostile and changing environment of an infected host.

Subsequent to the initial isolation of strain incognitus, M. fermentans was identified as the likely etiologic agent of an acute fatal disease in otherwise healthy adults (17). No other infectious agents were found. A similar wasting syndrome leading to death was reported in silvered leaf monkeys after experimental infection with this same agent (19). Many years prior to these recent studies, M. fermentans was isolated from bone marrow of leukemic patients (24) and other reports associated it with rheumatoid arthritis (2, 36). These reports prompted further investigations, including some experimental studies with animal models (9, 10, 26). None of these studies resulted in data proving a cause-and-effect relationship between M. fermentans infection and human disease. In fact, early serologic studies provided evidence that antibodies to this organism are common in adolescents and young adults (32).

Therefore, M. fermentans has been tentatively associated with disease throughout its history but the precise etiologic role of M. fermentans in disease remains unclear. This is, in part, due to the frequently unsuccessful attempts to isolate mycoplasmas in general by routine culture methods (6) and to the presence of individuals harboring the organism without signs of disease. Even though many cases have resulted in isolation of M. fermentans and each isolate has been assigned a new strain designation, there has been no attempt to assign molecular or functional characteristics to these strains which might assist in determining if there is a characteristic or group of characteristics which associate with specific diseases, or at least with sites of isolation.

In the present study, we were interested in defining methods to determine if specific strains exhibit characteristics which are more frequently associated with particular tissue sites within an infected host. We tested whether monoclonal antibodies (MAbs) developed against M. fermentans antigens could distinguish between isolates of M. fermentans to determine a possible correlation between the expression of these factors and the site of isolation. We also conducted the same correlative assessment for the chromosomal distribution of the M. fermentans insertion sequence (IS)-like element, hypothesizing a role for this potentially mobile element in the repression or activation of a specific gene expression.

MATERIALS AND METHODS

Sources of isolation.

The M. fermentans strains evaluated in this study were isolated from various sources (see Table 1). Strains were obtained as follows: E10 (24) and K7 (25) were obtained from W. H. Murphy; 16700, 12406, and DEPB were from the University of Alabama at Birmingham; AOU was from Luc Montagnier (Pasteur Institute, Paris, France); Z62 was from P. Hannan (Beecham Labs) (24); incognitus was from Shyh Lo (National Institute of Allergy and Infectious Diseases [NIAID]) (17, 18, 21); AMSO was from Ann Robinson (Laboratory of Immune Genetics, NIAID); MT2 was from W. J. Leonard and N. F. Halden (National Institute of Child Health and Human Development) (11); Elliman was from H. Elliman (University of Illinois, Chicago); 48429 was from Andy Lewis (NIAID); M51, M39, M52, M64, M73, and M70 were from R. Dular (Public Health Laboratory, Ottawa, Ontario, Canada); KL4 and KL8 were from P. Hannan (Beecham Labs); and PG18 was from Klieneberger-Nobel, Lister Institute, London, United Kingdom.

TABLE 1.

Relationships among the genotypes, phenotypes, and sources of isolation of different M. fermentans strains

Straina	PCR product size (bp)b	Genotype determined by Southern blottingc	Immuno- blotting resultd	Specimen origin
M51	Diffuse	II	−	Respiratory tract
M39	80	II	−	Respiratory tract
M52	Diffuse	II	−	Respiratory tract
M64	226	II	−	Respiratory tract
M73	Diffuse	II	−	Respiratory tract
M70	<80	II	−	Respiratory tract
KL8	>860	II	−	Rheumatoid arthritis, joint
KL4	>860	II	−	Rheumatoid arthritis, joint
PG18	Diffuse	II	−	Human genital tract
12406	>1,100	II	−	Human genital tract
E10	160	I	+	Leukemic patient, blood
16700	160	I	+	Human genital tract
AOU	160	I	+	AIDS patient, blood
Z62	160	I	+	Leukemic patient, blood
Incognitus	160	I	+	Kaposi’s sarcoma
AMSO	160	I	+	Lymphocyte culture
MT2	160	I	+	Human lymphocyte
Elliman	160	I	+	Hybridoma
DEPB	160	I	+	AIDS patient, blood
48429	160	I	+	Cell culture
K7	160	I	+	Leukemic patient, blood

^aSee Materials and Methods for descriptions. 

^bPCR was performed with the RS primers shown in Fig. 4. 

