New Insights into the Role of Mitochondrial Dysfunction in Diabetic Kidney Disease in the Omics Era Authors: Xiao ke Liu, Xiao lan Zhou, Wen rui Xia Published: 2026 ABSTRACT AbstractDiabetic kidney disease (DKD) is a major microvascular complication of diabetes and the leading cause of end-stage renal disease worldwide. Accumulating evidence identifies mitochondrial dysfunction as a central driver of DKD progression by linking metabolic reprogramming, oxidative stress, and inflammatory responses. Recent advances in spatial omics, genomics, transcriptomics, proteomics, metabolomics, and microbiome analyses have enabled the characterization of cell-specific and dynamic alterations in mitochondrial function and their interactions with the renal microenvironment. Integrative analyses across these molecular layers have revealed key mechanisms involving mitochondrial quality control, metabolic reprogramming, and inflammatory signaling, while identifying potential biomarkers and therapeutic targets. Furthermore, these approaches have provided mechanistic insights into the renoprotective effects of conventional therapies and bioactive compounds from traditional Chinese medicine. This integrative review synthesizes current evidence to clarify the role of mitochondrial dysfunction in DKD pathogenesis and discusses emerging opportunities for biomarker discovery and precision therapeutics. Introduction Diabetic kidney disease (DKD) is a chronic condition that stems from diabetes. According to the International Diabetes Federation, diabetes is a major contributor to chronic kidney disease (CKD), which is associated with a high risk of kidney failure, cardiovascular events, and premature mortality. Current therapies primarily delay disease progression rather than halt or reverse renal injury. Therefore, a better understanding of the molecular mechanisms underlying DKD is essential for identifying novel biomarkers and therapeutic targets.1 Mitochondrial dysfunction has emerged as a central feature of DKD pathogenesis. In DKD, hyperglycemia-induced excess energetic substrates (glucose, free fatty acids) lead to mitochondrial substrate selection shifts, which trigger defective quality control, ultimately resulting in renal cell damage and disease progression.2,3 Increasing evidence suggests that mitochondrial dysfunction serves as a critical link connecting metabolic disturbances and renal damage during DKD progression. Although substantial progress has been made, how mitochondrial dysfunction evolves across different renal cell populations and contributes to DKD progression remains incompletely characterized.4 In addition, most previous studies have focused on individual molecular layers, limiting the ability to reconstruct regulatory networks that connect genetic susceptibility, transcriptional alterations, protein dysregulation, metabolic reprogramming, and cellular heterogeneity.5 Integrated multi-omics research approaches can provide a comprehensive perspective on disease mechanisms and facilitate biomarker discovery.6 While large-scale GWAS have identified numerous loci associated with CKD susceptibility and kidney function, translating these genetic findings into clinically actionable therapeutic targets remains challenging.7 Recent advances in genomics, transcriptomics, proteomics, metabolomics, spatial omics, and microbiome analyses provide new opportunities to address these challenges. By integrating information across multiple biological layers, multi-omics approaches enable the characterization of cell-specific molecular alterations, spatial tissue heterogeneity, and intercellular communication networks that cannot be captured by single-omics studies alone.8,9 These approaches have expanded our understanding of mitochondrial dysfunction and facilitated the discovery of potential biomarkers and therapeutic targets in DKD. In this integrative review, we summarize current evidence regarding mitochondrial dysfunction in DKD from the perspectives of spatial omics, genomics, transcriptomics, proteomics, metabolomics, gut microbiome studies, and integrated multi-omics analyses. We further discuss the application of these approaches in biomarker discovery, therapeutic target identification, and the development of precision medicine strategies for DKD (Figure 1).10 By integrating these insights, we seek to provide a robust foundation for understanding the core pathogenesis of the disease and developing more efficacious therapeutic strategies. Figure 1Application of multi-omics in drug research and development.10Infographic in a circular wheel with Multi-omics at the center. The inner ring highlights three themes: Disease Study (top-left), Precision Medicine (top-right) and Drug Development (bottom). The outer ring is split into eight labeled wedges with simple icons: Target discovery, Disease understanding, Personalized medicine, Therapeutic response, Drug action network, Pathway prediction, Pharmacological model and Drug discovery. Colors group the wedges into blue (disease-focused), green (precision/clinical focus) and yellow (development and modeling).Wheel chart showing multi-omics links from disease study to precision medicine and drug development. Application of multi-omics in drug research and development.10 The Pathobiology of Mitochondrial Dysfunction is Incompletely Understood in DKD Hyperglycemia-induced mitochondrial dysfunction is a key driver of DKD progression. Excess glucose and lipid substrates increase the production of electron donors, including NADH and flavin adenine dinucleotide, overwhelming the mitochondrial electron transport chain and promoting excessive reactive oxygen species (ROS) generation.11 These ROS damage mtDNA in the inner mitochondrial membrane, impairing respiratory chain function.12 In addition, ROS reduce nitric oxide bioavailability through inhibition of endothelial nitric oxide synthase, thereby exacerbating oxidative stress.12 Oxidative mitochondrial injury also promotes the release of damage-associated molecular patterns (DAMPs), which activate inflammatory pathways and contribute to progressive renal injury.13 A vicious cycle between oxidative stress and Renin-Angiotensin-Aldosterone System (RAAS) hyperactivation accelerates renal damage in DKD. Angiotensin II, via AT1 receptors, induces mitochondrial dysfunction (increasing superoxide, activating UCP2) and inhibits eNOS.14 Concurrently, RAAS upregulates mineralocorticoid receptors, promoting further oxidative stress and fibrosis.15 Conversely, oxidative stress fuels RAAS via ROS-induced mtDNA damage and DAMP release, stimulating Ang II production.13 This bidirectional crosstalk perpetuates mitochondrial dysfunction and renal injury. Defective mitochondrial quality control is another hallmark of DKD. Impaired mitophagy reduces the clearance of damaged mitochondria, leading to the accumulation of dysfunctional organelles and persistent DAMP release.16,17 This process sustains inflammatory responses and further aggravates oxidative stress. In addition, profibrotic and senescence-associated factors, such as transforming growth factor-β (TGF-β), can further suppress mitophagy, creating