Targeting a conserved flagellin D1 epitope with 16G10 confers cross-serotype protection in murine Pseudomonas aeruginosa infection models Authors: Ying Wang, Yating Wang, Zifan Zhu, Qian Wang, Xinglong Liu, Qingli Cai, Yueyue Zhang, Yuhang Li, Hui Zhao, Xin Cheng, Jiang Gu, Quanming Zou, Jianxin Jiang Published: 2026 ABSTRACT ABSTRACTEscalating resistance in Pseudomonas aeruginosa (PA) underscores the need for resistance-agnostic immunotherapies. In this study, we generated a panel of 20 monoclonal antibodies against the flagellin protein FliC and identified 16G10 as a high-affinity lead. Epitope mapping revealed a conserved linear site within the D1 domain (Gly71–Ile88), a region essential for filament assembly. In vitro, 16G10 induced bacterial aggregation, suppressed swimming motility, and significantly reduced adhesion to and invasion of A549 cells, while limiting biofilm biomass by SEM and confocal 3D imaging. In vivo, in murine pneumonia, where 16G10 administered either after pre-incubation with imipenem-resistant PA strain or as post-infection therapy, it improved survival, lowered lung bacterial loads, reduced IL-6 and TNF-α concentrations, and attenuated histologic inflammation. Protection extended across serotype A and B flagellated strains, indicating broad coverage. These data nominate 16G10 as a promising candidate for immunotherapy of drug-resistant PA lung infections and establish a druggable flagellin epitope for future vaccine and antibody designs.KEYWORDS: Pseudomonas aeruginosa, flagellin, monoclonal antibody, pneumonia, immunotherapy Introduction Pseudomonas aeruginosa (PA) is a major driver of hospital‑acquired pneumonia and ventilator‑associated infections and increasingly threatens community populations with chronic lung disease. In 2024, the World Health Organization listed PA among the high‑priority bacterial threats due to global resistance and clinical impact.1 Extensively drug‑resistant (XDR) and multidrug‑resistant (MDR) strains now circulate widely. A recent analysis reported isolation rates of MDR and XDR PA as high as 11.5%–24.7% and 9.0%–11.2%,2 respectively, underscoring the shrinking efficacy of standard therapies. These resistant infections are linked to prolonged hospitalization, treatment failure, and excess mortality, particularly in immunocompromised patients and those requiring intensive care.3,4 The public‑health burden is substantial: the U.S. Centers for Disease Control and Prevention estimates at least 32,600 hospitalized cases and 2,700 deaths annually from PA infections, with attributable healthcare costs of approximately $767 million in the United States alone.2 There is a pressing need for alternative, resistance‑resilient interventions to prevent and treat PA lung infections effectively. PA pathogenesis is tightly linked to its flagellum, a helical propeller built from polymerized FliC subunits that exist as two antigenic serotypes, classically termed type A (FlaA) and type B (FlaB).5 Flagella drive swimming and swarming, promote initial attachment and facilitate biofilm formation and maintenance.6 Motility depends on precise packing of FliC monomers; perturbations in the extracellular domain or D‑region interfaces alter filament dynamics and dissemination within the host. The FliC protein of PA consists of four domains: D0, D1, D2, and D3. Among them, D0 and D1 are highly conserved, while D2 and D3 show significant sequence and structural variations. Beyond mechanics, FliC is a potent immunostimulant that activates Toll‑like receptor 5 (TLR5) on airway and sinonasal epithelia, indirectly modulating expression of bacterial adhesins.7 During chronic infection, selective pressures favor variants with altered flagellar function. Loss or dysregulation of motility increases intracellular c‑di‑GMP, upregulates Pel/Psl exopolysaccharides, and elaborate thick biofilm matrices with heightened antibiotic tolerance.5 Thus, FliC is considered as a central virulence determinant and an attractive target for anti‑virulence interventions. Vaccine studies have repeatedly shown that FliC‑based antigens can protect against lethal challenge. Intranasal immunization with a bivalent flagellin vaccine elicited robust mucosal immunity and protected mice from fatal pneumonia caused by heterologous flagellated strains.8 Broader formulations combining FlaA, FlaB, and pilin enhanced cross‑serotype protection and survival in multiple models.9,10 Antibody‑based strategies against FliC have likewise demonstrated efficacy. For example, therapeutic administration of an anti‑FlaA monoclonal antibody markedly improved survival in murine PA infection, approaching the benefit of imipenem treatment,11 and avian anti‑flagellin IgY reduced mortality and bacterial loads in respiratory and wound models.12 Another study develop a broad-spectrum monoclonal antibody WVDC-2109, which