Introduction
Osteoporosis is an important factor affecting global morbidity and mortality. Currently, OS is thought to be a complication of type 1 diabetes (T1D). This relationship is further confirmed by the decreased bone mineral density and increased risk of fracture in patients with T1D [
1]. As an independent risk factor for osteoporosis, T1D has attracted people’s attention in recent years [
2,
3]. Bone mineral density is decreased by 22 to 37% in patients with T1D [
4], and this effect is more obvious for patients with long-term T1D [
5]. Long-term hyperglycemia and the acclimation of advanced glycation end products lead to osteoporosis and fracture [
6]. Morbidity and mortality rates are more serious in T1D patients than in nondiabetic patients. Therefore, anti-osteoporosis treatment is necessary, and the treatment of diabetic osteoporosis requires further research.
Senescence has been recently shown to play an important role in the pathophysiological mechanisms of diabetic osteoporosis. Diabetes promotes senescence, reduces the turnover of bone and blood vessels, and then exacerbates changes in the bones of older mice [
7]. Compared with that in healthy controls, the proportion of senescent cells among MSCs in type 1 diabetes patients increased, as did the proportion of senescent cells among MSCs cultured under high-glucose (HG) conditions [
8]. Umbilical cord mesenchymal stem cells from gestational diabetes mellitus patients showed obvious premature cellular senescence accompanied by significantly decreased osteogenic potential [
9].
Increasing evidence has shown that melatonin exerts anti-senescence effects by preventing oxidative stress, inflammation, and mitochondrial dysfunction [
10,
11]. However, whether melatonin alleviates T1D- or HG- induced senescence has not been reported. With increasing age, a decrease in melatonin occurs, and this decrease is considered to be the key factor in osteoporosis [
12]. Additionally, there is an age-related reduction in the mRNA levels of the melatonin receptors MT1 and MT2 in mice [
11]. Pinealectomy in sheep increased trabecular separation but reduced thickness, leading to reduced bone mineral density and osteoporosis [
13]. The BMD of the femoral neck and spine in postmenopausal women with osteopenia increased in a dose-dependent manner after 1 year of melatonin treatment [
14]. However, whether melatonin can prevent diabetic osteoporosis has not been reported.
Therefore, we proposed the hypothesis that melatonin administration may prevent HG- and T1D-induced senescence, and the protective effects of melatonin against senescence may play a pivotal role in preventing osteoporosis. To test our hypothesis, we established a C57BL/6 mouse model of T1D and an HG-treated MC3T3-E1 cell model.
Materials and methods
Cell culture
MC3T3-E1 cells, which are preosteoblast cells with osteoblast differentiation ability, were used for the in vitro experiments and obtained from the Shanghai Cell Bank of the Chinese Academy of Sciences. The MC3T3-E1 cells were cultured in α-MEM medium containing 10% fetal bovine serum and 1% penicillin and streptomycin at 37 °C. When the cell density reached 80%, the cells were digested with 0.25% trypsin EDTA solution for the next experiment. The culture medium was changed every 3 days. The MC3T3-E1 cells were cultured with HG medium for 7 days and then treated with 1 μmol/l melatonin (M5250, Sigma-Aldrich, America) or not for 48 h before being harvested for senescence-associated β galactosidase staining, DCFH-DA analysis, immunofluorescence analysis, quantitative real-time PCR (RT-qPCR) analysis, and cell cycle analysis.
Cell Counting Kit-8 (CCK-8) assay
CCK-8 kit assay (CK04, Dojindo, Japan) was performed according to the manufacturer’s instructions after MC3T3-E1 cells were cultured with α-MEM medium containing various concentrations of glucose (5 mmol/l, 10 mmol/l, 15 mmol/l, 20 mmol/l, and 25 mmol/l) for 7 days. The cells were then treated with melatonin (0 nmol/l, 1 nmol/l, 10 nmol/l, 100 nmol/l, 1 μmol/l, 10 μmol/l, and 100 μmol/l; Sigma-Aldrich, America) for 48 h before being harvested. The absorbance of each well was measured at 450 nm with an enzyme-linked immunosorbent instrument after incubation at 37 °C for 1 h.
