Genomic comparison of carbapenem-resistant Enterobacteriaceae from humans and gulls in Alaska

Objectives: Wildlife may harbour clinically important antimicrobial-resistant bacteria, but the role of wildlife in the epidemiology of antimicrobial-resistant bacterial infections in humans is largely unknown. In this study, we aimed to assess dissemination of the bla KPC carbapenemase gene among humans and gulls in Alaska. Methods: We performed whole-genome sequencing to determine the genetic context of bla KPC in bacterial isolates from all four human carbapenemase-producing Enterobacteriaceae (CPE) infections reported in Alaska between 2013 – 2018 and to compare the sequences with seven previously reported CPE isolates from gull faeces within the same region and time period. Results: Genomic analysis of CPE isolates suggested independent acquisition events among humans with no evidence for direct transmission of bla KPC between people and gulls. However, some isolates shared conserved genetic elements surrounding bla KPC , suggesting possible exchange between species. Conclusion: Our results highlight the genomic plasticity associated with bla KPC and demonstrate that sampling of wildlife may be useful for identifying clinically relevant antimicrobial resistance not observed through local passive surveillance in humans. Published


Introduction
Carbapenems are considered to be antibiotics of last resort for treating multidrug-resistant bacterial pathogens, and resistance to these important antibiotics has become a surveillance target in humans and the environment in recent years.The first Klebsiella pneumoniae carbapenemase (KPC), an enzyme conferring carbapenem resistance, was identified in 1996 in a hospital in the eastern USA [1].The bla KPC gene has since spread globally within healthcare facilities, less commonly at the community level and, rarer still, to wildlife [2,3].This gene has been identified in different Gram-negative bacterial species and is highly mobile.Horizontal gene transfer of bla KPC between bacterial strains, species and even genera can occur through a variety of mechanisms, including transfer of conjugative plasmids.Specifically, the gene is commonly located on an 10-kb Tn4401 transposon, which itself is capable of transposition into different genomic regions [4].
A previous investigation reported the detection of seven carbapenem-resistant Escherichia coli isolates from gull (Larus spp.) faeces collected in Alaska during 2016, including four harbouring the carbapenemase gene bla OXA-48 and three harbouring bla KPC-2 [2].Four human clinical cases of carbapenemaseproducing Enterobacteriaceae (CPE) were also reported in Alaska between 2013-2018, all attributed to K. pneumoniae and reported to harbour bla KPC [5].In the current study, we performed wholegenome sequencing on bacterial isolates from these four human CPE isolates and compared the genomic data with the three bla KPC- 2 -positive E. coli isolates from gulls to assess dissemination of the bla KPC gene within Alaska.

Methods
Whole-genome sequencing was performed using Illumina HiSeq (Illumina Inc., San Diego, CA, USA) and PacBio Sequel (Pacific Biosciences, Menlo Park, CA, USA) platforms on four CPE isolates from humans isolated between 2013-2018 and three CPE isolates from gulls isolated in 2016.Briefly, draft genomes were assembled from short-and long-read sequences.Antimicrobial

Contents lists available at ScienceDirect
Journal of Global Antimicrobial Resistance j o u r n a l h o m e p a g e : w w w .e l s e v i e r .c o m / l o c a t e / j g a r resistance genes, multilocus sequence types and plasmid types were identified in silico, and core genome single nucleotide polymorphisms (SNPs) were used for phylogenetic inference.See Supplemental material for detailed methodology.All genomic data supporting the findings of this investigation are publicly available in the Sequence Read Archive (BioProject PRJNA622828).