exhibits in vitro complement-mediated bactericidal activity and enhances opsonophagocytic killing of PA.13 Together, these vaccine and antibody data validate FliC as a clinically relevant antigen for immunoprophylaxis and immunotherapy. Building on this foundation, our study maps a conserved D1‑domain epitope and develops 16G10, a monoclonal antibody that neutralizes flagellar function and mitigates disease, thereby extending the case for FliC‑targeted interventions against drug‑resistant PA. Methods & materials Bacterial strains and cell lines PA103 (serotype O11, ATCC 29260) and PAO1 (serotype O5, ATCC 47085) were obtained from the American Type Culture Collection (ATCC), clinical strains XN-22, XN-361, and XN-1 are stored in the research group’s laboratory strain bank and were revived from frozen stocks before use. Human lung epithelial cells (A549) (ATCC CCL-185™) were cultured in Dulbecco’s Modified Eagle Medium (DMEM; Gibco) supplemented with 10% fetal bovine serum (FBS; Gibco), and all cell cultures were maintained at 37°C in a humidified atmosphere containing 5% CO2. Antibody preparation We chose female BALB/c mice aged 6–8 weeks and acclimatized them for 7 d after purchase. Among them, some of the mice were not subjected to any treatment, and their serum was used as the non-specific antibodies, serving as the most negative control. We performed subcutaneous immunization at five sites on days 0, 14, and 21 using 10 μg of recombinant FliC antigen (20 μL per mouse) mixed with 500 μL of alum adjuvant (10 mg/mL). We collected serum from the mice 7 d post the third immunization, and antibody titer were measured using ELISA. These sera were used as positive anti-FliC polyclonal antiserum controls for anti-FliC in the subsequent experiments. Subsequently, spleen cells were isolated from the immunized mice and fused with SP2/0 myeloma cells in a 10:1 ratio with PEG-1500 facilitating fusion. The fused cells were then seeded into a 96-well plate with feeder cells and selected using HAT medium. On the 10th d, we switched the culture medium to HT medium, and we obtained stable antibody-secreting monoclonal cell lines through limiting dilution. Thereafter, the adherent monoclonal cell lines were adapted to suspension culture by slowly lowering the serum concentration and optimizing the culture conditions. After culturing the suspension cells for 4–7 d, we collected the supernatant, filtered it through a 0.22-μm filter, and purified the antibodies using Protein A/G affinity chromatography. The purity of the antibody product was initially assessed by SDS-PAGE, and the concentration was determined using the BCA method. Endotoxin levels in the purified monoclonal antibodies were then quantified using a kinetic turbidimetric LAL assay (BioProbe system), and acceptable samples were stored at −80°C for subsequent use. Swimming inhibition assay Activate the strains XN-22, XN-361, XN-1, PA103, and PAO1, and prepare a semi-solid culture medium. Add 200 μg of anti-FliC monoclonal antibody, 200 μg of mouse-derived anti-FliC polyclonal antibody, and 200 μg of non-specific antibody to the medium. Use a sterile disposable syringe to aspirate 1 mL of the overnight bacterial culture and inoculate it in the center of the semi-solid culture medium. Incubate overnight in a 37°C incubator; then take photos and record the data. Use ImageJ software to measure the swimming area of the bacteria in each group. Bacterial agglutination assay Cultivate the PA103 strain, and when the bacterial culture reaches an OD600 of 1.0, take 1 mL of the bacterial culture and incubate it with the corresponding concentration of antibody at room temperature for 1 hour. Prepare slides using the Gram‑staining method, and after drying the slides, check for bacterial aggregation using an oil‑immersion objective at 1000× magnification. Bacterial adhesion & invasion cell experiment Revive A549 cells, when they reach a concentration of 1 × 106 to 5 × 106 cells/mL, dilute them to 1 × 106 cells/mL. Then, use a multichannel pipette to transfer the cells to a 96-well cell culture plate. Place the plate in a 37°C, 5% CO2 incubator and incubate overnight to allow cell attachment. Activate the PA103 strain and pre-incubate it with antibodies according to the following groups: PBS control group, 16G10 group (200 μg/mL), non-specific antibody group, and anti-FliC polyclonal antibody group. Mix the secondarily activated bacterial suspension with each group of antibodies and incubate at 37°C for 1 hour. Then, add the bacteria and antibodies to the cells and incubate at 37°C with 5% CO2 for 1 hour (adhesion) or 3 hours (invasion). Aspirate the culture medium and wash each well three times with 250 μL of sterile PBS. Add 100 μL of gentamicin (100 μg/mL) to each well to kill extracellular bacteria; incubate at 37°C with 5% CO2 for 1 hour. Finally, wash