The miR-145-5p/NRF2/SLC7A11 axis regulates ferroptosis-mediated renal tubular epithelial cell injury in diabetic nephropathy.

Conceptualization, Zuopeng Lu; Methodology, Zuopeng Lu; Software, Yu Cao; Validation, Ganlin Wu, Yu Fang, and Yana Gao; Formal Analysis, Ganlin Wu and Yu Fang; Investigation, Yu Fang and Meirong Xu; Resources, Yu Cao; Data Curation, Ganlin Wu and Yu Fang; Writing – Original Draft Preparation, Ganlin Wu and Yu Fang; Writing – Review and Editing, Ganlin Wu and Meirong Xu; Visualization, Ganlin Wu and Zuopeng Lu; Supervision, Meirong Xu; Project Administration, Zuopeng Lu.

To investigate the involvement of miR‐145‐5p in DN, db/db mice were used as the experimental model, whereas age‐matched db/m mice served as controls. RT‐qPCR analysis showed that renal miR‐145‐5p expression was significantly increased in db/db mice compared with db/m mice ( P   0.05, Figure  1a ). Consistently, analysis of the GEO dataset GSE114477 , which included four healthy controls and four patients with DN, revealed elevated miR‐145 expression in patients with DN, although these findings should be interpreted cautiously because of the limited sample size (Figure  S1a ). miR‐145‐5p is upregulated in DN mice, and its inhibition improves ferroptosis and alleviates renal damage. (a) RT‐qPCR analysis of miR‐145‐5p and miR‐145‐3p expression in renal tissues of DN mice; (b) Detection of blood glucose levels in DN mice; (c) ELISA detection of urine ACR, serum creatinine, and BUN levels in DN mice; (d) Representative images of HE, PAS, and Masson staining to show the damage to renal tissues and renal fibrosis; (e) A quantitative analysis for HE, PAS, and Masson staining; (f) Immunohistochemical detection of NGAL and KIM‐1 expression in renal tissues of DN mice; (g) Assay kits to measure the levels of MDA, GSH, and Fe 2+ in renal tissues of DN mice; (h) Western blot analysis of GPX4, ACSL4, E‐cadherin, Vimentin, and α‐SMA protein expression in renal tissues of DN mice. N  = 6. Data were presented as mean ± SD. The normality of data in panels a–h was examined by the Shapiro–Wilk test. The F ‐test was used to test the homogeneity of variance regarding miR‐145‐3p expression in panel a and data in panel b, with the independent sample t ‐test applied for comparisons among groups. The Brown–Forsythe test was employed to examine the homogeneity of variance regarding miR‐145‐5p expression in panel a and data in panels c–h, with one‐way ANOVA and Tukey's multiple comparisons test applied for comparisons among groups. * P  < 0.05, ** P  < 0.01, *** P  < 0.001. Compared with db/m mice, db/db mice exhibited markedly elevated blood glucose levels ( P  < 0.001, Figure  1b ), accompanied by significant increases in the urinary ACR, serum creatinine, and BUN (all P  < 0.001, Figure  1c ), indicating impaired renal function. Histopathological examination further demonstrated characteristic DN‐associated lesions, including glomerular hypertrophy, mesangial matrix expansion, tubular epithelial vacuolar degeneration, and increased collagen deposition (all P  < 0.001, Figure  1d,e ). Immunohistochemical staining showed significantly increased expression of the renal tubular injury markers NGAL and KIM‐1 in db/db mice (both P  < 0.001, Figure  1f ), confirming substantial tubular injury. 53 Given the reported involvement of ferroptosis in renal tubular injury during DN, 20 , 21 ferroptosis‐related biomarkers were subsequently evaluated. Compared with db/m mice, renal tissues from db/db mice exhibited significantly increased MDA, Fe 2+ levels, and ACSL4 levels, together with reduced GSH content and GPX4 expression (all P  < 0.01, Figure  1g,h ), indicating enhanced ferroptotic activity. To determine the functional role of miR‐145‐5p, db/db mice were treated with a miR‐145‐5p antagomir via tail vein injection. As expected, miR‐145‐5p antagomir markedly reduced its renal expression ( P  < 0.001, Figure  1a ) and significantly improved renal function, as evidenced by decreased urinary ACR, serum creatinine, and BUN levels (all P  < 0.01, Figure  1c ). Histological analyses demonstrated attenuation of glomerular hypertrophy, mesangial matrix expansion, and collagen deposition (all P  < 0.01, Figure  1d,e ), accompanied by significantly reduced NGAL and KIM‐1 expression (both P  < 0.01, Figure  1f ). Moreover, miR‐145‐5p inhibition significantly reduced renal MDA and Fe 2+ levels while restoring GSH content and GPX4 expression (all P   0.05, Figure  1h ). Because ferroptosis has been implicated in EMT during DN progression, 54 EMT‐related proteins were further examined. Compared with db/m mice, db/db mice exhibited a reduction in the E‐cadherin expression together with increased Vimentin and α‐SMA expression, consistent with EMT activation. Notably, suppression of miR‐145‐5p largely reversed these changes (all P  < 0.05, Figure  1h ). Collectively, these findings demonstrate that miR‐145‐5p is aberrantly upregulated in diabetic kidneys and that its inhibition attenuates renal injury, suppresses ferroptosis, and alleviates EMT in DN. To determine whether ferroptosis mediates the protective effects of miR‐145‐5p inhibition, db/db mice treated with the miR‐145‐5p antagomir were further administered the ferroptosis inducer Erastin (30 mg/kg, intraperitoneally, every other day). Compared with vehicle‐treated mice, Erastin administration significantly increased renal MDA and Fe 2+ levels while reducing GSH content and GPX4 expression (all P  < 0.05, Figure  2a,b ), confirming successful ferroptosis induction. Correspondingly, renal function deteriorated following Erastin treatment, as reflected by significantly increased urinary ACR, serum creatinine, and BUN levels (all P  < 0.05, Figure  2c ). Activation of ferroptosis partially averts the beneficial effect of miR‐145‐5p inhibition in DN mice. (a) Measurement of MDA, GSH, and Fe 2+ levels in renal tissues using assay kits; (b) Analysis of GPX4 protein expression by Western blot; (c) Assessment of urine ACR, serum creatinine, and BUN levels via ELISA; (d) Representative images of HE, PAS, and Masson staining; (e) A quantitative analysis for HE, PAS, and Masson staining; (f) Detection