a self-perpetuating cycle of mitochondrial injury and inflammation.18 Mitochondrial dynamics are essential for maintaining cellular energy homeostasis. Under physiological conditions, mitochondrial fusion, fission, and mitophagy cooperate to preserve mitochondrial integrity by eliminating damaged mitochondria and maintaining mitochondrial quality.19 In DKD, hyperglycemia disrupts this: favoring fission, impairing mitophagy, causing mitochondrial fragmentation, reduced ATP, and dysfunction.4 The accumulation of dysfunctional mitochondria further amplifies oxidative stress and inflammatory responses, thereby aggravating renal injury. Metabolic reprogramming is another hallmark of mitochondrial dysfunction in DKD. Proximal tubular cells (PTCs), which have high energy demands, primarily rely on mitochondrial fatty acid β-oxidation for ATP production.20 However, mitochondrial dysfunction suppresses fatty acid β-oxidation and promotes a metabolic shift toward glycolysis, leading to reduced energy production and increased reactive oxygen species generation.4 Persistent oxidative stress and energy deficiency subsequently activate senescence-associated pathways, thereby promoting premature tubular cell senescence and accelerating DKD progression.21 Although these mechanisms have been extensively investigated, their spatial distribution, cell-type-specific heterogeneity, and dynamic changes during disease progression remain incompletely understood. Recent advances in spatial omics provide new opportunities to characterize mitochondrial dysfunction within its native tissue context, enabling a more comprehensive understanding of the molecular landscape of DKD. Spatialomics Spatial omics technologies have emerged as powerful tools for dissecting compartment-specific disruptions in mitochondrial function that drive the progression of DKD. These approaches, including spatial transcriptomics, spatial metabolomics, and spatial proteomics, collectively elucidate how spatially confined alterations in mitochondrial biology, lipid metabolism, inflammatory signaling converge to promote region-specific renal injury.22 Spatial transcriptomics has revealed marked cell-type and region-specific heterogeneity of mitochondrial dysfunction in DKD. Single-cell multi-omic and spatial profiling identified injured proximal tubular (iPT) cells, particularly the iPT_VCAM1+ subtype, as a key component of the fibrotic microenvironment (FMEs), characterized by impaired fatty acid oxidation (FAO), enhanced glycolysis, and increased SPP1/CXCL12-mediated profibrotic signaling.23–25 Spatial metabolomic analyses have identified distinct regional metabolic signatures associated with mitochondrial dysfunction in DKD.26 Ceramide synthase 6 (CerS6)-mediated ceramide accumulation promotes VDAC-dependent mitochondrial damage and activation of the cGAS–STING pathway, while depletion of tricarboxylic acid (TCA) cycle intermediates indicates impaired mitochondrial oxidative metabolism in renal tubules.27–31 Spatial proteomics further validates compartment-specific mitochondrial abnormalities at the protein level by demonstrating colocalization of pro-inflammatory proteins (CD45, CD68) with mitochondrial injury and enrichment of glycolysis-related proteins in sclerotic glomeruli, confirming the metabolic reprogramming identified by spatial transcriptomics.32–34 Collectively, spatial transcriptomics, metabolomics, and proteomics provide complementary insights into mitochondrial dysfunction by defining cell-specific transcriptional changes, regional metabolic remodeling, and protein-level alterations within their anatomical context. However, each platform captures only one molecular layer, highlighting the value of integrated spatial multi-omics approaches. A summary of potential biomarkers and mechanisms is provided in Table 1.Table 1Spatial Omics – Potential Biomarkers and MechanismsTechnologyPlatformClinically Used Biomarkers/TestsPreclinical Biomarkers (Not Clinically Used)Related PathwaysMechanismsReferencesSpatial TranscriptomicsVisium, CosMx, Seurat/UMAPAlbuminuria, estimated Glomerular Filtration Rate (eGFR)iPT_VCAM1+ cell gene signaturemTORFAO↓, Glycolysis↑, mt respiration↓, Fibrosis↑[23–25,31]Spatial MetabolomicsDESI-MS ImagingUrinary 8-hydroxydeoxyguanosine (8-OHdG)CerS6-derived ceramide (d18:1/16:0)cGAS-STINGmt permeability↑, mtDNA leakage → Inflammation↑[22,27–30]Spatial MetabolomicsMALDI-MSI PlatformN/ATCA cycle intermediates (α-ketoglutarate, succinate, fumarate)Mitochondrial Oxidative MetabolismOxidative metabolism↓, Energy depletion↑, Tubular injury↑[8]Spatial ProteomicsCODEX, IMCN/APro-inflammatory proteins (CD45, CD68) co-localizationInflammatory SignalingImmune infiltration↑, mt dysfunction → Local inflammation↑[32]Spatial ProteomicsLaser Microdissection, Thermo Orbitrap Fusion LumosN/AGlycolytic regulators (Aldolase A/B, Triosephosphate Isomerase, TPI)Glycolysis-Mitochondrial CrosstalkGlycolysis↑, mt energy production↓, Repair↓[33,34]Notes: ↑, increased or upregulated; ↓, decreased or downregulated. Table 1Spatial Omics – Potential Biomarkers and MechanismsTechnologyPlatformClinically Used Biomarkers/TestsPreclinical Biomarkers (Not Clinically Used)Related PathwaysMechanismsReferencesSpatial TranscriptomicsVisium, CosMx, Seurat/UMAPAlbuminuria, estimated Glomerular Filtration Rate (eGFR)iPT_VCAM1+ cell gene signaturemTORFAO↓, Glycolysis↑, mt respiration↓, Fibrosis↑[23–25,31]Spatial MetabolomicsDESI-MS ImagingUrinary 8-hydroxydeoxyguanosine (8-OHdG)CerS6-derived ceramide (d18:1/16:0)cGAS-STINGmt permeability↑, mtDNA leakage → Inflammation↑[22,27–30]Spatial MetabolomicsMALDI-MSI PlatformN/ATCA cycle intermediates (α-ketoglutarate, succinate, fumarate)Mitochondrial Oxidative MetabolismOxidative metabolism↓, Energy depletion↑, Tubular injury↑[8]Spatial ProteomicsCODEX, IMCN/APro-inflammatory proteins (CD45, CD68) co-localizationInflammatory SignalingImmune infiltration↑, mt dysfunction → Local inflammation↑[32]Spatial ProteomicsLaser Microdissection, Thermo Orbitrap Fusion LumosN/AGlycolytic regulators (Aldolase A/B, Triosephosphate Isomerase, TPI)Glycolysis-Mitochondrial CrosstalkGlycolysis↑, mt energy production↓, Repair↓[33,34]Notes: ↑, increased or upregulated; ↓, decreased or downregulated. Spatial Omics – Potential Biomarkers and Mechanisms Notes: ↑, increased or upregulated; ↓, decreased or downregulated. Transcriptomics Transcriptomic approaches, including bulk RNA sequencing (RNA-seq) and single-cell or single-nucleus RNA sequencing (sc/snRNA-seq), have substantially advanced our understanding of mitochondrial dysfunction in DKD by enabling cell-type-specific characterization of gene expression and regulatory networks.35 Compared with conventional transcriptomic analyses, single-cell technologies offer higher resolution for identifying cellular heterogeneity and mitochondrial alterations within distinct renal cell populations. Transcriptomic studies have consistently demonstrated that disruption of mitochondrial quality control (MQC) is a central feature of DKD. Altered expression of genes involved in mitochondrial biogenesis, mitophagy, and oxidative stress regulation, including