Gbp3 is associated with the progression of lupus nephritis by regulating cell proliferation, inflammation and pyroptosis

Abstract

Lupus nephritis (LN) is a major cause death in patients with systemic lupus erythematosus. We aimed to find the differentially expressed genes (DEGs) in LN and confirm the regulatory mechanism on LN. The mouse model of LN was constructed by subcutaneous injection of pristane. RNA-seq screened 392 up-regulated and 447 down-regulated DEGs in LN mouse model, and KEGG analysis found that the top 20 DEGs were enriched in arachidonic acid metabolism, tryptophan metabolism, etc. The hub genes, Kynu, Spidr, Gbp3, Cbr1, Cyp4b1, and Cndp2 were identified, in which Gbp3 was selected for following study. Afterwards, the function of Gbp3 on the proliferation, inflammation, and pyroptosis of LN was verified by CCK-8, ELISA, and WB in vitro. The results demonstrated that si-Gbp3 promoted cell proliferation and inhibited the levels of inflammatory factors (IL-1β, TNF-α and IL-8) and pyroptosis-related proteins (GSDMD, Caspase-1 and NLRP3) in a cell model of LN. In constrast, Gbp3 overexpression played an opposite role. In summary, Gbp3 promoted the progression of LN via inhibiting cell proliferation and facilitating inflammation and pyroptosis.

Keywords:

• transcriptome sequencing
• Gbp3
• pyroptosis

1. Introduction

Systemic lupus erythematosus (SLE) is a chronic inflammatory disease, which is caused by autoimmune dysfunction, behaving with complex-mediated lesions of blood vessels [1]. Loss of self-tolerance and chronic inflammation are critical markers of SLE pathogenesis [2]. Lupus nephritis (LN) is a common and serious complication of SLE [3,4], occurring in up to 50% of SLE patients during the development of the disease [5]. LN is probably characterised by urinary abnormalities or more obvious performance including nephrotic syndrome or rapidly progressive renal failure [6]. LN is one of the most important predictors of mortality in SLE [7]. Although studies have reported that the mortality of LN was reduced and the disease prognosis was improved, the percentage of patients progressing to end stage renal disease has remained stably [8]. Thus, there is an essential requirement to explore the disease pathogenesis and develop a new effective treatment.

With the establishment and improvement of the novel sequencing technology, next-generation sequencing (NGS) becomes more and more prevailing in the genomics researches. It is a high-throughput method allowing massive parallel sequencing consisting of simultaneous sequencing of multiple targeted genomic regions in multiple samples, and requests low input of DNA/RNA [9]. NGS enhances the accuracy and accelerates the speed of the diagnosis [10]. NGS has also been applied in the revelation of molecular mechanism and screening of potential therapeutic target of diseases. For examples, genetic mutations can be identified using NGS and used to develop strategies that will selectively kill cancerous cells in patients with pancreatic ductal adenocarcinoma [11]. NGS vastly improves the understanding of molecular mechanism of hematological malignancies [12]. It is also used to identify useful specific biomarker for the detection of LN among lupus patients, ensuring timely treatment [13]. Our study intended to screen EDGs in LN and found the candidate gene Gbp3.

The guanylate binding proteins (Gbps) are members of 65–73 kDa large GTPases. It is identified to be composed of seven Gbps (Gbp1-7) in humans [14]. GBPs is thought to contribute to host innate immunity because they mediate resistance against invading pathogens, with Gbp1, Gbp2, and Gbp5 being the most studied members [15,16]. Previous study has been reported that Gbps influence inflammasome responses [17]. Gbp1 activates caspase-4 to trigger the release of the pro-inflammatory cytokine, IL18, and lead to inflammatory pyroptosis and cell death [18]. In our study, Gbp3 showed differential expression in LN mice via transcriptome sequencing, which caught our attention. Gbp3, as a member of the family, is considered to originate from the Gbp1 gene [19]. At present, the role of Gbp3 has been extensively studied in glioma [20,21]. We pursue that it is worth figuring out the role of Gbp3 in the occurrence of LN and the influence on pyroptosis.

