Teriparatide ameliorates articular cartilage degradation and aberrant subchondral bone remodeling in DMM mice

Objective Knee osteoarthritis (KOA) is a highly prevalent musculoskeletal disorder characterized by degeneration of cartilage and abnormal remodeling of subchondral bone (SCB). Teriparatide (PTH (1–34)) is an effective anabolic drug for osteoporosis (OP) and regulates osteoprotegerin (OPG)/receptor activator of nuclear factor ligand (RANKL)/RANK signaling, which also has a therapeutic effect on KOA by ameliorating cartilage degradation and inhibiting aberrant remodeling of SCB. However, the mechanisms of PTH (1–34) in treating KOA are still uncertain and remain to be explored. Therefore, we compared the effect of PTH (1–34) on the post-traumatic KOA mouse model to explore the potential therapeutic effect and mechanisms. Methods In vivo study, eight-week-old male mice including wild-type (WT) (n = 54) and OPG−/− (n = 54) were investigated and compared. Post-traumatic KOA model was created by destabilization of medial meniscus (DMM). WT mice were randomly assigned into three groups: the sham group (WT-sham; n = 18), the DMM group (WT-DMM; n = 18), and the PTH (1–34)-treated group (WT-DMM + PTH (1–34); n = 18). Similarly, the OPG−/− mice were randomly allocated into three groups as well. The designed mice were executed at the 4th, 8th, and 12th weeks to evaluate KOA progression. To further explore the chondro-protective of PTH (1–34), the ATDC5 chondrocytes were stimulated with different concentrations of PTH (1–34) in vitro. Results Compared with the WT-sham mice, significant wear of cartilage in terms of reduced cartilage thickness and glycosaminoglycan (GAG) loss was detected in the WT-DMM mice. PTH (1–34) exhibited cartilage-protective by alleviating wear, retaining the thickness and GAG contents. Moreover, the deterioration of the SCB was alleviated and the expression of PTH1R/OPG/RANKL/RANK were found to increase after PTH (1–34) treatment. Among the OPG−/− mice, the cartilage of the DMM mice displayed typical KOA change with higher OARSI score and thinner cartilage. The damage of the cartilage was alleviated but the abnormal remodeling of SCB didn't show any response to the PTH (1–34) treatment. Compared with the WT-DMM mice, the OPG−/−-DMM mice caught more aggressive KOA with thinner cartilage, sever cartilage damage, and more abnormal remodeling of SCB. Moreover, both the damaged cartilage from the WT-DMM mice and the OPG−/−-DMM mice were alleviated but only the deterioration of SCB in WT-DMM mice was alleviated after the administration of PTH (1–34). In vitro study, PTH (1–34) could promote the viability of chondrocytes, enhance the synthesis of extracellular matrix (ECM) (AGC, COLII, and SOX9) at the mRNA and protein level, but inhibit the secretion of inflammatory cytokines (TNF-α and IL-6). Conclusion Both wear of the cartilage was alleviated and aberrant remodeling of the SCB was inhibited in the WT mice, but only the cartilage-protective effect was observed in the OPG−/− mice. PTH (1–34) exhibited chondro-protective effect by decelerating cartilage degeneration in vivo as well as by promoting the proliferation and enhancing ECM synthesis of chondrocytes in vitro. The current investigation implied that the rescue of the disturbed SCB is dependent on the regulation of OPG while the chondro-protective effect is independent of modulation of OPG, which provides proof for the treatment of KOA. The translational potential of this article Systemic administration of PTH (1-34) could exert a therapeutic effect on both cartilage and SCB in different mechanisms to alleviate KOA progression, which might be a novel therapy for KOA.

more aggressive KOA with thinner cartilage, sever cartilage damage, and more abnormal remodeling of SCB. Moreover, both the damaged cartilage from the WT-DMM mice and the OPG À/À -DMM mice were alleviated but only the deterioration of SCB in WT-DMM mice was alleviated after the administration of PTH . In vitro study, PTH  could promote the viability of chondrocytes, enhance the synthesis of extracellular matrix (ECM) (AGC, COLII, and SOX9) at the mRNA and protein level, but inhibit the secretion of inflammatory cytokines (TNF-α and IL-6). Conclusion: Both wear of the cartilage was alleviated and aberrant remodeling of the SCB was inhibited in the WT mice, but only the cartilage-protective effect was observed in the OPG À/À mice. PTH  exhibited chondroprotective effect by decelerating cartilage degeneration in vivo as well as by promoting the proliferation and enhancing ECM synthesis of chondrocytes in vitro. The current investigation implied that the rescue of the disturbed SCB is dependent on the regulation of OPG while the chondro-protective effect is independent of modulation of OPG, which provides proof for the treatment of KOA. The translational potential of this article: Systemic administration of PTH  could exert a therapeutic effect on both cartilage and SCB in different mechanisms to alleviate KOA progression, which might be a novel therapy for KOA.

