Novel monoclonal antibodies for immunodetection of AmpC β-lactamases

Cloning and expression of recombinant CMY-34

A synthetic and codon-optimized gene (Invitrogen, Waltham, MA, USA) encoding CMY-34 (GenBank accession no. EF394370.1) with BamHI restriction site at 5′-end and XhoI site at 3′-end was cloned into the respectively hydrolyzed pET28a(+) vector (Novagen, Irene, South Africa), fusing the gene to His-tag coding sequence at the 5′-end of the gene. The construct was verified by Sanger sequencing (GENEWIZ), and the plasmid was subsequently transformed into E. coli Tuner (DE3) (Novagen, Irene, South Africa). For recombinant CMY-34 (rCMY-34) expression, an overnight culture of E. coli Tuner (DE3) was diluted (dilution factor 1:100) and cultivated in 1 L flasks with Luria-Bertani (LB) medium supplemented with 30 µg/mL kanamycin by shaking (220 rpm) at 37 °C until the optical density (OD) at 600 nm reached 0.6. The expression was induced with 0.2 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) for 4 h at 25 °C by shaking. Then, the culture was pelleted at 3000× g for 15 min at 4 °C, washed with buffer (0.2 M NaCl, 20 mM Tris-HCl, pH 7.4), centrifuged again and suspended in the purification buffer (0.15 M NaCl, 20 mM imidazole, 10 mM β-mercaptoethanol, 20 mM Tris-HCl, pH 7.4) supplemented with 1 mM phenylmethylsulfonyl fluoride (PMSF). The suspension was stored at −70 °C until rCMY-34 purification.

Purification of rCMY-34

The frozen culture suspension was thawed on ice and disrupted by sonication. The supernatant was cleared by centrifugation at 15,000× g for 20 min at 4 °C, filtered through 0.45 µM polyvinylidene difluoride (PVDF) filter and diluted with purification buffer. Purification was performed using an ÄKTA™ start chromatography system (29022094; Cytiva, Marlborough, MA, USA) and 1 mL HisTrap™ HP column (29051021; Cytiva, Marlborough, MA, USA) according to the manufacturer’s recommendations. The elution step was performed with the imidazole gradient (80–500 mM). The purity of rCMY-34 was analyzed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) under reducing conditions, selected fractions were pooled, and the buffer was exchanged into storage solution (2 mM dithiothreitol (DTT), 0.2 M NaCl, 20 mM Tris–HCl, pH 7.4).

Generation of monoclonal antibodies against CMY-34

For generation of CMY-34 specific MAbs, three 6–8 week old female BALB/c mice were immunized subcutaneously with 50 µg of rCMY-34 three times as described previously (Zvirbliene et al., 2010). The sample size was determined based on standard laboratory practices and the hybridoma technology protocol following the guidelines specified in DIRECTIVE 2010/63/EU. No randomization, control group, or blinding procedures were implemented, and no statistical methods were applied for MAb generation. During immunization, the blood samples were collected by tail bleeding, and the titer of antigen-specific IgG in the blood was evaluated by indirect ELISA. Three days before hybridization, the mouse that generated the highest titer of antigen-specific IgG was boosted subcutaneously with 50 µg of rCMY-34 in phosphate–buffered saline (PBS). Animal care and experimental protocols were adhered to the ARRIVE and FELASA guidelines in accordance with European and Lithuanian legislations. ARRIVE study plan is available at MIDAS repository (http://dx.doi.org/10.18279/MIDAS.258798). All immunized mice were sacrificed without anesthesia by cervical dislocation with the requirements specified in ANNEX IV of DIRECTIVE 2010/63/EU. The killing of mice was completed by confirmation of the onset of rigor mortis. No live mice were left at the end of the experiment. The hybridization was performed as described by Köhler & Milstein (1975). Briefly, the spleen cells of immunized mouse were isolated and fused with mouse myeloma Sp2/0 (ATCC, CRL-1581) using polyethylene glycol PEG-4000 (P7306; Sigma-Aldrich, Burlington, MA, USA), and then cultivated in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 15% fetal bovine serum (FBS) and selection reagent containing hypoxanthine, aminopterin and thymidine (HAT, Sigma-Aldrich, Burlington, MA, USA, H0262). Then, secretion of antigen specific MAbs in the hybridoma cell culture supernatant was tested by indirect ELISA. Selected hybridoma clones were cloned using limiting dilution method, additionally tested by indirect ELISA, cryopreserved and cultivated for MAb purification.

Mice used for hybridization were obtained from Vilnius University, Life Sciences Center, Institute of Biochemistry (Vilnius, Lithuania), which has State Food and Veterinary Agency (Vilnius, Lithuania) permission to breed and use experimental animals for scientific purposes (vet. approval no. LT 59–13–001, LT 60–13–001, LT 61–13–004). Ethical approval to use BALB/c mice for experiments was granted by State Food and Veterinary Agency (Vilnius, Lithuania), permission no. G2–117, issued 11 June 2019. Mice care and experimental procedures were carried out by trained personnel in accordance with Directive 2010/63/EU. Mice were monitored daily and kept in controlled conditions (12 h light-dark cycle, temperature of 22 ± 1 °C, humidity 55 ± 3%, cardboard enrichment) and had constant access to standard rodent food and water ad libitum. No analgesia or anesthesia was administered during any procedures following the hybridoma technology protocol and bioethics approval.

MAbs were purified from hybridoma cell culture supernatants by affinity chromatography as described previously (Sližienė et al., 2022).

