Introduction

Degenerative diseases of the intervertebral disc, such as intervertebral disc disease (IVDD), encompass the physiological and pathological processes of natural aging and degeneration1, which can cause a series of clinical symptoms, such as waist pain and numbness of the lower limbs. Serious degenerative diseases of the intervertebral disc can cause dysphoria and sexual dysfunction2 and affect the ability to work and quality of life. Commonly used drugs to treat these diseases include nonsteroidal anti-inflammatory drugs3 and opioid analgesics4, but these only temporarily alleviate symptoms, and the pathogenesis of IVDD remains unclear. Therefore, it is necessary to gain a more in-depth understanding of the molecular mechanism underlying the development of IVDD19 and participate in the formation of the CCR4-NOT complex and contribute to various cell activities20.

Results

Expression of circPKNOX1 decreased in degenerative discs

We analyzed microchip data collected from previous research on circRNA and IVDD20 (Fig. 5B). Compared with that in normal tissues, the expression of KIAA0355 decreased significantly in degenerated tissues (Fig. 5C). Through TargetScan, we also identified the binding site of miR-370-3p and KIAA0355 (Fig. 5D).

Fig. 5: circPKNOX1 retarded the IVDD process via the miR-370-3p/ KIAA0355 axis.
figure 5

A miR-370-3p target genes were predicted using TargetScan, miRDB, and miRTarBase. B After knocking down miR-370-3p, the expression level of KIAA0355 increased; ***P<0.001. Data represent mean±SD, and the P values were determined by a two-tailed unpaired Student’s t test. C In degenerated tissues, the expression of KIAA0355 decreased; **P<0.01. Data represent mean±SD, and the P values were determined by a two-tailed unpaired Student’s t test. D The binding site of miR-370-3p and KIAA0355 was predicted using TargetScan. E After transfection with siKIAA0355, the expression of KIAA0355 decreased; **P<0.01. Data represent mean±SD, and the P values were determined by a two-tailed unpaired Student’s t test. The expression of related proteins (collagen II, aggrecan, SOX9, MMP3, MMP13, ADAMTS4, and ADAMTS-5) was observed by western blotting (F) after KIAA0355 knockdown and (G) after KIAA0355 and miR-370-3p knockdown. The fluorescence intensity of related proteins (collagen II, aggrecan, ADAMTS4, and MMP13) (scale bar, 100μm) was observed via an immunofluorescence assay (H) after KIAA0355 knockdown and (I) after KIAA0355 and miR-370-3p knockdown. RT-qPCR was used to detect the expression levels of related genes (J) after KIAA0355 knockdown (*P<0.05, **P<0.01. Data represent mean±SD, and the P values were determined by a two-tailed unpaired Student’s t test) and (K) after KIAA0355 and miR-370-3p knockdown (*P<0.05, **P<0.01, ***P<0.001. Data represent mean±SD, and the P values were determined by a two-tailed unpaired Student’s t test.). Top: the extracellular matrix was detected by Alcian staining (L) after KIAA0355 knockdown and (M) after KIAA0355 and miR-370-3p knockdown. Bottom: the Alcian staining was quantitatively analyzed. Data represent mean±SD, and the P values were determined by a two-tailed unpaired Student’s t test.

Full size image

Next, we explored whether KIAA0355 and IVDD were related and tested the efficiency of KIAA0355 siRNA (Fig. 5E). The knockdown of KIAA0355 was associated with a decrease in the expression of collagen II, aggrecan, and SOX9 and an increase in the expression of ADAMTS4, ADAMTS-5, MMP3, and MMP13 (Fig. 5F, H, J). Moreover, when the expression of KIAA0355 decreased, the degradation of the extracellular matrix was accelerated (Fig. 5L). Therefore, we concluded that KIAA0355 was a target gene related to IVDD.