of NGAL and KIM‐1 expression through immunohistochemistry; (g) Western blot analysis of E‐cadherin, Vimentin, and α‐SMA protein expression in renal tissues of DN mice. N  = 6. Data were presented as mean ± SD. The normality of data was examined by the Shapiro–Wilk test, with the Brown–Forsythe test to examine the homogeneity of variance. One‐way anova was adopted for comparisons among multiple groups, followed by Tukey's test. * P  < 0.05, ** P  < 0.01, *** P  < 0.001. Histopathological analyses further demonstrated that Erastin aggravated glomerular hypertrophy, mesangial matrix expansion, and collagen deposition (Figure  2d,e ). In addition, expression of the tubular injury markers NGAL and KIM‐1 was significantly increased (all P  < 0.05, Figure  2f ). Western blot analysis further showed enhanced EMT following Erastin administration, characterized by reduced E‐cadherin expression and increased Vimentin and α‐SMA levels (all P  < 0.05, Figure  2g ). Taken together, these findings indicate that pharmacological activation of ferroptosis partially abolishes the renoprotective effects conferred by miR‐145‐5p inhibition, suggesting that suppression of ferroptosis is a key mechanism underlying the protective role of miR‐145‐5p silencing in diabetic nephropathy. To further investigate the underlying mechanism in vitro , mouse RTECs (TCMK‐1) were exposed to HG for 48 h to establish a cellular model of DN, whereas NG treatment served as the control. Compared with NG‐treated cells, HG exposure significantly increased miR‐145‐5p expression ( P  < 0.001, Figure  3a ). Concurrently, HG treatment induced characteristic features of ferroptosis, including increased MDA and Fe 2+ levels together with markedly reduced GSH levels and GPX4 expression (all P  < 0.01, Figure  3b,c ). Transmission electron microscopy (TEM) further revealed typical ferroptotic mitochondrial alterations, including mitochondrial swelling, vacuolar degeneration, and marked loss of mitochondrial cristae (Figure  3d ), confirming ferroptosis induction under HG conditions. Suppression of miR‐145‐5p ameliorates HG‐induced TCMK‐1 cell damage by regulating ferroptosis. (a) RT‐qPCR detection of miR‐145‐5p expression in TCMK‐1 cells; (b) Detection of MDA, GSH, and Fe 2+ in TCMK‐1 cells; (c) Western blot analysis of GPX4 protein expression in TCMK‐1 cells; (d) Observation of the ultrastructure of TCMK‐1 cells using TEM. Cell experiments were independently repeated three times; the data were expressed as mean ± standard deviation. The normality of data was examined by the Shapiro–Wilk test, with the Brown–Forsythe test to examine the homogeneity of variance. One‐way anova was used for multiple group comparisons, followed by Tukey's multiple comparisons test. * P  < 0.05, ** P  < 0.01. *** P  < 0.001. To evaluate the role of miR‐145‐5p, HG‐treated TCMK‐1 cells were transfected with a miR‐145‐5p antagomir, which significantly reduced miR‐145‐5p expression ( P  < 0.001, Figure  3a ). miR‐145‐5p inhibition markedly decreased MDA and Fe 2+ levels while increasing GSH levels and GPX4 expression ( P  < 0.05, Figure  3b,c ). Moreover, TEM demonstrated partial restoration of normal mitochondrial morphology following miR‐145‐5p inhibition (Figure  3d ). To exclude the potential influence of osmotic stress, miR‐145‐5p expression and ferroptosis‐related markers were compared between the NG and mannitol (MT) groups. No significant differences were observed in miR‐145‐5p expression, GPX4 expression, or MDA, Fe 2+ , and GSH levels between these groups (Figure  S2 ), indicating that the observed changes were attributable to HG rather than osmotic effects. Collectively, these findings demonstrate that miR‐145‐5p inhibition effectively suppresses HG‐induced ferroptosis in TCMK‐1 cells. NRF2 is a master regulator of the cellular antioxidant response that suppresses oxidative stress and ferroptosis. 55 Consistent with previous reports showing that activation of the NRF2 pathway inhibits ferroptosis in DN, 37 RT‐qPCR and Western blot analyses demonstrated that HG treatment significantly reduced NRF2 mRNA and protein expression in TCMK‐1 cells, whereas miR‐145‐5p inhibition restored NRF2 expression (all P  < 0.05, Figure  4a,b ). Analysis of the GEO dataset GSE142025 (nine healthy controls and 28 DN samples) similarly revealed reduced NFE2L2 (NRF2) expression in patients with DN (Figure  S1b ). miR‐145‐5p regulates HG‐induced TCMK‐1 cell injury via NRF2/SLC7A11‐mediated ferroptosis. (a/i) RT‐qPCR detection of the mRNA level of NRF2 and SLC7A11 in TCMK‐1 cells; (b/j) Western blot analysis on the protein expression of NRF2, SLC7A11, E‐cadherin, Vimentin, and α‐SMA in TCMK‐1 cells; (c) Predictive analysis of targeted binding sites for miR‐145‐5p and NRF2 through TargetScan mouse ( https://www.targetscan.org/vert_80/ ); (d) The dual‐luciferase assay to verify the binding relationship between miR‐145‐5p and NRF2; (e) Assay kits to measure MDA, GSH, and Fe 2+ levels in TCMK‐1 cells; (f) Western blot technique to analyze GPX4 protein expression in TCMK‐1 cells; (g) TEM to observe the internal ultrastructure of TCMK‐1 cells; (h) ChIP‐qPCR to assess the impact of NRF2 on SLC7A11 transcription. Cell experiments were independently repeated three times. Data are presented as mean ± standard deviation. The normality of data in panels a–j was examined by the Shapiro–Wilk test. The F ‐test was conducted to examine the homogeneity of variance in panels d and h, with the independent sample t ‐test applied for comparisons among groups. The Brown–Forsythe test was performed to examine the homogeneity of variance in panels a–b, e–f, and i–j, with one‐way anova followed by Tukey's multiple comparisons test applied for multiple group comparisons. * P  < 0.05, ** P  < 0.01, *** P  < 0.001. TargetScan Mouse database predicted a conserved binding site between miR‐145‐5p and NRF2 (Figure  4c ). Dual‐luciferase reporter assays demonstrated that co‐transfection of miR‐145‐5p mimics significantly reduced luciferase activity of the wild‐type NRF2 reporter (NRF2‐WT) ( P   0.05, Figure  4d ), confirming that NRF2 is a direct downstream target of miR‐145‐5p. To determine