PGC-1α, TFAM, PINK1, and Parkin, indicates impaired mitochondrial maintenance in diabetic kidneys.36 In addition, restoring DNMT1 expression or activating the CDX2/CFTR signaling pathway has been shown to improve mitochondrial quality control and alleviate renal fibrosis, suggesting that these pathways are potential therapeutic targets.37,38 Investigating cellular heterogeneity, particularly mitochondrial dynamics, at the single-cell level is essential for elucidating their roles in disease progression and responses to therapy, thereby providing a foundation for subsequent validation through in vivo and in vitro experimental approaches. At the single-cell level, transcriptomic profiling has further revealed cell-specific mechanisms linking mitochondrial dysfunction to inflammation. CerS6 enrichment in podocytes promotes mitochondrial DNA leakage and cGAS–STING activation, while dysregulation of the PVT1–TRIM56–AMPKα axis enhances mitochondrial fission and NF-κB-mediated inflammatory responses.22,39–41 Transcriptomic analyses also consistently indicate suppression of oxidative phosphorylation and fatty acid oxidation with enhanced glycolytic programs, mediated by alterations in HIF-1α, SIRT3, and the PGC-1α/PPARα axis.42–46 These transcriptional abnormalities can be partially reversed by SGLT2 inhibition, suggesting potential metabolic targets for DKD intervention.47 Nevertheless, transcriptomic approaches primarily reflect regulatory changes rather than functional metabolic outcomes. Integration with metabolomic and proteomic analyses is therefore required to determine how transcriptional alterations translate into mitochondrial metabolic dysfunction and renal injury. Genomics Genomic approaches have provided important insights into the genetic basis of mitochondrial dysfunction in DKD by identifying susceptibility variants, mitochondrial DNA (mtDNA) alterations, and nuclear-encoded mitochondrial genes (NEMG) associated with disease progression. Unlike transcriptomic approaches that capture disease-associated expression changes, genomics focuses on inherited variation and provides evidence for genetic predisposition and potential causal relationships underlying DKD progression. Genomic studies consistently demonstrate that both mtDNA variants and NEMGs contribute to mitochondrial dysfunction and DKD susceptibility. Variants in mtDNA, together with NEMGs such as TBC1D32, SURF1, and AGXT2, impair mitochondrial oxidative phosphorylation and energy metabolism, thereby increasing susceptibility to kidney injury in patients with diabetes.48–50 Through Mendelian randomization (MR) and colocalization analysis with genetic instruments (cis-SNPs), it was found that mitochondrial-related genes such as TUFM and ISCA2 possess risk variants causally linked to diabetic nephropathy due to genomic perturbations.51 Beyond inherited genetic variation, genomic analyses have also identified regulatory molecules involved in mitochondrial homeostasis. Increased expression of miR-379 suppresses genes associated with mitophagy and mitochondrial quality control, leading to impaired mitochondrial function, oxidative stress, and progression of renal fibrosis and albuminuria.52,53 In addition, single-nucleus transcriptomic analyses identified a protective KLF6–ApoJ–CaMK1D signaling axis that mediates podocyte-to-tubule communication and preserves mitochondrial function during DKD progression.54 Collectively, genomic approaches define the inherited susceptibility and regulatory landscape of mitochondrial dysfunction in DKD, but their ability to capture dynamic molecular changes during disease progression remains limited. Integration with transcriptomic, metabolomic, and proteomic approaches is therefore essential to connect genetic risk with functional mitochondrial alterations. A summary of potential biomarkers and mechanisms is provided in Table 2.Table 2Transcriptomics and Genomics – Potential Biomarkers and MechanismsTechnologyPlatformClinically Used Biomarkers/TestsPreclinical Biomarkers (Not Clinically Used)Related PathwaysMechanismsReferencesSingle-cell/nuclear RNA-seq (sc/snRNA-seq)10x Genomics, Seurat/UMAPN/AYAP1, PGC-1α, TFAM, PINK1Mitochondrial Quality Control (MQC)Mt biogenesis↓, Mitophagy↓, Oxidative stress↑[36]scRNA-seqInteractive Kidney TranscriptomicsProteinuriaPodocyte-specific CerS6cGAS-STINGmtDNA leakage → Innate immunity↑, Inflammation↑[22,39,40]RNA-seqIllumina NovaSeq/HiSeqN/AlncRNA PVT1, DNMT1, CDX2/CFTRAMPK, Epigenetic RegulationEnergy sensing↓, Epigenetic dysregulation, Mt dysfunction[37,38,41]Genome-Wide Association Study (GWAS) and GenotypingUK Biobank Axiom Arrays, HaploGrep2Fasting insulin, obesity rateNuclear-encoded mitochondrial genes (NEMGs: TBC1D32, SURF1, AGXT2), mtDNA variants (m.A3943G, m.A10005G)Oxidative Phosphorylation (OXPHOS)OXPHOS disruption → Disease susceptibility↑[48–51]miRNA ProfilingIllumina miRNA MicroarraysN/AmiR-379, miR-668, miR-214Mitophagy, Mitochondrial DynamicsMitophagy↓, Mt dynamics imbalance[52,53]Notes: ↑, increased or upregulated; ↓, decreased or downregulated. Table 2Transcriptomics and Genomics – Potential Biomarkers and MechanismsTechnologyPlatformClinically Used Biomarkers/TestsPreclinical Biomarkers (Not Clinically Used)Related PathwaysMechanismsReferencesSingle-cell/nuclear RNA-seq (sc/snRNA-seq)10x Genomics, Seurat/UMAPN/AYAP1, PGC-1α, TFAM, PINK1Mitochondrial Quality Control (MQC)Mt biogenesis↓, Mitophagy↓, Oxidative stress↑[36]scRNA-seqInteractive Kidney TranscriptomicsProteinuriaPodocyte-specific CerS6cGAS-STINGmtDNA leakage → Innate immunity↑, Inflammation↑[22,39,40]RNA-seqIllumina NovaSeq/HiSeqN/AlncRNA PVT1, DNMT1, CDX2/CFTRAMPK, Epigenetic RegulationEnergy sensing↓, Epigenetic dysregulation, Mt dysfunction[37,38,41]Genome-Wide Association Study (GWAS) and GenotypingUK Biobank Axiom Arrays, HaploGrep2Fasting insulin, obesity rateNuclear-encoded mitochondrial genes (NEMGs: TBC1D32, SURF1, AGXT2), mtDNA variants (m.A3943G, m.A10005G)Oxidative Phosphorylation (OXPHOS)OXPHOS disruption → Disease susceptibility↑[48–51]miRNA ProfilingIllumina miRNA MicroarraysN/AmiR-379, miR-668, miR-214Mitophagy, Mitochondrial DynamicsMitophagy↓, Mt dynamics imbalance[52,53]Notes: ↑, increased or upregulated; ↓, decreased or downregulated. Transcriptomics and Genomics – Potential Biomarkers and Mechanisms Notes: ↑, increased or upregulated; ↓, decreased or downregulated. Proteomic Mitochondrial impairment is a key convergent mechanism, driving energy deficit, oxidative stress, and tubular injury in DKD. Proteomic approaches provide complementary insights by directly assessing mitochondrial protein abundance, functional alterations, and post-translational modifications that cannot be fully captured by genomic or transcriptomic analyses. Recent advances in proteomic technologies have been instrumental in elucidating the central role of mitochondrial dysfunction in the pathogenesis of DKD. Urinary alpha-1-microglobulin/bikunin precursor (AMBP) has been shown to preserve mitochondrial function and protect proximal tubular epithelial cells against oxidative stress, suggesting its potential as a biomarker of tubular injury.55,56 Moreover, mitochondrial