Angiopoietin‐like protein 8 expression and association with extracellular matrix metabolism and inflammation during intervertebral disc degeneration

Abstract Intervertebral disc degeneration (IDD) is considered the primary culprit for low back pain. Although the underlying mechanisms remain unknown, hyperactive catabolism of the extracellular matrix (ECM) and inflammation are suggested to play critical roles in IDD progression. This study was designed to elucidate the role of angiopoietin‐like protein 8 (ANGPTL8) in the progression of IDD, especially the relationship of ANGPTL8 with ECM metabolism and inflammation. A positive association between ANGPTL8 expression and degenerative grades of IDD was detected in the analysis of human nucleus pulposus tissue samples. Silencing of ANGPTL8 attenuated the degradation of the anabolic protein type collagen II, and reduced the expression of the catabolic proteins MMP3 and MMP9, and the inflammatory cytokine IL‐6 through inhibition of NF‐κB signalling activation. In addition, the effect of ANGPTL8 was evaluated in a rat model of puncture‐induced IDD. Based on the imaging results and histological examination in animal study, knockdown of ANGPTL8 was demonstrated to ameliorate the IDD progression. These results demonstrate the detrimental role of ANGPTL8 expression in the pathogenesis of IDD and may provide a new therapeutic target for IDD treatment.

Intervertebral disc (IVD) is an avascular organ composed of the inner semi-gelatinous nucleus pulposus (NP), the outer annulus fibrosus and cartilage endplate. 8 The extracellular matrix (ECM) of the inner disc is constituted by collagens and proteoglycans, which enable the normal function of the IVD. 9 The progression of IDD is closely related to NP cells degeneration. During IDD, proinflammatory cytokines, such as TNF-α and IL-1β could accelerate NP cells degeneration by promoting the release of matrix metalloproteinases (MMPs) and inflammatory cytokines. [10][11][12][13] It has been confirmed that ECM degradation and inflammation play a critical role in accelerating IDD progression. However, the involved factors or mechanisms in the regulation of ECM metabolism and inflammation are not completely understood.
The nuclear factor-kappa B (NF-κB) signalling pathway has been demonstrated as a crucial mediator of the IDD process and the therapeutic target in disc degeneration. 12,[14][15][16][17] Pro-inflammatory cytokines activate the NF-κB signalling and downstream pathways, thereby participating in inflammatory response, cell apoptosis and ECM metabolism. 10,17 In the NF-κB signalling pathway, TNF-α binds to its receptor and transduces the signal to activate the IκB kinase (IKK) complex, which is composed of three units: IκB kinase α (IKKα), IκB kinase β (IKKβ) and IκB kinase γ (IKKγ). Activated IKK complex results in the phosphorylation of inhibitor of κB alpha (IκBα) and p65, which alter the downstream gene expression in NP cells. [17][18][19] Hyperactivation of NF-κB signalling is an essential step in promoting ECM catabolism and inflammatory cytokines production, which in turn could accelerate the progression of IDD. 12 Accumulating evidence has showed that angiopoietin-like proteins (ANGPTL) are associated with inflammation, metabolism, cell apoptosis and cancer progression. 20,21 ANGPTL proteins have eight family members and mainly are secreted glycoproteins with a structure similar to that of angiopoietin family proteins. Our previous study has indicated the role of angiopoietin family proteins in IDD. 22 One of the members in ANGPTL family, angiopoietin-like proteins 8 (ANGPTL8), which is also known as lipasin or betatrophin, is a regulator of plasma lipid metabolism and a therapeutic target in diabetes mellitus. [23][24][25][26][27] Nevertheless, the role of ANGPTL8 in inflammation and related signalling pathway is unknown. A recent study has shown that ANGPTL8 has an intracellular location and was suggested to be related to the regulation of inflammation. 28 Our work is the first study to investigate the role of ANGPTL8 in ECM metabolism and inflammation during the IDD process, and may provide a new potential target for the development of therapeutic strategies for IDD.
So far, the effect of ANGPTL8 on NP cells, as well as its role in inflammation and ECM degradation, remains unclear. Thus, to investigate the role of ANGPTL8 in IDD in this study, we first measured the level of ANGPTL8 expression in normal and degenerative human NP tissue samples. Subsequently, we investigated the effects of ANGPTL8 silencing or over-expression on ECM degradation and inflammation in human NP cells, and elucidated the relationship between ANGPTL8 and the NF-κB signalling pathway. Meanwhile, a rat degenerative disc model of ANGPTL8-silencing was developed to clarify the effects in vivo.

