Myroides albus P34 was isolated from sewage discharged from an animal farm in Wenzhou, Zhejiang Province, China. The sewage samples were inoculated onto the blood plate, and incubated for 24 h at 35°C. Individual colonies of different color and size were randomly selected and preliminarily confirmed by a Vitek-60 microorganism autoanalysis system (bioMe’rieux Corporate, Craponne, France). Further identification was performed by sequencing the 16S rRNA gene and performing ANI analyses. E. coli DH5α and E. coli BL21 were used as the hosts for the resistance gene cloning and protein overexpression, respectively. The bla_WUS-1 gene was first cloned into the pUCP24 plasmid, and then the pCold I vector was used to induce cold shock expression of histidine-tagged WUS-1. All the strains and plasmids used in this work are listed in Table 1.

Antimicrobial susceptibility testing was conducted using the agar dilution method following the guidelines of the Clinical and Laboratory Standards Institute (CLSI). Antimicrobials used in this study included carbenicillin, ampicillin, ticarcillin, amoxicillin-clavulanate, ampicillin-sulbactam, piperacillin-tazobactam, cefazolin, cefoxitin, ceftazidime, cefepime, aztreonam, imipenem, meropenem, florfenicol, chloramphenicol, nalidixic acid, tetracycline, streptomycin, kanamycin, netilmicin, gentamicin, and amikacin. When neither the CLSI M100 (32nd Edition, 2022) nor European Enterobacteriaceae Antimicrobial Susceptibility Test Committee (European Committee on Antimicrobial Susceptibility Testing. [EUCAST], 2021) provided exact breakpoints for Myroides spp., the minimum inhibitory concentration (MIC) values were interpreted following the criteria for non-Enterobacteriaceae bacteria (such as A. faecalis, Elizabethkingia or Chryseobacterium; CLSI, 32nd Edition, 2022). E. coli ATCC 25922 was used as a quality control in each MIC test.

Genomic DNA extraction of Myroides albus P34 was performed using the AxyPrep Bacterial Genomic DNA Miniprep kit (Axygen Scientific, Union City, CA, United States) and was sequenced on both the Illumina HiSeq 2500 and PacBio RS II platforms by Shanghai Personal Biotechnology Co., Ltd. (Shanghai, China). The long PacBio reads were initially assembled by Canu (Koren et al., 2017), and then the sequence generated by the Illumina HiSeq 2500 platform was aligned to the primary assembly to correct for possible assembly errors using BWA v0.7.12 (Li and Durbin, 2009) and Pilon (Walker et al., 2014). Potential open reading frames (ORFs) were predicted using Prokka v1.14.6 (Seemann, 2014). Further functional annotation of predicted proteins was performed using DIAMOND (Buchfink et al., 2015) and searching against the NCBI nonredundant protein database with an e-value threshold of 1e-5. Resistance genes were annotated using the Resistance Gene Identifier (RGI) of the Comprehensive Antibiotic Resistance Database (CARD; McArthur et al., 2013). The ANI was calculated using FastANI v1.31 (Jain et al., 2018). Mobile genetic elements (MGEs) were detected using ISFinder (Siguier et al., 2006) and INTEGRALL (Moura et al., 2009). CGView Server (Petkau et al., 2010) was used to depict the circular map of the plasmid. Multiple amino acid sequence alignments and neighbor-joining phylogenetic tree constructions for WUS-1 and other class B1 MBLs sharing the closest amino acid sequence similarities to WUS-1 were performed using MAFFT (Katoh and Standley, 2013) and MEGAX (Kumar et al., 2018), respectively. Linear maps of genomic structure and structural comparisons of multiple genomes were portrayed using genoPlotR (Guy et al., 2010). The promoter region of bla_WUS-1 was predicted by BPROM.¹ The molecular weight and pI value of WUS-1 were predicted using the Expasy ProtParam Tool,² and the putative signal peptide cleavage site was predicted by SignalP 5.0 (Almagro Armenteros et al., 2019).

The strain Myroides albus P34 was screened for carbapenemase production by the modified Hodge test, in the meantime, the mCIM and eCIM tests following the CLSI M100 (32nd Edition, 2022) specifications were performed for carbapenemase phenotypic detection.

