Male wild-type (WT) mice (n=60, weight, 250±10 g) and AMPK knockout (AMPK^−/−) mice (n=60, weight, 250±10 g) were obtained from Jackson Laboratory (Bar Harbor, ME, USA). All mice were housed under standard laboratory conditions (27°C, 40–60% humidity, 12-h light/dark cycle) with free access to water and on a standard laboratory diet. Eight-week-old WT and AMPK^−/− mice were injected with streptozotocin (STZ; 50 mg/kg/day) for 5 days (20). Subsequently, melatonin (20 mg/kg/day; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) was administered intraperitoneally into WT and AMPK^−/− mice over the course of 12 weeks. The animal experiments were performed for a 12-week period (n=6/group). WT mice injected with PBS were termed the control group; WT mice injected with STZ were termed the STZ+WT group; WT mice treated with STZ and melatonin were termed the Mel+STZ+WT group; and, AMPK^−/− mice treated with STZ and melatonin were termed the Mel+STZ+AMPK^−/− group. At the end of the experiment, blood pressure was measured in conscious, acclimated mice using the tail-cuff method. Kidney tissues were harvested following the sacrifice of the mice. A part of the kidney was snap-frozen in liquid nitrogen and the other part was fixed with 10% PBS-buffered formalin, processed and embedded in paraffin. All experimental protocols were approved by the Ethics Committee of Chinese PLA General Hospital (Beijing, China).

The rat glomerular mesangial cell line, HBZY-1, was purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). Cells were cultured according to the supplier's protocols in RPMI medium (Thermo Fisher Scientific, Inc., Waltham, MA, USA) supplemented with 10% fetal bovine serum (HyClone; GE Healthcare Life Sciences, Logan, UT, USA) at 37°C in 5% CO₂. To mimic high glucose damage, cells were cultured in high glucose medium (25 mmol/l), as opposed to normal glucose medium (5.5 mmol/l), for ~12 h according to a previous study (26). The cells were incubated with 5 µM melatonin for ~12 h before high-glucose stress (18).

The kidneys were harvested, fixed for 1 h in 4% (w/v) paraformaldehyde at room temperature, and rapidly frozen in Optimal Cutting Temperature Compound (Agar Scientific, Ltd., Stansted, UK) for the preparation of cryosections (size, 4 µm) (27). Hematoxylin and eosin (H&E), Masson's trichrome (cat. no. KGMST-8004; Nanjing KeyGen Biotech Co., Ltd., Nanjing, China) and periodic acid Schiff-methenamine (PASM; cat. no. 395B; Sigma-Aldrich; Merck KGaA) staining were performed at room temperature, following which, samples were observed under an inverted microscope (magnification, ×40; BX51; Olympus Corporation, Tokyo, Japan).

Blood samples from tail vein were collected before the mice were sacrificed. Then, laminin (Mouse Laminin ELISA kit; cat. no. ab119572; Abcam, Cambridge, UK) and pre-collagen III (Mouse Collagen Type III ELISA kit; cat. no. MBS727234; MyBiosource, San Diego, CA, USA) in blood samples were measured via ELISA analysis, according to a previous study (28). Blood urea nitrogen (BUN; Mouse Blood Urea Nitrogen ELISA kit; cat. no. MBS751125; MyBiosource) and serum creatinine levels (Creatinine Assay kit; cat. no. ab65340; Abcam) were determined via ELISA based on the manufacturers' protocols.

After 12 h of fasting, the glucose levels in venous blood samples drawn from the tail vein were measured using a glucometer (Roche Diagnostics GmbH, Mannheim, Germany). Urine samples were collected prior to the sacrifice of the mice (29). To collect morning spot urine samples, animals were placed in metabolic cages at the beginning of the light cycle and were kept for 2 h with access to water, but not food. To obtain 24-h urine samples, animals were placed in metabolic cages at the beginning of the light cycle and were kept for 24 h with free access to water and on a standard laboratory diet (30). Urinary albumin content (Mouse Albumin ELISA kit; cat. no. ab108792; Abcam,) was determined according to the manufacturer's protocols.