whether NRF2 mediates the effects of miR‐145‐5p on ferroptosis, HG‐treated TCMK‐1 cells were co‐transfected with miR‐145‐5p antagomir and si‐NRF2. Compared with the miR‐145‐5p antagomir group, additional si‐NRF2 significantly reduced NRF2 mRNA and protein expression (all P  < 0.01, Figure  4a,b ), increased MDA and Fe 2+ levels, decreased levels of GSH and GPX4 (all P  < 0.05, Figure  4e,f ), and aggravated ferroptotic mitochondrial damage, as demonstrated by TEM (Figure  4g ). Because NRF2 transcriptionally regulates SLC7A11, 56 ChIP‐qPCR was performed to verify this interaction. The NRF2 antibody significantly enriched the SLC7A11 promoter in TCMK‐1 cells, and this enrichment was further enhanced following miR‐145‐5p inhibition ( P  < 0.01, Figure  4h ), indicating increased NRF2 binding to the SLC7A11 promoter. Consistently, RT‐qPCR and Western blot analyses showed that HG treatment significantly reduced SLC7A11 expression compared with NG treatment, whereas miR‐145‐5p inhibition restored SLC7A11 expression. This increase was partially abolished by NRF2 knockdown (all P  < 0.05, Figure  4i,j ). In parallel, HG‐induced EMT, characterized by reduced E‐cadherin expression and increased Vimentin and α‐SMA expression, was markedly attenuated following miR‐145‐5p inhibition. However, NRF2 silencing partially reversed these protective effects (all P  < 0.05, Figure  4j ). Taken together, these findings demonstrate that miR‐145‐5p promotes HG‐induced TCMK‐1 injury by suppressing the NRF2/SLC7A11 signaling pathway, thereby facilitating ferroptosis. Inhibition of miR‐145‐5p alleviates ferroptosis and EMT through restoration of NRF2/SLC7A11 signaling. To further verify the involvement of the miR‐145‐5p/NRF2/SLC7A11 axis in regulating ferroptosis, SLC7A11 expression was silenced using siRNA in HG‐treated TCMK‐1 cells following miR‐145‐5p inhibition. Consequently, si‐SLC7A11 significantly reduced SLC7A11 mRNA and protein expression (all P  < 0.01, Figure  5a,b ). This reduction was accompanied by significantly increased MDA and Fe 2+ levels together with decreased GSH levels and GPX4 expression (all P  < 0.05, Figure  5c,d ), indicating enhanced ferroptotic activity. Knockdown of SLC7A11 promotes ferroptosis and partially reverses the protective effect of miR‐145‐5p inhibition against HG‐induced injury in TCMK‐1 cells. (a) RT‐qPCR detection of the mRNA level of SLC7A11 in TCMK‐1 cells; (b/d) Western blot analysis on the protein expression of SLC7A11/GPX4 in TCMK‐1 cells; (c) Detection of MDA, GSH, and Fe 2+ in TCMK‐1 cells using assay kits; (e) TEM observation of the ultrastructure of TCMK‐1 cells; (f) Western blot analysis of E‐cadherin, Vimentin, and α‐SMA protein expression in TCMK‐1 cells. Cell experiments were independently repeated three times. Data are presented as mean ± standard deviation. The normality of data was examined by the Shapiro–Wilk test, with the Brown–Forsythe test to examine the homogeneity of variance. One‐way anova was used for multiple group comparisons, followed by Tukey's multiple comparisons test. * P  < 0.05, ** P  < 0.01, *** P  < 0.001. Consistently, TEM revealed more pronounced ferroptotic mitochondrial alterations following SLC7A11 knockdown (Figure  5e ). In addition, suppression of SLC7A11 further promoted EMT, as evidenced by significant changes in EMT‐related protein expression (all P  < 0.05, Figure  5f ). Collectively, these findings demonstrate that SLC7A11 deficiency promotes ferroptosis and partially reverses the protective effects of miR‐145‐5p inhibition against HG‐induced injury in TCMK‐1 cells, further supporting the involvement of the miR‐145‐5p/NRF2/SLC7A11 axis in regulating ferroptosis. To validate the regulatory role of the miR‐145‐5p/NRF2/SLC7A11 axis in vivo , the expression of NRF2 and SLC7A11 was first examined in mouse renal tissues by RT‐qPCR and Western blot analyses. Compared with db/m mice, db/db mice exhibited significantly reduced renal NRF2 and SLC7A11 expression, whereas inhibition of miR‐145‐5p partially restored the expression of both molecules ( P  < 0.05, Figure  6a,b ). Inhibition of miR‐145‐5p improves renal damage in DN mice by regulating ferroptosis through the NRF2/SLC7A11 axis. (a) RT‐qPCR detection of the levels of NRF2 and SLC7A11 in renal tissues of DN mice; (b/h) Western blot analysis on the expression of NRF2, SLC7A11, GPX4, E‐cadherin, Vimentin, and α‐SMA proteins in renal tissues of DN mice; (c) ELISA detection of urine ACR, serum creatinine, and BUN levels in DN mice; (d) Representative images of HE, PAS, and Masson staining to show renal tissue damage and fibrosis; (e) A quantitative analysis for HE, PAS, and Masson staining; (f) Immunohistochemical detection of NGAL and KIM‐1 expression in renal tissues of DN mice; (g) Detection of MDA, GSH, and Fe 2+ in the renal tissues of DN mice. N  = 6. Data were expressed as mean ± standard deviation. The normality of data was examined by the Shapiro–Wilk test, with the Brown–Forsythe test to examine the homogeneity of variance. One‐way anova was used for multiple group comparisons, followed by Tukey's multiple comparisons test. * P  < 0.05. ** P  < 0.01. *** P  < 0.001. To further investigate the functional role of NRF2, db/db mice were co‐treated with miR‐145‐5p antagomir and rAAV‐carried sh‐NRF2. Compared with mice receiving miR‐145‐5p antagomir alone, rAAV‐carried sh‐NRF2 significantly reduced renal NRF2 and SLC7A11 expression (both P  < 0.05, Figure  6a,b ). This was accompanied by significant increases in urinary ACR levels, serum creatinine, and BUN levels (all P  < 0.05, Figure  6c ), indicating deterioration of renal function. Histopathological analyses further demonstrated aggravated glomerular hypertrophy, mesangial matrix expansion, and collagen deposition following NRF2 silencing (all P  < 0.01, Figure  6d,e ). Moreover, renal expression of the tubular injury markers NGAL and KIM‐1 was significantly increased (both P  < 0.05, Figure  6f ). Consistent with enhanced ferroptosis, NRF2 knockdown resulted in elevated MDA and Fe 2+ levels together with reduced GSH levels and GPX4 expression (all P  < 0.05, Figure  6g,h ). In parallel, EMT was further aggravated, as