All the experimental protocols were approved by the Ethics Committee of Nanjing Drum Tower Hospital, the Affiliated Hospital of Nanjing University Medical School. With the informed consent of the patients, normal NP tissues were obtained from eight patients (Three males and five females, aged 12–23 years, mean age 17.8 years) who underwent surgery for congenital scoliosis. Degenerative NP tissues were collected from eight patients (Three males and five females, aged 40–65 years, mean age 53.7 years) who underwent posterior spinal fusion surgery. The human NP tissue samples were classified according to their degenerative grades (Song et al., 2018). After cutting into pieces, human NP tissues were enzymatically digested in 0.25% trypsin (Gibco; 15050065) and 0.2% type II collagenase (Gibco; 17101015) for 3 h. Following filtration and washing in PBS, the suspension was centrifuged. Isolated cells were grown in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS) (Gibco; 10099141) and 1% penicillin-streptomycin (Invitrogen; 10378016). The medium was replaced twice a week, and cells at 3–5 passage were used for the subsequent experiments.

NP tissue and NP cells were lysed in the radioimmunoprecipitation assay (RIPA) buffer. The protein concentration was determined using a BCA kit (Thermo Scientific). 5% skim milk (Solarbio, D8340) was added to TBST (Solarbio, T1081) for 60 min to block membranes. After blocking, the membranes were incubated overnight at 4°C with the corresponding primary antibodies of Ace-tubulin (1:800; Cell Signaling Technology, 5335s), Collagen 2 (1:500; Proteintech, 28459-1-AP), SOX9 (1:1,000; Cell Signaling Technology, 82,630), Smad3 (1:1,000; Cell Signaling Technology, 9,523), YAP (1:1,000; Cell Signaling Technology, 14,074), p-YAP (1:1,000; Cell Signaling Technology, 13,008), TAZ (1:1,000; Cell Signaling Technology, 72,804), LATS1 (1:1,000; Cell Signaling Technology, 3,477), LATS2 (1:1,000; Cell Signaling Technology, 5,888), and GAPDH (1:1,000; Cell Signaling Technology, 5,174). Then, HP-linked secondary antibodies (1:5,000; Cell Signaling Technology, 7,074) were applied. The immunoblotting was detected by UVP ChemiDoc-It Imaging System (UVP, CA, USA) with the enhanced chemiluminescence detection kit (Thermo Fisher; 34,580) added to the membranes. Each blot was measured using ImageJ software for its integrated density.

A total RNA extract was prepared with Trizol (Invitrogen, 15,596,018). Following reverse transcription, the RNA was converted into cDNA using PrimeScript RT Master Mix Kit (TaKaRa, RR036A). Real-time polymerase chain reactions (RT-PCRs) were carried out using the TaKaRa SYBR Premix Ex Taq (RR420A) and the ABI StepOnePlus Real-Time PCR System (Applied Biosystems, CA, USA). By normalizing by GAPDH, fold changes of interested mRNAs were calculated. The primers were designed as follows:

Col2: 5′-AAG​GGA​CAC​CGA​GGT​TTC​ACT​GG-3′, 5′-GGG​CCT​GTT​TCT​CCT​GAG​CGT-3′; SOX9: 5′-AGC​GAC​AAC​TTT​ACC​AG-3′; 5′-GGA​AAA​CAG​AGA​ACG​AAA​C-3; GAPDH: ACA​GCA​ACA​GGG​TGG​TGG​AC-3′, 5′- TTT​GAG​GGT​ACA​GCG​AAC​TT-3′; YAP: 5′-GCT​AGA​TAA​AGA​AAG​CTT​T-3′, 5′-CCA​ATA​GTT​CCG​ATC​CCT​T-3′; TAZ: 5′- GAA​TCA​GCC​TCT​GAA​TCA​T-3′, 5′-GTC​TGA​AGA​TCT​GAT​CCC​T-3′; LATS1: 5′-CCACCCTAC CCAAAACATCTG-3′, 5′- CGC​TGC​TGA​TGA​GAT​TTG​AGT​AC-3′; LATS2: 5′-GCC​AAA​GAC​TTT​TCC​TGC​CA-3′, 5′- TCT​TTG​CTC​CCC​AGG​ACT​TT-3′.

Using Ribobio’s EdU Reagent Kit (Guangzhou, China), the manufacturer’s instructions were followed to investigate cell proliferation. Following treatment with nocodazole and docetaxel for 3 days, the degenerative NPCs were exposed to 10 μM EdU for 2 h. Fluorescence images were captured using a fluorescence microscope (Zeiss, Heidelberg, Germany).

