The correlation between microvessel pathological changes of the endplate and degeneration of the intervertebral disc in diabetic rats

In this study, the pathological microvessel changes to the endplate and the degeneration of the intervertebral disc of diabetic rats were examined in order to identify the possible mechanism by which diabetes mellitus (DM) induces degeneration of the intervertebral disc. A total of 30 Sprague-Dawley rats were randomly divided into two groups. DM was induced in one of the groups by streptozotocin (STZ) administration. The rats were sacrificed 4, 8 and 12 weeks later. Five rats from each group were sacrificed at each time interval and lumbar disc and endplate tissue were obtained from each rat. The histological changes, collagen expression, microvessel density (MVD) and apoptosis of the disc were investigated by different methods. The expression of collagen I in the diabetic DM group was higher compared to the control group at the three time points (P<0.01), in contrast to the expression of collagen II. The factor VIII-related antigen (FVIII RAg) was expressed in the control and DM groups, while its expression was relatively low in the DM group. The MVD of the DM group was smaller compared to that of the control group at the three time points (P<0.01). The apoptotic index (AI) in the diabetic group was significantly higher compared to that of the control group at the three time points (P<0.01). A negative correlation was observed between the MVD of the endplates and the notochordal cell AI in the two groups. Compared to the control group, the endplate MVD decreased and the cavity became smaller or disappeared in the diabetic rats. In conclusion, there was a negative correlation between MVD and degenerative changes of the intervertebral disc in diabetic rats.


Introduction
Diabetes mellitus (DM) is a systemic endocrine and metabolic disease with chronic complications that damages almost every organ and system of the body (1)(2)(3)(4)(5). One of the characteristic complications of DM is microangiopathy, which constitutes the pathophysiological basis of a wide range of organ damage. In addition to causing diabetic retinopathy, nephropathy and cardiomyopathy, diabetic microangiopathy damages the intervertebral disc (6)(7)(8)(9)(10)(11). Although the disc is avascular, it depends on diffusion from microvessels at the endplate of the disc to supply the nutrients essential for cell activity and viability and to remove metabolic wastes (12)(13)(14). Changes in blood supply cause a deficiency of nutrient supply (10,(15)(16)(17)(18). This study examined the pathological microvessel changes to the endplate and the degeneration of the intervertebral disc in diabetic rats in order to identify the possible mechanism by which DM induces degeneration of intervertebral discs.

Materials and methods
Animals. A total of 30, three-month-old male adult Sprague-Dawley (SD) rats, obtained from the Experimental Animal Center of Wuhan University, Wuhan, China, and weighing 231-263 g, were used in this study. The rats were housed 5/cage under standard laboratory conditions (12/12-h light/dark cycle, at a temperature of 24-25˚C and humidity of 50-55%) and allowed free access to food and water during the study. The rats were fed on a normal pellet diet. All the experimental protocols adhered to the Guidelines for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health (NIH Publication, revised 1996). The study was approved by the Ethics Committee for Animal Research of Wuhan University, (Wuhan, China).
Grouping and treatment. The rats were randomly divided into the streptozotocin (STZ)-induced diabetes group (DM) and the control group (n=15 rats/group). The fasting blood glucose was measured by examining blood samples from the tail vein using a rapid blood glucose meter (One Touch Ⅱ; Johnson & Johnson, New Brunswick, NJ, USA). DM was then induced by a single intraperitoneal (i.p.) injection of STZ solution The correlation between microvessel pathological changes of the endplate and degeneration of the intervertebral disc in diabetic rats (50 mg/kg). In the control group, the rats were administered the same volume of sodium citrate buffer. Fasting blood glucose levels were measured 3 days later. The blood glucose was examined ≥3 times following STZ injection and every two weeks thereafter. The standard glucose measurement for the diagnosis of DM was >13.8 mmol/l. The rats were sacrificed with a lethal dose of sodium pentobarbital (60 mg/kg i.p.) at intervals of 4, 8 and 12 weeks later. Five rats in each group were sacrificed at each time interval and lumbar disc tissue and endplate was obtained from each rat.
