Genetic Environment Surrounding blaOXA-55-like in Clinical Isolates of Shewanella algae Clade and Enhanced Expression of blaOXA-55-like in a Carbapenem-Resistant Isolate

ABSTRACT Although Shewanella spp. are most frequently isolated from marine environments; more rarely, they have been implicated in human infections. Shewanella spp. are also recognized as the origin of genes for carbapenem-hydrolyzing class D β-lactamases. Due to the spread globally among Enterobacterales in recent years, risk assessments of both clinical and environmental Shewanella strains are urgently needed. In this study, we analyzed the whole-genome sequences of 10 clinical isolates and 13 environmental isolates of Shewanella spp. and compared them with those of Shewanella species strains registered in public databases. In addition, the levels of blaOXA-55-like transcription and β-lactamase activity of a carbapenem-resistant Shewanella algae isolate were compared with those of carbapenem-susceptible S. algae clade isolates. All clinical isolates were genetically identified as S. algae clade (S. algae, Shewanella chilikensis, and Shewanella carassii), whereas all but one of the environmental isolates were identified as various Shewanella spp. outside the S. algae clade. Although all isolates of the S. algae clade commonly possessed an approximately 12,500-bp genetic region harboring blaOXA-55-like, genetic structures outside this region were different among species. Among S. algae clade isolates, only one showed carbapenem resistance, and this isolate showed a high level of blaOXA-55-like transcription and β-lactamase activity. Although this study documented the importance of the S. algae clade in human infections and the relationship between enhanced production of OXA-55-like and resistance to carbapenems in S. algae, further studies are needed to elucidate the generalizability of these findings. IMPORTANCE Shewanella spp., which are known to carry chromosomally located blaOXA genes, have mainly been isolated from marine environments; however, they can also cause infections in humans. In this study, we compared the molecular characteristics of clinical isolates of Shewanella spp. with those originating from environmental sources. All 10 clinical isolates were genetically identified as members of the Shewanella algae clade (S. algae, S. chilikensis, and S. carassii); however, all but one of the 13 environmental isolates were identified as Shewanella species members outside the S. algae clade. Although all the S. algae clade isolates possessed an approximately 12,500-bp genetic region harboring blaOXA-55-like, only one isolate showed carbapenem resistance. The carbapenem-resistant isolate showed a high level of blaOXA-55-like transcription and β-lactamase activity compared with the carbapenem-susceptible isolates. To confirm the clinical significance and antimicrobial resistance mechanisms of the S. algae clade members, analysis involving more clinical isolates should be performed in the future.

in the supplemental material). Among the 22 isolates sequenced in this study, 14 formed a cluster with the type strains, and their bacterial species were determined. In addition, four isolates were type strains, and the remaining four isolates were either independent or clustered with other study isolates only: hence, the species could not be determined. Nine isolates were identified as S. algae, and all except one were clinical isolates, including ATCC 49138, a clinical isolate registered as S. haliotis by ATCC. In addition, two isolates respectively identified as S. carassii and S. chilikensis belonging to the S. algae clade were also clinically isolated. Taken together, all clinical isolates analyzed were identified as S. algae clade members, and all but one of the environmental isolates in this study were identified as Shewanella spp. outside the S. algae clade. Core genome SNP-based phylogenetic analysis of S. algae clade isolates. The core genome of the S. algae clade, which covered 60.6% (2,977,854 bp) of the genome of the reference isolate, was used for the single-nucleotide polymorphism (SNP)-based phylogenetic analysis. Isolates of the same species clustered into the same branch of the phylogenetic tree (Fig. 1). SNP differences ranged from 26,452 bp to 173,920 bp (median, 49,369 bp).
Comparison of bla OXA and the surrounding genetic environment. We compared the 25,000-bp genetic regions surrounding bla OXA-55-like among the chromosomes of S. algae isolates. The genetic structures around bla OXA-55-like were almost identical, except for minor differences, which included the presence of a gene for an IS4 family transposase in three of the nine isolates (Fig. 2a). The comparison of a 40,000-bp genetic region surrounding bla OXA-55-like of S. algae clade isolates, including non-algae species (S. chilikensis and S. carassii), is shown in Fig. 2b. These three species shared an approximately 12,500-bp common genetic region around bla OXA-55-like , including the adjacent C15 gene and lysR. However, beyond this shared region, the nucleotide sequences were unique to each species.  (Table S1).