The development of inflammatory diseases is usually accompanied with the activation of inflammasomes. The NLRP3 inflammasome has been linked with the pathogenesis and progression of some kidney diseases including LN [22], leading to pyroptosis and apoptosis [23]. Pyroptosis, a nonapoptotic form of programmed cell death, is closely related to the occurrence of kidney diseases [24]. It is worth noting to explore which genes regulate the pyroptosis in LN. In our study, we constructed a mouse model of LN and found many differentially expressed gene by transcriptome sequencing. Besides, we verified the function of Gbp3 in LN and explored its regulatory effect on cell proliferation, inflammation and pyroptosis, which may strengthen the understanding of pathogenesis of LN and provide a new mind for the therapy of the disease.

2. Methods

2.1. Animals and study design

Seven-week-old BALB/c mice fed in SPF environment were used to establish the LN model. They were purchased from Antaik Biotechnology (Beijing, China). A total of 12 mice were randomly divided in two groups: Control and model of LN (Model). The Model mice were given pristane (0.5 mL) (Sigma-Aldrich, St Louis, MO, USA) by intraperitoneal injection to induce LN, and the Control mice were injected with an equal volume of normal saline. Four weeks later, mice were euthanized by intraperitoneal injection of sodium pentobarbital (50 mg/kg), and blood, urine and kidney tissue samples were obtained from two groups.

2.2. Haematoxylin-eosin staining

The glomerulus tissue sections were prepared for Haematoxylin-eosin (HE) staining by Servicebio (Wuhan, China). They were stained in Haematoxylin (Solarbio, Beijing, China) for 5 min at room temperature, followed by Eosin (Solarbio) for 1 min. Then the images were obtained under an optical microscope (Leica, Weztlar, Germany).

2.3. Transcriptome sequencing and data analysis

The RNA was obtained from kidney tissue using TRIzol (Invitrogen, Carlsbad, CA, USA). The sequencing libraries were gained with the NEBNext UltraTM RNA Library Prep Kit for Illumina® (NEB, USA). We performed the Paired-end sequencing on Illumina platform.

Differential gene screening: The differentially expressed genes (DEGs) screening conditions in Control and Model groups were as follows: the multiple expression difference (|log₂FoldChange|) was >1 and the significance (P-value) was < 0.05. The R ggplots2 software package was used to draw the volcano map of DEGs. Two-way cluster analysis was performed on the union and samples of all the two groups of differential genes using the Pheatmap software package in R language.

Gene Ontology (GO) enrichment analysis was conducted with topGO. The degree of enrichment was measured by the Rich factor, FDR value, and the number of genes enriched to this GO Term. According to the Kyoto Encyclopaedia of Genes and Genomes (KEGG) enrichment analysis of DEGs, the pathways with the smallest P-value were selected for further presentation. The protein-protein interaction (PPI) network of 392 up-regulated and 447 down-regulated genes was constructed using the Search3 tool for the Retrieval of Interacting Genes/Proteins (STRING) database.

2.4. Quantitative real-time PCR (RT-qPCR)

The total RNA of kidney tissue was extracted by Trizol (Invitrogen). The cDNA library was gained by PrimeScript^TM RT Master Mix (TaKaRa, Dalian, China). The qPCR was performed by SYBR® Green Pro Taq HS premix (Accurate Biology, Changsha, China). The relative level of RNA expression was normalised to GAPDH using the 2^−ΔΔCt method. The procedure of qPCR was 95°C, 10 min; (95°C, 10 s; 60°C, 30 s), 36 cycles. The primers were provided by GenePharma, and the sequences were shown in the Supplementary Table S2.

2.5. Cell culture

Human glomerular epithelial cell (HGEC) (2 × 10⁵/well) (ATCC cell bank, Shanghai, China) was cultured in DMEM with 10% foetal bovine serum (Gibco, Grand Island, USA). Si-RNA was designed and synthesised by GenePharma Co., Ltd (Shanghai, China). The overexpression plasmids Gbp3 (pDNA3.1-Gbp3) were acquired from Shenggong Bioengineering Company (Shanghai, China). The cell was grouped as follows: (1) Control; (2) cell was treated with LPS (10 µg/mL) (Sigma-Aldrich); (3) cell was treated with LPS and si-NC; (4) cell was treated with LPS and si-Gbp3. (5) cell was treated with LPS and ov-NC; (6) cell was treated with LPS and ov-Gbp3. The siRNA and plasmids were transfected into cell using Lipofectamine 3000 (Invitrogen) according to the instruction, and the sequences were shown in the Supplementary Table S1.

