Overexpression of LINC00472 alleviates diabetic kidney disease progression via the miR-21-5p/STAT3 axis Authors: Pengcheng Sun, Yutong Hu, Zihui Xu, Fei Xiong Published: 2026 ABSTRACT AbstractTo reveal the diagnostic value and regulatory role of LINC00472 in diabetic kidney disease (DKD). A total of 103 patients with type 2 diabetes mellitus (T2DM) and 111 patients with DKD were enrolled. Serum and urine samples were collected within 24 h of admission. Logistic regression predicted factors inducing DKD, while ROC analysis evaluated the diagnostic value of lncRNA for DKD. Gene expression was detected by RT-qPCR. Cell function was assessed using CCK-8 and Transwell assays. Cytokine concentrations were measured by ELISA. Dual luciferase reporter assays and RNA immunoprecipitation validated intergenic targeting relationships. LINC00472 was downregulated in DKD patients and high-glucose (HG)-induced human glomerular mesangial cells (HGMCs). LINC00472 demonstrated diagnostic value for DKD and correlated with renal function indicators (inverse correlation with serum creatinine and urine albumin-to-creatinine ratio; positive correlation with estimated glomerular filtration rate). Transfection of oe-LINC00472 inhibited HGMC function while reducing expression of fibrosis genes and inflammatory factors. miR-21-5p, a downstream target of LINC00472, was upregulated in DKD patients. A miR-21-5p mimic could reverse the protective effects of LINC00472 overexpression on HGMC cells. STAT3, a downstream target gene of the LINC00472/miR-21-5p axis involved in DKD, was downregulated in DKD patients and after HG treatment. Overexpression of LINC00472 suppresses HGMC cell dysfunction by sponging miR-21-5p, thereby inhibiting fibrosis and inflammation and ameliorating DKD progression.Keywords: Diabetic kidney disease, LINC00472, miR-21-5p, ceRNA, STAT3, biomarker Introduction Diabetic kidney disease (DKD) represents the long-term accumulation of complications in individuals with diabetes. Statistics indicate that 30–40% of diabetic patients develop renal impairment to varying degrees [1]. Research demonstrates that DKD is one of the primary causes of end-stage renal disease (ESRD) progression in patients with type 2 diabetes mellitus (T2DM) [2–4]. With the global increase in diabetes patients, the incidence of DKD has also risen annually. This not only impacts the quality of life for DKD patients but also heightens the risk of cardiovascular disease [5]. Therefore, in-depth exploration of the pathogenesis of DKD, along with the identification of effective diagnostic biomarkers and therapeutic targets, is crucial for timely detection and prevention of DKD progression. In recent years, the role of long non-coding RNAs (lncRNAs) in the pathogenesis of DKD has garnered increasing attention [6,7]. Abnormal expression of multiple lncRNAs has emerged as potential DKD biomarkers and therapeutic targets [8]. Furthermore, lncRNAs participate in cellular functions and disease progression through various mechanisms, including gene expression regulation and involvement in signal transduction. In DKD development, lncRNAs can directly serve as causative agents. For instance, Yang et al. [9] identified that lncRNA GAS5 gene polymorphisms correlate with DKD progression. Qin et al. reported that lncRNA H19 drives glomerulosclerosis and tubular fibrosis. On the other hand, lncRNAs indirectly mediate certain nephropathic pathways. Studies indicate that lnc-Traf3ip2 exacerbates renal fibrosis in DKD mice via the NF-κB signaling pathway [10]. Liu et al. [11] found that dysregulation of lncRNA DANCR induces renal mesangial extracellular matrix accumulation through the TGF-β/SMAD pathway. LINC00472, as a specific lncRNA, exhibits abnormal expression in various diseases including cancer and cardiovascular diseases [12,13] and participates in the diagnosis and regulation of certain conditions. Research indicates that LINC00472 holds diagnostic value for patients with osteoarthritis [14] and chronic thromboembolic pulmonary hypertension [15]. Furthermore, reduced LINC00472 expression in renal cell carcinoma has been associated with poor patient prognosis [16]. Liu et al. [17] reported that LINC00472 overexpression induces cell cycle arrest in oral squamous cell carcinoma. However, its function in DKD remains incompletely understood. Preliminary studies indicate LINC00472 is downregulated in DKD patient samples and may hold potential for diagnosing DKD progression [18]. Our laboratory’s preliminary predictions identified binding sites between miR-21-5p and LINC00472. Although previous studies demonstrated that miR-21-5p is significantly expressed in DKD patients and correlates with renal fibrosis [19], However, the upstream regulatory mechanism of miR-21-5p in DKD remains unelucidated. Therefore, elucidating the regulatory role of LINC00472 in DKD and exploring the interaction mechanism between miR-21-5p and LINC00472 is crucial for comprehensively