In accordance with the manufacturer’s instructions, the viability of NPCs was evaluated by using CCK8 assay ((Solarbio). We subsequently incubated each well for 2 h at 37°C with DMEM/F12 containing 10% (v/v) CCK-8 solution after washing with PBS. Microplate readers (Thermo Scientific, Logan, UT, United States) were used to measure absorbance at 450 nm.

Analyses of immunofluorescence were performed as previously described (Wu et al., 2017). We fixed cells and tissues attached to slides with 4% paraformaldehyde and permeabilized them with 0.5% Triton X-100 in PBS. After washing the slides in PBS and blocking with 2% bovine serum albumin (BSA) in PBS for 2 h at 37°C, the slides were incubated with primary antibodies against: Ace-tubulin (1:800), Col2 (1:500), or YAP (1:200). A second goat anti-rabbit antibody (Boster) was then applied at 37°C for 2 h after the slides. A DAPI staining solution (Beyotime, Nantong, China) was applied to the nuclei for 5 min before images were taken under an Olympus BX53 microscope (Zeiss, Heidelberg, Germany). Using ImageJ software (version 1.8.0), five randomly selected microscopic fields per slide were analyzed to determine the average gray value.

According to the manufacturer’s instructions, immunohistochemical staining and immunofluorescent analysis were performed. The serial sections were incubated overnight at 4°C with primary antibodies against Col2 (1:500; ab34712-Abcam) and Ace-tubulin (1:800; 5335s-Cell Signaling Technology). The slides were stained immunohistochemically with secondary antibodies conjugated with HRP, incubated at 37°C for 1 h. Photographs of sections were taken using a Zeiss fluorescence microscope (Zeiss, Heidelberg, Germany).

The siRNA sequence was designed and synthesized to target the human YAP1 gene: 5-GGU​GAU​ACU​AUC​AAC​CAA​ATT-3 (Hippobio). We seeded NPCs in 6-well plates and grew them to a confluence of about 70%. Transfecting the cells with either 50 nM siRNA-YAP or a negative control, as directed by the manufacturer, took place for 12 h in Lipofectamine 3,000 (Thermo Fisher Scientific).

In accordance with the guidelines of the Institutional Animal Care and Use Committee at Nanjing University Medical School, all experiments were reviewed and approved by the committee. In the study, SD-Rats (10 weeks) were anesthetized with 10% chloral hydrate (0.3 mL/100 g) injected intraperitoneally. After placing the animals in a prone position, ethanol was applied to the tail skin for disinfection. A 21 G needle with a stopper was used to percutaneously puncture the discs Co6-7, Co7-8, and Co8-9 to the depth of 5 mm, after which the needle was rotated 360 degrees and held for 30 s. Four weeks after acupuncture, the caudal intervertebral discs of rats were examined by microCT and MRI. This study used MicroCT instead of X-ray to calculate the intervertebral disc height index (DHI), which is conducive to more intuitive indication of the changes in intervertebral disc height. The calculation method of DHI is as previously reported (Han et al., 2008). Degeneration grade of rat intervertebral disc was calculated and counted by MRI. We performed three IVDs on rats (Co6/7, Co7/8, and Co8/9) for the different treatments in order to evaluate the therapeutic effectiveness of microtubule stabilization for IDD. PBS, Nocodazole (3 μg/mL), and Docataxel (30 μg/mL) were administered intradiscally at the same total injection volume (2 μL) using 33-gauge needles (Hamilton, Benade, Switzerland) for all three IVDs (Co6/7, Co7/8, and Co8/9). During the 4 weeks treatment period, injections were given every week. After 4 weeks of treatment, microCT and MRI were performed again, and the intervertebral disc tissue of rats was stored in formalin for subsequent tissue staining.

Using CO2 inhalation, all animals were sacrificed and the vertebral segments were dissected from the coccygeal segments 6–7, 7–8, and 8–9. Specimens were fixed in 10% formalin for 20 h at room temperature, then decalcified in an ethylenediaminetetraacetic acid (EDTA)-glycerol solution for 4 weeks. An embedding of decalcified tissue in paraffin, sectioning sagittally to a thickness of 5 μm. Hematoxylin-eosin was used to stain the slides of each disc, and a histological grading system was used to grade them (Han et al., 2008). Meanwhile, Masson staining and COL2 immunohistochemical staining was performed on the sectioned tissues.

