Detection of Putative Virulence Genes alr, goiB, and goiC in Mycoplasma hominis Isolates from Austrian Patients

Abstract

In Mycoplasma hominis, two genes (alr and goiB) have been found to be associated with the invasion of the amniotic cavity, and a single gene (goiC) to be associated with intra-amniotic infections and a high risk of preterm birth. The syntopic presence of Ureaplasma spp. in the same patient has been shown to correlate with the absence of goiC in M. hominis. The aim of our study was to investigate the presence of alr, goiB, and goiC genes in two groups of M. hominis isolates collected from symptomatic and asymptomatic male and non-pregnant female patients attending an Outpatients Centre. Group A consisted of 26 isolates from patients with only M. hominis confirmed; group B consisted of 24 isolates from patients with Ureaplasma spp. as the only co-infection. We extracted DNA from all M. hominis isolates and analysed the samples for the presence of alr, goiB, and goiC in a qPCR assay. Additionally, we determined their cytotoxicity against HeLa cells. We confirmed the presence of the alr gene in 85% of group A isolates and in 100% of group B isolates; goiB was detected in 46% of the samples in both groups, whereas goiC was found in 73% of group A and 79% of group B isolates, respectively. It was shown that co-colonisation with Ureaplasma spp. in the same patient had no effect on the presence of goiC in the respective M. hominis isolate. We did not observe any cytotoxic effect of the investigated isolates on human cells, regardless of the presence or absence of the investigated genes.

1. Introduction

Mycoplasmas belong to the class Mollicutes, which are the smallest, self-replicating bacteria, lacking a rigid cell wall and possessing an extremely small genome [1]. These bacteria are mostly isolated from the mucosal surfaces of the genitourinary and respiratory tracts. Mycoplasma hominis, Ureaplasma urealyticum, and Ureaplasma parvum represent mycoplasmas associated with the normal genital flora, as well as with infection, and are very common [2,3]. These species do not appear to cause symptoms or adverse effects in non-pregnant women, and M. hominis seems to be associated with abnormal vaginal discharge only in women who also have bacterial vaginosis (BV) [3]. However, in pregnancy, colonisation with M. hominis or Ureaplasma spp. may be associated with several adverse pregnancy outcomes [4,5], and intraamniotic infections with these microorganisms have been associated with inflammation, preterm rupture of membranes, and preterm birth [6]. Allen-Daniels and colleagues identified two genes (alr and goiB) in M. hominis isolated from pregnant women to be associated with the invasion of the amniotic cavity and a single gene (goiC) to be associated with intra-amniotic infections and a high risk of preterm birth. The alr gene appears to encode an alanine racemase that converts L-alanine to D-alanine, which is a component of peptidoglycan. Although Mycoplasmas do not possess peptidoglycan, some species are able to encode alanine recamase [7]. However, there is no information available on the function of this enzyme in Mycoplasmas. The gene named by Allen-Daniels as goiB (gene of interest B) encodes a 379 amino acid protein of unknown function, which aligns over 97% of its length to a hypothetical protein from U. urealyticum. This gene is supposed to encode a secreted protease [7]. The goi C gene (designed by Allen-Daniels as gene of interest C) shows no similarity to proteins of known function, but appears, along with the previous two genes, to be unique to amniotic fluid/placental isolates. In that study, the absence of goiC in tested isolates was shown to correlate with the presence of Ureaplasma spp. in the vagina of the same patient [7].

The aim of the current study was to screen clinical M. hominis isolates from men and non-pregnant women for alr, goiB, and goiC and to evaluate if there is a correlation between the absence of goiC in tested isolates and the presence of Ureaplasma spp. in the same patient. Moreover, the cytotoxicity of the isolates against HeLa cells was investigated.

2. Results

2.1. Detection of the Putative Virulence Genes

The presence/absence of alr, goiB, and goiC in the respective isolates is listed in

Table 1. Presence of alr, goiB, and goiC in tested M. hominis isolates groups; F—female patients and M—male patients.