understanding the pathogenesis of DKD. Based on the above findings, we collected serum and urine samples from DKD patients and healthy controls to measure the expression levels of LINC00472 and miR-21-5p. We then validated their target relationship and their effects on mesangial cell function through cellular experiments. This study not only holds promise for identifying novel biomarkers for DKD diagnosis but may also provide new therapeutic strategies and approaches by modulating the LINC00472/miR-21-5p axis. Materials and methods Subject inclusion A total of 214 patients with T2DM (including 111 with DKD) were enrolled in the study. All subjects received treatment at Traditional Chinese and Western Medicine Hospital of Wuhan, Tongji Medical College, Huazhong University of Science and Technology between October 2023 and September 2024. Additionally, 98 healthy individuals undergoing physical examinations during the same period were assigned to the control group. Fasting peripheral venous blood and 24 h urine samples were collected within 24 h of admission. T2DM diagnosis followed the American Diabetes Association (ADA) criteria [20]. DKD was confirmed in patients meeting either: ≥2 instances of urinary albumin-to-creatinine ratio (UACR) ≥30 mg/g in three tests; or sustained estimated glomerular filtration rate (eGFR) <60 mL/min/1.73 m2 for ≥3 months. However, the following groups were excluded: patients with non-diabetic kidney disease, kidney transplant recipients, patients with coagulation disorders, those with a history of nonsteroidal anti-inflammatory drug (NSAID) or other nephrotoxic medication use, patients with organ failure, minors, and patients with impaired consciousness. This study was approved by the Ethics Committee of Traditional Chinese and Western Medicine Hospital of Wuhan, Tongji Medical College, Huazhong University of Science and Technology. (Approval Number:2021K-K063) Patients or their legal guardians provided written informed consent. Cell model establishment Human glomerular mesangial cells (HGMC, Catalog No.: CP-H067) were purchased from Procell Life Science Technology Co., Ltd., Wuhan. Cells were cultured in the corresponding complete medium (CM-H067, Procell, Wuhan). Cells were maintained at 37 °C in a 5% CO2 incubator. Passaging or subsequent procedures were performed when cell confluence reached 70% or higher. Glomerular mesangial cells were treated with 25 mmol/L D-glucose (50-99-7, MACKLIN, Shanghai) for 24 h to establish a hyperglycemic (HG) model. The control group was administered 5 mmol/L D-glucose plus 20 mmol/L mannitol to maintain osmotic balance with the HG group. The choice of a 25 mmol/L D-glucose concentration was based on previous studies simulating hyperglycemia in DKD models [21]. Cell transfection Transfection was carried out using Lipo8000™ (C0533FT, Beyotime, Shanghai). Briefly, cells were seeded in 24‑well plates and the medium was replaced just before transfection. In a sterile tube, 20 pmol of oligonucleotide (oe‑NC, oe‑LINC00472, mimic NC, or miR mimic) was mixed with 25 μL of basal medium, followed by the addition of 0.8 μL of Lipo8000 reagent. A 25‑μL aliquot of this complex was applied to each well, and the plates were incubated for 48 h. RT-qPCR RNA was extracted from serum and cultured cells via Trizol‑based phenol‑chloroform precipitation. Separate reverse transcription reactions were set up using the 1st cDNA Synthesis Kit (YEASEN, 12946) for lncRNA/mRNA and the miRNA cDNA First Strand Synthesis Kit (Accurate, AG11717) for miRNA. qPCR was carried out with Super Real Pre‑Mix Plus (TIANGEN, FP205) and Ultra SYBR Mixture (Cwbiotech, CW0957), respectively, on an ABI 7500 instrument. Melting curves were recorded, and relative quantification was performed by the 2^‑ΔΔCt method. β‑actin and U6 were used as internal controls for lncRNA/mRNA and miRNA, respectively. Primer details are available in Supplementary Table 1. CCK-8 Cell proliferation was evaluated using the Cell Counting Kit-8 (Beyotime, C0038, Shanghai) as previously described [22]. Briefly, transfected cells were trypsinized, pelleted, and resuspended in DMEM. Aliquots (100 µL, ∼2 × 10³ cells/well) were seeded and incubated for 24 h. At ∼70% confluence, 10 µL CCK‑8 was added per well (blank controls: medium + CCK‑8, no cells). After another hour, absorbance at 450 nm was measured using an Envision Nexus plate reader. Transwell For invasion assays, the transwell inserts were first coated with matrix gel and hydrated. Transfected cells were serum‑starved overnight. The Cell Migration and Invasion Staining Kit (Solarbio, G4740) was used for both migration and invasion. Lower chambers received 600 μL of medium with 10% FBS; upper chambers were seeded with 100 μL of cell suspension. After 24 h incubation, non‑migrated cells were removed with PBS. Inserts were fixed (500 μL ice‑cold fixative, 20 min), rinsed, and stained (1× staining solution, 10 min). After two water washes, membranes were examined