Lack of resistance to macrolides in Mycoplasma genitalium detected in South African pregnant women

Background Azithromycin regimens have been considered first-line treatment for Mycoplasma genitalium (M. genitalium), a sexually transmitted infection (STI) associated with adverse pregnancy outcomes. However, recent years have seen rapid emergence of macrolide resistance in M. genitalium as a result of widespread administration of azithromycin. Currently, there are limited data on macrolide resistance in pregnant women from KwaZulu-Natal (KZN), South Africa. This study investigated the prevalence of M. genitalium and emerging patterns of macrolide resistance in pregnant women from KZN. Methods This was a sub-study of a larger study which involved laboratory-based detection of STIs in pregnant women. In the main study, pregnant women provided urine samples for detection of STIs. For this study, deoxyribose nucleic acid (DNA) extracted from stored urine was used to determine emerging macrolide resistance by amplification of the 23S ribosomal ribonucleic acid (rRNA) gene of M. genitalium by polymerase chain reaction (PCR) and sequencing of amplicons to identify mutations associated with resistance. The Allplex™ MG & AziR assay was used as a confirmatory assay. Results The prevalence of M. genitalium in pregnant women was 5.9% (13 out of 221). Sequencing of PCR amplicons did not reveal the presence of the A2059G and A2058G mutations associated with macrolide resistance. These findings were confirmed by the Allplex™ MG & AziR assay. Conclusion Despite the lack of resistance to macrolides in this study population, continued antimicrobial resistance surveillance for M. genitalium in pregnant women is important because azithromycin is now part of the South African national STI syndromic management guidelines for vaginal discharge syndrome.


Introduction
Mycoplasma genitalium (M. genitalium) is an emerging sexually transmitted infection (STI) that is associated with non-gonococcal urethritis (NGU) in men 1,2 and cervicitis in women. 2,3 In addition, genital mycoplasma infections in women have been implicated in various pathological conditions and upper genital tract complications during pregnancy, consequently giving rise to a host of adverse outcomes. These include pelvic inflammatory disease, endometritis, chorioamnionitis and postpartum fever, resulting in complications such as infertility, spontaneous abortion, stillbirth, preterm birth, low birth weight and perinatal mortality. 4,5,6,7,8,9 However, there is still controversy regarding the role of M. genitalium in adverse pregnancy outcomes. 10,11,12 Further, M. genitalium has been reported to be a possible contributing factor for the acquisition and transmission of human immunodeficiency virus (HIV). 13,14 The prevalence of M. genitalium infection in women within the general population was reported between 1% and 4% but can reach 10% or higher in STI clinic attendees. 15,16,17 In a study conducted in Pretoria, South Africa, in 2012 and 2016, the prevalence of M. genitalium in termination of pregnancy attendees was reported to be 6.0% in 2012 and 7.7% in 2016. 18 Current South African Sexually Transmitted Infections Management Guidelines (2015) recommend the administration of a single oral dose (1 g) of azithromycin for the treatment of vaginal discharge syndrome and male urethritis syndrome. 19 However, recent years have seen the rapid emergence of macrolide resistance in M. genitalium as a result of this widespread administration of azithromycin. 18,20,21,22,23,24,25 The genetic basis for macrolide resistance was reported to be as a result of mutations occurring at positions 2058 and 2059 (E. coli numbering) within the V region of the 23S ribosomal ribonucleic acid (rRNA) gene of M. genitalium. 22 To date, there remains limited data available on the emergence of macrolide resistance in pregnant populations from KwaZulu-Natal (KZN). The present study provides data on the lack of macrolide resistance in pregnant women in our setting. In addition, this study provides data on the prevalence of M. genitalium in pregnant women, which is currently lacking in KZN.

Study population and samples
The study was a sub-study of the larger study which involved laboratory-based detection of STIs in pregnant women (BE214/17). In the main study, pregnant women provided urine samples for detection of STIs. The women were recruited from the King Edward VIII Hospital in Durban, KZN, during the period November 2017 to April 2018. All enrolled women were 18 years and older, willing to provide written informed consent, information on their demographics, sexual behaviour, clinical information and a urine sample to be tested for vaginal infections.
Deoxyribose nucleic acid (DNA) was extracted from the urine using the PureLink TM Microbiome DNA Purification Kit (ThermoFisher Scientific, Massachusetts, USA) in accordance with the manufacturer's instructions. Unused DNA was stored at -20 °C. For this sub-study, n = 221 stored DNA samples were used to determine the presence of emerging resistance to macrolides amongst the participants.