Amplification of the chromosomal bla_WUS-1 gene with its promoter region was performed using primers (Table 2) that contained BamHI and HindIII restriction endonuclease adapters at the 5′ ends. The PCR product was eluted from the agarose gel, digested with the restriction endonucleases BamHI and HindIII (Takara Bio, Inc., Dalian, China), and ligated into the pUCP24 vector using the T4 DNA ligase cloning kit (Takara Bio, Inc., Dalian, China). Recombinant plasmids were introduced into E. coli DH5α by the calcium chloride method. A single transformant was selected on a Luria-Bertani (LB) agar plate containing 40 mg/L gentamicin, and the cloned DNA fragment of the recombinant plasmid was confirmed by both PCR and PCR product Sanger sequencing (Shanghai Sunny Biotechnology Co., Ltd., Shanghai, China). Furthermore, antimicrobial susceptibility testing for the transformants was carried out as described above.

The PCR amplification product without a signal peptide but containing a thrombin cleavage site and a pair of flanking restriction endonuclease adapters (a BamHI site and a thrombin cleavage site were included in the orf-bla_WUS-1-F forward primer, and a HindIII site was included in the orf-bla_WUS-1-R reverse primer; shown in Table 2) was digested with BamHI and HindIII and then cloned into plasmid pCold I (digested with the same restriction enzymes). The recombinant plasmids were subsequently transformed into E. coli BL21 by the calcium chloride method, and the transformants were selected on LB agar plates containing 100 mg/L ampicillin. The transformant was further screened by PCR and sequencing and cultured in LB liquid medium with 100 mg/L ampicillin at 37°C overnight. Then, the overnight culture was diluted in LB liquid medium at a ratio of 1:100 and further cultured until it reached an OD₆₀₀ between 0.6 and 0.8. Protein expression was induced by incubating for 12 h at 16°C with isopropyl-β-d-thiogalactoside (IPTG; Sigma Chemicals Co., St. Louis, MO, United States) at a final concentration of 1 mM. Cells were harvested and resuspended and then disrupted sonically. The supernatant containing the β-lactamase was purified by a BeyoGold His-tag Purification Resin kit (Beyotime, Shanghai, China) according to the manufacturer’s instructions. Then, the His-tag was removed by incubating with thrombin (GenScript, Nanjing, China) for 12 h at 16°C. Finally, protein purification kits were applied to remove the free His-tag from the incubated protein. The purity and relative molecular mass of enzymes were checked using SDS–PAGE analysis, and the protein concentration was determined spectrophotometrically with a BCA Protein Assay Kit (Beyotime, Shanghai, China).

The enzymatic activities of purified WUS-1 were measured by a continuous spectrophotometric assay with a SpectraMax multifunctional microplate reader (M5, Molecular Devices, America). The rates of hydrolysis were calculated by monitoring the absorbance drop in a 200 μL reaction mixture containing 10 mM phosphate buffer (pH 7.4) and 50 μM ZnCl₂ at 30°C (Mammeri et al., 2002). Consequent data were analyzed by GraphPad Prism 9 software (GraphPad Software, San Jose, CA, United States) using nonlinear regression of initial reaction rates with the Michaelis–Menten equation to calculate the steady-state kinetic parameters (kcat and Km). The enzyme kinetic parameters were calculated following the formula.

Where v is velocity, Vmax is maximum velocity when all the enzymes are complexed to the substrate, [S] is substrate concentration and Km is Michaelis–Menten constant. The wavelengths are shown in Supplementary Table S1. To investigate the effects of inhibitors, the enzyme was preincubated with various concentrations of EDTA for 5 min at 30°C before the addition of substrate (100 μM imipenem; Mammeri et al., 2002). Fifty percent inhibitory concentrations (IC₅₀) were required to reduce the hydrolysis of 100 μM imipenem by 50%, and the results were determined by nonlinear regression in Prism software (Chen et al., 2019).

