Regulatory role of POSTN in keloid pathogenesis

Introduction

Keloids are a common, chronic and inflammatory skin disorder, characterized by benign fibroproliferative overgrowth of the skin (1). Keloids grow continuously and aggressively beyond the original wound boundaries, resulting from cutaneous injury or abnormal wound healing associated with inflammatory and fibrotic conditions (2–4). Although it is not a life-threatening condition, the incidence of keloids is relatively high, thus reducing the quality of life of patients due to pain, pruritus and cosmetic reasons (5,6). Various modalities, including surgery, radiation therapy and intralesional steroid injections, have been used for the treatment of keloids; however, they commonly relapse (7–9). In addition, the underlying molecular mechanisms of keloid formation remain largely unclear, thus restricting the development of therapeutic strategies.

Numerous studies have investigated genes and pathways associated with the pathophysiological processes involved in the development of keloids, thus providing novel insights into their onset (10–13). Notably, emerging evidence has suggested that inflammatory etiologies, particularly chronic inflammation, may be involved in the pathogenesis of keloids (14,15). Furthermore, previous studies have indicated that various pro-inflammatory cytokines/chemokines and their receptors could be notably upregulated in keloids and closely associated with keloid inflammation (16–18). For example, IL-17 can enhance the expression of C-X-C motif chemokine ligand (CXCL)12 in keloids, which in turn promotes the infiltration of C-X-C motif chemokine receptor (CXCR)4+ cells in keloid scars (18–20). In addition, existing data have suggested that T helper 2 (Th2)-related cytokines, such as IL-4 and IL-13, may act as crucial mediators of keloid inflammation (21). Mechanistically, IL-4 and IL-13 could be involved into skin fibrosis via excessive fibroblast proliferation and collagen deposition in keloids (21).

Previous cDNA microarray analyses have been performed to compare the gene expression profiles of fibroblasts between normal skin and keloid tissues. These analyses have identified several gene and pathway alterations, such as those in hypoxia-inducible factor 1α (HIF-1α) (11), the fibrosis and Wnt signaling pathway (22) and the extracellular matrix (23), all of which are closely associated with the pathogenesis of keloids.

Although gene chips could identify dysregulated genes in keloid tissues, the small sample size and considerable heterogeneity may affect data reproducibility. Therefore, the present study retrieved several microarray datasets from the Gene Expression Omnibus (GEO) database, with the aim of identifying common differentially expressed genes (DEGs) between keloid fibroblasts (KFs) and normal fibroblasts (NFs). Then, gene knockdown assays and RNA sequencing were performed to investigate the expression profiles, functions and potential pathogenic mechanisms of DEGs in keloids.

Materials and methods

Microarray data processing and screening of target DEGs

Three independent keloid datasets, namely GSE145725 (11), GSE7890 (22) and GSE44270 (23), were obtained from the GEO (https://www.ncbi.nlm.nih.gov/geo/). A total of 19 samples from GSE145725 (10 NF and 9 KF samples), 10 samples from GSE7890 (5 NF and 5 KF samples) and 12 samples from GSE44270 (3 NF and 9 KF samples) were selected for further analysis (details are shown in Table I). The original expression matrix was normalized and analyzed using R package (version 4.2.1) (24). DEGs were screened using the Limma package (25). Genes with P<0.05 and |log2 fold change (FC)|>1 were defined as DEGs in GSE145725, GSE7890 and GSE44270. TBtools was used to construct a Venn diagram to identify common DEGs from the three datasets (26).

Table I. Details of datasets used in the present study.

Table I.

Details of datasets used in the present study.

First author, year	GEO accession no.	Number of KF samples	Number of NF samples	KF samples	NF samples	Experiment type/platform	(Refs.)
Smith et al,	GSE7890	5	5	GSM194109;	GSM194118;	Expression	(22)
2008				GSM194110;	GSM194119;	profiling by
				GSM194111;	GSM194120;	array/
				GSM194112;	GSM194121;	GPL570
				GSM194113	GSM194122
Hahn et al,	GSE44270	9	3	GSM1081582;	GSM1081608;	Expression	(23)
2013				GSM1081583;	GSM1081609;	profiling by
				GSM1081584;	GSM1081610	array/
				GSM1081585;		GPL6244
				GSM1081586;
				GSM1081587;
				GSM1081588;
				GSM1081589;
				GSM1081590
Kang et al,	GSE145725	9	10	GSM4331585;	GSM4331594;	Expression	(11)
2020				GSM4331586;	GSM4331595;	profiling by
				GSM4331587;	GSM4331596;	array/
				GSM4331588;	GSM4331597;	GPL6244
				GSM4331589;	GSM4331598;
				GSM4331590;	GSM4331599;
				GSM4331591;	GSM4331600;
				GSM4331592;	GSM4331601;
				GSM4331593	GSM4331602,
					GSM4331603

[i] GEO, Gene Expression Omnibus; KF, keloid fibroblast; NF, normal fibroblast.

Patients and specimen collection

All subjects (three patients with keloids and three healthy controls) were recruited from February to May 2023 from the Department of Dermatology, Peking University Shenzhen Hospital (Shenzhen, China). The average age of patients with keloids was 28.3 years (range, 25–30 years), while the average age of healthy individuals was 28.7 years (range, 23–37 years). A total of two men and one woman were diagnosed with keloids, and sex-matched healthy controls (two men and one woman) were included in the present study. All of the patients with keloids enrolled in the current study did not receive any topical therapy or systemic treatment for ≥1 month before skin biopsy. Biopsies from the non-lesional skin of healthy volunteers were used as controls. Specimens were immediately placed in ice-cold 1X PBS pre-mixed with 1X penicillin and 1X streptomycin. The present study was approved by the Ethics Committee of Peking University Shenzhen Hospital and was conducted in accordance with The Declaration of Helsinki. Written informed consent was obtained from all participants.

Cell culture

Primary human dermal fibroblasts were isolated from skin tissues obtained from the three patients with keloids and the three healthy controls as previously described (27). Firstly, the epidermis and adipose tissues were removed from the specimens obtained from the healthy donors/patients with keloids. Subsequently, the dermis was cut into 0.25 cm2 sections and allowed to attach to a 10-cm culture dish. Fibroblasts began to migrate out and proliferate from the edges of the dermis onto the culture dish. This process yields a pure population of fibroblasts for subsequent studies. The fibroblasts obtained from keloid patients were designated KF#1, KF#2 and KF#3, while those derived from healthy donors were labeled NF#1, NF#2 and NF#3. The fibroblast cells were cultured in DMEM supplemented with 10% fetal bovine serum, 100 U/ml penicillin and 100 µg/ml streptomycin (all Gibco; Thermo Fisher Scientific, Inc.) in a CO2 incubator with 5% CO2 at 37°C.