Antimicrobial susceptibilities of S. algae clade isolates and E. coli DH5a carrying bla OXA-55-like from S. algae isolates. Among the S. algae clade isolates, only TUM17384 was nonsusceptible to piperacillin and imipenem, while all other isolates were susceptible to piperacillin, cefotaxime, ceftazidime, aztreonam, imipenem, and meropenem ( Table 2). TUM17377 (S. chilikensis) and TUM17378 (S. carassii) had lower MICs for cefazolin compared to S. algae isolates and also showed low MICs for ampicillin and ampicillin/clavulanic acid.
Comparison of the transcription levels of bla OXA-55-like . TUM17384 had an approximately 200-fold-higher level of bla OXA-55-like transcription than TUM4442 (the reference isolate) (Fig. 3). The remaining isolates had bla OXA-55-like transcription levels of 1.8-to 17.5-fold compared with that of TUM4442.
Comparison of b-lactamase activity in crude enzyme solution. We measured the b-lactamase activity of the crude enzyme solution of S. algae clade isolates. The hydrolytic activities for nitrocefin, benzylpenicillin, oxacillin, and meropenem of the crude enzyme solution of TUM17384, which had a high initial velocity of bla OXA-55-like enzyme, Arrow size is proportional to the predicted ORF length. The common genetic region among three species is enclosed in a square and presented in enlarged form below using S. chilikensis strain JC5 as the representative. The color code is as follows: magenta, LysR family transcriptional regulator; red, bla OXA-55-like ; green, protease; blue, hypothetical protein; purple, membrane protein; lime, transporter; cyan, carbamoyl phosphate synthase large subunit; yellow, IS4 family transposase; gray, others.

DISCUSSION
In this study, we analyzed the whole-genome sequences of Shewanella isolates recovered from clinical or environmental sources and publicly available Shewanella isolates and compared them with genomic data deposited in GenBank. All nine clinical isolates and one publicly available isolate originating from a clinical sample were identified as members of the S. algae clade. Although all isolates in the S. algae clade shared an approximately 12,500-bp genetic region harboring bla OXA-55-like , the genetic structures outside this region differed among the species. Among the S. algae clade, only one S. algae isolate showed carbapenem resistance, and this strain had high levels of bla OXA-55-like transcription and b-lactamase activity.
The difficulties with the correct identification of species of Shewanella are universally recognized. All S. algae clade isolates that had been identified in the clinical setting (those isolated at University of Tokyo Hospital [UTH]), were incorrectly identified by automated instruments and matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS). In addition, some of the isolates' draft whole-genome sequences, which were available in GenBank, were registered under the name of different Shewanella species. These results are expected, because it is difficult to identify Shewanella spp. using automated instruments and 16S rRNA sequencing (1,3). Although misidentification by MALDI-TOF MS has been reported previously (10), it is assumed that this was partly due to the insufficient registration of reference  spectrum data for Shewanella spp. in the database. The MALDI Biotyper Library version 9, which was used at UTH, has data for only four strains of S. putrefaciens and one strain each of S. algae, S. baltica, S. fidelis, S. frigidimarina, and S. profunda. We expect that enrichment of the data will facilitate the correct identification of Shewanella spp. with MALDI-TOF MS (11).
All clinical isolates analyzed in this study were identified as S. algae clade members, and all except one of the environmental isolates were identified as various Shewanella spp. outside the S. algae clade. Although most of the clinical isolates were from a single hospital, some of the isolates were different species (S. chilikensis and S. carassii), and 26,452 or more SNPs were found among the S. algae isolates. This suggests each patient acquired the S. algae clade isolates independently rather than by nosocomial transmission. In the past, human Shewanella species infections were mainly caused by S. algae and S. putrefaciens; however, assuming they have been correctly identified, most cases in recent years appear to have been caused by S. algae (1, 3). The fact that all of our clinical isolates were S. algae clade supports this assumption.
The amino acid sequences in OXA-SHE and OXA-55 sequences were converted from nucleotide sequences obtained from GenBank to amino acid sequences. The GenBank accession numbers of bla OXA-SHE and bla OXA-SHE are AY066004 and AY343493, respectively. b Of 13 OXA-type b-lactamase alignments, this table shows only the positions where amino acid substitution was detected. The amino acid position was counted from the initiation codon from OXA-SHE, including the estimated signal peptide. Boldface indicates the amino acid is different from OXA-SHE. c There were two silent mutations on bla OXA-55-like between TUM17378 and TUM17386. d There were two silent mutations on bla OXA-55-like between TUM4442 and TUM17379 and between TUM4442 and TUM17382. There were four silent mutations on bla OXA-55-like between TUM17379 and TUM17382.