^cSouthern blotting was performed with ORF-1 and ORF-2 probes. 

^dImmunoblotting was performed with MAb 1A2.6. −, negative; +, positive. 

Organisms and growth conditions.

Cultures of M. fermentans were grown in SP-4 medium (mycoplasma broth base, tryptone [Difco], peptone [Difco], arginine, phenol red [1%], DNA, and antibiotics for SP-4, supplemented with 10% fetal bovine serum, CMRL 1066, yeast extract, yeastolate, and glucose). The cultures were incubated at 37°C. Samples were harvested and washed in phosphate-buffered saline at pH 7.3. DNA was purified in accordance with standard protocols (phenol-chloroform-isoamyl alcohol) and concentrated by ethanol precipitation. DNA preparations were RNase A treated (Sigma).

Southern blotting and DNA hybridization.

For Southern blotting of the M. fermentans strains, 0.2 μg of genomic DNA was digested with 5 U of HindIII (Promega, Madison, Wis.) for 2 h at 37°C. The samples were electrophoresed on 0.8% Tris-borate-EDTA agarose gels (50 V for 16 h) and transferred to 1× Hybond N⁺ nylon membranes. DNA was UV cross-linked to the membrane in a UV Stratalinker 1800 (Stratagene). After prehybridization, the membranes were hybridized with 5′-end γ-³²P-labeled oligonucleotide probe RW006 (5′-GCT GTG GCC ATT CTC TTC TAC GTT-3′; see Fig. 3a) and probe ORF-1 (5′-GGA AAA CTC TTA TTC AGC C-3′; see Fig. 3b), located within the insertion sequence transposase gene and open reading frame 1 (ORF-1). Hybridization was performed at 42°C for 1 h in Rapid hyb (Amersham) hybridization buffer and followed by one washing in 5× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate)–0.1% sodium dodecyl sulfate (SDS) for 20 min at room temperature and two washings in 0.5× SSC–0.1% SDS for 15 min each at 45°C. DNA hybrids were visualized by autoradiography using Kodak XAR-5 film (Eastman Kodak Co., Rochester, N.Y.).

FIG. 3.

Southern blot analysis of the chromosomal distribution of the M. fermentans IS-like element. DNAs isolated from all of the strains were digested with HindIII, separated in 0.8% agarose, transferred to nylon membranes, and then probed. The Southern blots were hybridized with an oligonucleotide specific for ORF-1 (a), and the same strains were then hybridized with an oligonucleotide specific for the transposase gene (b). Approximate fragment sizes are indicated in kilobases. Differences between the two basic genotypes (I and II) and the identification of fragments common to the two panels are indicated by the numbered arrows (see text).

PCR.

After incubation and sufficient growth, the 21 strains were processed with proteinase K combined with buffer A (1 M Tris-HCl [pH 8.0], 1 M KCl, 1 M MgCl₂, Milli-Q distilled H₂O) and buffer B (1 M Tris-HCl [pH 8.0], 1 M MgCl₂, Triton X-100, Tween 20, Milli-Q distilled H₂O). One milliliter of culture was centrifuged for 20 min at 4°C, the supernatant was discarded, and the pellet was resuspended in proteinase K lysis buffer. The samples were incubated at 60°C for 1 h and then boiled for 10 min. Samples were incubated on ice for 10 min and then stored at −70°C until ready for PCR.

Amplification of the M. fermentans strains was performed by using four different primer pairs (see Fig. 2). Primers RS-47 and RS-49 and primers RW005 and RW004 were previously described by S.-C. Lo et al. (34); primers MF-1 (5′-GGA AAA CTC TTA TTC AGC C-3′) and MF-2 (5′-GGA AAA CTC TTA TTC AGC ATG C-3′) were synthesized by Keystone Laboratories. Amplification of DNA was performed in a total volume of 50 μl. Basically, PCR was performed with 40 cycles of denaturation (94°C, 25 to 30 s), annealing (60°C, 1 min), and extension (72°C, 1 min). Another primer, MF-4 (5′-GCG GCA CCA TCA ATC ACA TAT AC-3′), was used as the antisense primer along with the previously described RS-47 sense primer. For this primer pair, an initial denaturation at 94°C for 2 min was followed by 40 cycles as described above. PCR products were resolved on 2% agarose gels and visualized by ethidium bromide staining.

FIG. 2.

