Introduction

Antimicrobial resistance now ranks among the leading contributors to global mortality, yet the antibacterial pipeline has not kept pace; agents with genuinely novel mechanisms of action remain scarce, and the gap between resistance emergence and drug approval continues to widen.1–3P. aeruginosa sits at the intersection of these two trends. In the 2019 global analysis of pathogen-level mortality, P. aeruginosa was associated with approximately 559,000 deaths worldwide,4 and it remains a dominant cause of hospital-acquired bloodstream and lower respiratory tract infection in immunocompromised and critically ill patients. Case fatality scales with resistance. Pooled 30-day mortality after bloodstream infection with drug-resistant isolates approaches 32.0%, exceeding 40% for carbapenem-resistant P. aeruginosa (CRPA) and 50% for ceftazidime-resistant strains.5,6 Even in children — a population with greater physiological reserve — 60-day in-hospital mortality reaches 17.3%, and one in three develops septic shock.7,8 Reflecting this burden, the World Health Organization placed CRPA in the high-priority group of its 2024 Bacterial Priority Pathogens List.3,9,10 Where carbapenems fail, therapeutic options are few, and non-antibiotic approaches move from adjunct to necessity.

The obstacle, however, is not resistance alone. P. aeruginosa deploys a redundant virulence armamentarium whose components act on non-overlapping arms of host defence. PcrV, the tip protein of the type III secretion system (T3SS), licenses translocation of ExoS and ExoU into host cells and thereby disables phagocytes at the point of contact; exotoxin A (ETA), acting at a distance, ADP-ribosylates eEF2 to arrest host protein synthesis and precipitate apoptotic tissue injury.11,12 Redundancy of this order predicts that neutralizing any single factor will fall short. The clinical record is consistent with that prediction — though, on closer reading, not for the reason usually invoked. Although monoclonal antibodies have achieved marked success in oncology and immune diseases, progress in the anti-infective field has been slow. To date, only a limited number of monoclonal antibodies have been approved for microbial infections. For P. aeruginosa, four therapeutic mAbs have been reported that even entered Phase II/III clinical trials, yet none has been successfully brought to market.13–17 For instance, the anti-PcrV antibody KB001 inhibits the type III secretion system (T3SS) and confers protective effects in preclinical models, whereas the bispecific antibody MEDI3902 (Gremubamab)—targeting the extracellular polysaccharide Psl and the protein PcrV—exhibits potent neutralizing activity in preclinical studies. However, the subsequent Phase II clinical trials of both were unsuccessful. As for the underlying reasons, KB001 was designed as a Fab’ fragment lacking the Fc region; although this was intended to circumvent the risk of pulmonary inflammation, it resulted in the loss of key effector functions such as antibody-dependent cellular phagocytosis (ADCP) and complement-dependent cytotoxicity (CDC), rendering it unable to clear bacteria effectively.18 The failure of MEDI3902, in turn, may be attributable to several interrelated core deficiencies: compromised affinity, severe target-mediated drug disposition,19 and bias in indication selection. These factors compounded one another and together severely weakened its neutralizing and opsonophagocytic killing activities, ultimately leading to its failure to meet the primary efficacy endpoint.20 Taken together, these two programmes point somewhere specific: added valency does not rescue a candidate that lacks effector function or fails to sustain exposure at the infected site.

In addition, many monoclonal antibodies currently in early clinical development are of murine or chimeric origin and present significant translational barriers, such as high immunogenicity, potential induction of human anti-mouse antibody (HAMA) responses, short in vivo half-lives, and limited effector functions, all of which may compromise therapeutic efficacy and safety. These setbacks highlight the urgent need to develop novel, fully humanized antibody candidates capable of targeting alternative or multiple virulence mechanisms.21–23

Our team is dedicated to the development of vaccines and antibodies against P. aeruginosa. In earlier work, our laboratory, through screening based on a reverse vaccinology strategy, was the first to identify PA0833, a protective antigen of P. aeruginosa and an OmpA C-like protein. This protein not only participates in the maintenance of cell wall integrity, biofilm formation, and exopolysaccharide production, but also exacerbates pulmonary inflammatory responses through the TLR2-dependent NF-κB signaling pathway, and confers protection against P. aeruginosa infection.24

Building on this earlier work, we identified through screening that a vaccine combining the three antigens PA0833, ETA, and PcrV confers good protective efficacy. The three targets are complementary by mechanism rather than merely additive in number: PA0833 is surface-exposed and thus available for opsonization, anti-PcrV blocks contact-dependent intoxication, and anti-ETA neutralizes a secreted, diffusible toxin that operates beyond the bacterial surface. That complementarity, not multiplicity as such, motivated the present work.

In the present study, a fully human transgenic mouse platform was used to successfully screen and obtain fully human monoclonal antibodies against these three targets. To overcome the insufficient efficacy encountered by previous single-target antibody therapies in clinical translation, this study set out to physically mix the three fully human mAbs described above to prepare a multi-target combined antibody cocktail, and to evaluate the in vitro function of each antibody as well as the prophylactic and therapeutic effects of the antibody combination in murine systemic infection and pneumonia models, thereby providing a new approach for overcoming P. aeruginosa resistance and for developing therapeutic antibody drugs against multidrug-resistant pathogens.

Materials and Methods Study Design

The objectives of this study were: (i) screen for and identify fully human mAbs with high binding affinity and potent neutralizing or opsonophagocytic activity against PcrV, ETA, and PA0833; and (ii) evaluate the prophylactic and therapeutic efficacy of a triple-antibody combination against P. aeruginosa infection using well-established murine models of systemic infection and pneumonia.

