Abstract

   Hypertensive nephropathy (HN), mainly caused by chronic hypertension,
   is one of the major causes of end-stage renal disease. However, the
   pathogenesis of HN remains unclarified, and there is an urgent need for
   improved treatments. Gene expression profiles for HN and normal tissue
   were obtained from the Gene Expression Omnibus database. A total of 229
   differentially co-expressed genes were identified by weighted gene
   co-expression network analysis and differential gene expression
   analysis. These genes were used to construct protein–protein
   interaction networks to search for hub genes. Following validation in
   an independent external dataset and in a clinical database, POLR2I, one
   of the hub genes, was identified as a key gene related to the
   pathogenesis of HN. The expression level of POLR2I is upregulated in
   HN, and the up-regulation of POLR2I is positively correlated with renal
   function in HN. Finally, we verified the protein levels of POLR2I in
   vivo to confirm the accuracy of our analysis. In conclusion, our study
   identified POLR2I as a key gene related to the pathogenesis of HN,
   providing new insights into the molecular mechanisms underlying HN.

   Keywords: hypertensive nephropathy (HN), weighted gene co-expression
   network analysis (WGCNA), differentially expressed genes (DEGs),
   pathogenesis, key gene, POLR2I

Introduction

   Hypertension is a disease that usually leads to the impairment of
   target organs, especially kidney. Hypertensive nephropathy (HN), mainly
   caused by chronic hypertension, is one of the major causes of end-stage
   renal disease. It has been estimated that 60–90% of patients with
   chronic kidney disease are hypertensive ([35]Ku et al., 2019). However,
   existing data have demonstrated that even when blood pressure is
   reduced to the recommended goal, it can only slow, but not stop, the
   progression of HN ([36]Udani et al., 2011). Therefore, in addition to
   effectively reducing blood pressure, it is particularly important to
   understand the pathogenesis of hypertensive nephropathy. The
   tubulointerstitial compartment constitutes 95% of the total kidney mass
   ([37]Berthier et al., 2012), and the tubulointerstitial changes in HN
   patients are deemed as a major determinant in the development of renal
   damage ([38]Nath, 1992). Moreover, interstitial changes in
   hypertension-induced renal injury occurs before glomerular changes
   become apparent, suggesting that tubulointerstitial compartments may be
   the crucial initial site of injury ([39]Mai et al., 1993). However, the
   detailed molecular mechanism of tubulointerstitial lesions in HN is
   poorly understood.

   A growing body of evidence supported that genetic background affects
   the progression of nephropathy in HN patients ([40]Zhang et al., 2013;
   [41]Guo et al., 2019; [42]Sun et al., 2020). Weighted gene
   co-expression network analysis (WGCNA) ([43]Horvath and Dong, 2008) is
   an effective bioinformatics approach for constructing a co-expression
   network based on gene expression data profile, which provides new
   insights for predicting co-expressed genes related to clinical traits
   and the pathogenesis of diseases. Differential gene expression analysis
   is a widely used and excellent bioinformatics method to detect changes
   in gene expression levels between different groups ([44]Segundo-Val and
   Sanz-Lozano, 2016). Thus, WGCNA and differential gene expression
   analysis were combined to screen key genes related to the pathogenesis
   of tubulointerstitial lesions in HN. We hypothesized that the
   identification of key genes would provide a new insight into HN
   biomarker discovery.

   In the present study, datasets from Gene Expression Omnibus (GEO) were
   adopted to establish a gene co-expression network and to identify
   differentially expressed genes (DEGs) between HN tubulointerstitial
   tissues and matched controls. Then, the overlapping genes that are
   present in DEGs and the trait-related modules were used to construct a
   protein–protein interaction (PPI) network to select hub genes.
   Meanwhile, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and
   Genomes (KEGG) analysis were performed to assess the potential
   functions of the overlapping genes. After GEO and clinical validation,
   a key gene was screened from hub genes. Finally, the key gene was
   validated by in vivo experiments, and the potential biological
   functions of the key gene was investigated by gene set variation
   analysis (GSVA).

Materials and Methods

Data Collection and Data Pre-processing

   All HN and healthy control tubulointerstitial tissue samples were
   selected from the GEO database^[45]1 with the GSE numbers of
   [46]GSE37455 ([47]Berthier et al., 2012), [48]GSE104954 ([49]Grayson et
   al., 2018), and [50]GSE99325 ([51]Shved et al., 2017). These datasets
   were in accord with the following criteria: (1) containing both HN and
   healthy control tubulointerstitial tissues, (2) including at least six
   HN tubulointerstitial samples, (3) the species was Homo sapiens, and
   (4) complete expression profiles were available. The gene expression
   profiles of these datasets were collected from tubulointerstitial
   compartments of kidney biopsies from HN patients and healthy control
   donors ([52]Table 1), whose clinical information is provided in
   [53]Supplementary Tables 1–[54]3, respectively. More detailed clinical
   characteristics could be found in the corresponding references