Abstract

   Diabetic kidney disease (DKD) is a long-term major microvascular
   complication of uncontrolled hyperglycemia and one of the leading
   causes of end-stage renal disease (ESDR). The pathogenesis of DKD has
   not been fully elucidated, and effective therapy to completely halt DKD
   progression to ESDR is lacking. This study aimed to identify critical
   molecular signatures and develop novel therapeutic targets for DKD.
   This study enrolled 10 datasets consisting of 93 renal samples from the
   National Center of Biotechnology Information (NCBI) Gene Expression
   Omnibus (GEO). Networkanalyst, Enrichr, STRING, and Cytoscape were used
   to conduct the differentially expressed genes (DEGs) analysis, pathway
   enrichment analysis, protein-protein interaction (PPI) network
   construction, and hub gene screening. The shared DEGs of type 1
   diabetic kidney disease (T1DKD) and type 2 diabetic kidney disease
   (T2DKD) datasets were performed to identify the shared vital pathways
   and hub genes. Strepotozocin-induced Type 1 diabetes mellitus (T1DM)
   rat model was prepared, followed by hematoxylin & eosin (HE) staining,
   and Oil Red O staining to observe the lipid-related morphological
   changes. The quantitative reverse transcription-polymerase chain
   reaction (qRT-PCR) was conducted to validate the key DEGs of interest
   from a meta-analysis in the T1DKD rat. Using meta-analysis, 305 shared
   DEGs were obtained. Among the top 5 shared DEGs, Tmem43, Mpv17l, and
   Slco1a1, have not been reported relevant to DKD. Ketone body metabolism
   ranked in the top 1 in the KEGG enrichment analysis. Coasy, Idi1,
   Fads2, Acsl3, Oxct1, and Bdh1, as the top 10 down-regulated hub genes,
   were first identified to be involved in DKD. The qRT-PCR verification
   results of the novel hub genes were mostly consistent with the
   meta-analysis. The positive Oil Red O staining showed that the
   steatosis appeared in tubuloepithelial cells at 6 w after DM onset.
   Taken together, abnormal ketone body metabolism may be the key factor
   in the progression of DKD. Targeting metabolic abnormalities of ketone
   bodies may represent a novel therapeutic strategy for DKD. These
   identified novel molecular signatures in DKD merit further clinical
   investigation.

   Keywords: diabetic kidney disease, bioinformatics, ketone body
   metabolism, Mpv17l, HMGCS2, BDH1

Introduction

   The prevalence of diabetes and its related complications are increasing
   significantly globally. Diabetic kidney disease (DKD), a devastating
   long-term major microvascular complication of uncontrolled
   hyperglycemia, affects a large population worldwide. The prevalence of
   diabetes is projected to reach 578 million cases by 2030 in the world,
   30 to 40% of which develop DKD ([53]1). DKD is one of the principal
   causes of end-stage renal disease (ESDR) worldwide, and the disease
   progression contributes to irreversible damage to the kidney, impaired
   quality of life, and premature death ([54]2–[55]4). Although extensive
   studies focus on the DKD mechanism and oxidative stress, end-products
   of glycation, autophagy, and apoptosis have been identified to be
   involved in the pathogenesis of DKD, the exact mechanism of DKD remains
   to be elucidated ([56]5–[57]8). Currently, the effective therapy to
   completely halt DKD progression to ESDR is lacking.

   Bioinformatics is a new field of biological research aiming to
   synthesize mathematical, statistical, and computational methods to
   process biological data. Many public repositories, such as Gene
   Expression Omnibus (GEO) and ArrayExpress, provide an enormous quantity
   of data generated by genomic sequencing and microarray chips and
   comprehensive analyses can be performed by integrating multiple studies
   to achieve biological understanding. Integrative bioinformatics
   analysis with larger sample sizes and more minor potential biases is
   superior to finding new molecular signatures ([58]9). Different
   bioinformatics analyses for DKD have been reported ([59]10–[60]18). The
   datasets involved in those studies are either from type 1 diabetic
   kidney disease (T1DKD) ([61]10–[62]13) or type 2 kidney disease (T2DKD)
   samples ([63]14–[64]16). Some genes, such as connective tissue growth
   factor (CTGF) ([65]10), complement 3 (C3) ([66]12), complement 5 (C5)
   ([67]12), cyclin b2 (Ccnb2) ([68]14), and nuclear receptor subfamily 1
   group I member 2 (Nr1i2) ([69]14), have been identified to be molecular
   signatures and showed great potential as therapeutic targets. In 2020,
   systematic integrated analysis of genetic and epigenetic variation in
   DKD patients was reported, and functional annotation suggested the role
   of inflammation, specifically, apoptotic cell clearance and complement
   activation in kidney disease development ([70]17). More recently, Gao
   et al. performed integrative bioinformatics analysis of 3 human
   datasets associated with early T2DKD (pathologic stages I-III),
   glomerular DKD, and tubular DKD, respectively, and identified 7
   candidate genes (SPARC, POSTN, LUM, KNG1, FN1, VCAN, PTPRO)
   significantly associated with the progression of DKD ([71]18).

   The accelerator hypothesis for diabetes is emerging, which argues that
   type 1 diabetes mellitus (T1DM) and type 2 diabetes mellitus (T2DM) are
   the same disorder of insulin resistance set against different genetic
   backgrounds ([72]19). Based on that, we proposed that there should be
   key shared pathways and potential targets involved in the pathogenesis
   of both T1DKD and T2DKD. Unlike previous studies that focused on a
   single type of DKD, the present study enrolled more datasets of T1DKD
   and T2DKD, which were integrated with suitable meta-analysis to deeply
   mine shared molecular mechanism underlying T1DKD and T2DKD. The shared
   DEGs for T1DKD and T2DKD were used to perform the pathway enrichment
   analysis and hub gene screening. The study aims to identify novel
   molecular signatures and therapeutic targets for DKD. Understanding the
   crucial pathways in DKD could provide a new vision into DKD mechanism
   study and facilitate the development of novel therapeutic strategies.

Materials and Methods

Data Collection

   Datasets related to rodent T1DKD and T2DKD were obtained from the
   National Center of Biotechnology Information (NCBI) GEO. [73]Table 1
   showed the detailed information of datasets, and the
   inclusion-exclusion criteria are described as followed: (i) all samples
   are kidney tissues, (ii)datasets should include the control groups and
   experimental groups, and (iii) the datasets should have an appropriate
   sample size (the sample size of the experimental group is greater than
   2, and the sample size of the control group is greater than 2).

Table 1.

   The detailed information of datasets used in Meta-analysis.
   Type Datasets Platforms species samples Models References Experimental