2.6. Cell counting Kit-8 (CCK-8)

HGEC was seeded into 96-well microplates with 100 μL of cell suspension (4 × 10⁵/mL). CCK-8 kit (Solarbio, Beijing, China) was used to detect the proliferation in each group at 1 d, 2 d, 3 d and 4 d after different treatments. The OD₄₅₀ was analysed using a microplate reader (Bio-Rad, Hercules, CA, USA).

2.7. Enzyme-linked immunosorbent assay

The serum of mouse or culture supernatant from HGEC were used to detect the levels of TNF-α, IL-1β and IL-8 using Enzyme-linked immunosorbent assay (ELISA) kits (Esebio, Shanghai, China). Briefly, samples were added to 96-well plates and incubated for 2 h with corresponding antibody. Then the plate was washed and incubated with affinity streptomycin-HRP for 1 h. OD₄₅₀ was measured using a microplate reader (Thermo Fisher Scientific, Pittsburgh, PA, USA). Additionally, the concentrations of dsDNA and ANA in serum were also measured using ELISA kits.

2.8. Western blot

The proteins were extracted from kidney tissue or HGEC using RIPA lysis buffer (Beyotime,Nanjing, China), and the concentrations were measured by BCA Protein Assay kit (Beyotime). The proteins were separated by 10% SDS-PAGE, electrotransferred onto PVDF membrane (Zoman, Beijing, China), and incubated with specific primary antibodies of GSDMD, Caspase-1 and NLRP3 (ZenBioScience, Chengdu, China) at 4°C overnight. After washing, membrane was subjected to appropriate secondary antibody (1:1000) (Abcam, Cambridge, England) for 2 h at room temperature. Immunodetection was performed using enhanced ECL Chemiluminescence Detection Kit (Vazyme, Nanjing, China), and the images were obtained by Quantity One (Bio-Rad, Hercules, CA, USA).

2.9. Statistical analysis

All data were processed by GraphPad Prism 8.0 software, and the measurement data were expressed in the form of mean ± standard deviation. The t-test was used for comparison between two groups, and One-Way ANOVA was used for comparison between multiple groups. Tukey's multiple comparisons test was used for pairwise comparisons after ANOVA analysis, and p < 0.05 was considered statistically significant.

3. Results

3.1. The establishment of mouse model of LN

We first constructed a mouse model of LN induced by pristane and examined changes of related factors. The levels of inflammatory factors in serum including TNF-α, IL-1β and IL-8 of Model were all higher than that of Control (p < 0.001, Figure 1(A)). The concentrations of urine creatinine (UCR) and albumin (ALB) in urine of Model were upregulated compared with Control (p < 0.001, Figure 1(B)). Then we performed HE staining with kidney tissue sections to observe the pathological features. The results showed that in Control, the glomerular structure and cell layers were clear, the cells were arranged in order, and there was no interstitial oedema and inflammatory cell infiltration. The serum levels of ANA and dsDNA in LN model were significantly increased (Figure 1(C)). Compared with control, cells had a slightly disordered arrangement with unclear layers, and there was interstitial oedema and karyopyknosis with significant increasing number of inflammatory cells (Figure 1(D)). Moreover, the expression of pyroptosis-related proteins was detected by WB. We found that the levels of NLRP3, GSDMD and Caspase-1 in Model increased in comparison with Control (p < 0.01, Figure 1(E)).

Figure 1. The verification of mouse model of LN. (A) Detection of the inflammatory factor levels in serum by ELISA. (B) Detection of UCR and ALB in urine by ELISA. (C) Levels of ANA and dsDNA in serum detected by ELISA. (D) Observation of kidney histology by HE staining. E: Detection of pyroptosis-related protein levels by WB. ***P < 0.001 vs. Control. ** P < 0.01 vs. Control.

3.2. Screening differential genes

To obtain the DEGs, we performed transcriptome sequencing in Control and Model mice. Sequencing results showed that a total of 839 genes were significantly differentially expressed in the two groups. Compared with control, 392 genes were upregulated, and 447 genes were downregulated in model (Figure 2(A and B)).

Figure 2. The exhibition of DEGs of transcriptome sequencing. (A) Volcano plot. (Red dots indicate up-regulated genes, blue dots indicate down-regulated genes and grey dots indicate non-significantly DEGs. (B): Cluster map. (Horizontal indicates genes. One column indicates each sample. Red indicates highly expressed genes and green indicates low expressed genes.)

