Metformin inactivates the cGAS-STING pathway through autophagy and suppresses senescence in nucleus pulposus cells

Intervertebral disc degeneration (IVDD) is one of the most common skeletal muscle diseases. It can cause lower back pain (Luoma et al., 2000; Vergroesen et al., 2015), and affects mostly middle-aged and elderly people (de Souza et al., 2019). However, so far, the exact pathogenesis of IVDD has not been identified, and effective therapeutic drug treatments for IVDD are limited (Sakai and Grad, 2015). The intervertebral disc is mainly composed of two parts – the interior gelatinous core (nucleus pulposus) and the exterior hard-outer ring (annulus fibrosus) (Adams and Roughley, 2006). The nucleus pulposus, as the main functional unit of the intervertebral disc, secretes extracellular matrix to maintain the stability of the intervertebral disc, whereas the annulus fibrosus can protect the nucleus pulposus from external mechanical shocks (Rannou et al., 2004).

Although the specific mechanism of IVDD remains unclear, studies have reported that senescence is related to IVDD (Che et al., 2020; Roberts et al., 2006); the number of senescence-associated (SA)-β-gal positive cells (senescent cells) are increased in degenerative intervertebral disc tissue. Senescent cells can secrete a variety of cytokines, named the senescence-associated secreted phenotype (SASP) (Kuilman et al., 2008). SASP factors can affect the intracellular environment and lead to local or systemic dysfunction. Therefore, the senescence of nucleus pulposus cells is the main pathological basis of IVDD, and inhibiting the senescence of nucleus pulposus cells can be regarded as a therapeutic method for the treatment of IVDD (Le Maitre et al., 2007; Roughley, 2004).

The cGAS-STING signal is a key indicator of innate immune response with the major function of identifying and responding to cytoplasmic DNA. Upon recognition and binding to DNA in the cytoplasm by cGAS, STING (also known as STING1) is activated to initiate a series of downstream reactions (Su et al., 2020). In addition, studies have reported that it can stimulate the generation of inflammatory factors (Wan et al., 2020), lead to SASP factos secretion in senescent cells and accelerate senescence (Loo et al., 2020; Vizioli et al., 2020). These studies show that cGAS-STING is involved in the process of senescence.

Metformin (Met) is a common hypoglycemic drug that can reduce oxidative stress in diabetic nephropathy by activating autophagy (Ren et al., 2020). In addition, it can also exert a protective effect in the brain (Son et al., 2016) and heart (Xie et al., 2011) by activating autophagy. Autophagy is a dynamic process, during which the damaged proteins or organelles are wrapped by autophagy vesicles in a bilayer membrane structure to form autophagosomes, which then fuse with lysosomes to generate autolysosomes, resulting in the degradation of the wrapped contents (Klionsky et al., 2012). It is a self-protection mechanism for cells that can help maintain a stable environment of the cell and improve its own function (Levine and Kroemer, 2019). Some studies have revealed that metformin can slow down aging (Bharath et al., 2020; Kulkarni et al., 2020). Moreover, our previous research found that metformin can activate autophagy to inhibit the senescence of nucleus pulposus cells (Chen et al., 2016), but the specific mechanism of metformin in nucleus pulposus cells is not clear. This study is aimed at exploring the relationship between metformin-activated autophagy and cGAS-STING pathway.