Enzyme-linked immunosorbent assay (ELISA)

Indirect ELISA was used for testing of blood samples of immunized mice, selection of antigen-specific MAb producing hybridoma cells, cross-reactivity testing and determination of apparent dissociation constants (K_d). A total of 96-well plates (10547781; Thermo Scientific, Waltham, MA, USA) were coated with 50 µL/well of antigen at 3–5 µg/mL concentration in coating buffer (50 mM sodium carbonate, pH 9.5) for 16 h at 4 °C, and blocked with 250 µL/well of 2% bovine serum albumin (BSA) (w/v) in PBS for 1 h at room temperature (RT). After blocking, the wells were incubated with mouse blood samples (dilution ranging from 1:300 to 1:656,100 in phosphate–buffered saline containing 0.1% Tween-20 (v/v) (PBST)) or with undiluted hybridoma cell culture supernatant. For K_d determination, MAbs in the concentration range of 32–0.005 nM in PBST were tested. After incubation, plates were washed five times with PBST, followed by incubation with 50 µL/well of secondary horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG antibodies (1721011; Bio-Rad, Hercules, CA, USA) (dilution factor 1:5,000 in PBST) for 1 h at RT. Then, the plates were washed six times with PBST and incubated with 50 µL/well of 3,30,5,50-tetramethylbenzidine (TMB, Clinical Science Products, 01016-1-1000). The reaction was stopped with 25 µL/well of 1 N H₂SO₄ solution. The OD was measured with microplate spectrophotometer and calculated as difference between measured OD values at 450 nm and reference wavelength at 620 nm.

The K_d values of purified MAbs were determined from their titration curves. The determined value indicates MAb concentration (nM) at which the OD decreased by 50%.

Sandwich ELISA was used for development of CMY detection system. For the assay, 96-well plates (10547781; Thermo Scientific, Waltham, MA, USA) were coated with capture MAb (100 µL/well) at concentration of 2.5 µg/mL in coating buffer for 16 h at 4 °C. The wells were blocked with 250 µL/well of ROTI^® Block (A151.1; Carl Roth, Karlsruhe, Germany) for 1 h at RT. Then, the plates were incubated with 100 μL/well of serially diluted rCMY-34 as a standard in the range of 90–0.123 ng/mL or bacterial lysates with known total protein concentration in the range of 5,000–6.68 ng/mL in PBST for 1 h at RT. E. coli BL21 (Novagen, Irene, South Africa) lysate was tested as β-lactamase non-producing control. Subsequently, the wells were washed five times with PBST and incubated with 100 μL/well of HRP-labeled detection MAb for 1 h at RT. After incubation, the plates were washed six times with PBST, and 100 μL of TMB solution was added to each well to stop the reaction as described above.

Direct ELISA was used for testing of HRP-conjugated MAbs. For the assay, the antigen coating, blocking, washing and other procedures were performed as described above, in the protocol of indirect ELISA. The HRP-conjugated MAbs were serially diluted in the range of 1:100–1:72,900 with PBST (50 µL/well).

Competitive ELISA was used for selection of non-competing pairs of MAbs, suitable for sandwich-type immunoassays. For the method, antigen coating, blocking, washing and other procedures were carried out as described above, in the protocol of indirect ELISA. After blocking step, the plates were incubated with 50 µL/well of MAbs at concentration of 10 µg/mL for 1 h at RT. Then, HRP-conjugated MAbs (50 µL/well) were added, and the plates were incubated for 1 h at RT.

The isotypes of MAbs were determined using Mouse Immunoglobulin Isotyping ELISA Kit (550487; BD Biosciences, Franklin Lakes, NJ, USA) according to the manufacturer’s recommendations.

MAbs were conjugated with HRP as described previously (Stravinskiene et al., 2019).

The epitope of MAb 9D2 against CMY-34 was identified using a set of N-biotinylated synthetic peptides (GenScript) spanning a 22–49 amino acid (aa) region of CMY-34 (GenBank accession no. EF394370.1) and containing a linker sequence (-SGSG-) (Table 1). For testing, 96-well plates (10547781; Thermo Scientific, Waltham, MA, USA) were coated with 50 µL/well of Pierce™ avidin (21128; Thermo Scientific, Waltham, MA, USA) at concentration of 5 µg/mL in deionized water for 16 h at 4 °C, and blocked with 250 µL/well of 2% BSA (w/v) in PBS for 1 h at RT. After blocking step, the plates were washed and incubated with 100 µL/well of synthetic peptides at 10 µg/mL concentration in PBST for 1 h at RT. Then, the plates were washed and incubated with 100 µL/well of MAb 9D2 at 5 µg/mL concentration in PBST for 1 h at RT. The following steps of incubation with HRP-conjugated goat anti-mouse IgG antibodies and TMB were carried out as described above.

Western blot analysis (WB)

The reactivity of MAbs with β-lactamases was evaluated by western blot analysis (WB) as described previously (Kucinskaite-Kodze et al., 2020). Briefly, recombinant proteins (0.5–1 µg/line) or bacterial lysates (7–10 µg/line) were mixed with Pierce™ Lane Marker Reducing Sample Buffer (39000; Thermo Scientific, Waltham, MA, USA), boiled and fractionated by SDS-PAGE using 12% or 15% polyacrylamide gels. E. coli BL21 lysate was tested as β-lactamase non-producing control. After protein transfer to 0.45 µm PVDF membrane, the blocking of the membrane with 2% milk powder (w/v) in PBS was performed for 1 h at RT, followed by incubation with MAbs at 3 µg/mL concentration for 1 h at RT. Then, the membrane was washed with PBST and incubated with secondary goat anti-mouse IgG-HRP antibodies (1721011; Bio-Rad, Hercules, CA, USA) (dilution rate 1:4,000 with 2% milk powder in PBST) for 1 h at RT. The membrane was washed with PBST and NeA-Blue Precip reagent (Clinical Science Products, 01283-1-200) was used for detection.