away residual gentamicin with sterile PBS. For the invasion experiment, add 0.5% Triton X-100 to lyse the cells. For the adhesion experiment, add 0.25% trypsin-EDTA containing phenol red, treat for 5 minutes, and gently pipette to lyse the cells. Apply different concentrations of the suspension onto antibiotic-free LB plates, allow them to absorb for 10 minutes, then invert the plates and incubate overnight in a 37°C incubator for colony counting the next day. Confocal microscopy analysis of biofilm formation Each test sample–10H6 at concentrations of 0.5 mg/mL, gentamicin at concentrations of 0.3 mg/mL, and PBS as a control–was prepared at 500 μL and mixed with 1 mL of PA103 suspension expressing green fluorescent protein (GFP). The mixture was transferred to a glass-bottom confocal dish and incubated at 37°C for either 1 d or 7 d. After incubation, the bacterial suspension was removed, and the cover glass surfaces were gently rinsed three times with PBS. The samples were fixed with 2.5% glutaraldehyde for 4 hours, dehydrated using graded ethanol and tert-butanol solutions from 30% to 100%, and air-dried. The biofilm was observed with a Zeiss LSM 780 confocal laser scanning microscope, exciting GFP at 405 nm and detecting emission at 450 nm.14 Scanning electron microscopy (SEM) analysis of biofilm formation Sterile six-well plates were used, with each well containing a sterile cover glass of 10 mm diameter. The PA103 bacterial suspension was separately mixed with solutions of antibodies (500 μg/mL), antibiotics (300 μg/mL), or PBS, and incubated at 37°C for 36 hours. After washing with PBS to remove planktonic bacteria and residual medium, the samples were fixed with 2.5% glutaraldehyde at 4°C for 4 hours. The samples were then dehydrated through graded ethanol series from 30% to 100% and graded tert-butanol series from 30% to 100%, with each step lasting 15 minutes. After critical point drying, metal sputtering with gold was performed under vacuum, and bacterial biofilms were observed using scanning electron microscopy. Evaluation of the protective effect of monoclonal antibodies in mice All animal experiments were approved by the Animal Ethical and Experimental Committee of the Third Military Medical University (Approval No. AMUWEC20250100) and performed in accordance with ethical guidelines. Overnight cultures of PA 103 or PA XN-1,15 grown to the logarithmic phase (OD600 ≈ 1.2), were diluted with saline to create different infection doses. The lethal dose for PA 103 was 9.17 × 106 CFU/mouse; for PA XN-1, it was 1 × 107 CFU/mouse. The sub-lethal dose for PA 103 was 3.6 × 106 CFU/mouse. The groups were as follows: (1) Blank control group (PBS treatment); (2) Nonspecific antibody group; (3) Clinical frontline antibiotic group (Imipenem, 10 mg/kg); (4) 16G10 monoclonal antibody group (20 mg/kg). In the co-incubation treatment scheme, the 16G10 monoclonal antibody was pre-incubated with PA 103 at 37°C for 1 hour, and the other control groups were treated separately. The treated bacteria were then injected into the trachea of each mouse. In the lethal dose PA 103 group (n = 10/group), the survival rate was monitored daily for 7 d after infection. For mice infected with the sub-lethal dose, lung tissues were collected 24 hours after infection for hematoxylin and eosin (H&E) staining and pathological analysis (n = 3/group). Inflammation scores were assessed based on bleeding, neutrophil recruitment, and alveolar collapse in the lung tissue. Additionally, bacterial colonization and concentrations of inflammatory factors (TNF-α and IL-6) in lung tissue were measured (n = 5/group). The pathological section scoring was performed independently using a double-blind method. In the therapeutic treatment scheme, 1 hour after infection with the lethal dose, the treatments mentioned earlier were administered across five groups (n = 10/group), and the survival rate was monitored daily for 7 d. One hour after infection with the sub-lethal dose, the 16G10 monoclonal antibody group, nonspecific antibody group, antibiotic group, and PBS group were administered via the airway route, and body weight changes were measured over 7 d (n = 5/group). Lung tissues were collected 24 hours after infection for H&E staining and pathological analysis. Inflammation scores were assessed based on bleeding, neutrophil recruitment, and alveolar collapse in the lung tissue. Bacterial colonization and concentrations of inflammatory factors (TNF-α and IL-6) were also measured (n = 5/group). Dot-blot test The FliC protein sequence (394 amino acid residues) is divided into 42 overlapping peptides of 18 amino acids each, with each peptide overlapping the previous one by 12 amino acids; these peptides are named P1 through P42. The dot blot method is used for linear epitope identification. First, the peptides are dissolved in an