under an XDS‑440C microscope. Cells in five random fields were counted and averaged. ELISA ELISA kits for human TNF-α, IL-6, and IL-1β (cat. E-EL-H0109, H6156, H0149; Wuhan Elabscience) were used. Cell culture media were collected and centrifuged (1000 × g, 4 °C, 20 min), and the supernatants were kept. Per kit instructions, 100 μL each of dilution series, sample supernatant, or sample diluent were added to standard, sample, or blank wells, respectively. Plates were sealed and incubated at 37 °C for 90 min. After removing liquid, 100 μL of biotinylated antibody was added for 1 h. Wells were aspirated and washed three times (plate washer). Then 100 μL of HRP-conjugated antibody was added for another 30 min at 37 °C. After five washes, 90 μL of TMB substrate was added and developed in the dark for 15 min. Stop solution (50 μL) was added, mixed, and absorbance read at 450 nm. Assessment of cell hypertrophy Lysis the cells using 1× RIPA lysis buffer (Cat. No. #BL504A; Biosharp) containing 1% PMSF and a mixture of protease/phosphatase inhibitors (Cat. No. #PR20038; Proteintech). Determine total protein content using the BCA protein quantification kit (Cat. No. #PK10026; Proteintech), and use the ratio of total protein to cell number as an indicator of cell hypertrophy. Western blotting After separating the protein samples by SDS-PAGE, they were transferred to a nitrocellulose membrane. The membrane was first blocked at room temperature for 2 h with TBST (Tris-buffered saline containing Tween-20) supplemented with skim milk, followed by overnight incubation with the primary antibody at 4 °C. The membrane was then washed and incubated with horseradish peroxidase (HRP)-labeled secondary antibody at room temperature for 2 h. Protein bands were visualized using ECL Prime Western Blotting Substrate, and band intensity was quantified using ImageJ software. Full-length uncropped Western blot images for all figures are provided in Supplementary Figures S1 and S2. The Western blot procedures were performed as described previously [23]. Subcellular localization analysis Cellular fractionation was performed using a kit (E101, Vazyme) following the manufacturer’s instructions with reference to previously described procedures [22]. Cells were pelleted (900 × g, 4 °C, 3 min), resuspended in 200 μL Buffer A, vortexed (10 s), and incubated on ice (15 min). After adding 11 μL Buffer B (vortex 5 s, ice 4 min), centrifugation (17,000 × g, 30 s) gave the cytoplasmic fraction (supernatant). The pellet was washed with cold PBS, then 100 μL nuclear buffer was added, vortexed (30 s), and incubated on ice (30 min with vortexing every 10 min). Final centrifugation yielded the nuclear fraction. Dual luciferase reporter assay Transfection of wild-type/mutant vectors, oligonucleotides, and controls was done using Lipo8000™. Dual luciferase activity was measured with a Firefly & Renilla assay kit (LA11037, Bio-Lab, Beijing). After removing the culture medium, 50 μL of lysis buffer (with protease inhibitor) was added per well (48-well plate) and incubated at 37 °C for 15 min. Lysates were centrifuged at 16,000 × g for 5 min, and the supernatant was collected. Then, 20 μL of supernatant was mixed with 100 μL of firefly luciferase substrate, and luminescence was read. Subsequently, 100 μL of Stop & Renilla reagent was added to the same tube, and a second reading was taken. RIP Binding interactions of the target gene were assessed using an RIP kit (abs50071, Absin). About 2 × 107 cells were pelleted and lysed in RIP lysis buffer (sonication-assisted, final volume 400 μL). After high-speed centrifugation (4 °C, 10 min), the supernatant was collected. Two tubes (A and B) each received 30 μL of magnetic beads, washed five times with wash buffer (10× volume), then resuspended. Tube A was incubated with 5 μg target antibody, tube B with 5 μg isotype IgG (room temperature, 2 h). Beads were washed again five times. Then, 900 μL RIP buffer plus 150 μL cell lysate was added to tube A (and similarly to tube B). Incubation proceeded at 4 °C overnight. After discarding the supernatant and washing beads, each pellet was resuspended in 300 μL RIP wash buffer. From each, 100 μL was taken for protein analysis (elution buffer, boiling), and the remainder was used for RNA purification. Database sources miRNA target genes were aggregated from four distinct databases: ENCORI (https://rnasysu.com/encori/), miRWalk (http://mirwalk.umm.uni-heidelberg.de/), Targetscan (https://www.targetscan.org/vert_80/), and miRDB (https://mirdb.org/). mRNAs involved in DKD regulation were obtained from the CTD website (https://ctdbase.org/detail.go?type=disease&acc=MESH%3AD055113). The intersection of these two datasets yielded the overlapping gene information required for our study. Subsequently, we constructed a protein-protein interaction (PPI) network for the overlapping genes using the STRING software (https://cn.string-db.org/). Enrichment analysis of the target genes’ GO and KEGG pathways was performed via the Micro Bioinformatics website (https://bioinformatics.com.cn/). Statistical analysis Clinical data were analyzed using SPSS 26.0, and experimental results were plotted using GraphPad Prism 9.5. Differences between groups were compared using t-tests and analysis of variance (ANOVA). ROC curves assessed the diagnostic value of lncRNA for predicting DKD, while logistic regression analysis identified potential factors inducing DKD. Pearson correlation analysis examined the relationship between lncRNA expression and renal function indicators as well as miRNAs. p < 0.05 indicated statistically significant differences. Results Baseline characteristics and LINC00472 expression in healthy individuals, T2DM, and DKD patients Statistical analysis revealed (Table 1) that systolic blood pressure (SBP), total cholesterol (TC), triglycerides (TG), fasting blood glucose (FBG), and glycated hemoglobin (HbA1c) levels were significantly elevated in T2DM and DKD patients compared to normal ranges. However, age, sex, smoking history, and drinking history showed minimal variation across the three groups. Meanwhile, compared to patients with T2DM alone, those with DKD exhibited further elevations in blood pressure (SBP/DBP), lipid profiles (TC, TG, LDL-C), glycemic markers (FBG, HbA1c), inflammatory markers (hs-CRP), and renal function indicators (Scr, eGFR, UACR). Table 1.Comparison of clinical baseline characteristics among subjects. Control (n = 98)T2DM (n = 103)DKD (n = 111) Statistical characteristics Age (years)62.19 ± 16.0063.04 ± 14.8562.84 ± 14.89 Sex (man, n%)48 (48.98%)52 (50.49%)54 (48.65%)Smoking No56 (57.14%)53 (51.46%)53 (47.75%) Yes42 (42.86%)50 (48.54%)58 (52.25%)Drinking No49 (50%)53 (51.46%)51 (45.95%) Yes49 (50%)50 (48.54%)60 (54.05%) Blood pressure SBP (mmHg)121.52 ± 14.20130.92 ± 10.78a138.43 ± 8.10ab DBP (mmHg)81.23 ± 10.4383.30 ± 5.4192.13 ± 6.84ab Blood lipids TC (mmol/L)4.95 ± 0.285.18 ± 0.44a5.69 ± 0.44ab TG (mmol/L)1.61 ± 0.231.80 ± 0.30a2.29 ± 0.47ab HDL-C (mmol/L)1.32 ± 0.301.27 ± 0.491.17 ± 0.28a LDL-C (mmol/L)3.28 ± 0.103.30 ± 0.343.56 ± 0.38ab Blood glucose FBG (mmol/L)4.85 ± 0.467.13 ± 0.97a7.49 ± 0.48ab HbA1c (%)5.01 ± 0.496.46 ± 0.52a6.94 ± 0.69ab Renal function parameters hs-CRP (mg/L)1.53 ± 0.522.75 ± 1.123.62 ± 1.14ab Scr (μmol/L)84.98 ± 6.1885.95 ± 7.5995.28 ± 7.24ab eGFR (mL/min/1.73 m2)99.85 ± 7.4698.20 ± 7.9759.99 ± 4.33ab UACR (mg/g)25.44 ± 3.8526.38 ± 3.5238.79 ± 6.01abAbbreviations: T2DM, Type 2 diabetes mellitus; DKD, diabetic kidney disease; SBP, systolic blood pressure; DBP, diastolic blood pressure; TC, total cholesterol; TG, triglycerides; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; FBG, fasting blood glucose; HbA1c, glycated hemoglobin; hs-CRP, high-sensitivity C-reactive protein; Scr, serum creatinine; eGFR, estimated glomerular filtration rate; UACR, urine albumin-to-creatinine ratio. a p < 0.05 vs. control, b p < 0.05 vs. T2DM. Comparison of clinical baseline characteristics among subjects. Abbreviations: T2DM, Type 2 diabetes mellitus; DKD, diabetic kidney disease; SBP, systolic blood pressure; DBP, diastolic blood pressure; TC, total cholesterol; TG, triglycerides; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; FBG, fasting blood glucose; HbA1c, glycated hemoglobin; hs-CRP, high-sensitivity C-reactive protein; Scr, serum creatinine; eGFR, estimated glomerular filtration rate; UACR, urine albumin-to-creatinine ratio. a p < 0.05 vs. control, b p < 0.05 vs. T2DM. Diagnostic value of LINC00472 in patients with DKD Detection revealed significantly downregulated LINC00472 levels in serum samples from patients with T2DM and DKD (p < 0.0001, Figure 1A). Furthermore, serum LINC00472 demonstrated an AUC of 0.8228 (95% CI: 0.7646–0.8809, p < 0.0001, Figure 1B) for predicting T2DM, with a sensitivity of 85.44% and specificity of 70.41%. For predicting DKD, the AUC was 0.8600 (95% CI: 0.8131–0.9069, p < 0.0001, Figure 1C), with a sensitivity of 70.27% and specificity of 83.50%. Second, the analysis revealed (Figure 1D-F) that serum LINC00472 levels in DKD patients were inversely correlated with Scr (r = −0.7231) and UACR (r = −0.7160), and positively correlated with eGFR (r = 0.7386). Furthermore, Logistic regression analysis (Table 2) indicated that lower LINC00472 expression (OR = 0.372, p = 0.002), elevated SBP (OR = 2.068, p = 0.020), reduced eGFR (OR = 0.508, p = 0.035), and elevated UACR (OR = 2.163, p = 0.018) were all independent risk factors contributing to the progression from T2DM to DKD. Expression of LINC00472 in three populations and correlation analysis with renal function in DKD patients. A. Expression of LINC00472 in healthy individuals, T2DM, and DKD populations. ROC analysis of LINC00472 for predicting T2DM (B) and DKD (C). AUC denotes area under the curve; CI indicates confidence interval. Correlation analysis of serum LINC00472 expression with Scr (D), eGFR (E), and UACR (F) in DKD patients. