Transcriptomic insights into diabetic kidney disease: a systematic review of RNA sequencing studies

Abstract Background Diabetic kidney disease (DKD) is a leading cause of chronic kidney disease and end-stage kidney disease worldwide. Despite shared clinical risk factors, substantial inter-individual variability exists in DKD susceptibility and progression, suggesting an important contribution of molecular mechanisms beyond traditional clinical predictors. RNA sequencing (RNA-seq) technologies offer an unbiased approach to characterising transcriptomic alterations associated with DKD; however, findings remain fragmented across tissues, RNA classes, and analytical platforms. To date, no systematic review has comprehensively synthesised RNA-seq evidence with respect to the identification of DKD risk factors. Methods A systematic review was conducted in accordance with PRISMA 2020 guidelines. PubMed, Embase All, and Web of Science Core Collection entries from 2006 to April 2025 were searched for RNA-seq studies comparing transcriptomic profiles between diabetic individuals with and without DKD, or across defined DKD progression phenotypes with a further top-up search undertaken from April 2025 to May 2026. Eligible studies employed bulk, single-cell, long-read, or small RNA sequencing in human-derived tissues or biofluids (e.g. plasma, serum, urine). Data related to study design, tissue source, RNA species, sequencing methodology, differential expression findings, and pathway analyses were extracted. Methodological quality was assessed using the Newcastle–Ottawa Scale and an RNA-seq-specific reporting checklist. Results Twenty-eight studies met the inclusion criteria, encompassing kidney biopsy tissue, blood-derived samples, urinary extracellular vesicles, urine sediment, and patient-derived fibroblasts. Studies analysed diverse RNA classes, including messenger RNAs (mRNAs), long non-coding RNAs (lncRNAs), microRNAs (miRNAs), piwi-interacting RNAs (piRNAs), and other small non-coding RNAs (sncRNAs), using bulk, single-cell, and long-read RNA-seq approaches. While individual gene or RNA level overlap was limited, consistent convergence was observed among biological processes. Recurrent implicated themes included inflammatory and immune signalling, fibrosis and extracellular matrix remodelling, mitochondrial and metabolic dysfunction, altered intercellular communication, and cellular stress and repair pathways. Single-cell and spatial transcriptomic studies revealed cell-type–specific injury and fibrotic programmes not discernible in bulk analyses. Substantial heterogeneity in DKD definitions, tissue source, RNA class, comparison groups, sequencing platforms, and analytical pipelines limited cross-study integration necessitating a qualitative evaluation of key findings. Conclusions RNA sequencing studies have revealed widespread transcriptomic dysregulation in DKD across multiple tissues and RNA classes, with convergence on key biological processes underlying disease progression. Although methodological heterogeneity and limited validation constrain interpretation of individual transcriptomic markers, integrative use of bulk, single-cell, and long-read RNA-seq offers complementary insights into DKD pathogenesis. Improved standardisation, longitudinal designs, and multi-tissue integration will enable advanced transcriptomic discovery and translation for improved understanding of diabetic kidney disease mechanisms.