GSK’872 mitigates ischemia-reperfusion injury in rat lung transplants by regulating PANoptosis and inflammation in cell type specific manner

Abstract

Background

Ischemia-reperfusion injury (IRI) greatly impairs lung transplantation (LTx) outcomes, with no effective treatments. Although existing studies have confirmed that cell death and inflammation responses are critical in LTx-IRI, the specific cell death profiles of various parenchymal and inflammatory cells remain to be elucidated.

Methods

Using human single-cell RNA sequencing data from LTx-IRI, we identified activation of genes related to cell death and inflammation pathways. We examined the effects and mechanisms of a RIPK3 inhibitor, GSK’872, on IRI with a rat LTx model.

Results

GSK’872, added to lung preservation solution, injected to recipients, or in combination, reduced alveolar hemorrhage, perivascular edema, neutrophil infiltration and suppressed necroptosis, pyroptosis, and inflammation in lung tissues. GSK’872 decreased MLKL phosphorylation in type 2 alveolar epithelial cells and macrophages, and RIPK3 phosphorylation in neutrophils. GSK’872 induced apoptosis in RAW264.7 macrophages via RIPK1 and caspase 3 cleavage in a cold ischemia/warm reperfusion (CI/R) cell culture model. GSK’872 inhibited lipopolysaccharide (LPS)-stimulated neutrophil extracellular traps (NETs) formation with suppressed necroptosis and pyroptosis. GSK’872 did not rescue BEAS-2B lung epithelial cells from CI/R-induced cell death. However, conditioned medium from GSK’872-treated macrophages (after CI/R) or neutrophils (challenged by LPS) reduced CI/R-induced decrease in BEAS-2B cell viability.

Conclusions

GSK’872 alleviates LTx-IRI by inhibiting necroptosis and pyroptosis in macrophages and neutrophils directly, and protects lung epithelial cells via blocking soluble mediators indirectly. Administration of GSK’872 to lung preservation solution and/or injection to recipients may be new treatment options for IRI in LTx.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12964-026-03037-1.

Introduction

Lung transplantation (LTx) is the most effective treatment for patients with end-stage lung disease. Despite the rising number of LTx procedures in recent years, the median survival for first-time adult recipients remains at 6.7 years [1]. Primary graft dysfunction (PGD), characterized by pulmonary edema, thrombosis, and decreased pulmonary oxygenation [2], represents one of the most common and severe complications post-LTx, closely associated with early postoperative mortality [3, 4].

Lung ischemia-reperfusion injury (IRI) is a critical trigger for PGD [5]. However, our understanding of the mechanisms underlying IRI is limited, leading to a lack of targeted therapies for its prevention or treatment. Transcriptomic studies have revealed a significant enrichment of genes related to cell death and inflammatory responses in human LTx-IRI [6], providing potential therapeutic targets for treating IRI [7]. Among various types of cell death, pyroptosis (P), apoptosis (A), and necroptosis (N) have recently been collectively defined as PANoptosis [8]. A transcriptomics study shows that genes related to these three pathways are co-regulated during LTx-IRI [9]. Further evidence of PANoptosis may lead to effective treatment of LTx-IRI.

In PANoptosis, classical pyroptosis is mediated by the activation of the NLRP3 inflammasome [10], leading to the activation of caspase 1 or caspase 4/5/11, which subsequently cleave Gasdermin family proteins (e.g., GSDMD), causing pore formation on the cell membrane and cell death [11]. Classical apoptosis is mediated by activation of caspase 3. Classical necroptosis involves sequential phosphorylation of Receptor Interacting Protein Kinase 1 (RIPK1), RIPK3, and downstream Mixed Lineage Kinase Domain Like (MLKL). Phosphorylated MLKL oligomerizes and translocates to the plasma membrane, forming pores that lead to cell rupture and death [12]. Notably, RIPK3 not only regulates necroptosis but also mediates NLRP3 inflammasome activation [13], suggesting its role in both necroptosis and pyroptosis in PANoptosis. Additionally, RIPK3 regulates inflammatory cells independent from necroptosis: RIPK3 knockout reduces immune cell recruitment in West Nile virus-infected encephalitis mice [14] and regulates macrophage-mediated inflammation in folic acid-induced acute kidney injury in mice [15], both unrelated to necroptosis severity. Therefore, targeting RIPK3 emerges as a promising strategy for preventing and treating LTx-IRI.

In this study, using GSK’872, the most well-known small-molecule inhibitor of RIPK3, as a potential therapeutic, we investigated the relationship of the three cell death modes of PANoptosis and inflammation in a rat LTx-IRI model. The direct and indirect effects of GSK’872 on cell death in macrophages, neutrophils and lung epithelial cells were explored with cell culture and co-culture models.

Methods

Experimental animals and cell lines

Male pathogen-free Sprague-Dawley rats, 8–12 weeks of age and weighing 250–300 g, were purchased from the Zhejiang Laboratory Animal Center (Hangzhou, China). Human bronchial epithelial cells (BEAS-2B) and murine monocyte-macrophage leukemia cells (RAW264.7) were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). Primary circulating neutrophils were isolated from adult rats.

Rat orthotopic LTx model

The orthotopic rat left LTx model was used according to the methods previously described by our team [16]. The experimental protocol was approved by the Animal Care and Use Committee of Zhejiang University (Approval No. ZJU20220159).

A total of 104 rats (8–12 weeks old, 250–300 g) were randomly divided into five groups. The sham-operated rats (Sham group, n = 8) underwent endotracheal intubation with mechanical ventilation, followed by left thoracotomy and chest closure, and 1 h of postoperative mechanical ventilation, without LTx. The remaining 96 rats were randomly assigned as donors or recipients for 48 left LTxs. The grafts were subjected to cold ischemia for 18 h (CI 18 h) and followed by left LTx and 4 h reperfusion (R 4 h). The left lung grafts and recipient sera were collected at the end of reperfusion for subsequent analyses.

Administration protocols of RIPK3 inhibitor GSK’872

The 48 transplanted animals were further divided into four groups (n = 12 per group): vehicle only (without GSK’872), GSK’872 administered to both donor and recipient (Donor+Recipient, D + R), GSK’872 administered to donor lung preservation solution (Donor, D) and GSK’872 intraperitoneal injection to recipients (Recipient, R).