The cells were first washed in cold PBS, permeabilized with 0.1% Triton-X-100 for 10 min at 4°C and blocked with 10% goat serum albumin (Invitrogen; Thermo Fisher Scientific, Inc.) for 1 h at room temperature. Subsequently, samples were incubated with cytochrome c(cyt-c) antibody (1:1,000; cat. no. ab133504; Abcam) and cleaved caspase-3 antibody (1:1,000; cat. no. 9664; Cell Signaling Technology, Inc. Danvers, MA, USA), overnight at 4°C (31). Following three washes in PBS, Alexa-Fluor 116 488 donkey anti-rabbit secondary antibody (1:1,000; cat. no. A-21206; Invitrogen; Thermo Fisher Scientific, Inc.) was added to the samples for 1 h at room temperature (32). Images were observed with an inverted fluorescence microscope (magnification, ×40; BX51; Olympus Corporation).

The proteins isolated from kidney tissues with the help of RIPA buffer (Thermo Fisher Scientific, Inc.) were from three mice in each group. Proteins were also isolated from cells using RIPA buffer (Thermo Fisher Scientific, Inc.). The protein concentration was analyzed using the bicinchoninic acid protein assay (Thermo Fisher Scientific, Inc.). In the present study, a total of 40 µg cell proteins and 60 µg tissue proteins were separated by 12–15% SDS-PAGE. Following electrophoresis, the proteins were transferred onto a polyvinylidene fluoride membrane (Roche Applied Science, Penzberg, Germany). The membranes were subsequently blocked with 5% non-fat milk for 1 h at room temperature prior to incubation with the primary antibodies. The primary antibodies used were: Pro-caspase-3 (1:1,000; cat. no. 9662; Cell Signaling Technology, Inc.), cleaved caspase-3 (1:1,000; cat. no. 9664; Cell Signaling Technology, Inc.), cellular inhibitor of apoptosis protein 1 (c-IAP; 1:1,000; cat. no. 4952; Cell Signaling Technology, Inc.), B-cell lymphoma 2 (Bcl2; 1:1,000, cat. no. 3498; Cell Signaling Technology, Inc.), caspase-9 (1:1,000; cat. no. ab32539; Abcam), transforming growth factor (TGF)β (1:1,000; cat. no. 3711; Cell Signaling Technology, Inc.), matrix metalloproteinase 9 (MMP9; 1:1,000; cat. no. 13667; Cell Signaling Technology, Inc.), cyclophilin D (1:1,000; cat. no. ab181983; Abcam), p-cyclophilin D antibody (1:1,000, bioss, cat. no. bs-9878R), collagen I (1:1,000; cat. no. ab34710; Abcam), collagen III (1:1,000; cat. no. ab7778; Abcam), Bcl-2-associated X protein (Bax; 1:1,000; cat. no. ab32503; Abcam), AMPK (1:1,000; cat. no. ab131512; Abcam), phosphorylated (p)-AMPK (1:1,000, cat. no. ab23875; Abcam) and peroxisome proliferator-activated receptor γ coactivator 1-α (PGC1α; 1:1,000; cat. no. 2178; Cell Signaling Technology, Inc.) The secondary antibodies used in the present study were: Horseradish peroxidase (HRP)-coupled secondary antibodies (1:2,000; cat. nos. 7074 and 7076; Cell Signaling Technology, Inc.). Band intensities were normalized to the respective β-actin (1:2,000; cat. no. ab8224; Abcam) and/or GAPDH (1:1,000, cat. no. 5174; Cell Signaling Technology, Inc.) internal control signal intensity with the help of Quantity One Software (version 4.6.2; Bio-Rad Laboratories, Inc., Hercules, CA, USA) (29). Bands were detected using an enhanced chemiluminescence substrate (Applygen Technologies, Inc., Beijing, China). The experiment was repeated three times (33).

TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.) was used to isolate total RNA from tissues (34). Subsequently, the Reverse Transcription kit (Kaneka Eurogentec S.A., Seraing, Belgium) was applied to transcribe RNA (1 µg/group) into cDNA according to the manufacturer's protocol. RT-qPCR was performed with primers and matched probes from the Universal Fluorescence-labeled Probe Library (Roche Diagnostics, Basel, Switzerland) using SYBR™ Green PCR Master Mix (Thermo Fisher Scientific, Inc.) (35). The primers used in the present study were: Bax forward, 5′-GTCCTATTCTGATGGATCCC-3′ and reverse, 5′-GCTACGGTACCATGGCCTACG-3′; Bad, 5′-CCTAGATTCAGTGACTGAG-3′ and reverse, 5′-ACCTAACCGATGGCGGTCGAGTGC-3′; Survivin, 5′-ATCCGTTGTCCAGTCTTAGTCTA-3′ and reverse, 5′-GGTCGGTAGATTCATTAATGAT-3′ and GAPDH forward, 5′-GCGGATGAACGTGGACGTGAC−3′ and reverse, 5′-AACGTGGTCCAGACCAATGCG-3′). The cycling conditions were: 95°C for 8 min, followed by 35 cycles of 95°C for 10 sec and 72°C for 12 sec, for telomere PCR. The mRNA ratio of the target genes to GAPDH was calculated using the 2^−ΔΔCq method (36).

Electron transport chain complex I activity (Electron transport chain Complex I Assay kit; cat.no. MBS2540528; MyBiosource), Electron transport chain complex II activity (Complex II Enzyme Activity Microplate Assay kit; cat. no. ab109908; Abcam,) and Electron transport chain complex V activities (MitoTox™ Complex V OXPHOS Activity Assay kit; cat. no. ab109907; Abcam) were determined according to the manufacturers' protocols.

To observe cellular ROS levels, the dihydroethidium ROS probe (Molecular Probes; Thermo Fisher Scientific, Inc.) was incubated with the cells (1×10⁵) for ~30 min at 37°C in the dark. Subsequently, the cells were washed with PBS to remove the ROS probe. The cells were immediately analyzed under a fluorescence microscope (37). Image-Pro Plus version 6.0 (Media Cybernetics, Rockville, MD, USA) was used to obtain the fluorescence densities, which were normalized to that of the control group.

mPTP opening is an early event in mitochondrial apoptosis. In the present study, mPTP opening was measured via tetramethylrhodamine ethyl ester (TMRE) fluorescence. Samples were washed three times with PBS and were then stained with TMRE dye (38). The baseline fluorescence of TMRE was recorded and, after 30 min, TMRE fluorescence was recorded again. Image-Pro Plus version 6.0 (Media Cybernetics) was used to obtain the fluorescence densities of TRME fluorescence. According to a previous study (39), the mPTP opening rate was determined as the time for the fluorescence intensity to decrease by half of the baseline fluorescence intensity.

Mitochondrial potential was assessed using a JC-1 probe, a sensitive fluorescent dye used to detect alterations in mitochondrial potential (34). Following treatment, cells (1×10⁵) were incubated with 10 mg/ml JC-1 for 10 min at 37°C in the dark and monitored with a fluorescence microscope (magnification, ×100; BX51; Olympus Corporation). Red fluorescence was attributable to potential-dependent dye aggregation in the mitochondria. Green fluorescence, reflecting the monomeric form of JC-1, appeared in the cytosol following mitochondrial membrane depolarization.

ATP production was detected to monitor mitochondrial function. Firstly, the samples were washed three times with cold PBS. The samples were then lysed using the RIPA Lysis Buffer (Beyotime, China, cat. No:P0013E) and the Luciferase-based ATP Assay kit (Beyotime Institute of Biotechnology, Haimen, China) was used according to a previous study (25). ATP production was measured using a microplate reader (40).

LDH is released into the cell culture supernatant when cell membranes rupture (36,41). To evaluate LDH levels (cells density 1×10⁵), an LDH Release Detection kit (Beyotime Institute of Biotechnology) was used according to the manufacturer's protocol.

To analyze changes in caspase-3/9, caspase-3/9 activity kits (Beyotime Institute of Biotechnology) were used, according to the manufacturer's protocols (42). To analyze caspase-3 activity, 5 µl 4 mM DEVD-pNA substrate (final concentration, 200 µM) was added to the cells (1×10⁵) for 2 h at 37°C. To measure caspase-9 activity, 5 µl 4 mM LEHD-p-NA substrate (final concentration, 200 µM) was added to the samples for 1 h at 37°C. Subsequently, caspase activity was quantified by spectrophotometric detection at 400 nm using a microplate reader (43).

