Study approval and ethics statements

The human research protocol received ethical approval from the Biomedical Research Ethics Committee of the University of South China (Approval No. 2024077 and 2024107), with written informed consent obtained from all participants. All animal experiments were conducted in strict compliance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were reviewed by the Biomedical Research Ethics Committee of the University of South China (Approval No. 2024-491).

Study population and sample collection

This study enrolled 15 non-pneumonia patients requiring bronchoscopy (Control), 15 patients with non-PA pneumonia (Non-PA-pneumonia), and 15 patients with PA-induced pneumonia (PA-pneumonia) at the Second Affiliated Hospital, University of South China, between March 2023 and March 2024. Patients who are pregnant or breastfeeding, have severe underlying medical conditions, are in an immunosuppressed state, are currently receiving antibiotic therapy, or have chronic respiratory failure are excluded. Based on biological attributes, each cohort includes eight males and seven females. All patient characteristics were obtained at the time of initial diagnosis (Supplementary Table 1). During standard diagnostic bronchoscopy, sterile saline was instilled into the right middle lobe or left upper lobe lingula via the bronchoscope channel, achieving a fluid recovery rate of 60–70%. Polymorphonuclear neutrophils (PMNs) were isolated from bronchoalveolar lavage fluid (BALF) and peripheral blood samples using Percoll density gradient centrifugation. Seven control subjects, eight non-PA pneumonia patients, and six pneumonia patients were randomly selected for BALF collection. The baseline information of these subjects is shown in Supplementary Table 2. BALF-derived neutrophils were processed for RNA sequencing (RNA-seq), while peripheral blood neutrophils were utilized for functional in vitro assays. Human materials for organoid generation originated from adjacent non-cancerous lung tissues of three lung adenocarcinoma patients (Supplementary Table 3).

Cell culture and treatment

Isolated neutrophils were cultured in RPMI 1640 at 37 °C with 5% CO₂. Cell viability, confirmed by Ly6G/CD11b dual-labeling flow cytometry (A00-1-1102, Beckman Coulter, Fullerton, CA, USA), exceeded 95%. Neutrophils were pretreated with escalating doses of the copper ionophore elesclomol (ES) or dimethyl sulfoxide (DMSO; vehicle control) and challenged with PA strain PAO1 (ATCC #27853) at a multiplicity of infection (MOI) of 10 for 1 h⁶². For genetic manipulation, neutrophils were transducted with lentiviral-based short hairpin RNA (shRNA) knocking down TDRD9 (sh-TDRD9; 5′-TGGCAATAAGTCTCATGTATT-3′, LV-HO153046-sh, HonorGene), lentiviral TDRD9 overexpression construct (oe-TDRD9; LV-HO153046, HonorGene), lentiviral PD-L1 overexpression construct (oe-PD-L1; LV-HO014143, HonorGene), lentiviral-based shRNA knocking down CD80 (sh-CD80; 5′-CCTTAATCTCAGTAAATGGAA-3′, LV-HO005191-sh, HonorGene), and their corresponding negative control vectors (NC).

Prior to bacterial challenge, cells were incubated with 10 nM U-46199 (p38 MAPK agonist) for 24 h⁶³. To delineate TDRD9’s role, neutrophils subjected to sh-TDRD9 were exposed to 20 μM tetrathiomolybdate (cuproptosis inhibitor; 2 h), 50 μM deferoxamine (DFO; iron chelator; 1 h), 10 μM ferrostatin-1 (ferroptosis inhibitor; 1 h), 5 μM Z-VAD-FMK (pan-caspase inhibitor; 1 h), 1 μM necrosulfonamide (necroptosis inhibitor; 1 h), and 500 μM 3-methyladenine (3-MA; autophagy inhibitor; 1 h).

Cell counting kit-8 (CCK-8)

Cell viability was quantified using the CCK-8 assay (Dojindo Laboratories, Tokyo, Japan). Briefly, 100 μL of neutrophil suspension (5.5 × 10³ cells/well) was seeded into 96-well plates. Following a 4-h incubation (37 °C, 5% CO₂), 100 μL of complete medium containing 10% CCK-8 reagent was added to each well. Absorbance at 450 nm (OD₄₅₀) was measured using a Bio-Tek Microplate Reader (HEALES, Shenzhen, China).

