Kidney specimens were collected from patients with DN or minimal change disease (Normal) who were diagnosed at The Second Xiangya Hospital of Central South University and used for subsequent RNA extraction and immunohistochemistry (IHC) staining. Patients aged ≥18 years with nephropathy due to diabetes type 1 or type 2 were included in this study and provided written informed consent. These patients did not receive any treatment previously. Patients with other diseases, such as other rental diseases, autoimmune diseases, cancers and cardiovascular diseases were excluded. Our study got approval from the Ethics Committee of The Second Xiangya Hospital of Central South University.

Human renal tubular epithelial HK-2 cells and HEK-293T cells were provided by the Cell Bank of Chinese Academy Sciences (Shanghai, China). HK-2 cells were treated with normal glucose (NG, 5.5 mM), high glucose (HG, 30.5 mM) or the osmotic control high mannitol (HM, 5.5 mM glucose and 25 mM mannitol) for 24, 48 or 72 h (22), which were used for subsequent analysis of HG-induced cell injury. Glucose (G7021) and mannitol (M4125) were obtained from Sigma (St. Louis, MO, USA). For 5-aminoimidazole-4-carboxamide ribonucleotide (AICAR) or propranolol treatment, cells were treated with AICAR (S1802, Selleck, Houston, TX, USA) at 1 mM or propranolol (S4076, Selleck) at 100 µM.

MALAT1 and coding sequences of LIN28A and Nox4 were cloned into the pcDNA3.1 vector (V79020, ThermoFisher, Waltham, MA, USA) for overexpression of MALAT1 (oe-MALAT1), LIN28A (oe-LIN28A) and Nox4 (oe-Nox4). siRNA (10 nM) was selected for transient transfection thanks to its easy generation and introduction into cells with high efficiency. Specially, si-RNAs against MALAT1 (si-MALAT1), LIN28A (si-LIN28A) and Nox4 (si-Nox4) and scrambled negative control siRNAs (si-NC) were purchased from RiboBio (Guangzhou, China). HK-2 cells were transfected with vector, oe-MALAT1, si-MALAT1, oe-LIN28A, si-LIN28A, oe-Nox4, si-Nox4, si-LIN28A+vector, si-LIN28A+oe-Nox4, si-MALAT1+vector, si-MALAT1+oe-LIN28A or si-NC using the Lipo3000 reagent (L3000015, ThermoFisher) following the manual. Cells were harvested at 48 hours after transfection for subsequent assays. In some assays, cells were harvested and treated with NG or HG. For in vivo knockdown of MALAT1, viral vector-based shRNA with high transduction efficiency and long-term effect was used. Specially, the shRNA against MALAT1 (sh-MALAT1, Sigma) and scrambled shRNA (sh-NC, Sigma) were inserted into the pLKO.1 lentiviral vector (10878, Addgene, Watertown, MA, USA), sh-MALAT1 and sh-NC lentiviral particles were packaged in HEK-293T cells.

Sprague-Dawley male healthy rats (6-8-week-old) were provided by Hunan SJA laboratory animal Co., LTD (Changsha, Hunan, China) and blindly divided in to 4 groups: control, DN, DN+sh-NC and DN+sh-MALAT1. In the control group, rats were fed normally. In DN groups, rats were fed a high-glucose and fat diet for 8 weeks and intraperitoneally injected with streptozotocin (STZ, S0130, Sigma) at 55 mg/kg. Subsequently, the blood glucose was monitored every 3 days. DN was considered to be successfully established in rats which met following criteria: blood glucose > 16.7 mmol/L and 24 h urine volume and protein increased 150% after two weeks. DN rats were intravenously injected with 2×10⁷ sh-NC or sh-MALAT1 lentiviral particles and fed a high-glucose and fat diet for 5 months. Finally, kidneys were harvested for subsequent assays. Animal procedures were approved by The Animal Care and Use Committee of The Second Xiangya Hospital of Central South University.

Tissues from patients and rats were homogenized. Total RNA was extracted from HK-2 cells and tissue homogenates using Trizol from Beyotime (R0016, Shanghai, China). Subsequently, RNA was quantified and reversely transcribed into cDNA. The expression of MALAT1, LIN28A and Nox4 was determined using quantitative PCR with SYBR Green (QPK-201, TOYOBO, Tokyo, Japan) and normalized to GAPDH. Gene expression was calculated using the 2^−ΔΔCt method. Primers were listed in Table 1 .