The colony of Myroides albus P34 is medium sized with fruity odor and yellow pigment. It is positive for oxidase, urease and phosphatase, but negtative for catabolizing glucose, lactose, maltose, mannitol, sucrose, aesculin, or indole. It was preliminary identified as Myroides spp. The 16S rRNA gene of isolate P34 had the closest relationship with that of Myroides albus BIT-d1 (MK734183.1) with 100% coverage and 100% identity. ANI analysis showed that it shared the highest identity (99.38%) with the Myroides albus strain BIT-d1 Scaffold1 (NZ_WMJX01000001). Therefore, this isolate was designated Myroides albus P34. The complete genome of Myroides albus P34 consisted of a 3,701,500 bp chromosome and one circular plasmid of approximately 84,474 bp in length. The chromosome and the plasmid encoded 3,218 and 95 coding sequences (CDSs) with an average GC content of 34.18 and 31.07%, respectively (Supplementary Table S2). Sequence homology analysis showed that the plasmid shared the highest similarity to the plasmid p63039 of Myroides odoratimimus PR63039 (CP013691.1) at 91% coverage and 99.71% identity. Similar to most strains of the genus Myroides, in vitro susceptibility tests showed that Myroides albus P34 exhibited resistance to all antimicrobials tested except for chloramphenicol, with especially high MIC levels for carbapenems (imipenem >512 mg/L and meropenem >512 mg/L, Table 3). The modified Hodge test and the mCIM and eCIM tests (zone diameter of 6 mm for the mCIM test and 20 mm for the eCIM test with a zone diameter difference of 14 mm) were all positive, suggesting metallo-β-lactamase production. To investigate whether any novel MBL gene was encoded in the Myroides albus P34 genome, we checked the whole genome annotation result and found that in addition to one plasmid-encoded, function-characterized MBL gene (bla_MYO-1), the P34 genome also harbored six chromosomally encoded hypothetical MBL genes. We analyzed the sequence structures of the predicted MBL genes and found that one (finally designated bla_WUS-1) contained conserved motifs of an Ambler class B1 β-lactamase. We then cloned the gene and determined its function.

The bla_WUS-1 gene is 741 bp in length and encodes a predicted protein of 246 amino acids with a calculated molecular mass of ca. 28.64 kDa. The predicted cleavage site of the signal peptide is found between positions 19 and 20 (alanine and glutamine residues). After removal of the signal peptide, the enzyme is a ca. 26.33 kDa protein with a predicted pI value of 7.09. Nine functionally characterized protein sequences with the highest similarity (ranging from 38.36–70.73%) to WUS-1 were retrieved from the NCBI nonredundant (NR) protein database, and they were all class B1 β-lactamases, of which the β-lactamase MUS-1 (AAN63647.1) ranked the first. The phylogenetic tree for these proteins was constructed and showed that WUS-1 is most closely related to MUS-1 from M. odoratimimus CIP 103073 (100% coverage and 70.73% identity), TUS-1 from M. odoratus CIP 103105 (100% coverage and 70.32% identity) and MUS-2 from M. odoratimimus 35a (100% coverage and 69.92% identity; Figure 1). Alignment result of the deduced amino acid sequence of WUS-1 with the other 4 closest relatives (identities ranging from 42.33–70.73%) is shown in Figure 2. It showed that they shared two strictly conserved motifs of the subclass B1 metallo-β-lactamases that interact with the Zn²⁺ cofactor or with the water molecule located in the active site. The Zn1 binding site, also known as the 3H site, contains three histidine residues, His94, His96 and His157, while the ligands for the Zn2 or DCH site include Asp98, Cys176 and His218 (Garau et al., 2004).

When searching for WUS-1 homologous proteins (>70% amino acid similarity) in the NCBI NR database, 23 functionally uncharacterized sequences (percent identities between 70.2 and 98.37%) were retrieved (Figure 3), all of which were from the Myroides species. The closest relative of WUS-1 was a hypothetical class B enzyme from Myroides sp. Loew2-1 (WP_160339331.1), and they shared an identity of 98.37% with a coverage of 100%.