Reagents, small interfering (si)RNAs and antibodies

siRNAs targeting POSTN and IL-4R, as well as the controls (si-CTL, forward, 5′-CCUAAGGUUAAGUCGCCCU-3′ and reverse, 5′-AGGGCGACUUAACCUUAGG-3′) were from Guangzhou RiboBio Co., Ltd. The siRNA sequences targeting the POSTN gene were as follows: si-POSTN#1, forward, 5′-GUGACAGUAUAACAGUAAA-3′ and reverse, 5′-UUUACUGUUAUACUGUCAC-3′; si-POSTN#2, forward, 5′-GGAUCAGCGCCUCCUUAAA-3′ and reverse, 5′-UUUAAGGAGGCGCUGAUCC-3′. The siRNA sequences targeting the IL-4R gene were as follows: si-IL-4R#1, forward, 5′-GUCAACAUUUGGAGUGAAA-3′ and reverse, 5′-UUUCACUCCAAAUGUUGAC-3′; si-IL-4R#2, forward, 5′-CUUCCACCUUCGGGAAGUA-3′ and reverse, 5′-UACUUCCCGAAGGUGGAAG-3′. Antibodies against POSTN (cat. no. ab14041) and IL-4R (cat. no. ab203398) were from Abcam, antibodies against phosphorylated (p)-STAT1 (Tyr701) (cat. no. 9167), p-STAT1 (Ser727) (cat. no. 8826) and STAT1 (cat. no. 14994) were from Cell Signaling Technology, Inc., whereas the antibody against β-actin (cat. no. AF5003) was purchased from Beyotime Institute of Biotechnology. The SuperSignal West Femto Chemiluminescent Substrate Kit used for signal detection in western blotting was obtained from Thermo Fisher Scientific, Inc. (cat. no. 34095). All other reagents or chemicals were purchased from Thermo Fisher Scientific, Inc. unless otherwise mentioned.

Western blot analysis

Western blotting was performed as described previously (28). Briefly, protein samples from fibroblasts were harvested using ice-cold RIPA buffer containing 10 mM Tris pH 7.4, 100 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM NaF, 20 mM Na4P2O7, 1% Triton X-100, 10% glycerol, 0.1% SDS and 0.5% deoxycholate, and protease and phosphatase inhibitor cocktails. The protein concentration was determined using the BCA method (cat. no. 5000002; Bio-Rad Laboratories, Inc.). A total of 10 µg/lane protein was separated using a 10% SDS-PAGE gel, after which they were transferred onto PVDF membranes. The membranes were blocked with blocking buffer (cat. no. ST023; Beyotime Institute of Biotechnology) for 1 h at room temperature, and incubated with primary antibodies (1:1,000) as follows: phosphorylated (p)-STAT1 (Tyr701) (cat. no. 9167), p-STAT1 (Ser727) (cat. no. 8826) and STAT1 (cat. no. 14994; all Cell Signaling Technology, Inc.; β-actin (cat. no. AF5003) was from Beyotime Institute of Biotechnology. The membranes were incubated with primary antibodies overnight at 4°C. Subsequently, the corresponding secondary antibodies, goat anti-rabbit IgG H&L (HRP) (cat. no. ab205718; 1:10,000; Abcam) for p-STAT1 (Tyr701), p-STAT1 (Ser727) and STAT1 and m-IgGκ BP-HRP (cat. no. sc-516102; dilution: 1:10,000; Santa Cruz Biotechnology, Inc.) for β-actin, were incubated for 2 h at room temperature. The signals were detected using the SuperSignal West Femto Chemiluminescent Substrate Kit (cat. no. 34095; Thermo Fisher Scientific, Inc.). The images were analyzed using ImageJ software (version 1.54; National Institutes of Health) and the relative content of the target protein was expressed as the gray value of the target protein/β-actin gray value. A total of three biological replicates was performed for each treatment group.

Immunofluorescence staining

POSTN levels in KFs were detected using immunofluorescence staining as described previously (29). Briefly, cells were fixed using 4% paraformaldehyde at room temperature for 15 min, followed by permeabilization for 10 min at room temperature using 0.1% Triton X-100 (cat. no. P0096; Beyotime Institute of Biotechnology). Cells were blocked with blocking solution (cat. no. P0102; Beyotime Institute of Biotechnology) for 1 h at room temperature, followed by incubation with the primary antibody against POSTN (cat. no. ab14041; dilution: 1:100; Abcam) at 4°C overnight. After washing off the primary antibody, an Alexa Fluor555-labeled donkey anti-rabbit IgG (H+L) secondary antibody (cat. no. A0453; 1:200; Beyotime Institute of Biotechnology) was applied for 30 min at room temperature. For immunofluorescence staining, cells were counterstained with DAPI (cat. no. ab104139; Abcam) at room temperature for 15 min, and fluorescence microscopy (Nikon Inc., Melville, NY) was used for observation.

RNA isolation and reverse transcription-quantitative PCR (RT-qPCR)