MTT experiments were performed in 96-well plates to analyze cellular viability. Samples (1×10⁵) were washed three times with PBS and 50 µl MTT reagent was added to each well. The samples were incubated for 4 h at 37°C in a humidified atmosphere containing 5% CO₂. Subsequently, the MTT solution was removed, 200 µl dimethyl sulfoxide was added and the samples were incubated for 10 min at room temperature. Following the addition of Sorensen's buffer, the absorbance at 570 nm was measured (44).

To detect cell death, a TUNEL assay was performed using an In Situ Cell Death Detection kit (Roche Diagnostics GmbH), according to the manufacturer's protocol. DAPI was used to label the nuclei at room temperature for ~30 min and the cells were observed under an inverted fluorescence microscope (magnification, ×40; BX51; Olympus Corporation) (45).

To inhibit the expression of AMPK and PGC1α, small interfering (si)RNAs against AMPK and PGC1α were used to knockdown their expression. The siRNA against AMPK and PGC1α as well as the negative control siRNA were purchased from Yangzhou Ruibo Biotech Co., Ltd (Yangzhou, China) (46). The siRNA sequences were as follows: AMPK siRNA sense strand, 5′-GCTTACUGACTGACGT-3′ and antisense strand, 3′-AUGUUACCGUATTC-5′; PGC1α siRNA sense strand, 5′-GUAGGUACTACCTA-3′ and antisense strand, 3′-TUAUUTAGTTAACTGAT-5′; negative control siRNA sense strand, 5′-UUACCUTUCCATGATGCT′ and antisense strand, 3′-AUGTUGAAGUTCCGT-5′. To transfect siRNA into HBZY-1 cells, Opti-Minimal Essential Medium (Invitrogen; Thermo Fisher Scientific, Inc.) without serum or antibiotics was used to incubate cells for 24 h. Lipofectamine^® 2000 transfection reagent (Invitrogen; Thermo Fisher Scientific, Inc.) was added into the medium according to the manufacturer's protocol (47). Subsequently, siRNAs (70 nM/well of siRNA) were added into serum-free medium and incubated with cells for 72 h. Cells were collected and proteins were extracted (48).

All data are expressed as the means ± standard deviation and experiments were repeated ≥3 times in the present study. Statistical analyses were performed with SPSS software version 17.0 (SPSS, Inc., Chicago, IL, USA). Statistical analysis was performed using one-way analysis of variance followed by Tukey's post hoc test. P<0.05 was considered to indicate a statistically significant difference.

To verify the role of AMPK in diabetic renal fibrosis, western blotting was used to analyze AMPK expression. Compared with the control WT mice, STZ treatment significantly impaired AMPK activation; however, this effect was reversed by melatonin treatment (Fig. 1A and B). Furthermore, to confirm whether AMPK is necessary for melatonin-induced kidney protection under diabetic stress, AMPK^−/− mice were used. As shown in Fig. 1C, renal fibrosis was observed using Masson's trichrome staining. Compared with the control WT mice, STZ-treated mice presented with severe kidney fibrosis and this change was partially reversed by melatonin treatment. However, AMPK deletion inhibited the protective effects of melatonin on renal fibrosis. In addition, the expression levels of collagen I and collagen III were assessed in kidney samples. Chronic hyperglycemia significantly enhanced collagen I and collagen III expression (Fig. 1D-F). Notably, this effect was inhibited by melatonin, whereas absence of AMPK completely abrogated the inhibitory effects of melatonin on collagen accumulation.