For D and D + R groups, 5 µM GSK’872 (HY-101872, MedChemExpress, NJ, USA) was added to the Low-Potassium Dextran (LPD) solution during donor lung flushing and cold static storage. For R and D + R groups, GSK’872 was administered via intraperitoneal injection at a dose of 0.75 mg/kg, 1 h prior to reperfusion. In the Vehicle, D, and R groups, vehicle control (20 µL DMSO in 50 mL LPD solution, or 6 mL/kg vehicle solution via intraperitoneal injection) was administered without GSK’872.

Histopathological and immunohistochemical analysis

Rat left lung tissues were harvested, fixed, paraffin-embedded, and sectioned into 4 μm-thick slices, as previously described [17]. Hematoxylin and eosin (HE) staining was performed to assess general histopathological changes. Myeloperoxidase (MPO) immunohistochemical staining was carried out according to a reported protocol [18], with a specific antibody listed in Supplementary Table S1.

Five random, non-overlapping high-power fields (HPFs, 400× magnification) were selected for image capture. The perivascular cuff area was quantified as the percentage of the cuff zone relative to the total area encompassing the vessel and its surrounding edema [19]. The numbers of red blood cells and MPO-positive cells within each field were also counted to assess the extent of pulmonary hemorrhage and neutrophil infiltration, respectively. The average value from the five fields was used to represent the corresponding parameter for each sample.

Immunofluorescence and TUNEL staining

Commercial immunofluorescence kits based on tyramide signal amplification (TSA) technology (RecordBio, Shanghai, China) were used for immunofluorescent staining of paraffin-embedded lung tissue sections, according to the manufacturer’s instructions. Multiple immunofluorescent labeling was performed for phosphorylated MLKL (pMLKL), phosphorylated RIPK3 (pRIPK3) or cleaved caspase 8 combined with the following cell-specific markers: type 2 alveolar epithelial cells (AEC2) marker surfactant protein C (SPC), endothelial cell marker CD31, type 1 alveolar epithelial cells (AEC1) marker aquaporin 5 (AQP5), monocytes/macrophages marker CD68, and neutrophils marker MPO. Detailed antibody information is provided in Supplementary Table S1.

Fluorescent images were captured under a fluorescence microscope (BX51, Olympus, Japan). For each sample, five random non-overlapping HPFs were selected from each sample. The number and proportion of fluorescently labeled cells were quantified, and the average value across the five fields was used to represent the corresponding parameter for each sample.

TUNEL staining was performed on paraffin-embedded lung sections using a kit (Cat. No. RCT-50R, RecordBio, Shanghai, China), according to the manufacturer’s protocol. TUNEL-positive cells were considered as markers of cell death, and quantified as described above.

Cold Ischemia (CI)/Warm Reperfusion (R) cell culture model for LTx

A cell culture model simulating the hypothermic preservation and warm reperfusion of LTx was established in BEAS-2B cells and RAW264.7 cells, respectively, as previously described [20, 21].

In brief, BEAS-2B cells were cultured in RPMI-1640 medium (41402ES76, Yeasen, Shanghai, China) supplemented with 10% fetal bovine serum (FBS, 10099-141, Gibco, Thermo Fisher Scientific, MA, USA). RAW264.7 macrophages were maintained in DMEM (C11995500BT, Gibco, Thermo Fisher Scientific, MA, USA) with 10% FBS. All cells were incubated at 37 °C in a humidified atmosphere containing 5% CO₂. When cells reached approximately 80–90% confluence, culture medium was replaced with 4 °C LPD solution, and cells were transferred to a 4 °C environment with 50% oxygen to simulate the CI phase. After 18 h, cells were returned to normal culture conditions for 4 h to mimic the reperfusion.

Various concentrations of GSK’872 (0, 1.25, 2.5, 5, 10, 20 µM) were used with the following three regimens: CI/R combined (Cold Ischemia 18 h + Reperfusion 4 h), donor lung preservation solution (Cold Ischemia 18 h), and medium in the reperfusion phase (Reperfusion 4 h). For the control group, the vehicle solution was applied.

Cell viability was assessed using a CCK-8 assay kit (C0038, Beyotime, Shanghai, China), and lactate dehydrogenase (LDH) level in the supernatant was measured using an LDH detection kit (C0016, Beyotime, Shanghai, China), according to the manufacturer’s instructions, to evaluate cell injury.

Lipopolysaccharide (LPS) induced rat neutrophil activation model

Five minutes after intraperitoneal injection of 1,000 IU sodium heparin, peripheral blood was collected from the abdominal vein of SD rats. Rat neutrophils were isolated using a neutrophil isolation kit (Cat. No. P9203, Solarbio, Beijing, China), according to the manufacturer’s instructions. After cell counting, the neutrophils were divided into four groups (3 million neutrophils per group): Control, GSK’872, LPS, and LPS + GSK’872. The cells were cultured in RPMI-1640 medium supplemented with 10% FBS. The experiment was repeated six times.

Neutrophils in the GSK’872-treated groups were pre-incubated with 50 µM GSK’872 for 1 h [22], followed by stimulation with 20 µg/mL lipopolysaccharide (LPS, L2630, from Escherichia coli O111:B4, Sigma Aldrich, MA, USA) for 2 h. Control groups received the corresponding vehicle treatments. Following treatment, cells were centrifuged, and the resulting cell pellets were used for western blot analysis or DAPI staining to assess neutrophil extracellular traps (NETs) formation. The supernatants were collected to quantify the release of inflammatory cytokines with an enzyme-linked immunosorbent assay (ELISA).

The NETs formation was evaluated as follows: cell pellets were resuspended and smeared onto glass slides, followed by DAPI staining. For each sample, five random HPFs were selected, and the percentage of neutrophils forming NETs among all neutrophils was calculated.

BEAS-2B cell CI/R model treated with conditioned medium (CM)

CM from RAW264.7 macrophages or rat peripheral blood neutrophils was tested in the BEAS-2B cell CI/R model. BEAS-2B cells were treated with CM collected from four experimental groups: Control group, Control + GSK’872 group, CI/R group, and CI/R + GSK’872 group. The experiment was repeated six times.

