Needle-scalpel therapy inhibits the apoptosis of nucleus pulposus cells via the SDF-1/CXCR4 axis in a rat degenerative cervical intervertebral disc model

Animals and groups

Healthy male clean-grade Sprague–Dawley rats (200 ± 20 g) provided by Jinan Pengyue Experimental Animal Breeding (Shandong, China; certificate no. SCXK Lu 2019–0003) underwent housing at 23°C ± 2°C under a 12-h photoperiod with ad libitum food and water. Studies involving animals followed the Guide for the Care and Use of Laboratory Animals released by the United States National Institutes of Health. The animal ethics committee of the First Affiliated Hospital of Shandong First Medical University and Shandong Provincial Qianfoshan Hospital approved the protocol of this study ((2019) no. S177).

During model establishment, 5 rats died of wound infection or unknown causes. Finally, 35 rats completed 3 months of modeling. CS model rats were then randomized to the Model (CS model), M+Ns (CS model + Ns treatment), and M+Ms (CS model + celecoxib medicaments treatment) groups. All treatments lasted 4 weeks.

CS model establishment

Forty rats were used for CS model establishment as described in a previous study [27]. Briefly, following a 1-week adaptive housing, the animals underwent anesthesia by intraperitoneally injecting 3% pentobarbitone sodium at 30 mg/kg. After skin disinfection with Anerdian, a posterior cervical median incision exposed the tissue from the atlanto-occipital joint to the second thoracic vertebra. This was followed by a separation of fascia, platysma muscle, cervical trapezius, and rhomboideus, with muscle stripping to expose its deep layer. After fully exposing cervical vertebrae, the C2–C7 supraspinous and interspinous ligaments were snipped. Next, bilateral sacrospinal muscles and skin were sutured. Two random model animals underwent euthanasia 12 weeks following model establishment. C4/5 and C5/6 intervertebral discs underwent fixation. The sham group was treated by a sham operation, with only the skin along the nuchal incised and sutured.

Hematoxylin and eosin (H&E) staining was carried out to detect the histopathological features of the cervical intervertebral disc by light microscopy. The intervertebral discs in the normal and sham groups had normal morphological and structural features, but those of model rats were altered, with enhanced ECM production and shrinkage and cluster aggregation of the NP (Figure 1A–1C). Meanwhile, model rats had substantially reduced neck-skin response thresholds to mechanical stimuli applied using von Frey filaments 12 weeks after operation (Figure 1D). These observations indicated a successful CS model establishment.

Figure 1. Histomorphology of the cervical intervertebral disc and the neck-skin response thresholds to the mechanical stimuli 12 weeks after modeling. (A–C) H&E staining of cervical intervertebral disc in different groups 12 weeks after modeling, note: Scale bar = 100 µm. (D) The neck-skin response thresholds to the mechanical stimuli applied by von Frey filaments in different groups 12 weeks after modeling, note: values are means ± SEMs, n = 3 per group, ^**p < 0.01.

Interventions and sample preparation

Twelve weeks post-modeling, treatments were administered. The normal, sham, and model groups underwent oral administration of pure water at 2 mL/day for 28 days. The M+Ms group was intragastrically administered celecoxib (Celebrex; Pfizer, USA) in pure water at 10 mg/kg/day for 28 days. The M+Ns group received pure water at 2 mL/day for 28 days and administered Ns.

Four points were selected as fixed intervention points, including the superior nuchal line (muscle insertion such as the trapezius, semispinalis capitis, and splenius capitis) and the internal superior angle of the scapula (muscle insertion of the levator scapulae). Additionally, 3–4 knot nodes of extensor/flexor muscle fibers around the cervical vertebra were chosen considering individual condition. Following disinfection, an Ns (HZ series: 0.4 × 40 mm, Beijing Outstanding Huayou Medical Instrument) was pierced to release the above points as described in a previous work [23]. After receiving the indicated treatment for 28 days, magnetic resonance imaging (MRI) examination of the cervical vertebra and neck pain-related behavior tests were performed before rats were humanely euthanized.