While isolates of the S. algae clade, other than S. algae itself, have been mainly detected in environmental samples, there are sporadic reports of these being isolated from human specimens, including those suspected to contain causative pathogens of infectious disease (12,13). It is unclear whether the clinical characteristics of these species differ from those of S. algae due to the limited number of cases reported so far. However, the identification of an S. chilikensis isolate from clinical samples in this study suggested the ability of this species to cause human infections. We also verified that bla OXA-55-like is shared among the S. algae clade species, but the genetic structure surrounding bla OXA-55-like differs among species, except for in the vicinity of bla OXA-55-like . However, as the genomes of only a limited number of isolates, including those registered in GenBank, have been analyzed, it will be necessary to accumulate more information in the future.
The presence of bla OXA-55-like in S. algae has been repeatedly documented by previous studies (5,6). This study confirmed the universal occurrence of bla OXA-55-like in the chromosomes of S. algae isolates, and the diversity of bla OXA-55-like nucleotide sequences among the isolates is suggestive of the stable long-term persistence of the gene. Although several carbapenem-resistant isolates of S. algae have been reported, the mechanisms and the extent of involvement of bla OXA-55-like in carbapenem resistance are unknown (4,14). In this study, we cloned bla OXA-55-like genes with slightly different nucleotide sequences from nine S. algae isolates and introduced them into E. coli DH5a, but no differences in the MICs of broad-spectrum cephalosporins or carbapenems were observed. Therefore, there was no evidence that the minor differences in the bla OXA-55-like possessed by S. algae isolates have a direct effect on their susceptibility to carbapenems. However, the only carbapenem-resistant isolate showed high levels of bla OXA-55-like transcription and b-lactamase activity, indicating that the increased production of OXA-55like contributes to carbapenem resistance via an unknown mechanism.
Imipenem had a higher MIC value than meropenem in the S. algae clade isolates, but the crude enzyme hydrolysis activity of each S. algae clade isolate for imipenem was not significantly different from that for meropenem. The lower affinity of imipenem for PBP2 of the S. algae clade strains compared with meropenem is believed to have influenced the difference in MIC between imipenem and meropenem (Table 3). However, we have not been able to obtain data on the affinity of imipenem for PBP2. Even though the transcript level of bla OXA-55-like in strain ATCC 49138 was not high (Fig. 3), the strain's benzylpenicillin hydrolytic activity was higher than those of the other strains, except TUM17384, but the amino acid sequence alignment of OXA-55like did not reveal any amino acid substitutions characteristic of OXA-55-like in ATCC 49138 that would explain its hydrolytic activity.
There were several limitations to this study. First, only a small number of clinical isolates were analyzed, and most were collected from a single institution. Although all clinical isolates were S. algae clade members, it was unclear whether this predominance is a general feature of Shewanella species isolated in clinical settings. Because only one carbapenem-resistant isolate of S. algae was analyzed, we are unsure whether the high production of OXA-55-like observed in this isolate is a common characteristic of carbapenem-resistant isolates of S. algae. Second, clinical information on the patients was not available, and the detailed characteristics of the Shewanella infections could not be analyzed. Third, the mechanism that leads to the high production of OXA-55-like in the carbapenem-resistant S. algae isolate was not investigated. If bla OXA-55-like expression is regulated by other genes, differences in the genetic backgrounds of the different S. algae clade species may affect the frequency of carbapenem resistance.
In conclusion, we performed whole-genome sequencing analysis of Shewanella spp. detected in clinical and environmental samples and confirmed the dominance of the S. algae clade in the clinical isolates. In addition, we found that S. algae clade strains share an approximately 12,500-bp genetic region that harbors the gene bla OXA-55-like , but the genetic structures outside this region were different among the different clade species, and the expression of bla OXA-55-like was increased only in the carbapenem-resistant isolate. To confirm the clinical significance and antimicrobial resistance mechanisms of S. algae clade members, analysis involving more clinical isolates should be performed in the future.

