Intervertebral disc degeneration (IDD) is thought to be the cause of low back pain (LBP), which has been a heavy burden to the society.1 Recent data from southern Sweden showed that the prevalence of LBP in all patients who had a musculoskeletal problem was 17.1%, and the medical fee of these patients is higher than that of the general patients.1 Likewise, 16 million Americans are affected by the diabetes mellitus.2 Meanwhile, the diabetes may influence the outcome of spinal disorders associated with back pain. In the Spine Patient Outcomes Research Trial, a secondary analysis revealed patients with diabetes achieved less gains in function compared with the nondiabetes counterparts, who had spinal stenosis or degenerative spondylolisthesis.3 A cross-sectional analysis of 95 patients showed that diabetes mellitus may be a predisposing factor for the development of lumbar stenosis.4
Furthermore, diabetes may be an important cause factor in IDD. A prospective study of 200 patients showed that incidence of diabetic patients in lumbar disk disease was significantly higher than that in other diseases.5 Similar conclusion6 can be drawn among the patients who underwent surgery for cervical disc disease. Chinese hamsters that develop spontaneous diabetes had high incidence of spondylosis.7 The disc degeneration was also accelerated in the diabetes sand rats.8 And the biochemical constituents and metabolic way of the proteoglycan in IVDs of diabetic rats were changed.9 Won et al.10 reported that the IVDs showed premature degeneration with higher incidence of apoptosis of NP cell in diabetic rats.
However, the mechanism by which the intervertebral disc degeneration is induced in diabetic patients is not known. The decrease in the number of viable cells may be one of the initial triggers of disc degeneration.11, 12 Obviously, the apoptosis and senescence of IVD cells can all lead to the decrease of viable cells number. What is more, extensive studies have found that cell apoptosis and senescence was accelerated in the degenerative IVD and associated with the increased catabolic function of NP cells.13, 14 Meanwhile, it has been proved that various cells in other tissues became premature apoptotic and senescent in diabetic rats.10, 15 So our hypothesis is that diabetic induces disc degeneration by accelerating the apoptosis and senescence of IVD cells.
As mentioned, the decrease of viable cells is an important risk factor for IDD, but macroautophagy (referred to as autophagy in this article) can protect various cells during stress responses by turning over the intracellular organelles and molecules through the lysosomal pathway.16 Autophagy, known as another type of programmed cell death, is closely associated with and shares some molecular events and regulators with apoptosis in the pathological process of human diseases.17, 18 It has been demonstrated to be relevant with a wide variety of diseases, including: neurodegeneration, diabetes, ageing, and so on.19–21 Similarly, in the IVD of rats, Ye et al.22 found that autophagy was increased with age, and in osteoarthritic chondrocytes and cartilage, autophagy was also elevated.23 Furthermore, an increasing number of studies demonstrated that autophagy mechanisms could partly account for the cell senescence.24
The formation process of autophagy needs several autophagy-related genes (Atg), most of which were first found in yeast.25 Among the genes, Beclin-1 (also known as Atg6) and microtubule-associated protein 1 light chain 3 (also known as Atg8, LC3) are required for autophagosome formation and maturation, one of the important steps for autophagy.26 As LC3-II is involved in the expansion step of autophagosome formation, it is now regarded as the most promising autophagosomal marker in mammals.27 P62 protein called the autophagy-specific substrate can be degraded together with LC3-II. The impaired autophagy is accompanied by accumulation of p62.28
To investigate the mechanisms by which diabetes promotes IDD, the streptozotocin (STZ) induced diabetes model was applied in the study. The STZ, a nitrosurea derivative isolated from Streptomyces achromogenes, is taken up by the insulin-producing β cells of pancreas in vivo.29 It increases the blood glucose concentrations and reduces the insulin concentrations obviously by damaging the β cells and abolishing the β cell response to glucose.30, 31 However, the diabetes induced by the lage dose single injection of STZ is usually classified as type 1 diabetes characterized by a specific destruction of the pancreatic β cells.31
MATERIALS AND METHODS
All the produces about animals were performed under the approval and guidance of the animal Care and Use Committee at Wenzhou medical college.
