(P8) Survey of genomic features putatively contributing to antimicrobial resistance in Chlamydia trachomatis

Background:
Based on the polymorphism of the ompA gene Chlamydia trachomatis (CT) can be classified into 15 main genotypes (A – C, D – K, L1 – L3), which correlate with specific tissue tropisms. Chlamydial infections can be treated with antibiotics such as macrolides (azithromycin), tetracyclines (doxycycline) or fluoroquinolones. Post-treatment CT detection is often related with reinfection or lack of treatment compliance, but antimicrobial resistance (AMR) cannot be excluded. Still, CT-AMR phenotypic screening cannot be routinely performed, supporting the need to elucidate this phenomenon via molecular-based assays. In this study, a novel Next-Generation Sequencing (NGS)-based Multiplex-PCR approach was developed for analyzing target loci potentially driving AMR along with genes that, together with ompA, might infer C. trachomatis genomic backbone.

Methods:
To evaluate 503 CT strains (mainly isolated from anorectal specimens) of the anonymous Portuguese CT collection, two approaches were designed. 1) A nested-PCR protocol to amplify CT 23S rRNA and gyrA loci followed by Sanger sequencing of the amplified products. And, 2) a multiplex-PCR method involving 5 loci, including two potentially driving AMR, (gyrA, parC, the upstream region of CT105, CT442 and pmpH) followed by NGS, bioinformatics (namely, INSaFLU, version 1.5.2) and data analysis (R version 4.0.5).

Results:
No CT strain revealed 23SrRNA mutations in any of the documented positions in the peptidyl-transferase centre (A2057, A2058 and A2059), while four strains revealed a gyrA mutation linked to fluoroquinolone-resistance. The multi-locus approach further allowed inferring the potential genomic backbone and phylogeny, evidencing recombination patterns that warrant future investigation (namely by WGS). Indeed, strains were found with potential L2b genomic backbone that reverted to an L2 ompA-genotype, and one strain with a non-LGV ompA-genotype that displays LGV-like profiles in other loci. Furthermore, a putative LGV-specific indel was determined to be highly specific for LGV/non-LGV CT strain discrimination.

Conclusion:
Resistance to macrolides, specifically azithromycin, seems still uncommon amongst CT clinical strains and even fluoroquinolone-resistant mutants seem scarce.
ompA-genotyping nomenclature alone might not always reflect the genomic backbone of strains, nor their phylogenetic placement. On this behalf, the here described NGS-based Multiplex approach features to complement it, putatively potentiating enhanced molecular epidemiology studies of CT infections.