reflected by altered expression of EMT‐associated proteins (all P  < 0.05, Figure  6h ). To determine whether these detrimental effects were ferroptosis‐dependent, Ferrostatin‐1 was administered intraperitoneally to db/db mice receiving both miR‐145‐5p antagomir and rAAV‐carried sh‐NRF2. Ferrostatin‐1 treatment significantly reduced urinary ACR level, serum creatinine, and BUN levels (all P  < 0.05, Figure  6c ), indicating improved renal function. Histological examination showed marked attenuation of glomerular hypertrophy, mesangial matrix expansion, and collagen deposition (all P  < 0.01, Figure  6d,e ). In addition, Ferrostatin‐1 significantly decreased renal NGAL and KIM‐1 expression (all P  < 0.05, Figure  6f ), reduced MDA and Fe 2+ levels, restored GSH levels and GPX4 expression (all P  < 0.05, Figure  6g,h ), and markedly alleviated EMT (all P  < 0.05, Figure  6h ). Taken together, these findings demonstrate that inhibition of miR‐145‐5p alleviates renal injury in DN by suppressing ferroptosis through the activation of the NRF2/SLC7A11 signaling pathway.

The present study investigated the mechanistic role of the miR‐145‐5p/NRF2/SLC7A11 axis in HG‐induced RTEC injury during DN, with a particular focus on ferroptosis. Our findings demonstrate that inhibition of miR‐145‐5p attenuates ferroptosis and alleviates renal injury in DN by restoring NRF2/SLC7A11 signaling. These results provide new mechanistic insight into the molecular basis of ferroptosis‐mediated tubular injury and identify the miR‐145‐5p/NRF2/SLC7A11 axis as a potential therapeutic target for limiting DN progression. One of the major findings of this study is the identification of miR‐145‐5p as a promoter of ferroptosis in RTECs, which was consistently demonstrated in both in vitro and in vivo models of DN. miR‐145‐5p has been widely recognized as a diabetes‐associated miRNA 57 and has been reported to be upregulated in the serum of patients with diabetic retinopathy as well as in circulating exosomes derived from individuals with type 1 diabetes. 58 , 59 Differential expression analyses have further identified miR‐145 among several deregulated miRNAs associated with diabetic retinopathy, highlighting its potential as a biomarker of diabetes‐related vascular dysfunction. 60 Consistent with these observations, Wang et al . reported that miR‐145 was significantly upregulated in renal tubular tissues from patients with DN and that its inhibition attenuated renal injury by restoring taurine‐upregulated gene 1 (TUG1) expression. 29 In agreement with these findings, we observed significantly increased renal miR‐145‐5p expression in db/db mice and demonstrated that its inhibition improved renal function, reduced histopathological injury, and suppressed ferroptosis. Clinical evidence further supports a pathogenic role for miR‐145‐5p in DN. Platelet‐derived miR‐145‐5p levels have been shown to correlate positively with established indicators of renal injury, including urinary ACR, serum creatinine, and BUN, 61 suggesting that excessive miR‐145‐5p expression is closely associated with disease severity. Nevertheless, previous studies have reported conflicting findings. For example, Wei et al . proposed that miR‐145‐5p protected against DN progression by attenuating HG‐induced podocyte apoptosis. 32 However, that study was restricted to an in vitro podocyte model and did not evaluate renal pathology or tubular injury in vivo . By integrating both cellular and animal experiments, our study provides more comprehensive evidence supporting a pathogenic role for miR‐145‐5p in DN, particularly in renal tubular injury mediated by ferroptosis. Because EMT is a key driver of renal fibrosis, 62 we further examined the effects of miR‐145‐5p inhibition on EMT‐related proteins. Although regulation of EMT by miR‐145‐5p has been extensively reported in multiple malignancies, 63 , 64 its involvement in tubular EMT during DN remains largely unexplored. Our results demonstrate that miR‐145‐5p inhibition restores E‐cadherin expression while reducing Vimentin and α‐SMA expression, indicating attenuation of EMT. These findings suggest that suppression of miR‐145‐5p may alleviate renal fibrosis not only by inhibiting ferroptosis but also by limiting EMT progression. Similar cytoprotective effects have been reported in other disease models, where miR‐145‐5p inhibition suppresses oxidative stress and inflammatory responses through modulation of the Nurr1/TNF‐α signaling pathway, 65 further supporting its role as a regulator of cellular injury. Importantly, the rescue experiments performed in the present study further strengthened the mechanistic link between miR‐145‐5p and ferroptosis. Pharmacological activation of ferroptosis with Erastin substantially attenuated the renoprotective effects of miR‐145‐5p inhibition, whereas treatment with the ferroptosis inhibitor Ferrostatin‐1 markedly alleviated renal injury and fibrosis, even in the presence of NRF2 knockdown. These complementary gain‐ and loss‐of‐function experiments provide compelling evidence that ferroptosis is a critical downstream effector of miR‐145‐5p‐mediated renal injury. Collectively, our findings support the concept that the renoprotective effects of miR‐145‐5p inhibition may be mediated predominantly through suppression of ferroptosis, thereby limiting tubular injury and subsequent fibrotic remodeling during the progression of DN. Subsequent experiments focused on identifying the downstream target through which miR‐145‐5p exerts its regulatory effects. Dual‐luciferase reporter assays demonstrated that miR‐145‐5p directly targeted NRF2 and suppressed its expression. Although this regulatory relationship has previously been reported in chronic intermittent hypoxia‐induced lung injury, 41 its role in DN has not been elucidated. NRF2 is a master regulator of the cellular antioxidant response and plays a pivotal role in protecting against ferroptosis. 34 Upon activation, NRF2 promotes the transcription of antioxidant genes, including SLC7A11, which is essential for cystine uptake and GSH synthesis, as well as GPX4, a key enzyme that limits lipid peroxidation. 