RNA sequencing (RNA-seq)

RNA sequencing and bioinformatic analyses were conducted by Shanghai OE Biotech Co., Ltd. (Shanghai, China). Neutrophils isolated from human subjects were subjected to RNA-seq. Sequencing reads were aligned to the human reference genome using HISAT2 (v2.1.0). Read quantification was performed with HTSeq-count (v0.11.2). Differential gene expression analysis and visualization were carried out using DESeq2 (v1.22.2) with significance thresholds set at |log2(fold change)| >1 and an adjusted p-value 

Preparation and characterization of hyaluronic acid-coated p5RHH-si-TDRD9 NPs

Hyaluronic acid (HA)-coated and uncoated p5RHH-si-TDRD9 NPs were synthesized⁶⁴. Briefly, small interfering RNA targeting TDRD9 (si-TDRD9; 10 μM) or its negative control (si-NC) was mixed with 10 mM p5RHH peptide (VLTTGLPALISWIRRRHRRHC; Cat# SC1848, GenScript, Piscataway, NJ, USA) to facilitate self-assembly of p5RHH-siRNA NPs. HA-coated NPs (HA-si-TDRD9 NPs and HA-si-NC NPs) were prepared by incubating HA solution (10 mg; Cat# B8382, ApexBio, Houston, TX, USA) with pre-assembled p5RHH-siRNA NPs. Uncoated NPs (si-TDRD9 NPs and si-NC NPs) were generated by omitting HA. Nanoparticle size and morphology were assessed by transmission electron microscopy (TEM). Hydrodynamic diameter was determined via dynamic light scattering (DLS; Zetasizer Nano ZS90, UK). si-TDRD9 (sense: 5′-UGGCAAUAAGUCUCAUGUATT-3′, antisense: 5′-UACAUGAGACUUAUUGCCATT-3′) and si-NC (sense: 5′-UUCUCCGAACGUGUCACGUTT-3′, antisense: 5′-ACGUGACACGUUCGGAGAATT-3′) were obtained from HonorGene.

Uptake of p5RHH-siRNA NPs by neutrophils

Neutrophils were seeded into 12-well plates at a density of 5 × 10⁵ cells per well. After overnight adherence, cells were incubated with Cy5.5-labeled HA-si-TDRD9 NPs or naked si-TDRD9 NPs (HA omitted) for 4 h. Nanoparticle uptake was assessed by flow cytometry and immunofluorescence analysis of Cy5.5 and LY6G staining. To evaluate HA-mediated uptake, neutrophils were stained with anti-CD44 antibody (Cat# MU310, BioGenex, Fremont, CA, USA), followed by incubation with HA-si-TDRD9 NPs.

Animal model of pneumonia

A total of 286 male C57BL/6 mice (6–8 weeks old) were obtained from Hunan Slake Jingda Animal Experiment Company (Changsha, China) and housed at 22–26 °C with 40–60% humidity under a 12-h light/dark cycle. After mice were anesthetized via intraperitoneal injection of sodium pentobarbital (50 mg/kg), 50 μL of bacterial suspension of PAO1 (OD₆₀₀ = 0.6; 7 × 10⁶ CFU/mL in 0.9% sterile saline) was administered via oropharyngeal intratracheal instillation to simulate PA pneumonia⁶². Control mice received an equal volume of sterile saline.

Seventy-eight mice were randomly allocated to Control and PA groups (n = 39 per group). Three PA-group mice died during the seven-day infection period. On day 1 post-infection, twelve mice per group were randomly selected as Control-1d and PA-1d subgroups (n = 12 each). On day 3, twelve additional mice per group were randomly chosen from the same cohort as Control-3d and PA-3d subgroups (n = 12 each). The remaining mice were maintained until day 7, with 12 surviving PA-group mice designated as PA-7d and twelve randomly selected Control-group mice as Control-7d (n = 12 each). Euthanasia was performed via sodium pentobarbital overdose (100 mg/kg). Six mice per subgroup underwent pulmonary edema evaluation. From the remaining six mice per subgroup, BALF was collected through lung lavage with 1 mL ice-cold phosphate-buffered saline (PBS), while peripheral blood and lung tissues were harvested for downstream analyses.

To investigate the in vivo role of TDRD9 in neutrophils, seventy-eight mice were randomly allocated to six experimental groups: Control, Anti-Ly6G, PA, Anti-Ly6G+PA, Anti-Ly6G+PA+PMN^sh-NC, and Anti-Ly6G+PA+PMN^sh-TDRD9 (n = 12 per group). Mice in anti-Ly6G-treated groups were intravenously administered 50 μg anti-Ly6G antibody 48 h prior to PAO1 infection to deplete endogenous neutrophils⁶⁵. Neutrophils (1 × 10⁶ cells/mouse) transducted with sh-TDRD9 (5′- AGCTCAGAAATGGAGTATATT-3′, LV-MO029056-sh, HonorGene) or sh-NC were adoptively transferred via intravenous injection⁶⁶. Control mice received equivalent volumes of sterile saline. On day 3, BALF and lung tissues were collected for further evaluation.

