Phosphorylation of proteins is a key PTM for balancing glucose homeostasis, and relies on signaling cascades mediated by protein kinases and phosphatases. Unc-51-like kinase 1 (ULK1), a key serine/threonine protein kinase in autophagy, is modulated by AMPK and mTOR (56). The activated AMPK pathway and activated autophagy protect numerous types of renal cells in response to high glucose by directly activating ULK1 phosphorylation (57-59). Augmented phosphorylation levels of PI3K/Akt/mTOR induce renal apoptosis, glomerular injury and interstitial fibrosis by inhibiting ULK1 phosphorylation (60,61). The MAPK family is associated with the progression of DKD, and consists of p38 MAPK, JNK and ERK, which are involved in apoptosis of renal intrinsic cells in DKD (62-66).

Glycogen synthase kinase 3β (GSK3β) is a highly conserved, redox-sensitive serine/threonine protein kinase. Exposure to the diabetic condition leads to elevated expression of phosphorylated (p)-GSK3β at tyrosine 216 (p-GSK3β^Y216), but a decrease in the expression of inhibitory p-GSK3β at serine 9 (67,68). PI3K/Akt signaling inactivates GSK3β by phosphorylating GSK3β at serine 9 (69). Nrf2 is a key regulator in redox balance that can translocate into the nucleus and initiate the transcription of antioxidant genes, such as heme oxygenase-1 (HO-1) (70). Activated GSK3β directly phosphorylates Nrf2 and leads to its nuclear exclusion (71). Additionally, GSK3β phosphorylates Nrf2 at the Tyr568 site and promotes the translocation of Nrf2 out of the nucleus by increasing Fyn phosphorylation and its nuclear retention (71). Overactivated GSK3β phosphorylates Nrf2, thereby facilitating podocyte apoptosis and senescence in DKD (67,72). In addition, enhanced p-GSK3β^Y216 promotes phosphorylation and degradation of β-catenin, mediating podocyte apoptosis under diabetic conditions (73).

In addition, alteration of phosphorylation states of specific proteins participates in the pathogenesis of DKD. Activation of the JAK/STAT signaling pathway can affect cell senescence, autophagy, apoptosis and ferroptosis in DKD (74-76). The phosphorylation state of STAT3 is also modulated by diverse regulators. In renal tubular epithelial cells (RTECs), sulfhydrated SIRT1 induces dephosphorylation and deacetylation of STAT3, thereby decreasing high glucose-induced cell apoptosis, oxidative stress, inflammation response and EMT progression (77). Furthermore, increasing numbers of phosphorylated proteins have been shown to participate in the damage of renal cells in DKD (Table I) (78-87). This research contributes to a more comprehensive understanding of the pathogenesis of DKD and provides a theoretical basis for considering PTMs as a therapeutic target for DKD.

HAT p300/CBP is a key co-transcriptional activator that regulates the expression of numerous prooxidant, proinflammatory and profibrotic genes by mediating H3K27 acetylation, and is associated with mediation of diabetes-accelerated renal damage (88). Numerous targets of p300/CBP participate in the regulation of DKD development, including inducible nitric oxide synthase (NOS) and polo-like kinase 1 (PLK1) (89,90). ACSS2 epigenetically activates Raptor expression by histone H3K9 acetylation, promoting activation of the mTOR complex (mTORC)1 pathway and inhibiting podocyte autophagy (91). Upregulated acetylation levels of Beclin1 are associated with exacerbation of podocyte injury in diabetic conditions and a mutation at K414R suppresses hyperactivated autophagy, thus ameliorating podocyte impairment, highlighting the key role of PTMs in the pathological process of DKD (92).

Multiple HDACs regulate the activities of numerous proteins involved in cell death during the progression of DKD. Specifically, inhibition of transcription factor EB (TFEB) deacetylation induced by HDAC6 promotes TFEB activation and alleviates RTEC damage (93,94). Moreover, HDAC6-mediated deacetylation of α-tubulin suppresses autophagy and enhances motility of podocytes in DKD (95). HDAC4 suppresses podocyte autophagy and promotes podocyte apoptosis by acetylating STAT1 under diabetic conditions (96). SIRT1 activation can deacetylate forkhead box (Fox)O3a or regulate the NF-κB signaling pathway, while subsequently suppressing renal cell apoptosis (97-99). SIRT3 upregulation also antagonizes hyperglycemia-mediated tubular apoptosis by regulating the accumulation of ROS and modulating the ROS-sensitive Akt/FoxO signaling pathway (100). In parallel, SIRT6 inhibits the transcription of Notch1/4 by deacetylating histone H3K9, and protects podocytes against apoptosis and inflammation by enhancing autophagic flux (101). These results indicate that HDACs serve as potential therapeutic targets in DKD.