proteomic analyses identified multiple dysregulated mitochondrial proteins, including SLC25A16, ALDH3A2, NUDT19, and NDUFA4L2, indicating widespread disruption of mitochondrial transport, metabolism, and energy homeostasis during DKD progression.57 Proteomics has further demonstrated that upregulation of ACSL4 promotes mitochondrial damage and ferroptosis through activation of the C/EBPα–ACSL4 pathway, thereby accelerating renal injury.58,59 Furthermore, hyperacetylation of CPT1a impairs mitochondrial fatty acid β-oxidation via dysregulation of the SIRT3–CPT1a pathway, resulting in lipid accumulation and tubular injury.60 Proteomics also demonstrates that glycation-derived proteins such as CML downregulate mitochondrial CPT2, thereby impairing β-oxidation and promoting tubular fibrosis in DKD.61 Collectively, proteomics bridges molecular alterations and functional mitochondrial phenotypes by revealing protein-level dysregulation and post-translational modifications. However, limited protein coverage and technical dependence remain challenges, highlighting the need for integration with transcriptomic and metabolomic approaches to establish a comprehensive understanding of mitochondrial dysfunction in DKD. A summary of potential biomarkers and mechanisms is provided in Table 3.Table 3Proteomics – Potential Biomarkers and MechanismsTechnologyPlatformClinically Used Biomarkers/TestsPreclinical Biomarkers (Not Clinically Used)Related PathwaysMechanismsReferencesLC-MS/MSThermo Q Exactive, Sciex TripleTOFUrinary AMBP (α1-microglobulin/bikunin precursor)N/AOxidative Stress DefenseMt function↑, Oxidative damage resistance↑[55,56]Tandem Mass Tag (TMT)-based ProteomicsThermo Orbitrap Fusion LumosN/AACSL4 (upregulated)C/EBPα–ACSL4–FerroptosisMt damage↑, Ferroptosis↑ → Renal injury↑[58,59]Acetylome AnalysisThermo Orbitrap LC-MS/MSN/AHyperacetylated CPT1a (lysine residues 86, 639)SIRT3-CPT1a-FAOFAO↓, Lipid accumulation↑, Tubular injury↑[60]Mitochondrial ProteomicsThermo Orbitrap Astral MSN/ASLC25A16, CTNND1, C2CD2L, ALDH3A2, NEU1, APEH, CORO1A, NUDT19, NDUFA4L2Mitochondrial Transport/MetabolismMt transport/metabolism/function↓ → Disease progression↑[57]Notes: ↑, increased or upregulated; ↓, decreased or downregulated. Table 3Proteomics – Potential Biomarkers and MechanismsTechnologyPlatformClinically Used Biomarkers/TestsPreclinical Biomarkers (Not Clinically Used)Related PathwaysMechanismsReferencesLC-MS/MSThermo Q Exactive, Sciex TripleTOFUrinary AMBP (α1-microglobulin/bikunin precursor)N/AOxidative Stress DefenseMt function↑, Oxidative damage resistance↑[55,56]Tandem Mass Tag (TMT)-based ProteomicsThermo Orbitrap Fusion LumosN/AACSL4 (upregulated)C/EBPα–ACSL4–FerroptosisMt damage↑, Ferroptosis↑ → Renal injury↑[58,59]Acetylome AnalysisThermo Orbitrap LC-MS/MSN/AHyperacetylated CPT1a (lysine residues 86, 639)SIRT3-CPT1a-FAOFAO↓, Lipid accumulation↑, Tubular injury↑[60]Mitochondrial ProteomicsThermo Orbitrap Astral MSN/ASLC25A16, CTNND1, C2CD2L, ALDH3A2, NEU1, APEH, CORO1A, NUDT19, NDUFA4L2Mitochondrial Transport/MetabolismMt transport/metabolism/function↓ → Disease progression↑[57]Notes: ↑, increased or upregulated; ↓, decreased or downregulated. Proteomics – Potential Biomarkers and Mechanisms Notes: ↑, increased or upregulated; ↓, decreased or downregulated. Metabolomics Metabolomics and lipidomics provide functional insights into mitochondrial dysfunction by directly capturing alterations in metabolites and lipid species associated with disease progression. Compared with transcriptomic and proteomic approaches, these platforms more closely reflect the biochemical consequences of mitochondrial impairment and metabolic remodeling, offering valuable biomarkers and therapeutic targets for DKD.62,63 Metabolomics studies have demonstrated that mitochondrial dysfunction in DKD is marked by alterations in key metabolites.64,65 Altered acylcarnitine profiles indicate impaired mitochondrial fatty acid β-oxidation, while changes in tricarboxylic acid (TCA) cycle intermediates, glycolytic metabolites, and adipoylcarnitine reflect disrupted mitochondrial energy metabolism.66–68 In addition, alterations in urea cycle metabolites, including ornithine and citrulline, are associated with mitochondrial arginase 2 (ARG2), suggesting that dysregulated mitochondrial biogenesis, reactive oxygen species production, and mitochondrial dynamics contribute to DKD progression.69,70 Lipidomic analyses further demonstrate that phospholipid remodeling is a hallmark of mitochondrial dysfunction in DKD. Reduced cardiolipin levels together with increased lactosylceramides impair cardiolipin-dependent mitophagy and compromise mitochondrial membrane integrity.71,72 Furthermore, activation of the ChREBP–Gnpat pathway promotes mitochondrial fission through increased plasmalogen synthesis, whereas ALCAT1-mediated cardiolipin remodeling suppresses AMPK signaling, enhances lipid peroxidation, and aggravates podocyte injury.73–76 Collectively, metabolomics and lipidomics reveal the downstream metabolic consequences of mitochondrial dysfunction, including impaired fatty acid oxidation, altered energy metabolism, and lipid remodeling. However, metabolic profiles alone cannot fully determine the upstream regulatory mechanisms driving these changes, highlighting the necessity of integrating metabolomic data with transcriptomic and proteomic analyses. A summary of potential biomarkers and mechanisms is provided in Table 4.Table 4Metabolomics and Lipidomics – Potential Biomarkers and MechanismsTechnologyPlatformClinically Used Biomarkers/TestsPreclinical Biomarkers (Not Clinically Used)Related PathwaysMechanismsReferencesLC/GC-MSAgilent UHPLC-Q-TOF, Bruker GC-MSAcylcarnitines (long-chain), uric acidAcylcarnitines (short/medium-chain), TCA intermediates (succinate, fumarate)Mitochondrial FAOFAO↓ (Early/Late DKD), Oxidative metabolism↓[8,64–66]UPLC-MS/MSSciex QTRAP UPLC-MS/MSTrimethylamine N-oxide (TMAO)AdipoylcarnitineMitochondrial DysfunctionMt impairment → Renal dysfunction↑[68]Targeted LipidomicsSciex QTRAP 6500 PLUSN/ADecreased cardiolipin, increased lactosylceramidesCardiolipin-Dependent MitophagyMitophagy↓, Mt membrane integrity↓[71,72]UHPLC-MSThermo Orbitrap Fusion LumosN/AIncreased mitochondrial plasmalogens (plasmalogen-PC, plasmalogen-PE)ChREBP-GnpatMt fission↑ → DKD aggravation[73–75]LipidomicsUPLC-HRMS PlatformN/AIncreased oxidized cardiolipinALCAT1-AMPKLipid peroxidation↑, Energy sensing↓ → Podocyte injury↑[76]Metabolomics (NMR)Bruker NMR SpectrometerN/AUrea cycle metabolites (ornithine, citrulline)ARG2-Mitochondrial ROSUrea cycle disruption → Mt biogenesis and ROS affected[69,70]Notes: ↑, increased or upregulated; ↓, decreased or downregulated. Table 4Metabolomics and Lipidomics – Potential Biomarkers and MechanismsTechnologyPlatformClinically Used Biomarkers/TestsPreclinical Biomarkers (Not Clinically Used)Related PathwaysMechanismsReferencesLC/GC-MSAgilent UHPLC-Q-TOF, Bruker GC-MSAcylcarnitines (long-chain), uric acidAcylcarnitines (short/medium-chain), TCA intermediates (succinate, fumarate)Mitochondrial FAOFAO↓ (Early/Late DKD), Oxidative metabolism↓[8,64–66]UPLC-MS/MSSciex QTRAP UPLC-MS/MSTrimethylamine