Next, C5/6 cervical intervertebral disc samples were obtained and submitted to fixation with 4% paraformaldehyde and 2.5% glutaraldehyde for transmission electron microscopy and histological assessment. C4/5 disc tissue samples were obtained and kept at −80°C for immunoblot and quantitative real-time PCR.

von Frey test of mechanical allodynia

The pain-associated behavior of CS rats was assessed at 12 and 4 weeks after modeling and treatment, respectively, in a quiet environment at 25°C with low-intensity light from 9 AM to 6 PM. The animal’s neck and back were shaved with an HC1066 hair clipper (Philips, Amsterdam, Netherlands). Each rat was acclimatized to a black wire box for 30 min. Withdrawal responses to mechanical stimuli were examined with calibrated von Frey filaments applied from above the cage to the skin 2 mm away from the wound or the same area in control animals. The cutoff was 15 g, and the first filament evoking ≥3 responses (5 applications totally) was considered the threshold. In case of no response, 15 g was recorded. A response was defined as previously suggested [38].

MRI

MRI examination was performed 4 weeks after treatment on a Bruker BioSpec 94/70USR scanner (Bruker BioSpin MRI GmbH, Ettlingen, Germany) with an animal surface coil to detect alterations in the central sagittal image of the cervical intervertebral disc on T2-weighted (T2W) scans. Disc degeneration levels were assessed based on the Pfirrmann MRI grading scale (Table 1). The image acquisition parameters were: repetition/echo time, 2500/33 ms; field of view, 40 × 40 mm; acquisition matrix, 256 × 256; number of slices, 5; slice thickness, 0.4 mm. MRI data were independently assessed by two experienced radiologists in a blinded manner.

H&E staining of intervertebral disks

Intervertebral disks from rats underwent fixation with 10% formalin, followed by a 2-week decalcification with EDTA. After deparaffinization and rehydration, the sections were submitted to H&E staining. A microscope was used for analysis, based on histological grading criteria for intervertebral discs (Table 2).

Ultrastructural assessment by electron microscopy

Each intervertebral disc was excised without the vertebral body for transmission electron microscopy (TEM). Then, intervertebral discs underwent a 2-h fixation with 2% glutaraldehyde, and phosphate-buffered saline (PBS) was used to remove any excess fixative. Next, fixation with 1% osmium tetrachloride was performed, and graded concentrations of ethyl alcohol were employed to dehydrate the specimens. The samples were then embedded in epoxy resin and underwent staining with toluidine blue. After locating target areas in these sections, ultrathin sections were obtained for double staining with uranyl acetate and lead citrate before TEM.

Immunoblot

Total protein extraction from intervertebral disc samples utilized chilled RIPA buffer with phenylmethanesulfonyl fluoride (1 mM). Protein quantitation was carried out with the BCA Protein Assay Kit (P0012, Beyotime Biotechnology, Shanghai, China). Equal amounts of total protein (30 μg) were resolved by SDS-PAGE, with subsequent transfer onto a polyvinylidene fluoride (PVDF) membrane. After a 2-h blocking with 5% nonfat milk in TBST (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 0.05% Tween-20) at ambient, overnight incubation was carried out with primary antibodies targeting SDF-1 (1:2500, ab9797, Abcam, UK), CXCR4 (1:100, ab124824, Abcam), Bcl-2 (1:500, ab196495, Abcam), Bax (1:1000, ab32503, Abcam), type II collagen (1:1000, ab188570, Abcam) and GAPDH (1:2000, E-AB-20059, Elabscience, Wuhan, China) at 4°C. Next, a 2-h incubation was performed with HRP-linked goat antirabbit IgG (H + L) secondary antibodies (1:5000, EF0002, SparkJade, Shandong, China) at ambient. The enhanced chemiluminescence kit was employed for detection, and quantitation used ImageJ.

Quantitative real-time PCR

Total RNA extraction from cervical intervertebral disc employed the SPARKeasy Bone Tissue RNA Kit (SparkJade, AC1301, China) as directed by the manufacturer. RNA quantitation employed a Merinton^®SMA4000 spectrophotometer (Merinton, Beijing, China) and reverse transcription employed a Reverse Transcription kit (SparkJade, AG0304, China) as instructed by the manufacturer. qPCR was carried out with 2 × SYBR Green qPCR Mix (SparkJade, AH0104, China). The 2^−ΔΔCt method was used for data analysis, with GAPDH as a reference gene. Table 3 describes the primers (Accurate Biotechnology, China) used for qPCR.