Senescence-associated β galactosidase staining (SA-β-Gal staining)
As a canonical marker of cell senescence, SA-β-Gal staining was used to identify senescent cells. MC3T3-E1 cells were washed with PBS 3 times, fixed with 1X fixation solution for 15 min, and incubated with dimethylformamide (12767, CST, America) diluted 1X β-galactosidase staining solution (9860, 20 mg/ml, pH 6.0, CST, America) overnight with parafilm sealed in a 37 °C dry incubator without CO2. Images of each group were captured with an Olympus light microscope (Olympus, Tokyo, Japan).
DCFH-DA analysis
DCFH-DA analysis was used to evaluate the cellular ROS production. MC3T3-E1 cells were washed with PBS 3 times and incubated with 10 μmol/l PBS-diluted DCFH-DA (D6883, Sigma-Aldrich, America) for 30 min at 37 °C in dark field. Subsequently, the cells were washed 3 times with PBS. Fluorescence images were captured with an Olympus inverted microscope (Olympus, Tokyo, Japan).
Immunofluorescence analysis
γ-H2AX staining was used to identify cellular DNA damage. MC3T3-E1 cells were washed with cold PBS 3 times, fixed with 4% paraformaldehyde for 1 h, and incubated with primary antibody against γ-H2AX (ab81299, 1:1000, Abcam, America) overnight at 4 °C. The cells were washed with cold PBS 3 times and coincubated with secondary antibody (Rs3211, 1:1000, ImmunoWay, China) and phalloidin (ab176757, 1:1000, Abcam, America) at room temperature for 1 h. After washing with cold PBS 3 times, the cells were stained with DAPI (4083, 0.1 μg/ml, CST, America) for 5 min. After washing with cold PBS 3 times, images were captured with an Olympus inverted microscope (Olympus, Tokyo, Japan).
Quantitative real-time PCR analysis
Total RNA was isolated with TRIzol reagent (15596026, Thermo Fisher, America) and an RNA purification kit (12183555, Thermo Fisher, America). cDNA was synthesized with the PrimeScript™ RT Reagent Kit (RR037A, TaKaRa, Japan). RT-qPCR was performed with Fast Real-time PCR System (Roche 480, Switzerland) using a TB Green® Premix Ex Taq™ Kit (RR820A, TaKaRa, Japan). The following mRNAs that related to the induction of senescence were selected: p16 (ID: 12578), p21 (ID: 12575), IL-1β (ID: 16176), and TNF-α (ID: 21926). The following mRNAs that related to the regulation of senescence were selected: Sirt1 (ID: 93759), CDK2 (ID: 12566), CDK4 (ID: 12567), cyclinD1 (ID: 12443), Bmi1 (ID: 12151), Rnf2 (ID: 19821), Suz12 (ID: 52615), and Ezh2 (ID: 14056). The following mRNAs related to osteogenesis were selected: Runx2 (ID: 12393) and Osterix (ID: 170574). The relative gene expression was normalized to β-actin. Details of the primers (Thermo Fisher, America) are shown in Supplementary Table
1.
Animal experiments-animal group
All the animal experiments were approved by the Ethics Committee of the First Affiliated Hospital of China Medical University. Thirty-six 8-week-old C57BL/6 male mice from China Medical University were given free access to water and food and were housed under a 12-h light/dark cycle and constant temperature. Streptozotocin (V900890, Sigma-Aldrich, America) was diluted in 0.1 M citrate buffer (pH 4.5). The mice were treated intraperitoneally with vehicle or streptozotocin (160 mg/kg) in a single dose. One week after STZ injection, an intraperitoneal glucose tolerance test (IPGTT) was performed after overnight fasting and intraperitoneal injection of 1 g/kg glucose. Blood samples were collected from the tail vein at 0, 10, 20, 30, 60, 90, and 120 min. Blood glucose and insulin were measured by a OneTouch glucometer analyzer (Roche, Switzerland) and mouse insulin kit (Millipore, Germany), respectively. Once the fasting blood glucose concentration exceeded 200 mg/dl for more than 1 week, the mice were considered to be diabetic. The mice that did not conform to the criteria were excluded. The mice were treated with oral administration of vehicle or melatonin (M5250, 60 mg/kg/day, Sigma-Aldrich, America) for 2 months. Then, the bilateral femurs were harvested for micro-CT analysis and protein extraction after the mice were asphyxiated by CO2.