Total RNA was extracted with TRIzol® reagent (Invitrogen; Thermo Fisher Scientific, Inc.). Subsequently, RT-qPCR was performed as described previously (29). Briefly, total RNA was reverse-transcribed into cDNA using ReverTra Ace qPCR RT kit (cat. no. FSQ-201; Toyobo Life Science) with the following temperature protocol: 37°C for 15 min, 50°C for 5 min, and 98°C for 5 min. qPCR was performed using iTaq™ Universal SYBR® Green Supermix (cat. no. 1725121; Bio-Rad Laboratories, Inc.) with an initial denaturation at 95°C for 30 sec and the following thermocycling conditions: 95°C for 10 sec, 59°C for 25 sec and 72°C for 30 sec for 40 cycles. The relative mRNA expression levels of target genes were normalized to those of GAPDH using the 2−ΔΔCq method (30). All of the primers used in the present study were designed using the online software PrimerBank (https://pga.mgh.harvard.edu/primerbank/index.html). The primer sequences were as follows: GAPDH-forward (F), 5′-ACAACTTTGGTATCGTGGAAGG-3′, GAPDH-reverse (R), 5′-GCCATCACGCCACAGTTTC-3′; POSTN-F, 5′-CTCATAGTCGTATCAGGGGTCG-3′, POSTN-R, 5′-ACACAGTCGTTTTCTGTCCAC-3′; IL-4R-F, 5′-CGTGGTCAGTGCGGATAACTA-3′, IL-4R-R, 5′-TGGTGTGAACTGTCAGGTTTC-3′. CCNA1-F, 5′-GAGGTCCCGATGCTTGTCAG-3′, CCNA1-R, 5′-GTTAGCAGCCCTAGCACTGTC-3′; CCNA2-F, 5′-CGCTGGCGGTACTGAAGTC-3′, CCNA2-R, 5′-GAGGAACGGTGACATGCTCAT-3′; CCNB1 F, 5′-AATAAGGCGAAGATCAACATGGC-3′, CCNB1-R, 5′-TTTGTTACCAATGTCCCCAAGAG-3′; CCNB2 F, 5′-CCGACGGTGTCCAGTGATTT-3′, CCNB2-R, 5′-TGTTGTTTTGGTGGGTTGAACT-3′; CCNC-F, 5′-CCTTGCATGGAGGATAGTGAATG-3′, CCNC-R, 5′-AAGGAGGATACAGTAGGCAAAGA-3′; CCND3-F, 5′-TACCCGCCATCCATGATCG-3′, CCND3-R, 5′-AGGCAGTCCACTTCAGTGC-3′; CCNE2-F 5′-TCAAGACGAAGTAGCCGTTTAC-3′, CCNE2-R, 5′-TGACATCCTGGGTAGTTTTCCTC-3′; CDK1-F 5′-AAACTACAGGTCAAGTGGTAGCC-3′ and R, 5′-TCCTGCATAAGCACATCCTGA-3′; CDK2-F, 5′-CCAGGAGTTACTTCTATGCCTGA-3′, CDK2-R, 5′-TTCATCCAGGGGAGGTACAAC-3′; CDK3-F, 5′-GAAGGTAGAGAAGATCGGAGAGG-3′, CDK3-R, 5′-GTCCAGCAGTCGGACGATG-3′; CDK4-F, 5′-ATGGCTACCTCTCGATATGAGC-3′, CDK4-R, 5′-CATTGGGGACTCTCACACTCT-3′; CDK5-F, 5′-GGAAGGCACCTACGGAACTG-3′, CDK5-R, 5′-GGCACACCCTCATCATCGT-3′; JAK1-F, 5′-CTTTGCCCTGTATGACGAGAAC-3′, JAK1-R, 5′-ACCTCATCCGGTAGTGGAGC-3′; JAK2-F, 5′-TCTGGGGAGTATGTTGCAGAA-3′, JAK2-R, 5′-AGACATGGTTGGGTGGATACC-3′; STAT1-F, 5′-CGGCTGAATTTCGGCACCT-3′, STAT1-R, 5′-CAGTAACGATGAGAGGACCCT-3′; STAT2-F, 5′-CCAGCTTTACTCGCACAGC-3′, STAT2-R, 5′-AGCCTTGGAATCATCACTCCC-3′; IFNAR2-F, 5′-TCATGGTGTATATCAGCCTCGT-3′, IFNAR2-R, 5′-AGTTGGTACAATGGAGTGGTTTT-3′; IFNGR1-F, 5′-TCTTTGGGTCAGAGTTAAAGCCA-3′, IFNGR1-R, 5′-TTCCATCTCGGCATACAGCAA-3′; CXCL1-F, 5′-GCCAGTGCTTGCAGACCCT-3′, CXCL1-R, 5′-GGCTATGACTTCGGTTTGGG-3′; CXCL2-F, 5′-CAAACCGAAGTCATAGCCAC-3′ and R, 5′-TCTGGTCAGTTGGATTTGCC-3′; CXCL5-F, 5′-AGCTGCGTTGCGTTTGTTTAC-3′, CXCL5-R, 5′-TGGCGAACACTTGCAGATTAC-3′; CXCL8-F, 5′-TCTGTCTGGACCCCAAGGAA-3′, CXCL8-R, 5′-GCATCTGGCAACCCTACAACA-3′; CXCL12-F, 5′-ATTCTCAACACTCCAAACTGTGC-3′, CXCL12-R, 5′-ACTTTAGCTTCGGGTCAATGC-3′; CCL2-F, 5′-CAGCCAGATGCAATCAATGCC-3′, CCL2-R, 5′-TGGAATCCTGAACCCACTTCT-3′; CCL28-F, 5′-TGCACGGAGGTTTCACATCAT-3′, CCL28-R, 5′-TTGGCAGCTTGCACTTTCATC-3′; IL33-F, 5′-GTGACGGTGTTGATGGTAAGAT-3′ and R, 5′-AGCTCCACAGAGTGTTCCTTG-3′; IL34-F, 5′-CCTGGCTGCGCTATCTTGG-3′, IL34-R, 5′-AGTGTTTCATGTACTGAAGTCGG-3′; LIF-F, 5′-CCAACGTGACGGACTTCCC-3′, LIF-R, 5′-TACACGACTATGCGGTACAGC-3′; LIFR-F, 5′-TGGAACGACAGGGGTTCAGT-3′, LIFR-R, 5′-GAGTTGTGTTGTGGGTCACTAA-3′; TSLP-F, 5′-ATGTTCGCCATGAAAACTAAGGC-3′, TSLP-R, 5′-GCGACGCCACAATCCTTGTA-3′; GDF5-F, 5′-GCTGGGAGGTGTTCGACATC-3′, GDF5-R, 5′-CACGGTCTTATCGTCCTGGC-3′; FAS-F, 5′-AGATTGTGTGATGAAGGACATGG-3′, FAS-R, 5′-TGTTGCTGGTGAGTGTGCATT-3′; OSMR-F, 5′-AATGTCAGTGAAGGCATGAAAGG-3′, OSMR-R, 5′-GAAGGTTGTTTAGACCACCCC-3′; COL4A5-F, 5′-TGGACAGGATGGATTGCCAG-3′, COL4A5-R, 5′-GGGGACCTCTTTCACCCTTAAAA-3′; COL6A6-F, 5′-TTCAACATTGCTCCCCATAAGG-3′ and R, 5′-GTGTTTGTATTCCCACCCATCT-3′; DST-F, 5′-CTACCAGCACTCGAACCAGTC-3′, DST-R, 5′-GCCGAAGCTAATGCAAGAGTTG-3′; MMP3-F, 5′-CTGGACTCCGACACTCTGGA-3′, MMP3-R, 5′-CAGGAAAGGTTCTGAAGTGACC-3′.