3.3. Functional analysis of differentially expressed genes

GO enrichment analysis of DEGs (including up- and down-regulated EDGs) was classified according to molecular function (MF), biological process (BP) and cell component (CC). For up-regulated DEGs, they mainly enriched in extracellular region, neuron projection and dendrite in CC, vitamin binding, sulphur compound binding, anion binding in MF and carboxylic acid biosynthetic process, organic acid biosynthetic process and xenobiotic metabolic process in BP (Figure 3(A)). The top 20 GO terms showed that up-regulated EDGs in LN mainly played a role in extracellular region, ion transport, and lipid metabolic process, etc. (Figure 3(B)). KEGG enrichment analysis indicated that up-regulated EDGs mainly enriched in Peroxisome, Calcium signalling pathway, Chemical carcinogenesis, Drug metabolism, and Bile secretion, etc. (Figure 3(C)). The top 20 terms included Chemical carcinogenesis-receptor activation, Fluid shear stress and atherosclerosis, and Serotonergic synapse, etc. (Figure 3(D)).

Figure 3. Functional enrichment analysis of up-regulated DEGs. (A) Bar graph of GO enrichment analysis; (B) Bubble diagram of GO enrichment analysis; (C) Bar graph of KEGG Pathway enrichment; (D) Bubble diagram of KEGG Pathway enrichment.

For down-regulated DEGs, they concentrated on small molecule metabolic process, organic acid metabolic process and lipid metabolic process in BP, endoplasmic reticulum, extracellular region and chylomicron in CC and catalytic activity, acting on paired donors, with incorporation or reduction of molecular oxygen, oxidoreductase activity and cofactor binding in MF (Figure 4(A)). In the top 20 terms, the down-regulated DEGs in LN mainly participated in catalytic activity, catabolic process, organic substance catabolic process, etc. (Figure 4(B)). KEGG enrichment analysis suggested that down-regulated DEGs mainly enriched in Peroxisome, Chemical carcinogenesis, Steroid hormone biosynthesis, Drug metabolism, Fat digestion and absorption, etc. (Figure 4(C)). The top 20 mainly focused on Chemical carcinogenesis-receptor activation, Bile secretion, and Peroxisome, etc. (Figure 4(D)).

Figure 4. Functional enrichment analysis of down-regulated DEGs. (A) Bar graph of GO enrichment analysis; (B) Bubble diagram of GO enrichment analysis; (C) Bar graph of KEGG Pathway enrichment; (D) Bubble diagram of KEGG Pathway enrichment.

3.4. Protein network interaction analysis

In order to further screen the genes related to the occurrence and development of nephropathy, we constructed PPI network of 392 up-regulated genes and 447 down-regulated genes using STRING database (Supplementary Figure 1). A total of 133 lines were obtained with a composite score of >9.6, suggesting that there was a close interacted relationship between some of the up- and down-regulated genes.

3.5. The verification of the expression of DEGs

To test the reliability of sequencing results, we detected the expression of 4 up-regulated and 2 down-regulated DEGs by qRT-PCR. The results showed that the LN increased the expression levels of Kynu, Spidr, Gbp3 and Cbr1 (p < 0.001, Figure 5), and decreased the levels of Cyp4b1 and Cndp2 (p < 0.001), which was consistent with the sequencing results (Table 1).

Figure 5. qRT-PCR detection of the expression of DEGs. **P < 0.01, *** P < 0.001, vs. Control.

Table 1. The exhibition of Part DEGs.

Genes	Description	log2fc	pval	padj	Up/down
Kynu	Kynureninase	13.09495549	3.710939251	1.08703E-82	Up
Spidr	Scaffolding protein involved in DNA repair	2.874113188	1.91231E-70	1.04006E-67	Up
Gbp3	Guanylate binding protein 3	2.607491899	3.98924E-16	4.50836E-14	Up
Cbr1	Carbonyl reductase 1	2.732622864	3.29276E-66	1.59187E-63	Up
Cyp4b1	Cytochrome P450, family 4, subfamily b, polypeptide 1	−2.677624822	4.0472E–113	4.6959E-110	Down
Cndp2	CNDP dipeptidase 2 (metallopeptidase M20 family)	−2.370067605	1.02554E–88	8.11293E-86	Down

3.6. Effects of Gbp3 silencing in HGEC

To interpret the role of Gbp3 in LN, we silenced Gbp3 and explored the influence of it on LPS-treated HGEC. Compared with si-NC, HGEC transfected with three kinds of si-Gbp3 all reduced the expression of Gbp3 (p < 0.001, Figure 6(A)) and we chosed si-Gbp3-3 for the following experiments.

