Rapid PCR Detection of Mycoplasma hominis, Ureaplasma urealyticum, and Ureaplasma parvum

Objective. We compared laboratory developed real-time PCR assays for detection of Mycoplasma hominis and for detection and differentiation of Ureaplasma urealyticum and parvum to culture using genitourinary specimens submitted for M. hominis and Ureaplasma culture. Methods. 283 genitourinary specimens received in the clinical bacteriology laboratory for M. hominis and Ureaplasma species culture were evaluated. Nucleic acids were extracted using the Total Nucleic Acid Kit on the MagNA Pure 2.0. 5 μL of the extracts were combined with 15 μL of each of the two master mixes. Assays were performed on the LightCycler 480 II system. Culture was performed using routine methods. Results.   M. hominis PCR detected 38/42 M. hominis culture-positive specimens, as well as 2 that were culture negative (sensitivity, 90.5%; specificity, 99.2%). Ureaplasma PCR detected 139/144 Ureaplasma culture-positive specimens, as well as 9 that were culture negative (sensitivity, 96.5%; specificity, 93.6%). Of the specimens that tested positive for Ureaplasma species, U. urealyticum alone was detected in 33, U. parvum alone in 109, and both in 6. Conclusion. The described PCR assays are rapid alternatives to culture for detection of M. hominis and Ureaplasma species, and, unlike culture, the Ureaplasma assay easily distinguishes U. urealyticum from parvum.


Introduction
Mycoplasma hominis, Ureaplasma urealyticum, and Ureaplasma parvum are small, fastidious bacteria belonging to the Mollicutes class. They lack a cell wall (preventing staining with Gram stain) and are not sensitively detected on routine bacterial cultures. Optimal recovery requires specialized media and growth conditions. There are several human pathogens in the genera Mycoplasma and Ureaplasma which are responsible for a variety of clinical manifestations involving multiple body systems [1]. M. hominis causes septic arthritis and postpartum fever and has been associated with pelvic inflammatory disease and bacterial vaginosis [2]. Ureaplasma species can cause acute urethritis and have been associated with bacterial vaginosis, preterm birth, and neonatal respiratory disease [1,3].
Although M. hominis and Ureaplasma species can be cultured, this requires technical skill for interpretation of microscopic colonies and takes two to five days. U. urealyticum was the only Ureaplasma species until 2002, when U. parvum was described [4]. The two are not distinguished based on culture characteristics alone. Real-time PCR detection of these microorganisms from clinical samples circumvents technical issues related to culture and shortens turnaround time for detection and identification.
Few real-time PCR assays and associated studies have been described for M. hominis. A real-time PCR assay targeting M. hominis gap identified two positive cervical swabs from women being evaluated for infertility [5]. 153 urogenital specimens were tested with a real-time PCR assay targeting M. hominis yidC, of which 45 were PCR-and culture positive and 10 PCR positive only [6]. Finally, extragenital M. hominis infection was diagnosed in three patients using a real-time PCR assay targeting the M. hominis 16S ribosomal RNA gene [7]. There has been more work on real-time PCR assays for Ureaplasma species, although some have described assays but have not evaluated clinical specimens or clinical isolates [8]. A real-time PCR assay that detects and distinguishes U. urealyticum from parvum was described but used to assess 87 vaginal swabs [9]. Tang et al. used a real-time PCR assay that detects and distinguishes U. parvum and urealyticum to test 346 genitourinary swabs; 120 were positive for the former and 21 for the latter, including 5 positive for both [10]. Finally, Vancutsem et al. used a real-time PCR assay for detection and differentiation of U. urealyticum and parvum to evaluate 300 lower genital tract specimens; 132 were culture positive, of which all plus an additional 19 were PCR-positive (19, U. urealyticum; 120, U. parvum; 12, Ureaplasma species) [11].
Herein, we present one real-time PCR assay for the detection of M. hominis and another for the detection and differentiation of Ureaplasma species and report results of these assays on 283 genitourinary specimens in comparison to culture.

Clinical
Specimens. 283 genitourinary specimens (swabs, urine) submitted to the Mayo Clinic Clinical Microbiology Laboratory in transport medium (e.g., UTM, M5) for M. hominis and Ureaplasma culture were evaluated. No clinical data associated with these specimens was available. This study was approved by the Mayo Clinic Institutional Review Board.

Mycoplasma hominis Culture.
Samples were placed into arginine broth, incubated at 35 ∘ C, and monitored four times daily for up to five days. Color change (indicating an alkaline pH shift) in the arginine broth prompted subculture of 50 L to an A7 agar plate. Plates were incubated anaerobically at 35 ∘ C for up to five days and examined daily with an inverted light microscope for "fried egg" morphology colonies.