Furthermore, signaling factors associated with fibrosis, including TGFβ and MMP9, were more abundant in the kidney tissues of STZ-treated mice compared with the control WT mice (Fig. 1D, G and H). Treatment with melatonin significantly reduced TGFβ and MMP9 expression levels; however, the effects of melatonin were abolished by AMPK deletion. Furthermore, the serum concentrations of laminin and pre-collagen III were notably higher in STZ-treated mice compared with in control WT mice (Fig 1I and J). As expected, the high serum concentrations of laminin and pre-collagen III were significantly reduced by melatonin supplementation, whereas the effects of melatonin were reversed by AMPK deletion. Taken together, these results indicated that melatonin regulated diabetic renal interstitial fibrosis by activating AMPK.

Glomerular apoptosis is the primary factor associated with the development of diabetic renal fibrosis; therefore, the role of melatonin and AMPK in glomerular apoptosis was assessed. Firstly, caspase-3 staining demonstrated that the renal tissue in STZ-treated mice exhibited increased cleaved caspase-3 expression compared with WT mice (Fig. 2A and B). However, melatonin administration reduced caspase-3 expression, whereas the effects of melatonin were suppressed in response to AMPK deficiency. Similar results were obtained for the caspase-3 activity assay (Fig. 2C). Additionally, to investigate the protective role of melatonin and AMPK in glomerular apoptosis, western blotting was performed. Melatonin reduced the expression levels of pro-apoptotic proteins (caspase-3, −9 and Bax), but increased the expression of an anti-apoptotic factor (c-IAP), in an AMPK-dependent manner (Fig. 2D-H). Alterations in the expression levels of apoptosis-associated genes were also measured (Fig. 2I-L). The results demonstrated that the mRNA expression levels of Bcl2 and survivin were significantly reduced in the kidney tissues of STZ-treated mice compared with in the control WT mice. Conversely, Bax and Bcl2-associated agonist of cell death (Bad) mRNA expression levels were increased in STZ-treated mice compared with in control WT mice. Treatment with melatonin significantly reduced Bad and p53 mRNA expression, and increased Bcl2 and survivin mRNA expression in STZ-treated mice in an AMPK-dependent manner.

Diabetic kidney fibrosis is associated with renal hypertrophy. To evaluate renal hypertrophy, the mice were sacrificed and kidneys were weighed. Compared with the control WT mice, the kidneys of STZ-treated mice weighed more; however, this alteration was reversed by melatonin treatment (Fig. 3A). Loss of AMPK abolished the protective role of melatonin in kidney hypertrophy. These findings were further supported by the histopathological examination of H&E- and PASM-stained kidney tissues. The results revealed that STZ-treated mice exhibited moderate glomeruli atrophy and fragmentation, epithelial desquamation, renal tubule degeneration and kidney glomerular basement membrane thickening (Fig. 3B-F). However, these renal histopathological alterations were diminished by melatonin treatment in an AMPK-dependent manner.

Notably, the preservation of renal structural integrity was closely associated with an improvement in kidney function. Renal function was assessed by determining blood urea nitrogen (BUN) and serum creatinine levels at the end of the experiment. Compared with the control WT mice, STZ increased the levels of BUN (Fig. 3G) and serum creatinine (Fig. 3H). However, melatonin application reduced the BUN and serum creatinine levels, potentially by restoring AMPK activity. Additionally, urinary albumin content and the albumin-to-creatinine ratio (ACR) were measured to determine the severity of diabetic nephropathy. As shown in Fig. 3I-J, compared with the control WT mice, STZ-treated mice exhibited significantly increased urinary albumin content level and ACR. However, melatonin administration significantly reduced the urinary albumin and ACR in the STZ-treated mice, whereas the protective effect of melatonin was inhibited by AMPK deletion.

Diabetic renal fibrosis is involved in glomerular apoptosis; therefore, it was of interest to investigate whether melatonin may suppress hyperglycemia-induced glomerular damage via AMPK. Firstly, cellular viability and apoptosis were analyzed by MTT assay and TUNEL staining, respectively. Compared with the normal group, high glucose treatment significantly reduced cellular viability (Fig. 4A) and significantly increased apoptosis of HBZY-1 cells (Fig. 4B and C). However, melatonin treatment sustained cellular viability and suppressed hyperglycemia-induced apoptosis, whereas silencing AMPK resulted in a loss of these effects.