Finally, we evaluated the relationship between miR-370-3p and KIAA0355 through a rescue experiment. In NP cells, we knocked down miR-370-3p either alone or with KIAA0355. When miR-370-3p was knocked down alone, we observed a positive effect on NP cells. The expression of collagen II, aggrecan, and SOX9 increased, and the expression of ADAMTS4, ADAMTS-5, MMP3, and MMP13 decreased. Co-knockdown of KIAA0355 and miR-370-3p was not conducive to the growth of NP cells; the corresponding matrix synthesis was reduced, and the decomposition enzyme levels were increased (Fig. 5G, I, K). Notably, the Alcian staining results were consistent (Fig. 5M). Therefore, we concluded that circPKNOX1 regulated the growth, synthesis, and decomposition of NP cells through the miR-370-3p–KIAA0355 axis.

In the animal model, the circPKNOX1/miR-370-3p/KIAA0355 axis promoted IVDD development

The degree of bend in the tail of the model corresponded to the decrease in the disc height index (Fig. 6A). The expression of circPKNOX1 and KIAA0355 in the intervertebral disc of the tail was increased in the circPKNOX1-wt group (Fig. 6B, C). After several staining procedures (hematoxylin and eosin, Safranin-O/fast green, and toluidine blue), we found that the NP tissue was extruded by pressure and lost (Fig. 6D). We detected the expression of related anabolic and catabolic proteins in the intervertebral discs of the mice using western blotting (Fig. 6E). Anabolism was decreased in the degenerative group, and catabolism was increased, but this trend was reversed when circPKNOX1 expression increased. Finally, immunohistochemical staining showed that the expression of collagen II, aggrecan, ADAMTS4, and MMP13 decreased compared with that in the circPKNOX1-mut and circPKNOX1-wt groups (Fig. 6F). In summary, these results demonstrated that circPKNOX1 had a therapeutic effect on intervertebral disc degeneration.

Fig. 6: In the animal model, the circPKNOX1/miR-370-3p/ KIAA0355 axis promoted IVDD development.
figure 6

A Left: the tails of the mice were examined using X-ray imaging analysis. Right: the disc height index of the tails of the mice was quantitatively analyzed; *P<0.05. Data represent mean±SD, and the P values were determined by a two-tailed unpaired Student’s t test. B The expression of circPKNOX1 was detected using RT-qPCR; ***P<0.001. Data represent mean±SD, and the P values were determined by a two-tailed unpaired Student’s t test. C The expression of KIAA0355 was detected using RT-qPCR; ***P<0.001. Data represent mean±SD, and the P values were determined by a two-tailed unpaired Student’s t test. D The intervertebral disc tissue of mice was observed using hematoxylin and eosin, Safranin-O/fast green, and toluidine blue staining (scale bar, 200μm). E In the control, IVDD, circPKNOX1-mut, and circPKNOX1-wt groups, the expression of related proteins (collagen II, aggrecan, SOX9, MMP3, MMP13, ADAMTS4, and ADAMTS-5) was detected using western blotting. F Sections of the discs were analyzed using immunohistochemical staining for collagen II, aggrecan, ADAMTS4, and MMP13 (scale bar, 200μm). G The mechanism of action of circPKNOX1.

Full size image

Discussion

Although the mechanism through which circRNAs regulate the progression of IVDD is not clear, accumulating evidence indicates that circRNAs may act as miRNA sponges28, translating or interacting with transcription factors to regulate the transcription of their genes. Some studies have shown that circVMA21, like a sponge of miR-200c, can reduce the apoptosis of NP cells induced by inflammatory cytokines and the imbalance between extracellular matrix anabolism and catabolism through the XIAP pathway. The expression of XIAP in the NP cells is directly related to apoptosis and imbalances in the expression of anabolic and catabolic factors in the extracellular matrix. miR-200c, as an upstream molecule, regulates the activity and function of NP cells by inhibiting XIAP. As a sponge of miR-200c, circVMA21 plays a role in NP cells by targeting miR-200c and XIAP. In addition, IVDD in rats can be reduced by administering an intradiscal injection of circVMA21 in vivo29. The results of our bioinformatics analysis showed that circPKNOX1 had multiple potential binding sites for multiple miRNAs. The results of FISH and double-luciferase analysis showed that miR-370-3p directly combined with circPKNOX1, providing a potentially effective treatment strategy against IVDD.