66 Consequently, activation of the NRF2/SLC7A11/GPX4 signaling axis enhances antioxidant defenses and suppresses ferroptotic cell death. 67 Based on these observations, we investigated whether the miR‐145‐5p/NRF2/SLC7A11 axis mediates ferroptosis in RTECs during DN progression. Our findings demonstrated that knockdown of NRF2 partially abolished the anti‐ferroptotic effects of miR‐145‐5p inhibition in HG‐treated RTECs and attenuated its renoprotective role in db/db mice. These results are consistent with previous studies showing that NRF2 activation suppresses ferroptosis in HG‐treated RTECs, 68 whereas NRF2 inhibition promotes tubular cell death and inflammation in chronic kidney disease. 69 Likewise, NRF2 deficiency exacerbates ferroptosis and iron accumulation in vivo , thereby aggravating cisplatin‐induced acute kidney injury, whereas activation of the NRF2 pathway alleviates renal damage. 70 Increasing evidence further suggests that targeting NRF2 to suppress oxidative stress and ferroptosis may delay the progression of diabetic kidney disease. 71 Moreover, blockade of NRF2 aggravates HG‐induced mitochondrial dysfunction, podocyte ferroptosis, and podocyte injury. 72 Activation of the NRF2/SLC7A11/GPX4 pathway has also been shown to alleviate renal injury in DN mice, 73 protect RTECs from HG‐induced ferroptosis and tubular injury, 74 and suppress HG‐induced renal damage in experimental DN by inhibiting ferroptosis. 75 Collectively, these findings are in agreement with our results and support the conclusion that the renoprotective effects of miR‐145‐5p inhibition are mediated, at least in part, through restoration of NRF2/SLC7A11‐dependent antioxidant capacity and subsequent suppression of ferroptosis in RTECs. In summary, the present study identifies miR‐145‐5p as an important regulator of ferroptosis in DN, highlighting its potential as a therapeutic target for preventing renal injury. These findings provide mechanistic insight into the involvement of noncoding RNA‐mediated ferroptosis in DN and further advance our understanding of disease pathogenesis. Nevertheless, several limitations should be acknowledged. First, given the complexity of DN, the present study focused primarily on RTECs, and validation of the proposed mechanism in additional renal cell types, including podocytes and mesangial cells, would further strengthen the translational significance of our findings. Second, ferroptosis is regulated by multiple interconnected signaling pathways, and additional mechanisms beyond the miR‐145‐5p/NRF2/SLC7A11 axis, including direct regulation of GPX4 and other ferroptosis‐associated molecules, warrant further investigation. Third, the GSE114477 dataset used for bioinformatic validation included only four samples per group, limiting its statistical power and generalizability. Larger independent datasets and clinical cohorts are therefore needed to validate the differential expression of miR‐145‐5p and NRF2 in DN. Furthermore, analysis of clinical samples from patients with DN would help determine the association of miR‐145‐5p and NRF2 expression with disease severity and evaluate their potential as clinically relevant biomarkers. Finally, integration of transcriptomics, proteomics, metabolomic, and other multi‐omics approaches may provide a more comprehensive understanding of ferroptosis‐related molecular networks in DN and facilitate the development of precision therapeutic strategies.

Diabetes mellitus (DM) is one of the most prevalent metabolic disorders worldwide and is characterized by chronic hyperglycemia accompanied by disturbances in glucose, lipid, and protein metabolism. 1 , 2 According to the International Diabetes Federation, the global prevalence of DM among adults is projected to reach 12.2% by 2045. 3 Patients with DM are at increased risk of developing both macrovascular complications, such as cardiovascular disease, and microvascular complications, including diabetic kidney disease (DKD), diabetic nephropathy (DN), and diabetic neuropathy. 4 Among these complications, DN is one of the leading causes of chronic kidney disease and end‐stage renal disease, affecting approximately 40% of individuals with diabetes. 5 Progressive renal fibrosis is a pathological hallmark of DN and is driven by epithelial‐to‐mesenchymal transition (EMT) and endothelial‐to‐mesenchymal transition (EndMT), processes regulated by multiple signaling pathways, including transforming growth factor‐β (TGF‐β), Notch, and WNT. 6 , 7 Aberrant activation of pathways, such as angiopoietin‐like protein 4 (ANGPTL4), TGF‐β, WNT, and Notch, has been implicated in the progression of DN, 8 , 9 , 10 whereas glucocorticoid receptors, sirtuin 3 (SIRT3), and fibroblast growth factor receptor 1 (FGFR1) signaling exhibit renoprotective effects in diabetic kidneys. 11 , 12 , 13 In parallel, preclinical studies using animal models have demonstrated the therapeutic potential of several renoprotective agents, including the dipeptidyl peptidase‐4 inhibitor linagliptin, the sodium‐glucose cotransporter 2 inhibitor empagliflozin, 14 angiotensin‐converting enzyme inhibitors, 15 and the endogenous tetrapeptide N‐acetyl‐seryl‐aspartyl‐lysyl‐proline. 16 Nevertheless, the molecular mechanisms underlying DN remain incompletely understood, highlighting the importance of identifying novel regulatory pathways that may serve as therapeutic targets for delaying disease progression. Renal tubular injury is increasingly recognized as a pivotal event in the early pathogenesis of DN. Under diabetic conditions, renal tubules undergo a series of structural and functional alterations, including tubular atrophy, interstitial fibrosis, and peritubular capillary rarefaction, all of which contribute to progressive renal dysfunction. 17 , 18 Renal tubular epithelial cells (RTECs) are particularly susceptible to hyperglycemic injury because of their high energy demand and reliance on aerobic metabolism. 19 Accumulating evidence indicates that high glucose (HG) promotes ferroptosis of RTECs by increasing reactive oxygen species (ROS) production and oxidative stress, thereby contributing to renal tubular injury in DN. 20 , 21 Ferroptosis is a distinct form of regulated cell death characterized by iron‐dependent lipid peroxidation and the accumulation of lethal lipid hydroperoxides. 