| Collection of human NP tissues
The experimental protocols in this study were approved by the Ethics Committee of Tongji Medical College, Huazhong University of Science and Technology. Written informed consent was obtained from all participants included in our study. Human NP tissues were collected from 40 patients (men 16 and women 24; mean age 32.5 years, range 17-56 years) who underwent surgery for idiopathic scoliosis or degenerative disc disease. The Pfirrmann grades based on T2-weighted section images were used to assess the degree of IDD. 29 In brief, Grade II showed an inhomogeneous disc with or without horizontal bands and a hyperintense white signal; Grade III showed an inhomogeneous disc with intermediate gray signal intensity and a slightly decreased disc height; Grade IV showed an inhomogeneous disc with hypointense gray or black signal intensity and a moderately decreased disc height; Grade V showed an inhomogeneous disc with a hypointense black signal intensity and a collapsed disc space. The specimens were immediately sectioned for use in various experiments. One section fixed in 4% buffered formaldehyde (pH 7.4) was used in histological analysis. A second section was immediately immersed in RNAlater (Invitrogen, Carlsbad, CA) and frozen in liquid nitrogen for protein and RNA analysis. A third section was covered with phosphate-buffered saline (PBS) in a sterile tube for cell isolation.

| Isolation and culture of NP cells
Nucleus pulposus cells were isolated as previously described. 22 Briefly, NP tissues were cut into pieces and enzymatically digested in 0.2% type II collagenase and 0.25% trypsin for 3 hours. After filtering and washing with PBS, the suspension was centrifuged, and the isolated cells were cultured in Dulbecco's modified Eagle medium containing 15% foetal bovine serum (Gibco, Waltham, MA) and 1% penicillin/ streptomycin (Invitrogen) in a 5% CO 2 incubator. The culture medium was replaced twice a week, and the cells from the second or third passage were prepared for the subsequent experiments.

| Quantitative real-time polymerase chain reaction
Total RNA, extracted with Trizol reagent (Invitrogen) from NP tissues or cultured NP cells, was reverse-transcribed and amplified by Quantitative real-time polymerase chain reaction (qRT-PCR) according to the standard protocols. The qRT-PCR was performed to quantify ANGPTL8, COL2A1, MMP3, MMP9, IL-6 and GAPDH mRNA expression levels. GAPDH was used for normalization. The primers used for qRT-PCR are listed in Table 1. All the data were tested in triplicate.

| Immunofluorescence analysis
Immunofluorescence analysis was performed as previously described. 30 Briefly, NP cells or tissues attached to slides were fixed with

| Enzyme-linked immunosorbent assay
After the NP cells culture, supernatants were collected and centrifuged, the contents of ANGPTL8, MMP3 and MMP9 secreted by NP cells were analysed using the corresponding Enzyme-linked immunosorbent assay (ELISA) kit (R&D Systems, Inc, Minneapolis, MN) according to the standard protocol. The experiment was performed in triplicate.

| Knockdown or over-expression of ANGPTL8
Nucleus pulposus cells were seeded and cultured in six-well plates and reached 70%-80% confluence before infection.

| Animal experiments
The animal experiments were performed following a protocol ap- by a 20-gauge needle from the dorsal side. 31 The second puncture was performed 1 week later, and the rats were randomly divided into three groups: non-injection (IDD) group, group with injection of normal PBS (2 μL), and group with injection of ANGPTL8-siRNA plasmid (2 μL, 20 μmol/L) using a 33-gauge needle (Hamilton, Benade, Switzerland). 32,33 The injection procedure was repeated after 2 weeks.

| Radiography and magnetic resonance imaging examination
All rats underwent radiography (X-ray), and the disc height was measured using the ImageJ software (US National Institutes of Health) and calculated using the disc height index (DHI) according to a previously described method. 31

| Histologic analysis
Animals were euthanized and the discs were harvested. The specimens were decalcified and fixed in formaldehyde, dehydrated and embedded in paraffin. The slides of each disc were stained with haematoxylin and eosin (HE) staining and assessed using a histological grading scale. 34 The specimen sections were deparaffinized and rehydrated, and then microwaved in 0.01 mol/L sodium citrate for 15 minutes. Next, 3% hydrogen peroxide was used to block endogenous peroxidase activity for 10 minutes, and 5% BSA was used to

| Statistical analysis
All experiments were performed independently at least in triplicate and data are presented as the mean ± SD. Statistical analyses were performed using GraphPad Prism 7 software (La Jolla, CA).
Differences between group means were evaluated using Student's t test or one-way ANOVA. P < 0.05 was considered statistically significant.