Figure 6. Gbp3 affected the occurrence of LN by regulating cell proliferation, inflammation and pyroptosis. (A) Detection the efficiency of siRNA's transfection by qRT-PCR; (B) Detection of cell proliferation by CCK-8; (C) Detection of inflammatory factors by ELISA; (D) Detection of pyroptosis-related proteins by WB. **P < 0.01, ***P < 0.001, vs. Control. ## P < 0.01, ### P < 0.001, vs. LPS.

Then we assessed the effect of si-Gbp3 on cell proliferation using CCK-8. The cell viability was reduced after the induction of LPS (p < 0.01, Figure 6(B)). Adding LPS + si-NC did not cause significant change while LPS + si-Gbp3 recovered the cell viability (p < 0.01), illustrating that silencing Gbp3 relieved the suppression of LPS on HGEC proliferation.

Moreover, the expression of some inflammatory factors was measured by ELISA. We found that LPS promoted the levels of IL-1β, TNF-α and IL-8 in LPS group (p < 0.01, Figure 6(C)), but the addition of si-Gbp3 significantly inhibited the positive effects of LPS on them (p < 0.01) and si-NC had no effects on them. The results identified that silencing Gbp3 inhibited the overexpression of inflammatory factors caused by LPS in HGEC.

We further determined whether Gbp3 regulated pyroptosis using WB. The expression level of pyroptosis-related proteins (GSDMD, Caspase-1 and NLRP3) in the LPS group was higher than that in Control (p < 0.001, Figure 6(D)). There was no significant change between LPS + si-NC and LPS group, but the expression of GSDMD, Caspase-1 and NLRP3 was inhibited by si-Gbp3 compared to LPS group (p < 0.01). The results showed that silencing Gbp3 inhibited pyroptosis, which was increased by LPS.

3.7. Effects of Gbp3 overexpression in HGEC

In similar opposite way, to further explore the potential role of Gbp3 in LPS-treated HGEC, cells were transfected with pcDNA-Gbp3, which is verified to be upregulated (Figure 7(A)). The function of Gbp3 experiments were also further assessed. After overexpression of Gbp3, cell viability was significantly decreased in the LPS + ov-Gbp3 group (Figure 7(B)). Moreover, Gbp3 overexpression further resulted in increased levels of IL-6, IL-18, and TNF-α (Figure 7(C)). Western blotting was used to detect the pyroptosis-related proteins. The results demonstrated that overexpression of Gbp3 in LPS-treated HGEC caused an increase in GSDMD, Caspase-1 and NLRP3 protein levels (Figure 7(D)). Thus, these results indicated that Gbp3 overexpression promoted inflammatory response, pyroptosis and inhibited cell viability.

Figure 7. Effects of Gbp3 overexpression in HGEC. (A) Gbp3 expression was detected by WB; (B) Cell viability was detected by CCK-8; (C) The levels of proinflammatory cytokines was tested using ELISA; (D) Pyroptosis-related proteins were measured by WB.

4. Discussion

LN is a frequent and severe manifestation of SLE that evolves to ESRD in about 10% of patients within 5 years from the diagnosis [3]. The pathogenesis of LN is highly complex and mediated by multiple signalling pathways and networks of mediators [25]. Recent findings reveal that the pathogenesis of SLE is characterised by the hyper-activation of immunologic pathways related to the antiviral response, moreover, activation and differentiation of B cells are abnormally in SLE patients [26,27]. In our study, we established the mouse model of LN and screened the DEGs via transcriptome sequencing. A total of 839 DEGs showed significant differences, containing 392 upregulated genes and 447 downregulated genes. They were enriched in arachidonic acid metabolism, tryptophan metabolism, retinol metabolism, etc. The arachidonic acid metabolism emerging in the top 20 probably plays a paramount role in the kidney damage during the inflammation process [28]. It is reported that tryptophan metabolism is involved in the prediction of LN diagnosis [29]. Retinol binding protein levels has been proven on an increasing level in SLE patients [30]. Thus, we hypothesised that the DEGs involved in these processes might be associated with the pathogenesis of LN.