Immunoprecipitation (IP)

The reactivity of MAbs with β-lactamases was tested by IP. For analysis, 125 µL of rProtein G Sepharose™ Fast Flow resin (17127905; Cytiva, Marlborough, MA, USA) was calibrated by washing the resin four times with 0.1 M Tris-HCl (pH 8) buffer and centrifuging at 12,000× g for 30 s. After calibration, the resin was blocked with ROTI^® Block (A151.1; Carl Roth, Karlsruhe, Germany) for 1 h at RT with rotation. Then, the resin was washed with 0.1 M Tris-HCl (pH 8) buffer, mixed with 25 µg of MAb and incubated for 1 h at RT with rotation. Formed complexes were washed four times with PBST, divided into five separate tubes and incubated with bacterial lysates with known total protein concentration (40 µg/tube) or rCMY-34 (5 µg/tube) for 1.5 h at RT with rotation. After incubation, the complexes were washed four times with PBST, and the elution step was performed with 0.1 M glycine (pH 3) buffer. The eluted fractions were mixed with Pierce™ Lane Marker Reducing Sample Buffer (39000; Thermo Scientific, Waltham, MA, USA), boiled and analyzed by SDS-PAGE and WB. For WB, an HRP-conjugated MAb 9D2 against CMY-34 (dilution factor 1:200 with 2% milk powder (w/v) in PBST) was used for detection.

Epitope mapping

Epitope mapping of newly generated MAbs was performed using truncated overlapping CMY-34 fragments. In the first step, the potential linear B-cell epitopes of CMY-34 protein were predicted using the BepiPred-2.02 tool (https://services.healthtech.dtu.dk/services/BepiPred-2.0/) (Jespersen et al., 2017), which divided the antigen into six overlapping fragments (Table 2).

The DNA sequences of CMY-34 fragments were amplified by PCR using primers with additional BamHI and XhoI restriction sites (Table 2) and a synthetic CMY-34 gene (GenBank accession no. EF394370.1) as a template. The amplified products were hydrolyzed with restriction nucleases, cloned into the respectively hydrolyzed pET28a(+) vector and verified by Sanger sequencing (GENEWIZ). For expression of His-tag-fused fragments, an overnight culture of E. coli Tuner (DE3) was diluted (dilution factor 1:100) and cultivated in 5 mL of LB supplemented with 30 µg/mL kanamycin by shaking (220 rpm) at 37 °C until the OD at 600 nm reached 0.8–1. Protein expression was induced with 1 mM IPTG for 2.5 h at 37 °C by shaking. The cells were sedimented by centrifugation at 3,000× g for 5 min, mixed with 1% sodium dodecyl sulfate (SDS) (w/v) in PBS and boiled. Then, Pierce™ Lane Marker Reducing Sample Buffer (39000; Thermo Scientific, Waltham, MA, USA) was added, and samples were boiled repeatedly. Prepared cell lysates were evaluated by SDS-PAGE under reducing conditions and WB using in-house-generated mouse MAb 6C2 against His-tag at 3 μg/mL concentration or CMY-34 specific MAbs at 2 μg/mL concentration.

Sequencing of MAb variable regions

Determination of MAb variable heavy chain (VH) and light chain (VL) sequences was performed as described previously (Stravinskiene et al., 2020). Briefly, total RNA was extracted from hybridoma cells using GeneJET RNA Purification Kit (K0732; Thermo Scientific, Waltham, MA, USA), following the manufacturer’s protocol. Reverse transcription was carried out using the extracted RNA as a template and the RevertAid First Strand cDNA Synthesis Kit (K1621; Thermo Scientific, Waltham, MA, USA), according to the manufacturer’s recommendations. The variable regions were amplified by PCR using the Phusion Flash High-Fidelity PCR Master Mix (F548S; Thermo Scientific, Waltham, MA, USA) and previously described primers (Table 3) (Pleckaityte et al., 2011). Amplified DNA fragments were cloned using CloneJET PCR Cloning Kit (K1232; Thermo Scientific, Waltham, MA, USA) and verified by Sanger sequencing (GENEWIZ). The results were analyzed with the IgBlast tool of the National Center for Biotechnology Information (NCBI, https://www.ncbi.nlm.nih.gov/igblast/) (Ye et al., 2013), and VH and VL coding nucleotide sequences were identified. The sequences of complementarity-determining regions 1–3 (CDR1–3) were determined using a tool for standardized analysis of the antibody sequences IMGT/V-QUEST version 3.6.1 (https://www.imgt.org/IMGT_vquest/input) (Brochet, Lefranc & Giudicelli, 2008; Giudicelli, Brochet & Lefranc, 2011).