For the RAW264.7-derived CM, a CI/R model was first established in RAW264.7 macrophages by subjecting them to 18 h of CI in LPD solution at 4 °C, followed by 4 h of Reperfusion under standard culture conditions. GSK’872 (5 µM) was added to the LPD or culture medium of RAW264.7 macrophages during the CI and reperfusion periods, respectively, while vehicle controls received the corresponding solvent. The supernatants from both the CI and R phases were collected, diluted threefold with complete medium, and then applied to BEAS-2B cells in simulated CI/R conditions. Corresponding control supernatants were obtained from RAW264.7 macrophages cultured under standard conditions during the same time periods and were similarly diluted before use.

For the neutrophil-derived CM, rat peripheral blood neutrophils were isolated and pre-incubated with or without GSK’872 (50 µM) for 1 h. Following this, cells were stimulated with LPS (20 µg/mL) for 2 h. The resulting supernatants were collected, diluted threefold with complete medium, and used to treat BEAS-2B cells during the reperfusion phase. Control supernatants were obtained from neutrophils not exposed to LPS and were processed in the same manner.

Cell viability was assessed using the CCK-8 assay kit (C0038, Beyotime, Shanghai, China), and Propidium Iodide (PI)/DAPI staining was performed to evaluate the level of cell death in BEAS-2B cells and five random HPFs were selected to count for each sample.

Enzyme-Linked Immunosorbent Assay (ELISA)

The levels of HMGB1, IL-6, IL-1β, and MPO, in recipient rat serum, lung graft homogenates, and cell culture supernatants, were measured using commercial ELISA kits, according to the manufacturers’ instructions. Detailed information regarding to the ELISA kits is provided in Supplementary Table S2.

Western blotting analysis

Western blotting analysis was performed on rat lung tissue lysates and cultured cell lysates. In brief, tissues and cells were lysed on ice using NP-40 lysis buffer (Nonidet P-40, P0013F, Beyotime, Shanghai, China) supplemented with protease and phosphatase inhibitors and phenylmethanesulfonyl fluoride (PMSF, ST506, Beyotime, Shanghai, China). The protein concentration of the lysate was determined using a BCA Protein Assay Kit (P0010, Beyotime, Shanghai, China). After mixed with the loading buffer, the lysates were denatured by heating.

Equal amounts of protein (20–60 µg per lane) were separated by SDS-PAGE, following standard procedure. Detailed antibody information is provided in Supplementary Table S3. Band intensities of blots were quantified using Image Lab 6.1 software (Bio-Rad, CA, USA).

Analysis of single-cell RNA sequencing data (scRNA-seq) from human LTx

A publicly available single-cell RNA sequencing (scRNA-seq) dataset was retrieved from the Gene Expression Omnibus (GEO) database (accession number GSE220797 [23]), comprising single-cell transcriptomic profiles from six human lung transplants collected at the end of cold ischemic time and after 2 h of reperfusion post LTx.

For each annotated cell type, differentially expressed genes (DEGs) after reperfusion were identified using the FindMarkers function. Genes with an adjusted p-value < 0.05 and log₂ fold change > 0 were considered significantly upregulated. Functional enrichment analysis of these upregulated DEGs was performed using the enrichKEGG function in the clusterProfiler package to identify significantly enriched Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways.

Statistical analysis

Statistical analyses were performed using GraphPad Prism 9. Outliers were excluded using the ROUT test (Q = 1%). The Shapiro-Wilk test and Brown-Forsythe test were used to assess normality and homoscedasticity of the data, respectively. For data that met the assumptions of normality and homoscedasticity, they were described as mean ± standard deviation (SD), and statistical comparisons were conducted using one-way analysis of variance (ANOVA) or two-way ANOVA as appropriate. For data that did not meet these assumptions, they were presented as median ± interquartile range (IQR), and non-parametric tests (Kruskal-Wallis test) were applied for statistical comparisons. A P value < 0.05 is considered statistically significant.

Results

Cell death and inflammation-related genes were enriched in lung transplants during reperfusion

We analyzed single-cell sequencing data from 6 pairs of human LTx samples [23]. KEGG enrichment analysis shows multiple cell death- and inflammation-related pathways are activated in type 1 alveolar epithelial cells (AEC1, Fig. 1A), type 2 alveolar epithelial cells (AEC2, Fig. 1B), ciliated cells (Fig. 1C), alveolar macrophages (Fig. 1D), monocytes (Fig. 1E) and neutrophils (Fig. 1F), after 2 h reperfusion. This suggests that after LTx, both cell death and inflammation play important roles in the process, whether in lung parenchymal cells or inflammatory cells.

Fig. 1.

Cell death and inflammation played key roles during human lung transplantation. Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment of significantly upregulated genes in type 1 alveolar epithelial cells (AEC1) (A), type 2 alveolar epithelial cells (AEC2) (B), ciliated cells (C), alveolar macrophages (D), monocytes (E) and neutrophils (F) at 2 h post-transplantation, derived from human LTx single-cell RNA sequencing (scRNA-seq) data (GSE220797). All pathways shown have an adjusted p-value < 0.05

GSK’872 reduced lung IRI in rat LTx

We selected GSK’872, a classical RIPK3 inhibitor that can inhibit both cell death and inflammation, in this study. Five experimental groups were used to evaluate the effects of GSK’872 on IRI in rat LTx: Sham (surgery without ischemia-reperfusion), Vehicle (18 h ischemia + 4 h reperfusion, an IRI control), and three GSK’872-administered groups (donor-recipient combined [D + R], added to donor lung preservation solution [D, 5 µM], recipient intraperitoneal injection [R, 0.75 mg/kg]) (Fig. 2A).

Fig. 2.