| Expression of ANGPTL8 in human NP tissues during degeneration
The degree of IDD was assessed by MRI in T2-weighted images, and 10 specimens of each degenerative degree were selected in the present study. HE staining and Alcian blue staining were used to confirm the degenerative degree of selected IVD tissues ( Figure 1A). To investigate the expression of ANGPTL8 in the process of the IDD, the transcriptional level was measured by qRT-PCR in samples with different IDD degrees ( Figure 1B). A positive correlation was found between ANGPTL8 mRNA level and IDD grade ( Figure 1C). Moreover, Western blot analysis indicated that a higher ANGPTL8 protein expression level was related to a higher degenerative grade of NP tissues ( Figure 1D). Additionally, the positive rate of ANGPTL8-labelled cells was significantly greater in Grade IV or V than in Grade II or III according to the immunofluorescence analysis ( Figure 1E-F).
These results demonstrated that the level of ANGPTL8 expression in human NP tissues was increased during the IDD process.

| TNF-α treatment facilitate the expression of ANGPTL8 in NP cells
Previous study has demonstrated that pro-inflammatory cytokines promoted the NP cells degeneration and accelerated the progression of IDD. 18,35 In order to investigate the pathological effect of ANGPTL8 expression in IDD, NP cells were treated with TNF-α (50 ng/mL) for 0, 6, 12, 24 or 48 hours in vitro. In a time-dependent manner, TNF-α treatment up-regulated the transcriptional level of ANGPTL8 significantly ( Figure 2A). Similarly, the protein level of ANGPTL8 was also increased due to the stimulation of TNF-α ( Figure 2B). Moreover, the contents of ANGPTL8 in NP cells culture supernatant were increased in a time-dependent manner as analysed by ELISA ( Figure 2C). Overall,

| Knockdown of ANGPTL8 reduces TNF-αinduced ECM degradation and inflammation in vitro
To determine whether ANGPTL8 plays a role in TNF-α induced inflammation and ECM degradation in NP cells, knockdown of ANGPTL8 was realized by transfection with siRNA (si-ANGPTL8).
Three si-ANGPTL8 fragments were generated, and the inhibitory efficiency of ANGPTL8 in NP cells was analysed ( Figure 3A). The whereas that of type II collagen increased in the si-ANGPTL8 group compared with those in the control group ( Figure 3M-N). Hence, knockdown of ANGPTL8 attenuated the production of MMPs and inflammatory cytokines, and preserved type II collagen contents in NP cells.

| Over-expression of ANGPTL8 promotes TNF-α-induced ECM degradation and inflammation in vitro
In order to further assess the effect of ANGPTL8 during IDD, NP cells were transfected with lentivirus vector encoding homo- analysis also indicated that an elevated level of ANGPTL8 reduced the deposition of type II collagen and increased the MMP3, MMP9 and IL-6 protein level ( Figure 4M-N). These data demonstrated that ANGPTL8 enhanced the detrimental effects of TNF-α during the NP cells degeneration by promoting the release of MMPs and inflammatory cytokines.

| ANGPTL8 regulates ECM metabolism and inflammatory response through the NF-κB signalling pathway
The over-activation of the NF-κB signalling pathway is associated with the pathogenesis of IDD. 36 Pro-inflammatory cytokines, such as TNF-α, bind to specific receptors in the cytomembrane and activate the classical NF-κB pathway. 37 It was assumed that ANGPTL8

| Silencing of ANGPTL8 ameliorates IVD degeneration in vivo
To evaluate the effects of ANGPTL8 in vivo, intradiscal delivery of si-ANGPTL8 was performed in a rat degeneration disc model.
The extent of IDD was assessed by MRI and X-ray examination at 8 weeks ( Figure 6A,B). According to the MRI results, the Pfirrmann score, which was used to evaluate the degree of IDD, was lower in the si-ANGPTL8 group than in the IDD or PBS group ( Figure 6C). In addition, the %DHI was calculated based on the X-ray results. A low %DHI indicated a degenerative disc with a collapsed or narrowing intervertebral space, and it was noted that silencing of ANGPTL8 could restore the disc height to some extent ( Figure 6D). In order to further investigate the changes of IDD, histological staining and immunohistochemistry analysis were performed for the rat disc ( Figure 6E). Each group was subjected to HE staining and was assessed by histological grading ( Figure 6F). The si-ANGPTL8 group showed a higher level of type II collagen and a lower level of MMP3, and also showed improved histological grade compared with the IDD or PBS group. These results that knockdown ANGPTL8 could significantly retard the progression of IDD.