Cell transfection, IFNα, IFNβ, IL-4 and/or IL-13 treatment, cell counting Kit-8 (CCK-8) assay, EdU Cell Proliferation assay and cell cycle assay

Cells were cultured until they reached 70% confluence and were then transfected with 100 nM POSTN or IL-4R siRNAs at 37°C for 24 h using Lipofectamine® RNAiMAX Transfection Reagent (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. Cells were transfected with siRNAs for 24 h prior to subsequent experiments. Cells were treated with recombinant human IFNs or IL-4/IL-13. For IFNα and IFNβ treatment, cells were incubated with recombinant human IFNα (cat. no. 300-02AA; 50 ng/ml) and IFNβ (cat. no. 300-02BC; 50 ng/ml) (both from PeproTech; Thermo Fisher Scientific, Inc.) at 37°C for 15 min when cells reached 90% confluence, control cells were treated with PBS under the same conditions. For IL-4 and/or IL-13 treatment, cells were incubated with recombinant human IL-4 (cat. no. 200-04; 50 ng/ml) and/or IL-13 (cat. no. 200-13; 50 ng/ml) (both from PeproTech; Thermo Fisher Scientific, Inc.) at 37°C for 24 h when cells reaching 90% confluent, control cells were treated with PBS under the same conditions. Following the manufacturer's instructions, the Cell Counting Kit-8 (CCK-8) assay (cat. no. A311-01; Vazyme) was performed to evaluate cell proliferation rates. Cells were seeded on a 96-well plates with a density of 1×104 KFs/well for the POSTN knockdown experiment. After siRNA transfection for 24 h at 37°C, CCK-8 reagent was added to the KFs and incubated for 2 h at 37°C. Finally, a microplate reader (Infinite 200PRO; Tecan Group, Ltd.) was used to measure the absorbance at 450 nm. EdU Cell Proliferation assay was performed using an EdU Cell Proliferation Kit with Alexa Fluor 555 according to the manufacturer's instruction (cat. no. MA0425-1; Dalian Meilun Biology Technology Co., Ltd.). For nuclear staining, cells were counterstained with Hoechst at room temperature for 10 min and confocal microscopy was used to acquire images. The cell cycle assay was performed as described previously (31). Briefly, trypsinised cells were fixed in pre-cooled 70% ethanol at 4°C for 30 min. Following incubation in 50 µg/ml propidium iodide solution in the presence of RNase for 30 min at room temperature, the cells were analyzed using a flow cytometry instrument (FACS Calibur; Becton Dickinson, Franklin, NJ), and the data were processed with the MODFIT software (Verity Software House, Topsham, ME). At least three independent experiments were carried out for each treatment.

RNA sequencing (RNA-seq) and data analysis

Total RNA was extracted from si-POSTN- and si-CTL-transfected KFs using TRI Reagent® (cat. no. T9424; MilliporeSigma). Purified RNA was quantified and qualified, before library quality was assessed on an Agilent 2100 Bioanalyzer (Agilent Technologies, Inc.). The cDNA library was constructed using a Fast RNA-seq Lib Prep kit V2 (cat. no. RK20306; ABclonal) and the library was sequenced using Illumina Novaseq 6000 platform (Illumina, Inc.) with the Novaseq 6000 S4 reagent Kit (cat. no. 20027466; Illumina, Inc.) by Novogene Co., Ltd. Sequencing was carried out on the Illumina platform using paired-end sequencing with a read length of 150 bp. The final library was quantified using a Qubit 4.0 Fluorometer (Thermo Fisher Scientific, Inc.) and the final concentration was adjusted to 1.5 nM. The original expression matrix was normalized and analyzed using HiSaT2 (v2.0.5; daehwankimlab.github.io/hisat2/) and featureCounts (v1.5.0-p3; subread.sourceforge.net/). DEseq2 software (v1.20.0; http://bioconductor.org/packages/release/bioc/html/DESeq2.html) was used to detect the DEGs between the two groups; DEGs with |log2FC|>1 and FDR<0.05 were considered significant. Visualization of a boxplot, heatmap and volcano plot, as well as gene set enrichment analysis (GSEA) and Reactome analysis were performed using R software (v4.3.2; http://www.r-project.org/).

Statistical analysis

All experiments were repeated at least three times unless otherwise stated. Data are presented as the mean ± SEM. Statistical analysis was conducted using GraphPad Prism 9.0 software (Dotmatics). For two-group comparisons, unpaired Student's t-test was applied, while one-way ANOVA followed by Tukey's post hoc test was employed for multiple group comparisons. P<0.05 was considered to indicate a statistically significant difference.

Results

POSTN is upregulated in KFs

GSE7890, GSE44270 and GSE145725 datasets, encompassing the expression profiles of genes in control fibroblasts from healthy individuals and KFs from patients, were downloaded from the GEO database. DEGs were defined as those with significant differential expression between KFs and NFs, with P<0.05 and |log2FC|>1. Upregulated DEGs were considered those which were highly expressed in KFs compared with NFs. Notably, a total of 82,379 and 288 upregulated DEGs were identified in the GSE44270, GSE7890 and GSE145725 datasets, respectively (Fig. 1A). Additionally, a Venn diagram revealed that five upregulated DEGs overlapped among the aforementioned three datasets. Specifically, POSTN, CCDC80, RGS4, EBF1 and TBX15 were upregulated in KFs compared with NFs. Among these five genes, POSTN was more significantly upregulated in KFs (Fig. 1B). Previous studies have indicated that POSTN is markedly dysregulated in keloids, thus supporting its potential role in the pathogenesis of keloid (32,33). Therefore, POSTN was identified as a gene that could be potentially associated with keloid development. Subsequently, to detect the expression profile of POSTN in KFs and NFs derived from patients recruited for the present study, RT-qPCR, western blotting and immunofluorescence assays were performed. Consistent with the results of bioinformatics analysis, the mRNA expression levels of POSTN were significantly higher in KFs compared with those in NFs (Fig. 1C). Furthermore, western blot analysis (Fig. 1D) and immunofluorescence staining assays (Fig. S1) demonstrated that the protein expression levels of POSTN were markedly increased in KFs compared with those in NFs. Based on these results, it was hypothesized that POSTN could represent a potential modulator of keloid pathogenesis.