GSK’872 alleviated acute lung injury in rat lung transplants. (A) Experimental design. Rats were divided into five groups: Sham (surgery without ischemia-reperfusion, n = 8), Vehicle (LTx without treatment, n = 12), and three treatment groups (LTx with GSK’872 in the donor lung preservation solution and recipient injection [D + R, n = 12], donor lung with 5 µM GSK’872 in preservation solution [D, n = 12], and recipient intraperitoneally injected with 0.75 mg/kg GSK’872 [R, n = 12]). (B) Representative images of lung grafts, H&E staining, and myeloperoxidase (MPO) immunohistochemical staining of rat lung tissues. (C) Quantification of red blood cell counts in alveolar space per sample. (D) Quantification of perivascular edema level per sample. (E) Quantification of MPO-positive cell counts per sample. (F) IL-6 and (G) IL-1β levels in the serum were detected using ELISA. Data were represented as means ± SD in C-D and median ± IQR in E-G. *p < 0.05, **p < 0.01, ***p < 0.001, compared with Sham; #p < 0.05, ##p < 0.01, ###p < 0.001, compared with Vehicle

Gross observations showed that all GSK’872 regimens ameliorated transplant-induced lung congestion and edema, seen in vehicle group (Fig. 2B). Microscopically, LTx induced alveolar hemorrhage and perivascular edema (Fig. 2B). All three GSK’872 treatment protocols significantly reduced alveolar hemorrhage (Fig. 2C) and perivascular edema (Fig. 2D). Interestingly, neutrophil infiltration, detected as MPO positive cells by immunohistochemistry staining, was significantly reduced only in recipient intraperitoneal injection group (Group R) (Fig. 2E). GSK’872 reversed the increased IL-6 concentration also only in Group R (Fig. 2F). However, administration of GSK’872 to either the donor or recipient alone reversed the increased IL-1β concentration, while the combined administration showed no significant reversal in IL-1β levels (Fig. 2G). The result suggests that there may be an antagonistic effect when using GSK’872 at 5 µM in donor preservation solution and combination with 0.75 mg/kg for recipient intraperitoneal injection. We reduced the dose of GSK’872 for recipient to 0.15 mg/kg to avoid the potential toxicity. We found that combined low dose of GSK’872 (0.15 mg/kg) for recipient administration with 1 or 5 µM in donor preservation solution, significantly reduced LTx-induced IL-6 and IL-1β increase in the lungs (Supplementary Fig. 1).

GSK’872 attenuated necroptosis, pyroptosis and inflammatory response in rat lung transplants

The number of TUNEL-positive cells (as a sign of cell death in lung grafts) increased significantly after rat LTx, which was significantly reduced by GSK’872 treatment in all 3 regimens (Fig. 3A and B). Serum HMGB1, a marker of necrosis, increased significantly after LTx and was significantly reduced by GSK’872 only when it was added in the donor preservation solution (group D) (Fig. 3C).

Fig. 3.

GSK’872 reduced necroptosis, pyroptosis and inflammation in rat lung transplants. Rats were divided into five groups: Sham (surgery without ischemia-reperfusion), Vehicle (LTx without treatment), and three treatment groups (LTx with GSK’872 in the donor lung preservation solution and recipient injection [D + R], donor lung with 5 µM GSK’872 in preservation solution [D], and recipient intraperitoneally injected with 0.75 mg/kg GSK’872 [R]). (A) Representative TUNEL staining images showing cell death in rat lung sections. (B) Quantification of TUNEL-positive cell counts per sample. (C) Serum HMGB1 levels were determined using ELISA. (D) Representative Western blotting images. Quantification of protein levels of phosphorylated RIPK3 (pRIPK3) (E), phosphorylated MLKL (pMLKL) (F), NLRP3 (G), cleaved GSDMD (N-terminal GSDMD, N-GSDMD) (H), and cleaved caspase 3 (CC3) (I). Data were represented as median ± IQR. *p < 0.05, **p < 0.01, ***p < 0.001, compared with Sham; #p < 0.05, ##p < 0.01, ###p < 0.001, compared with Vehicle

Representative changes of proteins in different cell death pathways were analyzed with Western blotting and quantified with densitometry (Fig. 3D), and all unedited Western blotting images in this study are shown in supplementary materials. LTx significantly increased phosphorylated RIPK3 (pRIPK3) and phosphorylated MLKL (pMLKL), markers of necroptosis. pRIPK3 was significantly reduced only when GSK’872 was added to donor lung preservation solution (Fig. 3E), whereas pMLKL was suppressed by either donor or recipient administration (Fig. 3F). NLRP3 and cleaved GSDMD (N-terminal GSDMD, N-GSDMD), markers of pyroptosis, were elevated post-LTx. NLRP3 was reduced only with donor treatment (Fig. 3G), while N-GSDMD was suppressed by all three regimens (Fig. 3H). Taken together, administration of GSK’872 to lung preservation solution effectively decreased TUNEL-positive staining, the levels of HMGB1, pRIPK3, and pMLKL, suggesting this is the main route of its therapeutic effects. By contrast, GSDMD cleavage was reduced by GSK’872 either through lung preservation solution or injection to recipients. Interestingly, cleaved caspase 3 (CC3), an apoptosis marker, decreased after rat LTx with or without GSK’872 administration (Fig. 3I), suggesting 18 h cold preservation of donor lung led to a switch from apoptosis to necroptosis and pyroptosis.

GSK’872 inhibited necroptosis in AEC2, macrophages and neutrophils in rat LTx

To determine the effects of GSK’872 on cell death in different cell types, co-immunofluorescent staining was performed on rat lung tissue sections. LTx increased pMLKL-positive cells, co-localized with surfactant protein C (SPC) staining, a marker of AEC2 cells (Fig. 4A). The numbers of pMLKL/SPC double-positive cells were significantly reduced by GSK’872 in group D + R and D (Fig. 4A and B).

Fig. 4.

GSK’872 inhibited necroptosis and inflammatory responses in rat lung transplantation cell-type-specifically and route-dependently. Rats were divided into five groups: Sham (surgery without ischemia-reperfusion), Vehicle (LTx without treatment), and three treatment groups (LTx with GSK’872 in the donor lung preservation solution and recipient injection [D + R], donor lung with 5 µM GSK’872 in preservation solution [D], and recipient intraperitoneally injected with 0.75 mg/kg GSK’872 [R]). (A) Representative images of immunofluorescent co-staining for pMLKL (necroptosis marker, green) and surfactant protein C (SPC) (type 2 alveolar epithelial cell marker, red) in rat lung sections. (B) Percentage of pMLKL and SPC double-positive cells to SPC single positive cells per sample. (C-G) In addition to the five groups mentioned above, donor lungs (with or without GSK’872) after 18 h cold ischemia before reperfusion were also detected. (C) Necroptosis in macrophages were co-stained with antibodies for pMLKL (green) and CD68 (red). (D) Necroptosis in neutrophils were co-stained with antibodies for pRIPK3 (green) and MPO (red). (E) Quantification of CD68-positive cell counts per sample. (F) Percentage of pMLKL and CD68 double-positive cells to CD68 single-positive cells per sample. (G) Percentage of pRIPK3 and MPO double-positive cells to MPO single-positive cells per sample. (H) IL-1β levels in lung tissue homogenates were determined using ELISA. (I) MPO levels in serum were determined using ELISA. Data were represented as means ± SD in E-H and median ± IQR in B and I. *p < 0.05, **p < 0.01, ***p < 0.001, compared with Sham; #p < 0.05, ##p < 0.01, ###p < 0.001, compared with Vehicle