appropriate buffer and applied onto a dry nitrocellulose membrane (NC membrane), then left to dry at room temperature. Then, the membrane is immersed in Tris-buffered saline with 0.1% Tween 20 (TBST) containing 5% nonfat dry milk and blocked at room temperature for 1 hour to prevent nonspecific binding. After blocking, the membrane is washed with TBST and incubated in a solution containing monoclonal antibody 16G10 (50 μg/ml) at room temperature in the dark for 1 hour. Next, the membrane is washed 3 to 4 times with a sufficient volume of TBST buffer, each wash lasting 5 to 10 minutes. Subsequently, HRP-conjugated goat anti-mouse secondary antibody (1:7500 dilution in TBST with 5% nonfat dry milk) is added and incubated at room temperature in the dark for 1 hour with gentle shaking. After washing to remove the secondary antibody, ECL solution is added for chemiluminescent development, and the results are recorded. Finally, the grayscale values are analyzed using ImageJ software. Statistical analysis Statistical analysis was conducted using GraphPad Prism 9.5 software. Experimental data is expressed as mean ± standard deviation (xˉ ± s). Comparisons between two groups used unpaired t-tests, and comparisons among multiple groups used one-way ANOVA. Survival data was analyzed with Kaplan–Meier analysis. We considered a difference statistically significant if P < .05. Results Generation, binding kinetics, and functional activity of anti-PA FliC monoclonal antibodies To begin with, we generated 20 ELISA‑positive hybridoma clones and screened them to identify monoclonal antibodies targeting the flagellin protein FliC of PA. Among all candidates, clones 16G10 and 1D6 were identified as the IgG1κ isotype (Figure S1) and produced markedly higher OD450 signals than other clones and negative controls (Figure 1(A)). Bio‑layer interferometry confirmed specific, concentration‑dependent binding, yielding kinetic constants of KD = 7.46 × 10−9 M for 16G10 and KD = 7.65 × 10−8 M for 1D6, indicating higher affinity of 16G10 (Figure 1(B)). In swimming‑motility assays, both antibodies inhibited PA movement, with 16G10 producing a significantly smaller motility zone than 1D6 (P < .05), and both outperforming non‑specific antibodies (P < .05, Figure 1(C)). Together, these data supported selection of 16G10 as the lead candidate for further evaluation against PA infection. Identification and characterization of monoclonal antibodies against‑FliC of Pseudomonas aeruginosa. (A) ELISA screening of 20 hybridoma‑derived antibodies against PA FliC; bars show mean OD450 ± SD with appropriate positive (anti‑FliC polyclonal antiserum) and negative controls (PBS). (B) Bio‑layer interferometry sensorgrams for 16G10 and 1D6 across serial dilutions demonstrating concentration‑dependent binding. The globally fitted kinetic parameters are KD = 7.46 × 10−9 M, Kon = 3.34 × 105 M−1 × s−1, Koff = 2.49 × 10−3 s−1 for 16G10 and KD = 7.65 × 10−8 M, Kon =9.08 × 104 M−1 × s−1, Koff = 6.94 × 10−3 s−1 for 1D6. (C) Representative swimming‑motility plates and quantification of swimming areas for PA treated with 16G10, 1D6, or non‑specific antibodies, comparisons among multiple groups used one-way ANOVA, we considered a difference statistically significant if P < .05. Antibacterial and antibiofilm activity of the anti‑FliC monoclonal antibody 16G10 against PA PA103 Next, we assessed the antibacterial properties of the anti‑flagellin monoclonal antibody 16G10 using the flagellated PA strain PA103. Co‑culture with 16G10 induced marked bacterial aggregation and curtailed surface spread, consistent with impaired motility, whereas no obvious aggregation was observed with PBS or an non-specific antibodies (Figure 2(A)). Then we quantified adhesion to and invasion of A549 cells and found that 16G10 significantly reduced colony‑forming units in both assays compared with PBS and isotype controls (P < .05), with a comparable reduction observed with polyclonal anti‑FliC antibodies (Figure 2(B)). For biofilm evaluation, scanning electron microscopy revealed dense cell attachment and abundant extracellular polymeric substance forming mature biofilms in the PBS and non-specific antibodies, whereas cultures exposed to 16G10 displayed a visibly diminished EPS matrix and a sparser, looser architecture despite preserved bacterial morphology (Figure 2(C)). Consistent results were obtained by confocal laser scanning microscopy with 3‑D reconstruction at days 4 and 7. PBS controls showed intense fluorescence and thick, compact biofilms, 16G10 yielded punctate fluorescence with thinner, more dispersed structures at both time points, and the azithromycin (AZM) positive control showed minimal signal and biofilm formation (Figure 2(D)). Collectively, these findings indicate that 16G10 limits PA103 motility, epithelial adhesion and invasion, and biofilm development. Functional