Lateral flow immunoassay (LFIA)

Lateral flow immunoassay (LFIA) test strips were assembled of a nitrocellulose (NT) membrane Vivid 120 (VIV120SAMP; Cytiva, Marlborough, MA, USA), a 6,613 polyester fiber conjugate pad (Ahlstrom-Munksjo, Espoo, Finland), a 1,281 cotton sample pad (Ahlstrom-Munksjo, Espoo, Finland), a grade 243 wick pad (Ahlstrom-Munksjo, Espoo, Finland) and a self-adhesive base (nanoComposix, San Diego, CA, USA). The colloidal gold-conjugated MAb was prepared by covalent conjugation with BioReady™ 40 nm Carboxyl Gold spheres (AUXR40-5M; nanoComposix, San Diego, CA, USA), according to the manufacturer’s recommendations. The conjugate pad was prepared by its immersion in the treatment buffer (10% sucrose (w/v), 10 mM potassium phosphate, pH 7.4) for 10 min, followed by incubation with the gold-MAb conjugate (25 µL/cm²) for 1 h at RT, then drying for 5 h at 37 °C, followed by drying with desiccator overnight at RT. The test zone was prepared by immobilizing MAbs at 1 mg/mL concentration that were mixed with green dye: ProClin™ 300 (48912-U; Sigma-Aldrich, Burlington, MA, USA), dilution factor 1:40 with PBS, 3.3 mM Orange G (O7252; Sigma-Aldrich, Burlington, MA, USA), 0.56 mM xylene cyanole (B3267; Sigma-Aldrich, Burlington, MA, USA), dilution factor 1:10 with PBS. The control zone was formed by 0.2 mg/mL goat anti-mouse IgG Fc antibodies (103301; SouthernBiotech, Birmingham, AL, USA) mixed with blue dye (3.7 mM xylene cyanole, dilution factor 1:20 with PBS). The contour zone consisted of BSA conjugated with Remazol Brilliant Blue R dye (R8001; Sigma-Aldrich, Burlington, MA, USA). The antibodies were spotted onto the membrane (drop volume 21.6 nL) by the program-controlled dispensing system sciFLEXARRAYER S3 (Scienion, Volmerstraße, Germany), and then dried for 1 h at 37 °C. The sample pad was treated with blocking solution (2% BSA (w/v) in PBS) for 30 min at RT, rinsed two times with distilled water, incubated with treatment buffer for 30 min at RT, and dried for 1 h at 37 °C.

Before testing, bacterial isolates were grown on LB agar supplemented with 16 µg/mL ceftazidime for 20–24 h at 30 °C. For testing, a single colony was resuspended in 150 µL of extraction buffer (1% (w/v) 3-((3-cholamidopropyl)-dimethylammonium)-1-propanesulfonate (CHAPS, Carl Roth, Karlsruhe, Germany, 1479.1), 0.5% (w/v) BSA, 0.5% (v/v) Tween-20 in 1:1 diluted PBS) and vortexed. A total of 100 µL of sample was dispensed onto the sample pad, and the results were evaluated visually after 20 min.

Two-photon excitation (TPX) assay

Performance of MAbs against CMY-34 was evaluated in separation-free double sandwich immunoassay utilizing two-photon excitation fluorescence detection (Hänninen et al., 2000; Koskinen et al., 2021), as applied in commercial mariPOC^® test (ArcDia International Ltd., Turku, Finland). MAbs 9D2 and 2E11 were coated on microparticles, labelled with fluorescent molecules, and mixed together to obtain assay reagent cocktail as described previously (Koskinen et al., 2007).

For determination of analytical sensitivity of the assay, rCMY-34 was serially diluted with assay buffer (R-B01; ArcDia International Ltd, Turku, Finland) in the range of 0.5–1,500 ng/mL. For testing, bacterial isolates were grown on blood agar for 24 h at 30 °C. A single colony was suspended in 100 µL of extraction buffer (1% CHAPS, 0.5% (w/v) BSA, 0.5% (v/v) Tween-20 in 1:1 diluted PBS) and diluted with assay buffer. Diluted rCMY-34 samples and bacterial suspensions were mixed with the assay reagent cocktail and analyzed.

Testing of bacterial isolates by PCR

Bacterial isolates used for MAb characterization were additionally tested by PCR using previously described primers specific to bla_CMY (CMY-F1 5′-MTGGGGKAAAGCCGATATC-3′ and CMY-R1 5′-AGTTCAGCATCTCCCANCC-3′) (Mlynarcik et al., 2021) and bla_NDM (NDM-F 5′-GGGGATTGCGACTTATGC-3′ and NDM-R 5′-AGATTGCCGAGCGACTTG-3′) (Mlynarcik et al., 2019). During the PCR, a 701 bp fragment of bla_CMY and 258 bp fragment of bla_NDM were amplified. Obtained PCR products of isolates with previously unidentified β-lactamases were sequenced and analyzed by the Beta-Lactamase DataBase (BLDB) BLAST tool (http://www.bldb.eu:4567/) (Naas et al., 2017).

Bacterial isolates

In total, nine bacterial isolates with known β-lactamase profiles and two β-lactamase-negative isolates were tested. The isolates used for MAb characterization and LFIA testing were: CMY-34-positive Citrobacter portucalensis (3826Z08), which was a gift from Anette M. Hammerum and colleagues (Hammerum et al., 2011), NDM-1-positive Klebsiella pneumoniae (CCUG60138 and CCUG68728), Acinetobacter chinensis (CCUG74036T and CCUG74037), which were obtained from the Culture Collection University of Gothenburg (Sweeden). The isolates tested by LFIA and TPX assays were: CMY-6, NDM-1, OXA-1, CTX-M-15, SHV-11-positive K. pneumoniae (50627996), CMY-4, OXA-1, VIM-29, CTX-M-15-positive E. coli (50639799), CMY-16, NDM-1, OXA-10-positive Proteus mirabilis (50664164), CMY-2, OXA-1, OXA-181, CTX-M-15, TEM-1-positive E. coli (50816743), β-lactamase-negative Klebsiella aerogenes (50796520) and P. mirabilis (50793946), which were provided by Clinical microbiology department of Turku University Hospital (Finland).

