Bioinformatic Analysis Reveals VEGFA Promotes the Occurrence of PLA2R-associated Idiopathic Membranous Nephropathy by Angiogenesis via PI3K/AKT Signaling Pathway

Backround: PLA2R-associated IMN covers 70% of IMN, which is one of the main types of chronic kidney disease in adults and one of the most common causes of end ‐ stage renal disease. Vascular endothelial growth factor A (VEGFA), a homodimeric vasoactive glycoprotein, is the key mediator of angiogenesis, which lead to numerous kidney diseases, including TSHD7A-associated IMN. However, the role of VEGFA in PLA2R-associated IMN is still poorly understood. Methods: We downloaded the microarray data GSE115857 from GEO. The DEGs were identied with R software, and the functional and pathway enrichment analysis of DEGs was performed utilizing the DAVID and Cytoscape ClueGo plug-in. A comprehensive list of interacting DEGs was constructed using the STRING database and visualized by Cytoscape software. The Cytoscape MCODE and cytoHubba plug-in were used to identify clustered sub-networks, and hub genes from the protein-protein interaction network. Gene set enrichment analysis (GSEA) was used to identify signaling pathway in IMN. Results: There were 1422 genes (952 up-regulated genes and 470 down-regulated genes) were identied as DEGs in GSE115857. The BP of DEGs in GSE115857 was clustered in regulation of transcription from RNA polymerase II promoter, positive regulation of nuclear-transcribed mRNA poly(A) tail shortening, cell adhesion et al. The KEGG pathway of DEGs in GSE115857 was clustered in Rheumatoid arthritis, ABC transporters, PI3K/AKT signaling pathway et al. Then we got a huge PPI network from STRING. 6 modules were screen out to study the functional changes in IMN. The KEGG pathway of module 3 was enriched in soluble N-ethylmaleimide–sensitive factor attachment protein receptor (SNARE) interactions in vesicular transport. There were 3 hub genes screened out, namely, VEGFA, JUN,and FOS. Following the random walk, all genes were ranked and GSEA analysis showed that the signaling pathway of DEGs in GSE115857 was focused on angiogenesis, in which VEGFA

enriched in soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) interactions in vesicular transport. There were 3 hub genes screened out, namely, VEGFA, JUN,and FOS. Following the random walk, all genes were ranked and GSEA analysis showed that the signaling pathway of DEGs in GSE115857 was focused on angiogenesis, in which VEGFA acts as a core gene. CONCLUSION: In summary, this study reveals VEGFA promotes PLA2R-associated IMN by stimulating angiogenesis via PI3K/AKT signaling. Moreover, SNARE interactions in vesicular transport was involved in the development of PLA2R-associated IMN, which may offer a novel therapeutic strategy in treatment of IMN.

Background
Membranous nephropathy (MN) is one of the most common causes of nephritic syndrome in adults and is the second commonest glomerulopathy to progress to end stage renal disease (ESRD). Most cases are idiopathic (IMN), but ~ 20% are secondary to various causes, including cancers, infections, autoimmune diseases, and medications [1]. The M-type phospholipase A2 receptor (PLA2R) and thrombospondin type1 domain-containing 7A (TSHD7A) are two major autoantigens against podocyte antigens in idiopathic membranous nephropathy (IMN) [2,3]. PLA2R-associated IMN is a primary membranous nephropathy with positive serum anti-PLA2R antibody, which covers 70% of IMN [2]. The characteristic feature of IMN is immune complex (IC) formation/deposition on the sub-epithelial side of the glomerular basement membrane resulting in podocyte injury, in which membrane attack complex C5b-9 induces a variety of downstream pathways, including protein kinases, lipid metabolism, reactive oxygen species, growth factors/gene transcription, endoplasmic reticulum stress, and the ubiquitin-proteasome system [4].
The vascular endothelial growth factor A (VEGFA), a homodimeric vasoactive glycoprotein, is the key mediator of angiogenesis. Angiogenesis, the formation of new blood vessels, is responsible for a wide variety of physio/pathological processes in kidney diseases, such as diabetic kidney disease, and polycystic kidney disease [5].In the renal glomeruli, VEGFA is mainly expressed in and secreted from podocytes and tubular epithelial cells, which induces renal injury through binding to receptor VEGFR, mainly VEGFR2, which expressed on the surface of endothelial cells [6]. In recent decades, studies that investigated dysregulation of VEGFA expression during kidney diseases have led to the identi cation of a crucial role of this proangiogenic factor in the renal capillary network [7,8]. Ayumi Matsumoto et al. reported that VEGFA may be important for TSHD7A-associated IMN pathogenesis [9]. However, the role of VEGFA in PLA2R-associated IMN is poorly studied.
The phosphoinositide 3-kinase (PI3K)/AKT signaling plays an important role in physiological and pathological conditions of the organism. In the present study, we attempted to identify crucial genes and pathways that are involved in the pathogenesis of PLA2R-associated IMN. We found that VEGFA is highly expressed in kidney tissue from PLA2R-associated IMN patients. VEGFA may promote the occurrence of PLA2R-associated IMN by stimulating angiogenesis via PI3K/AKT signaling. Moreover, we found six modules that closely related to the development of PLA2R-associated IMN. In addition, SNARE interactions in vesicular transport was involved in the occurrence and progression of PLA2R-associated IMN, which has not been reported before. These results may provide new targets for the treatment of IMN.