Detection of Mycoplasma genitalium from urine
Commercially available primers and probes specific for M. genitalium (Ba04646249_sl) were used with the TaqMan quantitative polymerase chain reaction (qPCR) assay (ThermoFisher Scientific, USA). The assay was run on the Quant Studio 5 Real-time PCR detection system (ThermoFisher Scientific, USA). The PCR was performed in a final reaction volume of 5 µL comprising: 0.25 µL Fluorescein amidite (FAM)-labelled probe and/or primer mix, 1.25 µL Fast Start 4x probe master mix (ThermoFisher, Part No. 4444434), 1.5 µL template DNA and nuclease-free water. Non-template control reactions and positive controls (TaqMan™ Vaginal Microbiota Extraction Control; cat no. A32039, ThermoFisher Scientific, USA) were also included. Amplification was performed at 95 °C for 30 s followed by 45 cycles comprising of denaturation at 95 °C for 3 s and annealing at 60 °C for 30 s. Detection of amplified fluorescent products was carried out at the end of the annealing phase.
The raw fluorescent data included the C T mean values, which were automatically generated by the Quant Studio 5 Realtime PCR system software (ThermoFisher Scientific, USA).

Detection of mutations in the 23S ribosomal ribonucleic acid gene conferring macrolide resistance
Screening assay: Conventional polymerase chain reaction Macrolide resistance-associated mutations in the 23S rRNA gene were determined for all samples that tested positive for M. genitalium by the TaqMan assay. The unique 147 base pair (bp) region within the 23S rRNA gene of M. genitalium, which flanks mutations found in the V region of the gene, was amplified and sequenced. 22 The screening PCR was performed in a final volume of 25 µL and comprised of 12.5 µL Dream Taq (2x) master mix (ThermoFisher Scientific, USA), 1 µL of each (10 µM) primer, 5 µL template DNA and nuclease-free water. Amplification was performed at an initial denaturation of 95 °C for 2 min, followed by 40 cycles of 95 °C for 30 s (denaturation), 54 °C for 1 min (annealing) and 72 °C for 1 min (extension). A final extension step at 72 °C for 5 min was included. Polymerase chain reaction amplicons were separated on 1% gel electrophoresis and visualised using an ultraviolet transilluminator (Gene Genius, SYNGENE, Maryland, USA).

Deoxyribose nucleic acid sequencing of amplicons
The 147 bp amplicon was sequenced using the Sanger method at Inqaba Biotechnological Industries in Pretoria, South Africa. The amplicons were sequenced using an ABI3500XL genetic analyser and the raw sequence data were edited using Chromas software V2.6.5 (Technelysium, Queensland, Australia) and then subjected to the National Centre for Biotechnology Information's (NCBI) Basic Local Alignment Search Tool (BLAST) function for identity confirmation of the amplicons. This was followed by a multiple sequence alignment of study samples -V172, V193 and V208 -with strains of known mutations, LA141 (HF572938.1), LA088 (HF572933.1) and LA202 (HF572946.1), and a strain that does not include the mutations, that is, M. genitalium G37 complete genome (L43967.2), using ClustalW (https://www.genome. jp/tools-bin/clustalw).

Confirmatory assay -Allplex™ MG & AziR assay (Seegene)
The Allplex™ MG & AziR assay (Seegene, Seoul, South Korea) is a commercially available assay which allows for the simultaneous detection and identification of M. genitalium and six mutations (A2058C, A2058G, A2058T, A2059C, A2059G and A2059T) responsible for azithromycin resistance on validated sample types, which include genital swabs, urine and liquidbased cytology samples. The testing was performed as per the kit instructions and the samples were run on the Bio-Rad CFX96 equipped with Seegene interpretative software.

Statistical analyses
The statistical analysis was conducted using R software (version 3.6.1), a freely available software environment for statistical computing. Initially, the population characteristics were described using frequencies stratified by infection status of the STI investigated. All the tests were conducted at 5% level of significance.

Ethical consideration
Ethical approval to conduct the study was obtained from the Biomedical Research Ethics Committee (BREC), University of KwaZulu-Natal (UKZN), (reference no. BE214/17).

Characteristics of the population according to Mycoplasma genitalium status
Out of 221 samples, 13 tested positive for M. genitalium using the TaqMan assay. The prevalence of M. genitalium in the study population was 5.9%. The median age of all 13 of the women who tested positive for M. genitalium was 26 years (interquartile range [IQR] 21-35). The analysis showed no significant associations (p > 0.05) between socio-demographic, behavioural and clinical data with reference to M. genitalium status (Table 1). Despite the lack of statistical significance, more than 60% of the women who tested M. genitalium positive did not present with symptoms of abnormal vaginal discharge on the day of enrolment. Similarly, more than 75% of the women did not experience previous symptoms of STIs 3 months prior to the enrolment. All women who tested positive for M. genitalium (n = 13) were unmarried. More than half of these women (53.8%) had reported having between two and four life-time sexual partners. A larger proportion of the positive women (53.8%) had indicated that either their current sexual partners had no other partners, or they were not aware of their partners having other partners. With regard to behavioural practices, 84.6% of the M. genitalium positive women had not used a condom during their last sexual act. Clinically, the majority of women who tested positive were in the third trimester of pregnancy (53.8%).

