Molecular and Kinetic Characterization of MOX-9, a Plasmid-Mediated Enzyme Representative of a Novel Sublineage of MOX-Type Class C β-Lactamases

ABSTRACT

The MOX lineage of β-lactamases includes a group of molecular class C enzymes (AmpCs) encoded by genes mobilized from the chromosomes of Aeromonas spp. to plasmids. MOX-9, previously identified as a plasmid-encoded enzyme from a Citrobacter freundii isolate, belongs to a novel sublineage of MOX enzymes, derived from the resident Aeromonas media AmpC. The bla_MOX-9 gene was found to be carried on a transposon, named Tn7469, likely responsible for its mobilization to plasmidic context. MOX-9 was overexpressed in Escherichia coli, purified, and subjected to biochemical characterization. Kinetic analysis showed a relatively narrow-spectrum profile with strong preference for cephalosporin substrates, with some differences compared with MOX-1 and MOX-2. MOX-9 was not inhibited by clavulanate and sulbactam, while both tazobactam and avibactam acted as inhibitors in the micromolar range.

INTRODUCTION

AmpC-type enzymes (AmpCs), belonging to molecular class C serine β-lactamases, provide a notable contribution to intrinsic and acquired β-lactam resistance in several Gram-negative bacterial pathogens. AmpCs were initially described as β-lactamases encoded by chromosomal genes resident in several species of Enterobacterales (e.g., Enterobacter cloacae complex, Citrobacter freundii, Serratia marcescens, Morganella morganii, Providencia stuartii, Hafnia alvei, and Escherichia coli) and Gram-negative nonfermenters (e.g., Acinetobacter spp., Pseudomonas aeruginosa, and Aeromonas spp.), but have subsequently been detected also as plasmid-encoded enzymes acquired by horizontal gene transfer in several species endowed or not with chromosomal AmpCs (1–4).

AmpCs typically have molecular masses ranging from 34 to 40 kDa and isoelectric points ranging from 6.6 to >9.0 (1). Similar to other classes of serine-β-lactamases, AmpCs have two structural domains, including an α and an α/β domain, with the active site situated in a groove between these two domains (5, 6). Several conserved elements in the active site have been identified, which appear to be directly or indirectly involved in the substrate recognition and catalytic process (5–7). Studies carried on AmpC of C. freundii suggested the importance of Ser64 and Tyr150 in the catalytic activity of class C β-lactamases (8). Ser64, as Ser70 in class A β-lactamases, allows the nucleophilic attack onto β-lactam carbonyl. Tyr150 acts, during the acylation and deacylation steps, to increase the nucleophilicity of the active serine residue and water molecules (8).

From the functional standpoint, AmpCs typically exhibit a preference for cephalosporin substrates, including narrow-spectrum cephalosporins (e.g., cephalothin and cefazolin), third generation cephalosporins (e.g., cefotaxime and ceftazidime), and cephamycins (e.g., cefoxiti and moxalactam), while being usually less active against fourth-generation cephalosporins (e.g., cefepime) (9). AmpCs are also active on penicillins, whereas carbapenems are overall stable, behaving as inactivators of these enzymes (9). The older β-lactamase inhibitors (e.g., clavulanic acid and tazobactam) have limited activity against AmpCs, while diazabicyclooctanes (e.g., avibactam and relebactam) and boronates (e.g., vaborbactam) are good inhibitors of these enzymes (10, 11).

Over the years, several class C enzymes have been described, belonging in different lineages, and a new standard numbering scheme for this class of β-lactamases has recently been proposed (12). The MOX lineage was one of the first lineages identified as plasmid-mediated enzymes associated with acquired resistance to extended-spectrum cephalosporins and cephamycins in Enterobacterales (1, 13). The prototype of this group of enzymes is MOX-1, which was first isolated from a Klebsiella pneumoniae of clinical origin (14). Subsequently, additional enzymes of the MOX lineage were described (of which some were named CMY), and the genes encoding these enzymes were shown to be derived from the mobilization of resident chromosomal genes of some Aeromonas species (15).

The present study reports on the characterization of MOX-9, a novel member of the MOX lineage of AmpCs.