Sprague–Dawley Rat Model of Diabetes
To eliminate the influence of the overweight on the biomaterial of the intervertebral disc, we chose the streptozotocin injection to induce the diabetic model. Thirty-four 2-month-old Sprague–Dawley male rats, ranging in weight from 300 to 340 g, were obtained from Shanghai laboratory animal center, which were randomly divided into two groups. They were housed in a specific pathogen-free (SPF) room routinely and fed normal diet and water ad libitum. After adaptation, the diabetic group rats were injected intraperitoneally (i.p.) with 1% STZ (65 mg kg−1; Sigma, St Louis, MO),31, 32 dissolved in citrate buffer (pH 4.0–4.4), while the control group of rats were given vehicle citrate buffer (1 ml kg−1). After injection, all the rats still were given ad libitum access to the normal diet and tap water. Before injection and 3 days, 7 days, and 16 weeks after injection, the fasting plasma glucose (FPG) level was measured in the tail blood with an autoanalyzer (Roche, Mannheim, Germany). Only the rats with FPG level higher than >16.7 mM were kept in the diabetic group 1 week after injection.32 After injection, the weight of these rats was measured weekly.
Before injection and 16 weeks after injection, the lateral X-ray radiograph was taken to detect the disc height. We chose L4/5 disc mimicking the degenerated L4/L5 disc in human to evaluate disc height index (DHI).33 The preoperative imagings were used as a baseline measurement. Then these rats were euthanized, and the lumbar disc tissues containing the endplates from L1 to L6 were obtained. The cranial two discs of each rat were fixed in 10% neutral formalin and decalcified with 10% ethylenediaminetetraacetic acid solution, and then they were paraffin-embedded for midsagittal serial sectioning. L3/L4 discs were used to do transmission electron microscopy examination. The remaining discs were frozen immediately at −80°C for the Western blot, RT-PCR analysis, and DMMB assay.
Histology and Histochemistry
Four micrometers sections were stained with Hematoxylin–eosin and Masson–Trichrome for general histological examination. Then in order to analyze the proteoglycans expression in the extracellular matrix, the deparaffinized, and hydrated sections were stained in 1% Alcian blue solution for 30 min at 37°C, which was dissolved in 3% glacial acetic acid. Then the sections were washed with a tap water for 5 min and counterstained with 0.1% nuclear red solution dissolved in 5% aluminum sulfate for 20 min. After washing with Phosphate Buffered Saline (PBS), the sections were dehydrated routinely.
1,9-Dimethylmethylene Blue (DMMB) Colorimetric Assay
Using a dissecting microscope, the AF and NP tissue from the IVDs of the rats were separated. Then these tissues were separately pooled together and digested using papain at 60°C for 2 h. With a standard curve generated from chondroitin-6-sulfate (Sigma, C-8529), the glycosaminoglycan (GAG) content was measured in duplicate by DMMB procedure.34 The DNA concentration of each sample was measured using the PicoGreen assay (Molecular Probes, Eugene, OR) and used to normalize the GAG values.
The paraffin embedded sections were deparaffinized with xylene and rehydrated through graded ethanol. After washing, endogenous peroxidase activity was blocked with 3% H2O2 for 10 min. For immunohistochemistry for type II collagen, the sections were incubated with pepsin for 30 min to retrieve the antigen, but for cleaved caspase3, p16lnk4A35, and LC-3, these sections were incubated with trypsin for 20 min. Then the unspecific antigen was blocked by incubating with 5% bovine serum albumin and 1% Tween-20 in PBS for 30 min. Next, the sections were incubated with antibodies against collagen II (Abcam, Cambridge, MA; 1:200), cleaved caspase3 (Cell Signal Technology; 1:100), p16lnk4A (Abcam; 1:200), LC3 (Cell Signal Technology, Danvers, MA; 1:100) and PBS as negative controls overnight at 4°C. At last, the sections were incubated with corresponding HRP-conjugated secondary antibodies and counterstained with hematoxylin. In the microscopy, three fields of three sections chosen randomly from each tissue sample were imaged at 200× to quantify the percent p16lnk4A-positive and caspase3-positive cells. At least three sections from each specimen were used to analysis the expression of these proteins.