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. T2DM, type 2 diabetes mellitus; DKD, diabetic kidney disease; ROC, receiver operating characteristic; AUC, area under the curve; CI, confidence interval; Scr, serum creatinine; eGFR, estimated glomerular filtration rate; UACR, urine albumin-to-creatinine ratio. Table 2.Logistic regression analysis of factors inducing DKD.FactorsP valueOR95%CILINC004720.0020.3720.196-0.703Age (years)0.7541.1070.588-2.084Sex (man, n%)0.2441.4470.778-2.693Smoking0.9890.9950.530-1.869Drinking0.2681.4170.765-2.626SBP (mmHg)0.0202.0681.120-3.820DBP (mmHg)0.4401.2760.688-2.368TC (mmol/L)0.1401.5860.859-2.927TG (mmol/L)0.1771.5440.822-2.901HDL-C (mmol/L)0.1140.6050.324-1.129LDL-C (mmol/L)0.3221.3640.737-2.524FBG (mmol/L)0.0961.6890.912-3.131HbA1c (%)0.3581.3390.718-2.495hs-CRP (mg/L)0.0981.6800.908-3.109Scr (μmol/L)0.3750.7490.396-1.417eGFR (mL/min/1.73 m2)0.0350.5080.271-0.954UACR (mg/g)0.0182.1631.141-4.101Abbreviations: DKD, diabetic kidney disease; SBP, systolic blood pressure; DBP, diastolic blood pressure; TC, total cholesterol; TG, triglycerides; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; FBG, fasting blood glucose; HbA1c, glycated hemoglobin; hs-CRP, high-sensitivity C-reactive protein; Scr, serum creatinine; eGFR, estimated glomerular filtration rate; UACR, urine albumin-to-creatinine ratio. p < 0.05 indicates that the results are statistically significant. Logistic regression analysis of factors inducing DKD. Abbreviations: DKD, diabetic kidney disease; SBP, systolic blood pressure; DBP, diastolic blood pressure; TC, total cholesterol; TG, triglycerides; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; FBG, fasting blood glucose; HbA1c, glycated hemoglobin; hs-CRP, high-sensitivity C-reactive protein; Scr, serum creatinine; eGFR, estimated glomerular filtration rate; UACR, urine albumin-to-creatinine ratio. p < 0.05 indicates that the results are statistically significant. Effects of oe-LINC00472 on HGMC cell function and inflammatory levels Studies indicate that HG treatment reduces LINC00472 levels in HGMC cells, while transfection with oe-LINC00472 upregulates its expression (p < 0.0001, Figure 2A). Compared to the control group, HG treatment significantly enhances HGMC cell proliferation, migration, and invasion capabilities (p < 0.0001, Figure 2B-D). However, overexpression of LINC00472 markedly suppressed HGMC proliferation and significantly diminished their invasive and migratory capabilities. Furthermore, we observed that oe-LINC00472 transfection reduced fibrotic genes at both the mRNA and protein levels, and decreased inflammatory cytokine concentrations in HGMC cells (p < 0.01, Figure 2E-G). Effects of HG and oe-LINC00472 treatment on HGMC cell functions. A. Changes in LINC00472 expression in HGMC cells after HG treatment and establishment of an overexpression cell line. Effects of different transfection treatments on HGMC cell proliferation (B), migration (C), and invasion (D). E-F. Changes in fibrosis-related gene at both the mRNA and protein levels in HGMC cells after oe-LINC00472 transfection. G. Effects of LINC00472 overexpression on inflammatory levels in HGMC cells. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. HG, high glucose; oe-LINC00472, overexpression vector for LINC00472; HGMC, human glomerular mesangial cell; NC, negative control; FN, fibronectin; Col-IV, collagen type IV; TGF-β1, transforming growth factor-β1; TNF-α, tumor necrosis factor-α; IL-6, interleukin-6; IL-1β, interleukin-1β. miR-21-5p is a downstream target gene of LINC00472 Following nuclear-cytoplasmic separation, we detected that LINC00472 was predominantly localized in the cytoplasm (70.33%, Figure 3A). Bioinformatics analysis revealed a target binding site between LINC00472 and miR-21-5p (Figure 3B). Therefore, we constructed mutant plasmids to test their interaction. Results showed that in the presence of the wild-type plasmid, fluorescence intensity significantly increased upon transfection with the miR inhibitor; however, it was markedly suppressed upon transfection with the miR mimic (p < 0.0001, Figure 3C). Second, both LINC00472 and miR-21-5p were significantly enriched by target magnetic beads in RIP assays (p < 0.0001, Figure 3D). Furthermore, we observed significantly elevated miR-21-5p expression levels in serum samples from T2DM and DKD patients (p < 0.0001, Figure 3E). Additionally, a negative correlation was detected between LINC00472 and miR-21-5p expression in DKD patient serum (r = −0.7599, p < 0.0001, Figure 3F). Expression of miR-21-5p in DKD patients and validation of its targeting relationship with LINC00472. A. Subcellular localization analysis of LINC00472. B. Binding site information between LINC00472 and miR-21-5p. C. Dual luciferase reporter assay: miR-21-5p mimic significantly reduced the luciferase activity of the LINC00472-WT reporter compared with mimic NC, while miR-21-5p inhibitor significantly increased the activity compared with inhibitor NC; LINC00472-WT was unaffected. D. RNA immunoprecipitation (RIP) showing enrichment of LINC00472 and miR-21-5p by Ago2 antibody compared with IgG control. E. Expression of miR-21-5p in serum from three population groups. F. Correlation analysis of LINC00472 and miR-21-5p expression in serum from DKD patients. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. DKD, diabetic kidney disease; WT, wild-type; MUT, mutant; NC, negative control; RIP, RNA immunoprecipitation; Ago2, Argonaute 2; IgG, immunoglobulin G; T2DM, type 2 diabetes mellitus; miR inhibitor, miR-21-5p inhibitor; miR mimic, miR-21-5p mimic. Protective effect of miR-21-5p mimic against oe-LINC00472 on HGMC cells Results indicate that transfection with oe-LINC00472 suppresses miR-21-5p expression, whereas HG treatment and transfection with miR mimic upregulate this gene in HGMC cells (p < 0.0001, Figure 4A). Second, we observed enhanced proliferation capacity in HGMC cells following miR mimic transfection (p < 0.0001, Figure 4B), along with increased proportions of migrating and invasive cells (p < 0.01, Figure 4C, D). Furthermore, miR mimic transfection promoted the expression of fibrosis-related genes in HGMC cells (p < 0.05, Figure 4E) while simultaneously increasing inflammatory levels (p < 0.05, Figure 4F). Importantly, HG treatment significantly increased the protein-to-cell ratio, while oe-LINC00472 significantly attenuated this effect; miR-21-5p partially reversed the anti-hypertrophic effect of oe-LINC00472 (p < 0.01, Figure 4G). Effects of miR-21-5p mimics on HG-induced functional changes in HGMC cells. A. Effects of different transfection treatments on miR-21-5p expression in HGMC cells. Changes in HGMC cell proliferation (B), migration (C), and invasion (D) following miR-21-5p mimic transfection. E. Effects of miR-21-5p overexpression on fibrotic gene expression in HGMC cells. F. Expression of inflammatory cytokines in HGMC cells after miR-21-5p mimic transfection. G. Protein/cell ratio showing that miR-21-5p mimic reversed the anti-hypertrophic effect of oe-LINC00472. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. HG, high glucose; oe-LINC00472, overexpression vector for LINC00472; HGMC, human glomerular mesangial cell; miR-21-5p mimic, synthetic RNA mimic of miR-21-5p; NC, negative control; FN, fibronectin; Col-IV, collagen type IV; TGF-β1, transforming growth factor-β1; TNF-α, tumor necrosis factor-α; IL-6, interleukin-6; IL-1β, interleukin-1β. Prediction and analysis of miR-21-5p downstream target genes involved in DKD regulation Multi-database enrichment results revealed 55 common genes that are miR-21-5p target genes and participate in DKD regulation (Figure 5A). PPI analysis revealed an overlapping gene network comprising 55 nodes and 12 edges (p = 0.0007, Figure 5B). Key genes with high interaction frequency included STAT3, PIK3R1, PIKFYVE, and RASA1. Preliminary detection indicated only STAT3 was significantly downregulated in serum from DKD patients (p < 0.001, Figure 5C). GO analysis (Figure 5D) revealed that overlapping genes are involved in ubiquitin ligase complexes and intercellular junctions, exhibiting ubiquitin-protein transferase activity, DNA-binding transcription activator activity, and RNA-binding activity. They may participate in proteasome-mediated ubiquitination, mRNA polyadenylation shortening, and negative regulation of Ras signaling. KEGG analysis (Figure 5E) indicated that overlapping genes primarily participated in signaling pathways including axon guidance, Foxo, ubiquitin-mediated proteolysis, and cell cycle. Previous studies identified STAT3 as containing a binding site for miR-21-5p (Figure 6A). Validation confirmed direct binding between miR-21-5p and STAT3 in HGMC cells (p < 0.0001, Figure 6B, C). Furthermore, we observed significant downregulation of STAT3 mRNA in HG-induced cells (p < 0.0001, Figure 6D). Transfection with oe-LINC00472 increased STAT3 mRNA and protein levels in HGMC cells, while transfection with miR mimic suppressed this gene expression (Figure 6D-E). Prediction and analysis of miR-21-5p downstream target genes involved in DKD regulation. A. Integration of miR-21-5p target genes and DKD-regulated genes from five different databases. B. Construction of the PPI network for overlapping genes. C. Expression profiles of STAT3, PIK3R1, PIKFYVE, and RASA1 in healthy individuals and DKD patients. GO (D) and KEGG (E) enrichment analysis of overlapping genes. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. DKD, diabetic kidney disease; CTD, Comparative Toxicogenomics Database; PPI, protein-protein interaction; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes. Validation of the STAT3-miR-21-5p targeting relationship and effects of HG and transfection on STAT3 expression. A. Target sequence for miR-21-5p binding to STAT3. B. Dual luciferase assay and C. RNA immunoprecipitation validate the STAT3-miR-21-5p targeting relationship. D-E. Effects of HG and different transfection treatments on STAT3 mRNA and protein levels. HG, high glucose; 3′UTR, 3′-untranslated region; WT, wild-type; NC, negative control; MUT, mutant; RIP, RNA immunoprecipitation; HGMC, human glomerular mesangial cell; oe-LINC00472, overexpression vector for LINC00472; miR mimic, miR-21-5p mimic. STAT3 is the functional downstream effector of the LINC00472/miR-21-5p axis To directly establish the functional causal relationship of the LINC00472/miR-21-5p/STAT3 axis, we knocked down STAT3 in HGMC cells overexpressing LINC00472 and performed rescue experiments. As shown in Supplementary Figure S3A, si-STAT3 transfection effectively reduced STAT3 expression in HGMC cells, confirming the success of the knockdown (p < 0.001). Functionally, co-transfection with si-STAT3 significantly reversed the inhibitory effects of LINC00472 overexpression on HGMC proliferation, migration and invasion (p < 0.001, Figure S3B-D). Furthermore, si-STAT3 also abolished the inhibitory effects of LINC00472 overexpression on fibrosis-related genes (p < 0.001, Figure S3E) and on inflammatory cytokines (p < 0.01, Figure S3F). Additionally, as shown in Supplementary Figure S3F, si-STAT3 partially reversed the anti-hypertrophic effects of oe-LINC00472 (p < 0.01, Figure S3G). Discussion DKD remains an unresolved challenge and a particularly difficult issue within healthcare systems [24]. Due to the diverse causes and complex molecular mechanisms underlying DKD, current treatment approaches remain unsatisfactory. Additionally, research on extracellular vesicle therapy and stem cell therapy has gradually increased over the past decade [25]. However, the regulatory role of lncRNAs in diabetes and DKD, along with their diagnostic value, has also been progressively recognized. Therefore, continuously elucidating the molecular mechanisms regulating DKD may significantly aid future prevention and control efforts. This study found that LINC00472 was downregulated in DKD patients and HG-treated HGMC cells, consistent with preliminary research reports [18], suggesting LINC00472 may serve as a potential diagnostic biomarker for DKD. Analysis revealed that serum LINC00472 levels exhibit high diagnostic sensitivity and specificity for DKD. This provides a novel molecular marker for early clinical identification of DKD, facilitating timely intervention and treatment. Following oe-LINC00472 transfection, HGMC cells demonstrated significantly reduced proliferation, migration, and invasion capabilities, alongside decreased expression of fibrosis genes and inflammatory factors. These findings indicate that LINC00472 plays a crucial role in protecting HGMC cells from high-glucose injury. By inhibiting cell proliferation and migration, LINC00472 may slow the pathological progression of DKD and reduce the risk of renal fibrosis. Furthermore, its inhibitory effect on inflammatory factors suggests a potential role in counteracting the inflammatory response associated with DKD. miR-21-5p is a known-up regulator involved in the regulation of T2DM and DKD [26]. Furthermore, multiple studies have detected high levels of miR-21-5p in vesicles stimulated by hyperglycemia [27]. Furthermore, studies indicate that elevated miR-21-5p enhances renal epithelial cell proliferation and epithelial-mesenchymal transition (EMT)-mediated metastasis [28]. However, our findings suggest that this effect of miR-21-5p on HGMC cell function is primarily regulated by LINC00472. This is evidenced by the fact that transfection with oe-LINC00472 suppressed miR-21-5p expression in HGMC cells. Moreover, the protective effect of LINC00472 overexpression on HGMC cells was completely abolished by a miR-21-5p mimetic. Therefore, we preliminarily hypothesize that the LINC00472/miR-21-5p axis participates in DKD progression. Furthermore, bioinformatics analysis revealed STAT3 appears to mediate the LINC00472/miR-21-5p axis’s regulation of DKD. This is because we detected abnormal STAT3 expression in both DKD patient serum and HG-induced HGMC cells. Our KEGG and GO enrichment analyses indicate that the genes targeted by miR-21-5p are significantly enriched in pathways such as the FoxO signaling pathway, cell cycle regulation, ubiquitin-mediated proteolysis, and axon guidance. These pathways play well-defined, context-specific roles in the pathogenesis of diabetic nephropathy. The FoxO transcription factor family (particularly FoxO1 and FoxO3a) serves as a key regulator of oxidative stress, autophagy, and apoptosis in renal cells. Under normal glucose conditions, FoxO activity is suppressed by insulin/Akt signaling. However, in a diabetic environment, chronic hyperglycemia induces FoxO activation by reducing Akt-mediated phosphorylation, thereby leading to increased expression of pro-oxidative genes (such as catalase and superoxide dismutase 2) [29,30]. Regarding cell cycle regulation, mesangial cell hypertrophy and proliferation are early hallmarks of DKD. Hyperglycemia