RESULTS AND DISCUSSION

The bla_MOX-9 gene was previously identified in the C. freundii strain Cfr-FI-07, isolated from a hospital wastewater treatment plant in central Italy in 2013 (16). A BLAST search showed that MOX-9 was 97.9 to100% identical to the chromosomally encoded AmpCs of Aeromonas media strains deposited in public databases (n = 9, accessed 12 April 2022), confirming that this species is the natural host of this gene (15).

MOX-9 is 383 amino acids long, with a sequence identity of 72 to 81% compared to the other members of the MOX lineage of AmpCs. Interestingly, as previously noted (15), MOX-9 belongs to a sublineage outstanding the two major sublineages of MOX-type enzymes previously recognized (Fig. 1). Compared with MOX-1 and MOX-2, as representatives of the two major sublineages of the MOX lineage of AmpCs, MOX-9 differed by 74 and 78 amino acid residues, respectively.

FIG 1.

Phylogenetic tree showing the relationship of enzymes of the MOX lineage (of which some are named CMY). Sequences have been aligned with the MAFFT webserver (https://mafft.cbrc.jp/alignment/server/) and the tree has been built by using the neighbor-joining method and 1,000 bootstraps. The tree was rooted on the CMY-2 sequence used as an outgroup. Sequences were retrieved from the NCBI reference gene catalog (https://www.ncbi.nlm.nih.gov/pathogens/refgene).

The bla_MOX-9 determinant could be successfully transferred in E. coli DH5α by electrotransformation, revealing that the gene was plasmid borne. Analysis of the bla_MOX-9-containing contig derived from whole-genome sequencing (WGS) data (16) revealed a sequencing coverage much higher (approximate medium coverage ratio, 60×) than that of segments of chromosomal origin, in agreement with the plasmid location of the gene. Investigation of the genetic context of bla_MOX-9 in Cfr-FI-07 revealed that this determinant was carried on a 3.4-Kbp transposon, named Tn7469, composed by an ISKpn9-like insertion sequence (IS) (99.9% identical to ISKpn9 from K. pneumoniae plasmid pLRB01; accession number AJ276453) flanked by a 2,147-bp region at its 5′-end (Fig. 2). This transposon was inserted into an IS5-like element, likely following a transposition event, as suggested by the presence of 10-bp target site duplications. In turn, the IS5-like element was flanked by regions showing high similarity (>99%) to the Aeromonas caviae pKAM376_5 plasmid (AP024407.1). Interestingly, since ISKpn9-like was also found in the chromosome of some A. media strains (although apart from the resident bla_MOX-9-like genes), we could hypothesize that Tn7469 was originated in A. media and subsequently mobilized onto a pKAM376_5-like plasmid (Fig. 2). These findings overall provided novel insights into the mobilization mechanisms of bla_MOX-like genes.

FIG 2.

Genetic context of the bla_MOX-9 gene in C. freundii Cfr-FI-07 and its comparison to segments of the A. media R1-26 genome (region 3,313,893.3,318,039; accession number CP043579) and to the A. caviae pKAM376_5 plasmid. The Tn7469 transposon is indicated with a brace and its target site duplications are shown with yellow diamonds flanking the transposon. The ISKpn9-like element involved in the mobilization of bla_MOX-9 is boxed. Regions showing ≥99% nucleotide sequence identity are connected by blue areas. Hypothetical proteins are depicted in gray, while the IS5 element disrupted by Tn7469 transposition is in red.

The bla_MOX-9 gene was cloned into the pET-24 expression vector, and the recombinant E. coli strain BL21(DE3) (pET-24-MOX-9) was used to overproduce the MOX-9 enzyme. About 8 mg of pure MOX-9 protein were obtained from 1 L of an overnight culture, using two chromatographic steps as described in the methods section.

The experimental pI of MOX-9 was 7.1, in agreement with the theoretical value of 6.72, and different from that reported for MOX-1 (pI, 8.9) and MOX-2 (pI, 9.1) (13, 17).