Recruitment of EZH2, a methyltransferase that produces histone H3 lysine 27 tri-methylation (H3K27me3), at the FoxA1 promoter region promotes podocyte damage and apoptosis in the early stage of DKD (102,103). Inhibition of EZH2 by GSK126 attenuates podocyte injury and hyperglycemia-induced ferroptosis (104,105). By contrast, EZH2/H3K27me3 mitigates podocyte pyroptosis in DKD by increasing early growth response protein 1 (EGR1) in type 1 diabetic nephropathy (T1DN) (106). Moreover, AGEs induce downregulation of EZH2 in podocytes and decrease H3K27me3, which in turn leads to the upregulated expression of pathological factors (such as TGF-β1 and SNAI1) that contribute to podocyte injury in late DKD (107). KDM6A, a HDM that removes the di- and tri-methyl groups from histone H3K27, is activated in patients with DKD and mice with T1DN (108). Suppression of KDM6A by administration of GSK-J4 ameliorates the early lesions of db/db mice, including renal dysfunction, mesangial matrix accumulation, inflammation and apoptosis (109). The aforementioned studies indicate that levels of H3K27 are affected by multiple factors, such as cell types and environment, and the stage of DKD, thereby changing the transcription of its target genes.

Numerous altered ubiquitination statuses of proteins are associated with podocyte dysfunction. Specifically, TRIM29 promotes podocyte pyroptosis by interacting with inhibitor of NF-κBα (IκBα) to mediate its ubiquitination-dependent degradation, which triggers NF-κB activation (21). In addition, TRIM63 regulates PPARα ubiquitination and degradation, contributing to podocyte injury (110). The E3 ligase c-Cbl binds to podocin and increases the ubiquitination of podocin, leading to podocyte injury in DKD (111). Moreover, the deubiquitination of sperm-associated antigen 5 by USP14 activates Akt/mTOR signaling, exacerbating high glucose-induced autophagy and podocyte injury (112). The ubiquitination state of the receptor-interacting protein kinase (RIPK)1/RIPK3 pathway downregulated by ubiquitin C-terminal hydrolase L1, a crucial member of the deubiquitination family, serves a key role in podocyte necroptosis and apoptosis (113).

Lysine 63 ubiquitination (Lys63-Ub) is increased in the RTECs of DKD, and associated with autophagy deregulation and apoptosis activation (114). The apoptosis of RTECs is prevented by eukaryotic translation initiation factor 2α ubiquitination and degradation mediated by HMG-CoA reductase degradation 1 (115). Additionally, ARAP1 maintains persistent epidermal growth factor receptor (EGFR) activation by decreasing EGFR ubiquitination and subsequently activates TGF-β/Smad3 and hypoxia inducible factor 1α (HIF-1α) signaling, causing injury of RTECs and mesangial cells (MCs) (116,117). The aforementioned studies demonstrate that targets of ubiquitination, such as Lys63-Ub, may be a promising direction for DKD therapy.

SUMOylation and deSUMOylation serve vital roles in the pathogenesis of numerous nephropathic diseases (118-120). In DKD, Kruppel-like factor (KLF)15 upregulates the expression of SUMO1 and enhances the SUMOylation of P53, subsequently inhibiting the proliferation of MCs (121). However, the role of SUMOylation in other aspects of DKD has not yet been clarified.

Numerous studies have confirmed an increased number of O-GlcNAc-positive cells in the glomeruli and enhanced staining in the tubules in DKD (122,123). Hyperglycemia-driven elevation of O-GlcNAc modification contributes to DKD progression via the inhibition of Akt phosphorylation, thus abnormally activating endothelial NOS (eNOS) (124). A mutually reinforcing cycle is formed between activation of O-GlcNAcylation and the intrarenal renin-angiotensin-aldosterone system (RAAS), exacerbating glucose toxicity, while this effect is inhibited by RAAS blockers via increasing OGA levels (124). Akimoto et al (125) found that aberrant O-GlcNAcylation of α-actinin 4 and actin impairs cytoskeletal structure and adhesive function, leading to morphological changes in podocyte foot processes. This may trigger the early damage to the glomerular filtration barrier in DKD. Pharmacological inhibition of O-GlcNAcylation by OSMI1, an inhibitor targeting OGT, decreases podocyte apoptosis under diabetic conditions (126). Notably, O-GlcNAc modification exhibits a bidirectional paradox in tubular epithelial cells. During prolonged fasting conditions in RTECs, it is essential for renal lipolysis and exerts a protective effect against lipotoxicity (127). However, O-GlcNAcylation disrupts the homeostasis of retinol signals in HK2 cells (an immortalized proximal tubule epithelial cell line from normal adult human kidney) by extensively modifying the key molecules in the retinol signaling pathway, such as signaling receptor and transporter of retinol 6 and retinaldehyde dehydrogenases 1 (128). Moreover, diabetic conditions promote OGT-mediated O-GlcNAcylation of acyl-CoA synthetase long chain family member 4 (ACSL4), thereby stabilizing ACSL4 and facilitating tubular ferroptosis in DKD (129). This functional discrepancy may be attributed to modification targets, cell type and microenvironmental metabolic status. Additionally, as a subtype of glycosylation, inhibition of core fucosylation mitigates renal pathological changes, renal fibrosis and podocyte injury by downregulating the phosphorylation of Smad2/3 and ERK (130). Targeting protein O-GlcNAcylation may be a promising therapeutic approach for rescuing DKD progression.