N-oxide (TMAO)AdipoylcarnitineMitochondrial DysfunctionMt impairment → Renal dysfunction↑[68]Targeted LipidomicsSciex QTRAP 6500 PLUSN/ADecreased cardiolipin, increased lactosylceramidesCardiolipin-Dependent MitophagyMitophagy↓, Mt membrane integrity↓[71,72]UHPLC-MSThermo Orbitrap Fusion LumosN/AIncreased mitochondrial plasmalogens (plasmalogen-PC, plasmalogen-PE)ChREBP-GnpatMt fission↑ → DKD aggravation[73–75]LipidomicsUPLC-HRMS PlatformN/AIncreased oxidized cardiolipinALCAT1-AMPKLipid peroxidation↑, Energy sensing↓ → Podocyte injury↑[76]Metabolomics (NMR)Bruker NMR SpectrometerN/AUrea cycle metabolites (ornithine, citrulline)ARG2-Mitochondrial ROSUrea cycle disruption → Mt biogenesis and ROS affected[69,70]Notes: ↑, increased or upregulated; ↓, decreased or downregulated. Metabolomics and Lipidomics – Potential Biomarkers and Mechanisms Notes: ↑, increased or upregulated; ↓, decreased or downregulated. Gut Microbiome Increasing evidence indicates that the gut microbiome is an important regulator of mitochondrial dysfunction in DKD through the gut–kidney axis.77 DKD disrupts the balance of gut microbiota through three aspects: microenvironment, metabolic conditions, and immune status, via the chain reaction of “hyperglycemia-induced damage - renal function decline - inflammation activation”, featured by reduced abundance of beneficial bacteria like Bifidobacterium, Lactobacillus, and Prevotella, increased proportion of pathogenic bacteria such as Enterobacteriaceae and Klebsiella, and decreased anaerobic bacteria.78,79 Microbiome-associated metabolomic studies have identified several gut microbiota-derived metabolites that regulate mitochondrial function in DKD. Short-chain fatty acids (SCFAs) exert renoprotective effects by activating G protein-coupled receptors (GPR41/43) and modulating the HDAC/NF-κB signaling pathway, thereby improving mitochondrial function and suppressing inflammation.80 In contrast, harmful microbial metabolites, including imidazolepropionic acid, adipoylcarnitine, and phenyl sulfate, impair mitochondrial fatty acid β-oxidation and oxidative phosphorylation, activate inflammatory pathways, and promote albuminuria and glomerular injury.68 Metagenomic and metabolomic analyses have further demonstrated that depletion of butyrate-producing bacteria and reduced production of indole-3-propionic acid (IPA) impair mitochondrial biogenesis and antioxidant defense through suppression of the SIRT1/PGC-1α pathway, thereby aggravating DKD progression.81,82 Phenyl sulfate (PS), another gut microbiota-derived metabolite, suppresses mitochondrial oxidative phosphorylation in podocytes and promotes albuminuria and glomerular damage.83 Targeting the gut microbiota has emerged as a promising therapeutic strategy for preserving mitochondrial homeostasis. Restoration of beneficial bacteria enhances mitochondrial function through bile acid-mediated activation of TGR5 signaling, whereas interventions such as kidney tea alleviate mitochondrial damage by reshaping the gut microbiota, suppressing ferroptosis, and restoring mitochondrial homeostasis.84,85 However, microbiome-based approaches remain limited by substantial interindividual variability and the difficulty in establishing causal relationships between microbial alterations, metabolite changes, and renal injury. Collectively, multi-omics analyses demonstrate that gut microbiota-derived metabolites regulate mitochondrial energy metabolism, oxidative stress, inflammation, and ferroptosis. These findings highlight the gut–kidney axis as a promising source of biomarkers and therapeutic targets for DKD. Multi-Omics Multiomics platforms have emerged as powerful methodologies, enabling researchers to investigate the intricate molecular basis of DKD with depth and resolution. Through the comprehensive integration of diverse datasets, it becomes feasible to identify complex interactions, which in turn advances our comprehension of how a network of omics factors collectively perturbs biological pathways and exerts an impact on disease pathogenesis.63 These comprehensive analyses not only identify novel biomarkers and pathological mechanisms but also reveal the intricate crosstalk between cells, proteins, metabolic molecules, and transcription factors.86 This insight enhances our understanding of DKD progression and lays a foundation for the development of targeted therapeutic strategies. Integrated multi-omics studies consistently identify mitochondrial metabolic reprogramming as a central event in DKD progression. Genetic susceptibility loci, such as ELMO1, are associated with reduced expression of mitochondrial fatty acid β-oxidation proteins, including CPT1A and ACADM, accompanied by accumulation of acylcarnitines and TCA cycle intermediates, indicating impaired oxidative metabolism and lipotoxicity.8,87,88 Furthermore, integrated analyses using transcriptomics, metabolomics, and protein interaction networks have identified mitochondrial regulators such as MDM2, whose downregulation disrupts citrate metabolism and TCA cycle activity, thereby promoting podocyte injury and tubular damage.89 Multi-omics integration also provides mechanistic insight into metabolic reprogramming during DKD. Through multi-omics approaches, including transcriptomics, metabolomics, and spatial metabolomics, this study demonstrates that mitochondrial metabolic reprogramming in DKD is characterized by enhanced glycolysis, suppressed fatty acid oxidation, and dysregulated amino acid metabolism.42 These metabolic alterations are associated with dysregulation of the PKM2, AMPK–SIRT1–PGC-1α, and PINK1/Parkin pathways, ultimately leading to oxidative stress, inflammation, fibrosis, and progressive renal dysfunction.4,9,90–93 These findings illustrate how transcriptional changes translate into metabolic abnormalities and downstream renal damage. Another major advantage of multi-omics approaches is the ability to identify disease-associated biomarkers across multiple molecular layers. Integrated analyses combining transcriptomics, genomics, and scRNA-seq have identified mitochondrial dysfunction-related genes, including ACTN1, that participate in AGE–RAGE signaling, inflammatory responses, and cytoskeletal remodeling.94,95 Similarly, integrated proteomic and metabolomic studies demonstrate that glycation-induced mitochondrial dysfunction disrupts lipid metabolism through the AGE/RAGE and PPARα/Wnt pathways, thereby accelerating fibrosis and DKD progression.61,94,96 In addition, combined transcriptomic and proteomic analyses reveal that insulin resistance induces post-transcriptional downregulation of electron transport chain proteins, resulting in impaired oxidative phosphorylation, ATP depletion, and metabolic reprogramming toward glycolysis.97 Emerging multi-omics approaches further integrate spatial omics and microbiome-derived metabolomics to characterize mitochondrial dysfunction within the renal microenvironment. These studies demonstrate that excessive mitochondrial ROS production, mitochondrial-associated membrane remodeling, and altered gut microbiota-derived metabolites collectively contribute to immune activation, fibrosis, and