In vitro and in vivo targeting efficacy of HA-p5RHH-siRNA NPs

The regulatory effects of HA-si-TDRD9 NPs on the PD-L1/CD80/MAPK signaling axis and cuproptosis were assessed. Human peripheral blood neutrophils were incubated with 0.1 μM HA-si-NC NPs or HA-si-TDRD9 NPs for 4 h⁶⁴. Neutrophils were subsequently treated with 10 μM ES for 24 h, followed by infection with PAO1 for 1 h. For PD-L1 overexpression, neutrophils were transfected with oe-PD-L1 or oe-NC. Prior to bacterial challenge, cells were pre-treated with 1 μM SB203580 (p38 MAPK inhibitor) for 1 h^56,67.

The therapeutic efficacy of HA-si-TDRD9 NPs was evaluated in mice. Sixty-five mice were randomly allocated to five experimental groups: Control, Post-PA, Post-PA + HA-si-NC NPs, Post-PA + HA-si-TDRD9 NPs-low, and Post-PA + HA-si-TDRD9 NPs-high (n = 12 per group). Mice received intravenous injections of low-dose (0.1 μmol) or high-dose (0.2 μmol) HA-si-TDRD9 nanoparticles three days post-infection. Body weight was monitored daily for six consecutive days post-infection, with BALF and lung tissue samples collected on day 3 post-treatment.

To evaluate the preventive protective effects of HA-si-TDRD9 NPs, 65 mice were randomly allocated to five experimental groups: Control, Pre-PA, Pre-PA + HA-si-NC NPs, Pre-PA + HA-si-TDRD9 NPs-low, and Pre-PA + HA-si-TDRD9 NPs-high (n = 12 per group). Mice received intravenous injections of HA-si-TDRD9 NPs at low (0.1 μmol) or high (0.2 μmol) doses⁶⁴ 24 h pre-infection, with subsequent doses administered every 48 h. Body weight changes were monitored for 3 days post-infection. BALF and lung tissues were harvested on day 3.

Organoid culture

Fresh adjacent non-cancerous lung tissue samples from patients were collected and preserved in tissue preservation solution (Cat# OGCP-04-06, Guangzhou Wogen Biotechnology Co., Ltd., Guangzhou, China) at 4 °C. The tissues were minced into 1–2 mm³ fragments using surgical scissors in PBS containing 1% penicillin–streptomycin. Fragments were digested with organoid digestion solution (Cat# OGCP-04-01) for 1–2 h, filtered through a 100 µm cell strainer, and pelleted by centrifugation. The pellet was resuspended in Matrigel (Cat# OGCP-04-11) mixed 1:1 with organoid culture medium (Cat# OGCP-02-07). A 50 μL aliquot of the mixture was plated and solidified at 37 °C under 5% CO₂, followed by the addition of 500 μL organoid culture medium. Organoid growth was monitored daily for 7 days, with medium replaced every 2–3 days.

Bacterial growth assay

For bacterial inoculation, organoid culture medium, cells, and lung tissue homogenate were combined with sterile PBS and spiked with PA at an initial concentration of ~3 × 10² CFU/mL. A 100 μL aliquot of the mixture was seeded into LB medium and incubated at 37 °C with 5% CO₂ for 6, 12, 24, and 48 h. Bacterial colonies were visualized, and colony-forming units (CFU) were quantified at each timepoint.

Lung water content

Lung edema was quantified using the wet/dry weight ratio. Immediately after euthanasia, the lungs were exercised and weighed to determine wet weight. Tissues were then dehydrated at 60 °C for 72 h in a laboratory oven until constant dry weight was achieved. The ratio of wet weight to dry weight was calculated to assess pulmonary edema severity.

Hematoxylin and eosin (HE) staining

Fresh lung, hepatic, and renal tissues from all animal groups and HLOs were fixed in 4% PFA for 24 h at 4 °C, paraffin-embedded, and sectioned at 4 μm thickness. Sections were stained with hematoxylin (Cat# AWI0001a, Abiowell) and eosin (Cat# AWI0029a, Abiowell). Histopathological damage was evaluated under a BA210T bright-field microscope (Motic, Xiamen, China) at 20×, 100×, and 400× magnification. According to established protocols⁶⁸, pneumonia severity was quantified by assessing four parameters: (1) alveolar congestion, (2) hemorrhage, (3) leukocyte infiltration, and (4) alveolar wall thickness/hyaline membrane formation (0 = absent; 3 = severe). The total score ranges from 0 to 12 points.