Kinetic parameters of MOX-9 (k_cat, K_m or K_i, k_cat/K_m) were determined with a large panel of β-lactams and some β-lactamase inhibitors. As shown in Table 1, carbenicillin, ampicillin, and piperacillin were poor substrates, with k_cat values ranging from 0.07 s⁻¹ to 1.5 s⁻¹ and k_cat/K_m values around 10⁴ M⁻¹s⁻¹, while benzylpenicillin was a better substrate (k_cat, 18 s⁻¹; k_cat/K_m, 3.6 × 10⁵ M⁻¹s⁻¹), although MOX-9 appeared to be 10-fold less active than MOX-1 against this substrate. Cephalothin was the best substrate for MOX-9, with a k_cat value of 2,035 s⁻¹, and a k_cat/K_m value about 100-fold lower than that calculated for MOX-1 because of a higher K_m. Unlike MOX-1, the hydrolysis of cefepime by MOX-9 was measurable with a k_cat of 2 s⁻¹, a value overall similar to that reported for MOX-2 (1 s⁻¹) (13). However, it is important to highlight that the kinetic values for MOX-2 were determined by a microacidimetric assay (13). The affinity for cefepime, however, was very low, resulting in a low catalytic efficiency for this substrate (k_cat/K_m value of 5.6 × 10³ M⁻¹s⁻¹). MOX-9 showed much higher affinities for cefotaxime and ceftazidime, which combined with the respective turnover rates resulted in k_cat/K_m values around 10⁴ and 10³ M⁻¹s⁻¹, respectively. MOX-9 possessed a good activity for cefoxitin, with k_cat and k_cat/K_m values about 16- and 10-fold higher than MOX-1, respectively. Interestingly, MOX-9 hydrolyzed very poorly moxalactam with a k_cat of 0.004 s⁻¹, about 10- and 380-fold lower than that of MOX-2 and MOX-1, respectively. MOX-9 showed a slight hydrolysis against aztreonam which indeed acts as inhibitor for MOX-1 (17). In addition, MOX-9 showed k_cat and k_cat/K_m values for imipenem around 20- and 3-fold higher than MOX-1, respectively. MOX-9 was also tested against old inhibitors such as clavulanic acid, tazobactam, and sulbactam, as well as the new inhibitor avibactam. Clavulanic acid (similar to MOX-1) and sulbactam were unable to inhibit the hydrolytic activity of MOX-9. Avibactam and tazobactam showed both K_i values in the micromolar ranges, with the latter being able to inhibit MOX-9 better than MOX-1 (K_i = 82 μM versus K_i = 505 μM) (Table 2). It is important to emphasize that, compared to MOX-1, MOX-9 exhibits 51 amino acid substitutions distributed along the enzyme structure and some of them are positioned close to the active residues. This could explain the kinetic differences observed between MOX-9 and MOX-1. Furthermore, we remark that, unlike MOX-1, kinetic experiments for MOX-9 were performed in phosphate buffer. Antimicrobial susceptibility testing of a recombinant E. coli strain producing MOX-9 showed that enzyme production reduced susceptibility to various agents. The highest MIC increases were observed with cephalothin, cefoxitin, cefotaxime, and ceftriaxone, while susceptibility to ceftazidime, cefepime, penicillins, aztreonam, and carbapenems was only marginally or not affected (Table 3). The overall low MIC values observed with E. coli DH5α(pLBII-bla_MOX-9) was possibly due to a low level of enzyme production in the E. coli host. In fact, an in-silico prediction analysis, carried out using GenScript (https://www.genscript.com/tools/rare-codon-analysis), revealed that the codon usage of bla_MOX-9 (0.69 value), might be suboptimal for E. coli (ideal value 0.8–1.0). Antimicrobial susceptibility results with E. coli DH5α(pLBII-bla_MOX-9) were overall consistent with the relatively narrow-spectrum kinetic profile of the enzyme, with a strong preference for some cephalosporin substrates. The marginal or no changes observed with some penicillins, some expanded-spectrum cephalosporins (e.g., ceftazidime and cefepime), and aztreonam are consistent with the low turnover rate, possibly in combination with the complex in vivo interplay with outer membrane permeability and penicillin binding protein (PBP)-targeting. In the case of cefepime, a low enzyme affinity also contributed to antimicrobial effect.

TABLE 1.