Immunohistochemistry for SDF-1, CXCR4, Bcl-2, Bax and Type II collagen detection in the intervertebral disc

Paraffin-embedded sections underwent dewaxing with xylene and hydration using an ethanol gradient. After deparaffinization and rehydration, incubation was performed with 3% hydrogen peroxide, followed by antigen retrieval with citrate buffer for 15 min at 92–98°C. Next, samples underwent overnight treatment at 4°C with antibodies targeting SDF-1 (1:250, ab9797), CXCR4 (1:500, ab124824), Bcl-2 (1:100, ab196495), Bax (1:250, ab32503) and collagen type II (1:200, ab34712), all from Abcam. After rinsing, a 1-h incubation was performed with horseradish peroxidase-linked secondary antibodies at ambient. The DAB kit (Zhongshan Golden Bridge Biotechnology, Beijing, China) was employed for development, followed by hematoxylin counterstaining. A BX60 microscope (Olympus, Tokyo, Japan) was utilized for imaging, and data analysis was performed with Image-Pro Plus.

TdT-mediated dUTP nick-end labeling (TUNEL) staining

The TUNEL assay was performed to examine apoptosis in NPCs using the DAB (SA-HRP) TUNEL Cell Apoptosis Detection Kit (Servicebio, G1507, China). Paraffin-embedded sections underwent deparaffinization and rehydration with xylene and graded ethanol, respectively. After a 20-min treatment with proteinase K at ambient, specimens were incubated with 50 μL of TUNEL reaction mixture, shielded from light for 1 h at ambient. Next, a 30-min incubation was carried out with converter peroxidase (50 μL) away from light. After staining with 50 μL of 3,3′-diaminobenzidine for 10 min at ambient, hematoxylin counterstaining was performed for 3 min. Multiple high-power fields were randomly imaged with a microscope (Olympus) to determine the percentage of TUNEL-positive NPCs (yellowish-brown signals) relative to the total number of NPCs.

Statistical analysis

SPSS 23.0 (SPSS, USA) and GraphPad Prism 8.0.1 (GraphPad, USA) were employed for data analysis. Data are mean ± standard error of the mean (SEM). Group pairs were compared by unpaired Student’s t-test, while multiple groups were compared by one-way analysis of variance (ANOVA) or two-way repeated-measures ANOVA, followed by post hoc Tukey test. The Kruskal–Wallis nonparametric test was performed to compare data with skewed distribution. P < 0.05 indicated statistical significance.

Effect of Ns on the histological features of the intervertebral disc

Following 28 days of treatment, the collected cervical intervertebral disc samples underwent H&E staining and were imaged at 100×. As shown in Figure 2A, intervertebral discs in the normal control and sham groups exhibited regularly arranged annulus fibrosus and NP with a clearly delineated boundary between the annulus fibrosus and NP. The model groups, however, exhibited significant alterations, with scattered annulus fibrosus, clustered cellularity of the NP, and no distinguishable boundary between annulus fibrosus and NP. Intervertebral discs and annulus fibrosus in the M+Ns group were relatively regular, while slightly shrunken NP and visible cartilage endplates were detected. The M+Ms group had irregular intervertebral discs, with scattered annulus fibrosus and mixed cellularity, a normal appearance of some cell clusters of the NP, and a less clear boundary with loss of annular-nuclear demarcation.

Figure 2. The therapy of Ns affected intervertebral disc degeneration in the intervertebral disc of CS rats. (A) H&E staining of cervical intervertebral disc in different groups 4 weeks after intervening, note: scale bar = 100 um. (B) Transmission electron microscope of the nucleus pulposus in different groups 4 weeks after intervening, note: scale bar = 2 um. (C) The neck-skin response thresholds to the mechanical stimuli applied by von Frey filaments in different groups 4 weeks after intervening. (D) Pfirrmann MRI Grades in different groups 4 weeks after intervening. Note: values are means ± SEMs, n = 5 per group, ^*p < 0.05, ^**p < 0.01.

Histological grading criteria for the intervertebral disc were compared among groups based on staining signals, reflecting disc degeneration degree. The values obtained were markedly lower in the normal and sham groups than in the remaining groups (p < 0.01). The model group showed elevated values in comparison with the M+Ns and M+Ms groups (p < 0.01), with the M+Ms group having elevated values compared with the M+Ns group (p < 0.01).