Schematic diagram of the M. fermentans IS-like element and its flanking regions. The locations of all of the primer pairs used in this study are indicated, as are the sizes, in base pairs, of the respective amplicons. ORF-2 is the putative transposase. ORF-1 and ORF-3 have no assigned putative functions.

Immunoblotting.

SDS-polyacrylamide gel electrophoresis and Western blot analysis were performed as described previously, by using a 10% resolving gel and a 4% stacking gel, and then proteins were separated and transferred to nitrocellulose (Bio-Rad) by the method of Towbin et al. (33). Immunological reactions were visualized with peroxidase-labeled conjugates (Sigma).

MAbs.

MAbs directed to M. fermentans incognitus antigens were produced in conjunction with the Hybridoma Core Facility of the Multipurpose Arthritis Center at the University of Alabama at Birmingham. The basic procedure for the production and characterization of MAbs has been described previously in detail (35).

RESULTS AND DISCUSSION

Phenotyping of M. fermentans strains and isolates.

The most frequently colonized sites in a mycoplasma-infected host are epithelial cell surfaces (7, 12, 14, 23, 31). In the case of M. fermentans, the second most frequent association is with blood cells (17, 21). Are there characteristics that make some species or some strains within a species uniquely qualified for survival in one site as opposed to another? We previously identified an M. fermentans surface antigen (Pra) that can be divided into two distinct domains based on the immunoblot pattern obtained with MAbs, i.e., a domain that is resistant to degradation by trypsin, chymotrypsin, V-8 protease, and proteinase K and a second domain that is sensitive to these same proteinases (40). Our preliminary studies suggest that Pra is a complex surface network consisting of acylated proteins, but the nature of the membrane anchor and the noncovalent forces that mediate the interaction between the two domains have not been fully characterized (39). Nonetheless, we hypothesized that a correlation exists between the expression of the proteinase-resistant domain, which may play a protective role at the organism surface, and the association of the organism with particular cell types. Results in Fig. 1 show the variable expression of the Pra⁺ phenotype and the distinctive, diffuse immunoblot pattern of the Pra⁺ phenotype which correlated with isolation of the organism associated with cells of bone marrow origin and frequently from immunocompromised patients (Table 1). This broad distribution of the electrophoretic mobility of identical epitopes seen for the Pra⁺ phenotype is not uncommon for mycoplasmal antigens (38). The single exception to the above correlation was isolate 16700, which was isolated from the urethra of a patient with nongonococcal urethritis. Organisms lacking expression of the proteinase-resistant phenotype were most commonly isolated from the respiratory tract, from the genital tract, and from arthritic joints (Fig. 1 and Table 1). These Pra mutants were presumably epithelial cell associated. If the Pra⁺ phenotype does, in fact, provide the organism with protection from proteolytic degradation, then the above correlations support the possibility, although they certainly do not prove, that the Pra⁺ phenotype resides in a hydrolase-rich intracellular compartment. This niche may be represented by the professional-phagocyte-rich cellular environment found in the circulation.

FIG. 1.

Pra phenotypes of representative M. fermentans strains. Organism proteins were separated by SDS-polyacrylamide gel electrophoresis and then transferred to nitrocellulose. Reactivity with MAb 1A2.6 was visualized by using peroxidase conjugates. Strains incognitus and MT2 show the characteristic Pra⁺ pattern, while the Pra⁻ strains show no reaction with the MAb. Results for all strains are summarized in Table 1.

Genomic distribution of the IS element.