Glutathione (GSH) is a primary endogenous antioxidant (131). Protein glutathionylation is a key PTM involved in the pathogenesis of complex diseases, including cardiovascular disease, acute lung injury and renal disease (132-134). Evidence regarding functional changes induced by redox modifications as well as the increasing number of glutathionylated proteins indicate a role for glutathionylation in DKD (135,136). Notably, in patients with type 2 diabetes and microangiopathy, the levels of glutathionylated hemoglobin are markedly elevated whereas GSH levels are decreased (136). In normal podocytes, myosin 9A (Myo9A), Ras homolog family member A (RhoA) and actin are S-nitrosylated while the diabetic microenvironment induces the de-nitrosylation of Myo9A, actin and RhoA, causing increased RhoA activity and impaired podocyte migration (137). Moreover, oxidative modifications of serum albumin in patients with DKD not only trigger the activation of neutrophils but also result in inaccurate measurement of serum albumin (138). The aforementioned studies demonstrate that redox modification not only participates in the pathogenesis of DKD, but also affects the measurements of biochemical indexes.

Lysine succinylation is a naturally occurring PTM that alters the stability and function of substrate proteins. This modification is modulated by enzymes such as SIRT5 and serves a pivotal role in the pathogenesis of DKD. For example, upregulation of SIRT5 inhibits the succinylation of NIMA-related kinase 7 (NEK7), disrupts the interaction between NEK7 and NOD-like receptor protein 3 (NLRP3), and suppresses podocyte pyroptosis and oxidative stress-induced injury under hyperglycemia (139). However, whether protein succinylation is involved in the cell cycle of other renal cells remains unclear.

The PI3K/Akt pathway modulates FoxO3a activity by phosphorylating its three residues (Thr32, Ser253 and Ser315) and excludes FoxO3a from the nucleus (140). Klotho prevents podocyte injury against palmitate-induced oxidative stress by decreasing the phosphorylation of FoxO3a, promoting its nuclear translocation and upregulating the expression of antioxidants, including manganese superoxide dismutase (MnSOD) (140). However, in streptozotocin (STZ)-induced rats, activation of the PI3K/Akt/FoxO3 pathway alleviates inflammation and oxidative stress in MCs (141). Furthermore, activation of the AMPK/SIRT1/PPARγ coactivator 1α (PGC1α) pathway mitigates oxidative stress in db/db mice (142). Additionally, GSK3β knockdown enhances the antioxidant response driven by Nrf2 and suppresses oxidative stress, leading to the alleviation of oxidative stress and podocyte injury (67). Specifically, the elevated phosphorylation of GSK3β at serine 9 reduces the protein stability of GSK3β and diminishes its inhibitory effect on Nrf2, protecting podocytes and RTECs from oxidative stress (67,143). The insulin receptor (IR) is key for insulin action. Protein tyrosine-phosphatase 1B (PTP1B) attenuates insulin signaling by dephosphorylating IR and IR substrate 1, and knockdown of PTP1B protects podocytes from ER stress by improving insulin sensitivity (144). Targeting protein kinases and phosphatases may provide more options for DKD treatment.

Members of the SIRT family serve a pivotal role in DKD pathogenesis. Activation of the SIRT1/PGC1α/Nrf1 and SIRT1/NF-κB pathways effectively attenuates DKD, induces autophagic flux, mitigates oxidative stress and alleviates mitochondrial dysfunction in podocytes (77,145,146). Moreover, SIRT1 can interact with ACSS2. ACSS2 promotes mitochondrial oxidative stress and triggers renal tubular inflammation in DKD by modulating the SIRT1/carbohydrate responsive element binding protein (ChREBP) pathway (147). Polysulfides can attenuate diabetic renal pathological lesions via inactivation of p65 NF-κB and STAT3 phosphorylation/acetylation by sulfhydrating SIRT1, thereby reducing high glucose-induced oxidative stress, cell apoptosis, inflammation and EMT (77). Furthermore, the enhancement of mitochondrial oxidative stress in DKD is mediated by the reduction of SIRT3 activity and a subsequent increase in acetylated isocitrate dehydrogenase 2/SOD2 (148). General control of amino acid synthesis 5-like 1-mediated acetylation of MnSOD also exacerbates oxidative stress-induced renal injury in DKD (149). The aforementioned studies demonstrate the key role of the SIRT family in response to oxidative stress in DKD.

Numerous alterations in the levels of proteins involved in oxidative stress are regulated by the alteration of histone methylation status. Expression of monocyte chemoattractant protein-1 (MCP-1) is elevated due to the recruitment of SET7/9 and H3K4me1 to its promoters in db/db mice (150). Decreased levels of HIF-1α suppress oxidative stress and inflammation via HDM KDM3A in human umbilical vein endothelial cells exposed to hypoxia and diabetic conditions (151). Siddiqi et al (152) demonstrated that inhibition of EZH2 with DZNep can increase podocyte injury, oxidative stress and proteinuria in diabetic rats. Moreover, S-adenosylhomocysteine hydrolase inhibition induces podocyte injury and oxidative stress via the EZH2/EGR1/thioredoxin-interacting protein/NLRP3 signaling cascade in STZ-induced T1DN mice (153). At present, the fundamental mechanism by which EZH2 serves different roles in various DKD models and cells remains unclear. Due to the involvement of multiple cell types in the pathogenesis of DKD and the cell context-specific gene regulation mediated by EZH2, EZH2 may serve a dual role in different renal cells in DKD (154).