Effect of Ns on NP ultrastructure

As shown in Figure 2B, after 28 days of treatment, cells in the normal and sham groups were round or oval, with a visible plasma membrane and a centrally placed nucleus with a smooth and whole nuclear membrane. Nuclei had an even distribution, with no overt heterochromatin. Furthermore, multiple cellular organelles were observed, e.g., the rough endoplasmic reticulum, mitochondria, matrix vesicles, and vacuoles. Cells in the model group exhibited irregular shapes, with an uneven nuclear membrane structure; heterochromatin had higher abundance compared with euchromatin, forming a “crest” that surrounded the nuclear membrane. The cytoplasm showed shrinkage, increased intranuclear electron density, and a decreased number of organelles. Mitochondria appeared unclear and cavitated, the rough endoplasmic reticulum showed overt swelling, the numbers of lysosomes and vacuoles increased, and lipid droplets were detected. Moreover, some cells around the acupoints of M+Ns treated rats were mainly oval or irregularly shaped. The plasma membrane was round, while the nuclear membrane was whole but slightly uneven. The intranuclei showed an even distribution, with heterochromatin molecules surrounding the nuclear membrane and organelles (e.g., the rough endoplasmic reticulum and mitochondria) showing mild swelling. However, there were some lysosomes and vacuoles. Cells from rats of the M+Ms group exhibited irregular shapes, an uneven nuclear membrane structure, increased heterochromatin, shrunken cytoplasm, fewer organelles, swollen rough endoplasmic reticulum and mitochondria, and thickened periplasmic compartment.

Effect of Ns on the response threshold to mechanical stimuli induced by von Frey filaments

Neck-skin response thresholds non-significantly differed between the normal and sham groups at 4 weeks after intervention (p > 0.05). In comparison with sham animals, the model group exhibited reduced response threshold to mechanical stimuli (p < 0.01). Response thresholds to mechanical stimuli increased significantly (p < 0.01) in the M+Ns and M+Ms groups versus the model group, whereas the M+Ns and M+Ms groups had similar values (Figure 2C).

Effect of Ns on Pfirrmann MRI grades

As shown in Figure 2D, MRI scans revealed the whole configuration of the cervical vertebra. In the normal control and sham groups, T2W images showed homogenous and bright white intervertebral disc structure, with clear boundaries of the nucleus and annulus, hyperintense signal intensity, and normal height. In the model group, the T2W sequence showed inhomogeneous and black intervertebral disc structure, with unclear boundaries of the nucleus and annulus, hypointense signal intensity, and collapsed height. In the M+Ns group, T2W scans showed inhomogeneous and gray intervertebral disc structure, with unclear boundaries of the nucleus and annulus, intermediate signal intensity, and slightly decreased height. In the M+Ms group, the T2W sequence showed inhomogeneous and gray to black intervertebral disc structure, with unclear boundaries of the nucleus and annulus, intermediate-to-hypointense signal intensity, and moderately decreased height.

We compared Pfirrmann classes among groups according to cervical intervertebral degeneration level. The grades were markedly lower in the normal and sham groups in comparison with the model group (p < 0.01). Meanwhile, the model group had higher grades compared with the M+Ns and M+Ms groups (p < 0.01), with the M+Ms group showing higher values versus the M+Ns group (p < 0.01).

Ns therapy regulates SDF-1, CXCR4, Bcl-2, and Bax in the intervertebral disc of CS rats

Whether these treatments affect apoptosis in NPCs by regulating SDF-1/CXCR4, Bcl-2/Bax, and type II collagen was examined next. Western blotting revealed markedly increased SDF-1, CXCR4, and Bax protein amounts (p < 0.01) in the model group versus sham rats, while these proteins were starkly downregulated (p < 0.05) in the M+Ns and M+Ms groups relative to the model group. However, the mRNA expression levels of only CXCR4 were substantially decreased (p < 0.01) in the M+Ms group versus the model group. A comparison of both therapies revealed SDF-1, CXCR4, and Bax were markedly upregulated in the M+Ms group (p < 0.05) versus the M+Ns group; therefore, the Ns therapy showed a better tendency to inhibit SDF-1, CXCR4, and Bax upregulation. Conversely, Bcl-2, Bcl-2/Bax, and type II collagen expression levels starkly decreased (p < 0.01) in the model group versus sham animals, while the M+Ns group showed remarkably increased levels versus the model group (p < 0.01). Most notably, Bcl-2/Bax and type II collagen expression levels were similar in the M+Ms and model groups (p > 0.05). Comparing both therapies, starkly reduced Bcl-2/Bax and type II collagen expression levels were detected for the M+Ms group in comparison with the M+Ns group (p < 0.05), while Bcl-2 expression levels did not differ between the two groups. Therefore, the Ns therapy could better inhibit the downregulation of Bcl-2, Bcl-2/Bax, and type II collagen (Figure 3A).