Development of a statistically sound proof of the association suggested above requires a large number of isolates characterized with respect to Pra expression and sites of isolation. Therefore, if a genetic marker also existed (preferably associated with Pra in a nondissociable genotype-phenotype relationship) which correlated with the site of isolation, then it might be possible to develop a more sensitive and simple system for discerning the Pra phenotype of new isolates. We initially looked for a correlation between Pra expression and the genomic location of the previously published M. fermentans IS-like element. Disruption or activation of cryptic promoters is not an uncommon result of IS movement (1, 4). Although it is unproved, M. fermentans may use this system to accommodate its surface properties for survival in its current environment (13). Figure 2 is a schematic showing the basic structure of the previously described IS-like element, as well as the locations of the genotyping PCR primers and probes used in the current study. Figure 3 shows the distribution of IS-associated ORF-1 and ORF-2 (transposase gene) in HindIII genomic digests of the different M. fermentans strains. These restriction patterns are consistent with those previously shown for strain incognitus (13). Based on the distribution of these two ORFs, the M. fermentans strains can be grouped into two basic genotypes (I and II). The published sequence of the incognitus strain IS-like element contains no HindIII site (13). If the IS-like elements in the other M. fermentans strains are sufficiently similar to the IS-like element of strain incognitus and are intact, then ORF-1- and ORF-2-specific probes should cohybridize to the same HindIII restriction fragments. Examination of Fig. 3 by using this criterion (i.e., identification of fragments common to Fig. 3a and b) indicates that genotype I has at least 10 copies of the intact IS and genotype II has only 3. These are minimal estimates, since a single fragment could contain multiple probe sites. Identification of fragments that are not common to Fig. 3a and b indicate that (i) all genotype I strains have one copy of ORF-1 which is not associated with an IS (Fig. 3a, location 1); (ii) two of the genotype I strains, incognitus and 16700, have one additional non-IS-associated ORF-1 (Fig. 3a, location 2); (iii) 7 of the 10 genotype II strains have one non-IS-associated ORF-1 (Fig. 3a, location 3); and (iv) 6 of the latter 7 genotype II strains also have a non-IS-associated ORF-2 (Fig. 3b, location 4). The apparent non-IS-associated ORFs in these experiments may be the result of a common mutational event producing a HindIII site between ORF-1 and ORF-2. Strains KL4 and KL8 are unusual, since no IS elements are detectable in Fig. 3. Even though the latter two strains may actually constitute a third genotype, here we have placed them into genotype II based on additional parameters to be discussed below.

Table 1 shows genotypes I and II in the context of Pra expression and site of isolation. All strains in genotype I were Pra⁺, and all genotype II strains were Pra mutants.

Genotyping of M. fermentans strains by PCR.

The pattern of distribution of the IS-like element genomic insertion sites was a reliable reflection of Pra expression but required very cumbersome methodology. The asymmetric hybridization of the ORF-1 and ORF-2 probes seen in Fig. 3 indicated the presence of some differences within the IS-like element which may provide useful sequence markers for determining genotypes by using PCR. An M. fermentans-specific PCR primer pair has been previously described (34). Amplification with this primer pair, located within ORF-2 of the IS-like element (Fig. 2, RW005 and RW004), resulted in a 206-bp amplicon for the strains evaluated in that study. In the current study, amplification with these same primers produced a 206-bp amplicon for all of the strains listed in Table 1 (Fig. 4a). Although these primers performed as predicted, they do not allow discrimination of the two genotypes. Strains KL4 and KL8 showed no indication for the presence of either ORF-1 or ORF-2 (Fig. 3), even though a typical 206-bp product was obtained with the RW005-RW004 primer pair (Fig. 4a). This suggested that even though the RW005 and RW004 primer sites were present, the intervening sequence was sufficiently different to disallow hybridization of the RW006 probe. This may result in a dysfunctional transposase, which could explain the absence of multiple insertion sites in these two strains.

FIG. 4.

PCR analysis of representative M. fermentans strains from various sources (Table 1), using primers RW005 and RW004 (a) and primers RS47 and RS49 (b), which are located within the transposase gene and upstream of the IS-like element, respectively (Fig. 2). PCR amplicons were analyzed by electrophoresis in a 2% agarose gel stained with ethidium bromide. Sizes of products are indicated. Lanes: 1, M51; 2, KL8; 3, AMSO; 4, Elliman; 5, MT2; 6, DEPB; 7, M52; 8, AOU; 9, KL4; 10, M64; 11, E10; 12, 16700; 13, 12406; 14, Z62; 15, incognitus; 16, M39; 17, M73; 18, PG18; 19, K7; 20, 48429; 21, M70.

In the same previous study as that described above, another primer pair located immediately upstream of the IS-like element (Fig. 2, RS47 and RS49) produced a 160-bp product for only three of the six strains tested. The RS47-RS49 primer pair was not further evaluated, since those investigators were interested in defining PCR primers for detection of all strains of M. fermentans. When we evaluated the strains listed in Table 1 with the RS47-RS49 primer pair, the most common product was 160 bp (Fig. 4b). Other products ranged from ≤80 bp (M39 and M70) to over ≥860 bp (KL8, KL4, and 12406). Some strains also gave no distinct product (M51, M52, M73, and PG18). These results are recorded in Table 1, and there is a complete correlation between genotype I, Pra expression, and the IS-like element location resulting in a 160-bp product with the RS primers. PCR amplification of all genotype II strains resulted in different-sized amplicons or no amplicon. The lack of an RS product for strains M51, M52, M73, and PG18 implies that either one or both primer sites are missing (or lack sufficient homology) or that there is an insertion between the primer sites resulting in a template which was too large to amplify efficiently.