The means and standard deviations were calculated for all data. In GraphPad Prism 8, one-way ANOVA was used for the statistical analysis. In this study, statistical significance was defined as a p < 0.05.

In order to harvest the degenerated NP cells, we collected human NP tissue samples from patients who underwent posterior transforaminal lumbar interbody fusion surgery due to lumbar spinal stenosis. The degenerated intervertebral discs all demonstrated low signal intensity (Pfirrmann grade Ⅳ or Ⅴ) in T2 Weight images of Magnetic Resonance (Figure 1A). The normal NP tissues in this study were derived from patients with congenital scoliosis who underwent posterior hemivertebra resection surgery (Figure 1A). The histological characteristics of degenerated NP tissues and normal NP tissues were evaluated using HE staining, Safranin O-fast green staining and Masson staining, to specify the different degeneration grade and collagen expression (Figure 1B). According to the HE staining, Safranin O-fast green staining and Masson staining, decreased number of NP cells and decreased expression of collagen were shown in the degenerated NP tissue. Since loss of COL2 presented a very important phenotype in the progression of IDD, we then compared the expression of COL2 in degenerated NP cells and normal NP cells. The lower expression of COL2 in degenerated NP cells was confirmed by Western blot, Cell immunofluorescence and Immunohistochemistry (Figures 1C, D, F).

Microtubules are involved in secretory transport mechanisms in collagen and proteoglycan synthesis. Thus we explored the stability of microtubules in degenerated NP cells. As previously reported, acetylated tubulin (Ace-tubulin) was used as a marker and quantitative index for microtubule stabilization since tubulin acetylation was found mainly on stable microtubules resistant to depolymerization (Westermann and Weber, 2003). Tubulin acetylation is a consequence of microtubule stability. Here, in the results of Western blot, cell immunofluorescence and immunohistochemistry, the expression of Ace-Tubulin was enhanced in degenerated NP cells (Figures 1C, E, F). For the first time, we observed increased Ace-tubulin in the NP cells, representing enhanced microtubule stability.

Since the microtubule stability may regulate the cellular function of NP cells, we then investigated the effect of upregulation and the downregulation of microtubule stability on NP cells. Docetaxel, the microtubule stabilizer, and nocodazole, the microtubule depolymerizer, were used for the stabilization and destabilization of the microtubules, respectively. Regarding the cytotoxicity induced by high concentrations of docetaxel and nocodazole, we evaluated the proliferation and cell viability of NPCs treated with different concentrations of docetaxel and nocodazole using the CCK-8 and EdU staining. The EdU staining showed that the proliferation of NP cells was not significantly affected by docetaxel (0.5, 1, and 2.5 nM) or nocodazole (5, 10, and 25 nM), but was significantly inhibited by docetaxel at higher concentrations (5 and 10 nM) and enhanced by nocodazole at higher concentrations (50 and 100 nM) (Figures 2A, B). NP cells were assessed with the CCK-8 assay for cell viability (Figure 2C). Based on the results, 25 nM nocodazole slightly enhanced NP cells proliferation, while docetaxel within 2.5 nM had no significant effect on cell viability. The above results indicated that 2.5 nM docetaxel and 25 nM nocodazole were safe and efficacious for further investigation.

After treating NP cells with docetaxel and nocodazole for 1 week, the expression of Ace-tubulin was enhanced in NP cells treated with docetaxel compared to the untreated group, but no difference was found in NP cells treated with nocodazole, according to the results of Western blot and cell immunofluorescence (Figures 2D, E, G). As the MT stability increased using docetaxel, the results of Western blot, Q-PCR and cell immunofluorescence showed upregulated expression of COL2 (Figures 2D, F, H). Moreover, as the upstream regulators of COL2, the expression of SOX9 and Smad3 were also upregulated as microtubules stabilization increased (Figures 2D, E). As shown above, MT stabilization could promote the expression of COL2 in NP cells.