induces cell cycle arrest and hypertrophy in mesangial cells [31,32]. Our findings indicate that overexpression of LINC00472 inhibits hyperglycemia-induced increases in the protein-to-cell ratio—a surrogate marker of hypertrophy—and reduces mesangial cell proliferation, suggesting that the LINC00472/miR-21-5p axis may suppress pathological cell cycle entry. STAT3 has been shown to regulate the expression of cell cycle regulatory factors, such as cyclin D1 and p21. By restoring STAT3 expression, LINC00472 may help maintain normal cell cycle progression and prevent the transition to hypertrophy and excessive proliferation. Multiple studies indicate that STAT3 represents a promising therapeutic target for DKD [33]. Research by Wang et al. [34] suggests that miR-20a targets CXCL6 to mediate JAK/STAT3 pathway activation, thereby promoting renal tubular injury. Tang et al. [35] reported that miR-223-3p targets IL6ST/STAT3 to promote DKD progression. Ebrahim et al. [36] revealed the involvement of the PKC/NF-κB/STAT3 signaling pathway in regulating renal injury in diabetic patients. Additional studies indicate lysozyme promotes diabetic renal fibrosis via the JAK/STAT3 signaling pathway [37]. However, the role of STAT3 in DKD is complex and cell-type specific. While STAT3 is generally considered pro-fibrotic in tubular epithelial cells, its function in mesangial cells may differ. In our study, we observed downregulation of STAT3 in DKD patient serum and in HG-treated HGMCs, which may represent a feedback inhibitory mechanism or reflect an early adaptive response. The protective effect of LINC00472 was associated with restoration of STAT3 expression, suggesting that an optimal level of STAT3 activity is required for maintaining mesangial cell homeostasis. To directly establish a functional causal relationship, we performed rescue experiments by knocking down STAT3 in HGMCs overexpressing LINC00472. The results showed that STAT3 knockdown significantly reversed the protective effects of oe-LINC00472 on cell proliferation, migration, fibrosis, inflammation, and hypertrophy, thereby confirming that STAT3 is a necessary downstream effector of this axis. Although this study reveals a novel mechanism by which LINC00472 participates in DKD, several limitations exist. First, the homogeneity of the sample may affect the generalizability of the results. Second, the study primarily relies on in vitro cell experiments, necessitating further validation through in vivo animal studies. Third, and importantly, our experiments were confined to HGMCs. While mesangial cells play a central role in glomerular matrix accumulation and hypertrophy in DKD, the diabetic kidney injury also involves podocytes, glomerular endothelial cells, and tubular epithelial cells. The expression pattern and functional significance of LINC00472 in these other renal cell types remain entirely unexplored. Future studies will employ multicenter experiments combined with multi-omics technologies to comprehensively elucidate the pathogenesis of DKD, providing scientific basis for developing more effective therapeutic approaches. In summary, this study identified LINC00472 as a protective factor against DKD with diagnostic value. Specifically, LINC00472 overexpression suppressed proliferation, migration, and invasion in HGMC cells, reduced expression of fibrosis-related genes, and decreased inflammatory levels. This effect was mediated by miR-21-5p. However, the molecular mechanisms underlying the involvement of the LINC00472/miR-21-5p/STAT3 axis in DKD progression warrant further investigation. Supplementary Material Figure S1 p STAT3.tif Figure S3.tif Figure S2 GAPDH.tif Supplementary Materials.docx Supplementary Table 1 .docx Figure S2 Col 4.tif Figure S1 GAPDH.tif Figure S2 TGF β1.tif Figure S2 FN.tif Figure S1 STAT3.tif Acknowledgments Not applicable. Funding Statement The author(s) reported there is no funding associated with the work featured in this article. Ethics approval and consent to participate This study was approved by the Ethics Committee of Traditional Chinese and Western Medicine Hospital of Wuhan, Tongji Medical College, Huazhong University of Science and Technology. Patients or their legal guardians provided written informed consent. The procedures used in this study adhere to the tenets of the Declaration of Helsinki. Consent for publication Not applicable. Disclosure statement No potential conflict of interest was reported by the author(s). Availability of data and materials The datasets used and analyzed during the current study are available from the corresponding author on reasonable request. References Associated Data Supplementary Materials Figure S1 p STAT3.tif Figure S3.tif Figure S2 GAPDH.tif Supplementary Materials.docx Supplementary Table 1 .docx Figure S2 Col 4.tif Figure S1 GAPDH.tif Figure S2 TGF β1.tif Figure S2 FN.tif Figure S1 STAT3.tif Data Availability Statement The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.