Western blot analysis
Femurs were harvested, and the muscle and connective tissues were moved. The femurs were cut in the middle with sterilized scissors. The two parts of the femurs were placed in a centrifuge tube with the cut edges toward the bottom. Bone marrow was removed by high-speed centrifugation and PBS washing. Total femur samples containing the metaphysis and diaphysis were quickly frozen in liquid nitrogen and then crushed into small pieces with sterilized scissors. Following the manufacturer’s instructions of the Invent kit (SA-02-BT, America), the total proteins were extracted from the bone samples. Proteins were extracted from MC3T3-E1 cells using denaturing protein solution (WA-009, Invent, America) supplemented with a mixture of protease and phosphatase inhibitors (Beyotime, China). Protein separation was carried out under reducing conditions on a 6–12% Tris-acetate gel (Beyotime, China). The proteins were then transferred to the nitrocellulose membranes. The membranes were blocked in a buffer containing 5% milk for 1 h, and then incubated in 5% BSA buffer containing primary antibody at 4 °C overnight. After washing in TBST, the membranes were incubated with secondary antibody at room temperature for 1 h. The signals were detected after using ultrasensitive ECL luminescent solution (Proteintech, China). The following proteins that related to senescence identification and senescence modulation were selected. Rabbit antibody p16 (ab51243, 1:1000, Abcam, America), γ-H2AX (Ab81299, 1:5000, Abcam, America), p21 (ab188224, 1:1000, Abcam, America), p53 (ab33889, 1:1000, Abcam, America), ATM (ab199726, 1:2000, Abcam, America), MT1 (ab203038, 1:100, Abcam, America), MT2 (ab203346, 1:100, Abcam, America), Sirt1 (ab189494, 1:1000, Abcam, America), CDK2 (Ab32147, 1:1000, Abcam, America), CDK4 (ab108357, 1:1000, Abcam, America), cyclinD1 (ab40754, 1:1000, Abcam, America), Bmi1 (ab155069, 1 μg/ml, Abcam, America), Rnf2 (ab181140, 1:1000, Abcam, America), Ezh2 (ab191080, 1:500, Abcam, America), Suz12 (ab175187, 1:1000, Abcam, America), Bcl2 (ab182858, 1:2000, Abcam, America), and Runx2 (Ab236639, 1:1000, Abcam, America) were used to detect relative proteins. Mouse anti-β-actin (66009-1-Ig, Proteintech, China) was used as a load control. Goat anti-mouse (SA00001-1, Proteintech, China) and goat anti-rabbit (SA00001-2, Proteintech, China) antibodies were used as secondary antibodies. Details about the antibodies are shown in Supplementary Table
2.
Cell cycle analysis
MC3T3-E1 cells were digested and resuspended and then fixed with 1 ml cold 70% ethanol at − 20 °C overnight. After PBS washing, 50 μg/ml PI (4087, 50 μg/ml, CST, America) staining solution containing RNase A was added. Red fluorescence was detected at an excitation wavelength of 488 nm by flow cytometry (Beckman, America) after 30 min of incubation in a water bath at 37 °C, and at least 10,000 cells were counted for each sample. The results were analyzed by Flow Jo 10 (BD, America) software.
Gene transfection
MC3T3-E1 cells were cultured to 80% confluence under HG conditions for 7 days and then were treated with medium without FBS or antibiotic reagents for 6 h. Sirt1-specific SMART pool siRNAs (100 nmol/l, RiboBio, China) or MT1- or MT2-specific shRNA plasmids (GenePharma, China) were mixed with Lipofectamine 2000 reagent (11668027, Thermo Fisher, America) and then incubated for 15 min at room temperature. The cells were cultured with the conditioned medium for 6 h at 37 °C. The transfected cells were cultured with or without 1 μmol/l melatonin for 48 h before the next experiment.