Detection of macrolide resistance-associated mutations in the 23S ribosomal ribonucleic acid gene Screening assay: Conventional polymerase chain reaction
The 147 bp fragment associated with the 23S rRNA gene was detected in 8 out of 13 (61.5%) of the M. genitalium positive isolates analysed. Sanger sequencing confirmed that the PCR amplicons obtained were the 23S rRNA gene from M. genitalium (Accession number: Mk411362.1). Samples which produced good sequencing reads were aligned with strains with known mutations as well as a strain which lacked the mutations. According to the multiple sequence alignment, positions of known mutations associated with macrolide resistance, that is, A2058G, A2059C and A2059G, were not present in the study samples analysed. Similarly, these mutations were not present in the G37 strain (strain lacking the mutations) ( Figure 1). However, as expected, the A2058G mutation was present in strain LA141, the A2059G mutation was present in strain LA088 and the A2059C mutation was present in strain LA202 (Figure 1). The prevalence of macrolide resistant M. genitalium in our study population was 0%.

Confirmatory assay -Allplex™ MG & AziR assay (Seegene)
The Allplex™ MG & AziR assay produced results for 11 out of 13 samples tested (84.61%) showing that it is much more sensitive than the screening assay (conventional PCR). The conventional PCR detected the 23S rRNA in 61.5% of the samples tested (Table 1-A1). Two of the study samples V014 and V069 did not produce a positive result for the presence of M. genitalium on the Allplex™ MG & AziR assay despite testing positive on the TaqMan assay and yielding the 147 bp amplicon. The discrepancy between the assays may have been the result of loss of sample integrity (as a result of the freeze-thawing process) or loss of sample volume during shipment.
However, the Allplex™ MG & AziR assay confirmed that the prevalence of macrolide resistant M. genitalium in our study population was 0%. None of the study samples tested positive for any of the six mutations (A2058C, A2058G, A2058T, A2059C, A2059G and A2059T) responsible for azithromycin resistance (Table 2-A1). However, all assay controls yielded the expected results indicating that the results were valid.

Discussion
To the best of our knowledge, this is the first study to provide an estimate on the prevalence of M. genitalium in pregnant women from the Durban area in KZN, South Africa. We report a prevalence estimate of 5.9% for M. genitalium. Our data are consistent with previous reports on M. genitalium infection in South African pregnant women. 18 18 In our study, the factors that could be attributed to this prevalence are as follows: 67.4% of the study women did not use condoms during their last sexual act; 61.5% were not living with their regular partner; 70.1% were unaware if their partner had other partners and 75.1% had first experienced sex between 15 and 20 years of age. In addition, 51.1% had between two and four life-time sex partners and 43.4% were previously treated for STIs. Many of these factors have been shown to contribute to the prevalence of sexually acquired infections. 27,28,29,30 The emergence of macrolide resistance in M. genitalium has drawn recent attention. 18,20,21,22,23 The first study to report on macrolide resistance in M. genitalium in South Africa was conducted by Hay et al. in the Limpopo province of rural South Africa in 2015. 20 In that study, the authors reported detecting macrolide resistance-associated mutations in 4 out of 41 (9.8%) of M. genitalium positive isolates obtained from    21 Similarly, Le Roux et al. (2018) reported no patterns of macrolide resistance amongst their M. genitalium positive isolates obtained in 2012 from pregnant women in Pretoria, South Africa. 18 However, in that same study, the authors reported a macrolide resistance-associated mutation (A2059G) amongst their M. genitalium positive isolates obtained in 2016. The same mutation was present in 25% of their isolates tested. There is a possibility that the A2059G mutation is prevalent in South African populations; however, this needs further investigation to provide conclusive evidence.
Despite the lack of resistance to macrolides in the studied population, continued antimicrobial resistance surveillance for M. genitalium in pregnant women will be important since azithromycin is now part of the South African national STI syndromic management guidelines for vaginal discharge syndrome.
The present study was limited in that samples were collected from pregnant women attending a single antenatal facility. However, the hospital from which the women were sampled in this study serves as a central hospital for women from around the Durban area, thereby making the population more Note: The red block shows the positions of the mutations.  generalised. The strength of this study is that we now provide prevalence data as well as data on macrolide susceptibility of M. genitalium in a population of pregnant women from KZN, an area of research that has received limited attention.