Figure 1. POSTN is upregulated in KFs. (A) Venn plot of common upregulated differentially expressed genes identified from three independent datasets. (B) Relative mRNA expression levels of POSTN in the three datasets. (C) Relative mRNA expression levels of POSTN in KFs (n=3) and NFs derived from healthy donors (n=3). (D) Representative western blotting and semi-quantification of POSTN protein expression in KFs and NFs (n=3/group). Data are presented as the mean ± SEM. ***P<0.001, *P<0.05. KFs, keloid fibroblasts; NFs, normal fibroblasts; POSTN, periostin.

POSTN knockdown exerts limited effects on the proliferation of KFs

To assess the regulatory role of POSTN in the pathogenesis of keloids, the endogenous expression of POSTN was silenced in KFs using siRNA (si-POSTN) technology. As shown in Fig. 2A, POSTN expression was significantly decreased in KFs transfected with si-POSTN compared with that in the control cells; the knockdown efficiencies of si-POSTN#1 and si-POSTN#2 were 93 and 92%, respectively (Fig. 2A). Since the knockdown efficiency of si-POSTN#1 is higher than that of si-POSTN#2, we selected si-POSTN#1 for subsequent experiments unless otherwise noted. The CCK-8 assay demonstrated that POSTN silencing exerted no effect on the proliferation of KFs (Fig. 2B). Additionally, the EdU incorporation assay revealed that cell proliferation was comparable between KFs transfected with si-POSTN and si-CTL (Fig. 2C), thus suggesting that POSTN silencing had no notable effect on fibroblast proliferation. Furthermore, flow cytometry was utilized to examine whether POSTN knockdown could affect cell cycle distribution in KFs. As shown in Fig. 2D, no changes were detected in the proportion of cells in the G0/G1 or S phases of the cell cycle between the si-POSTN and si-CTL groups, thus indicating that POSTN knockdown did not affect the cell cycle distribution of KFs. To further investigate the effect of POSTN on KF proliferation, the expression levels of cell proliferation- and cell cycle-related genes were detected in si-POSTN- and si-CTL-transfected KFs. In line with the aforementioned results, the RT-qPCR analysis results showed that POSTN knockdown did not alter the expression levels of the majority of cell proliferation- and cell cycle-related genes in KFs, since the expression levels of CCNA1/A2/B1/B2/C/D3/E2 and CDK1/2/3/4/5 were comparable between the si-POSTN and si-CTL groups (Fig. 2E and F). Collectively, these findings suggested that POSTN knockdown exerted limited effects on the proliferation of KFs.

Figure 2. Knockdown of POSTN shows limited effects on the proliferation of KFs. (A) mRNA expression levels of POSTN in KFs transfected with si-POSTN; mRNA levels were analyzed using reverse transcription-quantitative PCR (n=4), whereas protein levels were detected by western blotting. (B) CCK-8 assay revealed that the proliferation was comparable between KFs transfected with si-POSTN and si-CTL (n=4). (C) EdU staining of KFs transfected with si-POSTN/si-CTL and their relative proliferation rate (n=5). Scale bar: 50 µm. (D) Cell cycle progression was assessed using flow cytometry after siRNA transfection, and the percentages of cells in different phases were calculated (n=3). (E) mRNA expression levels of cell cycle-related genes in KFs transfected with si-POSTN/si-CTL (n=3). (F) mRNA expression levels of cell growth-related genes in KFs transfected with si-POSTN/si-CTL (n=3). Data are presented as the mean ± SEM. ***P<0.001, NS, not significant; POSTN, periostin; si, small interfering.

POSTN knockdown inhibits the JAK/STAT signaling and the expression of proinflammatory factors in KFs

To further explore the role of POSTN in KFs, cells transfected with si-POSTN and si-CTL were subjected to RNA-seq. Notably, a total of 3,893 DEGs were identified between the two groups, including 1,856 upregulated and 2,037 downregulated genes (Fig. S2). Subsequently, GSEA was performed to investigate the potential biological functions of these DEGs, which could be involved in keloid pathogenesis. The results showed that the majority of genes were enriched in ‘JAK/STAT signaling pathway’, ‘cytokine-cytokine receptor interaction’, ‘NF-kappa B’, ‘natural killer cell-mediated cytotoxicity’, ‘antigen processing’ and ‘TNF’ signaling pathways (Fig. S3). Emerging evidence has suggested that the JAK/STAT signaling pathway and proinflammatory cytokines, such as CXCL12, C-C motif chemokine ligand (CCL)2 and thymic stromal lymphopoietin (TSLP), serve crucial roles in the pathogenesis of keloids (19,20,34). Therefore, the current study mainly focused on the regulatory effect of POSTN on the ‘JAK/STAT signaling pathway’ (Fig. 3A) and ‘cytokine-cytokine receptor interaction’ (Fig. 4A) in KFs. Heatmap analysis showed that the expression of the JAK/STAT pathway-related DEGs was evidently altered in si-POSTN-transfected KFs compared with in si-CTL-transfected KFs (Fig. 3B). In addition, the expression levels of JAK/STAT pathway-related genes, including JAK2, STAT1/2, IFNAR2 and IFNGR1, were significantly decreased in POSTN-depleted KFs (Fig. 3C, marked in red boxes). Furthermore, POSTN silencing markedly reduced the protein expression levels of p-STAT1 (Tyr701) and p-STAT1 (Ser727) in KFs, regardless of IFN-α/β treatment (Fig. 3D). Collectively, these results suggested that POSTN may be involved in the activation of JAK/STAT signaling in KFs.

Figure 3. Knockdown of POSTN blocks the JAK-STAT signaling pathway in KFs. (A) JAK-STAT pathway was revealed to be associated with POSTN knockdown in KFs by gene set enrichment analysis. (B) Heatmap showing DEGs related to the JAK-STAT pathway in si-POSTN and si-CTL KFs. Three independent RNA-sequencing experiments are shown in each group. (C) Reverse transcription-quantitative PCR analysis was performed to validate the DEGs related to the JAK-STAT pathway in si-POSTN- and si-CTL-transfected KFs (n=3). (D) Representative western blotting and semi-quantification of the protein levels of p-STAT1 (Tyr701) and p-STAT1 (Ser727) in KFs transfected with si-POSTN and si-CTL in the presence or absence of IFNα +β treatment (n=3). Data are presented as the mean ± SEM. ***P<0.001, **P<0.01, *P<0.05. DEGs, differentially expressed genes; KFs, keloid fibroblasts; NS, not significant; p-, phosphorylated; POSTN, periostin; si, small interfering.