Determination of kinetic constants of MOX-9 compared with MOX-1^a

Antibiotics	MOX-9d			MOX-1c
	Km (μM)b	kcat (s−1)	kcat/Km (μM−1 s−1)	Km (μM)	kcat (s−1)	kcat/Km (μM−1 s−1)
Benzylpenicillin	50 ± 2*	18 ± 1	0.360 ± 0.0055	40 ± 4	157 ± 0.04	3.925
Carbenicillin	5 ± 0.3*	0.07 ± 0.01	0.014 ± 0.001	NA	NA	NA
Ampicillin	29 ± 1*	0.77 ± 0.1	0.026 ± 0.0025	16 ± 0.6	2.36 ± 0.006	0.147
Piperacillin	90 ± 2*	1.5 ± 0.01	0.016 ± 0.0003	43.74 ± 0.0002	0.44 ± 0.02	0.010
Cephalothin	260 ± 6	2,035 ± 20	7.800 ± 0.105	34 ± 1.2	3415 ± 9.1	100.441
Cefoxitin	19 ± 1*	10 ± 0.5	0.526 ± 0.001	12.68 ± 0.14	0.63 ± 0.07	0.049
Cefepime	355 ± 8	2 ± 0.2	0.006 ± 0.0005	211 ± 0.04	ND	ND
Ceftazidime	14 ± 1*	0.02 ± 0.01	0.0014 ± 0.0006	311 ± 0.09	ND	ND
Cefotaxime	13 ± 0.5*	0.11 ± 0.01	0.0085 ± 0.00045	NA	NA	NA
Moxalactam	0.32 ± 0.05*	0.004 ± 0.001	0.012 ± 0.001	0.23 ± 0.006	1.52 ± 0.035	6.609
Aztreonam	0.47 ± 0.01*	0.08 ± 0.01	0.170 ± 0.017	2.85 ± 0.2	ND	ND
Meropenem	8 ± 0.5*	0.03 ± 0.005	0.004 ± 0.0005	NA	NA	NA
Imipenem	18 ± 0.7*	0.2 ± 0.01	0.011 ± 0.0003	3.2 ± 0.21	0.0098 ± 0.000014	0.003
Nitrocefin	90 ± 4	156 ± 5	1.733 ± 0.021	NA	NA	NA

^aNA, not available; ND, not determined.

^b(*), K_m was calculated as K_i using nitrocefin as the reporter substrate.

^cData from Alba et al. (17).

^dEach kinetic value is the mean of five different measurements; the error was below 5%.

TABLE 2.

Determination of K_i for MOX-9 compared with MOX-1^a

Inhibitors	MOX-9	MOX-1b
	Ki (μM)c	Ki (μM)
Clavulanic acid	>1,000	30,200 ± 3,000
Avibactam	52 ± 2	NA
Sulbactam	>1,000	NA
Tazobactam	82 ± 1	505 ± 0.14

^aNA, not available.

^bData from Alba et al. (17).

^cK_i values were determined using 100 μM nitrocefin as reporter substrate.

TABLE 3.

Antimicrobial susceptibility of C. freundii Cfr-FI-07 and E. coli DH5α (pLBII-bla_MOX-9) (producing the MOX-9 enzyme) to different β-lactams; susceptibilities of E. coli DH5α carrying the empty vector (pBC-SK) are also shown for comparison

		MIC (mg/L)
Antibiotic	E. coli DH5α (pBC-SK)	E. coli DH5α (pLBII-blaMOX-9)	C. freundii Cfr-FI-07
Ertapenem	≤0.015	≤0.015	16
Meropenem	0.015	0.015	16
Imipenem	0.125	0.125	>32
Ampicillin	2	4	>32
Amoxicillin	4	8	>32
Amoxicillin-clavulanic acid	4	8	>32
Piperacillin	1	1	>32
Piperacillin-tazobactam	1	1	>32
Aztreonam	0.125	0.125	8
Cephalotin	8	256	>256
Cefoxitin	4	16	>32
Ceftazidime	0.125	0.25	>32
Ceftazidime-avibactam	0.125	0.125	2
Ceftriaxone	0.03	0.5	>32
Cefotaxime	0.03	1	>32
Cefepime	0.03	0.03	4
Cefepime-taniborbactam	0.03	0.03	0.12
Ceftolozane-tazobactam	0.125	0.125	>32

Concluding remarks.