Microarray data information
We used "membranous nephropathy" as the keyword to search the Gene Expression Omnibus (GEO) database. We selected the expression pro ling of the array: the attribute name was tissue, and the organisms were Homo sapiens, and only the array with high expression of M-type phospholipase A2 receptor 1 (PLA2R1) was chosen to further analysis (Accession number: GSE115857). GSE115857 contains 11 renal biopsys from IMN patients, and 7 renal biopsys from living donors , which is based on the GPL14951 (Illumina HumanHT-12 WG-DASL V4.0 R2 expression beadchip) platform.
Because the data in this experiment were all from public databases, there was no need for approval by the ethics committee.

Identi cation of differentially expressed genes
The R software (Version 3.6.2) was utilized to screen differentially expressed genes (DEGs) between the IMN group and living donors group. An adjusted p-value of <0.05 and |logFC| >1.0 were considered statistically signi cant. The ggplot2 and RColorBrewer package in R was used to plot the vocano map and the heat map in the two groups.
Gene ontology and pathway enrichment analysis Gene Ontology (GO) includes three categories: molecular function (MF), biological process (BP), and cellular component (CC). GO and KEGG pathway analysis were performed via Database for annotation, Visualization and Integrated Discovery (DAVID) [10] and Cytoscape ClueGo plug-in. Gene Set Enrichment Analysis (GSEA) determines whether an a priori de ned set of genes has statistically signi cant differences in expression under two different biological conditions [11]. This analysis, performed using GSEA software 4.0.3 from the Broad Institute. The gene set of "h.all.v7.1.symbols.gmt", which summarizes and represents speci c, well-de ned biological states or processes, was downloaded from the Molecular Signatures Database (http://software.broadinstitute.org/gsea/msigdb/index.jsp). The normalized enrichment score (NES) was determined by analysis of 1000 permutations. A gene set was considered signi cantly enriched when the P-value was less than 0.05 and the false discovery rate (FDR) was less than 0.25.

Protein-protein interaction network and hub gene identi cation
To evaluate the interactive relationships among DEGs, the STRING online database (https://string-db.org) was used [12]. The protein-protein interaction (PPI) network of DEGs was visualized with Cytoscape software, and central genes were identi ed by cytoHubba to select the hub genes, which were intersected by Betweenness, Bottleneck, Closeness, and Degree method [13]. Signi cant modules in the PPI network were identi ed by Molecular Complex Detection (MCODE), a plug-in of Cytoscape software that clusters a network based on topology to recognize closely connected regions [14]. The MCODE algorithm sorts and identi es each identi ed module. The higher the score, the stronger the genes association in this module.

