Molecular Characterization of Antimicrobial Resistance Genes in Neisseria gonorrhoeae Isolates from Selected Regions in Kenya through Whole Genome Sequencing

Abstract

Gonorrhea, a sexually transmitted infection is caused by an obligate gram-negative bacterium called Neisseria gonorrhoeae (N. gonorrhoeae). N. gonorrhoeae infect various parts of the body and when left untreated, gonoccocal infections can lead to infertility in women and sterility in men. With the absence of a gonorrheal vaccine, different antibiotics have been used to treat gonorrhea since 1930’s. While initially sensitive to newly introduced antibiotics, some strains develop single or multiple mechanisms of resistance over time. In Kenya, a few studies have reported resistance to penicillins, tetracyclines, and recently to fluoroquinolones. Consequently, information on antimicrobial resistance (AMR) in Kenyan gonococci is limited. Drug resistant N. gonorrhoeae isolates have been observed in a sexually transmitted infection (STI) surveillance study in Kenya. The underlying molecular mechanisms causing this antimicrobial resistance had not been determined. Therefore, the aim of this study was to determine the molecular mechanisms of antimicrobial resistance as well as the genetic diversity of N. gonorrhoeae isolates obtained from Kenya under the STI surveillance program through whole genome sequencing. Forty one archived N. gonorrhoeae isolates exhibiting varying antibiotic resistance profiles were obtained and sub-cultured on Modified Thayer-Martin (MTM) agar medium. DNA was extracted and the whole genomes sequenced using the Illumina Miseq platform. Sequence reads were assembled de novo using CLC Genomics Workbench. Contigs were mapped to FA1090 reference and searched against the NCBI nucleotide database. Genome annotation was performed using Rapid Annotation Subsystem Technology (RAST). Phylogenetic analyses were performed using both Bayesian Evolutionary Analysis Sampling Trees (BEAST) and core genome multi-locus sequence typing (cgMLST). Sequence typing was done using both N. gonorrhoeae multi antigen sequence typing (NG-MAST) and multi-locus sequence typing (MLST). Mutations and amino acid alterations were identified using Bioedit sequence alignment editor. Provided antimicrobial susceptibility results were interpreted with reference to European Committee on Antimicrobial Susceptibility Testing standards (EUCAST) version 8.0, 2018 standards. Core genome phylogeny revealed five distinct clusters among the study isolates all characterized by varied MLST and NG-MAST sequence types (STs). Region based clustering was not observed. Temporal Bayesian phylogenetic tree clustered Kenyan isolates closely together into distinct lineages with USA being the most probable ancestral location. A total of 25 MLSTs STs were identified. Of the 41 sequences, 26 belonged to 14 known MLST STs of which ST-1932 was the most common (12.2%), while 15 belonged to 11 new MLSTs STs. Thirty three NG-MAST STs were identified. Thirty four sequences belonged to 28 novel NG-MAST STs while 6 belonged to 5 previously reported STs. A new NG-MAST, ST-19168 was the most common (10%). Twenty five isolates had both β-lactamase (TEM) and TetM encoding plasmids whereas 3 isolates lacked either of the plasmids. TEM encoding plasmids were identified in 29 isolates (penicillinase producing N. gonorrhoeae; PPNG) of which 28 had an African type origin (pDJ5) while 1 had an Asian type origin (pDJ4). Two genotypes of β-lactamase were identified; TEM-1 and a recently described TEM-239. Mosaic PenA patterns associated with ceftriaxone and cefixime resistance was not observed in the present study. All analyzed isolates had non mosaic penA alleles. PonA L421P was identified in 19 isolates. Of the 41 isolates, 34 had TetM encoding plasmids, 33 of which had American TetM determinants, whereas 1 had a Dutch TetM determinant. All isolates had S10 V57M mutation. Isolates expressing TEM and tetM had significantly higher penicillin, tetracycline and doxycycline minimum inhibitory concentrations (MICs) respectively. Double GyrA amino acid substitutions; S91F and D95G/A were identified in 39 of the 41 isolates. All ciprofloxacin resistant isolates expressed these GyrA alterations. Of these 39 isolates 29 had an additional E91G or S87R ParC amino acid substitution and significantly high ciprofloxacin MICs. No genetic determinants specifically associated with azithromycin, spectinomycin and gentamycin resistance were identified. R228S substitution in dihydropteroate synthase associated with sulphonamide resistance in gonococci was identified in 40 of the 41 isolates. G120D/N, A121G/S, N122K, and N deletion at position 122 alterations in PorB were identified in 26 isolates. Two isolates had mutated mtrR promoters while G45D, T86A, D79N, H105Y and A39T substitutions were identified in MtrR of 38 isolates. No significant relation was observed between identified PorB, MtrR and PonA alterations and antibiotic MICs. The observed high numbers of different MLST and NG-MAST STs indicate that the sampled gonococci are genetically diverse. The observed non-regional distribution of both MLST and NG-MAST STs indicate a heterogeneous gonococcal population in Kenya. Kenyan gonococci are closely related and have evolved into distinct lineages when compared to other global gonococcal strains. High penicillin and tetracycline resistance in the analyzed gonococcal isolates is mainly mediated by plasmid-borne blaTEM, and tetM genes. The observed high ciprofloxacin resistance and high norfloxacin MICs is mediated by GyrA S91F, D95G/D95A and ParC E91G or S87R substitutions while the observed low level azithromycin resistance could be caused by reduced drug accumulation. From the findings of this study, the combination of azithromycin and ceftriaxone recommended by both Centre for Disease Control and Kenya Ministry of Health is still useful for treatment of gonoccoacal infections in Kenya, while cefixime, gentamycin and spectinomycin are suitable alternatives. The findings further provide current genotypic information on circulating gonococcal strains in Kenya which is useful in planning of containment and gonorrhea treatment measures. The sequence data generated in this study will allow comparative studies between Kenyan gonococci and other global isolates as well as characterization of other molecular markers in the study isolates.