The total RNA was extracted from NP (three rats each group) using TRIzol reagent (Invitrogen, Carlsbad, CA). TRIzol chemical solution is a method of guanidinium thiocyanate–phenol–chloroform extraction technique. One microgram of total RNA was used to synthesize cDNA (MBI Fermantas, Sankt Leon-Rot, Germany). For PCR amplification, 20 µl of reaction volume included 10 µl of 2× SYBR Premix Ex Taq mixture (Takara, Shiga, Japan), 0.2 µmol/L each primer, 2 µl of twofold diluted cDNA and sterile distilled water. Then reaction and detection was carried out in a light-cycle a light cycle (Roche). The primers used were shown in Table 1. The cycle threshold (Ct) values were collected and normalized to the housekeeping gene GAPDH. TheΔΔCt method was used to calculate the relative mRNA levels of each target gene.36
Table 1. Primer Sequences for Aggrecan, Collagen II, Collagen I, and GAPDH
Primer Sequences (5′–3′)
GAPDH, glyceraldehyde phosphate dehydrogenase.
TUNEL is the technique of terminal deoxynucleotidyl transferase dUTP nick end labeling for detecting DNA fragmentation from apoptotic signaling cascades. When the IVD sections were dewaxed, they were incubated with 15 µg/ml of proteinase K for 15 min at 37°C. After using 3% H2O2 to quench endogenous peroxidase for 5 min at room temperature and washing three times with PBS, in situ cell death detection kit (Roche) was applied to the sections according to the manufacturer's instructions. Meanwhile, the positive and negative controls were carried out in these sections which were treated with DNase (2 U/ml) for 1 h or processed without terminal transferase separately.
Transmission Electron Microscopy (TEM)
After fixing in 2.5% glutaraldehyde for one night, the NP was postfixed in 2% osmium tetroxide and block-stained with 2% uranyl acetate. Following the dehydration in series acetone, these cubes were embedded into Araldite. Semi-thin section and Toluidine Blue staining were performed for observation of location. At last, ultra-thin sections of at least three blocks from one disc were cut and then they were all observed in a Hitachi TEM.
The protein of NP was isolated using RIPA Lysis Buffer (Beyotime, Nantong, Jiangsu, China), whose concentration was determined using an Enhanced BCA Protein Assay Kit (Beyotime). The proteins were separated by the SDS–PAGE and transferred to polivinyledene fluoride (PVDF) membranes, which were incubated with primary antibodies against caspase-8, caspase-9, caspase-3 (Cell Signal Technology, 1:1,000), p16lnk4A, P62, Beclin-1 (Abcam, 1:500), LC3 and β-actin (Cell Signal Technology, 1:1,000) and probed with the respective secondary antibodies. Then the bands were detected with ECL plus reagent (Invitrogen) by the system of enhanced chemiluminescence detection (PerkinElmer, Waltham, MA). At last, the intensity of these bands was quantified by the software of AlphaEaseFC 4.0.
Statistical analyses were carried out using the SPSS 15 statistical software program (SPSS, Inc., Chicago, IL). The Student's t-test was used to analysis the difference between groups. Differences were considered statistically significant at an error level of p < 0.05.
At 1 week after injection, the FPG level of two rats in the diabetic group was less than 16.7 mM, but the STZ increased significantly the FPG of the remaining rats (n = 15) in the diabetic group, compared with control (n = 17; 20.7 ± 3.5 mM vs. 7.0 ± 1.7 mM p < 0.01). At 16 weeks after injection, the FPG of diabetic rats remained higher than control rats (21.9 ± 4.4 mM vs. 7.8 ± 2.3 mM p < 0.01), but the body mass of diabetic rats, as expected, was lower than control rats (352 ± 79 g vs. 513 ± 89 g p < 0.01). One weak rat in the diabetic group died one month after injection and the mortality rate of the diabetic rats was 6.7%. At last, in diabetic group, 28 discs were distributed to do histology, histochemistry, immunohistochemistry, and TUNEL assay; 14 discs were distributed to do TEM; 28 discs were distributed to do Western blot, RT-PCR and DMMB. In the control group, the sample number of each assessment has two more discs than the diabetic group. The diabetic rats were exposed to hyperglycemia for 16 weeks, which models chronic hyperglycemia in humans (10–15 years). When the exposed time was extended, the mortality rate of diabetic rats increased due to complications of severe hyperglycaemia.37 As a result, the time point of 16 weeks was chosen to do the study.