In summary, circPKNOX1 regulated the expression of KIAA0355 through the miR-370-3p pathway, thus affecting the progression of IVDD (Fig. 6G). Therefore, IVDD treatment based on circPKNOX1 may be an effective therapeutic strategy. A limitation of this study was the age gap between IVDD and non-IVDD patients, which may have introduced bias; further in vitro and in vivo experiments are needed to eliminate the influence of age. In addition, as miRNA sponges, circRNAs may perform other functions in IVDD that require further investigation. For instance, circRNAs act as dynamic scaffolds for regulating protein-protein interactions30,31. Advances in chemical and biological research are needed to improve the current techniques for recognizing these interactions.

Materials and methods

Human intervertebral disc tissue

According to the procedures of the Ethics Committee of Shaw Hospital in Zhejiang, we collected human nucleus pulposus (NP) tissue after surgery. Before the operation, all the patients, after being informed of all risks and possible outcomes, signed informed consent forms. All specimens were stored at −80 °C, according to the specifications. The control group comprised patients with lumbar vertebrae trauma (n = 3), and the pathological group comprised of patients with herniated discs (n = 49). Patient information is provided in Supplementary Table S132.

Mouse disc degeneration model

C57 mice were purchased from Shanghai SLAC Laboratory Animal Co., Ltd. (Shanghai, China). All animal experiments were conducted according to the principles and procedures of the National Institutes of Health Guide for the Care and Use of Laboratory Animals and the guidelines of Sir Run Run Shaw Hospital (Zhejiang University, Hangzhou, Zhejiang, China) for animal treatment.

After anesthesia was administered, the tails of the mice were fixed with a thin wire, which was then bent into a ring to simulate the chronic degeneration of the intervertebral disc under pressure33. One month after the tail was fixed, the mice were divided into four groups. Wild-type (wt) and mutant (mut) circPKNOX1 adenoviruses were constructed and provided by Hanbio Biotechnology Co., Ltd. (Shanghai, China). After successful modeling, the experimental and control groups were injected intraperitoneally with adenovirus solution and normal saline, respectively, once a week for 8 weeks. At the end of the experiment, the mice were euthanized, and the tails were removed for X-ray imaging analysis.

NP cell culture

To extract NP cells, intervertebral disc tissue was isolated from humans and mice, and digested with collagenase type II (Sigma, USA). The NP cells were cultured at 37 °C in a humidified environment containing 5% carbon dioxide. During cultivation, nutrients were supplied by Dulbecco’s Modified Eagle Medium (DMEM) containing 10% fetal bovine serum (ThermoFisher Scientific, Waltham, MA, USA).

Bioinformatics analysis

We predicted the downstream miRNA of circPKNOX1 using three analysis tools: TargetScan (http://www.targetscan.org/), miRanda (http://www.microrna.org/), and RNAhybrid. We used the intersection of data from three online databases, TargetScan, miRDB (http://mirdb.org), and miRTarBase (http://mirtarbase.mbc.nctu.edu.tw/php/download.php), to predict the target gene of miR-370-3p. The binding site was analyzed using TargetScan and CircInteractome (https://circinteractome.nia.nih.gov/index.html).

RNA extraction and RT-qPCR analysis

The total cellular RNA was extracted from cultured NP cells using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and the Ultrapure RNA Kit (CWBIO, Bei**g, China) according to the manufacturer’s instructions. miRNA was extracted using a miRNA extraction kit (CWBIO). After reverse transcribing RNA into cDNA using a reverse transcription kit (Accurate, Hunan, China), miRNA and mRNA expression was analyzed using RT-qPCR. The reaction mixture contained 1 µl of cDNA, 5 µl of SYBR Green Master Mix (Yeasen Biotech Co., Ltd., Shanghai, China), 1 µl of primer (TsingKe, Hangzhou, China), and 3 µl of water (10 µl total). The reaction was carried out in an ABI 7500 Sequencing Detection System (Applied Biosystems, Foster City, CA, USA) under the following conditions: 40 cycles of denaturation at 95 °C for 5 s and amplification at 60 °C for 24 s. The primer details are provided in Supplementary Table S2.