Histopathology. The lumbar 5/6 disc was fixed in 4% paraformaldehyde-0.1 M phosphate buffer (pH 7.4), followed by decalcification with 10% ethylenediaminetetraacetic acid (EDTA)-0.1 M phosphate buffer (pH 7.4). Following decalcification, the tissues were dehydrated in graded ethanol, embedded in paraffin, cut into 6-µm sections in the coronal plane and processed for routine hematoxylin and eosin staining for the evaluation of degeneration under a light microscope. The sections were assessed blindly by two independent authors.
Ultrastructure observation. At 12 weeks, two rats were randomly selected from each group for ultrastructure observation. The lumbar 2/3 disc was fixed with 3% glutaraldehyde solution and then sent to the Microscope Center of Wuhan University for ultrastructure observation.
Immunohistochemistry. Immunohistochemical staining was conducted using the streptavidin-peroxidase complex method. Rabbit anti-mouse collagen Ⅰ and Ⅱ, and the factor Ⅷrelated antigen (FⅧ RAg) (dilution, 1:200) (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) polyclonal antibody was used as the primary antibody. Color development was achieved with 3,3'-diaminobenzidine (DAB), which stained positive cells brown. Cells with brown particles in the plasma were evaluated as positive results. Five views of each slice were randomly selected for analysis. Images were captured on a Zeiss Axioskop 40 microscope equipped with a Canon Eos 10D digital camera (Canon, New York, NY, USA). The optical density of collagen Ⅰ and Ⅱ was analyzed by a high resolution graphic analysis system and the average optical density of the five views was recorded as the expression of collagen in the sample.
Microvessel density (MVD). Microvessels were immunohistochemically marked by FⅧ RAg staining. Positive results were determined by Weidner's method (19). After screening the areas with microvessel spots at low-power field (magnification, x100), microvessels in the area were counted in a x400 field. Five separate areas were assessed and the mean was calculated to determine the MVD of each section.
Appotosis of notochordal cells. Apoptotic notochordal cells were detected using the TUNEL method, with an in situ cell death detection kit (Roche Diagnostics, Mannheim, Germany). The assay was performed according to the manufacturer's instructions. Briefly, following routine deparaffinization and treatment with H 2 O 2 (3%), sections were digested with proteinase K (20 µg/ml, pH 7.4, 12 min) at 25˚C and incubated with the reaction mixture (1:40, 60 min) at 37˚C. Incorporated fluorescein was detected with horseradish peroxidase following a 30-min incubation at 37˚C and subsequently dyed with DAB. Brown nuclei were assessed as positive apoptotic cells. The apoptotic index (AI) was evaluated for one section of 5 high-power fields.
Statistical analysis. Data were presented as the mean ± standard error of the mean. Statistical analysis was performed using SPSS 13.0 software (SPSS, Inc., Chicago, IL, USA). One-way analysis of variance (ANOVA) with Tukey's post hoc test was used to examine differences between groups. The correlation between the endplate MVD and the nucleus pulposus cell AI was examined by Pearson's correlation analysis. P<0.01 was considered to indicate a statistically significant difference.

Results
Body weight and serum glucose levels. No accidental deaths occurred during the experiment. The STZ-induced diabetic rats showed significantly smaller body weights compared to the control animals. The typical symptoms of DM in the rats were polydypsia, polyphagia, polyuria and emaciation. The data presented in this study are the blood glucose measurement obtained the third time after the STZ injection. Following STZ injection, the serum glucose levels in the DM group were significantly higher compared to those in the control group (P<0.01) ( Table I).