Actinomycin D, a widely used RNA transcription inhibitor for shutting off mRNA transcription, was used to investigate the decay rates of endogenous mRNAs. Briefly, HK-2 and HEK-293T cells were treated with actinomycin D (S8964, Selleck) at 5 µg/mL for indicated time. Subsequently, RNA was extracted and reversely transcribed into cDNA. The abundance of MALAT1 and Nox4 mRNA were analyzed by qRT-PCR.

The culture medium was removed. 100 µL of fresh culture medium and 10 µL of MTT (M6494, ThermoFisher) were mixed well and added into each well. Cells were then incubated for 4 hours, and 50 µL of DMSO was added, and the absorbance (490 nm) was measured.

Kidney specimens from patients and rats were fixed, embedded and sliced into 5-µm sections. Slices were subsequently deparaffinized and rehydrated. H&E and Masson’s trichrome staining were applied to analyze renal histological change and collagen deposition. For IHC staining, antigen was retrieved in antigen retrieval solution, and slices were incubated with anti-LIN28A (1:50, ab175352, Abcam, Cambridge, UK) or anti-Nox4 (1:100, ab133303, Abcam). Subsequently, slices were incubated with an HRP-conjugated secondary antibody for 1 h. DAB (P0203, Beyotime) was used to visualize the signal. Slices were then stained with hematoxylin, mounted and imaged with a BX51 microscope from Olympus (Tokyo, Japan).

The Click-iT TUNEL Assay kit (C10337) was obtained from ThermoFisher and used for analyzing cell apoptosis. HK-2 cells were fixed and permeabilized in 0.2% Triton X-100. Kidney tissues from rats were embedded in paraffin and sliced into 5-µm sections. Slices were subsequently deparaffinized, fixed and permeabilized in 0.2% Triton X-100. TdT rection was performed in cells and slices, and the Click-iT reaction was performed for detecting apoptotic cells. Cells and slices were stained with DAPI (C1005, Beyotime), mounted and imaged under a Leica confocal microscope (Wetzlar, Germany).

The culture supernatants of HK-2 cells and rat serum were harvested and stored at -80˚C until use. The concentrations of tumor necrosis factor-α (TNF-α), interleukin-6 (IL-6) and albumin were determined by ELISA kits. Human and rat TNF-α (ab181421 and ab108913), IL-6 (ab178013 and ab100772) and albumin (ab108788 and ab108789) ELISA kits were obtained from Abcam.

HK-2 cells were stained with using the fluorescent probes dihydroethidium (DHE; Invitrogen, Carlsbad, CA, MA) at 5 µM for 30 min at 37°C in the dark. DHE can be oxidized and then intercalates within DNA, staining the nucleus a bright fluorescent red. Subsequently, cells were washed with PBS for 10 min. Then, Intracellular ROS levels were measured using a fluorometer with excitation at 535 nm and emission at 610 nm (Varioskan™ LUX Multimode Microplate Reader; Thermo Fisher Scientific Inc.). Images were acquired at room temperature and DHE fluorescence intensity was measured using the ImageJ software (NIH, Bethesda, MD).

A drop of blood was collected from rat tails and the level of blood glucose was examined using the AlphaTRAK glucose meter (Zoetis, Parsippany, NJ, USA) following the manufacturer’s recommendation. Serum was prepared and urine was collected. The concentrations of blood urea nitrogen (BUN), serum and urinary creatinine were measured using the BUN colorimetric detection kit (EIABUN, ThermoFisher) and creatinine assay kit (ab65340, Abcam), respectively.

HK-2 and HEK-293T cells were harvested and resuspend in nuclear isolation buffer (20 mM MgCl2, 40 mM Tris-HCl, 1.28 M sucrose and 4% Triton X-100). The nuclei were pelleted by centrifugation and resuspended in RIP buffer (5 mM EDTA, 0.5 mM DTT, 25 mM Tris, 150 mM KCl, 0.5% NP40, RNase and protease inhibitors). Chromatin was sheared, and the supernatants were collected by centrifugation. The anti-LIN28A (ab175352) and normal IgG control (ab37415, Abcam) were pre-coated on magnetic beads, mixed with the supernatants and incubated overnight. Subsequently, immunoprecipitated RNA was recovered using TRIzol, and the abundance of MALAT1 and Nox4 mRNA were analyzed by qRT-PCR.