Adenovirus overexpression

The construction and extraction of adenovirus circPKNOX1 and insertion of PCR-cloned circ gene cDNA (Ad-circPKNOX1) and green fluorescent protein gene cDNA (Ad-GFP) into a pAdEasy-EF1-MCS-CMV-GFP vector were performed by Han Biological, Shanghai, China. We infected cells with the virus and adjusted the amount of virus according to the titer and cell number (virus amount = MOI × cell number/virus titer (PFU/ml) × 1000). After 4 h of infection, the medium was changed, and the cells were observed under a fluorescence microscope to determine whether the infection was successful. Finally, the expression levels were detected using PCR34.

RNA interference and overexpression

We suppressed the expression of circPKNOX1 through a siRNA-mediated gene and used specific inhibitors and mimetics (Ribobio, Guangzhou, China) to inhibit or induce the expression of miR-370-3p, respectively. We introduced siRNA into cells, using the Lipofectamine RNAiMAX transfection reagent (ThermoFisher), and detected related expression levels after 48 h.

Alcian blue staining

Alcian staining was performed with an Alcian blue stain kit (G1563, Solarbio). NP cells were seeded onto 24-well plates and fixed with 10% neutral formalin after transfection. The cells were acidified for 5 min and then stained for 30 min. Finally, excess Alcian dye was removed, and the cells were washed and observed under the scanner (V600, EPSON).

Western blotting

Human or mouse NP cells were lysed with a radioimmunoprecipitation assay buffer (Beyotime, China) to extract cellular proteins. The proteins were separated using 10% SDS-PAGE. The separated proteins were transferred from the gel to a polyvinylidene fluoride membrane (Bio-Rad, Hercules, CA, USA) and blocked with skimmed milk powder for 1 h. Subsequently, the membrane was immersed in the relevant protein antibody solution (collagen II, aggrecan, SOX9, MMP3, MMP13, ADAMTS4, or ADAMTS-5; 1:1000; Abcam) at 4 °C for 8 h. After washing away excess antibody, the membrane was incubated with a secondary antibody at room temperature (25 °C) for 2 h. The protein bands were observed with an Amersham lmager 600 (General Electric Company, USA) and analyzed with the ImageJ software (National Institute of Health, Bethesda, MD, USA).

Immunofluorescence (IF) assay

The treated NP cells were fixed with 4% paraformaldehyde for 20 min. After the paraformaldehyde was washed off, the cell membrane was permeabilized with 0.3% Triton X-100 and then blocked with 5% BSA for 60 min. The cells were immersed in the relevant antibody solution (collagen II, aggrecan, MMP3, or ADAMTS-5; 1:1000; Abcam) at 4 °C for 8 h. After excess antibody was washed off, the cells were incubated in goat anti-rabbit IgG conjugated to fluorescent Cy5 dye (1:100; Abcam), for 2 h in the dark. Nuclei were stained with DAPI (Life Technologies, Carlsbad, CA, USA). A Zeiss LSM780 confocal microscope was used to observe cell fluorescence. IF images were captured using a Nikon Eclipse TI and merged using Image-Pro Plus 6.0 (National Institute of Health).

RNA fluorescence in situ hybridization (FISH)

A Cy3-labeled circPKNOX1 probe and a 488-labeled locked nucleic acid miR-370-3p probe were designed and synthesized by HaoKe, Wuhan, China. A FISH kit (RiboBio, Guangzhou, China) was used to detect the NP cell fluorescence signal. These images were captured using the Nikon A1Si Laser Scanning Confocal Microscope (Nikon Instruments Inc., Japan).

Statistical analysis

Statistical analysis was performed using SPSS v22.0. The unpaired data between the two groups were analyzed using a t test (with a 95% confidence interval for the differences between groups). Values of P < 0.05 were considered statistically significant.