Histological investigation. The histological structure of the disc appeared to be normal at 4, 8 and 12 weeks in the control group (Fig. 1A, C and E). Numerous microvessels were evident in the endplate. The discs of the controls consisted of a large amount of extracellular matrix interspersed with a small number of cells comprising ~1% of the total volume, where cell morphology varied. Those in the annulus fibrosus and cartilage endplate were more elongated and fibroblast-like compared to those of the nucleus pulposus, which were more rounded or oval and chondrocyte-like, sometimes with a capsule around them. Misalignment was not observed in the inner or intermediate layers of the annulus fibrosus. Cracks and tears were also not observed. By contrast, in the DM group, the histological structure of the discs gradually exhibited more degeneration at 4, 8 and 12 weeks (Fig. 1B, D and F). Fewer microvessels were evident in the endplate compared to the control group. Decreases in notochordal cells and increases in fibroblasts were observed. Enlarged chondrocytes were observed in the inner and intermediate layers of the annulus fibrosus. The annulus fibrosus demonstrated disrupted alignment and formation. There was also a reduction in the nucleus pulposus matrix accompanied by fibrosis and hyalinization, suggesting depletion of the extracellular matrix.
Electron microscopic findings. In the control group, organelles such as the golgi apparatus, mitochondria, lysosomes and the cell membrane were intact and orderly with abundant glycogenosome and sparse lipid droplets in the plasma ( Fig. 2A and C). Collagenous fibers were scattered neatly and closely in the plasma, while the mesh between the fibers was relatively small. The ultrastructure of the notochordal cell was different between the control and DM groups ( Fig. 2B and D).
The membrane of the cell and organelles was disrupted, while the organelles swelled and vacuolized and the nucleus structure was destroyed. Collagenous fibers were scattered loosely and messily in the plasma, while the meshes between the fibers were relatively large.
Expression of collagen Ⅰ and Ⅱ. The optical density of collagen Ⅰ and Ⅱ was used to perform a semi-quantitative analysis (Figs. 3 and 4). It was found that the collagen composition changed greatly in the nucleus pulposus, such that the proportion of collagen Ⅰ increased, while the proportion of collagen Ⅱ   (Table II). However, the expression of collagen Ⅱ in the DM group was lower compared to the control group at the three time points (P<0.01) ( Table III).
MVD of the endplates. FⅧ RAg was selected as a marker of vascular endothelial cells to reveal the microvessels of the endplates. FⅧ Rag was expressed in the control and DM groups, although the expression in the DM group was relatively low. MVD decreased in the two groups over time. The   MVD of the DM group was smaller compared to that of the control group at the three time points (P<0.01) ( Table IV).
AI of notochordal cells. AI was measured by TUNEL assay (Fig. 5). A brown nucleus was assessed to be a positive apoptotic cell. Cell apoptosis occurred in the two groups. There were more apoptotic cells/high-power field in the DM group compared to the control group. AI increased over time in the two groups. AI in the diabetic group was significantly higher compared to the control group at the three time points (P<0.01) ( Table V).

Correlation analysis between MVD and notochordal cell AI.
A decrease in MVD occurred while AI of the disc increased over time in the two groups. A negative correlation was observed between the endplate MVD and the notochordal cell AI in the control group (Pearson's correlation coefficient, r=-0.953, P<0.01) (Fig. 6A). The same was observed in the DM group, with a correlation coefficient of -0.936 (P<0.01) (Fig. 6B).

Discussion
There are two major nutrient transportation pathways in the intervertebral disc. One is the endplate pathway by     which nutrients pass through the bone marrow cavity-blood sinus-cartilage endplate route to support the nucleus pulposus. The other is the annulus pathway. Previous studies have suggested that the endplate route is the major pathway for nutrient transfer to the intervertebral disc (20,21). There is a significant decrease in cortex thickness over the central portion of endplates and shells, with a mean minimum thickness of 0.40 mm, a mean maximum thickness of 0.86 mm and an overall mean of 0.64±0.41 mm. Increased porosity is also observed along the central regions of the cortical shells (22). The porous structure accounts for ~7-10% of the endplate area (23), providing direct contact between the vertebral blood sinus and the endplate. There is a continuous capillary bed that is most dense in the area adjacent to the nucleus pulposus. There are a number of microvessel plexuses in the center with less MVD at the boundary (24). The vascular branches issuing from the microvessel plexuses pass through the porous structure and form microvessel buds in the endplate (25). These vessels drain either into the subchondral post-capillary venous network or directly into the veins of the marrow spaces in the vertebral bodies (26). The decrease in nutrients constituent in the disc is considered a key factor in disc degeneration (27)(28)(29). Loss of nutrient supply may lead to cell death, loss of matrix production and increase in matrix degradation, eventually leading to disc degeneration (10).