characterization of 16G10 against Pseudomonas aeruginosa PA103. (A) Representative bright‑field micrographs (200×) of PA103 after incubation with PBS, non-specific antibodies, or 16G10, illustrating bacterial distribution; scale bars as indicated. (B) Quantification of PA103 adhesion to and invasion of A549 cells following treatment with PBS, non-specific antibodies, polyclonal anti‑FliC antiserum, or 16G10. Results expressed as CFU/mL with statistical comparisons (Comparisons among multiple groups used one-way ANOVA, P values shown on the graphs). (C) Scanning electron microscopy of biofilms formed under the indicated conditions. PBS and non-specific antibodies groups show dense networks with abundant EPS, while samples treated with 16G10 display reduced matrix and a more open structure; scale bars, 1 μm. (D) Confocal laser scanning microscopy with 3‑D reconstructions of biofilms at 4 and 7 d for 16G10, azithromycin (AZM), and PBS groups; images depict overall biomass and thickness; scale bars, 20 μm. Pre-treatment of bacteria with 16G10 attenuates virulence in vivo To determine in vivo efficacy, PA103 was preincubated with the anti‑flagellin monoclonal antibody 16G10 for 1 hour at 37°C. Then, the pre-incubated bacteria were used to challenge mice, after which survival was monitored and, at 24 hours, lungs were collected for bacterial burden, cytokine measurements, and histopathology (Figure 3(A)). Treatment with 16G10 significantly improved survival compared with PBS and non-specific antibodies, with protection comparable to imipenem (log‑rank, P < .05; Figure 3(B)). Lung bacterial loads were reduced by roughly three orders of magnitude in the 16G10 group relative to PBS and non-specific antibodies (P < .05). Pro‑inflammatory cytokines were concurrently decreased. The IL‑6 levels were lower than in PBS and non-specific antibodies (P < .05), and TNF‑α was similarly reduced (P < .05; Figure 3(D–E)). Hematoxylin and eosin staining showed attenuated inflammatory infiltration and better preserved alveolar architecture in 16G10‑treated lungs, with significantly lower inflammation scores than PBS and non-specific antibodies (P < .05; Figure 3(F)). In vivo prophylactic protective activity of 16G10 in a murine Pseudomonas aeruginosa pneumonia model. (A) Experimental scheme: PA 103 was pre-incubated with 16G10 before mouse challenge. Survival was recorded over time, and at 24 hours lungs were harvested for bacterial counts, cytokine assays, and histology. (B) Kaplan–Meier survival curves for PBS, 16G10, imipenem, and non‑specific antibody groups. (C) Lung bacterial burden at 24 hours expressed as log10 CFU per mouse with P values indicated. (D–E) The concentrations of IL‑6 and TNF‑α in lung homogenates across treatment groups. (F) Representative H&E‑stained lung sections (20×; scale bars, 50 μm) and quantified inflammation scores. Comparisons among multiple groups used one-way ANOVA. Survival data was analyzed with Kaplan–Meier analysis. We considered a difference statistically significant if P < .05. Post‑exposure therapeutic efficacy of 16G10 in a murine PA pneumonia model To test post‑exposure protection in vivo, mice were intratracheally infected with PA 103 for 1 hour and then intratracheally administered with 16G10 (Figure 4(A)). Relative to phosphate‑buffered saline (PBS), non-specific antibodies, and the antibiotic control imipenem, 16G10 significantly improved survival (log‑rank, P < .05; Figure 4(B)) and mitigated infection‑associated weight loss over 7 d (versus PBS, P < .05; Figure 4(C)). At 24 hours, lung bacterial burden was markedly reduced in the 16G10 group compared with PBS (P < .05) and non-specific antibodies (P < .05; Figure 4(D)). As expected, pro‑inflammatory cytokines in lung homogenates were also diminished. The IL‑6 concentrations were lower than in PBS (P < .05) and non-specific antibodies (P < .05; Figure 4(E)), and TNF‑α levels were reduced relative to PBS (P < .05) and non-specific antibodies (P < .05; Figure 4(F)). Histopathological examination with hematoxylin and eosin staining demonstrated fewer inflammatory infiltrates and better preserved alveolar architecture in 16G10‑treated lungs, with significantly lower inflammation scores than PBS (P < .05) and non-specific antibodies (P < .05; Figure 4(G)). Post‑exposure therapeutic efficacy of 16G10 in a murine Pseudomonas aeruginosa pneumonia model. (A) Study design: mice were infected intratracheally with PA 103 and treated 1 hour later with 16G10 via intratracheal instillation. Outcomes included survival/weight monitoring, lung colony counts, cytokine assays, and histology at 24 hours. (B) Kaplan–Meier survival curves for PBS, imipenem, nonspecific antibodies, and 16G10 groups. (C) Percent change in body weight over 7 d post‑infection. (D) Lung bacterial burden at 24 hours expressed as log10 CFU per