The recombinant strain pUCP24-bla_WUS-1/DH5α showed increased MIC levels for ampicillin (16-fold), carbenicillin (256-fold), ticarcillin (32-fold), imipenem (8-fold) and meropenem (4-fold) in comparison with the control strains (DH5α carrying the vector pUCP24), whereas no MIC level change was observed for expanded spectrum cephalosporins or aztreonam (Table 3). Clavicularulanic acid and sulbactam did not restore the activity of β-lactams. WUS-1 displayed extremely low catalytic efficiencies (kcat/Km) for cefazolin and cefoxitin (Table 4). In accordance with other subclass B1 β-lactamases, the catalytic efficiencies of WUS-1 for cephalosporins were lower than those of penicillins or carbapenems (Mammeri et al., 2002; Vessillier et al., 2002; Naas et al., 2003). The highest catalytic efficiency was observed with ampicillin (kcat/Km rate of 10.06 μM⁻¹·s⁻¹). The catalytic efficiency for carbenicillin (1.93 μM⁻¹·s⁻¹) was lower than that of carbapenems. Carbapenems behaved as good substrates, and the high kcat/Km ratios observed with imipenem (3.99 μM⁻¹·s⁻¹) and meropenem (3.24 μM⁻¹·s⁻¹) always resulted from very high turnover rates. Cefoxitin (0.32 μM⁻¹·s⁻¹) and cefazolin (0.57 μM⁻¹·s⁻¹) were poorly hydrolyzed by WUS-1, and aztreonam was not hydrolyzed at all. The result of the β-lactamase activity inhibition analysis, as measured by the IC₅₀ (50% inhibitory concentration), demonstrated that the activity of WUS-1 was inhibited by EDTA (IC₅₀ of 71.28 μM for WUS-1).

The bla_WUS-1 gene was located on the chromosome of Myroides albus P34. To compare the genetic environment of bla_WUS-1 with those of homologous genes, a comparative genomic analysis was carried out on a nucleotide sequence of approximately 21 kb in length containing the upstream and downstream flanking regions of the bla_WUS-1 gene. Five chromosome segments with bla_WUS-1 equivalent genes that shared the highest sequence similarities in the center were found from the NCBI database (similarities ranging between 78.89 and 81.01%), and they were all from Myroides spp. strains (Figure 4). The genes upstream and downstream of the bla_WUS-1-related region were quite different from those sequences carrying counterpart genes. However the orfD-bla_WUS-1-orfP locus was relatively conserved and flanked by a pair of 9 bp imperfect inverted repeats (IRs), suggesting that sequence variation might result from rearrangement mediated by the inverted repeats.

Analyzing the complete genome sequence of Myroides albus P34, we found that the isolate contained a circular plasmid designated pMA84474, which shared the highest identity with the plasmid p63039 of Myroides odoratimimus PR63039 (CP013691.1, 91% coverage and 99.71% identity) retrieved from the NCBI nucleotide database. The annotation result showed that there were six drug-resistance genes and a series of mobile genetic elements densely distributed in the multidrug resistance (MDR) region (20.2–37.5 kb) of pMA84474 (Figure 5), and the six resistance genes included two β-lactam resistance genes (bla_OXA-347 and bla_MYO-1), one aminoglycoside drug resistance gene (aadS), two macrolide resistance genes (ereD and ermF), and one sulfonamide resistance gene (sul2). Comparative genomic analysis showed that the MDR region of pMA84474 shared the highest nucleotide sequence similarity with sequences in the plasmid p63039 (CP013691.1, 59.0% coverage and 93.69% identity) and the chromosomes of Myroides odoratimimus PR63039 (CP013690.1, 77.0% coverage and 94.83% identity) and Myroides odoratimimus G13 (CP037427.1, 66.0% coverage and 99.56% identity). The five resistance genes were clustered in a mobile genetic element-related unit flanked by a pair of IS91 family transposase ISWz1 sequences (ISWz1-aadS-bla_MYO-1-IS4-ereD-ermF-bla_OXA-347-intI-ISWz1). IS91 is the prototype element of a family of bacterial insertion sequences and is able to perform one-ended transposition with high efficiency, which raises the possibility of IS91-mediated gene spreading without the need to assemble a compound transposon (Garcillan-Barcia and de la Cruz, 2002).