Ureaplasma
Culture. Samples were placed into U9 broth, incubated at 35 ∘ C, and monitored four times daily for up to five days. A color change (indicating an alkaline pH shift) in the U9 broth prompted subculture of 100 L to an A7 agar plate. Plates were incubated anaerobically at 35 ∘ C for up to 48 hours and examined with an inverted light microscope for small, circular to irregular colonies growing into the surface of the agar, with a surrounding red zone. Confirmation of Ureaplasma species was indicated by golden-brown stained colonies with the addition of 0.167 M CO(NH 2 ) 2 and 0.04 M MnCl 2 in water.  (Table 1) were designed using the LightCycler Probe Design Software, version 2.0 (Roche Diagnostics, Indianapolis, IN, USA) and DNA Workbench, version 5.7.1 (CLC Bio, Cambridge, MA, USA). Positive control plasmids were constructed for the three target-specific genes (Table 1) using the pCR 2.1 TOPO TA Cloning Kit (Invitrogen Corporation, Carlsbad, CA, USA). Sources for the inserted target sequences were M. hominis ATCC 23114, U. urealyticum ATCC 27618, and U. parvum ATCC 27815D. Plasmids were purified using the High Pure Plasmid Isolation Kit (Roche Applied Science). Sizes of the cloned inserts were confirmed by EcoR1 digestion. Plasmid inserts were sequenced using M13 forward and reverse primers included in the cloning kit, to confirm proper insert orientation. Plasmids were diluted in Tris-EDTA buffer (pH 8.0) and stored at 4 ∘ C.

Polymerase Chain Reaction Assay. Primers and probes
The two assays were independently optimized on the LightCycler 480 II platform employing LightCycler 480 Software version 1.5 (Roche Applied Science). 15 L of PCR master mix, containing final concentrations of 1X Roche Genotyping Master (Taq DNA polymerase, PCR reaction buffer, deoxyribonucleoside triphosphate with dUTP substituted for dTTP and 1 mM MgCl 2 ), 1 mM (additional) MgCl 2 , and 1X of each of the LightCycler primer-probe sets (Table 1) were added to a 96-well LightCycler 480 plate. Extracted nucleic acid (5 L) was then added to each well. The cycling program was as follows: denaturation at 95 ∘ C for 10 min; amplification for 45 cycles of 10 s at 95 ∘ C, 15 s at 55 ∘ C (single acquisition), and 15 s at 72 ∘ C; melting curve analysis for 30 s at 95 ∘ C, 10 s at 59 ∘ C, 15 s at 45 ∘ C (ramp rate of 0.1 ∘ C/s), and 0 s at 80 ∘ C (ramp rate of 0.14 ∘ C/s and continuous acquisition); and cooling for 30 s at 40 ∘ C. Positive and negative controls were included in each run. The positive control consisted of the abovementioned plasmids in S.T.A.R. buffer : sterile water (1 : 1) at a concentration of 1,000 targets/ L. The negative control consisted of 1,000 colony forming units of Escherichia coli ATCC 25922 S.T.A.R. buffer : sterile water (1 : 1) at a concentration of 1,000 targets/ L.

Polymerase Chain Reaction Sensitivity and Specificity.
Predicted amplified product, primer, and probe sequences were subjected to BLAST searches using the National Center for Biotechnology Information (NCBI) genomic database (http://www.ncbi.nlm.nih.gov/). Analytical sensitivity was assessed by spiking a series of six tenfold dilutions of quantified genomic DNA from M. hominis ATCC 23114, U. urealyticum ATCC 27816, and U. parvum ATCC 27815D into genitourinary samples. Each dilution was extracted in triplicate and each extract was assayed in duplicate. The limit of detection was the lowest dilution where all six replicates were detected. Inclusivity and cross-reactivity were assessed using a panel organisms (Table 2), including 16 members of the Mollicutes class.
Clinical sensitivity and specificity were assessed by assaying the aforementioned clinical specimens and comparing results to those of culture. Discordant samples were tested courtesy of Dr. Stellrecht, at an independent clinical laboratory (Albany Medical Center) with a previously described assay [12].
The ability of the Ureaplasma assay to differentiate urealyticum from parvum was assessed as follows. Cultured isolates from clinical samples were directly subjected to PCR with species differentiation based on melting curve analysis; sequence variations underlying the probed regions of U. urealyticum and parvum result in separation of the melting temperature of the two species (Figures 1 and 2). Results were compared to those of a previously described conventional PCR speciation method targeting the multiplebanded antigen using primers UMS-57 and UMA222 for U. parvum and UMS-170 and UMA263 for U. urealyticum [13].