According to the results of our previous experiments (Chen et al., 2016), metformin has no effect on cell viability at concentrations of 10–100 μM, so the concentration gradient 10 μM, 50 μM and 100 μM was selected for the current experiment. Fig. 1A shows that treatment of nucleus pulposus cells with tert-butyl hydroperoxide (TBHP), which is a known inducer of oxidative stress and cellular damage (Wedel et al., 2020), decreased the cell viability of cells in a dose-dependent manner, which was reversed by metformin treatment (Fig. 1B,C). As shown by western blotting, the level of the markers P16, P21 and P53 (also known as CDKN2A, CDKN1A and TP53, respectively), which are more highly expressed during nucleus pulposus cell senescence (Chen et al., 2021), was inhibited by Met treatment in a dose-dependent manner (Fig. 1D,E). In addition, the staining of SA-β-gal further confirmed the protective effect of Met on nucleus pulposus cells. The number of SA-β-gal-positive aging nucleus pulposus cells increased significantly after the treatment with TBHP. However, Met reversed this phenomenon (Fig. 1F,I). We also found after the treatment of nucleus pulposus cells with TBHP that the expression of cGAS-STING increased, according to western blot detection, and the phosphorylation of downstream P65 (also known as RELA, an NF-κB subunit) was enhanced, which could be reversed by Met in a dose-dependent manner (Fig. 1G,H,K). Immunofluorescence results showed that the treatment of TBHP facilitated P65 protein transfer to the nucleus, whereas silencing of STING protein and Met treatment inhibited the effect of TBHP. Furthermore, we also found that overexpression of STING reversed the effect of Met on TBHP, and P65 protein re-translocated from the cytoplasm to the nucleus (Fig. 1J,L). These results reveal that Met can suppress the senescence of nucleus pulposus cells and regulate the activation of cGAS-STING pathway.

Lentiviruses were used to overexpress or silence STING to determine its effect on senescence in nucleus pulposus cells. We found that knockdown of STING (Sh-STING) significantly inhibited the expression of the TBHP-induced senescence marker proteins P16 and P21 compared with the control (Sh-NC) group, whereas STING silencing alone did not affect the expression of these proteins. However, overexpression of STING protein increased the expression levels of P16 and P21, indicating a positive correlation between STING and senescence, and that STING promotes senescence (Fig. 2A–D). This result was further confirmed by the results of SA-β-gal staining (Fig. 2E,K); the number of senescent nucleus pulposus cells was significantly decreased when STING protein was knocked down, whereas the number of senescent nucleus pulposus cells was increased when STING protein was overexpressed. Subsequently, we explored the relationship between Met and STING protein. As shown in Fig. 2G, Met inhibited TBHP-induced expression of P16 and P21, but this phenomenon was reversed after overexpression of STING, the SA-β-gal staining showed that the number of senescent cells was decreased after Met treatment, but senescent nucleus pulposus cells were instead increased after overexpression of STING (Fig. 2I,J), indicating that STING is involved in the regulation of Met against TBHP-induced cell senescence. In addition, we also examined the gene expression of inflammatory factors IL-1β and IL-6. As shown in Fig. 2F, the expression of inflammatory factors decreased after Met treatment, but overexpression of STING protein (Lv-STING) reversed the inhibitory effect of Met on inflammatory factors. These results illustrate that Met reduces inflammatory responses by inhibiting the activation of cGAS-STING pathway and ultimately delaying senescence of nucleus pulposus cells.

Nucleus pulposus cells were treated with metformin of different concentrations 10, 50 and 100 μM of Met for 24 h. The ratio of LC3-II/LC3-I (LC3-II represents the lipidated forms of LC3 protein, LC3-I the non-lipidated form; LC3 proteins are also known as MAP1LC3 proteins) and the expression of P62 (also known as SQSTM1) were detected by western blot. As shown in Fig. 3A,B, as the concentration of Met increases, the ratio of LC3-II/LC3-I increased in a dose-dependent manner, and the level of p62 decreased in a dose-dependent manner. In addition, we observed the generation of autophagosomes using transmission electron microscope, known as the gold standard for the detection of autophagy activation (Klionsky et al., 2016). Met-treated nucleus pulposus cells exhibited more autophagosomes in the cytoplasm, indicating that met activated autophagy (Fig. 3C,D). By double immunofluorescence staining for LC3-II (a marker of the autophagosome; Singh and Bhaskar, 2019) and LAMP1 (a marker of the lysosome; Cheng et al., 2018), we observed that Met treatment facilitated autophagosome–lysosome fusion (Fig. 3E,F). Rapamycin (an autophagy activator; Spang et al., 2014) and 3-methyladenine (3-MA; an autophagy inhibitor; Wu et al., 2010) were used to activate or inhibit the occurrence of autophagy. Western blot detection showed that the expression of cGAS, STING, phosphorylated (p)-P65, P16 and P21 in nucleus pulposus cells under the treatment of TBHP increased, while autophagy activated by rapamycin or Met can inhibit the phenomenon induced by TBHP. In addition, after adding 3-MA to nucleus pulposus cells treated with Met, we found that 3-MA suppressed the protective effect of Met on nucleus pulposus cell senescence (Fig. 3G–I). The results of SA-β-gal staining also verified this phenomenon; the number of positive senescent nucleus pulposus cells decreased after the treatment with Met, which was reversed by treatment with 3-MA (Fig. 3J,K). In addition, we used chloroquine (CQ) for verification, as shown in Fig. S2. To sum up, Met inactivates the cGAS-STING pathway and inhibits senescence through inducing autophagy.

