Emergence of rare carbapenemases (FRI, GES-5, IMI, SFC, and SFH-1) in Enterobacterales isolated from surface waters in Japan

Objectives Carbapenemase-producing Enterobacterales (CPE) pose serious threats to public health. Compared with clinical CPE, the genetic characteristics of environmental CPE are not well understood. This study aimed to characterize the genetic determinants of carbapenem resistance in CPE isolated from environmental waters in Japan. Methods Eighty-five water samples were collected from rivers and a lake in Japan. CPE were identified using selective media, and genome sequencing was performed for the obtained isolates (n = 21). Results Various rare/novel carbapenemases were identified: GES-5 in Raoultella planticola (n = 1), FRI-8 and FRI-11 in Enterobacter spp. (n = 8), IMI-22 and IMI-23 in Serratia ureilytica (n = 3), and SFC-1, SFC-2 and SFH-1 in Serratia fonticola (n = 9). Genomes of 11 isolates could be closed, allowing the elucidation of the genetic contexts of the carbapenemase genes. The blaGES-5 gene was located within a class 1 integron, In2071 (cassette array, blaGES-5-aacA3-aadA16), on a 33 kb IncP6 plasmid. The blaFRI-8 genes were carried on IncFII(Yp) plasmids ranging in size from 191 kb to 244 kb, and the blaFRI-11 genes were carried on 70 kb and 74 kb IncFII(pECLA)/IncR plasmids. The blaIMI-22 and blaIMI-23 genes were colocated on a 107 kb plasmid. The blaSFC and blaSFH-1 genes were found on putative genomic islands inserted at tRNA-Phe genes in chromosomes. Conclusions This study revealed the presence of rare/novel carbapenemases among CPE in aquatic environments, suggesting that the environment may act as a potential reservoir of these minor carbapenemases.


40
The worldwide spread of carbapenemase-producing Enterobacterales (CPE) is a major threat 41 to public health. CPE can be resistant to carbapenems and many other classes of commonly    143 Twelve isolates and nine isolates were obtained from lake water samples and river water 144 samples, respectively ( Table 1). Lake isolates were identified as Enterobacter spp. (n = 8),

145
Serratia ureilytica (n = 3), and Raoultella planticola (n = 1), and all nine river isolates were 146 identified as Serratia fonticola clade III (discussed in detail below) by ANI and dDDH analyses 147 ( Table 1). 148 Among the 21 isolates, 12 were subjected to both DNBSEQ short-read sequencing and ONT 149 long-read sequencing. Nine genomes were completed by Unicycler, and two of the remaining 150 three genomes were completed by Flye. The genome of JBIWA006 could not be completed by 151 either approach. The completed genomes contained one to 12 plasmids (see Table S2 for details). The draft genome assemblies of the remaining isolates had 28 to 217 contigs and 153 189,689 bp to 2,472,998 bp N50 values after discarding contigs shorter than 100 bp (see Table   154 S3 for details).

183
IS1326, tniBΔ2, tniA, and IRt (inverted repeat at tni end). A BLASTN search against the 184 GenBank database using pJBIWA001_5 as a query sequence identified some closely related 185 plasmids. pN260-3, which was carried by an E. roggenkampii strain detected in a patient with 186 cholecystitis in Japan in 2019, 38 showed the highest query coverage (97%) and also showed the 187 high percent identity (>96%) (see Figure S1A for the comparison of these two plasmids).
The Enterobacter genus comprises 22 species and 22 additional genomospecies. 39 The lake

192
Enterobacter isolates in the present study were identified as E. asburiae (n = 4), E.
193 quasiroggenkampii (n = 1), and Enterobacter genomospecies 8 (n = 1) based on the ANI and 194 dDDH analyses (also see Figure S2 for the Enterobacter phylogenetic tree JBIWA009, while a pJBIWA008_3-type genetic context was detected in JBIWA010.   Although three Serratia isolates recovered from the lake could be classified as S. ureilytica, 264 nine Serratia isolates from rivers could not be classified based on the ANI and dDDH analyses.

265
The pairwise ANI values and dDDH values among the nine river isolates were all above 96% and 70%, respectively, indicating the river isolates belong to a single species. The river isolates 267 seem to be related to S. fonticola (Figure S3), but the ANI and dDDH values between the river 268 isolates and the type strain of S. fonticola ranged from 95.5% to 95.7% and 65.1% to 65.4%, 269 respectively, which are lower than the 96% ANI and 70% dDDH cutoff values. These results 270 suggest the 9 river isolates could belong to a novel species whose closest relative is S. fonticola.

271
It should be noted that genome sequences for type strains were not available for some In order to gain further insight into the phylogenetic characteristics of river Serratia isolates, 278 we downloaded RefSeq assemblies labelled S. fonticola and built a phylogenetic tree ( Figure   279 4, also see Figure S4 for phylogenetic characteristics of these strains in relation to other