Macrophages were identified with an anti-CD68 antibody. Total CD68-positive macrophages showed no significant change between sham and post-transplant, but GSK’872 in donor lung preservation solution (group D + R and D) slightly yet significantly reduced macrophage numbers (Fig. 4C and E). We then examined vehicle and GSK’872 treated donor lungs prior to transplantation, similar drop of CD68⁺ cells was noted after 18 h preservation. By contrast, cold preservation significantly increased pMLKL/CD68 double-positive macrophages compared with Sham, which was significantly reduced by GSK’872 treatment in the preservation solution before reperfusion or after reperfusion (group D + R and D) (Fig. 4C and F), indicating suppression of macrophage necroptosis.

We found that MLKL expression is very low in rat neutrophils (Fig. 6D). The necroptosis in neutrophils is therefore determined with pRIPK3/MPO co-staining. Cold ischemia alone increased pRIPK3/MPO double-positive cells, which was reduced by GSK’872 post-preservation. After reperfusion, the total neutrophils and pRIPK3/MPO double-positive cells were significantly increased. Recipient intraperitoneal administration of GSK’872 (group D + R and group R) significantly reduced pRIPK3/MPO double-positive neutrophils (Fig. 4D and G), indicating inhibition of necroptosis in recipient neutrophils.

Fig. 6.

GSK’872 modulated macrophage PANoptosome assembly and drove the phenotypic shift from necroptosis to apoptosis in rat lung transplantation. Rats were divided into five groups: Sham (surgery without ischemia-reperfusion), Vehicle (LTx without treatment), and three treatment groups (LTx with GSK’872 in the donor lung preservation solution and recipient injection [D + R], donor lung with 5 µM GSK’872 in preservation solution [D], and recipient intraperitoneally injected with 0.75 mg/kg GSK’872 [R]). (A) Representative images of immunofluorescent co-staining for cleaved caspase 8 (green), phosphorylated RIPK3 (pRIPK3, red) and CD68 (macrophage marker, white) in rat lung sections. (B) Percentage of cleaved caspase 8 (CC8) and CD68 double-positive cells among CD68 single-positive cells per sample. (C) Percentage of pRIPK3 and CD68 double-positive cells among CD68 single-positive cells per sample. (D) Number of CC8⁺ CD68⁻ cells per sample. Data were represented as means ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, compared with Sham; ###p < 0.001, compared with Vehicle

LTx significantly increased IL-1β (mainly macrophage-derived [24]) in lung tissue homogenates and MPO (mainly neutrophil-derived [25]) in serum. GSK’872 in preservation solution reduced IL-1β elevation in group D (Fig. 4H), while recipient intraperitoneal administration reduced serum MPO levels in group R (Fig. 4I).

Effects of GSK’872 on epithelial and macrophage cell death in a CI/R-treated cell culture model

Since adding GSK’872 in the lung preservation solution effectively reduced reperfusion-induced necroptosis in rat lung AEC2 cells and macrophages, we next aim to explore the mechanisms in these cell types with a CI/R-treated cell culture model [20, 21].

Human lung epithelial BEAS-2B cells were treated with different doses of GSK’872 (from 0 to 20 µM) under standard culture condition, or subjected to simulated CI/R. GSK’872 was added to CI and/or R phases. GSK’872 did not affect cell viability, and CI/R significantly reduced cell viability to approximately 80%. None of the treatment regimens of GSK’872 rescued cells from CI/R-reduced cell viability (Fig. 5A). GSK’872 at 5 µM also showed no significant effects on CI/R-induced HMGB1 release in BEAS-2B cells (Supplementary Fig. 2).

Fig. 5.

GSK’872 promoted apoptosis in macrophages while inhibiting their necroptosis during cold ischemia/warm reperfusion (CI/R). BEAS-2B cells (A) and RAW264.7 macrophages (B) were treated with various concentrations of GSK’872 using three different regimens: supplying GSK’872 throughout the entire CI 18 h/R 4 h process, adding it in Low-Potassium Dextran (LPD) lung preservation solution during cold ischemia, and including it only in the culture medium during reperfusion. The cell viability was detected using the CCK-8 kit. Data were represented as means ± SD of ten independent experiments. (C-H) RAW264.7 macrophages were divided into six groups: control (no CI/R), control + GSK’872 [5 µM], vehicle (with CI 18 h/R 4 h), and three GSK’872-treated groups (supplying GSK’872 throughout the entire CI/R process [D + R], adding GSK’872 in the donor lung preservation solution [D], and adding GSK’872 in the culture medium during reperfusion [R]). Data were represented as means ± SD of six independent experiments. (C) LDH levels in the cell supernatant were determined using a commercial reagent kit. (D) TNF-α levels in the cell supernatant were determined using ELISA. (E) Representative Western blots of apoptosis-related proteins in RAW264.7 macrophages. Quantification of protein levels of RIPK1 (F), Cleaved RIPK1 (G) and cleaved caspase 3 (H). *p < 0.05, **p < 0.01, ***p < 0.001, compared with control; #p < 0.05, ##p < 0.01, ###p < 0.001, compared with vehicle (CI/R without GSK’872)

RAW264.7 macrophages were treated with the same experimental protocols. Again, GSK’872 did not affect cell viability. However, CI/R-induced cell viability decrease was enhanced by GSK’872 in a dose-dependent manner, especially when it was added to both CI and/or R phases (Fig. 5B), which helps to explain the GSK’872-induced reduction of macrophage numbers after 18 h cold preservation in rat lungs. GSK’872 at 5 µM was used in the subsequent experiments with macrophages.