Decomposition and rupture can occur after cell senescence, resulting in a large number of DNA damage fragments; γ-H2AX (phosphorylated H2AX) is a marker of DNA damage (Kuo and Yang, 2008; Rothkamm et al., 2015). Damaged DNA fragments recognized by the cGAS protein then activate STING protein to initiate downstream nuclear factor (NF)-κB activation and other reactions (Dunphy et al., 2018). Western blot results showed that TBHP led to the upregulation in the level of γ-H2AX and cGAS and STING over a period of time in a time-dependent manner (Fig. 4A,B). However, Met reversed this phenomenon, and Met activates autophagy to degrade the damaged fragments, thus inhibiting the γ-H2AX formation (Fig. 4C,D). In addition, we found that the autophagy-promoting effect of Met was significantly inhibited after treatment with 3-MA, thereby eliminating the inhibitory effect on γ-H2AX (Fig. 4E–G). Furthermore, immunofluorescence analysis showed that Met inhibited the induction of γ-H2AX stimulated by TBHP (Fig. 4H,I). These results suggest that Met inhibits the activation of cGAS-STING pathway by activating autophagy to clear damaged DNA fragments.

An acupuncture-induced rat model of IVDD was established. After operation, the rats were visualized through X-ray imaging, and Hematoxylin and Eosin (H&E) and Safranin O-fast green (S-O) staining, and the imaging and histopathological changes were observed. X-ray analysis showed that at 4 weeks after disc puncture surgery, the loss of disc height in the Met group was less than that in the IVDD group (Fig. 5A,B). After 4 weeks, the number of nucleus pulposus cells in the nucleus pulposus of the IVDD group was gradually reduced and replaced by fibro chondrocytes. However, these histopathological changes were significantly delayed after Met treatment, and the histological score of the Met group was significantly lower than that of the IVDD group (Fig. 5C,D). The administration of Met and 3-MA alone did not affect disc degeneration, as shown in Fig. S1. In addition, it is worth noting that according to the immunohistochemical staining results, Met treatment could increase the level of LC3-II and inhibit the expression of P16, P62, cGAS and STING in the intervertebral disc of rats (Fig. 5E,F). To sum up, Met ameliorates disc degeneration by inactivating the cGAS-STING pathway through autophagy.

Due to the aging population, the incidence of age-related IVDD is gradually increasing (Vo et al., 2016). IVDD is mainly characterized by lower back pain and has a serious influence on daily life. In the progression of IVDD, the release of inflammation factors is the main cause of excessive senescence of nucleus pulposus cells. Some studies have shown that there is oxidative stress in the degenerative intervertebral disc, and the release of reactive oxygen species (ROS) can destroy cell homeostasis and cause inflammation to accelerate the senescence of nucleus pulposus cells (Chen et al., 2014; Lin et al., 2020). Herein, we treated nucleus pulposus cells with TBHP to release a more-stable exogenous ROS to establish an IVDD model (Chen et al., 2020), inhibiting the progression of IVDD by inhibiting the occurrence of oxidative stress.