For in vitro experiments, data were derived from three independent biological replicates. For in vivo experiments, the number of mice per group was determined to be 10, which was considered sufficient to detect differences between groups. Mice were randomly assigned to experimental groups. Blinding was not employed in this study. Statistical analyses were performed using appropriate methods, as specified in the respective figure legends.

Animals

Female BALB/c mice, aged 6 to 8 weeks, were purchased from Hunan SJA Laboratory Animal Co., Ltd. and housed in our laboratory’s specific pathogen-free (SPF) facility. Mice were kept ten per cage under a 12-hour light/dark cycle, with free access to water and food. All in vivo experiments were conducted in strict accordance with the guidelines of the Animal Ethics and Experiment Com-mittee of the Third Military Medical University (Approval No. AMUWEC20223336). All procedures complied with national and institutional guidelines for the care and use of experimental animals.

Clinical isolates were obtained from residual blood samples and other clinical specimens collected during routine diagnostic procedures at six hospitals. No additional specimens were prospectively collected specifically for this study. The study protocol was reviewed and approved by the Institutional Review Boards of all participating hospitals. As all specimens were residual samples obtained from routine clinical practice and were anonymized prior to use, the requirement for informed consent was waived by the Ethics Committees.

Anesthesia was induced by intraperitoneal injection of Avertin (20 μL/g; Dowobio, DW3101). At the conclusion of the experiment, mice were placed in a dedicated sealed chamber and euthanised by carbon dioxide inhalation at a rate of 30–70% of the chamber volume per. The euthanasia procedure conducted in accordance with the American Veterinary Medical Association’s Guidelines for the Humane Euthanasia of Animals (2020 edition).

All passive immunizations were administered by intraperitoneal (i.p.) injection at a total injection volume of 500 μL per mouse. In the high-dose combination group, mice received U0151, U0462, and U0057 monoclonal antibodies at a total antibody dose of 150 mg/kg, with each antibody administered at 50 mg/kg (final concentration: 2 mg/mL for each antibody). In the low-dose combination group, mice received a total antibody dose of 15 mg/kg, with each antibody administered at 5 mg/kg (final concentration: 0.2 mg/mL for each antibody). For the single-agent groups, mice were administered the respective mAb alone at the dose defined for each antibody in the corresponding experiments.

To evaluate the prophylactic efficacy of monoclonal antibodies. Mice received an intraperitoneal injection of 500 μL of monoclonal antibodies 2h before infection. Subsequently, they were challenged with wild-type PAO1 or clinical through two routes: intravenously (7.0 × 107 CFU) and intratracheally (1.0 × 107 CFU). Following infection, the mice were continuously monitored for 7 days, and mortality was recorded every 12h. To assess the therapeutic efficacy of the monoclonal antibodies, mice were infected with the wild-type PAO1 strain either intravenously (7.0 × 106 CFU) or intratracheally (1.0 × 106 CFU). Two h post-infection, 500 μL of monoclonal antibodies was administered intraperitoneally. Body weights of the mice were recorded throughout the infection period. For sample collection, mice were euthanized 24h post-infection; bronchoalveolar lavage fluid was collected from intratracheally infected mice, and the proportion of neutrophils in the samples was determined by flow cytometry. Additional mice were euthanized on day 5 post-infection to collect lung tissue homogenates from intratracheally infected mice, as well as liver and kidney tissues and homogenates from intravenously infected mice. Tissue samples were serially diluted, plated on Pseudomonas isolation agar, incubated at 37°C for 12h, and colonies were counted. Liver and kidney sections were stained with hematoxylin and eosin (HE) to evaluate inflammation.

Strains

PAO1 and clinical strains were provided by H. Zen (Army Medical University, Chongqing, China), Clinical strains of 6 P. aeruginosa isolates were collected from 6 hospitals in different districts of China (Supplementary Table S1). Unless otherwise specified, all bacterial strains were cultured in Luria–Bertani broth (Merck, 1.10285) or on Pseudomonas isolation agar (Merck, 17208). Bacterial strains were cultured, washed and diluted with sterile PBS to an appropriate cell concentration determined spectrophotometrically at 600 nm (OD600).

Generation of Fully Human Monoclonal Antibodies

Hybridoma generation and screening. Human immunoglobulin transgenic mice (CAMouseHG; Camsehg Biotechnology, Nanjing, China) were immunized with recombinant PcrV, ETA, or PA0833. Splenocyteswere isolated and fused with myeloma cells to generate hybridomas. Antigen-specific hybridomas were identified by enzyme-linked immunosorbent assay (ELISA), and stable monoclonal lines were established by limiting dilution cloning. Clones were ranked by antigen-binding activity, as determined by ELISA, and on the stability of antibody secretion during continuous culture. Representative clones were taken forward for further characterization, including anti-PcrV antibodies (U0151, U0543, U0091, U0402, U0445, U0396, U0271, U0245, and U0077), anti-exotoxin A (ETA) antibodies (U0836, U0251, U0256, U0223, U0462, U0060, U0322, U0674, U0549, and U0138), as well as anti-PA0833 antibodies (U0227, U0277, U0120, U0057, U0172, and U0153).

Vector construction. Total RNA was extracted from the selected hybridoma clones and the immunoglobulin heavy-chain (VH) and light-chain (VL) variable-region sequences were determined (GenScript, Nanjing, China). Verified fragments were cloned into a modified pcDNA3.4 mammalian expression vector contain human IgG1 heavy-chain and light-chain constant regions. The heavy-chain and light-chain plasmids were propagated separately in E. coli DH5α competent cells (Tiangen, Cat. No.: CB101-01), and endotoxin-free plasmid DNA was prepared using an endotoxin-free plasmid DNA extraction kit (Omega, Cat. No.: D692-01).