Preparation of bacterial lysates

For preparation of lysates, K. pneumoniae (CCUG60138, CCUG68728) and A. chinensis (CCUG74036T, CCUG74037) isolates were grown in LB media supplemented with 16 µg/mL meropenem and C. portucalensis (3826Z08) isolate was grown in LB supplemented with 16 µg/mL ceftazidime for 16–20 h at 30 °C. The cells were disrupted by sonication, and the suspension was cleared by centrifugation at 15,000× g for 20 min at 4 °C. Total protein concentration was determined using Pierce™ Bradford Protein Assay Kit (23200; Thermo Scientific, Waltham, MA, USA), according to the manufacturer’s protocol. The supernatants were aliquoted and stored at −70 °C.

Cefinase test

The enzymatic activity of β-lactamases was tested with chromogenic cephalosporin (nitrocefin) impregnated BBL™ Cefinase™ paper discs (BD Biosciences, Franklin Lakes, NJ, USA). The discs were dispensed into sterile petri dishes. Then, 2.5 μg of rCMY-34 or 5 μg of bacterial lysate with known total protein concentration was diluted with PBS to the total volume of 20 μL. The samples were dispensed onto the discs, incubated for 1 h at RT, and evaluated visually.

Structure prediction of MAb-antigen complex

Sequences of MAb Fab fragment were constructed by concatenating determined MAb 2E11 and 9D2 VH and VL sequences with the corresponding UniProtKB (Bateman et al., 2023) sequences of mouse IgG2a Ig-like domain 1 for the 2E11 heavy chain (UniProt accession no. P01863), mouse IgG1 sequences for 9D2 heavy chain (UniProt accession no. P01868) and mouse immunoglobulin kappa for the light chain (UniProt accession no. P01837).

Structure modeling of the interaction between CMY-34 β-lactamase (GenBank accession no. ABN51006.1) and Fab-variable fragments of the MAbs was performed using AlphaFold2-Multimer (Jumper et al., 2021; Evans et al., 2022). In the first step, models were generated using standard AlphaFold and ColabFold pipelines as described previously (Olechnovič et al., 2023). No reliable models were produced, therefore the AFsample protocol was used to generate 6,000 models (Wallner, 2023). AFsample models having self-assessment score (ranking_confidence) higher than 0.8 were relaxed using OpenMM (Eastman et al., 2017). In addition to the AlphaFold quality self-estimation, the antibody-antigen interfaces in the predicted structures were also evaluated by the VoroIF-jury procedure using the FTDMP software (Olechnovič et al., 2023, 2025), as well as VoroMQA (Olechnovič & Venclovas, 2017) and VoroIF-GNN (Olechnovič & Venclovas, 2023) scores. For further analysis of MAb 2E11 epitope, the model having minimum predicted-aligned error (PAE) value for the antibody-antigen interface was selected as the most reliable model.

Additionally, we decided to use AlphaFold3 (Abramson et al., 2024) for the antibody-antigen interaction modeling. The models were generated using the default parameters of AlphaFold Server with or without the templates, and the most reliable model was selected according to the reported ranking score values.

The protein interaction interfaces were compared using contact area difference (CAD)-score (Olechnovič & Venclovas, 2020). The contacts between antibody and antigen molecules were analyzed using the VoroContacts server (Olechnovič & Venclovas, 2021).

Statistical analysis and visualization of the data

The data was analyzed, and the graphs were generated using GraphPad Prism (Dotmatics, Boston, MA, USA) and OriginPro 8 (OriginLab) softwares. The ELISA results of MAb reactivity testing with β-lactamases and CMY-34 fragments are presented as mean of OD values ± standard deviation (SD) of three replicates (n = 3). The cut-off values were calculated as sum of three SD (3SD) of negative control. The ELISA results of MAb testing with synthetic peptides are demonstrated as mean ± SD, n = 3. The indirect ELISA results of MAb cross-reactivity testing are presented as mean ± SD, n = 3, cut-off value of negative control. Results of developed sandwich ELISA are presented as mean ± SD, n = 3, cut-off value of negative control (E. coli BL21 lysate). The results of dose-response curve and testing of bacterial isolates by TPX assay are shown as mean, n = 3, cut-off value of negative control. K_d values were calculated with OriginPro 8 (OriginLab) software, and the results are demonstrated as mean ± standard error of mean (SEM), n = 3. The limit of detection of the assays was defined as 3SD from intercept as described previously (Armbruster & Pry, 2008).

Generation of MAbs against CMY-34 and determination of their specificity and affinity

For generation of MAbs, recombinant His-tagged β-lactamase CMY-34 (rCMY-34) was expressed in E. coli and purified by nickel-affinity chromatography (Fig. S1). The enzymatic activity of the purified protein was confirmed by cefinase test (Fig. S2). For this purpose, rCMY-34 was applied on the paper disc that is impregnated with chromogenic cephalosporin—nitrocefin. During the testing, nitrocefin turned from yellow to red, indicating the hydrolysis of this β-lactam and the proper folding of the recombinant enzyme.