22 In addition to directly inducing cell death, ferroptosis can amplify inflammatory responses and exacerbate kidney injury, thereby accelerating the progression of renal dysfunction. 23 , 24 Consequently, elucidating the molecular mechanisms that regulate ferroptosis may provide novel insights into the pathogenesis of DN and identify potential therapeutic strategies for preserving renal function. MicroRNAs (miRNAs) are endogenous small noncoding RNAs that regulate gene expression by binding to the 3′‐untranslated region (3′‐UTR) of target messenger RNAs, thereby suppressing mRNA translation or promoting mRNA degradation. 25 Increasing evidence indicates that miRNAs play critical roles in the pathogenesis of DN, particularly through regulating EMT, EndMT, inflammation, oxidative stress, and renal fibrosis. 26 Moreover, interactions between miRNAs and long noncoding RNAs (lncRNAs) further modulate the progression of DN by regulating genes involved in these pathological processes. 27 For example, knockdown of lncRNA H19 attenuates renal fibrosis through miR‐29a‐mediated suppression of fibroblast‐specific protein‐1 (FSP‐1), an important mediator of EndMT. 28 Among these regulatory miRNAs, miR‐145‐5p has attracted increasing attention because of its involvement in diabetic renal injury. Under HG conditions, miR‐145‐5p is upregulated and promotes fibrosis and inflammatory responses in RTECs by suppressing dual‐specificity phosphatase 6 (DUSP6). 29 In addition, urinary exosome‐derived miR‐145‐5p from patients with diabetic kidney disease induces podocyte apoptosis. 30 However, these findings contrast with reports demonstrating protective effects of miR‐145‐5p against HG‐induced apoptosis in human renal tubular epithelial HK‐2 cells 31 and podocytes, 32 suggesting that the biological function of miR‐145‐5p in DN may be context‐dependent and remains incompletely understood. Notably, recent evidence has shown that miR‐145‐5p promotes ferroptosis in endometriosis by targeting MUC1. 33 Nevertheless, whether miR‐145‐5p regulates ferroptosis in RTECs during DN progression and the underlying molecular mechanisms remain unclear. Nuclear factor erythroid 2‐related factor 2 (NRF2) is a master transcription factor that orchestrates the cellular antioxidant response by regulating the expression of multiple cytoprotective genes, including solute carrier family 7 member 11 (SLC7A11). 34 , 35 SLC7A11 is a key component of the cystine/glutamate antiporter system Xc − and promotes intracellular glutathione (GSH) synthesis by mediating cystine uptake, thereby protecting cells against oxidative stress and ferroptosis. 36 Activation of the NRF2/heme oxygenase‐1 (HO‐1) signaling pathway has been shown to inhibit ferroptosis in both db/db mice and HG‐treated renal tubular epithelial cells. 37 Mechanistically, NRF2‐mediated upregulation of SLC7A11 and glutathione peroxidase 4 (GPX4) expression is essential for maintaining redox homeostasis and suppressing ferroptotic cell death. 38 , 39 Consistently, activation of the SLC7A11/GPX4 pathway has been reported to alleviate HG‐induced ferroptosis in podocytes and ameliorate renal injury in streptozotocin‐induced diabetic mice. 40 A previous study has identified NRF2 as a potential target of miR‐145‐5p, 41 and our preliminary bioinformatics analysis using the TargetScan Mouse database likewise predicted NRF2 as a downstream target of miR‐145‐5p. However, whether miR‐145‐5p regulates ferroptosis through the NRF2/SLC7A11 signaling axis during the development of DN has not yet been elucidated. Therefore, the present study investigated the molecular mechanism by which the miR‐145‐5p/NRF2/SLC7A11 axis regulates ferroptosis‐mediated RTEC injury in DN, with the aim of identifying potential therapeutic targets for preventing the progression of DN.

The authors have no conflicts of interest to declare. Ethics approval and consent to participate: All animal experiments were approved by the Animal Ethics Committee of Xianning Medical College, Hubei University of Science and Technology. All experimental procedures were conducted in accordance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals (8th Edition, 2011), and every effort was made to minimize animal suffering and reduce the number of animals used. No written consent has been obtained from the patients as there is no patient‐identifiable data included. Funding: The authors have nothing to report.

All animal experiments were approved by the Animal Ethics Committee of Xianning Medical College, Hubei University of Science and Technology. All experimental procedures were conducted in accordance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals (8th Edition, 2011), and every effort was made to minimize animal suffering and reduce the number of animals used. No written consent has been obtained from the patients as there is no patient‐identifiable data included. Six‐week‐old male C57BL/KsJ‐db/m and db/db mice were purchased from Hubei Biont Biological Technology Co., LTD. (Wuhan, China; animal license number No. SCXK [Hubei] 2021–0027). Animals were housed under specific pathogen‐free conditions at 18–22°C with 40–60% relative humidity under a 12‐h/12‐h light/dark cycle, with free access to standard chow and water. Mice were randomly assigned to the following experimental groups ( N  = 6 per group): (1) db/m group, serving as the nondiabetic control; (2) db/db group (C57BL/KsJ‐db/db), representing the DN group; (3) db/db + miR antagomir and db/db + antagomir negative control (NC) groups; (4) db/db + miR antagomir + Erastin and db/db + miR antagomir + Vehicle groups; (5) db/db + miR antagomir + sh‐NRF2 and db/db + miR antagomir + sh‐NC groups; and (6) db/db + miR antagomir + sh‐NRF2 + Vehicle and db/db + miR antagomir + sh‐NRF2 + Ferrostatin‐1 groups. For miR‐145‐5p intervention, db/db mice received 50 μg of miR‐145‐5p antagomir or antagomir NC (GenePharma, Shanghai, China) via tail vein once weekly. 