Transient expression. HEK 293F cells (Thermo Fisher Scientific, Catalog No. R79007) were co-transfected with the heavy-chain and light-chain plasmids at a mass ratio of 2:1 (10 ug total DNA per mL of culture) using PEI transfection reagent (BIOENGINE, Catalog No. TP0200903, DNA: PEI =1:5, w/w). Cells were cultured in HEK293 expression medium (OPM Biosciences, Catalog No. 81075–001) at 37°C and 5% CO2, with shaking at 120 rpm, and the corresponding feed supplement was added on days 1 and 2 post-transfection.

Purification and characterisation. Culture supernatants were harvested 7 d post-transfection by centrifugation, clarified by depth filtration (0.8–1.2 um) and passed through a 0.22-um membrane. Antibodies were captured on Protein A and eluted with 25 mM citrate-phosphate buffer (pH 3.6); the eluates were immediately neutralized with 2 M Tris-HCl (pH 9.0) to pH 5.0. Eluates were then concentrated and buffer-exchanged into PBS (0.02 mM sodium phosphate, pH [6.5]) using 30-kDa MWCO centrifugal ultrafiltration devices (Cat. No. UFC203024, Millipore Sigma, Burlington, MA, USA). Protein concentration was determined by absorbance at 280 nm (NanoDrop, Thermo Fisher Scientific) using an extinction coefficient of 1.4 mL mg−1 cm−1; size-exclusion HPLC (Thermo Fisher Scientific) determined monomer content on a TSKgel G3000SWXL column. All preparations used in this study had a concentration of ≥6 mg/mL, a monomer content of ≥95%, and an endotoxin level of <1 EU/mg.

Enzyme-Linked Immunosorbent Assay (ELISA) for Binding Activity

The binding activity of mAbs to P. aeruginosa PcrV, ETA, or PA0833 antigens was determined by enzyme-linked immunosorbent assay (ELISA). Briefly, microplate wells (Thermo Labsystems) were coated overnight at 4°C with 200 ng per well of PcrV, ETA, or PA0833 protein in 0.05 M carbonate buffer (pH 9.5). After washing three times with PBS containing 0.05% Tween-20 (PBST), the wells were incubated with serial dilutions of mAbs as primary antibodies for 1 h at 37°C. Following three additional washes, the wells were incubated with HRP-labeled goat anti-human IgG (Tianjin Sunory Biotechnology Co., Ltd., Tianjin, China) as the secondary antibody for 1 hour at 37°C. After a final wash, the signal was developed using TMB substrate solution, and the reaction was stopped by adding 1 M H2SO4. The optical density at 450 nm was measured using a microplate reader.

Biolayer Interferometry (BLI) Experiments

Antibody-antigen binding affinity was evaluated using an Octet RED96 system (ForteBio). Antibody sensor (Octet, 18–5142) was used. Binding kinetics were analyzed by globally fitting association and dissociation data obtained from antigen concentrations ranging from 3.13 to 100 nM. Using Octet Analysis Studio 12 software, applying a 1:1 binding model to determine the equilibrium dissociation constant (Kd).

Hemolysis Inhibition Assay for Evaluating mAb Neutralizing Activity

wPcrV and mAb were diluted with PBS. U0543, U0245, U0396, U0271, U0077, U0445, U0151, U0091, and U0402 were each diluted to 200 μg/mL, corresponding to approximately 1.334 nmol/L (calculated based on an IgG1 molecular weight of 150 kDa; 1 mg IgG1 ≈ 6.67×10−9 mol). Subsequently, 12 gradients of 2-fold serial dilutions were performed, with a final volume of 200 μL per well. In two 96-well round-bottom plates, 75 μL of diluted wPcrV (10 μg/mL) was added to each well in rows A–F. Each mAb was then added to the corresponding wells according to a continuous dilution gradient from left to right, from high to low concentration, at 75 μL per well, with three replicate wells set for each concentration. The positive control was wPcrV at the optimal hemolytic concentration (150 -μL, 5 μg/mL), and the negative control was 150 μL PBS. After incubation at 37°C for 30 min, 150 μL of rabbit red blood cell suspension was added to each sample well, and incubation continued at 37°C for 60 min. Subsequently, centrifugation was performed at 1800 rpm for 5 min, 200 μL of supernatant was transferred to new 96-well flat-bottom plates, and absorbance at 540 nm (OD540) was measured.

The red blood cell hemolysis rate was calculated according to the following formula:

Hemolysis Rate = 100%- [(Mean OD540 of sample wells − Mean OD540 of negative control wells)/(Mean OD540 of positive control wells − Mean OD540 of negative control wells)] × 100%.

Cell Viability Assay for Neutralizing ETA-Induced Cytotoxicity

Each mAb was serially diluted 2-fold in PBS to generate seven concentration gradients, with a final volume of 150 μL per gradient. Then, 150 μL of ETA (10 μg/mL) was added to the wells containing the correspondingly diluted mAb. All samples were incubated at 37°C for 30 min. Subsequently, 100 μL of the antibody-antigen mixture was transferred to cell-containing wells and incubated for an additional 6 h. Following the instructions of the CCK-8 Cell Viability Assay Kit, 20 μL of CCK-8 reagent was added to each well. After incubation at 37°C for 1 h, absorbance at 450 nm was measured to assess cell viability. The control wells included cells cultured in DMEM and treated with CCK-8 solution. The blank control wells consisted of DMEM and was treated with CCK-8 solution.