Intervertebral Disc Degeneration in Diabetic Rats
The X-ray could not found obvious decrease of the disc height in each group, and no difference in DHI score of L4/5 disc was seen between the diabetic group and the controls (results not shown). But the physicochemical properties of the groups were different. In the NP of most diabetic rats, some notochordal cells disappeared, especially in the border between AF and NP, which were replaced by the chondrocyte-like cells (Fig. 1A, B, E, and F). What is more, in some region of NP from diabetic rats, there were some cell clusters, which was a degeneration mark.38 However, the cell clusters cannot be found in the nucleus pulposus of control rats. Using the method Alcian blue which dyeing the disc blue color, a substantial reduction of PGs was observed in diabetic rats compared with the control rats (Fig. 1C, D, G, and H). The similar result was observed in the DMMB assay (Fig. 1I). The GAG content of NP and AF was reduced significantly in diabetic rats (387 ± 31 µg GAG/ng DNA in NP and 55 ± 9 µg GAG/ng DNA in AF) compared with control rats (496 ± 56 µg GAG/ng DNA in NP and 77 ± 16 µg GAG/ng DNA in AF).
The expression of collagen II was mainly observed in transition zone and the pericellular area in NP of control rats. At the border of NP and AF, the collagen II formed the lacuna surrounding the NP cells (Fig. 2A and B). The intensity of collagen II expression in the extracellular matrix of diabetic disc was decreased obviously, compared with controls (Fig. 2A–H). But the chondrocyte-like cells in transition zone of diabetic disc can be stained intensively (Fig. 2E and F). Consistent with the result of immunohistochemistry and DMMB, the aggrecan and collagen II mRNA level in the NP cells of diabetic rats was significantly less than that in control rats (aggrecan p = 0.028, collagen p = 0.016; Fig. 2I). But the difference of collagen I mRNA in the two groups was not significant (p = 0.253; Fig. 2I).
Apoptotic Effect of Diabetes on Nucleus Pulopsus Cells
A few apoptotic cells were stained green in NP of control rats, but apoptosis developed significantly in the NP of diabetic rats (Fig. 3A and C, p < 0.01). Immunohistochemistry for detection of cleaved caspase3 demonstrated that cleaved caspase3-positive NP cells were more apparent in diabetic rats (Fig. 3A and C, p < 0.01). Further more, diabetes led to activations of caspases-8 (initiator caspase in the extrinsic pathway), caspase-9 (initiator caspase in the intrinsic pathway), and caspase-3 (executioner caspase in both intrinsic and extrinsic pathways; Fig. 3B and C, active caspase8/β-actin p = 0/019, active caspase9/β-actin p < 0.01, active caspase3/β-actin p = 0.034).
Increased Level of Nucleus Pulposus Cell Senescence in Diabetic Rats
To detect the level of nucleus pulposus cell senescence, we used immunochemistry and Western blot to analysis the expression of p16lnk4A protein, a senescence marker. Immunoreactivity for p16lnk4A was restricted to the nucleus of the NP cells (Fig. 4A). The discs in diabetic group showed significantly higher proportions of p16lnk4A immunopositive cells (23 ± 9%), compared with the controls (13 ± 6%; p = 0.013; Fig. 4A and C). When studying the relative expression of p16lnk4A protein using Western blot, it also increased in the NP of diabetic rats (Fig. 4B).
Increases of Autophagy in NP Cells of Diabetic Rats
When observing the ultrastructure of the NP cells from control rats, numerous large vesicles and glycogen depositions always can be found in the cytoplasm, but rare autophagosomal vacuoles were observed (Fig. 5A and B). In diabetic rats, there were some autophagosomes in the cytoplasm, which were double-membrane structures containing parts of cytoplasmic organelles (Fig. 5C and D).
To exactly quantify the amount of autophagy expression in diabetic rats relative to the controls, we next examined the LC3, Beclin-1, and P62 expression by immunochemistry and Western blot. Using immunochemistry, LC3-positive cells was mainly observed in NP of discs. Compared with control rats (Fig. 6B and C), the intensity of LC3 positive-staining increased obviously in the NP of diabetic rats (Fig. 6E and F). As LC3-II has been regarded as an indicator of autophagy activity, we also compared the level of LC3-II expression between the two groups. As shown in Figure 5G, the degree of LC3 conversion (LC3-I to LC3-II) was increased in the NP of diabetic rats. The relative optical quantification revealed that LC-3II/LC-3I and Beclin-1/β-actin increased significantly in diabetic rats (Fig. 6H, LC-3II/LC-3I p < 0.01, Beclin-1/β-actin p = 0.012). However, the P62 protein expression decreased in diabetic rats compared with control (Fig. 6H, p = 0.078).