Hyperglycemia prompts Von Hippel-Lindau tumor suppressor E3 ubiquitin ligase to ubiquitinate glucose-6-phosphate dehydrogenase, leading to ROS production and podocyte injury (155). Moreover, under diabetic conditions, sustained activation of PH domain and leucine-rich repeat protein phosphatase 1 promotes the nuclear retention of FoxO1 via prevention of its ubiquitination, inducing aberrations in renal gluconeogenesis and the activation of the apoptotic cascade, and exacerbating oxidative stress in diabetic rats (156). Oxidative stress and apoptosis are decreased in RTECs by promoting SIRT1 expression via suppressing its ubiquitination (157). Moreover, the phosphorylation of c-Cbl at Tyr731 facilitates the combination of c-Cbl and SIRT1, which triggers polyubiquitination of SIRT1 by c-Cbl and promotes SIRT1 degradation, decreasing the antioxidant effects of FoxO3a in DKD (158). Connexin32 (Cx32) upregulates SIRT1 expression by inhibiting the ubiquitination of Lys335 of SIRT1 by suppressing Smad specific E3 ubiquitin protein ligase 1 (Smurf1), thus alleviating oxidative stress in DKD (157). In addition, Cx32 decreases renal oxidative stress levels and ameliorates the pathological progression of diabetic renal fibrosis by promoting NADPH oxidase 4 polyubiquitination and degradation (159). Nrf2 is an important antioxidant in response to oxidative stress. Suppression of the ubiquitination of Nrf2 ameliorates experimental DKD through antioxidation and regulation of the Keap1/Nrf2 signaling pathway (160,161). Additionally, ubiquitination participates in ferroptosis induced by oxidative stress. Ginkgolide B alleviates oxidative stress and ferroptosis by inhibiting glutathione peroxidase (GPX)4 ubiquitination to improve changes in renal structure in mice with DKD (162). Stimulator of interferon genes protein inhibition alleviates ferroptosis and oxidative stress in DKD via stabilization of ferroportin 1 (FPN1) protein levels by decreasing FPN1 ubiquitination for proteasomal degradation (163). In addition, 2-deoxy-d-ribose induces ferroptosis in RTECs by degrading the cystine/glutamate antiporter SLC7A11 protein via the UPS, resulting in decreased intracellular cystine uptake (164). These findings underscore the key role of ubiquitination and deubiquitination in DKD development, and highlight the potential therapeutic targets.

In early stage DKD, ectonucleoside triphosphate diphosphohydrolase 5 (ENTPD5), a nucleotide hydrolase located in the ER, modulates the N-glycosylation of proteins and facilitates renal cell proliferation (165). Separately, in late stage DKD, sustained hyperglycemia activates the hexosamine biosynthesis pathway to increase the levels of UDP-GlcNAc, which triggers a feedback mechanism that suppresses transcription factor SP1 activity and downregulates ENTPD5 expression, aggravating ER stress (165). Additionally, O-GlcNAcylation of the mineralocorticoid receptor directly enhances both the protein abundance levels and transcriptional activity of the receptor under diabetic conditions (166). These findings provide novel directions for the diagnosis and targets of DKD.

Hypoglutathionemia and elevated oxidative stress levels contribute to the early biochemical abnormalities in diabetes (167). Additionally, reactive carbonyl derivate levels increase in patients with diabetes, and this elevation is more pronounced in diabetic patients undergoing hemodialysis, suggesting that both diabetic state and hemodialysis contribute to the enhancement of protein oxidation (168).

A previous study showed that K99 succinylation of hydroxysteroid 17β dehydrogenase 10 (HSD17B10) maintains mitochondrial RNA ribonuclease P (RNase P) stability (169). Astragaloside IV alleviates hyperglycemia-induced oxidative stress and mitochondrial dysfunction in HK2 cells by upregulating carnitine palmitoyltransferase 1A-mediated K99 succinylation of HSD17B10 to maintain RNase P activity (169).

Extensive alterations of cell signaling pathways serve a broad role in maintaining mitochondrial homeostasis and optimizing oxidative phosphorylation. Inactivated AMPK and PI3K/Akt signaling and abnormal activation of the JNK pathway participate in mitochondrial dysfunction under diabetic conditions, modifying mitochondrial dynamic homeostasis and energy metabolism disorder in DKD (170-174). Finerenone treatment can reduce mitochondrial fragmentation and restore mitophagy via PI3K/Akt/eNOS signaling in HK2 cells exposed to diabetic conditions and tubular cells of mice with DKD (172). In addition, elevated SIRT1 phosphorylation at Ser47 is associated with mitochondrial dysfunction in podocytes (175). Inhibiting SIRT1 phosphorylation-mediated ubiquitin-proteasome degradation restores the capacity of SIRT1 to promote PGC1α deacetylation and nuclear translocation, and thereby upregulates genes associated with mitochondrial biosynthesis and antioxidant defense in DKD (176).