Flow cytometry

Lung tissues were dissociated by mincing followed by enzymatic digestion in RPMI-1640 medium containing collagenase IV and DNase I at 37 °C for 30 min. The homogenate was filtered through a 70 μm cell strainer and centrifuged at 300 × g for 5 min. Red blood cells were lysed using ACK buffer, and single-cell suspensions were prepared in PBS supplemented with 2% FBS. For immunophenotyping, cells were stained for 30 min at 4 °C with the following fluorochrome-conjugated antibodies: Ly6G-PC7 (1A8-Ly6g) (Cat# 25-9668-82, eBioscience, San Diego, CA, USA), F4/80-PC5.5 (BM8) (Cat# 45-4801-82, eBioscience), CD19-APC (eBio1D3 (1D3)) (Cat# 17-0193-82, eBioscience), CD3 (145-2C11) (Cat# 12-0031-82, eBioscience), CD49b (DX5) (Cat# 11-5971-82, eBioscience). Data acquisition was performed on a CytoFLEX flow cytometer (A00-1-1102, Beckman Coulter). Neutrophils (Ly6G+), macrophages (F4/80+), B cells (CD19+), NK cells (CD49b+), and T cells (CD3+) were quantified using CytExpert v2.5 software (Beckman Coulter).

For intracellular cytokine staining, single-cell suspensions were stimulated with Cell Stimulation Cocktail (Cat# 00-4975-93, eBioscience) for 4 h at 37 °C under 5% CO₂. Cells were surface-stained with CD4 (GK1.5) (Cat# 11-0041-82, eBioscience) and intracellularly stained with: IL-17A (eBio17B7) (Cat# 12-7177-81, eBioscience), CD25 (PC61.5) (Cat# 17-0251-82, eBioscience), and Foxp3 (FJK-16s) (Cat# 12-5773-82, eBioscience). Th17 (CD4 + IL-17A+) and Treg (CD4+CD25+Foxp3+) populations were gated using CytExpert v2.5 software, and the Th17/Treg ratio was calculated.

Enzyme-linked immunosorbent assay (ELISA)

Proinflammatory cytokines (IL-6, IL-1β, TNF-α) in BALF and HLOs were quantified using commercial mouse ELISA kits (IL-6: Cat# CSB-E04639m; IL-1β: Cat# CSB-E08054m; TNF-α: Cat# CSB-E04741m; CUSABIO Co., Ltd., Wuhan, China) and human ELISA kits (IL-6: Cat# KE00139; IL-1β: Cat# KE00021; TNF-α: Cat# KE00154; Proteintech). BALF and HLO samples were centrifuged to remove cellular debris. Soluble cluster of differentiation 80 (sCD80), p53, and programmed death ligand-1(PD-L1) in cell culture medium were measured using commercial human ELISA kits (sCD80: CSB-E15768h-IS; p53: Cat# CSB-E08334h-IS; PD-L1: CSB-E13644h; CUSABIO Co., Ltd.). Assays were performed according to the manufacturer’s protocol, with absorbance measured at 450 nm using a Bio-Tek microplate reader.

Quantitative real-time PCR (qRT-PCR)

Total RNA was isolated from neutrophils, lung tissues, and HLOs using TRIzol reagent (Cat# 15596026, Thermo Fisher Scientific). RNA purity and concentration were determined using a Micro Drop spectrophotometer (Thermo Fisher Scientific). Complementary DNA (cDNA) synthesis was performed with 2 μg total RNA using the HiFiScript cDNA Synthesis Kit (Cat# CW2569M, CWBIO, Taizhou, China), following the manufacturer’s protocol. qRT-PCR amplification was conducted in a QuantStudio 1 Real-Time PCR System (Thermo Fisher Scientific) using UltraSYBR Mixture (Cat# CW2601M, CWBIO). Gene-specific primers (Supplementary Table 4) were designed using NCBI Primer-BLAST. Relative mRNA expression levels were calculated using the 2^−ΔΔCt method, normalized to the endogenous control β-actin.