Although it has been found the association between diabetes and IDD,5, 8 higher rates of older age, obesity, and other complications found in diabetic patients may confuse the conclusion of the association. As a result, the animal model was chosen to study the effect of diabetes on IVDs, whose age and complications can be controlled. The diabetes animal models include: (1) genetically induced spontaneous diabetes models; and (2) experimentally induced non-spontaneous diabetes models. Some studies have used the first model to explore the influence of diabetes on IVDs.10, 39 However, to our knowledge, the second model has not been used to study the IVD. The type 1 diabetes induced by the large dose single injection of STZ is characterized by the significant high blood glucose concentration, low insulin concentrations, polyuria, and continued weight loss. Due to the fact that the insulin is essential to the transport of glucose across the cell membrane, the type 1 diabetic rats with absence of insulin can be usually regarded as in starvation from an energy storage standpoint.40 The weight lost in the diabetic rats could eliminate the influence of obesity which is an independent contributor to IVD degeneration.41
The reasons for the aminal death after STZ injection can be classified into two categories: acute STZ toxicity and complications of severe hyperglycaemia.37 When the animals died more than 10 days after STZ injection, the reason is always the complications of severe hyperglycaemia.37 Bloch et al.42 found a survival of 70 + 7% over their 4-week study in diabetic control ICR mice after a large single injection of STZ. Deeds et al.37 observed a 10% mortality rate of the diabetic mice. In the present study, because the rats were housed in a specific pathogen-free (SPF) room and raised carefully, most diabetic rats could survive during the 16-week period which is commensurate with 10–15 years in humans.
Changes in physicochemical properties can discern disc deterioration before it is apparent on histological sections or radiological images. Although Videman et al.43 found the MRI representation of IVDs did not differ between diabetic patients and their twins, the hydration, fixed charge density (FCD), and hydration under various osmotic pressures in diabetic rats decreased compared with the control rats.44 Similarly, in the present study, the X-ray representation was not different between the two groups, but we found the expression of the protein and mRNA of PGs and collagen II in the IVD cells of diabetic rats decreased significantly. Because extracellular matrix (ECM) changes and gene expression of tissue-specific ECM components largely reflect alterations in the biology of the cells, the function of IVD cells in diabetes was damaged obviously. On the other side, the expression of extracellular matrix metalloproteinases including MMP-1, -2, -3, and -13 also increased in the disc of diabetic rats.10 IDD is biochemically characterized by the decrease of ECM synthesis and increase of ECM degradation,11 so we suggested that diabetes can break the balance of ECM metabolism in disc degeneration.
The apoptosis can be induced in various models of disc degeneration such as: mechanical models, structural, and genetically induced models.45, 46 Consistent with the previous studies,10 we also found the apoptosis of NP cells was increased in diabetic rats. Three signal pathways have been found to be related with disc degeneration, which include death receptor (extrinsic pathway), mitochondrial (intrinsic pathway), and endoplasmic reticulum pathways.47 Won et al.10 found the expression of Fas leading to the extrinsic pathway was significantly increased in the disc of diabetic rats. In the present study, the expression of caspase9 and caspase8 was both increased in diabetes rats, indicating that both the extrinsic and intrinsic pathways were activated by diabetes. Wang et al.47 demonstrated that apoptosis occurred via these three apoptosis pathways together in degenerative discs of patients.
Cell senescence has been considered as a correlative factor of disc degeneration in humans.13, 35 Several cellular senescent markers have been investigated in the degenerated disc and cartilage, such as: senescence-associated-beta-galactosidase (SA-beta-Gal) and p16lnk4A.13, 48, 49 P16lnk4A plays an important role in regulating cell growth and senescence.49 It can inactivate the cyclin dependent kinase that phosphorylates the retinoblastoma (RB) protein to slow down the progression of the cell cycle.50
The NP cells exposed to higher levels of glucose showed decreased proliferation compared to those in culture with normal levels of glucose.51 Ziv et al.44 found that the physicochemical properties of young diabetic sand rat disc were similar with that of the discs from old rats. In our study, the diabetic rats exhibited greater percentage of cells expressing p16lnk4A in their discs. The Western blot also revealed the higher expression of p16lnk4A protein in the diabetic rats. Further, the transition of notochordal cells to chondrocyte-like cells in NP indicating disc became premature was observed. Meanwhile, the cellular senescence can also reduce the PG and collagen II synthesis in diabetic disc. One of the mechanisms for induction of cellular senescence is stress-induced premature senescence.35 The stress includes reactive oxygen species (ROS),52 mechanical load,53 or cytokines.54 So whether the diabetes can intensify the ROS in NP cells will be the topic of our further study.