Phosphorylation of multiple proteins causes disruption of mitochondrial dynamics and leads to loss of mitochondrial voltage potential under high-glucose conditions. Dynamin-related protein 1 (Drp1) is a key regulator of mitochondrial fission. Cyclin-dependent kinase 5 phosphorylates Drp1 at Ser616 and thus produces excessive ROS, leading to EMT progression in HK2 cells (177). Enhanced phosphorylation of Drp1 at Ser637 (p-Drp1^Ser637) by Rho-associated coiled coil-containing protein kinase 1 promotes the transposition of Drp1 to the mitochondrial surface and accounts for excessive mitochondrial fission in mouse podocytes (178-180). By contrast, resveratrol decreases Drp1 levels while increasing p-Drp1^Ser637 levels, blocking mitochondrial fission in MCs by inhibition of phosphodiesterase-4D/protein kinase A signaling (181). The functional consequences of serine phosphorylation of Drp1 may be dependent on cell type and stimulation. Mitophagy is a specialized form of autophagy that mediates the selective elimination of damaged or dysfunctional mitochondria (182). The phosphorylated form of FUN14 domain-containing 1 (FUNDC1) inhibits the induction of mitophagy by blocking the interaction between the FUNDC1 LC3-interacting region and LC3 (183,184). In addition, activated Src kinase serves as a negative modulator of mitophagy in DKD by inducing the phosphorylation of FUNDC1 at Tyr18, which impairs the ability of podocytes to clear damaged mitochondria (185). Studies have demonstrated the key role of aberrant protein phosphorylation in the progression of DKD (Table II) (81,186-190). However, further validation at different stages of human DKD are needed to fully understand the impact of PTMs.

Protein acetylation is a key component of diverse metabolic reactions. SIRT1 sustains mitochondrial homeostasis by mediating mitochondrial biogenesis and mitophagy in DKD (191,192). It has been hypothesized that the SIRT3/SOD2/GPX4 signaling pathway participates in the regulation of ferroptosis in DKD via maintenance of mitochondrial redox homeostasis (193). SIRT6 upregulation alleviates mitochondrial dysfunction and podocyte apoptosis via AMPK activation mediated by its deacetylation of H3K9 and H3K56 (194). In addition, activation of SIRT1 exerts a renoprotective role in restoring mitochondrial homeostasis, providing a preclinical research basis for small molecule drugs targeting the SIRT family.

TRIM22, a E3 ubiquitin ligase, is highly expressed in patients with DKD, interacts with optic atrophy 1 and induces its ubiquitination, thus altering mitochondrial fusion-associated proteins involved in respiration/ATP synthesis, influencing ROS production and mitochondrial function in DKD (195). Long non-coding RNA PVT1 is involved in mitochondrial dysfunction by interacting with TRIM56 at the post-transcriptional level to induce AMPK ubiquitination, leading to aberrant mitochondrial biology and increasing mitochondrial DNA leakage in podocytes in DKD (196). Additionally, interference with Smurf2 inhibits both RUNX family transcription factor 3 ubiquitination and the TLR4/NF-κB signaling pathway, which alleviates mitochondrial dysfunction and tubular injury (197). Another E3 ubiquitin ligase, Cullin3, directly interacts with mitochondrial ribosomal protein L12 to induce its ubiquitination, resulting in mitochondrial biosynthesis dysfunction in RTECs (198). However, the effect of ubiquitination on the mitochondrial homeostasis of MCs and endothelial cells needs further investigation.

DeSUMOylation of RNA binding motif protein X-linked serves a key role in determining the microRNA (miRNA/miR) composition of renal cell exosomes, which prevents the protective miRNAs from inhibiting mitochondrial damage in DKD (199).

In chronic diabetes, mitochondrial proteins are susceptible to PTMs triggered by glycation and oxidation. Oxidative and nitrosative stresses promote mitochondrial oxidative dysfunction in STZ-induced diabetic rats (200). Carbonyl-mediated modifications selectively target key protein components of major mitochondrial cycles, including oxidative phosphorylation and fatty acid β-oxidation (201). Methylglyoxal, a dicarbonyl compound that accumulates to high levels in the hyperglycemic environment, exerts an inhibitory impact on both the tricarboxylic acid cycle and the electron respiratory chain (202). Notably, such modifications are specific to certain mitochondrial proteins and trigger disturbances in mitochondria involved in renal cellular toxicity and the progression of DKD (202).

The expression of lysine lactylation is notably elevated in renal tissues from patients with diabetes as well as db/db mice (43,203,204). Lactylation of acyl-CoA synthetase family member 2 at lysine 182 contributes to tubular mitochondrial dysfunction, which accelerates the progression of DKD (203). However, research on the roles of lactylation in other renal cell types is lacking.

NF-κB, interacting with IκB and IκB kinase (IKK), is a key intracellular molecule regulating inflammation and is abnormally activated in DKD (205,206). FoxM1 transcriptionally activates SIRT4, and suppresses phosphorylation of NF-κB (Ser536) and the levels of NLRP3 inflammasome to ameliorate renal damage and podocyte pyroptosis in DKD (207). Inflammation in DKD is decreased by activation of the Nrf2-mediated antioxidant pathway and inhibition of the MAPK-mediated inflammatory pathway, such as ERK1/2, JNK and MAPK (205,208,209). Treatment with salidroside triggers the phosphorylation of Akt and GSK3β; suppressed expression of p-Akt (Ser473) and p-GSK3β (Ser9) inhibits oxidative stress and inflammation in DKD rats (210). Upregulated phosphorylation of SH2 domain-containing protein-tyrosine phosphatase-2 (SHP2) is detected in macrophages in both diabetic patients and mouse models (211,212). Macrophage SHP2 deficiency alleviates DKD via the suppression of MAPK/NF-κB-dependent inflammation, subsequently attenuating renal dysfunction, collagen deposition, fibrosis and inflammatory response in STZ-treated mice (211). Dephosphorylation of STAT3 (Tyr705) ameliorates tubulointerstitial inflammation and glomerulosclerosis in DKD (213,214). Monitoring the dynamic changes of the phosphorylated status of proteins during DKD progression may improve understanding of the pathogenesis of DKD.