Western blot analysis

Total protein was extracted from neutrophils, lung tissues, and HLOs using RIPA lysis buffer (Cat# R0010, Solarbio, Beijing, China). Equal amounts of protein were resolved on 10% SDS-polyacrylamide gels and transferred to nitrocellulose membranes. After blocking with 5% non-fat dry milk, membranes were incubated with primary antibodies overnight at 4 °C and with horseradish peroxidase (HRP)-conjugated secondary antibodies for 1 h at room temperature (Supplementary Table 5). Immunoreactive bands were imaged on a ChemiScope 6100 imaging system (CLiNX Science Instruments, Shanghai, China) and quantified using ImageJ v1.51 software (NIH, Bethesda, MD, USA) normalized to β-actin expression.

Immunofluorescence (IF) and immunohistochemistry (IHC)

For IF, lung tissue sections were subjected to microwave-mediated antigen retrieval in Tris-EDTA buffer. Autofluorescence was quenched by incubating sections with 0.1% sodium borohydride, followed by endogenous peroxidase blockade using 0.3% hydrogen peroxide. Sections were blocked with 5% BSA and incubated with anti-LY6G (1A8) (1:200; Cat# 65078-1-Ig, Proteintech, Rosemont, IL, USA), DLAT polyclonal (1:100; Cat# 13426-1-AP, Proteintech), and FDX1 polyclonal (1:100; Cat# 12592-1-AP, Proteintech) primary antibodies and HRP-conjugated secondary antibody (1:200; Cat# AWI0629, Abiowell). Tyramide signal amplification (TSA) was performed using 520 and 570 fluorophores (Abiowell). Nuclei were counterstained with 4’,6-diamidino-2-phenylindole (DAPI). Fluorescent images were acquired using a Motic BA410 fluorescence microscope.

For IHC, dewaxed HLO sections underwent heat antigen retrieval followed by endogenous peroxidase blockade with 1% periodic acid. Tissue sections were incubated with primary antibodies against Ki67 (45b13) (1:200; Cat# AWA11025, Abiowell), NKX2.1 (EP1584Y) (1:200; Cat# ab76013, Abcam), and SOX9 (45A09) (1:200; Cat# AWA10351, Abiowell) at 4 °C overnight and species-matched HRP-conjugated secondary antibodies at 37 °C for 1 h. Sections were stained with 3,3’-diaminobenzidine (DAB) and hematoxylin. Images were captured using a bright-field microscope and analyzed using Image-Pro Plus software v6.0 (Media Cybernetics, Rockville, MD, USA).

TUNEL assay

Paraffin-embedded HLOs were dehydrated and treated with 100 μL of 1 × Proteinase K working solution. Subsequently, 100 μL of 1 × Equilibration Buffer was added and incubated at room temperature for 10–30 min, followed by the addition of 50 μL TdT incubation buffer. The samples were protected from light and incubated for 60 min. After washing with PBS, nuclei were counterstained with DAPI. Fluorescence microscopy was performed to capture images, and the percentage of TUNEL-positive cells was quantified.

Co-immunoprecipitation (Co-IP) assay

Total cellular proteins were extracted using ice-cold IP lysis buffer (Abiowell) supplemented with protease inhibitor cocktail. For immunoprecipitation, lysate was incubated overnight at 4 °C with Rabbit anti-PD-L1 (Cat# 28076-1-AP, Proteintech) or Rabbit IgG isotype control (Cat# B900610, Proteintech), followed by 4-h incubation with Protein A/G agarose beads. Beads were washed with lysis buffer, and bound proteins were eluted in IP lysis buffer. Eluates were resolved by SDS-PAGE and immunoblotted using Rabbit anti-PD-L1 and Rabbit anti-CD80 antibodies.

Detection of reactive oxygen species (ROS) and Fe²⁺ levels

Intracellular ROS levels were quantified using a CytoFLEX flow cytometer with the DCFH-DA Cellular ROS Assay Kit (Cat# S0033S, Beyotime, Shanghai, China). Fe²⁺ concentrations (μmol/10⁶) were determined using the Cell Ferrous Iron Colorimetric Assay Kit (Cat# E-BC-K881-M, Elabscience, Wuhan, China).

Statistical analysis

Data are presented as mean ± standard deviation (SD). Statistical comparisons were performed in GraphPad Prism v9.0 (GraphPad Software, San Diego, CA, USA). Two-group comparisons were performed using unpaired two-sided t-test. Multi-group comparisons were performed using two-sided ANOVA with Tukey’s post hoc test. Pearson’s coefficients for TDRD9 and cuproptosis-related protein expression were analyzed. A p-value 

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.