Radiological study
Micro-CT (SkyScan-1276) was used to record the microstructure of the metaphyses of the femur shaft at 5 μm per layer, and the X-ray parameters were set at a voltage of 55 kV with a current of 72 μA together with a 0.4-degree rotational step. The trabecular region of 17-week-old mice was approximately 0.5 mm from the growth plate level in the direction of the metaphysis, and extended from this position for an additional 2 mm. The trabecular regions of 8-week-old and 9-week-old mice were approximately 1 mm from the growth plate level in the direction of the metaphysis, and extended from this position for an additional 2 mm. The cortical region of 17-week-old mice was approximately 6 mm from the growth plate level in the direction of the metaphysis, and extended from this position for an additional 1 mm. The cortical regions of 8-week-old and 9-week-old mice were approximately 4 mm from the growth plate level in the direction of the metaphysis, and extended from this position for an additional 1 mm. The data were reconstructed by NRecon-1.7.3.1 with a pixel size of 5.0 μm and angular step of 0.4 degree. The trabecular bone mineral density (BMD), trabecular bone volume fraction (BV/TV), trabecular number (Tb. N), trabecular thickness (Tb. Th), trabecular separation (Tb. SP), trabecular pattern factor (Tb.Pf), structure model index (SMI), cortical bone mineral density (BMD), cortical bone volume fraction (BV/TV), cortical number (Ct. N), cortical thickness (Ct. Th), and cortical porosity (Ct. Porosity) were recorded. Details about the materials and reagents involved in this paper are shown in Supplementary Table
3.
Statistics
All the data presented as the mean ± SEM were from experiments performed at least three times with biological and technical replicates. The significant differences between groups in all cases were analyzed by two-tailed, unpaired Student’s t test with GraphPad Prism 8.2.0. P values < 0.05 were considered statistically significant.
Discussion
Compared with that of control rats, the volumetric BMD of diabetic Sprague Dawley rats decreased significantly, while the BV/TV of proximal tibial trabeculae decreased by 60% [
15]. Additionally, diabetic F344 and Sprague Dawley rats had delayed cortical development compared with control rats [
15]. Furthermore, the cortical thickness and cross-sectional area of the proximal femur were lower in T1D patients [
6]. Moreover, the decrease in bone formation in T1D patients occurred in early stages of the disease, especially in patients with poor glycemic control [
16]. Moreover, at the tibial shaft, after adjusting for age, BMI, and tibial length, the cortical thickness of early-onset T1D patients was less than that of adult-onset T1D patients [
17]. Senescence has been considered one of the pathological mechanisms underlying diabetic osteoporosis [
8]. T1D and HG can both lead to senescence, as previously reported [
8,
18,
19]. Compared with those of nondiabetic mice, the mRNA levels of p53, p21, and p16 were significantly increased in the retina extracts of diabetic mice [
18]. Immunofluorescence staining revealed that high-glucose (30 mmol/l) conditions significantly increased the levels of p16 and p21 [
20]. Thus, targeting cellular senescence has been an anti-osteoporosis therapeutic strategy [
21]. As an endocrine molecule secreted by the pineal gland, melatonin exhibits anti-senescence effects [
11,
22]. To explore whether administration of melatonin can alleviate T1D-induced diabetic osteoporosis, we performed micro-CT analysis.
After melatonin administration, the amount of BMD increased [
23,
24], the width and length of cortical bone formation increased [
25], the trabecular number and trabecular thickness increased, and trabecular defects were reversed [
26]. By activating MT2, melatonin upregulates osteogenic-related protein expression and inhibits osteolysis-related factor expression [
12]. The number and volume of trabecular bones in MT2
−/− mice were significantly reduced, and the separation of trabecular bones was increased [
26]. The micro-CT results indicated that melatonin prevented T1D-induced bone loss and promoted bone formation in sham mice. To explore the anti-bone loss and osteogenesis mechanisms of melatonin, we tested the total proteins extracted from mouse femurs. In vivo experiments revealed that melatonin reduced the level of senescence-associated proteins in T1D mice and increased the level of osteogenesis-related proteins.
To explore the mechanism underlying the anti-bone loss effect of melatonin and its role in promoting bone formation, and to determine whether its anti-senescence effects play a major role in its anti-bone loss mechanisms, we performed in vitro experiments. HG culture medium was used to imitate high circulating blood glucose levels, and melatonin treatment was applied. In brief, melatonin prevented HG condition–induced ROS overproduction, DNA damage, hypo-proliferation, and senescence of MC3T3-E1 cells. Additionally, melatonin promoted MC3T3-E1 cell proliferation and increased the levels of osteogenesis-related proteins. To further explore the anti-senescence and osteogenesis mechanisms of melatonin, the study focused on the process of senescence and the proteins or factors that participate in the regulation of senescence. There are melatonin receptors expressed in the bone. Therefore, we investigated whether the anti-senescence effects of melatonin depend on its receptors, and the potential interactions between melatonin receptors and senescence-regulating proteins.