Patients Isolates	Total Number F/M	alr (%) F(%)/M(%)	goiB (%) F(%)/M(%)	goiC (%) F(%)/M(%)
M. hominis only (group A)	26 15/11	22/26(85) 12(80)/10(90)	12/26(46) 6(40)/6(54)	19/26(73) 9 (60)/10(90)
M. hominis + Ureaplasma spp. (group B)	24 12/12	24/24(100) 12(100)/12(100)	11/24(46) 3(25)/8(67)	19/24(79) 10(83)/9(75)

.

Twelve of the patients (six from group A and six from group B; four women and eight men) reported no symptoms; in all their M. hominis isolates, the presence of goiC was confirmed. The other 38 patients reported various symptoms, including pain, itching, vaginitis, and prostatitis.

2.2. Co-Occurrence of Putative Virulence Genes

The co-occurrences of the alr, goiB, and goiC genes in the two M. hominis groups are listed in

Table 2. Co-occurrences of the putative virulence genes in tested M. hominis groups A and B.

Patients Isolates	alr+/goiB+/goiC+ (%)	alr+/goiB+/goiC- (%)	alr+/goiB-/goiC+ (%)	alr+/goiB-/goiC- (%)	alr-/goiB-/goiC+ (%)	alr-/goiB-/goiC- (%)
M. hominis only (group A)	9/26 (34.6)	3/26 (11.5)	8/26 (30.8)	2/26 (7.7)	2/26 (7.7)	2/26 (7.7)
M. hominis + Ureaplasma spp. (group B)	10/24 (41.7)	1/24 (4.2)	9/24 (37.5)	4/24 (16.6)	0/24 (0)	0/24 (0)

alr+goiB+goiC+: all genes detected. alr+goiB+goiC-: alr and goiB genes detected. alr+goiB-goiC+: alr and goiC genes detected. alr+goiB-goiC-: only alr gene detected. alr-goiB-goiC+: only goiC gene detected. alr-goiB-goiC-: no putative virulence gene detected.

.

2.3. Lactate Dehydrogenase Assay (LDH)

In the current study, none of the M. hominis isolates investigated, regardless of the presence or absence of the mentioned genes, showed toxicity to HeLa cells. After 48 h of incubation of M. hominis collected from LDH assay plates in PPLOa medium, a change in the colour of the PPLOa medium was observed in all samples collected, indicating that all M. hominis isolates survived and were capable of growing.

2.4. Growth Assay

There were no significant differences in growth rates between the study groups (A and B.

2.5. Comparison of the Ct-Values

Comparison of the Ct-values of hitA and goiC within the same M. hominis sample revealed no differences. Both were constant between cycles 16 and 20, except for two M. hominis isolates, where hitA emerged at cycle 13 and 16 and goiC at cycles 23, and 19, respectively.

3. Discussion

We compared two groups of patients with M. hominis isolates and showed that the putative virulence genes (alr, goiB, and goiC) were present in both investigated groups. Co-colonisation with Ureaplasma spp. in the same patient had no effect on the presence of goiC in the corresponding M. hominis isolate. This is in contrast to the report that the absence of goiC in M. hominis isolates correlated with the co-existence of Ureaplasma spp. in the vagina [7]. We compared growth rates of the M. hominis isolates of groups A and B and revealed no growth impairment due to the presence or absence of any of the investigated genes, or the presence of Ureaplasma spp. in the same patient. The co-occurrence analysis of these genes also showed no significant differences between the tested groups.

In our collection of “M. hominis only” isolates (Group A), 85% of samples were positive for alr, 46% for goiB, and 73% for goiC. We compared our results with those available in public databases (n = 60) and found that only 16 M. hominis strains were positive for alr (26%), 5 for goiB (8%), and 11 for goiC (18%). Unfortunately, for these isolates, no information about the patients, their medical history, or co-infections was available. Moreover, we compared the specific sequences of the primers used in this study, as well as their products, with those available in public databases and could confirm that they are highly conserved, mainly alr with identities 100–99.35% and goiB with identities 100–99.42%, while goiC presented identities 100–97.40%. The significantly higher rates of putative virulence genes in our isolates can be explained by the likelihood of mixed infections with different M. hominis genotypes within the same patient (as all colonies of M. hominis on the agar plates were used), which cannot be completely excluded. On the other hand, a comparison of Ct values of the control gene hitA and the target gene goiC within the same sample of M. hominis revealed no difference, resulting in the assumption that both genes were present in the same number in each M. hominis isolate tested. Only in two of the 50 M. hominis isolates investigated, notably from the group where Ureaplasma spp. was also detected in the same patient, was there a discrepancy in the Ct values with regard to goiC emerging later than hitA. This could be an indication of genetic variation in these two isolates, or growth inhibition of goiC-positive M. hominis in the presence of Ureaplasma spp. Growth assays revealed no difference, regardless of the original presence or absence of Ureaplasma spp. in the same patient. Moreover, as this was only observed in two out of 24 isolates, where M. hominis and Ureaplasma spp. were detected, genetic variation seems more likely.