mitochondrial injury through pathways including Nrf2/PINK1, mTOR, and TXNIP.93,98 These findings indicate the importance of considering both tissue-level spatial organization and systemic metabolic interactions when investigating mitochondrial dysfunction in DKD. There are still many challenges to be addressed in multi-omics technologies. Currently, most studies integrate only two or three omics levels and are primarily cross-sectional in nature, which limits the analysis of how mitochondrial dysfunction evolves over time during the course of DKD. Furthermore, standardized analytical workflows and independent external validation mechanisms remain underdeveloped, restricting the clinical application of candidate biomarkers. Future research should integrate spatial omics, single-cell sequencing, metabolomics, and longitudinal clinical cohort data. A summary of potential biomarkers and mechanisms is provided in Table 5.Table 5Integrated Multi-Omics and Gut Microbiome – Potential Biomarkers and MechanismsTechnologyPlatformClinically Used Biomarkers/TestsPreclinical Biomarkers (Not Clinically Used)Related PathwaysMechanismsReferencesGenomics+Transcriptomics+Proteomics+MetabolomicsCytoscape, KEGG, Illumina NovaSeqN/AELMO1 SNPs → decreased CPT1A/ACADM → increased acylcarnitinesFAO-OXPHOS CrosstalkFAO pathway disruption → Lipotoxicity↑[8]Transcriptomics+Metabolomics+Spatial OmicsSeahorse Analyzer, MALDI-MSIProteinuria, eGFRBranched-chain amino acid (BCAA) catabolic defects, PKM2PKM2-OXPHOSOXPHOS↓, Glycolysis↑[91–93]Microbiome+MetabolomicsIllumina NovaSeq 6000, Bruker UHPLC-QTOFShort-chain fatty acids (SCFAs), TMAOPhenyl sulfate (PS), Indoxyl sulfate (IS), Indole-3-propionic acid (IPA)NLRP3, SIRT1/PGC-1αInflammasome activation /Mt biogenesis↑[68,80–83]Transcriptomics+ProteomicsThermo Orbitrap, Illumina HiSeqN/AInsulin resistance → decreased ETC Complex I/IV proteinsOXPHOS-Glycolysis ShiftOXPHOS↓, Glycolysis↑[97]Microbiome+LipidomicsUPLC-MS/MS, 16S rRNA SequencerN/AGut microbiota-derived bile acidsFXR/TGR5-AMPK/PGC-1αFAO↑, Mt biogenesis↑[85]Metabolomics (NMR)Bruker NMR SpectrometerN/AUrea cycle metabolites (ornithine, citrulline)ARG2-Mitochondrial ROSUrea cycle disruption → Mt biogenesis and ROS affected[69,70]Notes: ↑, increased or upregulated; ↓, decreased or downregulated. Table 5Integrated Multi-Omics and Gut Microbiome – Potential Biomarkers and MechanismsTechnologyPlatformClinically Used Biomarkers/TestsPreclinical Biomarkers (Not Clinically Used)Related PathwaysMechanismsReferencesGenomics+Transcriptomics+Proteomics+MetabolomicsCytoscape, KEGG, Illumina NovaSeqN/AELMO1 SNPs → decreased CPT1A/ACADM → increased acylcarnitinesFAO-OXPHOS CrosstalkFAO pathway disruption → Lipotoxicity↑[8]Transcriptomics+Metabolomics+Spatial OmicsSeahorse Analyzer, MALDI-MSIProteinuria, eGFRBranched-chain amino acid (BCAA) catabolic defects, PKM2PKM2-OXPHOSOXPHOS↓, Glycolysis↑[91–93]Microbiome+MetabolomicsIllumina NovaSeq 6000, Bruker UHPLC-QTOFShort-chain fatty acids (SCFAs), TMAOPhenyl sulfate (PS), Indoxyl sulfate (IS), Indole-3-propionic acid (IPA)NLRP3, SIRT1/PGC-1αInflammasome activation /Mt biogenesis↑[68,80–83]Transcriptomics+ProteomicsThermo Orbitrap, Illumina HiSeqN/AInsulin resistance → decreased ETC Complex I/IV proteinsOXPHOS-Glycolysis ShiftOXPHOS↓, Glycolysis↑[97]Microbiome+LipidomicsUPLC-MS/MS, 16S rRNA SequencerN/AGut microbiota-derived bile acidsFXR/TGR5-AMPK/PGC-1αFAO↑, Mt biogenesis↑[85]Metabolomics (NMR)Bruker NMR SpectrometerN/AUrea cycle metabolites (ornithine, citrulline)ARG2-Mitochondrial ROSUrea cycle disruption → Mt biogenesis and ROS affected[69,70]Notes: ↑, increased or upregulated; ↓, decreased or downregulated. Integrated Multi-Omics and Gut Microbiome – Potential Biomarkers and Mechanisms Notes: ↑, increased or upregulated; ↓, decreased or downregulated. Treatment Advances in multi-omics technologies have further deepened our understanding of the mechanisms underlying current treatments for DKD. In addition to their established hypoglycemic or hemodynamic effects, integrated transcriptomic, proteomic, metabolomic, and spatial omics analyses indicate that various drugs also regulate mitochondrial dysfunction by restoring mitochondrial biosynthesis, fatty acid β-oxidation, oxidative phosphorylation, mitochondrial autophagy, and redox homeostasis. Renin-angiotensin-aldosterone system inhibitors (RAASi) have been consistently demonstrated in multiple studies to significantly attenuate both functional and structural kidney damage in individuals with DM.99 The RENAAL trial (losartan), the Collaborative Study Group trial (captopril), the MICRO-HOPE substudy (ramipril), and the IDNT trial (irbesartan) demonstrated that ARBs and ACEi significantly slow the progression of DKD by reducing renal function decline and the risk of adverse kidney outcomes.100 Also, a study demonstrates that losartan treatment modulates key mitochondrial processes in renal tubular cells, thereby contributing to the amelioration of DKD.101 SGLT2i demonstrate kidney-protective effects in DKD, and these renoprotective actions are mediated through both receptor-dependent and receptor-independent mechanisms. Multi-omics studies consistently demonstrate that SGLT2 inhibitors improve mitochondrial integrity by alleviating mitochondrial fragmentation, restoring mitochondrial biogenesis and mitophagy, enhancing oxidative phosphorylation, and suppressing oxidative stress through pathways including AMPK/mTOR, Nrf2/HO-1, and nitric oxide signaling.102–106 These mitochondrial protective effects are now considered important contributors to their renoprotective benefits beyond glucose lowering. Nonsteroidal selective MRAs belong to a new category of drugs that provide cardio-renal protection. Metabolomics analysis revealed that the nonsteroidal MRA finerenone attenuates mitochondrial dysfunction, oxidative stress, apoptosis, mitochondrial biogenesis and ameliorates tubular injury through the same PI3K/Akt/eNOS pathway, thereby improving DKD.107 These findings provide mechanistic support for the clinical efficacy observed in patients with DKD. GLP-1 receptor agonists have shown considerable potential for treating DKD. Their therapeutic benefits are achieved through multiple mechanisms, including enhanced blood glucose management, promotion of weight loss, reduction of inflammation, and regulation of kidney blood flow.20 Spatial metabolomics and single-cell transcriptomics studies demonstrate that GLP-1 agonists regulate lipid metabolism through activation of the PI3K/AKT signaling pathway and upregulation of PPAR-α and CPT1 expression, promoting fatty acid oxidation, reducing lipid accumulation, and alleviating mitochondrial dysfunction and renal injury in DKD.9 The FLOW trial (semaglutide) and the LEADER trial (liraglutide) demonstrated that these GLP-1 receptor agonists provide significant benefits for diabetic kidney disease by slowing the decline in kidney function and reducing the risk of major cardiovascular events.108,109 Active Ingredients of Traditional Chinese Medicine To fully leverage the systemic nature of herbal medicine, multi-omics approaches are pivotal for decoding the synergistic