PPI network and hub gene analysis
To explore the relationship between these DEGs and to identify hub genes, a PPI network of DEGs was constructed using the STRING online database and visualized using Cytoscape. There were 1199 nodes and 5490 edges in the PPI network, including 760 up-regulated genes, 350 down-regulated genes. In addition, 6 modules from the PPI network were selected using the MCODE plug-in of Cytoscape: module 1 (score=24), consisting of 24 nodes and 276 edges (Figure 4  For the BP, genes from module 1 were most signi cantly clustered in positive regulation of mitotic metaphase/anaphase transition (GO:0045842, 5DEGs), there is no BP enrichment in module 2, genes from module 3 were most signi cantly clustered in organelle fusion (GO:0048284, 15DEGs) and regulation of glomerular ltration (GO:0003093, 4DEGs), genes from module 4 were most signi cantly clustered in ribonucleoprotein complex export from nucleus (GO:0071426, 10DEGs), genes from module 5 were most signi cantly clustered in cell adhesion mediated by integrin (GO:0033627, 8DEGs) and maturation small subunit ribosomal RNA (SSU-rRNA) (GO:0030490, 6DEGs), genes from module 6 were most signi cantly clustered in pentose-phosphate shunt, oxidative branch (GO:0009051, 4DEGs) ( Figure  5, Table 3).
In the present study, we used cytoHubba to choose hub genes. According to the ve classi cation methods in cytoHubba, the top 30 hub genes selected by these ranked methods in cytoHubba are shown in Table 5. Finally, three central genes were identi ed by overlapping the rst 30 genes, as shown in Figure   9. VEGFA is the most excellent central genes based on ve ranked methods. JUN and FOS were also selected as hub genes ( Table 6).

Discussion
In recent years, with the use of vascular endothelial growth factor A (VEGFA) inhibitor in oncology, the number of patients with kidney disease is increasing, mainly manifested as thrombotic microangiopathy, minimal change nephropathy, and collapsing focal segmental glomerulosclerosis [8,15,16], however, there is no membrane nephropathy reported. Additionally, rituximab, a B cell-target monoclonal antibody, which has become the rst line therapy in IMN according to KDIGO CLINICAL PRACTICE GUIDELINE ON GLOMERULAR DISEASES 2020, showed its ability to decrease the VEGFA level in plasma in patients with recurrent mantle cell lymphoma [17]. These data suggest that VEGFA may get involved in the development of MN. In this study, we identi ed 1422 genes as DEGs in kidney from IMN patients compared to living donors, 3 hub genes, one up-regulated gene VEGFA, two down-regulated genes JUN and FOS, and six signi cant modules selected from a PPI network, which partially revealed the molecular mechanism in IMN and may be used to develop novel targets for IMN treatment. Moreover, we found that soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) interactions in vesicular transport involved in the occurrence and progression of PLA2R-associated IMN, which was rst reported in this paper.
Angiogenesis is a process of new blood vessel formation from existing vessels, which are essential for the formation of normal kidney structures and for the functions of glomeruli [18]. Abnormal capillary formation in the kidney may cause morphological changes in glomeruli, as well as in ltration of in ammatory cells [19]. Angiogenesis is tightly regulated by a balance of proangiogenic and antiangiogenic factors, in which the most potent proangiogenic factor, vascular endothelial growth factor A (VEGFA),has been intensively investigated [15].VEGFA is the predominant isoform of VEGF in humans and one of the most potent proangiogenic factors, which is mainly expressed in and secreted from podocytes [7] and renal tubular epithelial cells in glomeruli [5]. Expression of VEGFA promotes endothelial cell proliferation, migration, and survival, but can also be associated with vascular hyperpermeability in in ammatory conditions including in ammatory kidney diseases via binding to receptor VEGFR2 [20].Neuropilin 1 (NRP1) functions as a co-receptor for VEGFA and is required for complete activation of VEGFR2 [21]. As VEGFR2 is expressed on the surface of endothelial cells, VEGFA-mediated epithelialendothelial crosstalk in the glomeruli plays an essential role in numerous renal injury [22]. In the present study, we found that VEGFA and NRP1 are highly expressed in kidney tissue from IMN patients. GSEA analysis con rmed that the up-regulated VEGFA promotes angiogenesis in IMN group. Pathway enrichment analysis found that the biological process induced by VEGFA is associated with PI3K/Akt signaling. In addition to, the biological process of modules 3 was clustered in cell migration involved in sprouting angiogenesis, which is a main type of angiogenesis. Thus, VEGFA activates PI3K/Akt signaling via binding to VEGFR2 to accelerate angiogenesis, and leads to vascular hyperpermeability, which result in increased ltration of in ammatory factors, complement, and cytokines, eventually lead to immune complex (IC) and MAC formation. These results suggest that regulation of the VEGFA/PI3K/Akt signalling pathway may be a treatment strategy for IMN.
Complement-mediated cell injury has been demonstrated to contribute to IMN [4]. With complement activation, the formation of membrane attack complex C5b-9 in plasma membranes on the soles of the foot processes of visceral glomerular epithelial cells, which has been shown to induce kidney cell injury through activating a variety of downstream pathways including focal adhesion kinases [23], lipid metabolism, reactive oxygen species, growth factors/gene transcription, endoplasmic reticulum stress, the ubiquitin-proteasome system [4] and DNA damage [24], and it impacts the integrity of the cytoskeleton and slit diaphragm proteins, and contributes to the pathogenesis of mesangial-proliferative glomerulonephritis, thrombotic microangiopathy, and acute kidney injury [4]. In our study, six modules were found from PPI network. Our nding remains consistent with previous studies. The biological process of genes from six modules were clustered in positive regulation of mitotic metaphase/anaphase transition, regulation of glomerular ltration, cell adhesion mediated by integrin, pentose-phosphate shunt, oxidative branch, intrinsic apoptotic signaling pathway in response to endoplasmic reticulum stress, nucleotide-excision repair, DNA incision, and fatty acid metabolism. The KEGG pathway enrichment was clustered in ubiquitin mediated proteolysis, and ECM-receptor interaction. Thus, it seems that the vascular hyperpermeability induced by VEGFA/PI3K/Akt signaling plays an initial role in the development of IMN ( Figure 12).
Notably, we found that SNARE interactions in vesicular transport was involved in the development of IMN, which was the rst reported in this paper. Soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) proteins are essential for exocytosis, mediating the fusion of vesicles with their target membrane. Study reported that SNARE proteins mediate fusion of autophagosomes with endolysosomal vesicles, which mediates autophagosome maturation, resulting in autophagy [25]. And inhibiting autophagosomes induced excessive autophagy ameliorates proteinuria and protect against glomerular and podocyte injury in nephrotic syndrome rats [26].Moreover, with accumulation of autophagosomes, cell viability altered, which directly induces cellular toxicity [27]. In our report, we found that genes, including STX6, STX3, VAMP1, VAMP3, VAMP4, STX1B, and SEC22B, encoding SNARE proteins are highly expressed in kidney biopsies from IMN patients, which means inhibitor of SNARE interaction may provide a new therapeutic strategy in treating IMN,however, it needs to be veri ed experimentally.
It should be noted that there were some limitations to our research. First, recent study demonstrated that there are three non-HLA loci, namely, NFKB1, IRF4 and PLA2R1 revealed in IMN, which means that the canonical NF-κB pathway plays an active role in IMN [28], however, there is no signi cantly different expression of NFKB1 and IRF4 in this study. The shortage may be that the samples from these two groups are too small. Second, focusing on validated target genes may exclude potential targets that have not been validated by the experiment. Finally, due to its descriptive nature, we were not able to validate our hypothesis in this study.