We can exclude the growth of other Mycoplasma species, because the agar plates on which M. hominis was cultured from swabs are designed to diagnose and distinguish between M. hominis and Ureaplasma spp. Isolates of M. hominis were checked for the presence of the putative virulence genes by PCR, always using hitA as a reference gene that is specific for M. hominis. Moreover, each patient was also tested for the presence of M. genitalium by specific PCR at the OCD, and only M. hominis isolates of M. genitalium-negative patients were included in the current study. In addition, testing more than 200 vaginal swabs collected at the OCD in our laboratory using the same 16S rRNA-specific primers revealed only Candidatus Mycoplasma girerdii, while no other mycoplasma species were detected except M. hominis, M. genitalium, Ureaplasma parvum, or U. urealyticum (own unpublished data).

Comparative studies on mycoplasmas have confirmed that each species possesses a unique or specific set of virulence genes, such as the lipoprotein genes in M. penetrans or M. agalactiae [8,9]. It is interesting to note that some of the major immunogenic lipoproteins (such as P35 in M. penetrans) can be completely switched off at the transcriptional level, making it difficult to detect these organisms in the state of disease. This means that infections caused by these microorganisms may go undetected [8]. The genes alr, goiB, and goiC are exceptional, as they cannot be switched on or off, but they are present or absent. However, it remains to be elucidated if these genes are indeed related to virulence, as in the current study, none of the M. hominis isolates investigated, regardless of the presence or absence of the mentioned genes, showed toxicity to HeLa cells. All isolates collected after co-incubation with HeLa cells were able to grow, and microscopic examination revealed no monolayer disruption of HeLa cells after co-incubation with the M. hominis isolates, indicating that there was no pathogenic effect on any of the M. hominis isolates. However, it was demonstrated that M. hominis mainly adhered to the cell surface of the HeLa cells, and mycoplasmal invasion was also observed. The ability of M. hominis to invade HeLa cells was first described by Taylor-Robinson [5]. We must point out that the ability of M. hominis to invade is not limited to HeLa cells, but has also been observed in spermatozoa [10,11] and Trichomonas vaginalis, which act as a ‘Trojan horse’ for M. hominis to replicate and distribute to other human hosts [12]. Thi Trung Thu et al. [13] demonstrated that a high number of M. hominis intracellularly associated with T. vaginalis possessed the goiC gene in association with arl and goiB. They also showed that metronidazole treatment of M. hominis-infected T. vaginalis allows the delivery of live intracellular goiC-positive M. hominis from antibiotic-killed protozoa and that these free M. hominis were able to infect human cell cultures, but no cytotoxicity assays were performed. Adherence and proteins involved in adherence are important for Mycoplasma pathogenicity, which is why most studies have focused on this [14,15,16,17]. Unfortunately, detailed studies on the cytotoxicity of M. hominis isolates carrying putative virulence genes are scarce, as are data on alr and goiB (which are also unique to M. hominis isolated from amniotic fluid or placenta) and their role in M. hominis virulence. In general, data on the role of M. hominis as a pathogen are rare and cannot be pinned down to single genes or proteins to date.