Results and discussion
All four human K. pneumoniae isolates were sequence type 258 (ST258), a widespread epidemic clone [6], and harboured either bla KPC-2 or bla KPC-3 .The bla KPC gene was located within Tn4401 transposons, which were found in different genetic backgrounds.Isolates KPC1808800061 and KPC142200207 harboured an IncFII g plasmid in which bla KPC-3 was located within a Tn4401d transposon (Table 1).Isolate KPC1709400028 harboured an IncFII K2 plasmid with two inverted identical Tn4401a transposons, each harbouring bla KPC-3 (Table 1).This plasmid shared >99% DNA sequence similarity to the widespread plasmid pKpQIL [7].Isolate KPC1715200133 harboured bla KPC-2 within a chromosomallyencoded Tn4401a transposon that was inserted into a putative guanine deaminase gene (Table 1).In addition to the different genetic backgrounds of bla KPC and different target site duplication flanking regions, the four human ST258 isolates had different antimicrobial resistance gene profiles (Fig. 1).
In comparison, the three bla KPC-2 positive E. coli isolates from gulls in Alaska harboured bla KPC-2 within Tn4401a transposons on IncC plasmids (Table 1).The 9906-bp Tn4401a transposons had identical sequences to the transposons from two human isolates, with the exception of a single SNP differentiating bla KPC-2 from bla KPC-3 .Furthermore, one K. pneumoniae plasmid from a humanorigin isolate shared a >20-kbp DNA segment with an E. coli plasmid isolated from a gull (Fig. 2).Thus, comparisons among KPC-positive CPE isolates originating from humans and gulls in Alaska revealed varying levels of genetic similarity at discrete genetic loci.
The finding of KPC genes within a highly conserved Tn4401a transposon in different plasmid types and bacterial genera from different hosts highlights the genomic plasticity associated with this resistance gene.Sheppard et al. previously illustrated the dissemination of bla KPC within a single healthcare facility over a period of 5 years and analogised it to a nested Russian doll-like mobility [8].Our results demonstrate comparable nested genetic levels of bla KPC diversity between bacterial isolates from humans and gulls in Alaska, although without direct epidemiological links between hosts.
The diversity of antimicrobial resistance gene profiles and genetic backgrounds of the KPC gene among human-origin multidrug-resistant K. pneumoniae isolates in Alaska is congruent with independent acquisition of CPE, which likely occurred outside of Alaska given that all patients had previously been hospitalised outside of the state prior to diagnosis [5].Similarly, given the migratory connectivity of birds from Southcentral Alaska with areas within the Pacific Northwest and California and the tendency of gulls to inhabit landfills during the non-breeding season [9], gulls may have acquired bla KPC outside of Alaska and maintained it within individuals or within the gull population throughout the breeding and post-breeding period (i.e.2-4 months).This scenario seems plausible given that colistin-resistant E. coli was detected in the environment of experimentally challenged gulls 29 days after they were inoculated with mcr-1-positive E. coli [10].Alternatively, humans and/or gulls may have acquired CPE through direct or indirect routes from local healthy carriers (see Ref. [11]) or out-ofstate visitors, given incomplete information on community carriage of antimicrobial resistance within Southcentral Alaska and the limited scope of research and surveillance for CPE within this region.
We find it striking that only four CPE isolates have been reported in humans in Alaska over a 6-year period and seven gull   faecal samples were positive for CPE in a single year.We recognise that the number of isolates compared in this study is low, however we compared all isolates reported in a state with minimal agriculture, relatively low human population density, and where gulls are presumed to acquire antimicrobial-resistant bacteria from human sources.Wildlife may serve as indicators of clinically important antimicrobial resistance in the environment [12] and may be associated with human-acquired antimicrobial-resistant bacterial infections [13].Through the application of a One Health genomic approach that elucidated the genetic context of bla KPC at multiple genetic levels, we demonstrate that sampling wildlife can provide information about critically important antibiotic resistance not identified through passive surveillance at a similar spatiotemporal scale in the human health sector.

Ethical approval
Not required.

Funding
This work was funded in part by the US Geological Surveythrough the Environmental Health and Wildlife programs of the Ecosystems Mission Area.Any use of trade, firm or product names is for descriptive purposes only and does not imply endorsement by the US Government.

Fig. 1 .
Fig. 1.Core genome midpoint-rooted maximum-likelihood phylogenetic tree of four carbapenem-resistant Klebsiella pneumoniae human isolates harbouring a bla KPC carbapenemase gene (*).Shaded squares in the matrix to the right indicates presence of antimicrobial resistance genes, in which different colours represent different antibiotic classes.

Fig. 2 .
Fig. 2. Genetic background of bla KPC in three Escherichia coli isolates from gulls and four Klebsiella pneumoniae isolates from humans sampled in Alaska.Dotted boxes indicate duplicated inverted sequence and shaded regions indicate identical nucleotide identity (with the exception of a single single nucleotide polymorphism differentiating bla KPC-2 from bla KPC-3 ).Tn4401 transposons (dark blue) with triangles indicating deletions, antimicrobial resistance genes (red), hypothetical proteins (orange) and transposon elements (light blue) are indicated.

Table 1
Genetic context of bla KPC in Enterobacteriaceae isolates from humans and gulls in Alaska.
a Inverted duplicated transposon.