Imbalance between ECM metabolism and inflammatory response
in NP tissue of IVD is the critical culprit in the development of IDD. 30 In the present study, we found that the expression level of ANGPTL8 was increased in degenerative disc and correlated with the degree of IDD. Knockdown of intracellular ANGPTL8 expression in NP cells reduced the expression of MMP3 and MMP9, and the production of the pro-inflammatory cytokine, IL-6. On the contrary, the content of type II collagen was increased. Accordingly, TNF-α treatment promoted the expression of ANGPTL8 and over-expression of ANGPTL8 under TNF-α stimulation showed the opposite effects to that with ANGPTL8 silencing. Our further study indicated that ANGPTL8 promoted the TNF-α-induced ECM degradation and inflammation through the NF-kB signalling pathway in NP cells ( Figure 7). In addition, the in vivo study in a rat tail puncture-induced degeneration model also demonstrated that silencing ANGPTL8 could partly restore the degenerative IVD. These findings indicated that ANGPTL8 plays an adverse role during the IDD process.
The ECM of the NP tissue is rich in collagens and proteoglycans. 17 Approximately 70% of NP dry weight, which is involved in water absorption, nutrition diffusion and osmotic regulation in NP tissues, is composed of type II collagen and aggrecan. 38 In a healthy IVD, there is a harmonious balance between the catabolism and anabolism of ECM gradients, which is crucial for normal structure and function of IVD. 39,40 During the IDD process, the imbalance of ECM metabolism, presented as the progressive loss of type II collagen and aggrecan, induces changes of the morphology and structure of IVD, including disc height reduction and collapsed disc space, or decreased water content and mechanical buffering ability. 38,[41][42][43][44] In molecular analysis, catabolic enzymes, primarily MMPs and a disintegrin and metalloprotease with thrombospondin motifs (ADAMTSs), are highly expressed in degenerative IVD. 45,46 MMPs, such as MMP-3, MMP-9 and MMP-13, degrade the type II collagen and aggrecan in ECM, resulting in the hyperactive catabolic effect. 38,47 Previous studies indicated that proinflammatory cytokines could promote the expression of MMPs and ADAMTSs, and accelerate the progression of IDD. 16,[48][49][50] In the present experiment, TNF-α treatment induced the expression of MMP3, MMP9 and IL-6, as well as the decrease of type II collagen. It was demonstrated that TNF-α accelerated the NP cells degeneration, and ANGPTL8 may serve as a pro-degeneration molecule by enhancing the effects of pro-inflammatory cytokines.
The NF-κB signalling pathway has been verified to play a role in promoting the apoptosis and senescence of NP cells, and increasing matrix-degrading enzyme activity and mediating inflammatory response in IDD. 16,51 Thus, inhibiting the NF-κB signalling pathway has a therapeutic potential in the progression of IDD. 12,14,17,51,52 In (F), Histological grades were measured according to the HE staining results. Data were presented as the mean ± SD (n = 3). *P < 0.05 vs control group, # P < 0.05 vs IDD group. $ P < 0.05 vs PBS group solution injection. 32 The degree of IDD was assessed based on the histological changes or the morphology, water content, and disc height of IVD. Consistent with the in vitro study results, delivering si-ANGPTL8 plasmids ameliorated the disc degeneration in vivo, which was presented as an improved histological grade in HE staining. The collagen II expression was increased and MMPs were decreased in the si-ANGPTL8-treated group compared with the IDD or PBS group. In addition, the conclusion that silencing of ANGPTL8 reversed the IDD process could also be supported by the MRI or X-ray assessment and histological results. Thereby, our animal study suggested that silencing of ANGPTL8 slowed the progression of IDD to some degree.
Members of angiopoietin-like proteins are associated with inflammation and metabolism, and most of them are pro-inflammatory molecules. 20 For example, ANGPTL2 and ANGPTL4 are acknowledged for their adverse pro-inflammatory properties, and play a crucial role in mediating acute inflammation. 56 In conclusion, our results indicated that ANGPTL8 plays an adverse role in the progression of IDD. Silencing of ANGPTL8 could reduce ECM degradation and inflammation in human NP cells via inhibition of the NF-κB signalling pathway. Moreover, over-expression of ANGPTL8 enhanced the effects of TNF-α and the activation of NF-κB signalling pathway. We also demonstrated that knockdown of ANGPTL8 ameliorates IDD progression in a rat model of IDD. Thus, ANGPTL8 may be a potential therapeutic target for the treatment of IDD.