The changes of collagen, proteoglycan and water content are the pathological characteristics of intervertebral disc degeneration caused by the dysfunction and quantitative reduction of the disc cells (7,23,29). The collagen ingredient changes that occur during the degeneration process include the decrease of collagen Ⅱ and the increase of collagen I. A positive correlation reportedly exists between the amount of collagen reduction and the degree of intervertebral disc degeneration (30,31). The reason for the decrease in cell density is cell apoptosis, which is thought to be the principal cause of the extracellular matrix degradation (32,33). However, a high inverse correlation between the density of openings in the osseous endplate (in particular the size of the capillary buds) and the morphologic degeneration grade of the disc, supports the hypothesis that occlusion of these openings may deprive the cells of nutrients, leading to insufficient maintenance of the extracellular matrix and disc degeneration (15,34,35).
In this study, the expression of collagen Ⅰ increased along with the progression of DM. When compared to the control, the difference was significant (P<0.01). The expression of collagen Ⅱ follows the opposite pattern, decreasing with the progression of DM. In this instance, the difference between the experimental and the control groups was also significant (P<0.01). In addition, this study has demonstrated that the disc cell AI in the DM group was significantly higher compared to the control group (P<0.01).
Microangiopathy is one of the characteristic complications of DM and the pathophysiological basis of multi-organ damage. In this study, the pathological changes of microvessels in the endplate increased over time in the DM group. The cavity of the microvessels became smaller and less dense. When compared to the control, the difference was significant. The AI of the DM group was also significantly higher compared to the control group, resulting in changes in the extracellular matrix that included the decrease of collagen Ⅱ and the increase of collagen I. The difference was statistically significant compared to the control (P <0.01). The histological findings confirmed the observations that the disc degeneration of the DM rats was greater compared to that of the control. The statistical analysis demonstrated a negative correlation between the endplate MVD and the notochordal cell AI in the DM and control groups.
It can be hypothesized that hyperglycemia, low oxidative stress and advanced glycosylation end products in DM rats would cause microvessel endothelial cell injury of the disc endplates, thereby leading to either occlusion or a decrease in the number of microvessels. In the immediate aftermath, the blood supply of the disc would decline sharply, causing damage to the disc in two major ways. First, the disc nutrition would be decreased or lost entirely. Second, cellular metabolic wastes and toxins would not be able to be excreted and would accumulate in the disc leading to ischemia, hypoxia, acidosis and eventually to cell necrosis and apoptosis. Cell products would also change. The contents of proteglycan, collagen Ⅱ and water would decrease, while the contents of collagen I would increase. Calcification and disc fissures would occur and the mechanical properties of the disc would degrade.
Limitations in using a rat lumbar disc model include a smaller disc size and a different cell composition of the rat lumbar disc compared to a human disc. In addition, unlike the human disc, the rat disc is subjected to less mechanical load as it stands on four legs. Due to the limited applicability of the rat model, experiments are necessary to determine the pathological changes of the disc in human beings suffering from DM.
In conclusion, compared to the control, the endplate MVD decreased and the cavity became small or disappeared in the DM rat. DM accelerated the degeneration process of the disc. The results of this study have shown that a negative correlation exists between the mircrovessel density and the degenerative changes of the intervertebral disc within diabetic rats. The aforementioned limitations should be taken into consideration when extrapolating the results to humans.