MATERIALS AND METHODS
Bacterial isolates. Nine isolates of Shewanella spp. were detected in different patients at the University of Tokyo Hospital (UTH) between November 2014 and August 2016, seven of which were stored at 280°C and used in this study. The strains collected at the hospital did not have any information that could identify the patient, and only the year of isolation was recorded. All isolates were identified as S. putrefaciens with the automated Microscan WalkAway system (Beckman Coulter, Brea, CA, USA) and as S. putrefaciens or Shewanella sp. by retest using the MALDI Biotyper with library version 9 (Bruker Daltonics, Bremen, Germany) at the hospital. Additionally, we used a clinical isolate that had been previously reported but not genetically analyzed (14). In total, nine clinical isolates, each detected from a different patient, from hospitals in Japan were analyzed in this study (Table 1).
In addition, a clinical isolate and 13 environmental isolates of Shewanella spp., provided by institutions in Japan or purchased from the National Collection of Type Cultures, were included in the analysis ( Table 1).
Whole-genome sequencing analysis and identification of bacterial species. Draft whole-genome sequencing analysis of the study isolates was performed with Illumina MiSeq (Illumina, Inc., San Diego, CA), except for S. algae JCM 21037 (=ATCC 51192), for which the draft whole-genome sequence data have already been registered (GenBank accession no. JAAXPX000000000.1). Library preparation, sequencing, and de novo assembly for MiSeq were performed as previously reported (15).
For the purpose of comparison, we collected the whole-genome sequencing data of Shewanella species isolates, including type strains, deposited in GenBank in October 2019. We employed whole-genome sequencing data in which the 16S rRNA gene nucleotide sequence was more than 1,300 bp and that were identified as Shewanella sp. by BLAST search. As a result, whole-genome sequencing data of 20 type strains and 74 other isolates were adopted (Table S1).
The draft genome sequence of the study isolates and registered isolates were compared with fastANI (https://github.com/ParBLiSS/FastANI), and isolates with an average nucleotide identity (ANI) value of 95% or more were clustered (16). If a cluster of isolates included a type strain of a specific species, the isolates within the same cluster were designated species of the type strain.
Additionally, long-read nucleotide sequences were obtained using the MinION sequencer (Oxford Nanopore Technologies, Oxford, United Kingdom) for isolates identified as S. algae clade members (S. algae, S. carassii, or S. chilikensis) to determine the complete whole-genome sequences. DNA extraction, library preparation, sequencing, and de novo assembly for MinION were performed as previously published (17).
Comparison of bla OXA and the surrounding genetic environment. The bla OXA genes in the draft genome sequences of the study isolates were identified and compared with bla OXA reference sequences using ResFinder (version 4.1). The genetic environments surrounding bla OXA-55-like of S. algae clade isolates identified in our analysis were compared with the reference S. algae JCM 21037 (=ATCC 51192) (type strain) using Easyfig (version 2.2.2). Because only one strain each of S. carassii and S. chilikensis was identified in the study isolates, the draft whole-genome sequencing data registered in GenBank for these species (S. carassii, NZ_NGVS00000000.1; S. chilikensis, NZ_MDKA00000000.1 and NZ_NIJM00000000.1) were also included in the analysis.
Core genome single-nucleotide-polymorphism-based phylogenetic analysis of S. algae isolates. Core genome single nucleotide polymorphism (SNP)-based phylogenetic analysis was performed using the complete genome sequence of S. algae TUM17379 (accession no. AP024613.1) as the reference. Core genome SNP analysis was performed as previously described (18).
Cloning of bla OXA-55-like genes. The bla OXA-55-like genes of S. algae isolates were amplified by PCR using Platinum Taq DNA polymerase High Fidelity (Invitrogen, Carlsbad, CA, USA) with a bla OXA-55-like forward primer incorporating an EcoRI digestion site (59-GATGCATCGAGAATTCATGAATAAAGGTTTGC-39) and a bla OXA-55-like reverse primer incorporating a BamHI digestion site (59-ATGGACACAGGATCCTCAAGGCAGCAGCTGTTC-39). The PCR product was purified by the Wizard SV Gel and PCR Clean-Up system (Promega, Madison, WI) and cloned into pCR4-TOPO using the TOPO TA cloning kit for sequencing (Invitrogen) and One Shot TOP10 chemically competent E. coli (Invitrogen). The accuracy of the nucleotide sequences of the inserts was confirmed by Sanger sequencing using the M13 primer (Invitrogen). Subsequently, pCR4-TOPO carrying bla OXA-55-like was digested with EcoRI and BamHI (TaKaRa Bio, Inc.) and ligated to the pHSG298 DNA plasmid (TaKaRa Bio Inc.) pretreated with EcoRI and BamHI. The resulting plasmids were chemically transformed into E. coli DH5a. Transformants carrying pHSG298 harboring bla OXA-55-like were selected on agar plates containing 50 mg/ml of kanamycin at 37°C for 24 h, and the presence of bla OXA-55-like was confirmed with PCR. This experiment was approved by the Toho University Safety Committee for Recombinant DNA Experiment (approval no. 21-52-458).
