Systematic identification of knowledge gaps in whole-genome sequence analysis of multi-resistant thermotolerant Campylobacter spp

Background: Campylobacter spp. is the most frequent cause of bacterial food-borne gastroenteritis and a high priority antibiotic resistant bacterium according to the World Health Organization (WHO). European monitoring of thermotolerant Campylobacter spp. does not reflect the global burden of resistances already circulating within the bacterial population worldwide. Methods: We systematically compared whole genome sequencing with comprehensive phenotypic antimicrobial susceptibility, analyzing 494 thermotolerant Campylobacter poultry isolates from Vietnam and Germany. Any discrepancy was checked by repeating the wet lab and improving the dry lab part. Selected isolates were additionally analyzed via long-read Oxford Nanopore technology, leading to closed chromosomes and plasmids. Results: Overall, 22 different resistance genes and gene variants (e. g. erm (B), aph(3’)-IIIa , aph(2'')-If , catA , lnu (C), bla OXA , sat4 ) and point mutations in three distinct genes ( gyrA , 23S rRNA, rpsL ) associated with AMR were present in the Campylobacter isolates. Two AMR genes were missing in the database and one falsely associated with resistance. Bioinformatic analysis based on short-read data partly failed to identify tet (O) and aadE , when the genes were present as duplicate or homologous gene variants. Intriguingly, isolates also contained different determinants, redundantly conferring resistance to gentamicin, kanamycin, streptomycin, chloramphenicol and lincomycin. We found a novel inactive tet (W) and analysis based on assemblies from short-read data was impaired to identify full-length aad9 , which was apparently phase variable. One German isolate contained a yet unexplained gentamicin resistance. GyrT86I led to a rare atypical phenotype of ciprofloxacin resistance but nalidixic acid sensitivity. Long-read sequencing revealed AMR gene localization occasionally on plasmids but mainly on the chromosome, which was frequently inconsistent with predictions from short-read sequencing. AMR genes were often organized in multidrug resistance islands (MDRI) and partially located in proximity to transposase genes, suggesting main mobilization of resistance determinants via natural transformation and transposition in Campylobacter . Conclusions: The revealed gaps of knowledge suggest consideration of frequent duplicate and mosaic genes, gene mutations leading to (transiently) truncated proteins and gene variants missing in databases. Furthermore, there is a need for deciphering yet unknown resistance mechanisms and resistance spread in thermotolerant Campylobacter that may pose a challenge to global food safety.