Among massive DEGs, 6 candidate DEGs were confirmed by qRT-PCR, containing 4 up-regulated (Kynu, Spidr, Gbp3 and Cbr1) and 2 down-regulated genes (Cyp4b1 and Cndp2). The results proved the reliability of transcriptome sequencing results. It has been reported that in patients with inflammatory diseases, such as atopic dermatitis, Kynu is preferentially upregulated [31]. Tat-Cbr1 protein has anti-inflammatory properties in vitro and in vivo through inhibition of NF-κB and MAPK activation [32]. Increased Cyp4b1 mRNA is associated with the inhibition of dextran sulphate sodium-induced colitis by caffeic acid in mice [33]. Cndp2 is related to the susceptibility of diabetic nephropathy in type 2 diabetes [34]. These genes were related to the development of inflammation, and probably involved in the occurrence of LN via regulating inflammation. However, their regulatory mechanism in LN was needed for further investigation.

Gbp protein elicits a unique role in host defense mechanisms and displays clear associations with specific disease pathogenesis. Gbp1 up-regulation can retard acute viral myocarditis-related cardial damage by inhibiting inflammatory responses and cardiomyocyte apoptosis while increasing cardiomyocyte viability [35]. Similarly, an increased expression of Gbp1 was found in skin lesions of cutaneous lupus erythematosus and in islets of type I diabetes patients [36]. A recent report has indicated that Gbp1 is a key protein in inflammatory pyroptosis [18]. Gbp5 has potential to recover cellular homeostasis and weakened inflammation and tissue destruction in rheumatoid arthritis [37]. By interrogating the functions of individual genes, we hypothesised that Gbp3 may be a pathogenic factor affecting LN. We had evidenced that the expression of Gbp3 was higher in LN mice model than that in Control, which was consistent with the sequencing result. It is reported that Gbp3 is involved in the proliferation of glioma cells through regulating SQSTM1-ERK1/2 pathway [20]. Feng and his colleague evidence that Gbp3 contributes to pathogen-selectivity towards F. novicida, mediating activation of the inflammasome [38]. Then we wondered to prove the contribution of Gbp3 to LN via silencing and overexpression in HGEC model induced by LPS. We found that the knockout of Gbp3 improved the HGEC proliferative potential and reduced the overexpression of inflammatory factors in LN. In contrast, overexpression Gbp3 promoted inflammatory responses and suppressed cell viability. Consistent with previous findings, our results suggest that Gbp3 plays a pathogenic role in LN through regulating cell proliferation and inflammation.

Moreover, silencing Gbp3 was negative related to the expression with the pyroptosis-related proteins, proving that Gbp3 participated in the process of pyroptosis. The development of pyroptosis is always companied with LN. A detailed study performed in 2017 confirmed that the inflammatory body NLRP3, a pyroptosis-related factor, was activated in patients and mice with LN, leading to Sertoli cell injury and severe albuminuria [39]. Piperine is reported to suppress LN by blocking the pyroptosis of tubular epithelial cells, indicating a possible role of pyroptosis in the progression of LN [40]. In addition, Gbp3 governs caspase-4 activation, which could trigger the occurrence of pyroptosis [41]. The above evidence indicated that Gbp3 was associated with pyroptosis and had a positive effect on it to contribute to LN.

Though characterising the role of Gbp3 in LN, the study has several limitations that open the door to future investigation. On the one hand, we only certain the functional effect of Gbp3 on LN but lack of clinical verification, thus further mechanistic studies of it are needed to identify. On the other hand, the role of other screened DEGs in the pathogenesis of LN is worth some attention.

In summary, this study screened the DEGs in the mouse between normal health and LN in transcriptome sequencing results. The expression of Gbp3 had a high level in LN. Gbp3 inhibited the cell viability and improved inflammation and pyroptosis to promote the procession of LN.

Authors' contributions

Zhongfeng Zhang conceived and designed the present study.

Wenyu Song and Run Yan performed the experiments, analysed the data, and drafted the article.

Zhongfeng Zhang revised the article critically for important intellectual content.

All authors read and approved the final manuscript.

Ethics approval and Patient consent

This study adhered to the tenets of the Declaration of Helsinki and was approved by the ethics committee of The Affiliated Hospital of Guizhou Medical University. The animal experiments were approved by the ethics committee of The Affiliated Hospital of Guizhou Medical University. All methods were carried out in accordance with relevant guidelines and regulations. All methods are reported in accordance with ARRIVE guidelines.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The datasets generated and/or analysed during the current study are available in the [Run Yan] repository, [http://www.ncbi.nlm.nih.gov/bioproject/926917].

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.