In the next step, rCMY-34 was used for immunization of mice and generation of CMY-specific MAbs. Following hybridization, more than 130 hybridoma clones were selected in total for further testing of MAb reactivity with the antigen, of which fewer than a half were cloned and had their isotype determined. Following the screening, the collection of 14 IgG class MAbs was created (Table S1). To evaluate the binding affinity of generated MAbs with rCMY-34, their apparent dissociation constants (K_d) were determined by indirect ELISA (Table S1). Experimentally obtained K_d values of the MAbs were in the range of 0.05–2.5 nM, indicating high-affinity interaction with rCMY-34 (Table S1). In the following step, noncompeting MAb pairs suitable for sandwich-type assays were selected. For this purpose, six high affinity MAbs were conjugated with HRP and tested by competitive ELISA (Fig. S3). Selected candidate pairs of MAbs were additionally tested by sandwich ELISA to evaluate their ability to detect rCMY-34 with the highest sensitivity (Fig. S4). Based on these results, two non-competing, high-affinity MAbs produced by hybridoma clones 9D2 (IgG1, K_d 0.24 nM) and 2E11 (IgG2a, K_d 0.10 nM) were selected for further investigation by a wide range of immunoassays (Table S1, Fig. S5). Their reactivity with rCMY-34 was additionally confirmed by indirect ELISA, western blot (WB) and immunoprecipitation (IP) methods (Fig. 1, Table 4).

To further characterize the reactivity of selected MAbs and their ability to recognize β-lactamases, several antibiotic-resistant β-lactamase producing bacterial isolates were analyzed. In the first step, a previously described (Hammerum et al., 2011) C. portucalensis isolate (3826Z08) producing CMY-34 β-lactamase was tested. The presence of CMY-34 coding gene was additionally confirmed by PCR, with a set of previously described primers (Mlynarcik et al., 2021) targeting a 701 bp region of bla_CMY. Moreover, the PCR analysis of two NDM-1 β-lactamase producing K. pneumoniae isolates (CCUG60138 and CCUG68728) with bla_CMY and bla_NDM specific primers (Mlynarcik et al., 2019, 2021) revealed that these isolates also encode a CMY β-lactamase (Fig. S6). The sequencing of amplified PCR products confirmed that both K. pneumoniae isolates contained bla_CMY-4. Therefore, these isolates were used as CMY-4 producers for further characterization of MAbs and the development of CMY-specific immunoassays. As a negative control for these assays, E. coli BL21 strain and two NDM-1 producing A. chinensis isolates (CCUG74036T and CCUG74037) were utilized.

The ability of MAbs to recognize CMY β-lactamases was tested using lysates of above described CMY-34, CMY-4 and NDM-1 producing bacterial isolates with known total protein concentration. After preparation of lysates, the retained enzymatic activity of β-lactamases was confirmed by cefinase test (Fig. S2). The analysis revealed nitrocefin hydrolysis, indicating that the lysate preparation conditions did not disrupt structure of β-lactamases, which is important for their catalytic activity. The testing of C. portucalensis (3826Z08) lysate by WB, IP and indirect ELISA confirmed the reactivity of MAbs 9D2 and 2E11 with CMY-34 (Fig. 1, Table 4). Moreover, cross-reactivity of MAbs 9D2 and 2E11 with CMY-4 was revealed by WB and IP when lysates of bla_CMY-4 harboring K. pneumoniae (CCUG60138 and CCUG68728) were analyzed (Fig. 1, Table 4). Protein bands of approximately 38–39 kDa corresponding to CMY-34 and CMY-4 in the respective bacterial lysates were identified by WB, along with a 40 kDa band of rCMY-34, which was used as a positive control (Figs. 1A, 1B). Furthermore, both MAbs immunoprecipitated CMY-34 and CMY-4 from C. portucalensis (3826Z08) and K. pneumoniae (CCUG60138 and CCUG68728) lysates, respectively (Fig. 1C, Table 4). The distinct protein bands visible in WB below the full-length natural CMY-34 in the C. portucalensis lysate and recombinant CMY-34 may correspond to partially degraded forms of CMY-34 that still retain MAb recognition site (Fig. 1C). Thus, due to their high affinity, MAbs 9D2 and 2E11 were able to capture and immunoprecipitate all forms of these β-lactamases—both full-length and partially degraded. Protein band observed in WB slightly below the 25 kDa corresponds to the light chain of the MAb used for IP (Fig. 1C). Considering this, a control IP reaction with only IP MAb (no cell lysate or target proteins) was analyzed to identify bands derived solely from the antibody (Fig. 1C, lane PBST). This control demonstrates that a protein band corresponding to the light chain was derived from the antibody used for IP.

Determination of MAb variable sequences

Variable regions of heavy (VH) and light (VL) chains of the MAbs 9D2 and 2E11 were determined. Total RNA isolated from the hybridoma cells (Fig. S7A) was used to synthesize single-stranded cDNA. The VH and VL coding sequences were amplified by PCR using primers specific to mouse immunoglobulin gene heavy and light chains localized in the framework region 1 (Orlandi et al., 1989; Wang et al., 2000; Pleckaityte et al., 2011) (Figs. S7B, S7C). The amplified PCR products were cloned into the cloning vector, sequenced and analyzed with the IgBlast (https://www.ncbi.nlm.nih.gov/igblast/) (Ye et al., 2013) and IMGT/V-QUEST (https://www.imgt.org/IMGT_vquest/input) (Brochet, Lefranc & Giudicelli, 2008; Giudicelli, Brochet & Lefranc, 2011) tools. After analysis, the VH and VL coding sequences and the sites of complementarity-determining regions 1–3 (CDR1–3) were determined. The obtained VH and VL sequences confirmed the presence of clonal differences between MAbs 9D2 and 2E11. The sequences are doposited in the ABCD (AntiBodies Chemically Defined) Database (Lima et al., 2020) with assigned accession numbers ABCD_BD767 and ABCD_BD768 for MAb 9D2 and 2E11, respectively.

