Rates of fluoroquinolone resistance in domestically acquired Campylobacter jejuni are increasing in people living within a model study location in Canada

: Antimicrobial resistance was evaluated in Campylobacter jejuni isolated from 1,291 diarrheic people 2 over a 15 year period (2004-2018) in Southwestern Alberta, a model location in Canada with a high rate of 3 campylobacteriosis. The prevalence of resistance to chloramphenicol, clindamycin, erythromycin, and 4 gentamicin was low during the examination period (≤4.8%). Resistance to tetracycline remained consistently 5 high (41.6-65.1%), and resistance was primarily conferred by plasmid-borne tetO (96.2%). Resistance rates to 6 ciprofloxacin and nalidixic acid increased substantially over the examination period, with a maximal 7 fluoroquinolone resistance (FQR) prevalence of 28.9% in 2016. The majority of C. jejuni isolates resistant to 8 ciprofloxacin (93.9%) contained a C257T mutation within the gyrA chromosomal gene. Follow up with infected 9 people indicated that the observed increase in FQR was primarily due to domestically acquired infections. Moreover, the majority of FQR C. jejuni subtypes (82.6%) were endemic in Canada, primarily linked to cattle 11 and chicken reservoirs; 18.4% of FQR isolates were assigned to three subtypes, predominantly associated with 12 cattle. Study findings indicate the need to prioritize FQR monitoring in C. jejuni infections in Canada, and to 13 elucidate the dynamics of the emergence and transmission of resistant C. jejuni strains within and from cattle 14 and chicken reservoirs. [ 197 ]

147 Genetics of fluoroquinolone resistance D r a f t Inglis,Taboada,Boras Canadian Journal of Microbiology Page 8 179 analyzed and visualized using the BioCalculator v3.2 software (Qiagen Inc.), and converted to binary values (i.e. 180 where a '0' represented the absence of the ancillary gene locus, and a '1' indicated its presence).

Data analysis
D r a f t Inglis,Taboada,Boras Canadian Journal of Microbiology Page 9 211 Resistance to tetracycline but not to clindamycin, chloramphenicol, erythromycin, or gentamicin was 212 common in Campylobacter jejuni D r a f t Inglis,Taboada,Boras Canadian Journal of Microbiology Page 10 243 observed increase in infections with C. jejuni resistant to ciprofloxacin was due to infections that occurred 244 within Canada.
245 Resistance to ciprofloxacin corresponded to mutations in the Campylobacter jejuni gyrA gene
248 jejuni isolates that did not carry the C257T mutation all carried other SNPs in the gyrA gene (n=2 for A21G; n=2 249 for A64G; n=4 for T72C; n=6 for C243T; n=9 for T357C; n=5 for C360T; n=1 for C408A; n=7 for C471T; and n=5 250 for T483C). None of the 12 C. jejuni isolates that possessed intermediate resistance to ciprofloxacin (i.e. MIC of 251 2 µg/mL) carried the C257T mutation, nor did any of the 46 C. jejuni isolates that were susceptible to 252 ciprofloxacin (i.e. MIC ≤1 µg/mL). All the C. jejuni isolates that carried the Thr86Ile mutation were also resistant 253 to nalidixic acid. There was no relationship between the number of mutations in the gyrA gene with subtype 254 (Fig. S3). In resistant isolates not exhibiting the Thr86Ile mutation in the gyrA gene, plasmid-mediated 255 quinolone resistance was not detected, and a mutation in the cmeABC operon indicative of an efflux pump 256 mechanism of resistance was observed in only one C. jejuni strain isolated from a human in 2006 (data not 257 shown). Nonetheless, the likely mechanism of FQ and/or quinolone resistance could be ascertained for the 258 majority of resistant isolates, with only a small number of Thr86Ile mutation-negative isolates undefined. 273 ciprofloxacin (Fig. 3). This contrasted with C. jejuni subtypes resistant to tetracycline, which were more widely 274 distributed (Fig. 4). Although the subtype map of isolates resistant to nalidixic acid was similar to ciprofloxacin 275 (Fig. S5), the diversity of subtypes was higher for nalidixic acid (H=4.61; P=0.010) and for tetracycline (H=5.23; D r a f t Inglis,Taboada,Boras Canadian Journal of Microbiology Page 11 276 P<0.001) as compared to ciprofloxacin (H=4.24) (Fig. S6). In total, 656 CGF subtypes were identified (100% 277 resolution), and 118 subtypes contained isolates that were resistant to ciprofloxacin (18.0%) (Fig. 5 375 from genomic information (Zhao et al. 2016). In addition to the C257T mutation, we observed a trend for a 376 larger number of single nucleotide polymorphisms within the gyrA gene for C. jejuni isolates exhibiting a MIC of 377 ≥32 µg/mL ciprofloxacin (i.e. ranging from 3.8 to 5.2 SNP mutations). Of these SNP mutations, only one confers 378 an amino acid substitution (i.e. A64G). Although the mechanisms are not well understood, additional 379 mutations within the gyrA gene have been reported to be associated with increased FQR (Wieczorek and Osek 380 2013). We did not investigate this in FQR isolates, and a mutation in the intergenic cmeR-cmeABC intergenic 381 region in conjunction with the C257T mutation within the gyrA gene has previously been linked to higher MIC 382 values (Yang et al. 2017). An examination of the ten FQR C. jejuni isolates that did not possess the C257T 383 mutation in the current study, revealed that none of these isolates exhibited PMQR or mutations in the 384 cmeABC operon; CmeABC is a resistance-nodulation-division type of efflux pump that contributes to resistance 385 to a variety of antimicrobial agents in Campylobacter . Thus, the mechanism conferring FQR in 386 these isolates was not identified.