Figure 4. Knockdown of POSTN decreases the expression of proinflammatory factors in KFs. (A) Cytokine-cytokine receptor interaction pathway was revealed to be associated with POSTN knockdown in KFs by gene set enrichment analysis. (B) Heatmap showing DEGs related to cytokine-cytokine receptor interaction in si-POSTN and si-CTL KFs. The red box highlights the downregulated expression of IL-4R in si-POSTN KFs compared with si-CTL KFs. Three independent RNA-sequencing experiments are shown in each group. (C) Reverse transcription-quantitative PCR was performed to validate the DEGs related to cytokine-cytokine receptor interaction in si-POSTN- and si-CTL-transfected KFs (n=4). KF#1 and KF#2 indicate KFs derived from two different patients. ***P<0.001, **P<0.01, *P<0.05. DEGs, differentially expressed genes; NES, normalized enrichment score; IL-4R, IL-4 receptor; KFs, keloid fibroblasts; POSTN, periostin; si, small interfering.

The expression pattern of DEGs associated with ‘cytokine-cytokine receptor interaction’ is presented in a heatmap in Fig. 4B. To further validate the significance of POSTN expression in the production of proinflammatory cytokines/chemokines, POSTN was silenced in KFs derived from two different patients with keloids using two sets of specific si-POSTN constructs. RT-qPCR indicated mRNA expression levels of keloid inflammation-related DEGs, including CXCL1/CXCL2/CXCL5/CXCL8/CXCL12, CCL2/CCL28, IL33/IL34, LIF/LIFR, TSLP, GDF5, FAS and OSMR, were similar between the two patients (Fig. 4C). Notably, the majority of DEGs were downregulated in POSTN-depleted KFs compared with in control cells, thus indicating that POSTN could modulate the expression of inflammatory cytokines in KFs. Taken together, the aforementioned findings suggested that POSTN could serve a key role in maintaining keloid inflammation.

IL-4 and/or IL-13 upregulates POSTN in KFs

Previous studies have demonstrated that IL-4 and IL-13, two signature type 2 cytokines, can induce the expression of POSTN in inflammation-related diseases such as atopic dermatitis and bronchial asthma (35–37). Notably, another study showed that Th2 markers are significantly upregulated in keloid lesions compared with in normal skin, thus suggesting that keloid pathogenesis could be triggered by Th2 cytokines (32). Therefore, the present study explored whether IL-4 and/or IL-13 could induce POSTN expression in KFs. As shown in Fig. 5A and B, treatment of primary fibroblasts with IL-4 and/or IL-13 notably induced POSTN expression, at both the mRNA (Fig. 5A) and protein levels (Fig. 5B). Additionally, the immunofluorescence staining results further verified that IL-4 and IL-13 could induce the protein expression of POSTN in KFs (Fig. 5C). These results suggested that IL-4 and/or IL-13 could upregulate POSTN in KFs.

Figure 5. Expression of POSTN is upregulated by IL-4 and/or IL-13 in KFs. (A) Reverse transcription-quantitative PCR results revealed that IL-4 and IL-13 increased the mRNA expression levels of POSTN in NFs and KFs derived from different donors (n=4/group). (B) Western blotting showed that IL-4 and/or IL-13 induced the protein expression of POSTN in KFs derived from two different patients with keloids. (C) Immunofluorescence staining of the protein expression of POSTN in KFs in the presence or absence of IL-4 or IL-13 stimulation. Data are presented as the mean ± SEM. ***P<0.001 vs. CTL. KFs, keloid fibroblasts; NFs, normal fibroblasts; POSTN, periostin

IL-4R is essential for IL-4/IL-13-induced POSTN upregulation in KFs

Notably, Reactome enrichment analysis revealed that the ‘IL-4 and IL-13 signaling’ pathway was significantly enriched in DEGs between POSTN-depleted and control KFs (Fig. 6A, highlighted in red). This finding was consistent with previous reports suggesting that POSTN is strongly involved in Th2 cell-mediated inflammation (38,39). Notably, the mRNA (Fig. 6B) and protein (Fig. 6C) expression levels of IL-4R, a cytokine receptor for both IL-4 and IL-13, were notably inhibited in POSTN-depleted KFs derived from three patients with keloids. Additionally, protein expression levels of IL-4R and POSTN were higher in KFs compared with NFs, suggesting a positive association of the expression between these two proteins in primary fibroblasts (Fig. S4A). The aforementioned finding indicated that POSTN could positively regulate IL-4R expression in KFs. Subsequently, to verify the role of Th2 signaling in inducing POSTN expression, IL-4R was silenced in KFs using specific siRNAs (Fig. S4B) and POSTN expression was then detected in the presence or absence of IL-4/IL-13 treatment. As shown in Fig. 6D, treatment of KFs with human recombinant IL-4/IL-13 upregulated POSTN, while its mRNA levels were notably reduced in IL-4R-depleted KFs induced by IL-4/IL-13. In line with this finding, western blot analysis showed that stimulation of KFs with IL-4/IL-13 markedly increased the protein levels of POSTN, which were reduced following IL-4R knockdown (Fig. 6E). Subsequent immunofluorescence analysis further verified that the protein levels of POSTN were reduced in IL-4R-depleted KFs compared with control KFs in the presence of IL-4/IL-13 stimulation (Fig. 6F). Collectively, these results indicated that IL-4R, which could be positively regulated by POSTN, was necessary for the IL-4/IL-13-induced upregulation of POSTN in KFs, thus supporting the presence of a positive feedback loop between POSTN and Th2 signaling.

Figure 6. IL-4R is positively regulated by POSTN and essential for IL-4/IL-13-induced POSTN upregulation in KFs. (A) A bubble chart showing the top 20 signaling pathways associated with POSTN in KFs based on Reactome analysis. The IL-4 and 13 signaling pathway is highlighted in red. (B) RT-qPCR results showing the mRNA expression levels of POSTN and IL-4R in KFs transfected with si-POSTN/si-CTL (n=4). (C) Western blotting of the protein expression of POSTN and IL-4R in KFs transfected with si-POSTN/si-CTL. (D) RT-qPCR results showing the mRNA expression levels of POSTN in KFs transfected with si-IL-4R/si-CTL, in the presence or absence of IL-4/IL-13 treatment (n=4). (E) Representative blots showing protein expression of POSTN in KFs transfected with si-IL-4R/si-CTL, in the presence or absence of IL-4/IL-13 treatment. (F) Immunofluorescence staining showing protein expression of POSTN in KFs transfected with si-IL-4R/si-CTL, in the presence or absence of IL-4/IL-13 treatment. Data are presented as the mean ± SEM. ***P<0.001, *P<0.05. IL-4R, IL-4 receptor; KFs, keloid fibroblasts; POSTN, periostin; RT-qPCR, reverse transcription-quantitative PCR; si, small interfering.