CI/R exposure significantly increased LDH release from RAW264.7 macrophages, while combined GSK’872 administration in both cold lung preservation solution and warm reperfusion medium (D + R) further promoted cell LDH release (Fig. 5C). The CI/R exposure also significantly increased TNF-α release from RAW264.7 macrophages, and the two groups administered with GSK’872 in lung preservation solution (D + R, D) had even higher TNF-α levels (Fig. 5D). It has been proposed that TNF-α is mainly released by apoptotic cells [26]. These findings prompted us to investigate the effects of GSK’872 on apoptosis in CI/R-exposed RAW264.7 macrophages. Western blotting analysis (Fig. 5E) showed that CI/R significantly decreased the RIPK1 level (Fig. 5F), whereas cleaved RIPK1 (Fig. 5G) and cleaved caspase 3 (Fig. 5H) levels were significantly increased, thereby confirming increased apoptosis in RAW264.7 macrophages. The GSK’872 administration in lung preservation solution (D + R, D) significantly enhanced these changes. These results suggest that GSK’872 administration in lung preservation solution inhibits necroptosis but promotes apoptosis in CI/R-exposed RAW264.7 macrophages.

GSK’872 modulated macrophage PANoptosome assembly and drove the phenotypic shift from necroptosis to apoptosis in rat LTx

To elucidate the coordinated PANoptotic signaling that drives the shift from necroptosis to apoptosis in macrophages, we performed immunofluorescent co-staining of cleaved caspase 8 and pRIPK3 (key components of PANoptosomes) in CD68-positive macrophages in rat lung tissue (Fig. 6).

The Sham group exhibited baseline, near-absent signals for cleaved caspase 8 and pRIPK3. Upon LTx (Vehicle), both markers were sharply increased and predominantly co-localized within CD68⁺ cells (Fig. 6A), indicating the extensive assembly of PANoptosome complexes within macrophages post-LTx.

Compared with the Vehicle group, administration of GSK’872 via the donor preservation solution (D + R, D) significantly reduced the proportion of pRIPK3 and CD68 double-positive macrophages while markedly increasing the proportion of cleaved caspase 8 and CD68 double-positive macrophages (Fig. 6B and C), confirming that GSK’872 effectively shifts the macrophage PANoptosome from a pro-necroptotic toward a pro-apoptotic state. In contrast, no such shifting effect was observed with recipient-only administration (Fig. 6B and C), which is consistent with our in vitro cell experiments (Fig. 5G and H) and suggests the effect of GSK’872 on the macrophage PANoptosome depended on its administration via donor preservation solution.

In addition, compared to the remaining groups, combined donor and recipient treatment (D + R) also significantly increased the number of cleaved caspase 8⁺ CD68⁻ non-macrophage cells throughout the rat lung tissue (Fig. 6D), suggesting that this excessive pro-apoptotic effect may account for the compromised efficacy of the combined treatment.

Direct effects of GSK’872 on neutrophils in a cell culture model

To determine the effects of GSK’872 on neutrophils, we isolated neutrophils from rat peripheral blood, and challenged them with LPS to simulate the inflammatory responses of IRI during LTx. LPS induced significant formation of NETs detected with DAPI staining on neutrophil smeared slides (Fig. 7A and B), and release of MPO in the cell supernatant (Fig. 7C). These changes were significantly reduced by GSK’872 administration.

Fig. 7.

GSK’872 inhibited lipopolysaccharide (LPS)-induced neutrophil extracellular traps (NETs) formation, necroptosis and pyroptosis in rat neutrophils. Rat neutrophils were divided into four groups: control, GSK’872 [50 µM, 3 h, including 1 h of pretreatment], LPS [20 µg/mL, 2 h], and LPS + GSK’872. (A) Representative images of NETs formation in rat neutrophil smears. (B) Percentage of NETs formation in total cells. (C) MPO in the cell supernatant was determined using ELISA. (D) Representative Western blotting images of cell death-related proteins. Quantification of protein levels of pRIPK3 (E), NLRP3 (F), GSDMD (G) and cleaved GSDMD (N-GSDMD) (H) in rat neutrophils. Data were represented as means ± SD of six independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, compared with Control; #p < 0.05, ##p < 0.01, ###p < 0.001, compared with LPS

Western blotting showed that LPS treatment significantly increased the level of pRIPK3, which was inhibited by GSK’872 (Fig. 7D and E). Although LPS did not influence the levels of NLRP3 (Fig. 7F) and GSDMD (Fig. 7G), GSK’872 administration significantly decreased them in both control and LPS-challenged conditions. On the other hand, LPS significantly increased the level of cleaved GSDMD (N-GSDMD), a sign of pyroptosis, which was inhibited by GSK’872 (Fig. 7H). These results suggest that GSK’872 may inhibit the NETs formation, MPO release, necroptosis and pyroptosis in neutrophils.

CM from GSK’872-treated macrophages or neutrophils reduced CI/R-induced lung epithelial cell death

GSK’872 protected AECs in the LTx-IRI model; however, it failed to confer protection in the CI/R cell culture system. We hypothesize that this discrepancy is due to the inhibition of cell death and inflammatory responses in macrophages or neutrophils, thereby indirectly preserving AEC viability. To test this hypothesis, we first treated CI/R-exposed RAW264.7 macrophages or LPS-stimulated rat neutrophils with or without GSK’872, collected the CM and then tested them on either control or CI/R-exposed BEAS-2B epithelial cells (Fig. 8A).

Fig. 8.

Conditioned medium from GSK’872-treated macrophages or neutrophils reduced CI/R-induced epithelial cell death. (A) Conditioned medium (CM) were collected from RAW264.7 macrophages exposed to CI/R or from LPS stimulated rat neutrophils, with or without GSK’872 treatment. BEAS-2B cells were incubated under 4 conditions: Control or CI/R, with different CM collected as described above. (B) Propidium Iodide (PI) positive staining indicates cell death. (C) Quantification of PI-positive cell counts per sample. (D) Cell viability of BEAS-2B cells was detected by CCK-8 assay. Data were represented as means ± SD of six independent experiments. **p<0.01, ***p<0.001, compared with Control; ##p<0.01, ###p<0.001, compared with CI/R

BEAS-2B cell death was measured by propidium iodide (PI) staining. Under control conditions, CM with or without GSK’872 did not induce cell death, while under CI/R conditions cell death was significantly less in cells exposed to CM from GSK’872-treated RAW264.7 macrophages or rat neutrophils, than that without GSK’872 (Fig. 8B and C). These CM also significantly inhibited CI/R-induced reduction in viability in BEAS-2B cells, as measured by CCK assay (Fig. 8D).