Statistical
Analysis. Assessment of the assays' sensitivity and specificity, with associated 95% confidence intervals (CI), compared to that of culture for M. hominis and Ureaplasma species was made using SAS software version 9.1 (SAS, INC, Cary, NC, USA). genitourinary specimen. Amplified product, primer, and probe sequences were subjected to NCBI database searches using BLAST software; no significant homology was noted outside of the genera targeted by these assays. Nucleic acid material from members of the Mollicutes class, excluding M. hominis and the Ureaplasma species, was not detected ( Table 2).
The Ureaplasma PCR assay had a clinical sensitivity and specificity of 96.5% (95% CI: 92.1%, 98.9%) and 93.8% (95% CI: 88.1%, 97.0%), respectively ( Table 3). The 14 discordant results were tested at Albany Medical Center; [12] five of nine specimens that were PCR positive/culture negative were PCR positive, and all five specimens that were PCR negative/culture positive were PCR negative. Of the specimens that tested positive for Ureaplasma species by PCR and were culture positive, U. urealyticum alone was detected in 28, U. parvum alone in 109, and both in 2. Among the PCR positive/culture-negative specimens, U. urealyticum was detected in 3 and U. parvum in 6.
Thirty-one culture isolates of Ureaplasma species were tested with the Ureaplasma assay and a previously reported PCR method that differentiates between the two species [13]. The reference method yielded species-level identification for 20 isolates, including 4 U. urealyticum and 16 U. parvum, with identical results to the assay described herein. The remaining 11 isolates were speciated by the assay described herein but not by the reference method; they were confirmed to be Ureaplasma species by partial 16S ribosomal RNA gene sequencing [14]. All partial 16S ribosomal RNA gene sequences were identical to one another and were perfect matches to bases 145,365 through 145,845 of GenBank AF222894.1    International Journal of Bacteriology (U. parvum) and bases 40 through 520 of GenBank L08642.1 (U. urealyticum).

Discussion
We describe two rapid real-time PCR assays, one for detection of M. hominis and the other for detection of Ureaplasma species; they have comparable performance to culture but yield results in three hours, instead of two to five days for culture. These assays are performed on a standard platform and are adaptable to automation, a potential advantage over other described methods, especially for large reference laboratories that process large numbers of specimens.
We are not aware of other real-time PCR studies that have assessed M. hominis and Ureaplasma species using the same set of clinical samples. Overall, 14% of tested specimens were PCR positive for M. hominis and 52% for Ureaplasma species. A multiplex PCR enzyme-linked immunosorbent assay was used to detect M. hominis and U. parvum and urealyticum in cervical swabs from 175 Australian women with and without cervicitis; 16% tested positive for M. hominis and 68% for Ureaplasma species [15]. Multiplex PCR and autocapillary electrophoresis were used to detect M. hominis and Ureaplasma species (without differentiating U. parvum from urealyticum) in genitourinary specimens from 113 South Koreans with sexually transmitted infections; 12% were positive for M. hominis and 43% for Ureaplasma species [16]. These findings are similar to ours [15,16].
Our PCR assay not only detects Ureaplasma species but also differentiates U. parvum from urealyticum. As in prior studies, U. parvum was more common than U. urealyticum, [10,11,15,17] with 41% of the genitourinary specimens testing positive for the former and 12% for the latter. In one prior study, 63% of specimens were positive for U. parvum and 7% for U. urealyticum [15]. Another study showed, using a multiplex PCR-reverse line blot assay, that 48% of first voided urine specimens from women attending sexual health clinics in Australia were positive for U. parvum and 25% for U. urealyticum [17]. In the study by Tang et al., 36% of genitourinary swabs collected from hospitalized males and females in China were positive for U. parvum and 8% for U. urealyticum [10]. Finally, in study by Vancutsem et al., 44% of lower genital tract specimens obtained from healthy women at their first prenatal visit in Belgium were positive for U. parvum and 10% for U. urealyticum [11]. Despite different geographic locales and clinical status, these numbers are strikingly similar.
In addition to the advantage of speed, the described assays overcome the challenges of detection of these organisms by culture. Although culture is considered a gold standard method (and was so considered in this study), colonial identification is challenging and subjective because it is done using the human eye and a dissecting microscope. Artifacts may be misidentified as colonies, yielding false-positive results, or colonies may be overlooked, yielding false-negative results. Although PCR may be considered more technically complex, in a laboratory where technologists are familiar with PCR, this approach is more user-friendly (and generalizable among assays for various microorganisms) than culture.
The described assays may be useful for investigating epidemiology and pathogenesis of infections with U. parvum and urealyticum [2,18]. Although extra-genital specimens were not tested, the described M. hominis assay may be useful to detect extra-genital M. hominis infections [7].