The cGAS-STING signal transduction axis consists of the synthase of the second messenger ring GMP-AMP (cGAS) and the cyclic GMP-AMP receptor stimulator (STING) of the interferon gene. It is an innate immune pathway, and an important axis of inflammatory response and cell senescence (Hopfner and Hornung, 2020). In addition, the cGAS-STING signaling pathway is a newly found pathway that responds to cytosolic DNA. Under normal circumstances, intracellular DNA is stable, existing in nuclei and mitochondria. When DNA is damaged, followed by the phosphorylation of histone γ-H2AX (a DNA damage marker) (Kuo and Yang, 2008), the damaged DNA fragment can enter the cytoplasm and be recognized by cGAS (Gekara, 2017; Song et al., 2019), thus activating STING and its downstream molecules. NF-κB has been confirmed as a downstream molecule of cGAS-STING (Fang et al., 2017). When the damaged DNA fragment induces the activation of cGAS-STING, it can phosphorylate P65 and promote its transfer to the nucleus. This will facilitate the generation of inflammatory cytokines (Aarreberg et al., 2019) and then induce oxidative stress, resulting in the degradation and destruction of extracellular matrix (Wang et al., 2019), as well as the acceleration of cell senescence.

Furthermore, it has been reported that the activation of cGAS-STING can activate downstream pro-inflammatory responses, and the release of inflammatory factors triggers the generation of SASP factor (Dou et al., 2017; Glück et al., 2017; Yang et al., 2017), thus inducing the senescence, and the loss of cGAS protein or STING protein will cause SASP factors to decrease (Dou et al., 2017). The above findings indicate that the cGAS-STING signaling pathway is abnormally activated in senescent cells. This present study has also found that silencing the STING gene can inhibit the senescence of nucleus pulposus cells, while the overexpression of STING can promote the senescence of nucleus pulposus cells. This suggests that the cGAS-STING signal can be used as a new therapeutic target to inhibit senescence, thus alleviating the progress of IVDD.

As a self-protective mechanism for cells, autophagy has an active protective effect on neurodegenerative diseases (Luo et al., 2020), cardiac infarction (Zhang et al., 2019), osteoarthritis (Zhang et al., 2015) and other diseases. In addition, previous studies in our group also demonstrated that Met could inhibit the senescence of nucleus pulposus cells by activating autophagy (Chen et al., 2016); however, the molecular mechanism remains unclear. The abnormal activation of cGAS-STING pathway and the increased expression of autophagy-related genes in senescent cells (Gruber et al., 2015) lead us to speculate on the relationship between autophagy and the cGAS-STING pathway.

Our study showed that autophagy induced by Met inhibited the abnormal activation of cGAS-STING. We found that the level of γ-H2AX (an DNA damage marker) was improved in TBHP (as a ROS donor)-induced IVDD, and TBHP induced the formation of γ-H2AX in a time- and dose-dependent manner. However, Met inhibits the formation of γ-H2AX by activating autophagy via the AMPK pathway (Chen et al., 2016). After using 3-MA to suppress autophagy, the inhibitory effect of metformin on γ-H2AX was abolished. Damaged DNA fragments in the cytoplasm of senescent nucleus pulposus cells are identified by cGAS and then recruit downstream STING to activate the NF-κB signaling pathway, thus promoting the release of inflammatory cytokines. After being treated with Met, autophagy is activated to degrade the damaged DNA fragments and downregulate the levels of γ-H2AX (Gkirtzimanaki et al., 2018), thus inhibiting the activation of cGAS-STING signaling pathway and the occurrence of downstream pro-inflammatory responses. The reduction of inflammatory factor release in turn inhibits the SASP, finally inhibiting the senescence of nucleus pulposus cells. The results of in vivo experiments were consistent with the in vitro experiments results; Met injection delayed the pathological changes associated with IVDD, and Met inhibited the expression of cGAS and STING, as shown by immunohistochemical staining.