HK-2 and HEK-293T cells were lysed, and cell lysates were harvested. Biotinylated MALAT1 or Nox4 RNAs (RiboBio) were added into cell lysates and incubated for 5 h at 4 °C for forming RNA-Protein complexes. Subsequently, streptavidin-conjugated magnetic beads (88816, ThermoFisher) were added, and samples were incubated for 2 h with gentle rotation at 4 °C. The RNA-Protein complexes were pulled down by magnetic beads and eluted for western blotting analysis of LIN28A.

Rats were sacrificed, and kidneys were excised and weighed. Parts of kidneys were homogenized in lysis buffer. The supernatants of tissue homogenates and cell lysates were collected and quantified using the BCA kit (ab102536, Abcam). Protein (20 µg) was loaded each lane and electrophoresed prior to transfer to PVDF membranes (88518, ThermoFisher). Membranes were then blocked and incubated with anti-LIN28A (1:500, ab279647), anti-Nox4 (1:1000, ab133303), anti-AMPK (1:2000, ab80039), anti-p-AMPK (1:1000, 2531), anti-mTOR (1:1000, ab2732), anti-p-mTOR (1:500, 2971), anti-Bcl-2 (1:1000, ab196495), anti-Bax (1:500, ab32503) and anti-β-actin (1:6000, ab8227) overnight. Next day, membranes were incubated with an HRP-conjugated secondary antibody for 1 h. The bands were visualized using ECL substrate (1705061) from Bio-Rad (Hercules, CA, USA) and analyzed using the Image J software. Antibodies were from Abcam and Cell Signaling Technology (Danvers, MA, USA).

Data from at least three independent assays were expressed as mean ± standard deviation. The variance in two and multiple groups were analyzed by the Student’s t test and one-way analysis of variance (ANOVA), respectively. P<0.05 was statistically significant. *P<0.05, **P<0.01 and ***P<0.001.

To explore the roles of MALAT1 in DN, we analyzed its expression and found that DN tissues showed increased MALAT1 expression ( Figure 1A ). Subsequently, we examined MALAT1 expression in HK-2 cells treated with normal glucose (NG), high mannitol (HM) or high glucose (HG). Compared to NG and HM, HG significantly enhanced MALAT1 expression ( Figure 1B ). Moreover, MALAT1 was overexpressed by transfection of oe-MALAT1 and silenced by si-MALAT1 ( Figure 1C ). Our results showed that HG further enhanced MALAT1 expression in cells transfected with oe-MALAT1 but failed to induce MALAT1 expression in cells transfected with si-MALAT1 ( Figure 1D ). Compared to NG, HG reduced HK-2 cell viability ( Figure 1E ). Overexpression of MALAT1 accelerated the reduction of cell viability, whereas silencing of MALAT1 maintained cell viability in response to HG treatment ( Figure 1E ). Furthermore, we found that HG induced cell apoptosis, ROS generation and the secretion of TNF-α and IL-6, which were further enhanced by overexpression of MALAT1 but suppressed by knockdown of MALAT1 ( Figures 1F–H ). Our findings demonstrated that MALAT1 enhanced HK-2 cell apoptosis, ROS generation and inflammatory cytokine secretion in response to HG treatment.

As the expression of LIN28A and MALAT1 have been reported to be positively correlated (17), we examined LIN28A expression in DN tissues. Elevated LIN28A expression was validated by qRT-PCR, western blotting and IHC staining ( Figures 2A, B ). Moreover, HG treatment upregulated LIN28A in HK-2 cells ( Figure 2C ). LIN28A was overexpressed by transfection of oe-LIN28A and knocked down by transfection of si-LIN28A ( Figure 2D ). HG-induced LIN28A expression was further promoted by transfection of oe-LIN28A in HK-2 cells but inhibited by transfection of si-LIN28A ( Figure 2E ). Overexpression of LIN28A obviously impaired HK-2 cell viability and promoted cell apoptosis in response to HG treatment, whereas knockdown of LIN28A significantly weakened these effects ( Figures 2F, G ). Besides, HG-induced ROS generation was strengthened by overexpression of LIN28A but reduced by silencing of LIN28A in HK-2 cells ( Figure 2H ). Also, overexpression of LIN28A enhanced HG-induced secretion of TNF-α and IL-6 in HK-2 cells, but knockdown of LIN28A significantly reduced their secretion ( Figure 2I ). Collectively, these results demonstrated that LIN28A promoted HG-induced injury in HK-2 cells.