mouse. (E–F) Concentrations of IL‑6 and TNF‑α in lung tissue across groups. (G) Representative H&E‑stained lung sections (20×; scale bars, 50 μm) with corresponding inflammation scores. Comparisons among multiple groups used one-way ANOVA. Survival data was analyzed with Kaplan–Meier analysis. We considered a difference statistically significant if P < .05. Epitope mapping and cross‑serotype efficacy of 16G10 A total of 908 PA isolates with full-length FliC sequences were retrieved from NCBI. Multiple sequence alignment of the P3 epitope was performed using OmicStudio tools (https://www.omicstudio.cn/tool), and the results are presented as sequence logos. To map the 16G10 epitope, we screened an overlapping peptide microarray and validated candidates by dot‑blot analysis; densitometric quantification identified a single dominant linear epitope (P3) within FliC corresponding to residues Gly71–Ile88 (Figure 5(A)). Structural modeling of the 16G10 Fab in complex with FliC, integrated with sequence annotation, located this epitope on a surface‑exposed region and highlighted key contacts mediated by the light‑chain complementarity‑determining regions CDR1–CDR3 (Figure 5(B)). Sequence logo analysis indicated that critical epitope residues are highly conserved across strains (Figure 5(C)). On the basis of this conservation, we hypothesized that 16G10 would inhibit flagellar function in both A‑ and B‑type flagellins. Accordingly, we assessed swimming motility in four additional PA strains (XN‑22, XN‑361, PAO1, and XN‑116) and observed that 16G10 markedly reduced swimming areas relative to isotype IgG controls (Figure 5(D)). We then established a murine infection model using the B‑type flagellin strain XN‑1 and found that 16G10 conferred significant protection when mixed with bacteria prior to challenge (pre‑incubation) and when administered after infection (therapeutic dosing), with survival superior to PBS and isotype controls and comparable to imipenem (Figure 5(E–F)). These results suggest that the 16G10 recognized a conserved epitope and conferred cross‑serotype protection. Epitope mapping and cross‑serotype efficacy of 16G10. (A) Peptide microarray and confirmatory dot‑blot identifying the FliC P3 segment (Gly71–Ile88) as the 16G10 epitope; bars show grayscale quantification of peptide reactivity. (B) Structural model of FliC bound to the 16G10 Fab with the mapped epitope highlighted. Schematic of the FliC domain map and annotated sequences of the antibody light‑chain CDRs. (C) Sequence logo illustrating conservation of residues Gly71–Ile88 across FliC variants. (D) Swimming‑motility assays in clinical serotype a PA (XN‑22, XN‑361) and serotype B (PAO1, XN‑1) strains. Representative plates and quantification of swimming area. (E) Kaplan–Meier survival after pre‑incubating XN‑1 with 16G10 before mouse challenge, alongside PBS, non-specific antibodies, and imipenem controls. (F) Kaplan–Meier survival in a therapeutic model where 16G10 was administered post‑infection, with PBS, non-specific antibodies, and imipenem comparators. Computational modeling of antibody–antigen interaction The variable region sequences of 16G10 were analyzed using IMGT/V-QUEST for complementarity-determining region (CDR) annotation. High-confidence tertiary structures were generated through structural modeling in SWISS-MODEL. Molecular docking of 16G10-Fab and FliC was performed using ClusPro 2.0 with rigid-body parameters and optimized electrostatic and desolvation weighting factors tailored for antibody–antigen systems.16 The top 10 ranked complexes from ClusPro’s balanced scoring mode were evaluated for binding free energy (ΔG) and dissociation constant (KD) using PRODIGY with default interfacial contact criteria. The optimal predicted complexes were selected for visualization, which was carried out using PyMOL.17 Discussion In this study, we generated a panel of 20 monoclonal antibodies against the PA FliC and identified 16G10 as the lead clone based on its nanomolar‑range affinity and robust functional activity. In vitro, 16G10 promoted bacterial aggregation, markedly curtailed flagellum‑driven motility, and significantly reduced adhesion to, invasion of, and biofilm formation on eukaryotic cells. Epitope mapping revealed a previously unreported linear determinant within FliC, which is highly conserved across A‑ and B‑type flagellins. Consistent with this conservation, 16G10 inhibited motility of strains expressing either serotype. In acute murine pneumonia models challenged with A‑ or B‑type flagellated strains, 16G10 conferred significant protection in both prophylactic (pre‑incubation) and therapeutic (post‑infection treatment) settings. Our research indicates that 16G10 has strong potential as an antibody drug for treating or preventing PA lung infections. The protective activity of 16G10 is plausibly explained by