Neutralization rate was calculated according to the following formula:

Neutralization rate =100%- [(Mean OD450 value of control wells -Mean OD450 value of antibody-antigen mixture wells)/(Mean OD450 value of control wells -Mean OD450 value of blank control wells] × 100%.

Opsonophagocytic Killing Assay

The opsonophagocytic killing (OPK) activity of monoclonal antibodies was evaluated using U-bottom 96-well plates with 1× Hank’s Balanced Salt Solution (HBSS) as the assay buffer. Each reaction well contained 20 μL of phosphate-buffered saline (PBS) or monoclonal antibody at specified concentrations, 10 μL of P. aeruginosa PAO1 suspension (6 × 105 CFU/mL), and 50 μL of rabbit or mouse whole blood supplemented with 1000 U/mL recombinant hirudin to inhibit coagulation. After incubation at 37°C for 1 h, 10 μL aliquots from each reaction mixture were serially diluted and plated onto Pseudomonas isolation agar. Colony-forming units were counted after overnight incubation at 37°C. The bactericidal efficacy of the monoclonal antibodies was calculated as the percentage reduction in CFU counts compared to those in the PBS-treated control wells.

Analysis of Pulmonary Inflammation in a Mouse Pneumonia Model

To evaluate the prophylactic efficacy of monoclonal antibodies, mice received an intraperitoneal injection of 500 μL antibody solution 2 hours prior to infection. Subsequently, the mice were intratracheally challenged with 5×106 CFU, corresponding to the LD50 dose. At 24 h post-infection, bronchoalveolar lavage fluid (BALF) was collected from each mouse. Briefly, 0.5 mL sterile normal saline was slowly instilled into the lungs via a catheter and then gently aspirated. This lavage procedure was performed three additional times, yielding a total BALF volume of 1.5 mL per mouse. The pooled BALF samples were first incubated with anti-CD16/CD32 antibodies to block nonspecific Fc receptor binding. The cells were subsequently stained with PE/Cy7-conjugated anti-mouse CD45 and APC/Cy7-conjugated anti-mouse Ly-6G antibodies, which serve as leukocyte and neutrophil markers, respectively. Neutrophils (CD45+ Ly-6G+) were quantified by flow cytometry, and their absolute numbers and proportions were determined.

Histological Analysis

Mice were euthanized on day 5 post-infection for histopathologic assessment. Kidneys and liver were collected from the systemic infection model; lung were collected from the pneumonia model, inflated before fixation, and all fixed in 10% neutral-buffered formalin. All tissues were embedded in paraffin, sectioned, and stained with hematoxylin and eosin (H&E). Sections were examined at 20 × magnification by pathologys blinded to group allocation. Grading followed predefined organ-specific criteria adapted from the semiquantitative principles described by Klopfleisch (2013).25 Kidney sections were scored for inflammatory cell infiltration, tubular injury, interstitial edema, and structural disruption; liver sections for inflammatory cell infiltration, hepatocellular degeneration or necrosis, sinusoidal congestion, and disruption of hepatic architecture; lung sections for hyperemia, edema, hemorrhage, and neutrophil infiltration. A score of 0 indicated normal histology, whereas scores of 1, 2, 3, and 4 represented minimal, mild, moderate, and severe pathological changes.

Statistical Analysis

Survival was estimated by the Kaplan-Meier method, each treatment group was compared with the saline group control by the Log-rank (Mantel-Cox) test. Normally distributed variables are reported as mean ± SD, or as mean ± SEM that indicated in the figure legends, and were compared by unpaired two-tailed Student’s t test for two groups or by one- or two-way ANOVA for multiple groups, with Dunnett correction for comparisons against a single control and Tukey correction for all pairwise comparisons. Variables that departed from normality, including the ordinal histopathologic scores, are reported as median (IQR) and were compared by the Mann–Whitney U-test with Dunnett correction. P-value < 0.05 was considered statistically significant. All statistical analyses were performed using GraphPad Prism 8.0.2 software (GraphPad Software, San Diego, CA, USA) and SPSS 28.0 (IBM Corp, Armonk, NY, USA).

Statement on the Use of AI-Assisted Writing

ChatGPT version 5.5 was used solely for language polishing and grammatical improvement during manuscript preparation. The authors confirm that all scientific content, data interpretation, conclusions, and references are original and accurate. The authors take full responsibility for the integrity of the entire manuscript and have verified the terms of use of the AI tool to ensure its suitability for publication.

Results Preparation of Humanized Monoclonal Antibodies Against Pseudomonas aeruginosa Antigens PcrV, ETA, and PA0833

To obtain humanized monoclonal antibodies targeting the key antigens PcrV, ETA, and PA0833 of P. aeruginosa, this study employed a fully human transgenic mouse platform (CAMouseHG, Chongqing Jinmaibo Biotech Co., Ltd.) for antibody screening and production. Approximately 2000 hybridoma clones were initially generated, constituting a heterogeneous cell population with diverse antibody-secreting capacities (Figure 1A). Following the removal of unfused cells using HAT medium, enzyme-linked immunosorbent assay (ELISA) was further applied to screen the binding activity of hybridoma supernatants, which rapidly eliminated a large number of clones lacking antigen-binding capability. Ultimately, 57 specific hybridoma clones were obtained.