and network-based therapeutic effects of its active components. This is powerfully exemplified by studying compounds like Astragaloside IV, Arctigenin, Berberine, and Poricoic acid A, providing a holistic understanding of their intervention in DKD. The active ingredients in them are the key components for treating DKD. This real-world study, analyzing data from 86,960 patients with overall DKD and 2844 with advanced DKD, found that CHMs provide time-dependent renal and survival benefits.110 Astragalus root (Astragalus membranaceus (Fisch.) Bunge) is one of the most widely prescribed herbs for DKD, and its major bioactive constituent, astragaloside IV (AS-IV), has shown considerable renoprotective effects.111 Integrative molecular docking and functional genomics revealed that Astragaloside IV attenuates mitochondrial dysfunction and podocyte apoptosis in DKD by activating the Nrf2/TFAM signaling pathway.112 Proteomics revealed that astragaloside IV ameliorates DKD by modulating the mitochondrial quality control network, suppressing Drp-1 and PINK1/Parkin signaling.113 In addition, integrated proteomic and lipidomic analyses demonstrate that AS-IV inhibits the HIF-1α/HMOX1 signaling pathway, suppresses ferroptosis, and restores mitochondrial metabolic homeostasis by increasing GPX4 expression and regulating ACSL4.114 Fructus Arctii is traditionally recognized for dispelling wind-heat, enhancing detoxification, and alleviating sore throat symptoms, while laboratory investigations have demonstrated that it possesses a range of pharmacological properties, including antioxidative effects, antibacterial and antifungal capacities, blood glucose-lowering abilities, lipid-regulating functions, anti-inflammatory actions, pain-relieving activities, antiviral potentials, and anti-cancer properties.115 Arctigenin (ATG) is the main component of Fructus Arctii, and ATG was found to bind and activate protein phosphatase 2A (PP2A) through integrated transcriptomic and proteomic analyses. This activation suppresses NF-κB-mediated inflammatory signaling and stabilizes the podocyte cytoskeleton by promoting dephosphorylation of drebrin-1 (DBN1), with additional benefits through AMPK activation which improves mitochondrial function and reduces oxidative stress, leading to amelioration of DKD.116 Berberine, the major alkaloid isolated from Coptis chinensis, has been extensively investigated as a mitochondrial-targeted therapeutic agent. Metabolomic studies demonstrate that berberine activates the AMPK/PGC-1α signaling pathway, enhances mitochondrial fatty acid β-oxidation, reduces lipid accumulation and reactive oxygen species production, and restores podocyte energy metabolism.117 Integrated transcriptomic analyses demonstrated that berberine exhibits strong binding affinity and stability to EZH2, and suggested that berberine may modulate ferroptosis in DKD by suppressing the SLC7A11 pathway, thereby attenuating lipid peroxidation and potentially alleviating mitochondrial dysfunction.118 Poria cocos has long been esteemed in traditional applications for its diuretic action, ability to dispel dampness, spleen-invigorating effect, and mind-calming property. Poricoic acid A (PAA), a major triterpenoid component found in the surface layer of Poria cocos (SPLC), is recognized as the core bioactive constituent responsible for its renoprotective effects. Metabolomic analyses indicate that PAA alleviates podocyte injury and renal fibrosis by restoring mitophagy and regulating mitochondrial dynamics, partially through suppression of FUNDC1-mediated mitochondrial stress pathways119 (Figure 2). Figure 2Application of multi-omics in drug discovery and drug development.10 multi-omics aids in identifying new targets, facilitating drug screening. Multi-omics holds immense potential in drug repurposing and natural product drug discovery. Multi-omics utilizes expression maps after drug treatment to identify key genes and predict treatment pathways. Multi-omics constructs a drug action atlas, enabling the localization of drug effects and the study of drug resistance and interactions. By analysing the cell composition of pharmacological models, assessing model similarity to cases, and studying intercellular interactions, multi-omics assists in building pharmacological research models. Multi-omics is being applied across various stages of drug discovery and development, and these applications are enhancing the probability of drug discovery success and improving therapeutic outcomes.The first section, ′Drug discovery′, includes target discovery, drug screening, drug repurposing and nature products. The second section, ′Pathway prediction′, involves treatment, multi-omics and pathway prediction. The third section, ′Drug action atlas′, covers therapy effect location, drug resistance and interaction network. The fourth section, ′Pharmacological model construction′, includes cellular composition, similarity assessment and cellular interaction. Below, a flowchart shows ′Drug discovery′ leading to ′Disease understanding′, ′Target discovery′ and ′Drug screening′. This connects to ′Drug development′, which includes ′Drug pathway′, ′Drug action network′, ′Pharmacological model construction′ and ′Precision medicine′.A diagram illustrating stages in drug discovery and development using multi-omics. Application of multi-omics in drug discovery and drug development.10 multi-omics aids in identifying new targets, facilitating drug screening. Multi-omics holds immense potential in drug repurposing and natural product drug discovery. Multi-omics utilizes expression maps after drug treatment to identify key genes and predict treatment pathways. Multi-omics constructs a drug action atlas, enabling the localization of drug effects and the study of drug resistance and interactions. By analysing the cell composition of pharmacological models, assessing model similarity to cases, and studying intercellular interactions, multi-omics assists in building pharmacological research models. Multi-omics is being applied across various stages of drug discovery and development, and these applications are enhancing the probability of drug discovery success and improving therapeutic outcomes. Discussion The integration of multi-omics technologies, including spatial omics, genomics, transcriptomics, proteomics, metabolomics, and microbiome-associated omics, has fundamentally reshaped our understanding of mitochondrial dysfunction in DKD. Rather than representing isolated pathological events, mitochondrial abnormalities constitute an interconnected regulatory network involving metabolic reprogramming, impaired mitochondrial quality control, oxidative stress, inflammatory activation, and altered mitochondrial dynamics. Collectively, these mechanisms drive progressive renal injury across different renal cell types and disease stages, establishing mitochondrial dysfunction as a central pathogenic hub in DKD. One of the major advances provided by multi-omics technologies is the ability to characterize mitochondrial dysfunction from both spatial and temporal perspectives. From a spatial perspective, spatial omics technologies have revealed