Conclusion
The present study suggests that the PI3K-Akt signaling pathway is involved in the pathogenesis of IMN. VEGFA overexpressed in IMN, and is an important therapeutic target for IMN, which may lead to vascular hyperpermeability via PI3K-Akt signaling pathway, resulting in increased ltration of in ammatory factors, complement, and cytokines, eventually contribute to immune complex (IC) and MAC formation in IMN. Moreover, SNARE interactions in vesicular transport involved in the development of IMN, which needs further basic studies to validate our results and to illuminate the molecular mechanism in IMN.
The red dots represented the up-regulated genes, and the blue dots denoted the down-regulated genes. The grey dots indicated the genes with a |log2Fc|<1 and/or P value>0.05 Biological process enrichment analysis of modules genes.
Page 27/31  Molecular function enrichment analysis of modules genes.

Figure 8
Page 28/31 KEGG enrichment analysis of modules genes.   Signaling pathway activated in kidney biopsies from IMN patients.

Figure 11
Expression pro le of core genes involved in angiogenesis in GSE115857. The red nodes represent upregulated DEGs with a p-value of <0.05 and logFC >1.0; the green nodes represent down-regulated DEGs with a p-value of <0.05 and logFC <-1.0.

Figure 12
The possible mechanisms by which VEGFA/PI3K/Akt is involved in IMN.