Mapping of MAb epitopes

It was demonstrated, that selected pair of MAbs 9D2 and 2E11 do not compete for rCMY-34 binding (Fig. S3), which is indicative of their distinct epitopes. In the next step, analysis of MAb recognition sites within the CMY-34 β-lactamase was performed. For this purpose, six overlapping His-tagged CMY-34 fragments (CMY-34_1–CMY-34_6) were expressed in E. coli (Fig. 2A). The efficiency of protein expression was confirmed by SDS-PAGE and WB when E. coli lysates after protein synthesis induction were analyzed (Fig. S8). For epitope mapping, the same E. coli lysates were tested with MAbs 9D2 and 2E11 by WB and sandwich ELISA on microtiter plates coated with in house-generated His-tag-specific MAb 6C2 (Figs. 2B, 2C). Both ELISA and WB revealed the reactivity of MAb 9D2 with the overlapping fragments CMY-34_1 and CMY-34_2, indicating MAb 9D2 epitope localization at 22–49 amino acids (aa) of CMY-34 (according to CMY-34 sequence: GenBank accession no. ABN51006.1), as this region is present in both CMY-34_1 and CMY-34_2 fragments but absent in the non-reactive CMY-34_3 fragment (Table 2). However, it was determined that MAb 2E11 did not react with any of CMY-34 fragments, according to ELISA and WB results (Figs. 2B, 2C), suggesting that epitope is not strictly linear and is disrupted by dividing the CMY-34 protein into the fragments.

For fine epitope mapping of MAb 9D2, the overlapping synthetic peptides (P0–P16) spanning a determined MAb 9D2 epitope (22–49 aa region of CMY-34) were analyzed (Fig. 3A). MAb 9D2 reactivity with synthetic peptides was evaluated by indirect ELISA, which revealed the reactivity of the MAb with P0 (22–49 aa region of CMY-34 coresponding peptide) and P9–P13 peptides (Fig. 3B). Based on these results, MAb 9D2 epitope consists of 13 aa-long segment, which is localized at 22–34 aa region of CMY-34 (according to CMY-34 sequence: GenBank accession no. ABN51006.1) (Fig. 3C).

Computational prediction of MAb 2E11 epitope

Since the localization of MAb 2E11 epitope within the CMY-34 protein has not been experimentally determined using truncated CMY-34 fragments, computational modeling of CMY-34 and MAb 2E11 interaction was performed to determine the putative epitope. The modeling was performed using the sequences of CMY-34 and MAb 2E11 Fab fragment. In total, 6000 AlphaFold2 (Jumper et al., 2021; Evans et al., 2022; Wallner, 2023) models were generated. Only 20 models had high self-assessment score (ranking_confidence >0.8, Fig. S9), and 19 of them were highly similar to each other, having an average antibody-antigen interface CAD-score value of 0.82 (Fig. S10A). Interestingly, this binding mode was also observed as a top ranked AlphaFold model when only the VH and VL fragments of the MAb were used for modeling of interaction (Fig. S10B) and in the top ranked AlphaFold3 (Abramson et al., 2024) model. These 19 models were also the most reliable models when evaluating them with VoroIF-jury (Olechnovič et al., 2023, 2025), which is a protocol independent of AlphaFold quality self-estimation. The evaluation suggested that the prediction was accurate enough for epitope estimation, therefore, a model having minimal average PAE value at the MAb-CMY-34 interface was selected for further analysis (Fig. 4A). According to the most reliable AlphaFold model, the putative MAb 2E11 epitope is conformational and consists of three short segments located at 38–44, 299–311 and 362–379 aa of CMY-34 (according to CMY-34 sequence: GenBank accession no. ABN51006.1) (Fig. S11). Numerous hydrogen bonds are present between polar residues of antibody and antigen in every fragment of the interaction interface, and there are several possible salt bridges identified (Fig. S10C).

The spatial arrangement of the putative epitope of MAb 2E11 is in close proximity to the MAb 9D2 recognition site (Fig. 4B). AlphaFold2-based AFsample protocol did not generate any reliable models of the MAb 9D2 and CMY-34 interaction (Fig. S9), but in the AlphaFold3 model of CMY-34 and MAb 9D2 complex, the antibody is bound to the antigen in different possition than MAb 2E11 (Fig. S12). The overlap between the epitopes is low, thus the results of structure modeling are in line with the experimentally observed independence of the epitopes.

Analysis of MAb cross-reactivity

The identified binding site of MAb 9D2 (22–34 aa region of CMY-34) was further analyzed by Beta-Lactamase DataBase (BLDB) BLAST tool (http://www.bldb.eu:4567/) (Naas et al., 2017) to determine its similarity across the members of CMY family. The analysis revealed 100% sequence identity among 173 of 181 currently identified enzymes of the CMY family, including in this study analyzed CMY-4 variant (Fig. 5A). The remaining seven members of CMY family (CMY-71, CMY-80, CMY-96, CMY-102, CMY-104, CMY-155, CMY-179) had only one or two aa differences within this region (Fig. 5B) (according to BLDB, last update of the database: January 22, 2025).