POSTN knockdown inhibits the expression of genes associated with extracellular matrix formation in KFs

Reactome enrichment analysis also revealed significant enrichment of extracellular matrix organization-related signaling in DEGs between si-POSTN- and si-CTL-transfected KFs (Fig. 6A). This was further supported by a heatmap analysis (Fig. S5A), which highlighted the significance of POSTN in regulating extracellular matrix remodeling. Additionally, RT-qPCR analysis verified that the mRNA expression levels of extracellular matrix-related DEGs, including COL4A5, COL6A6, DST, and MMP3, were significantly decreased in POSTN-depleted KFs (Fig. S5B). Overall, POSTN could positively regulate extracellular matrix formation in KFs, a process involved in the pathogenesis of keloids.

Discussion

Keloids are a chronic and inflammatory skin condition, characterized by fibroproliferative overgrowth of the skin (1). Keloids are considered to result from abnormal wound healing in susceptible individuals following trauma, inflammation, surgery and burns (2). Keloids are not life-threatening; however, they can affect the quality of life of patients due to pain and itching, and they can easily relapse even if surgically removed (5,6,40). The precise mechanism underlying keloid pathogenesis remains unknown; therefore, identifying the pathological mechanisms of keloids is of importance for the development of novel and efficient therapeutic strategies.

Previous microarray analyses have identified several genetic alterations associated with keloid pathogenesis, such as HIF-1α (11), Wnt signaling (22) and extracellular matrix (23). However, the sample sizes in the aforementioned studies were relatively small and the heterogeneity was high, possibly resulting in biased results. To accurately screen genes involved in keloid formation, three datasets, namely GSE7890, GSE44270 and GSE145725, from the GEO database were compared in the current study to identify common DEGs between KFs and NFs. By comprehensively analyzing the gene expression profiles of these three GEO datasets, a total of five common DEGs overlapping between the three datasets were identified, namely POSTN, CCDC80, RGS4, EBF1 and TBX15, which were all upregulated in KFs compared with NFs. Among these five genes, it has been reported that POSTN is markedly dysregulated in keloids, thus indicating its possible role in keloid pathogenesis (32,33). Consistently, the mRNA and protein levels of POSTN were also elevated in KFs isolated from patients with keloids. To the best of our knowledge, only a few studies have systematically investigated the role of POSTN in keloid formation (32,33,41). Therefore, POSTN, a notable candidate gene associated with keloid development, was selected for further investigation.

POSTN, a matricellular protein that belongs to the fasciclin family, serves key roles in skin development, skin remodeling/repair and skin-related diseases (42–44). A previous study demonstrated that POSTN is substantially expressed in the dermis of patients with psoriasis, eventually leading to epidermal hyperplasia and the pathogenesis of psoriasis (36). Another study on atopic dermatitis also suggested that POSTN may be involved in the pathogenesis of allergic skin inflammation by inducing TSLP production and disrupting barrier function (37). However, the role of POSTN in keloids has not been systematically investigated. In the present study, RNA interference technology was applied to evaluate the role of POSTN in KFs, which are the most common cells involved in keloid formation. Notably, POSTN knockdown had limited effects on the proliferation of KFs, as evidenced by the comparable cell counts, EdU incorporation rates and cell cycle distribution between KFs transfected with si-POSTN and those transfected with si-CTL. Keloid is featured by unrestricted fibroproliferative overgrowths of the skin (45). The current study revealed that POSTN silencing could not affect the growth and proliferation of KFs; this could be due to the limited sensitivity of the examined fibroblasts to growth inhibition. Additionally, certain types of fibroblasts may exert distinct biological functions. It would therefore be of interest to investigate other functions of POSTN, such as its involvement in fibrosis, extracellular matrix production and inflammation, within these particular types of KFs.

To further explore the role of POSTN in keloid pathogenesis, KFs transfected with si-POSTN or si-CTL were subjected to RNA-seq. The results demonstrated that the JAK/STAT pathway, which has previously been involved in the pathogenesis of keloids (46), was markedly affected by POSTN knockdown in KFs. Additionally, GSEA further revealed that the cytokine-cytokine receptor interaction pathway was significantly altered by POSTN silencing. These findings were further validated using two sets of si-POSTN constructs in KFs derived from two different patients with keloids. Specifically, the expression levels of genes associated with keloid inflammation, including CXCL1/CXCL2/CXCL5/CXCL8/CXCL12, CCL2/CCL28, IL33/IL34, TSLP, GDF5 and OSMR, were markedly decreased in POSTN-deficient KFs, when compared with control cells. Based on these observations, it was hypothesized that POSTN could be involved in the pathogenesis of keloids, at least in part, by regulating the inflammatory response within KFs.

It has been well established that chronic inflammation and immune responses serve key roles in keloid formation (14–16). Several studies have reported the upregulation of proinflammatory cytokines in keloid lesions, including TSLP, IL-4, IL-13, CXCL8, CXCL12 and CCL2 (4–7,19,34,47,48). Furthermore, Wu et al (32) reported that keloid tissues are characterized by an inflammatory microenvironment, including activated Th2 (IL-4R and TSLP), Th1 (CXCL10/11) and Th17/Th22 (CCL20) pathways, underscoring their roles in keloid pathogenesis. Among the aforementioned cytokines, CXCL12 is considered a critical mediator, and a previous study showed that CXCL12 is upregulated in keloid scars compared with normal skin (19). However, the underlying mechanism remains unclear. Herein, the current study demonstrated that POSTN knockdown could downregulate CXCL12 in KFs, thus providing a potential explanation for its dysregulation in keloids. Consequently, CXCL12 accumulation could be involved in keloid pathology via the CXCL12/CXCR4 axis, thus promoting cell chemotaxis and the recruitment of inflammatory cells into keloid tissues, eventually sustaining chronic inflammation (20). Notably, TSLP, which was also shown to be positively regulated by POSTN in KFs in the present study, is known to induce the infiltration of CXCR4+ fibrocytes by upregulating CXCL12 in keloid scars (19). A previous study also revealed the hyperactivation of MCP-1 (CCL2) in keloid lesions (49), which is in consistent with the findings of the current study. In addition, elevated CCL2 release in keloid tissues could augment fibroblast proliferation, thus initiating keloid development (49). The aforementioned findings collectively support the role of inflammatory cytokines and chemokines in promoting the pathogenesis of keloids.