To sum up, we previously showed that Met inhibits disc degeneration, but in this study, we explored deeper mechanisms. We report the that cGAS-STING signal could be used as a therapeutic target for IVDD, and that Met inhibits the activation of the cGAS-STING signal by activating autophagy, thus having a therapeutic effect on diseases such as IVDD (Fig. 6). Our research shows that Met might become a potential therapeutic agent for cGAS-STING-related diseases.

The experimental procedures and animal use procedures are in line with the guidelines for the use of Experimental Animals of the National Institutes of Health and were approved by the Animal Care and Use Committee of Wenzhou Medical University (WYDW2019-0027).

Metformin (purity≥99%), type II collagenase, 3-methyladenine (3-MA), tert-butyl hydroperoxide solution (TBHP) and chloroquine (CQ; used at 50 μM) were purchased from Sigma-Aldrich (Missouri, USA). ABCAM (Cambridge, Massachusetts, USA) provided primary antibodies for P16 (ab51243), P21 (ab86696), γ-H2AX (ab81299), cGAS (ab179785), STING (ab179775), Beclin-1 (ab207612), P62 (ab240635), LC3 (ab62721), P65 (ab16502). The antibody dilution used was 1:1000. Cell Signaling Technology (USA) provided primary antibodies for p-P65 (#3033) and GAPDH (#2118). The antibody dilution used was 1:1000. DAPI was obtained from Beyotime (Shanghai, China).

Sprague–Dawley (SD) rats (male, 200–250 g, 4 week) were used for cell collection. Under sterile conditions, the L1–L6 spine was carefully removed and lumbar intervertebral discs were collected. Gelatinous nucleus pulposus tissue was isolated from lumbar intervertebral disc under a microscope and digested with 0.25% type II collagenase (Sigma-Aldrich, USA) at 37°C for 2 h. After centrifugation at 95 g for 5 min, the cell suspensions were cultured in Dulbecco's modified Eagle's medium (DMEM)/F12 (Gibco) with 10% fetal bovine serum (FBS; Gibco) and antibiotics (1% streptomycin/penicillin; Gibco) in the incubator at 5% CO₂ at 37°C. The culture medium was replaced every 2 days. Given that nucleus pulposus cells did not change significantly before the passage 2, only used cells from passage 2 in all experiments to avoid phenotypic changes.

The senescence model of nucleus pulposus cells was simulated by adding TBHP (100 μM) to the culture medium for 24 h. To observe the effect of Met on cell senescence, cells was pretreated with metformin of different concentrations (10, 50, 100 and 200 μM) for 24 h before adding TBHP (100 μM). To observe the function of STING in nucleus pulposus cells, nucleus pulposus cells were exposed to Lv-STING or Sh-STING (see below) plus TBHP (100 μM) for 12 h.

Cell viability was determined with cell counting kit 8 (CCK-8; Dojindo Co, Kumamoto, Japan). Briefly, nucleus pulposus cells were evenly plated in 96-well plates at a cell density of 10⁵ cells/ml incubated with DMEM containing 10% FBS for 24 h at 37°C. Cells were then treated with Met or TBHP alone, or with both of them. After treatment, the cells were washed three times for 5 min with phosphate-buffered saline (PBS), followed by adding 100 μl solution that had a combination of 90 μl DMEM and 10 μl CCK-8 solution to each well. The absorbance of each group was finally measured using a microplate reader at 450 nm.