It has previously been suggested that PGCα1 is the downstream mediator of AMPK, and activated PGCα1 favors cellular survival via inhibition of the mitochondrial apoptosis pathway (49). Therefore, PGCα1 activation may be required for the kidney protection exerted by melatonin/AMPK pathways. The present results demonstrated that PGCα1 expression was reduced by high glucose-induced stress and was reversed back to normal levels with melatonin treatment (Fig. 4D-F). However, silencing AMPK abolished PGCα1 expression, thus suggesting that PGCα1 was activated by melatonin/AMPK. To understand the role of PGCα1 in glomerular survival, PGCα1 expression was knocked down in melatonin-treated HBZY-1 cells (Fig. 4D-F). Following inhibition of PGCα1 expression, the protective role of melatonin on cellular apoptosis was inhibited, as indicated by the LDH release assay (Fig. 4G), which was similar to the results obtained in AMPK-silenced cells. In addition, these findings were further supported by the caspase-3 activity assay (Fig. 4H). Therefore, these data indicated that the AMPK/PGCα1 pathway was necessary for melatonin-induced glomerular survival under high glucose treatment.

Considering the well-established role of mitochondrial damage in glomerular apoptosis and diabetic renal fibrosis, the role of the melatonin/AMPK/PGCα1 pathway in mitochondrial homeostasis was investigated. Mitochondrial apoptosis is activated by ROS overproduction, membrane potential collapse, pro-apoptotic factor release and caspase family activation (4,37). Compared with the normal group, high glucose evoked excessive ROS production (Fig. 5A and B), which was accompanied by a decline in mitochondrial potential (Fig. 5C and D). In addition, the mitochondrial pro-apoptotic protein cyt-c was rapidly released into the nucleus (Fig. 5E and F), in turn leading to an upregulation of other pro-apoptotic proteins, including caspase-3, caspase-9 and Bax (Fig. 5G-J). Conversely, the mitochondrial anti-apoptotic factor Bcl2 was downregulated under high glucose-induced stress (Fig. 5G and K). These results indicated that mitochondrial apoptosis was activated by hyperglycemia. Conversely, melatonin treatment neutralized ROS generation (Fig. 5A and B), maintained mitochondrial potential (Fig. 5C and D), suppressed cyt-c release (Fig. 5E and F), and corrected the imbalance between mitochondrial anti- and pro-apoptotic proteins (Fig. 5G and K). Notably, silencing of AMPK or PGCα1 attenuated the beneficial effects of melatonin on mitochondrial damage. Therefore, these data indicated that melatonin sustained mitochondrial homeostasis by activating the AMPK/PGCα1 pathway, thereby protecting the glomerulus against glucotoxicity.

Besides mitochondrial apoptosis, mitochondrial function and structure are essential for cellular viability (10,50). The primary function of mitochondria is to supply sufficient energy for cellular metabolism. The ATP content was reduced by high glucose treatment (Fig. 6A) and was reversed to normal levels in response to melatonin administration via activation of the AMPK/PGCα1 pathway. Mechanistically, mitochondria produce ATP via the respiratory electron-transport chain (ETC) (38,51). The activity of ETC was decreased by high glucose treatment (Fig. 6B-D) and was reversed to normal levels with melatonin application. However, inhibition of the AMPK/PGCα1 pathway abolished the protective effects of melatonin on ETC activity.

Renal functional impairment is closely associated with structural damage. Compared with the normal group, high glucose medium promoted the opening of mPTP, an early indicator of mitochondrial damage. Notably, melatonin was able to activate the AMPK/PGCα1 pathway and thereby modify the mPTP opening rate (Fig. 6E). At the molecular level, mPTP opening is regulated by cyclophilin D, whereby p-cyclophilin D enhances mPTP opening (52,53). Based on this, the present study investigated whether melatonin and the AMPK/PGCα1 pathway controls mPTP opening via cyclophilin D. As shown in Fig. 6F and G, cyclophilin D phosphorylation was increased in response to high glucose treatment and was decreased with melatonin supplementation. Knockdown of AMPK and PGCα1 upregulated p-cyclophilin D expression, confirming that melatonin may maintain mitochondrial structure and function via the AMPK/PGCα1 pathway.