Herein, we have provided novel insights into the genetic context of bla_MOX-9, a member of the bla_MOX lineage of AmpC β-lactamase genes, derived from the resident chromosomal AmpC gene of A. media. Kinetic characterization of the MOX-9 enzyme, belonging to a separate sublineage of the MOX lineage of class C β-lactamases, revealed an overall narrow substrate profile, with some differences in comparison to MOX-1 and MOX-2 (the only other MOX-type enzymes for which kinetic data have been reported), including a remarkably lower activity against moxalactam. Previous studies demonstrated that moxalactam is a better substrate for MOX-1 than for other AmpCs because MOX-1 has more space in the substrate-binding site, facilitating the deacylation step by water molecule (5, 18). Moreover, compared to MOX-1, MOX-9 exhibited better hydrolysis of cefoxitin and, although at low levels, of carbapenems. These differences underscore that functional variability can be observed among MOX-type enzymes and could contribute to a better understanding of the structure-function relationships of these enzymes.

MATERIALS AND METHODS

Strains.

C. freundii Cfr-FI-07 was isolated from a hospital wastewater plant in central Italy in September 2013 (16). E. coli NovaBlue (Novagen Inc., Madison, WI), E. coli DH5α [supE44 ΔlacU169 (φ80lacZΔM15) hsdR17 recA1 endA1 gyrA96 thi-1 relA1; Bethesda Research Laboratories, Bethesda, MD], and E. coli BL21(DE3) (Novagen Inc.) competent cells were used for bacterial transformation experiments.

Gene transfer experiments.

Electroporation experiments were performed with electrocompetent E. coli DH5α using a plasmid DNA preparation from Cfr-FI-07, as described previously (16). Transformants were selected on Mueller-Hinton agar containing cefoxitin (16 mg/L).

Analysis of the bla_MOX-9 genetic environment in Cfr-FI-07.

The sequence upstream of bla_MOX-9 was determined by a primer walking strategy using primers designed on the bla_MOX-9 contig and another with similar coverage and characterized by an ISKpn9 fragment at one end (MOX-9_upstream_R 5′-AAGGTGTGCTGGATGTGAAG-3′ and ISKpn9_ downstream _F 5′-TGGCGATGTTATTCACGCTG-3′). Sanger sequencing of PCR amplicons obtained for gap closing was performed at an external sequencing facility (Eurofins Genomics, Germany). Nucleotide sequence comparisons of data from WGS from Cfr-FI-07 (16) were carried out using the BLAST web server at the NCBI web site (http://blast.ncbi.nlm.nih.gov/). Structural comparisons of genetic structures were obtained using EasyFig (19).

Cloning and expression of the bla_MOX-9 gene.

The bla_MOX-9 gene, with the signal peptide, was amplified by PCR from C. freundii Cfr-FI-07 using the following primers: Mox_for 5′-GGGGGCATATGATGCAACAACAGGTGCG-3′ and Mox_rev 5′-GGGGGGAATTCTCACTCGGCCAGCTTGCT-3′. The purified fragment was cloned in pET-24a (+) vector using the NdeI and BamHI restriction sites and transformed in E. coli NovaBlue cells. The authenticity of the bla_MOX-9 gene carried by the recombinant plasmid, named pET-Mox-9, was verified by Sanger sequencing of both strands using an automated sequencer (ABI Prism 3500, Life Technologies, Italy). To overproduce the enzyme, the recombinant plasmid pET-Mox-9 was transferred into E. coli BL21(DE3). For antimicrobial susceptibility testing experiments, the bla_MOX-9 gene was amplified by PCR from Cfr-FI-07 using the following primers: MOX-9-cf-fwd 5′-GGAATTCCATATGCAACAACAGGTGCGGATGA-3′ and MOX-9-cf-rev 5′-GAAGATCTCATCACTCGGCCAGCTTGCT-3′. The purified fragment was cloned in the pLBII multicopy expression vector (a derivative of pBC-SK) (20) using the NdeI and BglII restriction sites, to obtain recombinant plasmid pLBII-bla_MOX-9. E. coli DH5α transformants carrying the pLBII-bla_MOX-9 were selected on Luria-Bertani (LB) agar plates supplemented with chloramphenicol (85 mg/L). The authenticity of the cloned bla_MOX-9 gene carried by the pLBII-bla_MOX-9 recombinant plasmid was confirmed by Sanger sequencing as described above.

Purification of MOX-9 β-lactamase.

