Carriage of the zoonotic organism Streptococcus suis in chicken flocks in Vietnam

Streptococcus suis infections are an emerging zoonotic agent causing severe disease in humans and a major pig pathogen worldwide. We investigated the colonization of S. suis in healthy chickens in different flocks (n = 59) as well as in‐contact pigs in farms with S. suis‐positive chickens (n = 44) in the Mekong Delta of Vietnam. Streptococcus suis was isolated from 20 (33.9%) chicken flocks and from all pigs investigated. Chicken isolates formed a distinct genotypic cluster compared with pig and human strains, although two chicken isolates (10%) clustered with pig isolates. Chicken isolates had unusually high levels of resistance against tetracycline (100%), clindamycin (100%) and erythromycin (95%); and intermediate resistance against penicillin (35%) and ceftriaxone (15%). Our findings suggest that chickens may potentially represent a source of S. suis infection to in‐contact humans and pigs.


| RE SULTS
Of the 59 chicken farms investigated, 10 (16.9%) raised both pigs and chickens; the median distance between the chicken house/pen and the nearest pig pen/s was 10 m . In 29 (49.2%) farms, pigs were not raised in the farm itself, but there were nearby pig farms (located within 1 km). The remaining 20 flocks (33.9%) were located >2 km from a pig farm. A total of 20/59 (33.9%) chicken and 44/44 (100%) pig samples (from the 10 farms raising both pigs and chickens) yielded S. suis (20 chicken and 160 pig isolates). Of the 160 pig isolates, 39 (24.4%) were isolated from 23 pigs in the 10 farms where S. suis was also isolated from chickens. There was no difference in the prevalence of S. suis in chickens from pig-raising farms (5/10; 50%) and those in farms without pigs (15/49; 30.6% if nearby and remote farms combined; overall p = .26; Table 1).
None of the 59 (20 chicken, 39 pig) isolates investigated were serotype 2. PCA of whole-genome sequence data (available in ENA BioProject PRJNA626534) showed that the majority of chicken isolates formed a distinct population compared with pigs, and there was less overall diversity among chicken isolates. Despite this, two chicken isolates grouped near pig isolates (Figure 1a). Our chicken isolates formed a distinct cluster from pig and human isolates reported previously ( Figure 1b). Furthermore, the PCA distance between chicken and pig isolates in same farm was similar to the distance between chicken and pig isolates from different farms ( Figure 1c), suggesting that direct contact with pigs is not the main source of infection to chickens.

Impacts
• This paper reports a high prevalence of infection with Streptococcus suis (33.9%) in chickens in the Mekong Delta of Vietnam.
• Results from this study should help increase awareness of good farming practices to limit transmission of S. suis from animals to humans.  (Table 2). Overall, we did not find evidence of differences in the prevalence of multidrug resistance between chicken (95%) and pig isolates (87%) (p = .63).
Chicken isolates were fully resistant against three of 10 anti-microbials tested, in contrast to pig isolates (eight anti-microbials). Notably, chicken isolates had higher levels of resistance against levofloxacin than pig isolates, but this should be interpreted cautiously since these analyses ignored farm clustering. In one of the two chicken-pig clusters (cluster 1, Figure 1a), isolates from both species had identical AMR profile although chicken and pig isolates came from different farms. In contrast, AMR profile of pig isolates had considerable diversity even if they clustered together (data not shown), suggesting that AMR profiles are probably driven by differences in plasmid composition. No significant associations between AMU in flocks and phenotypic AMR in chicken S. suis were found (Table A1).  (Table A2, summarized in Table A3). Compared to pig and human isolates, chicken isolates had a significantly higher prevalence of resistance against penicillin (p-values = .02 and <.01, respectively), clindamycin (both p-values < .01) and erythromycin (both p-values < .01).

| D ISCUSS I ON
The relatively high prevalence of S. suis (~34%) in chickens even in the absence of pigs demonstrates that chickens can be a selfsustaining reservoir of S. suis. The prevalence of AMR in chickens was higher than observed in human patients, notably with regard to penicillin, ceftriaxone, erythromycin and clindamycin, which is of concern, given that β-lactams (including penicillin and ceftriaxone) are currently the most effective anti-microbials to treat human infections (Dutkiewicz et al., 2017). The lack of association between AMU in flocks and the observed profiles suggests that chickens may not harbour these organisms long-term and may acquire S. suis

ACK N OWLED G EM ENTS
We would like to thank the staff at the Sub-Department of Animal 110085/Z/15/Z).

CO N FLI C T O F I NTE R E S T
The authors have no conflicts of interest to declare.