YAP/TAZ plays a critical role in response to mechanical stress, perturbation of the cytoskeleton and matrix remodeling (Calvo et al., 2013). YAP/TAZ nuclear accumulation is a key determinant of their function by inducing gene expression through interactions with partner proteins; while YAP/TAZ could also be phosphorylated and inactivated, leading to their cytoplasmic retention and degradation. We investigated the expression and phosphorylation level of YAP/TAZ in NP cells to determine whether MT stabilization was involved in the expression and activity of YAP. The Western blot analysis showed that both total YAP and phosphorylated YAP were significantly decreased in NP cell after treatment of docetaxel, which was confirmed by Q-PCR (Figures 3A–C). Moreover, the proportion of phosphorylated YAP in total YAP ascended after treatment of docetaxel compared to the untreated group (Figures 3A, B), indicating that MT stabilization not only inhibits the expression of YAP, but promoting phosphorylation of YAP. MT destabilization by nocodazole treatment had no significant effect on total YAP and phosphorylated YAP in NP cells. The expression of TAZ decreased in docetaxel group and unchanged in nocodazole group, which was similar to that of YAP. The immunofluorescence results also confirmed that the docetaxel treatment significantly decreased total YAP, while no noteworthy change in YAP was observed after nocodazole treatment (Figure 3D). The above results showed that MT stabilization inhibited the expression and activity of YAP/TAZ in nucleus pulposus cells.

Hippo pathway regulates YAP/TAZ primarily through LATS kinases (LATS1 and LATS2)-mediated phosphorylation. Moreover, Hippo pathway is regulated by actin cytoskeletal factors, including capping proteins and cofilin (Gaspar and Tapon, 2014; Meng et al., 2016). A recent study also suggested that Patronin, a microtubule stabilizer, interacted with Hippo signaling in Drosophila (Yang and Choi, 2021). Therefore, we further investigated that whether the inhibition of YAP/TAZ by MT stabilization was mediated by Hippo pathway in NP cells. Our results in NP cells revealed that the docetaxel treatment significantly increased the expression of LATS1 and LATS2 (Figures 4A–C), which was consistent with previous findings that the components of the Hippo pathway promoted ciliogenesis and MT stabilization (Kim et al., 2014). In addition, the expression of LATS1/2 was not affected when NP cells were treated with nocodazole. Meanwhile, the enhanced expression of Ace-tubulin and inhibited expression of YAP was also found after docetaxel treatment by Western blot analysis. The above results illustrated that MT stabilization inhibited the expression and activity of YAP/TAZ via regulating Hippo pathway.

To clarify whether YAP plays a role in the up-regulated COL2 expression promoted by MT stabilization, YAP was knocked down by small interfering RNA (siRNA). As the total and phosphorylated YAP both decreased following YAP siRNA transfection, the nucleus pulposus related proteins, including COL2, SOX9 and Smad3, were all upregulated with and without docetaxel treatment; while the expression of LATS1, LATS2 and Ace-tubulin were not significantly altered (Figures 5A, B). Moreover, Lysophosphatidic acid (LPA) was utilized to promote the nucleus re-localization of YAP in the NP cells, which would up-regulate the activity of YAP (Yu et al., 2012). The results revealed that the increased expression of COL2, SOX9 and Smad3 by docetaxel were diminished with concurrent treatment of LPA, while the LPA treatment alone to the NP cells also significantly decreased the expression of COL2, SOX9 and Smad3 (Figures 6A–C). Therefore, the synthesis effect of COL2 induced by MT stabilization was via the regulation on YAP.

To further investigate the therapeutic effects of MT stabilization on IDD, we used a classic rat caudal intervertebral disc acupuncture animal model for this study. PBS, nocodazole and docetaxel were injected into three consecutive independent intervertebral discs of each rat 4 weeks after acupuncture. MRI and microCT examinations of rat IVD were performed at 4 weeks after the modeling and 4 weeks after the treatment to determine the degree of degeneration. A percent disc height index (%DHI) was calculated based on the microCT results. With a low percent DHI in all rats 4 weeks after needle puncture, the intervertebral disc showed intervertebral space stenosis or collapse, indicating degenerative changes of IVD (Figures 7A, B). After 4 weeks of treatment with docetaxel and nocodazole, DHI in docetaxel group recovered significantly, while DHI in control group and nocodazole group did not increase significantly (Figure 7A). In addition, the Pfirrmann score was derived from MRI results and used to assess the degree of IDD. Four weeks after needle puncture, the Pfirrmann scores of IVD in all rats showed no significant difference. After 4 weeks of treatment with docetaxel and nocazole, Pfirrmann scores were significantly lower in the docetaxel group compared to the PBS group and the nocodazole group. Intervertebral discs were collected for HE staining, Masson staining, and COL2 immunohistochemical staining after treating rats for 4 weeks. The results showed that the Docetaxel group had a significantly better degenerative score than the PBS group and the Nocodazole group, and the expression of COL2 was also significantly increased compared with PBS group and Nocodazole group (Figure 7C).