Figure 1 The figure caption outlines the process of preparing monoclonal antibodies, as well as the characterization of the binding properties of candidate antibodies targeting the three antigens PcrV, ETA and PA0833 through the enzyme-linked immunosorbent assay (ELISA). (A) and binding characterization of antibodies targeting PcrV (B), ETA (C), and PA0833 (D) via ELISA. (A) Schematic Illustration of Monoclonal Antibody Preparation. Following the primary immunization of experimental mice, bone myeloma cells and immune spleen cells were isolated and fused to generate hybridoma cells. Subsequent cloning, screening, and functional evaluation yielded positive hybridoma clones, which were then subjected to antibody cloning and recombinant expression for final antibody production. (B) ELISA binding curves of anti-PcrV antibodies. All nine antibodies (U0151, U0543, U0091, U0402, U0445, U0396, U0271, U0245, and U0077) exhibited concentration-dependent binding, with U0151 showing the strongest binding ability. (C) ELISA binding curves of anti-ETA antibodies. All ten antibodies (U0836, U0251, U0256, U0223, U0462, U0060, U0322, U0674, U0549, and U0138) demonstrated concentration-dependent binding, and U0462 had the strongest binding capacity. (D) ELISA binding curves of anti-PA0833 antibodies. All six antibodies (U0227, U0277, U0120, U0057, U0172 and U0153) showed concentration-dependent binding. In the figure, all negative control are IPI, and blank control are PBS. All concentrations in the figure are expressed in (µg/mL).

On this basis, with an OD450 value greater than 2.0 set as the inclusion criterion, the in vitro binding capacity and neutralizing activity of these clones were further evaluated. As a result, 9, 10, and 6 unique antibody sequences targeting PcrV, ETA, and PA0833, respectively, were ultimately screened out. The confirmed antibody sequences were cloned into the pcDNA3.4 eukaryotic expression vector via homologous recombination, and transient transfection was performed in 293F cells to achieve milligram-level expression of monoclonal antibodies.

To evaluate the antigen-binding activity of the obtained antibodies, ELISA was performed to determine the binding capacities of different antibody concentrations to their corresponding target antigens. The results showed that all nine anti-PcrV antibodies (U0151, U0543, U0091, U0402, U0445, U0396, U0271, U0245, and U0077) exhibited concentration-dependent binding behavior. Specifically, as the antibody concentration (log10 μg/mL) increased, the absorbance at 450 nm (A450) gradually increased until reaching a plateau (Figure 1B). Among them, the saturated A450 value of U0151 approached 2.0, indicating a significantly higher binding capacity than that of the other eight antibodies.

For the ETA antigen, ten anti-ETA monoclonal antibodies (U0836, U0251, U0256, U0223, U0462, U0060, U0322, U0674, U0549, and U0138) also showed concentration-dependent binding. At equivalent antibody concentrations, U0462 exhibited the highest A450 value, suggesting the strongest binding capacity (Figure 1C). Meanwhile, the six antibodies targeting the PA0833 antigen (U0227, U0277, U02120, U0057, U0172, and U0153) similarly displayed concentration-dependent binding curves (Figure 1D).

In control experiments, neither human IgG1 (hIgG1) nor 1% PBS buffer produced significant binding signals with the three aforementioned antigens (A450 < 0.2), demonstrating that the selected antibodies possess high specificity for their respective target antigens.

Characterization of Monoclonal Antibodies Against Pseudomonas aeruginosa Proteins PcrV, ETA, and PA0833

Anti-PcrV monoclonal antibodies showed varying neutralizing capabilities against PcrV-mediated erythrocyte lysis in hemolysis inhibition experiments. With a half-maximal inhibitory concentration (IC50) of 5.80 nM, U0151 exhibited the strongest inhibitory action. It was followed by U0543 (IC50 = 34.7 nM), U0091 (IC50 = 57.3 nM), U0402 (IC50 = 82.4 nM), U0445 (IC50 = 121.5 nM), and U0396 (IC50 = 174.1 nM). In contrast, under the experimental circumstances, U0271, U0245, and U0077 were unable to produce detectable neutralization (Figure 2A).

Figure 2 Functional Screening and Characterization of Monoclonal Antibodies. (A) To evaluate the ability of monoclonal antibodies to neutralize the membrane-perforating activity of the virulence factor PcrV in red blood cells. Nonlinear regression was used to fit the data and determine the half-maximal inhibitory concentration (IC50). (B) Monoclonal antibody neutralizing effects on ETA-induced cytotoxicity in A549 cells are compared. The IC50 was determined by fitting the data using nonlinear regression. (C) Anti-PA0833 monoclonal antibodies’ relative OPK activity. (D) Preventive protection effectiveness of potential monoclonal antibodies in a model of systemic infection in mice. A systemic infection model was established by injecting a bacterial suspension of the wild-type PAO1 strain (7.0 × 107 CFU/mL) into the tail veins of female BALB/c mice aged 6–8 weeks (n = 10 per group). A prophylactic dose of 100 mg/kg of monoclonal antibodies was given before the bacterial challenge. For seven days, survival was tracked. The Mantel-Cox Log rank test was used to assess statistical significance between groups. (E) Bio-layer interference (BLI) technology was used to measure the antigen–antibody binding affinity of candidate antibodies (U0151, U0462, and U0057).

All ten anti-ETA antibodies demonstrated the ability to reduce ETA-induced cell death in cytotoxicity neutralization tests using A549 cells. Interestingly, five clones exhibited dose-dependent neutralizing efficacy: U0836 (IC50 = 65.56 μM), U0251 (IC50 = 35.36 μM), U0256 (IC50 = 210.4 μM), U0223 (IC50 = 62.4 μM), and U0462 (IC50 = 7.95 μM). Neutralization was not statistically significant for the remaining five antibodies (Figure 2B).