Antimicrobial susceptibility testing. Antimicrobial susceptibility testing was performed for the S. algae clade isolates and E. coli DH5a isolates carrying bla OXA-55-like cloned from S. algae isolates by the broth microdilution method using BBL Mueller-Hinton II broth, which was cation adjusted (Becton Dickinson and Co., USA) according to the Clinical and Laboratory Standards Institute (CLSI) guidelines (19). The following antimicrobial agents were used for antibiotic susceptibility testing: ampicillin, piperacillin, cefazolin, cefotaxime, ceftazidime (Sigma-Aldrich, St. Louis, MO, USA), aztreonam (Tokyo Chemical Industry Co., Ltd., Tokyo, Japan), imipenem (Banyu Pharmaceutical, Tokyo, Japan), clavulanic acid, and meropenem (Wako Pure Chemical Industry, Ltd., Tokyo, Japan). E. coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853 were used as quality control strains. The results were interpreted according to CLSI guidelines (20).
Quantitative RT-PCR for bla OXA-55-like . Quantitative reverse transcription-PCR (RT-PCR) was performed for S. algae clade isolates to compare bla OXA-55-like transcription levels. The isolates were grown in LB broth (Becton Dickinson and Co.) for 24 h at 37°C with shaking at 160 rpm and harvested at an optical density at 600 nm (OD 600 ) of 1.0. The RNA was extracted using the RNeasy minikit (Qiagen, Hilden, Germany), then used to generate cDNA with PrimeScript RT master mix (TaKaRa Bio Inc.). Quantitative PCR was performed using SYBR green PCR master mix (Applied Biosystems, Foster City, CA) with the primer pair bla OXA-55-like _Forward_Primer (59-GTTGGTTGGAGTTGGACGAC-39) and bla OXA-55-like _Reverse_Primer (59-TGCTTGAGCACCTGTTTCAC-39) on the Applied Biosystems 7500 Fast system (Applied Biosystems). Amplification of the rpoB gene was simultaneously performed with the rpoB_Forward_Primer (59-TTTGATCCCATTCCTTGAGC-39) and rpoB_Reverse_Primer (59-CC ACCAGAGGCTTCTCTGAC-39). The bla OXA-55-like transcription levels of the isolates were compared using the threshold cycle (DDC T ) method (21). The amplification efficiency of the quantitative PCR for rpoB and bla OXA-55-like was verified with 10-fold serially diluted TUM17384 total RNA ranging from 10 0 ng to 10 25 ng per assay, which demonstrated that the ratio of amplification efficiency of bla OXA-55-like to rpoB was 0.98.
b-Lactamase activity assay. b-Lactamase activity in the S. algae clade isolates was evaluated. First, the isolate of interest was inoculated into LB broth (Becton Dickinson and Co.) and incubated for 24 h with shaking at 160 rpm and 37°C. After incubation, 10 ml of the culture medium was centrifuged at 3,500 Â g for 15 min at 4°C. After the supernatant was discarded, the pellet was washed with 500 ml of phosphate-buffered saline (PBS) (pH 7.0) and recentrifuged at 13,000 Â g for 1 min at 4°C. After resuspending the pellet in 500 ml of PBS, the crude enzyme solution was prepared by sonication and subsequent centrifugation at 13,000 Â g for 30 min at 4°C. The protein concentration was measured by the Bradford method using bovine serum albumin (BSA) (Bio-Rad Laboratories, Inc., Hercules, CA) as a standard. The change in absorbance over time caused by the hydrolysis of b-lactam by b-lactamase was measured using a Shimadzu UV-2500 spectrophotometer (Shimadzu, Kyoto, Japan). The b-lactams used as the substrates for b-lactamase were adjusted to a final concentration of 100 mM in PBS. All reactions were performed in a Bandpass 10-mm cuvette with a total volume of 5 ml of enzyme added to 500 ml of substrate solution at 30°C.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. TABLE S1, XLSX file, 0.1 MB.