Furthermore, whether substituting LPS with HMGB1 to stimulate rat neutrophils or utilizing the NR8383 cells instead of RAW264.7 cells, the resulting CM from GSK’872-treated cells consistently reversed CI/R-induced cell viability reduction in BEAS-2B cells, but the combined media (CI/R+BothGSK) did not provide significantly enhanced protection compared to the individual treatments (Supplementary Fig. 3). Consistent with the lack of additive effects observed in the previous in vivo dual-administration experiments, this suggests a ceiling effect for GSK’872.

Discussion

In this study, we found that GSK’872 effectively reduced IRI in rat LTx, and its efficacy varied depending on the administration regimen. As a known RIPK3 inhibitor, it showed broader regulation on PANoptosis in a cell-type specific manner. Moreover, we propose a crosstalk from macrophages/neutrophils to lung epithelial cells based on cell biological studies.

GSK’872 has demonstrated therapeutic potential across a diverse range of disease models, including subarachnoid hemorrhage [27], systemic inflammation-induced seizures [28], stroke [29], various retinopathies [30, 31], pneumonia [32], and lupus nephritis [33]. These protective effects are primarily mediated through the classical inhibition of necroptosis. Beyond its well-established role, GSK’872 also exerts therapeutic effects by modulating non-classical functions of RIPK3. These functions include reducing IL-1β release induced by SARS-CoV-2 [34] and inhibiting rheumatoid arthritis driven by malate shuttle-dependent mitochondrial respiration [35]. Furthermore, GSK’872 has been shown to prevent ConA-induced immune-mediated hepatitis by promoting the infiltration of myeloid-derived suppressor cells [36]. GSK’872 also protects against myocardial fibrosis by restoring autophagic flux [37] and by inhibiting the CaMKII δ-mediated pathway [38]. Other non-canonical roles include the protection of spinal cord injury by enhancing mitochondrial antioxidative capacity [39].

Regimen-specific therapeutic effects of GSK’872 on LTx-IRI

GSK’872 attenuated IRI in rat LTx by reducing alveolar hemorrhage, edema, neutrophil infiltration, and pro-inflammatory cytokines, while suppressing necroptosis, pyroptosis, and associated inflammation. Moreover, different administration regimens of GSK’872 yielded distinct therapeutic effects in a cell-type specific manner.

When added to preservation solution, GSK’872 reduced TUNEL⁺ cells, HMGB1, pRIPK3, and NLRP3, and suppressed necroptosis in AEC2 cells and macrophages. In contrast, recipient injection primarily targeted neutrophils, reducing infiltration, serum MPO, and NETs formation. Both routes inhibited necroptosis and pyroptosis, but in a cell-type-specific manner—macrophages during cold ischemia, neutrophils during reperfusion.

Administration route dictates cellular targeting: airway delivery favors epithelial and alveolar macrophages [40], while systemic routes target endothelial cells [41] and intravascular leukocytes [42]. IRI mechanisms also differ by phase—some interventions are effective during ischemia [43, 44], others during reperfusion [45].

GSK’872’s efficacy via donor or recipient administration highlights its therapeutic potential. Of course, when combined application is considered, timing and dosing require further optimization to avoid drug toxicity to donor lungs and/or recipients. This toxicity is driven by high concentrations of GSK’872 binding to RIPK3 to form complexes, which subsequently recruit and activate caspase 8, thereby mediating apoptosis [46–48]. This is directly evidenced by immunofluorescent staining in our present study, which showed that the combined donor–recipient treatment significantly increased apoptosis in non-macrophage cells.

The distinct efficacy of GSK’872 across different regimens underscores the need for careful design in preclinical studies for drugs, including animal species (large animals should be considered), donor lung injury models (prolonged cold preservation, acid aspiration, warm ischemia), and administration regimens.

Effects of GSK’872 on PANoptosis in LTx-IRI

Among programmed cell death pathways, pyroptosis [49, 50], apoptosis [40, 51], necroptosis [18, 19, 52], and ferroptosis [43, 53, 54] are key in LTx-IRI. PANoptosis—co-regulation of pyroptosis, apoptosis, and necroptosis [55]—is supported by transcriptomic data from lungs after LTx [9] and ex vivo lung perfusion [56]. However, these profiles required validation at the protein level across cell types, and PANoptosis’ role in IRI-LTx needed experimental confirmation.

Here, bioinformatics analysis revealed enriched expression of inflammation- and cell death-related genes in lung epithelial cells, macrophages, and neutrophils. We then demonstrated GSK’872 (a known RIPK3 inhibitor) protected rat lungs from IRI. Interestingly, beyond inhibiting necroptosis, GSK’872 also suppresses pyroptosis and regulates apoptosis. These results provide experimental evidence on the regulatory relationships among the three cell death modes in IRI.

Beyond inhibiting necroptosis, GSK’872 also reduces GSDMD cleavage (a pyroptosis marker) in post-LTx lung tissue. In cell culture, GSK’872 similarly inhibits pRIPK3 (necroptosis marker) and pyroptosis markers (NLRP3, GSDMD, N-GSDMD) in neutrophils. These results suggest that necroptosis and pyroptosis can be co-regulated by IRI in the lung, especially in neutrophils in a GSK’872-related mechanism.

PANoptosome-mediated apoptosis-to-necroptosis shift in LTx-IRI

In rat LTx, prolonged hypothermic preservation of donor lungs led to the dominant cell death mode shifts from apoptosis to necrosis [57]. Aligning with this dynamic shift, we observed reduced cleaved caspase 3 in rat lung tissue post-reperfusion. Further mechanistic exploration in macrophage cultures revealed that GSK’872 induced cleavage of RIPK1 (a crucial step for limiting necroptosis [58]), and cleavage of caspase 3 (a crucial step for apoptosis). This discrepancy strongly underscores that the three cell death modalities within PANoptosis do not necessarily follow a synchronized or parallel trend, but rather undergo dynamic shifts between one another.

What is the key mechanism driving this shift in cell death modalities? Inhibition of caspase 8 led to embryonic lethality in mice by shifting cell death from apoptosis to necroptosis and pyroptosis [59]. Caspase 8 and RIPK3 are universally recognized as core constituents across various PANoptosomes [60]. Both molecules are intimately regulated by GSK’872: apart from inhibiting RIPK3 phosphorylation, GSK’872 physically binds to RIPK3, establishing a complex that recruits caspase 8 to drive apoptosis and inhibit necroptosis [46, 48, 61].