OPK assays revealed that, in the presence of phagocytes and complement, U0057 elicited robust bactericidal activity against PA0833-expressing P. aeruginosa, significantly surpassing U0172 (P = 0.0269), U0277 (P = 0.0170), U0227 (P = 0.0127), U02120 (P = 0.0198), and U0153 (P = 0.0114) (Figure 2C).

The prophylactic protective effect of antibodies was evaluated in a mouse systemic infection model (antibody dose of 100 mg/kg): the protection rates of the anti-PcrV antibody U0151, the anti-ETA antibody U0462, and the anti-PA0833 antibody U0057 were 50%, 44%, and 37.5%, respectively (Figure 2D). Further analysis of the binding affinity of these three antibodies using biolayer interferometry (BLI) showed KD values of 2.69 nM for U0151, 1.75 nM for U0462, and 0.34 nM for U0057 (Figure 2E), indicating high affinity and potential therapeutic value.

Protective Efficacy of Monoclonal Antibodies in a Mouse Systemic Infection Model of Pseudomonas aeruginosa PAO1

To evaluate the prophylactic effects of anti-PcrV monoclonal antibody U0151, anti-ETA monoclonal antibody U0462, and anti-PA0833 monoclonal antibody U0057, this study employed an acute lethal dose systemic infection model with P. aeruginosa PAO1 (2 × LD50, 7.0×107 CFU/mouse, intravenous tail injection). Mice were observed continuously for 7 days after challenge, and deaths were recorded every 12 h. Survival rate data were plotted using GraphPad Prism 8.0.2, and differences between groups were analyzed by Log-rank (Mantel-Cox) test (Figure 3A).

Figure 3 The preventive and therapeutic effects of monoclonal antibodies in a mouse systemic model. (A) BALB/c mice (n = 10) were immunized with monoclonal antibodies and challenged with PAO1 at 7.0×107 CFU/mouse by intravenous tail injection. The survival rate was monitored for 1 week. P-value was calculated using the Mantel–Cox Log rank test, with the saline group used as the control. (B and C) BALB/c mice (n = 10) were infected with PAO1 (7.0×106 CFU/mouse, intravenous tail injection) and treat with monoclonal antibodies. (B) Weight in infected mice (n = 10) immunized with monoclonal antibodies. The data are presented as mean ± SD. Body weight was analyzed using two-way repeated-measures ANOVA followed by Dunnett’s multiple comparisons test, with the saline group used as the control. Statistical significance was defined as P ≤ 0.05 (*), P ≤ 0.01 (**), and P ≤ 0.001 (***), compared with the Saline group. (C) The number of viable bacteria in the kidney and liver of mice (n = 10) at 5-day post-infection are shown. Data are presented as mean ± SD. Bacterial burdens were analyzed using one-way ANOVA followed by Dunnett’s multiple comparisons test, with the saline group used as the control. (D) Shows six representative images of pathological sections of kidney and liver: three for kidney and three for liver, from the control group (left), 15 mg/kg group (middle), and 150 mg/kg group (right), respectively. (E) Shows semi‑quantitative analysis of renal and hepatic inflammation in infected mice. The scores of kidney and liver inflammation severity in immunized and control mice at 5 days after infection are shown (n=10). Data are presented as median (interquartile range); P values were from Mann‑Whitney U-test, compared with the saline control group.

In the normal saline negative control group, only one mouse survived. The survival rates of mice passively immunized in the single-drug groups (U0151, U0462, or U0057, all at a dose of 5 mg/kg) were 30%, 30%, and 20%, respectively, showing no significant difference compared to the control group (P >0.05). The combination drug group (U0151 and U0462 and U0057) exhibited dose-dependent protection: the survival rate of the low-dose combination group (15 mg/kg, total dose) was 50%, significantly higher than that of the control group (P = 0.0429); the survival rate of the high-dose combination group (150 mg/kg, total dose) was 80%, with an even more significant difference (P = 0.0026). The protective rates of high-dose combination group, low-dose combination group, single U0151 group, single U0462 group, and single U0057 group were 78%, 44%, 22%, 22%, and 11%. Combination therapy significantly extended the survival time of mice, demonstrating a synergistic prophylactic effect.

In the systemic P. aeruginosa infection model evaluating the therapeutic efficacy of monoclonal antibodies, mice were intravenously infected with 7.0×106 CFU/mouse of PAO1. Body weight changes were monitored during the first 5 days after infection (Figure 3B).

Monotherapy at 5 mg/kg conferred little protection against challenge-induced weight loss. U0151 separated from saline only at day 3 (adjusted P = 0.037) and U0462 only at day 5 (adjusted P = 0.047); no other single antibody reached significance at any time point. The three-antibody combination behaved differently. Because the low-dose combination delivered 5 mg/kg of each antibody — matching the monotherapy dose on a per-antibody basis — it provides the interpretable test of whether combining the three specificities adds benefit. At 15 mg/kg total, weight loss was attenuated relative to saline from day 3 onward (D3, adjusted P = 0.048; D4, P = 0.024; D5, P = 0.0009). The high-dose combination (150 mg/kg total; 50 mg/kg per antibody) advanced the onset of protection to day 2 and produced the largest separation (D2: adjusted P = 0.0490; D3: adjusted P = 0.0014; D4: adjusted P = 0.0003; D5: adjusted P = 0.0015). Two-way repeated measures ANOVA across six groups and six time points (n = 10 per group; all animals survived to day 5 and contributed complete series) revealed significant main effects of time (F(3.189, 172.2) = 150.9, P < 0.0001), treatment group (F(5, 54) = 4.556, P = 0.0015), and a significant interaction between time and treatment (F(25, 270) = 2.527, P = 0.0001), indicating that the effects of the treatments varied over time.