High-glucose stimulation increases H3K9/14Ac at the receptor for AGEs, plasminogen activator inhibitor-1 and MCP-1 promoters, performing key roles in rat MCs in DKD, whereas losartan reverses the H3K9/14Ac marks at targeted genes (215). PCAF serves an essential role in regulating inflammatory molecules through H3K18ac, providing a potential therapeutic target for inflammation-associated renal diseases (216-218). HDACs also participate in the inflammatory response in DKD. Apelin-13 inhibits diabetes-induced elevation of inflammatory factors and histone hyperacetylation by upregulation of HDAC1 (219). Gene silencing of HDAC4 decreases the inflammatory response and apoptosis induced by hyperglycemia in podocytes (96). Upregulation of SIRT1 inhibits inflammation through decreasing the induction of inflammatory cytokines and reducing acetylated-NF-κB (220-223). SIRT6 protects podocytes from inflammation by inhibiting the Notch pathway (101). Moreover, upregulation of SIRT7 decreases inflammation and improves renal function in glomerular endothelial cells by regulating the H3K18ac level of death-associated protein kinase-3 (224). These findings support the impact of acetylation on DKD and the potential mechanisms of existing therapeutic drugs, such as losartan.

Different histone modifications are involved in the inflammatory response in DKD. SET7/9 and H3K4me1 expression are markedly increased, whereas H3K9me2 and H3K9me3 are decreased by inflammation induced by hyperglycemia (225,226). Moreover, the diabetic environment attenuates SET domain-containing protein 8 (SETD8) levels, as well as their downstream target H4K20me1. Upregulation of H4K20me1 inhibits endothelial inflammation in DKD by occupying the promoter regions of diverse target genes, including IL-1 receptor-associated kinase 1, Wnt family member 5A and PTP1B (227-229). Additionally, yes-associated protein 1 promotes hyperglycemia-induced inflammation and ECM deposition by triggering the activation of NF-κB/jumonji domain-containing protein-3 signaling in MCs (230). In addition, KDM6A regulates the transcription of inflammatory genes in a manner dependent on its demethylase activity (109). These results support that epigenetic alternations are associated with sustained pro-inflammatory pathways and partly explain the phenomenon of 'hyperglycemic memory' in DKD. This refers to the persistent susceptibility of diabetic patients to develop complications stemming from early hyperglycemic exposure, even following effective implementation of blood glucose control (231).

Ubiquitin-modifying enzymes and deubiquitinases act in conjunction to regulate the transmission of intracellular ubiquitin signaling to maintain normal cell activities. OTUD5, a deubiquitinating enzyme, deubiquitinates K63-linked TGF-β-activated kinase 1 (TAK1) at the K158 site through its active site C224, which prevents TAK1 phosphorylation and decreases downstream inflammatory responses in podocytes during DKD (232). E3 ubiquitin ligase speckle-type BTB-POZ protein promotes NLRP3 degradation by elevating K48-linked polyubiquitination of NLRP3 (233). TGF-β1 is profibrogenic in renal fibrosis (234). Latent TGF-β1, unlike the active form of TGF-β1, protects against inflammation and fibrosis by blocking the E3-ligase Arkadia-mediated Smad7 ubiquitin proteasomal degradation pathway in STZ-induced T1DN (234). Obstruction of ubiquitin degradation of IKK induced by decreased ubiquitin ligase NEDD4L and inhibition of TNF receptor-associated factor 6 K63 polyubiquitination mediated by USP25 can decrease the activation of NF-κB and relieve inflammation (19,235). Parkin inhibits pathological progression of DKD by promoting the ubiquitination of GATA-binding protein 4 (GATA4) and downregulating GATA4/growth arrest-specific protein 1 signaling to inhibit premature senescence, renal inflammation and fibrosis (236). The aforementioned studies demonstrate that protein ubiquitination performs an important role in inflammation in DKD.

In diabetic mice, the expression of SENP6 is decreased in glomerular tissue; this downregulation exacerbates injury to the glomerular filtration barrier. Mechanistically, SENP6 enhances the ubiquitination of the Notch1 intracellular domain (N1ICD) by deSUMOylating Notch1, subsequently reducing N1ICD and inhibiting Notch1 signaling activation in podocytes (237). Moreover, SENP6 deSUMOylates KDM6A and decreases the binding affinity of KDM6A to endothelin-1 (Edn1) via upregulation of H3K27me2/3 at its promoter (237). Numerous studies have indicated that hyperglycemia-induced activation of NF-κB inflammatory signaling is mediated by the SUMO E3 ligase protein inhibitor of activated STAT y and IκBα SUMOylation (238,239). These findings indicate that the combined effect of various PTMs regulate the pathogenesis of DKD.

Protein glycosylation serves an important role in protein stability, binding, folding and activity, and is a key PTM of proteins. In mouse kidney endothelial cells, hyperglycemia causes increased methylglyoxal modification of the corepressor mSin3A, which results in increased recruitment of OGT and enhanced O-GlcNAcylation of Sp3 (240). This modification of Sp3 causes an increase in angiopoietin 2 expression, sensitizing microvascular endothelial cells to the proinflammatory effects of TNFα (240).