The factors leading to senescence include telomere shortening, ROS overproduction, DNA damage, and cytokine production, and these factors are associated with the senescence-associated secretory phenotype (SASP) [
27‐
30]. Senescence arrest mainly occurs in the G1 phase of the cell cycle. Most of the senescence-inducing factors activate the p53–p21 and/or pRB–p16 pathway. P16 and p21 are both cyclin-dependent kinase inhibitors [
29,
31]. Melatonin pretreatment prevented the dexamethasone-induced downregulation of CDK2 in progenitor cells obtained from the adult rat hippocampus [
32]. We previously found that melatonin at a concentration of 1 mM inhibited the proliferation of the hFOB 1.19 human osteoblastic cell line and downregulated cyclin D1 and CDK4 expression. Additionally, melatonin inhibited the proliferation of the hFOB 1.19 human osteoblastic cell line at high concentrations but increased proliferation at low concentrations [
33]. Collectively, these results suggest that the effect of melatonin on cell cycle–related gene expression is related to drug concentration and target cells. The present work found that melatonin promotes cell cycle progression and prevents HG-induced G1 arrest in an MT1- and MT2-dependent manner. The inhibition of cyclin D1 or CDK4/6 promotes the senescence of a variety of tumor cell types [
34‐
36]. Overexpression of CDK4 and CDK6 delays senescence of human diploid fibroblasts [
37]. It has been reported that melatonin at a concentration of 1 mM decreased the protein levels of CDK2, CDK4, and Cyclin D1 in H1975 non-small-cell lung cancer cells, B65 neuroblastoma cells, and X02 cancer stem cells [
38‐
40]. We found for the first time that low concentrations of melatonin increased the expression of CDK2, CDK4, and CyclinD1 through the MT1 and MT2 pathways, which may be one of the anti-senescence mechanisms of melatonin.
In proliferating cells, INK4 sites (encoding p16 and p15) are maintained in an inhibitory chromatin state by PRC1/2. It has been reported that PRC inhibits senescence and prevents bone osteoporosis [
41]. As a component of PRC1, Bmi1 overexpression promotes osteogenic differentiation potential and inhibits cellular senescence [
42]. Additionally, Bmi1 knockout mice presented with obvious premature senescence and osteoporosis [
43]. As a component of PRC2, EZH2 plays an important role in the inhibition of key senescence-related proteins in early puberty. The absence of EZH2 leads to premature cellular senescence, low bone formation, and osteoporosis [
44]. Interestingly, it has been reported that melatonin can increase the mRNA level of Bmi1 in C6 glioma cells [
45]. For the first time, we found that melatonin increased the expression of Bmi1, Rnf2, Ezh2, and Suz12 via the MT1 and MT2 pathways, which may be one of the anti-senescence mechanisms of melatonin.
As an NAD
+-dependent enzyme closely related to senescence, Sirt1 prevents glucose-induced cellular senescence of endothelial cells [
46]. Additionally, Sirt1 gene knockdown increases the expression of p53 and p21 in renal adenocarcinoma cells [
47]. It has been reported that melatonin inhibits the production of ROS by upregulating Sirt1 [
48]. Moreover, the melatonin receptor antagonist luzindole partially abrogated the upregulation effect of melatonin on Sirt1 [
49]. Then, we explored whether Sirt1 participated in the anti-senescence effect of melatonin. Here, we found that melatonin induced anti-ROS, anti-DNA damage, and anti-senescence effects by upregulating Sirt1 via the MT1 and MT2 pathway.
Collectively, melatonin prevented senescence and bone loss of T1D mice, which may be attributed to the multiple anti-senescence effects. Further to this, melatonin promoted the bone formation of sham mice, which may be attributed to the promotion of the proliferation of preosteoblasts. The limitation of this work is that the mechanisms by which the melatonin receptors exert anti-senescence effects were based on preosteoblast MC3T3-E1 cell studies.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.