At the end of the 5-day observation period, substantial bacterial colonization remained in the liver and kidneys of untreated mice, whereas bacterial loads in the livers and kidneys of monoclonal antibody–treated mice were significantly reduced (Figure 3C).

Histopathology gave a partially concordant picture. Severe renal lesions were observed in the saline-treated mice, characterized by extensive inflammatory cell infiltration and marked disruption of the normal renal architecture; hepatic showed minor inflammatory changes, limited to slight sinusoidal congestion and scattered portal infiltrates, with architecture broadly retained. After combination treatment, exhibited substantially reduced inflammatory cell infiltration, and the normal tissue architecture of both the kidneys and liver was largely preserved without obvious necrosis or severe structural damage (Figure 3D). Consistent with these histological observations, semi-quantitative histopathological scores of both the liver (Low dose vs Saline, P =0.021; High dose vs Saline, P<0.001) and the kidney (High dose vs Saline, P =0.024) were significantly lower in the combination treatment groups than in the saline group (Figure 3E). The high-dose group (150 mg/kg, total dose) cleared more bacteria than the low-dose group (15 mg/kg, total dose), yet histopathological scores in the two groups were indistinguishable, suggesting that the lower dose was sufficient to confer effective therapeutic protection. These histopathological findings were consistent with the bacterial burden assay, demonstrating that combination therapy effectively reduced bacterial colonization in infected organs, attenuated infection-induced inflammatory responses, and protected host tissues from pathological damage. Taken together, the combination reduced organ bacterial burden and alleviated infection-induced tissue injury at doses at which the individual antibodies were, highlighting the protective efficacy of these antibodies against P. aeruginosa infection in vivo.

Protective Efficacy of Monoclonal Antibodies in a Mouse Pneumonia Model of Pseudomonas aeruginosa PAO1

To further explore the protective effects of anti-PcrV monoclonal antibody U0151, anti-ETA monoclonal antibody U0462, and anti-PA0833 monoclonal antibody U0057, mice were challenged with PAO1 (2×LD50, 1.0×107 CFU/mouse, intratracheal injection) 2 h after passive immunization and observed continuously for 7 days, with deaths recorded every 12 h. Survival rate data were plotted using GraphPad Prism 8.0.2, and differences between groups were analyzed by Log-rank (Man tel-Cox) test (Figure 4A).

Figure 4 Protective efficacy of monoclonal antibodies in a murine P. aeruginosa pneumonia model. (A) BALB/c mice (n=10) were immunized with monoclonal antibodies and challenged with PAO1 at 1.0×107 CFU/mouse by intratracheal injection. The survival rate was monitored for 1 week. P-value was calculated using the Mantel–Cox Log rank test, with the saline group used as the control. (B–F) The immunized mice and control mice were infected intratracheally with 1.0×106 CFUs/mouse of PAO1. (B) Weight in infected mice (n = 10) immunized with monoclonal antibodies. The data are presented as mean ± SD. Body weight was analyzed using two-way repeated-measures ANOVA followed by Dunnett’s multiple comparisons test, with the saline group used as the control. Statistical significance was defined as P ≤ 0.01 (**) compared with the Saline group. (C) The number of viable bacteria in the lungs of mice (n = 10) at 5-day post-infection are shown. Data are presented as mean ± SD. Bacterial burdens were analyzed using one-way ANOVA followed by Dunnett’s multiple com-parisons test, with the saline group used as the control. (D) Evaluation of neutrophil infiltration in infected mice (n=10). The bar represents the percentage and the number of neutrophils in the BALF of immunized mice at 24 h post challenge. The P-value using one-way ANOVA followed by Dunnett’s multiple comparisons test, with the saline group used as the control. (E) Three representative images are shown. They are, respectively, the control group (left), the 15 mg/kg group (middle), and the 150 mg/kg group (right). (F) Semiquantitative analysis of pulmonary inflammation in infected mice. Lung severity scores of immunized mice and control mice at 5 days post-infection are shown (n = 10). Data are shown as median (IQR); P values are from Mann–Whitney U-tests versus saline control.

In the normal saline negative control group, only two mice survived. The survival rates of mice passively immunized in the single-drug groups (U0151, U0462, or U0057, all at a dose of 5 mg/kg) were 50%, 50%, and 40%, respectively, showing no significant difference compared to the control group (P >0.05). The combination drug group (U0151 and U0462 and U0057) exhibited dose-dependent protection: the survival rate of the low-dose combination group (15 mg/kg, total dose) was 70%, significantly higher than that of the control group (P = 0.0184); the survival rate of the high-dose combination group (150 mg/kg, total dose) was 90%, with an even more significant difference (P = 0.0026). The protective rates of high-dose combination drug group, low-dose combination drug group, single U0151 group, single U0462 group, and single U0057 group were 88%, 63%, 38%, 38%, and 25%. Combination therapy significantly extended the survival time of mice, demonstrating a synergistic prophylactic effect.