HK-2 cells were transfected with si-LIN28A or si-LIN28A in combination with oe-LIN28A. We found that knockdown of LIN28A reduced the abundance of MALAT1 in HK-2 cells, and simultaneous overexpression of LIN28A restored the expression of MALAT1, suggesting that LIN28A maintained MALAT1 expression in HK-2 cells ( Figure 3A ). RIP assays showed that MALAT1 was efficiently enriched in the anti-LIN28A-immunoprecipitated fractions from HK-2 and HEK-293T cells ( Figure 3B ). Besides, LIN28A was pulled down by the biotinylated MALAT1 probe ( Figure 3C ). These data indicated that MALAT1 directly interacted with LIN28A in HK-2 and HEK-293T cells. As LIN28 functions as an RNA binding protein (RBP) to stabilize target RNAs (23), we hypothesized that LIN28A might regulate MALAT1. To validate the hypothesis, HK-2 cells transfected with si-LIN28A or si-LIN28A in combination with oe-LIN28A were treated with actinomycin D. Silencing of LIN28A impaired MALAT1 stability with a half-life of ~5 hours in HK-2 cells, and simultaneous overexpression of LIN28A restored MALAT1 stability ( Figure 3D ), implying that LIN28A interacted with and stabilized MALAT1.

HK-2 cells were transfected with si-MALAT1 or si-MALAT1 in combination with oe-LIN28A and subsequently treated with HG. The expression of LIN28A were suppressed by si-MALAT1 transfection but enhanced by simultaneous overexpression of LIN28A ( Figures 4A, B ). Knockdown of MALAT1 improved the viability of HG-treated HK-2 cells, but it was abrogated by LIN28A overexpression ( Figure 4C ). Furthermore, knockdown of MALAT1-mediated alleviation of cell apoptosis, ROS generation and secretion of TNF-α and IL-6 were reversed by overexpression of LIN28A ( Figures 4D–F ). Our findings suggested that MALAT1-mediated aggravation of HG-induced injury was dependent on LIN28A in HK-2 cells.

As ROS-generating NADPH oxidase Nox4 serves key roles in DN, we examined whether LIN28A interacted with Nox4 in HK-2 cells. RNA pull-down assays showed that LIN28A was pulled down by the Nox4-sense probe, but not by the Nox4-antisense probe ( Figure 5A ). In addition, we found that Nox4 was enriched in the anti-LIN28A-immunoprecipitated fractions from HK-2 cells ( Figure 5B ). HK-2 cells transfected with si-LIN28A showed reduced expression of LIN28A and Nox4 ( Figure 5C ). Furthermore, knockdown of LIN28A reduced the stability of Nox4 mRNA in HK-2 cells in response to actinomycin D ( Figure 5D ). These observations suggested that LIN28A targeted Nox4 to maintain its stability in HK-2 cells. Subsequently, we investigated whether MALAT1 was implicated in the interaction between LIN28A and Nox4. As same as GAPDH, MALAT1 was primarily localized in the cytoplasm of HK-2 cells ( Figure 5E ). Besides, Nox4 expression was inhibited by knockdown of MALAT1 and increased by overexpression of MALAT1 ( Figures 5F, G ). RIP assays showed that Nox4 mRNA was enriched by anti-LIN28A in HEK-293T and HK-2 cells, and overexpression of MALAT1 significantly increased the enrichment ( Figure 5H ). Knockdown of MALAT1 reduced Nox4 mRNA stability, whereas overexpression of MALAT1 increased its stability in HEK-293T and HK-2 cells in response to actinomycin D ( Figure 5I ). HG-induced expression of Nox4 was suppressed by knockdown of MALAT1, but it was partially reversed by simultaneous overexpression of LIN28A in HK-2 cells ( Figures 5J, K ). To conclude, MALAT1 promoted the interaction of LIN28A and Nox4 to maintain Nox4 mRNA stability in HK-2 cells.