The current study showed that POSTN expression was increased in KFs, while the expression of inflammatory factors was reduced in POSTN-depleted KFs, thus indicating a positive association between POSTN and inflammatory signaling in keloid pathogenesis. Therefore, targeting the inflammatory microenvironment by reducing the expression of these cytokines and/or inhibiting POSTN, could reverse the inflammatory process and offer a potential treatment strategy for keloid therapy. However, due to the lack of a robust in vivo model for keloids, it remains challenging to definitively conclude that POSTN could be directly involved in keloid formation through regulating the inflammatory responses. Therefore, the establishment of an effective animal model for keloids could help in addressing this limitation. It has been reported that IL-4 and IL-13 generation is enhanced in keloid lesions compared with in normal skin (50). However, herein, abnormal expression of IL-4/IL-13 in POSTN-deficient KFs was not observed (data not shown), possibly due to the fact that IL-4/IL-13 are primarily secreted by immune cells rather than fibroblasts. In terms of scar formation, the primary cellular sources of IL-4 and IL-13 are CD4+ Th2 cells, basophils, mast cells, eosinophils, natural killer T cells and type 2 innate lymphoid cells (51,52). These cytokines can promote collagen deposition, thus contributing to the pathogenesis of scars. Additionally, IL-4 and IL-13 can enhance POSTN expression in fibroblasts, thus facilitating scar formation and progression (53). Consistent with this, Nangole and Agak (54) showed that the plasma levels of IL-4 and IL-13 are significantly elevated in patients with keloids compared with in healthy subjects, supporting the notion that IL-4/IL-13 signaling and immune responses could serve key roles in keloid formation.

Emerging evidence has suggested that IL-4 and/or IL-13, two signatures of type 2 cytokines, can induce the expression of POSTN in allergic diseases (35–37). The present study demonstrated that IL-4 and/or IL-13 upregulated POSTN in KFs, thus indicating that Th2 signals could upregulate POSTN in keloids. In line with this result, a previous study showed that POSTN is expressed in Th2-associated fibroblasts in atopic dermatitis (55), further confirming a crosstalk between Th2 immunity and keloid pathogenesis. Notably, the IL-4 and IL-13 signaling pathway was significantly enriched in POSTN-related DEGs in KFs in the present study. Additionally, POSTN knockdown could inhibit the expression of IL-4R, a cytokine receptor for both IL-4 and IL-13, thus suggesting that POSTN may positively regulate IL-4R expression in KFs. Notably, IL-4R was necessary for IL-4/IL-13-induced POSTN upregulation, underlying a positive feedback loop between POSTN and Th2 signals in keloids. Previous studies have also revealed that targeting IL-4Ra using dupilumab results in both atopic dermatitis improvement and shrinkage of keloids (50,56). Moreover, IL-4/IL-13 signaling is activated and the mRNA levels of Th2-related factors have been shown to be increased in keloid lesions (21,56). In addition, it has been reported that keloids and other Th2-skewed diseases, including atopic dermatitis (57) and asthma (58), share several overlapped clinical features. RNA-seq data have also linked keloids to a Th2-skewed immune profile, thus indicating a close association between keloid pathogenesis and Th2-related gene expression signatures (32). Collectively, IL-4 and/or IL-13 could induce POSTN expression, while, in turn, the elevated POSTN expression could promote IL-4R expression, thus suggesting a positive feedback loop and a novel mechanism that could be involved in the pathogenesis of keloids. Due to the lack of an effective mouse model for keloids, the regulatory effect of IL-4/IL-13 on POSTN expression could not be investigated in vivo. This represents a limitation of the present study. Nonetheless, an in vivo model for keloid research should be established in the future, potentially through an organoid-based approach, to further explore the interaction between POSTN and IL-4/IL-13 signaling in vivo.

In conclusion, the results of the present study demonstrated that POSTN was upregulated by IL-4 and/or IL-13 in KFs. Functionally, POSTN could promote inflammation in KFs without affecting their proliferation. Furthermore, POSTN could positively regulate IL-4R expression, which may be essential for IL-4/IL-13-induced POSTN upregulation in KFs, thus suggesting that a positive feedback loop between POSTN and Th2 signaling could be involved in keloid development. However, further studies are needed to verify the clinical relevance of POSTN in keloid pathogenesis, and to assess the therapeutic potential of targeting POSTN and/or Th2 signaling in keloid treatment.

Supplementary Material

Supporting Data

Acknowledgements

Not applicable.

Funding

The present study was supported by grants from the National Natural Science Foundation of China (grant nos. 82103726 and 82203900), the Guangdong Basic and Applied Basic Research Foundation (grant nos. 2023A1515010575 and 2025A1515010947), the Shenzhen Science and Technology Program (grant nos. JCYJ20210324110008023 and JCYJ20230807095809019), the Shenzhen Sanming Project (grant no. SZSM202311029), the Shenzhen Key Medical Discipline Construction Fund (grant no. SZXK040) and the Shenzhen High-level Hospital Construction Fund and Peking University Shenzhen Hospital Scientific Research Fund (grant no. KYQD2024378).

Availability of data and materials

The sequencing data generated in the present study may be found in the NCBI Gene Expression Omnibus under accession number GSE307210 or at the following URL: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE307210. The other data generated in the present study may be requested from the corresponding author.

Authors' contributions

BJ, FZ and CH contributed to the experimental design, performance, data collection/analysis and manuscript preparation. XL, KZ, JG and JW helped to perform in vitro experiments. WZ, YZ and BY contributed to data analysis and discussed the results. BY and CH supervised the study. WZ, BY and CH contributed to funding acquisition. FZ and CH confirm the authenticity of all the raw data. All authors read and approved the final manuscript.

Ethics approval and consent to participate

The studies involving human participants were reviewed and approved by the Ethics Committee from the Peking University Shenzhen Hospital (protocol code 2022-070). Written informed consent to participate in the study was provided by the participants.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References