E. coli BL21(DE3) cells carrying pET24-bla_MOX-9 were grown in 1 L of LB medium with 50 mg/L kanamycin at 37°C in an orbital shaker (180 rpm). The isopropyl-β-thiogalactoside (IPTG), at a concentration of 0.4 mM, was added when the culture reached an optical density at 600 nm of 0.8. After addition of IPTG, the culture was incubated for 16 h at 22°C under aerobic conditions. Cells were harvested by centrifugation at 10,000× g for 10 min at 4°C and, after washing with normal saline, the pellet was resuspended in 30 mM Tris HCl + 27% sucrose (pH 8.0). Crude enzyme was obtained by sonication in ice (5 cycles at 60 W with 2 min intervals). The lysate was centrifuged at 100,000× g for 40 min at 4°C. The cleared supernatant was dialyzed overnight at 4°C against 20 mM sodium acetate buffer (pH 5.0) and loaded onto SP Sepharose Fast Flow equilibrated with the same buffer. The column was extensively washed to remove unbound proteins, and the β-lactamase was eluted with a linear gradient of NaCl (0 to 1 M) in the same buffer. Active fractions were pooled and dialyzed against 25 mM sodium phosphate buffer, pH 7.0, and loaded onto a Sephacryl S-100 column (XK16/70, bed volume 130 mL) equilibrated with 25 mM sodium phosphate buffer (pH 7.0), 0.15 M NaCl. The active fractions were recovered and stored at −40°C for further experiments. The isoelectric point (pI) of 20 μg of pure MOX-9 enzyme was determined by isoelectric focusing analysis using 5% polyacrylamide gels containing ampholynes (pH range, 3.5 to 9.5) in the Multiphor II apparatus (Pharmacia LKB Biotechnology). The gels were focused at 4°C and 25 W for 180 min.

Determination of kinetic parameters.

Kinetic parameters were determined following the hydrolysis of each substrate at 25°C in 25 mM sodium phosphate buffer (pH 7.0). Data were collected with a Perkin-Elmer Lambda 25 spectrophotometer (Perkin-Elmer Italia, Monza, Italy). Steady-state kinetic parameters were determined under initial-rate conditions using the Hanes linearization method (21). Each kinetic value is the mean of three different measurements; error was below 10%. Competitive inhibition assays were monitored directly using 100 μM nitrocefin as the reporter substrate. K_i values were calculated using the following equation: v₀ / v_i = 1 + (K_m × I)/(K_m + S) × K_i, where v_i and v₀ represent the initial rates of hydrolysis of nitrocefin with or without inhibitor, respectively; I is the concentration of inhibitor or poor substrate, K_i is the inhibition constant, K_m is the Michaelis-Menten constant, and S is the concentration of reporter substrate. The plot v₀ / v_i versus [I] yielded a straight line of slope K_m/(K_m + S) × K_i (22, 23).

Antimicrobial susceptibility testing.

MICs for C. freundii Cfr-FI-07 and E. coli DH5α carrying either pLBII-bla_MOX-9 or the empty vector (pBC-SK) were determined by reference broth microdilution method using a bacterial inoculum of 5 × 10⁵ CFU/mL, according to Clinical and Laboratory Standards Institute (CLSI) performance standards (24). The medium used for MIC determination was cation-adjusted Mueller-Hinton (MH) broth, and the bacterial inoculum was prepared in the same medium from overnight cultures in MH agar plates (containing chloramphenicol, 50 μg/mL, in case of E. coli carrying the pLBII-bla_MOX-9 or the pBC-SK plasmid vectors). It should be noted that, in the pLBII multicopy plasmid vector, the cloned gene is expressed under the control of the Plac promoter, and when the E. coli host is grown in media of low glucose concentration (such as MH broth) expression of the cloned gene is essentially constitutive and does not depend on addition of IPTG. In fact, by measuring the β-lactamase activity in early stationary-phase cultures of E. coli DH5α(pLBII-bla_MOX-9) in MH broth, either induced with IPTG or not, we found no significant differences of enzymatic activity using both cephalothin and nitrocefin as substrates. Tazobactam and avibactam were used at a fixed concentration of 4 mg/L. Clavulanic acid was used in combination with amoxicillin at the rate 1:2.

Data availability.

The nucleotide sequence data of the bla_MOX-9 gene appear in GenBank under accession number KJ746495.