its recognition of a linear epitope (G71–I88) located in the N‑terminal D1 domain of FliC, which is critical for both flagellar filament assembly and innate immune recognition.18 The conserved D0/D1 core of flagellin forms the inner filament and contains residues indispensable for protofilament formation; mutational and structural studies have mapped a TLR5‑binding “hot spot” within D1 that is simultaneously required for bacterial motility, placing our epitope squarely in a nexus of structural and immunostimulatory functions.19 By occupying this site, 16G10 would be expected to sterically hinder subunit packing and destabilize filament growth, thereby arresting swimming motility and promoting bacterial aggregation – phenotypes consistent with our in vitro observations. In PA flagellum‑driven motion facilitates initial adhesion to surfaces, thus neutralization of D1 function attenuates epithelial adhesion, aligning with the reduced adhesion/invasion we observed.20 Collectively, these insights support a mechanism in which 16G10 neutralizes a conserved functional domain in the N‑terminal D1 of FliC, disabling motility‑dependent virulence while dampening TLR5 engagement. One interesting finding is that 16G10 was able to suppressed the PA103 biofilm formation in vitro. Here are two possible explanations. First, by targeting a functional epitope in FliC, 16G10 diminishes flagellum‑driven motility, thereby blocking the initial surface approach and microcolony nucleation steps that depend on coordinated flagellar and twitching motility.21 Second, flagella act as surface mechanosensors that trigger c‑di‑GMP signaling upon engagement; disrupting filament function can blunt this signal, lowering intracellular c‑di‑GMP and downstream expression of biofilm matrix genes, resulting in thinner, less cohesive biofilms–an effect consistent with our SEM/CLSM readouts.22 Third, FliC can adhere to epithelial cells. Epitope masking by 16G10 would sterically hinder this interaction, reducing initial attachment and the subsequent transition to sessile growth.23 Together, three pathways provide a coherent mechanistic framework for the inhibition of PA103 biofilm development by 16G10. In this study, we found that 16G10 protected against an imipenem‑resistant PA103 challenge, which suggests that a therapeutic strategy that combines 16G10 with antibiotics is well‑suited for drug‑resistant PA. To date, the mapped epitope of lies in the D1 region that is required for filament assembly, escape by mutating this site is likely to carry a steep fitness cost. This makes 16G10 a strong candidate for anti‑virulence therapy that applies lower selective pressure than bactericidal antibiotics. Combining 16G10 with agents that penetrate or destabilize biofilms could yield additive or synergistic effects on bacterial burden and biofilm biomass while suppressing selection for resistant variants. In support, enzymatic disruption of extracellular DNA with DNase weakens the matrix and enhances antibiotic killing within biofilms.24,25 By neutralizing motility and early adhesion while antibiotics kill metabolically active cells, combination therapy addresses complementary facets of PA pathogenesis–motility, biofilm tolerance, and planktonic growth–and may extend efficacy against multidrug‑resistant strains where monotherapy is inadequate. Hybridoma technology remains a workhorse for antibody discovery, offering high specificity, stable secretion, and straightforward screening against native antigens. Although hybridomas yield murine antibodies require humanization, CDR grafting and related strategies to mitigate immunogenicity, its strengths include rapid generation of functional binders and easy access to material for detailed kinetic and functional profiling, as achieved for 16G10.26 Further investigations are needed to improve the druggability of 16G10. First, sequence‑guided humanization and developability optimization informed by AI (Artificial Intelligence) predictors and structure models can preserve affinity while improving stability and manufacturability.27 Second, nucleic‑acid-based delivery offers alternative pharmacology, such as lipid‑nanoparticle mRNA encoding antibodies has shown safe in‑human expression and could enable rapid, dose‑tunable “in vivo manufacturing” during outbreaks.28 Third, AI‑assisted design can propose Fc engineering for extended half‑life or enhanced effector functions, and optimize paratope diversity to resist escape. Together, these advances suggest a clear path to translate 16G10 into a clinically viable anti‑PA therapeutic. Anti-flagellin strategies have demonstrated significant therapeutic or diagnostic efficacy across diverse bacteria beyond PA. In non-typhoidal Salmonella, specific monoclonal antibodies (mAbs) promote phagocytosis and, when combined, mediate complement-dependent bactericidal