42 To modulate ferroptosis, mice were intraperitoneally administered Erastin (30 mg/kg; MedChemExpress, Monmouth Junction, NJ, USA), 43 , 44 Ferrostatin‐1 (1 mg/kg; MCE, Monmouth Junction), 45 , 46 or an equivalent volume of vehicle every other day. For NRF2 silencing, recombinant adeno‐associated virus (rAAV) carrying short hairpin RNA targeting NRF2 (sh‐NRF2) or a negative control vector (sh‐NC) (1 × 10 11 virus particles, R&S Biotechnology) was administered via tail vein. 47 A total of 60 mice were included in the study, comprising 10 experimental groups with six mice per group. Animals used in different experiments and result sections were derived from the same experimental cohorts whenever applicable. At 22 weeks of age, 48 urine and blood samples were collected before euthanasia by intraperitoneal administration of 0.5% pentobarbital sodium (120 mg/kg; P3761, Sigma‐Aldrich). Kidneys were subsequently harvested for downstream analyses. One kidney from each mouse was fixed in 4% paraformaldehyde, embedded in paraffin after dehydration, and sectioned for histological analyses. The contralateral kidney was snap‐frozen and stored at −80°C for reverse transcription‐quantitative polymerase chain reaction (RT‐qPCR), Western blot assay, enzyme‐linked immunosorbent assay (ELISA), and other biochemical analyses. Whole blood samples were centrifuged at 3,000 ×  g for 15 min to obtain serum. Blood glucose concentrations were measured using a glucose analyzer (Roche, Germany) according to the manufacturer's instructions. Urinary albumin, urinary creatinine, serum creatinine, and blood urea nitrogen (BUN) levels were measured using commercially available ELISA kits according to the manufacturers' instructions. Urinary BUN was quantified using a mouse BUN detection kit (BC1503, Solarbio), whereas urinary and serum creatinine levels were determined using a mouse creatinine ELISA kit (SBJ‐M0081, SenBeiJia Biological Technology Co., Ltd., Nanjing, China). The urinary albumin‐to‐creatinine ratio (ACR) was subsequently calculated to evaluate renal function. Renal tissues or cultured cells were homogenized and centrifuged, and the supernatants were collected for biochemical analyses. Ferrous iron (Fe 2+ ), GSH, and malondialdehyde (MDA) levels were determined using an iron detection kit (BC5410, Solarbio), a GSH assay kit (S0053, Beyotime, Shanghai, China), and an MDA assay kit (BC0025, Solarbio), respectively, according to the manufacturers' protocols. Kidney tissues were fixed in 10% neutral‐buffered formalin, paraffin‐embedded, and sectioned at a thickness of 4 μm. Following deparaffinization and rehydration, sections were stained with hematoxylin and eosin (H&E) using standard protocols. Histopathological alterations were examined and photographed under a light microscope (Olympus Corporation, Tokyo, Japan). Paraffin‐embedded renal sections were incubated with 1% periodic acid solution (S0124, Bioss) for 10 min, followed by staining with Schiff reagent ( R20524 , Shanghai Yuanye Bio‐Technology Co., Ltd.) in the dark for 10 min. After sodium hydrogen sulfite rinses and hematoxylin counterstaining ( B25380 , Shanghai Yuanye Bio‐Technology Co., Ltd.), sections were dehydrated, cleared, and mounted with neutral resin (C0173, Beyotime). Images were acquired using a light microscope (Olympus, Japan). Glomerular volume and PAS‐positive areas were quantified using Image‐Pro Plus 6.0 software (Media Cybernetics, Rockville, MD, USA). 49 , 50 , 51 Renal fibrosis was evaluated using a Masson trichrome staining kit (G1006, Wuhan Servicebio Technology Co., Ltd., Wuhan, China) following the manufacturer's instructions. Briefly, paraffin sections were deparaffinized, rehydrated, stained sequentially with hematoxylin, Ponceau‐acid fuchsin, phosphomolybdic acid, and aniline blue, followed by dehydration and mounting. Collagen deposition was quantified as the percentage of Masson‐positive area using the Image‐Pro Plus 6.0 software. 49 , 50 Paraffin‐embedded renal sections were deparaffinized, rehydrated, and subjected to heat‐induced antigen retrieval using citrate buffer (PR30001, Proteintech, Wuhan, China). Endogenous peroxidase activity was blocked with 3% hydrogen peroxide, followed by blocking with 5% goat serum (C0265, Beyotime). Sections were incubated overnight at 4°C with primary antibodies against kidney injury molecule‐1 (KIM‐1; ab78494, Abcam, Cambridge, UK) and neutrophil gelatinase‐associated lipocalin (NGAL; ab216462, Abcam). After washing, sections were incubated with a horseradish peroxidase (HRP)‐conjugated secondary antibody (ab6721, Abcam), and immunoreactivity was visualized using 3,3′‐diaminobenzidine (DAB; P0202, Beyotime). Sections were counterstained with hematoxylin, dehydrated, mounted, and examined under a light microscope (Olympus, Japan). The percentages of KIM‐1‐ and NGAL‐positive cells were quantified from randomly selected microscopic fields. Mouse RTECs (TCMK‐1; Icellbioscience, Shanghai, China; icell‐m089) were cultured in Minimum Essential Medium (HyClone, Logan, Utah, USA; Cat.No. SH30024.02) supplemented with 10% fetal bovine serum (Ausgenex, Gold Coast, Queensland, Australia; C0227) and 1% penicillin–streptomycin (Gibco, Grand Island, New York, USA; Cat.No. 15140‐122). Cells were maintained at 37°C in a humidified incubator containing 5% CO 2 and passaged every 48–72 h. TCMK‐1 cells were seeded into six‐well plates (Corning Inc., Corning, NY, USA; Cat.No. 3516) at a density of 2 × 10 6 cells per well. After reaching approximately 70–80% confluence, cells were transfected with 200 nmol/L miR‐145‐5p antagomir and/or 100 ng NRF2‐specific small interfering RNA (si‐NRF2), SLC7A11‐specific siRNA (si‐SLC7A11), or the corresponding negative controls using Lipofectamine 2000 (Thermo Fisher, Waltham, MA, USA; Cat.No. 11668030), according to the manufacturer's instructions. All oligonucleotides were purchased from GenePharma (Shanghai, China). Forty‐eight hours after transfection, cells were harvested for subsequent experiments. TCMK‐1 cells were assigned to the following experimental groups: Normal glucose (NG) group: cells cultured in medium containing 5.5 mM glucose (Y0001745, Sigma‐Aldrich) for 48 h; Mannitol (MT) group: cells cultured in 5.5 mM glucose supplemented