The participation of Mycoplasma spp. and Ureaplasma spp. in adverse pregnancy outcomes is increasingly accepted, and colonisation with both microorganisms at the same time has been linked with various adverse effects, including infertility, low birth weight, stillbirth, preterm delivery, pelvic inflammatory disease (PID), etc. [18,19]. Colonisation values of M. hominis on the mucosal surfaces of the cervix or vagina range between 20 and 30% around the world [18,19,20] and seem to indicate a commensal character of the microorganisms. Thus, the sole presence of these microorganisms in the vaginal flora may not cause pathological problems, but only their combination with other factors, such as bacterial vaginosis or cervical insufficiency, may induce preterm labour [21]. It has been shown that the colonisation of the lower genital tract with M. hominis and U. urealyticum was associated with placental tissue infection in women with spontaneous abortion. The authors assumed that the direct action of U. parvum on placental tissue altered the gestational tolerogenic state, reducing the immune response against pathogens and activating the extrinsic apoptotic pathway, causing spontaneous abortion [22]. As M. hominis and Ureaplasma spp. can be detected in both healthy women and BV patients, questions still arise as to whether M. hominis should be classified as a pathogen at all. To date, no evidence has been found that M. hominis is a vaginal pathogen in adults [23], and according to the recommendation of the European Academy of Dermatology and Venereology and the European STI Guidelines Editorial Board, patients should not be screened for M. hominis, U. urealyticum, and U. parvum [24]. However, it has been shown that the prevalence of M. hominis in semen samples of infertile and fertile men varies from 3.2% to 18.2% and from 0.9% to 14.3%, respectively [25]; the frequency of genital ureaplasmas and mycoplasmas detected in semen samples of infertile men in Tunisia was 19.2% and 15.8%, respectively [10]. Thus, the role of these microorganisms in disease is still not clear, and therefore, the question of whether M. hominis should be considered a pathogen still requires further research. Of the patients investigated in the current study, 12 did not report symptoms, despite all being carriers of M. hominis with goiC. This and the lack of cytotoxic activity towards HeLa cells in vitro suggests that perhaps goiC is not a virulence gene per se, or M. hominis with this gene is only virulent in pregnant women. Allen-Daniels and colleagues [7] focused on symptoms and isolation of M. hominis from the placenta to reflect the virulence of the isolates; no cytotoxicity assay was performed then. They described their results on the basis of just a few isolates. In our study also, no statistics were performed on the correlation of symptoms and the presence or absence of the genes tested, as a total of 50 samples is too small a number to allow reliable correlation statistics. Therefore, future studies involving a larger number of patients are needed to complete the picture. On the other hand, it might also be supposed that our negative lactate dehydrogenase assay results indicate that this test is not suitable for testing the virulence of M. hominis isolates, which has to be clarified in future studies. Furthermore, other cell lines, such as human placental cells or vaginal epithelial cells, should be tested in more detail to observe the possible cytotoxic effect of M. hominis. Also, testing of gene or protein expression levels during bacteria-human cell co-cultures could explain more about the role of alr, goiC, and goiB genes. However, in our study, we first focused on the presence or absence of these genes in two different M. hominis groups (with concomitant Ureaplasma spp. in the same patient and without). Exploration of the cytotoxicity of M. hominis isolates in other types of cells is planned for future projects.

Here, this question also arises: can M. hominis with putative virulence genes harm the foetus, as they can obviously enter amniotic fluid and pass the placenta? M. hominis has been isolated from various sites of the human body [1], including the central nervous system in neonates [1,26], indicating infection in utero or during birth. It still has to be elucidated if this extragenital M. hominis possess (more) virulence genes, and the question of whether to screen for M. hominis in pregnancy or not currently remains unanswered [27].

4. Materials and Methods

4.1. Sample Collection

Only patients for whom a complete STD screening (including tests for Trichomonas vaginalis, Neisseria gonorrhoeae, Chlamydia trachomatis, and Mycoplasma genitalium (Alinity m system (STI Amplification Kit; Abbott; North Chicago, IL, USA)), as well as Candida spp., aerobic and anaerobic bacteria including M. hominis and Ureaplasma spp. (by culture), or bacterial vaginosis (by Gram staining) was performed were considered for this study.

Genital swabs taken from symptomatic and asymptomatic patients visiting the Outpatients Centre for Diagnosis of Venero-Dermatological Diseases (OCD), Vienna, Austria, were smeared onto Mycoplasma/Ureaplasma agar plates (Thermo Fisher Scientific; Waltham, MA, USA) and cultivated anaerobically at 37 °C for two days before microscopic examination. Mycoplasma genitalium would not grow on these plates within the time of observation.