The sequence analysis of putative MAb 2E11 recognition site was subsequently performed. The alignment of MAb 2E11 binding regions (38–44, 299–311 and 362–379 aa of CMY-34) with CMY β-lactamase family, revealed high sequence homology to most of the CMY enzymes. It was determined that 38–44 aa region of CMY-34 was identical to all identified members of CMY family. Moreover, there were only one or two aa mismatches in 19 CMY variants observed, when 299–311 aa region of CMY-34 was aligned (Table S2). Analysis of 362–379 aa region revealed that 48 CMY variants had one aa and five CMY variants had two aa differences in this segment (Table S3).

For further analysis of MAb cross-reactivity, MAbs 9D2 and 2E11 were tested with 12 recombinant β-lactamases by indirect ELISA and WB (Figs. 5C, 5D). Selected recombinant enzymes represent different classes of β-lactamases. According to Ambler’s classification (Ambler, 1980), for cross-reactivity testing used recombinant IMI-1, KPC-2, SHV-42 and SME-3 β-lactamases correspond to class A, recombinant IMP-1 and NDM-1—class B, recombinant ACT-14, ADC-144, DHA-1, PDC-195—class C, recombinant OXA-48 and OXA-134 variants—class D β-lactamases. After analysis, weak cross-reactivity of MAb 9D2 with recombinant PDC-195 (rPDC-195) β-lactamase was determined by both WB and ELISA (Figs. 5C, 5D). The alignment of MAb 9D2 epitope (22–34 aa region of CMY-34) with PDC-195 aa sequence revealed a 152–164 aa region of PDC-195 containing one or two aa segments identical to those of CMY-34 and CMY-4 (Fig. 5A). In contrast to MAb 9D2, there was no cross-reactivity of MAb 2E11 with rPDC-195 and the remaining recombinant β-lactamases observed (Figs. 5C, 5D).

Application of MAbs for development of immunoassays

Characterized MAbs 9D2 and 2E11 were applied in sandwich ELISA and rapid analytical tests, such as lateral flow immunoassay (LFIA) and two-photon excitation (TPX) assay (Koskinen et al., 2007), for immunodetection of CMY β-lactamases in bacterial isolates.

In the first step, the ability of MAbs to capture and detect CMY enzymes in bacterial lysates was evaluated by sandwich ELISA. For testing and optimization of the assay, the microtiter plates were coated with MAb 2E11 for antigen capture, while HRP-conjugated MAb 9D2 was utilized as detection antibody, and rCMY-34 was tested as target antigen. The optimal plate type, MAb concentration, immobilization and blocking conditions showing the highest sensitivity of the assay were selected. A standard calibration curve with rCMY-34 was generated (Fig. 6A), and observed limit of detection for rCMY-34 was 43 pg/mL. Moreover, the developed sandwich ELISA was tested for detection and quantification of CMY β-lactamases in C. portucalensis (3826Z08, CMY-34 producer) and K. pneumoniae (CCUG60138 and CCUG68728, CMY-4 and NDM-1 producers) lysates with known total protein concentration of 555 μg/mL. Lysates of E. coli BL21 strain and A. chinensis isolate (CCUG74036T, NDM-1 producer) were analyzed as negative controls (Fig. 6B). The determined CMY-34 concentration in the C. portucalensis (3826Z08) lysate was 34.5 ng/mL. Meanwhile, CMY-4 concentration in two analyzed K. pneumoniae (CCUG60138 and CCUG68728) lysates were 9.3 and 10.6 ng/mL, respectively.

In the next step, MAbs 9D2 and 2E11 were applied in rapid analytical test, such as LFIA. For the assay, MAb 2E11 was immobilized on the nitrocellulose membrane and used as a capture antibody. The MAb 9D2 was labeled with 40 nm gold nanoparticles and used for visual detection. The combination of materials used for assembly of the test strips and concentration of MAbs showing the highest sensitivity were selected. Blue contour lines were added onto the test strip for visual identification of the control and test line positions during the testing (Fig. 7A). The limit of detection for rCMY-34 was determined visually and reached 33 ng/mL. Due to limited sensitivity of the camera, a positive test line is not clearly visible in the picture when 33 ng/mL of rCMY-34 was tested (Fig. 7B). The ability of optimized LFIA-based test to detect CMY β-lactamases was examined with seven bacterial isolates producing CMY-2, CMY-4, CMY-6, CMY-16 and CMY-34 (Table 5). For analysis, a single colony of bacteria was mixed with extraction buffer and loaded onto the strip. The LFIA test was able to identify all tested CMY-positive isolates. No reactivity with CMY-negative bacteria, such as NDM-1 producing A. chinensis (CCUG74036T and CCUG74037) or β-lactamase-negative E. coli BL21, K. aerogenes (50796520) or P. mirabilis (50793946) was observed (Fig. 7C, Table 5).

MAbs 9D2 and 2E11 were further tested with TPX assay technique. The combination of MAbs showing the highest sensitivity was selected—MAb 9D2 coated microparticles were used for capturing the antigen, and fluorescently labelled MAb 2E11 was utilized for detection (Fig. 7D). The analytical sensitivity of the assay for rCMY-34 was 0.33 ng/mL, as defined by the value intercept of dose-response curve with three standard deviations (Fig. 7E). Positive results were obtained for all CMY producing isolates with high signal intensity (Fig. 7F, Table 5). No reactivity was observed towards the β-lactamase-negative K. aerogenes (50796520) and P. mirabilis (50793946) isolates.