Figure 3. The therapy of Ns affected apoptosis of nucleus pulposus cells mediated by specifically regulating SDF-1/CXCR4 signal axis in the intervertebral disc of CS rats. (A) Western blot assay of SDF-1/CXCR4, Bcl-2/Bax and type II collagen in different groups 4 weeks after intervening. (B) Real-time PCR analysis of SDF-1/CXCR4, Bcl-2/Bax and type II collagen in different groups 4 weeks after intervening. (C) Immunohistochemical staining and optical density value of SDF-1/CXCR4, Bcl-2/Bax and type II collagen in different groups 4 weeks after intervening. (D) TUNEL staining to confirm the effect of 4 weeks of intervention on apoptosis of nucleus pulposus cells in different groups. Note: scale bar = 5 um, values are means ± SEMs, n = 5 per group, ^*p < 0.05, ^**p < 0.01.

Quantitative real-time PCR yielded similar results. SDF-1, CXCR4, and Bax mRNA amounts were markedly increased (p < 0.01) in the model group versus sham animals, while the M+Ns group had starkly decreased amounts versus the model group (p < 0.01). However, the mRNA amounts of only CXCR4 were decreased significantly (p < 0.01) in the M+Ms group compared with the model group. Comparing the M+Ns and M+Ms groups, SDF-1, CXCR4, and Bax mRNA amounts were decreased in the M+Ns group (p < 0.05). Thus, the Ns therapy could better inhibit the increases of SDF-1, CXCR4, and Bax mRNA amounts. Conversely, Bcl-2, Bcl-2/Bax, and type II collagen mRNA amounts were starkly decreased (p < 0.01) in the model group versus sham animals, while the M+Ns group had remarkably increased amounts versus the model group (p < 0.01). However, Bcl-2/Bax and type II collagen mRNA amounts were similar in the M+Ms and model groups (p > 0.05). Comparing both therapies, Bcl-2/Bax and type II collagen mRNA amounts were starkly reduced in the M+Ms group versus the M+Ns group (p < 0.05), whereas the mRNA expression levels of Bcl-2 did not differ between the two groups. Therefore, the Ns therapy could better inhibit the decreases of Bcl-2, Bcl-2/Bax, and type II collagen mRNA levels (Figure 3B).

Immunohistochemistry staining further confirmed these findings. A comparison of the M+Ns and model groups corroborated western blot and quantitative real-time PCR data (Figure 3C). Specifically, SDF-1, CXR4, and Bax expression levels were markedly decreased in the M+Ns group compared with the model group (p < 0.01). Meanwhile, SDF-1, CXR4, and Bax expression levels were starkly higher in the M+Ms group versus the M+Ns group (p < 0.05). In contrast, Bcl-2 and type II collagen expression levels were markedly higher in the M+Ns group in comparison with the model group (p < 0.01) and remarkably reduced in the M+Ms group versus the M+Ns group (p < 0.05). These results suggested that the Ns therapy relieved apoptosis in the intervertebral disc by suppressing the SDF-1/CXCR4 axis in the NPCs of CS rats.

Effect of Ns on TUNEL staining in NPCs

To further evaluate Ns’ effect on SDF-1/CXCR4 axis-mediated inhibition of apoptosis in NPCs, TUNEL staining was performed. As shown in Figure 3D, TUNEL staining revealed markedly elevated apoptotic rate for NPCs in the model group in comparison with sham animals (p < 0.01), while this rate remarkably decreased (p < 0.01) in the M+Ns group versus the model group. However, similar apoptotic rates for NPCs (p > 0.05) were detected in the M+Ms and model groups. Comparing both therapies, NPC apoptosis was substantially enhanced in the M+Ms group versus the M+Ns group (p < 0.01); therefore, the Ns therapy could better reduce the increase in the apoptotic rate of NPCs.