Total RNA was isolated from cells using TRIzol reagent (Invitrogen, Grand Island, NY). 1 μg of total RNA was used to synthesize cDNA (MBI Fermantas, Germany). For the quantitative real-time PCR (qPCR), 10 μl of reaction volume was used, including 5 μl of 2×SYBR Master Mix, 0.25 μl of forward primer, 0.25 μl of reverse primer, and 4.5 μl of diluted cDNA. Reactions were performed using the CFX96Real-Time PCR System (Bio-Rad Laboratories, California, USA) under the following conditions: 10 min 95°C, followed by 40 cycles of 15 s 95°C and 1 min 60°C. The cycle threshold (Ct) values were collected and normalized to β-actin levels. Relative mRNA levels of each target gene were calculated using the 2-ΔΔCt method. IL-1 and IL-6 primer sequences are as follows (Hammad et al., 2021): for IL-1β, Forward 5′-AAGGTTCGGGAGTATCTGTCTG-3′, Reverse 5′- GGGTTGGAGCTGATATGTAGCA-3′; for IL-6, Forward 5′-CGTCCTGGACAAGACCAAGT-3′, Reverse 5′-ATTGCTGTCCCGAATGTCTC-3′.

The second-generation rat nucleus pulposus cells after treatment were fixed in 2.5% glutaraldehyde overnight, then fixed in 2% samarium tetroxide for 1 h and stained with 2% uranyl acetate for 1 h, and then dehydrated with acetone in an ascending order. The sample was embedded in araldite, cut into semi-thin sections and stained with Toluidine Blue for cell localization. Finally, the sections were observed by transmission electron microscope (Hitachi, Tokyo, Japan).

When the second-generation nucleus pulposus cells were cultured under 80% fusion degree, protein from nucleus pulposus cells were lysed in radioimmunoprecipitationassay (RIPA) buffer with 1 mM PMSF and then centrifuged for 20 min at 13,680 g at 4°C. Then the protein concentration was determined by BCA protein analysis kit (Beyotime). 40 ng of purified protein were separated by SDS-PAGE and transferred to PVDF membrane (Bio-Rad, USA). After blocking with 5% non-fat milk at room temperature for 2 h, the primary antibody was incubated overnight at 4°C, and rinsed with Tris-buffered saline containing 0.05% Tween 20 (TBST) three times, and the secondary antibody was incubated for 2 h at room temperature. Then an enhanced chemiluminescence (ECL) kit was used to develop the membrane, and the gray value of the membrane was quantitatively determined using a ChemiDoc XRS+ with ImageLab3.0 software (Bio-Rad). Original blot images are shown in Fig. S3.

The second-generation nucleus pulposus cells were inoculated in six-well plate slides, treated with Met (100 μM) or DMEM for 24 h, then incubated with TBHP for 24 h, rinsed with PBS, fixed with 4% paraformaldehyde, and permeated using 0.1% Triton X-100 diluted in PBS for 10 min. The cells were blocked by 5% bovine serum albumin, rinsed by PBS, and incubated overnight with primary antibody at 4°C. The next day, the cells were incubated with Alexa Fluor^® 488-conjugated secondary antibodies (1:300) for 1 h at room temperature and labeled with DAPI for 5 min. Three regions were randomly selected on the glass slides using confocal laser scanning microscope (Leica Microsystems, Germany), and the fluorescence intensity was measured by observers were blind to the experimental treatment using ImageJ software 2.1 (Bethesda, Maryland, USA).

The senescence level was determined using a SA-β-gal staining kit (Beyotime, Shanghai, China) according to the manufacturer's instructions. The nucleus pulposus cells were inoculated in a six-well plate, the cells were fixed with 0.2% glutaraldehyde at room temperature for 15 min and rinsed by PBS three times. The cells were then stained overnight with X-gal staining solution at pH 6.0. Five visual fields were randomly selected on each slide for observation using a microscope (Olympus, Tokyo, Japan), and the percentage of SA-β-gal-positive cells were calculated and statistically analyzed.

Lenti (Lv)-STING and Sh-STING were purchased from GeneChem (Shanghai, China). For transfection, the cells were seeded at 30% confluence and infected with lentivirus at a multiplicity of infection (MOI) of 40. After 12 h of transfection, the culture medium was changed every other day. When the cell confluency reached 90% density, the transfected cells were passaged for further experiments. The sequence of Sh-STING is 5′-GATGTTCAATCAGCTACACAACTTCAAGAGAGTTGTGTAGCTGATTGAACATC-3′. The sequence of Sh-NC is 5′-GGGTGAACTCACGTCAGAATTCAAGAGATTCTGACGTGAGTTCACCC-3′. The transfection efficiency was detected by western blotting.