compartment-specific disruptions in mitochondrial function. For example, podocyte injury is closely associated with ceramide accumulation, VDAC1 oligomerization, and cGAS–STING activation, whereas injured proximal tubular cells exhibit impaired fatty acid oxidation, enhanced glycolysis, and defective mitochondrial respiration, ultimately promoting tubulointerstitial fibrosis. In addition, spatial metabolomics has identified region-specific lipid remodeling and mitochondrial metabolic disturbances in the renal medulla, highlighting the importance of tissue microenvironment in regulating mitochondrial function. From a temporal and disease progression perspective, multi-omics analyses delineate the evolution of mitochondrial dysfunction throughout DKD. In early stages, transcriptomic and metabolomic profiling reveal upregulated glycolysis and impaired fatty acid oxidation, accompanied by accumulation of long-chain acylcarnitines and alterations in TCA cycle intermediates, indicating a shift in energy substrate utilization. As DKD advances, persistent hyperglycemia and oxidative stress lead to epigenetic dysregulation, mitochondrial hyperacetylation, and defective mitophagy, exacerbating mitochondrial fragmentation, bioenergetic failure, and renal cell senescence. This progression is marked by a transition from adaptive metabolic shifts to maladaptive mitochondrial decay, which in turn accelerates renal functional decline. Genomic and proteomic studies further identify genetic variants and post-translational modifications that accelerate mitochondrial decay, reinforcing mitochondrial dysfunction as both a cause and consequence of DKD progression. Concurrently, gut microbiome research, integrated via metabolomics, highlights that microbial metabolites (eg, TMAO, IPA, SCFAs) can either exacerbate or ameliorate DKD by directly influencing mitochondrial energetics, oxidative stress, and inflammatory signaling. Multi-omics integration provides a powerful framework to synthesize these spatiotemporal insights. By correlating data across omics layers, the research can summarize the sequence of mitochondrial defects—from genetic susceptibility and transcriptional reprogramming to protein-level modifications and metabolic disturbances—across different renal cell types and disease stages. This holistic view reveals how mitochondrial dysfunction serves as a converging mechanism in DKD pathogenesis. The strength of multi-omics research lies not in a single molecular dataset, but in integrating multiple biological levels into a unified regulatory network. Genomics identifies genetic susceptibilities; transcriptomics captures dynamic mechanisms of gene regulation; proteomics reflects changes in protein function; metabolomics reveals characteristics of metabolic remodeling; and spatialomics preserves information on tissue structure and cell-cell interactions. The mechanistic insights provided by this systems-level research perspective cannot be achieved by any single omics platform alone. These advances also have important therapeutic implications. The American Diabetes Association (ADA) guidelines recommend that for DKD, first-line drugs include Angiotensin-converting enzyme inhibitors (ACEI) /angiotensin II receptor blockers (ARB) for patients with hypertension or albuminuria, SGLT2i and GLP-1RAs for those with type 2 diabetes, and nonsteroidal mineralocorticoid receptor antagonists (MRA) as add-on therapy to ACEI/ARB for renoprotection. Multi-omics analyses have clarified how these therapies restore mitochondrial quality control, improve oxidative phosphorylation and fatty acid oxidation, reduce oxidative stress, and suppress inflammatory signaling. Similarly, multi-omics elucidates the mechanisms of active components from traditional Chinese medicine—such as astragaloside IV, arctigenin, berberine, and poricoic acid A—which are shown to preserve mitochondrial function, reduce ferroptosis, suppress inflammation, ameliorate fibrosis, and counteract oxidative stress through pathways involving AMPK/PGC-1α, HIF-1α, NF-κB, and others. These findings not only explain their molecular mechanisms but also facilitate the identification of novel mitochondrial-targeted therapeutic strategies and predictive biomarkers, providing an important foundation for precision medicine in DKD. (Figure 3). Figure 3Multi-omics in personalized medicine.10 ST identifies biomarkers of drug response by analysing the multi-omics information of different drug-responsive populations. The spatial transcriptome was used to analyse the spatial expression information of patient biomarkers to achieve personalized medicine.The first diagram illustrates the therapy process for patients, leading to effects, no effects, or adverse effects. These outcomes connect to multi-omics analysis for biomarker identification. The second diagram shows patients undergoing multi-omics analysis, including metabolomics, transcriptomics, proteomics, genomics and spatialomics, leading to personalized medicine. Both diagrams emphasize the role of multi-omics in understanding therapy outcomes and tailoring personalized medicine approaches.Two diagrams showing multi-omics in therapy and personalized medicine processes. Multi-omics in personalized medicine.10 ST identifies biomarkers of drug response by analysing the multi-omics information of different drug-responsive populations. The spatial transcriptome was used to analyse the spatial expression information of patient biomarkers to achieve personalized medicine. Despite these advances, significant knowledge gaps persist. The precise hierarchy and temporal sequence of mitochondrial defects across different renal cell types remain incompletely mapped. While multi-omics powerfully identifies associative networks, establishing definitive causality requires rigorous functional validation in cell-type-specific models. Moreover, the integrated application of multi-omics platforms—particularly combining spatial, temporal, and single-cell data—remains limited, hindering the construction of dynamic, multi-layered regulatory networks. The translational potential of identified biomarkers and targets is also tempered by technical and cohort-related challenges, including sample heterogeneity, batch effects, and the complexity of integrating high-dimensional datasets from diverse omics layers. Future research should prioritize longitudinal multi-omics studies in well-phenotyped human cohorts to delineate the spatiotemporal evolution of mitochondrial dysfunction in DKD. The functional interrogation of candidate pathways using advanced in vitro and in vivo models—such as organoids, conditional knockouts, and spatial functional genomics—is essential to move from correlation to causation. Furthermore, developing strategies to therapeutically target the mitochondrial proteome and metabolome, potentially by leveraging insights from both synthetic and natural compounds, represents a critical next step. Finally, the implementation of machine learning and network medicine approaches will be key to integrating multi-omics data, identifying robust biomarkers, and achieving precision nephroprotection in DKD.