Microangiopathy was suggested to account for the above change of NP cells.5, 6 The vasoconstriction induced by the smoking has been suggested to reduce the activity of IVD cells and lead to IDD.55 As we all know, diabetic retinopathy and nephropathy are the two major diabetic microvascular complications. The microvascular abnormalities of type 1 diabetes induced by STZ include the thickening of the basement membrane and vascular occlusion, which can all lead the microvascular stenosis.56, 57 The nutrition of disc is mainly supplied by diffusion through the endplates and the decrease of nutrient in disc may be the cause of IDD.58 Some studies have demonstrated that the endplate permeability is due to microvascular network in the central endplate.59 Obviously, the narrowing of the vascular lumen in the vicinity of the endplate can lead to the decrease of nutrient and waste exchange between discs and the surrounding vascular. Furthermore, the apoptosis and senescence of the NP cells are influenced by the nutrients supply.60
Autophagy is currently considered as an important protective mechanism in age-related diseases or degenerated diseases is a cellular catabolic pathway leading to lysosomal degradation and recycling of proteins and organelles. In osteoarthritis, Carames et al.61 found that autophagy activated by rapamycin reduced the severity of cartilage degradation in the mice. But there are fewer studies about autophagy in intervertebral disc degeneration. On one hand, autophagy can prevent annulus fibrosus cells from apoptosis in vitro under serum starvation condition.62 The permeability of the endplate can be blocked by the microangiopathy induced by diabetes, leading to block nutrition supply of the NP cells. The increased level of autophagy in diabetic NP cells may partly account for the cell apoptosis. To our knowledge, this is the first article to report the coexistence of autophagy and apoptosis in disc in vivo. On the other hand, autophagy in the rat disc wass increased with age.22 In the diabetic rats, when the disc became premature10 and NP cells became senescent, the LC3-II/LC3-I and Beclin-1/β-actin protein which represented the autophagy flux increased, but the P62 protein which was degraded through the autophagy-lysosomal pathway decreased. Using EM, we also demonstrated the increased autophagosomal vacuoles.
Autophagy has received interest recently due to its potential in regulation of the aging process. Multiple studies have reported that the activity of autophagy was reduced in aged tissues.63, 64 In the cartilage of mice, Carames et al.65 demonstrated that the activity of autophagy was decreased with age and early degeneration. However, the conflicting observations were reported by Sasaki et al.23 in human cartilage. In the discs of rats, Ye et al.22, 62 also found that the activity of autophagy was increased in the degenerated and aged discs. Several possible explanations can account for the tissue difference of autophagy activity. First, the discs used in the studies may be in the different periods of the progressive course of IDD. Sasaki et al.23 suggested autophagy may increase as an adaptive response to protect cells in the early stage of osteoarthritis and failure of the adaptation may lead to further progression of osteoarthritis. The aged discs from 24-month old rats may still be in the early stage of IDD. Second, the fact that NP cells suffer from starvation in the degenerated and aged disc can explain the controversy. Autophagy can be induced under serum starvation in the annulus fibrosus cells.62
A major limitation of this study is that the streptozotocin-induced diabetic model is a type 1 diabetes, which cannot perfectly reflect all aspects of human disc degeneration in type 2 diabetic patients. The second, owing to the higher mortality rate of diabetic rats, the time course during diabetes develops has not been established in the study. The third, the accurate factor in diabetes that influences the disc degeneration still is not elucidated absolutely.
In conclusion, the present study establishes a direct cause effect relationship between diabetes and IDD. Diabetes induced by STZ can cause disc degeneration by aggravating nucleus pulposus cells apoptosis and senescence. Microangiopathy may play a major role in the change of IVD cells. What is more, the cell autophay as a response for cell apoptosis and senescence increased in the disc of diabetic rats, which may be a protective mechanism.