activity. Furthermore, passive transfer of a flagellin-specific IgG protects mice from lethal Salmonella infection.29 Similarly, against Campylobacter jejuni, an engineered single-domain antibody resist gastrointestinal proteases while potently inhibiting motility, highlighting their potential as oral therapeutics.30 Anti-flagellin mAbs also improve rapid detection via immunomagnetic separation for Campylobacter jejuni in clinical specimens31 and Vibrio vulnificus in food matrices.32 Collectively, these diverse applications confirm that targeting flagellin is a highly effective, broad-spectrum strategy against multiple flagellated pathogens. At the current stage, here are limitations of this study that merit future work. First, most experiments used acute murine pneumonia models that only partially recapitulate human airway disease, where chronic biofilms, mucus, translation from mice to patients is notoriously imperfect. Future studies should incorporate chronic infection models in multiple independent models. Second, although we tested several strains, PA displays extensive genomic and phenotypic heterogeneity across clinical isolates. A broader panel of recent clinical isolates, which spans A or B flagellins, mucoid status, and resistance genotypes, should be profiled to define coverage and identify predictors of response. Third, the 16G10 is murine and has not undergone humanization. We plan AI‑aided humanization with early developability screening and computational liability mapping to support clinical translation. In conclusion, we discovered and characterized 16G10, an anti‑FliC monoclonal antibody that recognizes a conserved D1 epitope (G71–I88). 16G10 blocks flagellar motility and diminishes adhesion, invasion, and biofilm formation in vitro. In mouse pneumonia models, it reduces lung burden, inflammation, and mortality, protecting against A‑ and B‑type–and imipenem‑resistant–strains. These findings establish 16G10 as a cross‑serotype anti‑virulence candidate and map a druggable flagellin epitope for therapeutic and combination strategies. Supplementary Material Supplemental Material Acknowledgments We thank Rong Hu from Chongqing Institute for Food and Drug Control for assistance with BLI. Biographies Jiang Gu is the Deputy Director of the National Engineering Research Center of Immunological Products at the Army Medical University. He serves as the Chief Scientist for a national key research and development program, focusing his primary research on the fundamental and clinical applications of structural vaccinology, particularly in immune interventions against multidrug-resistant bacteria for more than 10 years. Professor Gu has been awarded the First Grade Prize for Technological Invention in Chongqing and the Second Grade Prize for Military Science and Technology Progress. Notably, he has spearheaded the development and commercialization of a Class 1.1 innovative new drug, a genetically engineered Pseudomonas aeruginosa vaccine, achieving a technology transfer agreement valued at 2.8 billion yuan. Quanming Zou is a Professor at the Army Medical University and Director of the National Engineering Research Center for Immunobiological Products. His research primarily centers on the development of vaccines against multidrug-resistant bacteria. His honors include the National Second-Class Award for Technological Invention, the Guanghua Engineering Science and Technology Award (the highest honor in China’s engineering field), the FirstClass Natural Science Award of Chongqing, and the First-Class Technological Invention Award of Chongqing. Jianxin Jiang is a member of the Chinese Academy of Engineering, Director of the National Key Laboratory of Trauma and Chemical Poisoning. He is mainly engaged in the research on the treatment of high-explosive weapon injuries and traumatic infections. He has been awarded 4 Second-Class National Prizes for Science and Technology Progress, the Ho Leung Ho Lee Prize for Science and Technology Progress, the Wu Jieping Medical Innovation Award, and the Chongqing Distinguished Talents Award. He has successively served as President of the East Asian Region of the International Traffic Medicine Association, and Secretary-General of the Asian Trauma Association. Funding Statement This work was supported by the National Natural Science Foundation of China [32200771], Chongqing Natural Science Foundation General Program [CSTB2024NSCQ-MSX0901], Postdoctoral Fellowship Program of CPSF [GZB20230998], and Chongqing Postdoctoral Innovation Talent Support Program [CQBX202403]. Disclosure statement No potential conflict of interest was reported by the author(s). Supplementary material Supplemental data for this article can be accessed online at https://doi.org/10.1080/21645515.2026.2689214 References Associated Data Supplementary Materials Supplemental Material