387
To determine if the increasing FQR rates observed in SWA were due to internationally acquired infections, 388 infected individuals were interviewed following the standard protocol used by AHS; in Canada 389 campylobacteriosis is a reportable disease mandating follow up with afflicted individuals. Examination of travel 390 information indicated that international travel did not disproportionately contribute to the increased FQR rates 391 that we observed, and that increased resistance due to domestically acquired infections was responsible, 395 ) in order to place resistant genotypes in the context of the wider population structure. It 396 is noteworthy, the CGF method was found to be more discriminatory than MLST for the detection of clusters of 397 C. jejuni cases , and the CGF method has previously been used to elucidate the molecular 408 large quantities in their feces (Inglis et al. 2004 World-Health-Organization. 2017. WHO publishes list of bacteria for which new antibiotics are urgently needed. https://www.who.int/news-room/detail/27-02-2017-who-publishes-list-of-bacteria-for-whichnew-antibiotics-are-urgently-needed. Wu, N., Qiao, M., Zhang, B., Cheng, W.D., and Zhu, Y.G. 2010. Abundance and diversity of tetracycline resistance genes in soils adjacent to representative swine feedlots in China. Environ Sci Technol 44 (  Grey circles indicate C. jejuni isolates recovered from individuals who were infected outside of Canada and the USA (i.e. traveler's diarrhea), and white squares indicate C. jejuni isolates that could not be designated as either domestically or internationally acquired. Histogram bars denoted with an asterisk were higher (P≤0.050) than the prevalence of resistance observed in 2004. Solid lines indicate the linear relationship between time and resistance, and dotted lines are 95% confidence limits.

Fig. 2. Relative sensitivities of Campylobacter jejuni isolates recovered from diarrheic people in Southwestern
Alberta to ciprofloxacin during three 5-year time periods (2004-2008, 2009-2013, and 2014-2018). The arrow indicates the breakpoint minimum inhibitory concentration, and asterisks indicate conspicuous increases in resistance.

Fig. 3.
Minimum spanning tree of Campylobacter jejuni comparative genomic fingerprinting subtypes recovered from diarrheic people (2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018) showing subtypes possessing isolates resistant to ciprofloxacin. Black fill denotes resistant isolates, grey fill indicates isolated possessing intermediate resistance, and white fill denotes susceptible isolates. The thickness of lines connecting subtypes represent mismatched loci (i.e. one to three loci), and subtypes with no line represent ≥ four mismatched loci between respective subtypes. Stars indicate major subtypes recovered from diarrheic people in Southwestern Alberta that are associated livestock reservoirs, and white asterisks within markers indicate subtypes attributed to internationally acquired infections. Highlighted clusters A to J illustrate prominent clusters (95% level of resolution) that contain subtypes resistant to ciprofloxacin, where orange highlighting indicates clusters containing resistant subtypes primarily associated with cattle reservoirs, blue highlighting indicates clusters containing resistant subtypes associated with both cattle and chickens, and green highlighting indicates clusters with an accumulation of subtypes associated with traveler's diarrhea (see Table 2 for subtype-specific information).  D r a f t a MIC, minimum inhibitory concentration determined by dilution plating (µg/mL); n, number of isolates; Avg, average number of SNP mutations. b Two ciprofloxacin resistant isolates were not characterized for SNP mutations due to poor amplification of the gryA gene. c For SNP nomenclature the first letter represents the non-mutated nucleotide, the number represents the nucleotide position in the GyrA gene, and the second letter represents the mutated nucleotide. d Resulted in an amino acid substitution (Ser22Gly), where Ser is serine, 22 is the amino acid position, and Gly is glycine. e Resulted in an amino acid substitution (Thr86Ile), where Thr is tyrosine, 86 is the amino acid position, and Ile is isoleucine.
D r a f t   Fig. 3; CipR size, number of isolates within the same CGF subtype resistant to ciprofloxacin. *Dominant animal reservoir. **Primary subtypes infecting people in SWA (primary subtypes are designated with stars in Fig. 3).
D r a f t isolates that could not be designated as either domestically or internationally acquired. Histogram bars denoted with an asterisk were higher (P≤0.050) than the prevalence of resistance observed in 2004. Solid lines indicate the linear relationship between time and resistance, and dotted lines are 95% confidence limits.
D r a f t Fig. 2. Relative sensitivities of Campylobacter jejuni isolates recovered from diarrheic people in Southwestern Alberta to ciprofloxacin during three 5-year time periods (2004-2008, 2009-2013, and 2014-2018). The arrow indicates the breakpoint minimum inhibitory concentration, and asterisks indicate conspicuous increases in resistance.
D r a f t Fig. 3. Minimum spanning tree of Campylobacter jejuni comparative genomic fingerprinting subtypes recovered from diarrheic people (2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018) showing subtypes possessing isolates resistant to ciprofloxacin. Black fill denotes resistant isolates, grey fill indicates isolated possessing intermediate resistance, and white fill denotes susceptible isolates. The thickness of lines connecting subtypes represent mismatched loci (i.e. one to three loci), and subtypes with no line represent ≥ four mismatched loci between respective subtypes. Stars indicate major subtypes recovered from diarrheic people in Southwestern Alberta that are associated livestock reservoirs, and white asterisks within markers indicate subtypes attributed to internationally acquired infections. Highlighted clusters A to J illustrate prominent clusters (95% level of resolution) that contain subtypes resistant to ciprofloxacin, where orange highlighting indicates clusters containing resistant subtypes primarily associated with cattle reservoirs, blue highlighting indicates clusters containing resistant subtypes associated with both cattle and chickens, and green highlighting indicates clusters with an accumulation of subtypes associated with traveler's diarrhea (see Table 2 for subtypespecific information).