Histone Kcr, a PTM, was first identified in 2011 (45). Sodium crotonate (NaCr) may mitigate DKD through an antidiabetic effect, as well as by inducing ACSS2 and P300-induced histone Kcr (241).

TGF-β1 induces profibrotic and inflammatory genes, which serve key roles in glomerular dysfunction and the mesangial matrix expansion associated with DKD (242,243). Multiple protein kinases and phosphatases have an essential role in renal fibrosis. TGF-β1 stimulation results in the phosphorylation/activation of PKCβII, a direct substrate of mTORC2, thus modulating renal fibrosis in DKD (244). Activation of AMPK and Akt signaling alleviates renal injury and fibrosis in DKD (245-247). Activation of MAPK signaling aggravates fibrosis under diabetic conditions (248-250). Activation of the GSK3β and Nrf2/HO-1 pathways also causes inhibitory regulation of EMT and exerts anti-renal fibrosis activity, delaying the progression of DKD (251,252). The suppression of phosphorylation of EGFR and NF-κB are involved in amelioration of renal tubulointerstitial fibrosis (253-255). Targeting shared profibrotic pathways via modulation of protein phosphorylation may serve as a novel therapeutic strategy for DKD (Table III) (130,190,256-260).

Different acetylated states of proteins perform different functions. For example, p300-dependent H3K27 acetylation on the PLK1 gene promoter ameliorates renal fibrosis of DKD (90). Moreover, sterol regulatory element-binding transcription factor 1a K333 acetylation mediated by the acetyltransferase CBP is key for Smad3 association and they cooperatively mediate TGF-β transcriptional responses (261). In addition, numerous deacetylases contribute to renal fibrosis in DKD. HDAC2 serves a key role in the development of ECM accumulation, EMT and renal interstitial fibrosis in diabetic kidneys by regulating the acetylation of substrates, including Edn1 and miR-205 (262-264). SIRT1 activation markedly suppresses endothelial-mesenchymal transition (EndMT) progression, and attenuates albuminuria and glomerulopathy via regulation of the acetylation of NF-κB, FoxO1 and FoxO3a (265-267). SIRT3 deficiency in endothelial cells stimulates TGF-β/Smad3-dependent mesenchymal transformation in RTECs (268). SIRT6 has been demonstrated to directly interact with Smad3, a key downstream mediator of TGF-β, and inhibits its nuclear accumulation and transcriptional activity by deacetylating it in HK2 cells (269). Furthermore, FoxO3a exerts a renoprotective effect against diabetic kidney injury via the SIRT6/Smad3 pathway (Table III) (269,270). These studies reveal the vital role of imbalanced acetylation in renal fibrosis in DKD.

EZH2/H3K27me3 recruitment at the promoters of profibrotic genes is downregulated in rat MCs in T1DN and reciprocally upregulates expression of these profibrotic genes, such as connective tissue growth factor and Serpine1 (271). Moreover, EZH2 alleviates the progression of renal interstitial fibrosis in T1DN by regulating its downstream genes, such as MMP9 (272). In type 2 diabetic nephropathy, OGT stabilizes EZH2 by promoting its glycosylation and then inhibiting MC hyperproliferation and fibrosis by enhancing the enrichment of EZH2/H3K27me3 in the hairy and enhancer of split 1 promoter (273). However, the role of EZH2 in DKD remains controversial. In early DKD, enhanced expression of EZH2 is associated with decreased DEPTOR levels and increased mTOR activity, inducing MC hypertrophy and matrix expansion (274). Under diabetic conditions, recruitment of EZH2 inhibits SOX6, induces cell proliferation, fibrosis and inflammatory cytokine release in MCs (275). The method used to establish DKD models and context-dependent factors affects the function of EZH2 in renal fibrosis. To explain the role of EZH2 in renal fibrosis, the association between EZH2 expression and histological characteristics of patients with DKD should be assessed.

The enhanced recruitment of SET7/9 and elevated H3K4me at the p21 promoter, concurrent with the decreased H3K9me level are observed in the glomeruli of diabetic rats, resulting in increased mesangial hypertrophy (276). Histone H2AK119 mono-ubiquitination (H2AK119-Ub) downregulates SET7/9 (277). Genetic suppression of SET7/9 decreases profibrotic gene expression and prevents EndMT by regulating insulin-like growth factor-binding protein 5 (278,279). Additionally, SETD8/H4K20me1 regulates EndMT in DKD by modulating methylation of its downstream targets, such as profilin 2 and enolase 1 (280-282). However, additional studies are required to examine the effects of SET7/9 and SETD8 in other cell types in the progression of DKD.

Beyond methyltransferases that act on the lysine sites of proteins, elevated protein arginine methyltransferase 1 expression activates activating transcription factor 6 by recruiting H4R3me2as to the promoter, promoting ER stress and EMT activation in HK2 cells (283).

In addition, demethylases are associated with renal fibrosis. Expression of fibrotic proteins and dickkopf-1 is negatively regulated by the KDM6A inhibitor GSK-J4, attenuating glomerulosclerosis and renal fibrosis in mice with DKD (284,285). The HDM KDM3A is recruited to the CTGF promoter to activate transcription, augmenting hyperglycemia-induced CTGF induction in RTECs (286). Additionally, histone lysine-specific demethylase 1 aggravates renal fibrosis by decreasing SIRT3 expression and activating the TGF-β1/Smad3 pathway (Table III) (287).