After microscopic examination, all colonies of a positive culture of Mycoplasma hominis (grown in typical fried egg shape) from each respective patient were further transferred and cultivated micro-aerobically in PPLOa, consisting of PPLO broth (Difco, Thermo Fisher Scientific) and supplemented with 10% arginine solution (Sigma-Aldrich, 100 mL/L), 20% yeast extract (Difco, Thermo Fisher Scientific, 50 mL/L), 0.2% phenol red indicator (VWR International GmbH; Radnor, Pennsylvania, USA), horse serum (Sigma-Aldrich; St. Louis, Missouri, USA, 100 mL/L), and an antibiotic mixture (Sigma-Aldrich, 10 mL/L); pH 6.5 at 37 °C. After two rounds of sub-culture at the ISPTM, the samples were used for LDH assay and PCR testing.

Altogether, only samples from patients with only M. hominis or M. hominis and Ureaplasma spp. were included in the current study. Out of those 50 isolates, 26 isolates were collected from patients with only M. hominis infection (11 from males and 15 from females), designated as group A, and 24 isolates were from patients with both M. hominis and Ureaplasma spp. as only co-infection (12 from males and 12 from females), designated as group B.

4.2. DNA Isolation

Mycoplasma hominis cultures in PPLOa were centrifuged at 10,000× g for 10 min, and after two washing steps with PBS (phosphate-buffered saline), the QIAGEN DNA Mini Kit (QIAGEN GmbH, Hilden, Germany) was used for the extraction of total DNA from the cultures. The DNA was stored at −20 °C until further use.

4.3. Detection of the Virulence Genes

Primers specific for alr, goiB, and goiC developed by Allen-Daniels [7] were used in a qPCR assay to detect these genes. The M. hominis-specific gene hitA was used as standard control [28].

qPCRs were performed in a CFX96 thermocycler (Bio-Rad; Hercules, CA, USA) in a total volume of 15 µL/sample, consisting of 7.5 µL Takyon No Rox SYBR Master Mix dTTP Blue (Eurogentec; Seraing, Liège, Belgium, USA ), 0.25 µL of each primer (10 µM), 5 µL water, and 2 µL DNA (10 ng) using following protocol: starting with 50 °C for 10 min and 95 °C for 10 min, followed by 35 repeats of 15 s at 95 °C and 1 min at 60 °C. The qPCR runs were conducted in duplicate, twice independently.

4.4. Lactate Dehydrogenase Assay

Determination of cytotoxicity was achieved with the Lactate Dehydrogenase Assay (CyQUANT^TM LDH Cytotoxicity Assay Kit reaction mixture, Thermo Fisher Scientific-Life Technologies), which is based on the release of lactate dehydrogenase (LDH) upon cell lysis. HeLa cells were used as target cells and cultured in 96-well plates (10*4 cells per well) in IMDM Medium (Gibco, Thermo Fischer Scientific), with 10% fetal bovine serum (Sigma Aldrich) at 5% CO₂ for 24 h. For the test, the medium was replaced by 180 µL of IMDM medium without FBS, and 20 µL of M. hominis isolates (between 1 × 10⁷ cells/mL and 1 × 10¹⁰ cells/mL) were added to the HeLa cells and incubated for 2, 24, and 48 h at 37 °C in 5% CO₂ atmosphere. Ten representatives from each study group (A and B) were chosen for the assays repeated three times, and mean cytotoxicity was calculated according to the manufacturer’s protocol. After distinct incubation times (2, 24, and 48 h), M. hominis samples were collected from the wells (total one well volume 200 µL) and transferred to PPLOa medium (content of each well in separate 2 mL medium) to evaluate survival; a change in the colour of the medium indicated the presence of viable M. hominis.

4.5. Growth Assay

Ten representatives of each study group (A and B) were cultivated in PPLOa medium to determine their growth, using colour-changing units (CCUs). Growth was determined after 4, 24, and 48 h of incubation, using serial 1:10 dilution in 96-well microtiter plates. Plates were sealed and incubated at 37 °C for 48 h. Each isolate was tested twice in duplicate. The number of bacteria was the highest dilution at which there was a visible colour change.

5. Conclusions

We presented that the presence of Ureaplasma spp. in the same patient had no effect on the presence of goiC in the respective M. hominis isolate. No cytotoxic effect of the tested isolates on human cells was found. Thus, the question of whether or not M. hominis should be considered a pathogen still requires further research.