We propose that GSK’872 shifts the catalytic profile of PANoptosomes, thereby transitioning the macrophage death modality from a necroptotic toward a pro-apoptotic state—a mechanism successfully validated by immunofluorescent staining on rat lung slices.

It is important to note that the rat LTx only partially recapitulates the complexity of human LTx. Clinically, most donor lungs have experienced injury related to donor factors (age, cause of death, comorbidities) and preservation conditions. The activation of different cell death and inflammatory pathways vary from case to case. The PANoptosis can be presented in different manners.

Programmed cell death mediated crosstalk between inflammatory and parenchymal cells in LTx-IRI

At the cellular level, we found GSK’872 regulated cell death in a cell-type-specific manner and revealed crosstalk between inflammatory cells and lung epithelial cells. In macrophages [62] and neutrophils [63], apoptosis is immune-silent and anti-inflammatory, whereas necroptosis and pyroptosis promote proinflammatory mediator release.

GSK’872 in preservation solution reduced donor-derived pulmonary macrophages and their necroptosis in vivo. In vitro, it promoted cleavage of caspase-3 and RIPK1, favoring apoptosis over necroptosis, thereby suppressing damage-associated molecular patterns (DAMPs) release [62, 64] and macrophage-related inflammation.

Since it is currently difficult to distinguish between macrophages and monocytes at the tissue-section level, GSK’872 may also promote apoptosis and inhibit necroptosis in donor-derived intravascular non-classical monocytes (NCMs). These are critical, deleterious cells involved in recruiting recipient-derived inflammatory cells [65, 66]. Because our organ preservation solution containing GSK’872 is delivered via vascular perfusion, it allows for more thorough drug contact with intravascular NCMs and facilitates their clearance.

Recipient neutrophil infiltration is a hallmark of LTx-IRI [67]. While neutrophil apoptosis is protective [18], NETs formation exacerbates injury [68]. Intraperitoneal GSK’872 in recipients specifically inhibited necroptosis in recipient-derived neutrophils and directly suppressed NETs, necroptosis, and pyroptosis in cultured neutrophils.

An intriguing observation in our study is the inhibition of GSK’872 on necroptosis in AEC2 cells after LTx, but no effects on CI/R induced reduction of viability of lung epithelial cells in culture. A conditioned medium study demonstrated, when macrophages or neutrophils were treated with GSK’872, CI/R or LPS induced cell injury was inhibited, and the “death signal” in the conditioned medium was significantly reduced. These results further suggest that GSK’872 not only reduced necroptosis and pyroptosis from macrophages and neutrophils directly, but also protected other cell types from further damage by reducing the release of DAMPs. The crosstalk between inflammatory cells and parenchymal cells in the lungs is indicated for IRI in LTx.

This study possesses several limitations. First, to induce more severe injury and thereby validate the efficacy of GSK’872 more clearly, we utilized a rat lung transplantation model with cold storage at 4 °C for 18 h. However, 10 °C cold storage is increasingly becoming the clinical standard for human lung transplantation [69]. Second, there is a significant discrepancy between small animal models and human pathophysiology. So, future studies will utilize 10 °C cold storage large animal models and incorporate more clinically significant parameters to confirm the drug’s therapeutic effects. Finally, simple cellular models cannot accurately reflect the complex multicellular interactions involved in lung transplantation; the adoption of organoid models represents a key direction for our future research [70].

In conclusion, administration of GSK’872 to donor lung preservation solution protected macrophages from necroptosis, whereas its injection to recipients inhibited necroptosis of neutrophils in rat LTx. Mechanistic studies suggested a crosstalk between inflammatory cells and lung parenchymal (e.g. epithelial) cells, which help to explain the potential therapeutic benefits of GSK’872 on LTx-IRI. The direct and indirect effects of GSK’872 on pyroptosis, necroptosis and apoptosis revealed by this study indicate that PANoptosis and related inflammation are targets for further investigation.

Abbreviations

AEC1

Type 1 alveolar epithelial cells

AEC2

Type 2 alveolar epithelial cells

ANOVA

Analysis of variance

AQP5

Aquaporin 5

CC3

Cleaved caspase 3

CI/R

Cold ischemia/warm reperfusion

CM

Conditioned medium

DEGs

Differentially expressed genes

ELISA

Enzyme-linked immunosorbent assay

FBS

Fetal bovine serum

GEO

Gene Expression Omnibus

GSDMD

Gasdermin D

HPFs

High-power fields

IQR

Interquartile range

IRI

Ischemia-reperfusion injury

KEGG

Kyoto Encyclopedia of Genes and Genomes

LDH

Lactate dehydrogenase

LPS

Lipopolysaccharide

LPD

Low-Potassium Dextran

LTx

Lung transplantation

MLKL

Mixed Lineage Kinase Domain Like

MPO

Myeloperoxidase

NCMs

Non-classical monocytes

N-GSDMD

N-terminal GSDMD

NETs

Neutrophil extracellular traps

PGD

Primary graft dysfunction

PI

Propidium Iodide

PMSF

Phenylmethanesulfonyl fluoride

pMLKL

Phosphorylated MLKL

pRIPK3

Phosphorylated RIPK3

RIPK1

Receptor Interacting Protein Kinase 1

RIPK3

Receptor Interacting Protein Kinase 3

scRNA-seq

Single-cell RNA sequencing

SD

Standard deviation

SPC

Surfactant protein C

TSA

Tyramide signal amplification

Authors’ contributions

JSL, YBL, MW designed the experiments. JSL, XJM, JHH, XLH, QYF, YCW performed the animal and pathological experiments. JSL, XJM, JHH, YFY, ZLL, FXL performed the cell and biochemical experiments. JSL, YFY, JHH analyzed the data. JSL, YBL, MW wrote the paper. MYL revised the manuscript. MW supervised the project. All authors have read and agreed to the submitted version of the manuscript.

Funding

This work was supported by grants from the National Natural Science Foundation of China (82572504), the National Key R&D Program of China (2023YFF1204404), the Zhejiang Provincial Natural Science Foundation of China (LZ25H310001, LZ20H010001), and the Zhejiang Provincial Medical and Health Science and Technology Plan (2025KY068).

Data availability

All the data of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The rat experimental protocol was approved by the Animal Care and Use Committee of Zhejiang University (Approval No. ZJU20220159).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.