Adult male SD rats (200–230 g, n=36) were purchased from the Animal Center of Shanghai Chinese Academy of Sciences. According to the IVDD model induced in the previous studies (Han et al., 2008), the rats were weighed and injected intraperitoneally with 2% (w/v) pentobarbital sodium (40 mg/kg body weight). The caudal disc (Co7/8) at the experimental level was located by palpation on the upper caudal vertebrae and confirmed by counting the vertebrae in the sacral area in the test piece. We used a needle (21 g) to penetrate the full layer of the annulus fibrosus from the tail skin. To avoid an excessively deep acupuncture, the length of the needle was determined according to the size of the annulus fibrosus and nucleus pulposus, and the length measured in the preliminary experiment was ∼5 mm. All needles were kept in place for 1 min. After the operation, the rats were randomly divided into four groups: Sham operation group (Sham group), IVDD+vehicle group (IVDD group), IVDD+Met group (Met group), and IVDD+metformin+3-MA group (Met+3-MA group). The sample size of rats was predicted by PASS software (NCSS LLC, USA), with an Alpha level of 0.05 and Power of 0.9, we predicted a sample size of 7 per group. To take into account a 20% dropout rate, the final sample size of rats we chose was 9 per group. Metformin was diluted with normal saline to make the final concentration of metformin reach 20 mg/ml. The Met group was injected intraperitoneally with metformin (50 mg/kg body weight/day) and the Met+3-MA group was injected intraperitoneally with a mixture of metformin (50 mg/kg body weight/day) and 3-MA (15 mg/kg body weight/day). At the same time, rats in the Sham and IVDD group were injected with an equal volume of solvent. All rats were killed at 4th week after the disc puncture operation. During this period, all rats were free to move and were provided with food and water regularly to ensure their survival.

The rats were killed with an overdose of 4% pentobarbital sodium, and the tails were collected. Rat caudal intervertebral disc tissue was isolated, fixed, decalcified and embedded in paraffin. After cutting the tissue into 5 μm slices, the sections were stained with Hematoxylin and Eosin (H&E) and Safranin O-fast green (S-O) and observed under a microscope. The cells and morphology of nucleus pulposus and annulus fibrosus were assessed under a microscope by another group of experienced histological researchers who were blind to the experimental conditions and were evaluated in accordance with the grading criteria described previously (Mao et al., 2011). The histological score of a normal intervertebral disc is 5 points, a score for moderate degenerative intervertebral disc is 6–11 points, and the score for severe degenerative intervertebral disc is 12–15 points.

X-ray images of all rats were obtained 4 weeks after operation. The height of intervertebral disc was measured using ImageJ (NIH) software and expressed as the disc height index (DHI) (Zhang et al., 2009). The DHI changes of punctured IVDD were expressed as DHI% [DHI%=(post-punctured DHI/pre-punctured DHI)×100%].

The tissue was fixed with 4% paraformaldehyde, decalcified and embedded in paraffin. Then, it was sliced, dewaxed and rehydrated. The activity of endogenous peroxidase was assessed by incubation with 3% H₂O₂ for 10 min, and non-specific binding sites were blocked for 30 min with 5% bovine serum albumin. The slices were then incubated with the primary antibody at 4°C overnight. On the second day, the secondary antibody was incubated with the samples at 4°C for 1 h. Images were analyzed using Image-ProPlus6.0 version (media cybernetics) software.

These experiments were performed at least three times. The results are presented as the mean±s.d. and statistically analyzed using GraphPad Prism 5.0 software (GraphPad software, San Diego, CA, USA). One-way analysis of variance (ANOVA) and Tukey's post hoc test were used for comparison of two groups. A P<0.05 was considered statistically significant.

The peer review history is available online at https://journals.biologists.com/jcs/article-lookup/doi/10.1242/jcs.259738.