Various PCR primer combinations (Fig. 2) were used to determine if there were any differences in the sequence within the IS-like element which might allow the detection of a stable PCR product that would be representative of the genotype II strains. Amplification with the primers MF-1 and MF-2 suggested that ORF-1 was not significantly different among the strains (Fig. 5a and b). Strains KL4 and KL8 gave no product, and strain 16700 consistently produced a weak amplicon. Similarly, by using primers MF-1 and RW004, we found no differences in the linkage between ORF-1 and ORF-2 among the strains (Fig. 5c and d). Once again, strains KL4 and KL8 gave no product and strain 16700 was amplified poorly. The linkage between the RS genomic site and the ORF-1 IS site was evaluated by using primers RS-47 and MF-2 (Fig. 5e and f). Amplification of all genotype I strains resulted in a 546-bp amplicon, reaffirming the stability of this particular insertion site for the genotype I strains. The genotype II strains were inconsistently amplified with this primer pair, indicating the instability of this insertion site in these strains.

FIG. 5.

PCR analysis of M. fermentans strains. Amplification of representative M. fermentans strains from various sources (Table 1) by using primers MF-1 and MF-2, MF-1 and RW004, and RS-47 and MF-2 (Fig. 2) is shown. PCR amplicons were analyzed by electrophoresis in 2% agarose gels and stained with ethidium bromide. Lanes in a, c, and e: 1, M51; 2, KL8; 3, AMSO; 4, Elliman; 5, MT2; 6, DEPB; 7, M52; 8, AOU; 9, KL4; 10, M64; 11, E10. Lanes in b, d, and f: 12, 16700; 13, 12406; 14, Z62; 15, incognitus; 16, M39; 17, M73; 18, PG18; 19, K7; 20, 48429; 21, M70. Sizes of products are indicated on the left.

Since Fig. 4b, 5e, and 5f suggested that the genotype II strains do not have an IS-like element consistently present near the RS site, we used primers RS-47 and MF-4, which flank the putative insertion site, to ascertain if the IS was consistently absent from this site. Figure 6 shows a stable 450-bp product for all genotype II strains, indicating that the genotype II strains have a stable linkage between the RS site and ORF-3 with no intervening IS-like element.

FIG. 6.

PCR analysis of the M. fermentans strains by using primer sites which flank the IS-like element. Primer pair RS-47–MF-4 (Fig. 2) amplicons were analyzed by electrophoresis in a 2% agarose gel and stained with ethidium bromide. Sizes of products are indicated. Genotype II strains are represented by the 450-bp product. Genotype I strains were not consistently amplified.

These analyses indicate that the two M. fermentans genotypes consistently differed in the sites of insertion of the IS-like element but not in the sequence of the element itself. Also, due to potential sequence instability within the IS-like element, the only reliable genotyping markers reside outside the IS-like element and in the genomic sequences immediately upstream and downstream of the IS junctions. Therefore, the method of choice for species detection appears to be direct detection of specific sequences within the IS-like element. In contrast, for genotype distinction, detection of stable genomic insertion sites is required, i.e., with primers RS47 and RS49 for genotype I and primers RS47 and MF-4 for genotype II. These primer pairs should allow easy and rapid genotyping of M. fermentans in clinical samples and thus obviate the need for the frequently unsuccessful isolation and culturing of this organism.

Aside from the immediate usefulness for diagnostic genotyping, these data also will help to understand how this species may adapt for survival in a particular host population. At this time, we have only a suggestive link between the sites of insertion of the IS-like element and Pra expression, and as previously stated by Lo et al. (13), there is no proof that this element is mobile. Conclusive evidence will have to await a more direct connection between the pra gene sequence and a specific insertion site. We have not analyzed a sufficient number of isolates to say that proteinase resistance is always associated with isolation from cells of a blood-related compartment or from immunocompromised patients, but completion of these analyses will determine if M. fermentans uses the mobility of this IS-like element coupled to Pra expression as a primary means of maintaining a specific niche in its host.