Nox4 was upregulated, and more Nox4 positive cells were observed in DN tissues ( Figures 6A, B ). Besides, Nox4 was upregulated in HK-2 cells treated with HG ( Figure 6C ). Moreover, Nox4 was upregulated in cells transfected with oe-Nox4 and downregulated in cells transfected with si-Nox4 ( Figure 6D ). HG-induced Nox4 expression was significantly increased in HK-2 cells transfected with oe-Nox4 but inhibited in cells transfected with si-Nox4 ( Figure 6E ). Importantly, overexpression of Nox4 further reduced HK-2 cell viability in response to HG, whereas knockdown of Nox4 maintained cell viability ( Figure 6F ). Additionally, HG-induced cell apoptosis, ROS generation and secretion of TNF-α and IL-6 were considerably increased by overexpression of Nox4 but reduced by knockdown of Nox4 ( Figures 6G–I ). These data demonstrated that Nox4 enhanced HG-induced injury in HK-2 cells.

HK-2 cells were transfected with si-LIN28A or si-LIN28A in combination with oe-Nox4 and treated with HG. HG treatment increased the expression of LIN28A and Nox4, which was suppressed by transfection of si-LIN28A, and Nox4 expression was increased by transfection of oe-Nox4 ( Figures 7A, B ). Knockdown of LIN28A improved HG-treated HK-2 cell viability, but simultaneous overexpression of Nox4 abolished this effect ( Figure 7C ). Intriguingly, silencing of LIN28A mitigated cell apoptosis, ROS generation and secretion of TNF-α and IL-6 in HG-treated HK-2 cells, which were largely abrogated by Nox4 overexpression ( Figures 7D–F ). These data demonstrated that LIN28A targeted Nox4 to aggravate HG-induced injury in HK-2 cells.

As Nox4/AMPK/mTOR signaling has been reported to be implicated in diabetic kidney disease (21), we hypothesized that the AMPK/mTOR signaling might be implicated in Nox4-mediated regulation of HG-induced injury. HK-2 cells were transfected with si-Nox4 and treated with HG in combination with the AMPK activator AICAR or the mTOR activator propranolol. HG treatment increased the expression of Nox4 and Bax and phosphorylation of AMPK (p-AMPK) and mTOR (p-mTOR) and reduced Bcl-2 expression in HK-2 cells, which were largely abolished by knockdown of Nox4 ( Figure 8A ). AICAR and propranolol reversed Nox4 knockdown-mediated regulation of p-AMPK and p-mTOR and expression of Bax and Bcl-2 in HG-treated HK-2 cells ( Figure 8A ). Moreover, knockdown of Nox4 increased the viability of HG-treated HK-2 cells, but it was reversed by AICAR or propranolol treatment ( Figure 8B ). Besides, knockdown of Nox4-mediated alleviative effects on HG-induced cell apoptosis, ROS generation and secretion of TNF-α and IL-6 were abolished by AICAR or propranolol ( Figures 8C–E ). Our findings indicated that knockdown of Nox4 attenuated HG-induced injury through suppression of the AMPK/mTOR signaling in HK-2 cells.

A rat model of DN was established. Subsequently, rats with DN were intravenously injected with shMALAT1 or sh-NC lentiviral particles and fed a high glucose and fat diet for 5 months. Rats with DN showed ascending body weight in the first 8 weeks and declining body weight after 8 weeks, but the degree of weight loss was obviously slowed down in rats injected with shMALAT1 ( Figure 9A ). Similarly, elevated kidney weight (KW)/body weight (BW), fasting blood glucose, albumin-to-creatinine ratio (ACR), blood urea nitrogen (BUN) and creatinine were observed in rats with DN, whereas all of them were significantly reduced by shMALAT1 injection ( Figures 9B–E ). Furthermore, we observed severe renal tubule damage and increased collagen deposition and cell apoptosis in renal tubules, and injection of shMALAT1 markedly attenuated these manifestations ( Figures 9F–H ). The expression of MALAT1 was dramatically promoted in DN rats and greatly reduced by injection of shMALAT1 ( Figure 9I ). These observations demonstrated that knockdown of MALAT1 alleviated renal tubular epithelial injury in DN rats. Elevated levels of LIN28A, Nox4, p-AMPK and p-mTOR and Bax and reduced Bcl-2 expression in DN rats were largely reversed by knockdown of MALAT1 ( Figure 9J ). Increased secretion of TNF-α and IL-6 was also significantly inhibited by knockdown of MALAT1 in DN rats ( Figure 9K ). Collectively, these results implied that knockdown of MALAT1 alleviated renal tubular epithelial injury through suppression of LIN28A and the Nox4/AMPK/TOR signaling axis in DN.