Numerous inhibitors (such as tazemetostat, GSK126 and GSK-J4) of methyltransferases and demethylases are undergoing clinical trials for cancer treatment (288,289). Such preclinical research utilizing pharmacological drugs that target methylation may advance DKD treatment.

During the progression of renal fibrosis, numerous proteins undergo ubiquitination and deubiquitination (290). E3 ubiquitin ligase of the TRIM subfamily of RING-containing proteins is notably associated with renal fibrosis in DKD (195,291). TRIM18 promotes EMT, inflammation and fibrosis in HK2 cells via ubiquitination of PTP1B and activates STAT3 signaling (292). Upregulation of TRIM13 suppresses mesangial collagen synthesis in DKD by promoting ubiquitination of C/EBP homologous protein, providing insight into the application of histone ubiquitination in the management of DKD (293). Smurf1/2, HECT-type E3 ubiquitin ligases, participate in renal fibrosis by ubiquitinating TGR5 and ChREBP (294-296). Additionally, serum creatinine enhances the interaction between c-Cbl and CKIP-1 by promoting the phosphorylation of c-Cbl, thereby increasing c-Cbl-mediated ubiquitination of CKIP-1 to downregulate its expression, which exacerbates renal inflammatory fibrosis in diabetic mice (297). USP modulates EMT and the production of ECM components by deubiquitinating and stabilizing their respective substrates (22,298). For example, USP9X decreases Nrf2 ubiquitination and deubiquitinates Cx43 to regulate the EMT process (299-301). Ubiquitination and degradation of KDM3A increases TGF-β-induced factor 1 transcriptional activity, inactivating TGF-β1/Smad2/3 signaling and suppressing the progression of DKD (302). TAK1 mediates the phosphorylation of Ski-related novel protein N (SnoN), leading to SnoN ubiquitination and degradation, which enhances EMT and ECM deposition to promote renal fibrosis during DKD (Table III) (303,304). Modulation of ubiquitination may play a promising role in the treatment of DKD.

High glucose enhances the SUMOylation of STAT1, which prevents STAT1 from exerting an effective protective function in inhibiting EMT (305). Furthermore, diabetic conditions activate TGF-β/Smad signaling via SUMO2/3 mediated SUMOylation of Smad4 in MCs (306,307).

O-GlcNAc augments the protein stability, transcriptional activity and nuclear translocation of ChREBP. Diabetic conditions increase the levels of O-GlcNAcylated ChREBP, which further lead to lipid accumulation and upregulation of fibrotic proteins in MCs (308). In addition, O-GlcNAc is, in part, coupled to the profibrotic MAPK signaling pathway via inhibition of Akt phosphorylation and potentially through ROS (309).

The hyperglycemia-induced increases in phosphorylation and oxidation of mitochondrial proteins contributes to tubular dysfunction during DKD (310). S-nitrosylation serves a role in the precise regulation of glomerular homeostasis by modulating multiple important signaling pathways in DKD models. Specifically, S-nitrosylation of laminin prevents the development of glomerular nodules, while denitrosylation of S-nitrosoglutathione reductase and increased S-nitrosylation of β3-integrin collectively result in diffuse glomerulosclerosis in podocytes (311,312).

An elevation in histone lactylation is observed in mice with DKD. H3K14la promotes the transcription of KLF5 in RTECs of DKD (313,314). Notably, disruption of the lactate/H3K14la/KLF5 pathway mitigates renal dysfunction and DKD pathology (313). In addition, 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3 (PFKFB3), a key glycolytic enzyme, is associated with renal fibrosis and dysfunction. Lactate generated from PFKFB3-mediated tubular glycolytic reprogramming significantly enhances histone lactylation, particularly H4K12la, which is enriched at the promoters of NF-κB signaling genes (such as Ikbkb, Rela and Relb), activating their transcription, and facilitating the inflammatory response and renal fibrosis (39).

NaCr exerts an antidiabetic effect, decreases blood glucose and serum lipid levels, and alleviates renal function and DKD-associated inflammatory and fibrotic damage. NaCr induces histone Kcr and H3K18 crotonylation. However, NaCr and Cr-CoA-induced histone Kcr and renoprotective effects are abrogated by inhibiting the activity of ACSS2 or histone acyltransferase p300 in vitro (241). Moreover, ACSS2 notably increases H3K9cr levels in renal tissues and tubular epithelial cells (48,91). Genetic and pharmacological suppression of ACSS2 inhibits H3K9cr-mediated IL-1β expression, which alleviates IL-1β-dependent macrophage activation and tubular senescence to delay renal fibrosis (48). Targeting ACSS2 may serve as a potential therapeutic intervention for the management of DKD, but warrants further preclinical and clinical investigations.

Beyond its role in energy supply, BHB serves as a bioactive molecule that exerts numerous protective effects, including in DKD. BHB antagonizes glomerulosclerosis in diabetic rats via upregulation of MMP2 production through elevation of H3K9bhb at the MMP2 promoter (315). Moreover, BHB ameliorates hyperglycemia-induced podocyte injury in vitro (316).