In the pneumonia model evaluating the therapeutic efficacy of monoclonal antibodies, mice were intratracheally infected with 1.0×106 CFU/mouse of PAO1. Body weight changes were monitored during the first 5 days after infection (Figure 4B). Our results indicated that while single-agent treatments at 5 mg/kg were in-effective separated from saline at any time point, the combined administration of U0151, U0462, and U0057 significantly mitigated the challenge-induced body weight reduction. The high-dose combination (150 mg/kg total) weight loss was attenuated relative to saline from day 3 onward with the Saline group (D3: adjusted P = 0.0047; D4: adjusted P = 0.0012; D5: adjusted P = 0.0020). The low-dose combination (15 mg/kg total) — the arm that matches monotherapy on a per-antibody basis —separation emerged only at the final time point (D5: adjusted P = 0.0019).

At the end of the 5-day observation period, substantial bacterial colonization remained in the lungs of untreated mice, whereas bacterial loads in the lungs of monoclonal antibody–treated mice were significantly reduced (P <0.0001; Figure 4C).

Furthermore, neutrophil counts in bronchoalveolar lavage fluid (BALF) were analyzed 24 h post-infection. In the saline control group, neutrophils accounted for nearly 60% of the total leukocyte population. Regarding the neutrophil proportions, the high-dose combination group (150 mg/kg), low-dose combination (15 mg/kg), U0151, and U0462 monotherapies significantly reduced the percentage of neutrophils compared with the saline group (all P < 0.0001); the U0057 monotherapy also significantly reduced the proportion (P = 0.0015; Figure 4D, upper panel).

Consistent with the changes in proportions, the absolute neutrophil counts were also significantly decreased by the high-dose combination (P < 0.0001), the low-dose combination (P < 0.0001), as well as the U0151 (P = 0.0025) and U0462 (P = 0.0053) monotherapies (Figure 4D, lower panel). However, treatment with U0057 alone did not lead to a statistically significant reduction in absolute neutrophil counts compared with the saline control (P = 0.1583; Figure 4D, lower panel). Collectively, these data demonstrate that the antibody combination therapy, particularly the high-dose combination, effectively alleviated pulmonary inflammatory responses.

Histological analysis revealed severe pulmonary lesions in the saline-treated mice at 5 days post infection, characterized by extensive inflammatory cell infiltration, thickened and congested of the alveolar septa, and loss of alveolar architecture (Figure 4E). In contrast, treatment with the combination of U0151, U0462, and U0057 markedly alleviated these pathological changes. Both combination arms reduced infiltration and septal thickening, with alveolar architecture broadly retained. (Figure 4E and F). Consistent with these histological observations, semi-quantitative histopathological scores were significantly lower in both combination treatment groups than in the saline group, with significant reductions observed in the 15 mg/kg group (P = 0.0287) and the 150 mg/kg group (P < 0.001) (Figure 4F). Although the high-dose group showed a trend toward lower histopathological scores and less severe pulmonary lesions than the low-dose group, no statistically significant difference was observed between the two treatment groups, suggesting that both dosing regimens effectively protected against infection-induced lung injury. These histopathological findings were consistent with the bacterial burden assay (Figure 4C), demonstrating that combination therapy effectively reduced bacterial colonization, attenuated infection-induced pulmonary inflammation, preserved lung tissue architecture, and protected the host from pathological damage. Collectively, these findings indicate that the combination of U0151, U0462, and U0057 markedly alleviated lung injury caused by P. aeruginosa infection, reduced infection-associated inflammatory responses, and preserved pulmonary architecture, highlighting its protective efficacy against P. aeruginosa-induced pneumonia in vivo.

Evaluation of the Broad-Spectrum Protective Efficacy of the Triple Antibody Combination Therapy Against Clinical Isolates of Pseudomonas aeruginosa

To assess the broad-spectrum protective potential of the triple antibody combination—comprising anti-PcrV antibody U0151, anti-ETA antibody U0462, and anti-PA0833 antibody U0057—we established mouse models of systemic infection and pneumonia using clinical P. aeruginosa isolates, and systematically evaluated the in vivo protective efficacy of the combination therapy.

Following challenge with clinical isolates XN-1, ZJ-008, ZJN-04, BJ-15, GZ-18, and KM-9, mice in the saline control group exhibited a progressive decline in survival over the 7-day observation period. In stark contrast, mice treated with the triple antibody combination demonstrated survival rates of 70%, 60%, 50%, 70%, 60%, and 50% against the respective isolates (Figure 5A). Log rank tests confirmed that the survival curves of the combination therapy groups were statistically distinct from those of the control group for all tested isolates (P < 0.05).

Figure 5 Broad-spectrum protective efficacy of the triple antibody combination therapy against clinical isolates of P. aeruginosa. (A) Systemic infection model: BALB/c mice (n = 10 per group) were challenged intravenously with different clinical isolates (1.0 × 107 CFU/mouse). The control group received saline, while the prophylactic group received U0151, U0462, and U0057. (B) Pneumonia model: Mice were challenged with an equal bacterial inoculum (1.0 × 107 CFU/mouse) via intratracheal instillation, with the same prophylactic regimen as in (A). Survival was monitored for 1 week, with the normal saline group serving as the control, and P values were calculated using the Mantel–Cox Log rank test.

Consistent with the findings from the systemic infection model, similar protective effects were observed in the pneumonia models. Mice treated with the triple-antibody combination achieved survival rates of 60%, 70%, 50%, 50%, 60%, and 70% against the six isolates, respectively (Figure 5B). Furthermore, Kaplan–Meier survival analysis combined with Log rank testing demonstrated that the combination therapy significantly improved survival outcomes compared with those in the control group across all infection conditions (P <0.05).

Collectively, the triple antibody combination (U0151, U0462 and U0057) exerted significant protective effects against all tested clinical P. aeruginosa isolates in both systemic infection and pneumonia models. Notably, protective rates in the p