Transcriptome-based network analysis reveals renal cell type-specific dysregulation of hypoxia-associated transcripts [glomeruli]. Homo sapiens

Accumulating evidence suggests that dysregulation of hypoxia-regulated transcriptional mechanisms is involved in development of chronic kidney diseases (CKD). However, it remains unclear how hypoxia-induced transcription factors (HIFs) and subsequent biological processes contribute to CKD development and progression. In our study, genome-wide expression profiles of more than 200 renal biopsies from patients with different CKD stages revealed significant correlation of HIF-target genes with eGFR in glomeruli and tubulointerstitium. These correlations were positive and negative and in part compartment-specific. Microarrays of proximal tubular cells and podocytes with stable HIF1α and/or HIF2α suppression displayed cell type-specific HIF1/HIF2-dependencies as well as dysregulation of several pathways. WGCNA analysis identified gene sets that were highly coregulated within modules. Characterization of the modules revealed common as well as cell group- and condition-specific pathways, GO-Terms and transcription factors. Gene expression analysis of the hypoxia-interconnected pathways in patients with different CKD stages revealed an increased dysregulation with loss of renal function. In conclusion, our data clearly point to a compartment- and cell type-specific dysregulation of hypoxia-associated gene transcripts and might help to improve the understanding of hypoxia, HIF dysregulation, and transcriptional program response in CKD. Overall design: Kidney biopsies were microdissected and hybridized on Affymetrix HGU133A and HGU133 Plus 2 arrays. The expression data was normalized using RMA and CustomCDF v 18 ENTREZG in batches, as indicated. Batch correction was performed using ComBat in GenePattern. The disease groups are Diabetic Nephropathy (DN, n=14), Focal and Segmental Glomerulosclerosis (FSGS, n=22), Focal and Segmental Glomerulosclerosis and Minimal Change Disease (FSGS&MCD, n=6), Hypertensive Nephropathy (HT, n=15), IgA Nephropathy (IgA, n=26), Minimal Change Disease (MCD, n=13), Membranous Glomerulonephritis (MGN, n=21), Rapidly Progressive Glomerulonephritis (RPGN, n=23), Systemic Lupus Erythematosus (SLE, n=30), Thin Membrane Disease (TMD, n=3), and Tumor Nephrectomies (TN, n=14). This dataset has substantial overlap with GEO datasets GSE47184,GSE35489, GSE37463

越来越多的证据表明，缺氧调控的转录机制失调参与了慢性肾脏病（chronic kidney diseases, CKD）的发生发展。然而目前尚不清楚缺氧诱导转录因子（hypoxia-induced transcription factors, HIFs）及其后续的生物学过程如何推动CKD的发生与进展。本研究对200余例不同分期CKD患者的肾活检组织开展全基因组表达谱分析，结果显示肾小球与肾小管间质中HIF靶基因的表达与估算肾小球滤过率（estimated glomerular filtration rate, eGFR）存在显著相关性；此类相关性兼具正向与负向，且部分具有组织特异性。对稳定敲低HIF1α和/或HIF2α的近端肾小管细胞及足细胞进行基因芯片分析，结果呈现出细胞类型特异性的HIF1/HIF2依赖性，同时观测到多条通路的失调情况。加权基因共表达网络分析（Weighted Gene Co-expression Network Analysis, WGCNA）鉴定出模块内高度共调控的基因集；对这些模块的表征分析揭示了共通的、以及细胞群与条件特异性的通路、基因本体（Gene Ontology, GO）术语及转录因子。对不同分期CKD患者体内缺氧相关通路的基因表达分析显示，随着肾功能丧失，这些通路的失调程度显著升高。综上，本研究数据明确表明缺氧相关基因转录本存在组织与细胞类型特异性的失调，这或有助于加深人们对CKD中缺氧、HIF失调以及转录程序应答的理解。

总体实验设计：对肾脏活检组织进行显微解剖后，在Affymetrix HGU133A及HGU133 Plus 2芯片上完成杂交。表达数据采用鲁棒多芯片平均法（Robust Multi-array Average, RMA）结合CustomCDF v18 ENTREZG进行批次标准化，具体如前文所述；随后采用GenePattern中的ComBat工具完成批次校正。

本研究纳入的疾病组别包括：糖尿病肾病（Diabetic Nephropathy, DN, n=14）、局灶节段性肾小球硬化症（Focal and Segmental Glomerulosclerosis, FSGS, n=22）、局灶节段性肾小球硬化症合并微小病变病（FSGS&MCD, n=6）、高血压肾病（Hypertensive Nephropathy, HT, n=15）、IgA肾病（IgA Nephropathy, IgA, n=26）、微小病变病（Minimal Change Disease, MCD, n=13）、膜性肾小球肾炎（Membranous Glomerulonephritis, MGN, n=21）、急进性肾小球肾炎（Rapidly Progressive Glomerulonephritis, RPGN, n=23）、系统性红斑狼疮（Systemic Lupus Erythematosus, SLE, n=30）、薄基底膜肾病（Thin Membrane Disease, TMD, n=3）以及肿瘤肾切除术标本（Tumor Nephrectomies, TN, n=14）。

本数据集与基因表达综合数据库（Gene Expression Omnibus, GEO）数据集GSE47184、GSE35489及GSE37463存在大量重叠。