with 44.5 mM mannitol as an osmotic control; HG group: cells cultured in 50 mM glucose for 48 h 52 ; HG + miR antagomir group or HG + antagomir NC group: cells cultured under HG conditions and transfected with the corresponding NC or miR‐145‐5p antagomir; HG + miR antagomir + si‐NRF2 group, HG + miR antagomir + si‐SLC7A11 group, or HG + miR antagomir + si‐NC group: cells cultured under HG conditions and co‐transfected with miR‐145‐5p antagomir together with si‐NRF2, si‐SLC7A11, or si‐NC for 48 h. Total RNA was extracted from TCMK‐1 cells and mouse renal tissue using a Total RNA Extraction Kit (Axygen, Union City, CA, USA; AP‐MN‐MS‐RNA‐250). Messenger RNA was reverse‐transcribed into complementary DNA (cDNA) using the PrimeScript RT Reagent Kit (Takara, RR037Q), whereas miRNA was reverse‐transcribed using the StarScript III miRNA RT Kit (by Stem‐loop) (GenStar, A237‐10). Quantitative PCR detection was performed using SYBR® Premix Ex TaqTM II (Takara, RR390Q) on an ABI7900HT Real‐Time PCR System (Applied Biosystems, Foster City, CA, USA). The amplification protocol consisted of an initial denaturation at 95°C for 10 min, followed by 40 cycles of 95°C for 10 s, 60°C for 20 s, and 72°C for 34 s. Relative gene expression was calculated using the 2 −ΔΔCt method, with glyceraldehyde‐3‐phosphate dehydrogenase or U6 serving as the internal controls for mRNA and miRNA, respectively. Primer sequences are listed in Table  1 . Primer sequences Potential binding sites between miR‐145‐5p and the 3′‐untranslated region (3′‐UTR) of NRF2 were predicted using the TargetScan database ( https://www.targetscan.org/vert_80/ ). The wild‐type (NRF2‐WT) and mutant (NRF2‐MUT) NRF2 3’‐UTR sequences containing the predicted miR‐145‐5p binding sites were cloned into the pmiR‐GLO luciferase reporter vector (Promega, Madison, WI, USA). TCMK‐1 cells were co‐transfected with NRF2‐WT or NRF2‐MUT plasmids together with either miR‐145‐5p mimic or mimic NC using Lipofectamine™ 2000 (Thermo Fisher Scientific). After 48 h, firefly and Renilla luciferase activities were measured using a Dual‐Luciferase Reporter Assay System (Promega; E1910), and relative luciferase activity was calculated according to the manufacturer's instructions. Chromatin immunoprecipitation (ChIP) was performed using a ChIP assay kit (P2078, Beyotime, Shanghai, China) according to the manufacturer's instructions. TCMK‐1 cells at 80–90% confluence were cross‐linked with 1% formaldehyde for 10 min, followed by quenching with 125 mM glycine for 5 min. After washing with PBS, cells were lysed in lysis buffer containing protease inhibitors, and chromatin was fragmented by sonication to an average size of 200–1,000 bp. Following 10‐fold dilution, an aliquot of chromatin was reserved as the input control, whereas the remaining chromatin was immunoprecipitated overnight at 4°C with an anti‐NRF2 antibody (ab325240, Abcam) or normal rabbit IgG as the NC. Immune complexes were captured using Protein A/G magnetic beads, washed sequentially, and subjected to reverse cross‐linking with 200 mM NaCl and proteinase K (50 μg/mL) at 65°C for 4 h. DNA was subsequently purified and analyzed by quantitative PCR. TCMK‐1 cells were fixed in 2% glutaraldehyde for 12 h, postfixed with 1% osmium tetroxide (OsO4) for 2 h, dehydrated through a graded ethanol series, embedded, sectioned, and stained with toluidine blue. Mitochondrial ultrastructural characteristics of ferroptosis were examined using a transmission electron microscope (HT7700, Hitachi). Total protein was extracted from mouse renal tissues or TCMK‐1 cells using radioimmunoprecipitation assay (RIPA) buffer (P0039; Beyotime) supplemented with protease inhibitor cocktail (P1006; Beyotime). Following centrifugation at 13,000 rpm for 10 min, protein concentrations were measured using a bicinchoninic acid (BCA) protein assay kit (P0012; Beyotime). Equal amounts of proteins were separated by 10% sodium dodecyl sulfate‐polyacrylamide gel electrophoresis (SDS‐PAGE) and transferred onto polyvinylidene fluoride membranes (Millipore, Burlington, MA, USA; IPVH20200). Membranes were blocked with 5% nonfat milk for 2 h and incubated at 4°C with the following primary antibodies: β‐Actin (CST, Beverly, MA, USA, #4970), NRF2 (CST, #12721), SLC7A11 (Proteintech, 26,864‐1‐AP), GPX4 (Abcam, ab125066), ACSL4 (ab155282, Abcam), E‐cadherin (ab308347, Abcam), Vimentin (ab92547, Abcam), and α‐SMA (ab124964, Abcam). After washing, membranes were incubated with an HRP‐conjugated goat anti‐rabbit IgG secondary antibody (Servicebio, Wuhan, China; GB23303) for 1 h at room temperature. Protein bands were visualized using enhanced chemiluminescence reagent (Millipore; WBKLS0500) and quantified with Image‐Pro Plus 6.0 software (Media Cybernetics, Rockville, MD, USA). β‐Actin was used as the internal loading control. All experiments were performed independently at least three times. Statistical analyses were performed using GraphPad Prism 9.5 software (GraphPad Software, San Diego, CA, USA). Measurement data are presented as the mean ± standard deviation (SD). Normality was assessed using the Shapiro–Wilk test, whereas homogeneity of variance was evaluated using the F ‐test or Brown–Forsythe test, as appropriate. Comparisons between two groups were performed using the independent sample t ‐test, while comparisons among multiple groups were analyzed using one‐way analysis of variance ( anova ) followed by Tukey's post hoc multiple comparison test. All statistical tests were two‐tailed, and P  < 0.05 was considered statistically significant.

Figure S1. Expression of miR‐145‐5p and NRF2 in clinical DN samples from GEO datasets. (a) The volcano plot of differentially expressed miRNAs in the GSE114477 dataset (4 healthy controls and 4 DN samples); (b) The volcano plot of differentially expressed mRNAs in the GSE142025 dataset (9 controls and 28 DN samples). The screening criteria for differentially expressed miRNAs were set at fold change ≥ 1.5 and FDR < 0.05, with volcano plots generated using Sangerbox ( http://sangerbox.com/home.html ). Figure S2 . High osmotic pressure has no effect on ferroptosis and miR‐145‐5p expression in TCMK‐1 cells.