Introduction
Knee osteoarthritis (KOA) is a highly prevalent musculoskeletal disorder, which has impaired activities and quality of life of the patient and is ranked one of the most common orthopedic disorders [1]. It is a whole-joint disorder with cartilage erosion and subchondral bone (SCB) disturbance, the pathogenesis and treatment of which is affected by multi-factors such as senility, gender, genetics, obesity, and trauma, etc [2]. Cartilage is highly-organized connective tissue and chondrocytes are the unique cellular component to maintain the balance of the extracellular matrix (ECM) [3], and the SCB plays a crucial role in KOA progression [4,5]. Synovial cells can secret the related cytokines by inducing the expression of receptor activator of nuclear factor ligand (RANKL) to promote osteoclast formation and bone resorption [6]. The integrity and physiology of cartilage and SCB is important for the maintenance of the joint, which is a challenging aspect in KOA treatment. Lesions of these structures would result in KOA initiation with clinical symptoms while the treatment of KOA is challenged with less desirable effects. Current strategies such as exercise and analgesic drugs are common choices with no ideal effect [7]. Therefore, disease-modified medicines for the prevention and treatment of OA are highly desirable.
Teriparatide (PTH (1-34)), the recombinant human parathyroid hormone , is applied for the treatment of osteoporosis (OP) and bone fracture [8]. Investigations of PTH (1-34) on wild-type (WT) KOA mice have been extensively studied and a multitude of research revealed that PTH (1-34) might play its role in regulating the differentiation of chondrocytes [9,10]. Nevertheless, the mechanism of PTH (1-34) functions in KOA is still uncertain. Osteoprotegerin (OPG) is capable of binding the RANKL to regulate the differentiation of osteoclast [11] and maintain the integrity of articular cartilage [12]. PTH (1-34) is well applied in OP treatment and regulated OPG/RANKL/RANK signaling [13], but whether the mechanisms of PTH (1-34) involved the role of OPG during the KOA treatment is still uncertain and remains to be explored.
Therefore, in this study, both WT and OPG-knockout (OPG À/À ) mice models were used to evaluate the potential therapeutic distinction of PTH (1-34) including cartilage-protective, chondro-regenerative, and SCB amelioration. We hypothesized that PTH (1-34) might attenuate the KOA progression by exerting chondro-protective and SCB-alleviated function, which mechanisms may correlate with the OPG signaling.

Animals
All experimental animals were approved by the Animal Care and Use Committee of Peking University Shenzhen Hospital (No. 2021-501) and followed Animal Research: Reporting of in vivo Experiments (ARRIVE) guidelines [14]. Eight-week-old male mice including the WT mice (n ¼ 54) and the OPG À/À mice (n ¼ 54) were purchased from the Shanghai Research Center for Biomodel Organisms (Shanghai, China) and genotyped by polymerase chain reaction (PCR). The WT mice were randomly assigned to three groups: the WT-sham group (n ¼ 18), the WT-DMM group (n ¼ 18), and the WT-DMM þ PTH (1-34) group (n ¼ 18). Similarly, OPG À/À mice (n ¼ 54) were randomly assigned to three groups as well. Mice in the sham group and the destabilization of medial meniscus (DMM) group were administered with an equal volume of saline. All mice were housed in Topbio-technology (Shenzhen, China) under standard laboratory conditions (24 C, 12 h light/dark cycle) with food and water. Moreover, none of the mice caught death or infection during the experiment.

DMM surgery and treatment protocol
The mice were anesthetized with isoflurane (RWD life science, Shenzhen, China) and performed with DMM according to the previous protocol, which was widely accepted to establish post-traumatic KOA [15]. To relieve the pain and remove the sutures before starting the drug treatment, the mice were administrated with anesthesia one week postoperatively [16,17]. Specifically, all mice were treated by subcutaneous injection (S.I) with Teriparatide (PTH (1-34), HY-P0059, MCE, USA) (100 uL Â 40 μg/kg/day), which was confirmed by Ultimate 3000 U system High Performance Liquid Chromatography (HPLC) (Thermo Scientific) (Fig. S1). The mice were euthanized in the 4th, 8th, and 12th weeks for relevant evaluation.

Gross observation
All mice were euthanized at the designed time points for the harvest of the knee joints. The tibia plateaus were photographed with a Zoomstereoscopic microscope (Leica, Germany) and evaluated according to the grading system (Table S1) [18].

Microcomputer tomography evaluation
The region of interest (ROI) was defined between the SCB plate and the epiphysis [19]. The SCB was scanned and three-dimensionally

Biosafety evaluation
The primary organs including the heart, liver, spleen, lung, and kidney were obtained. Further, we performed hematoxylin and eosin (HE) staining to evaluate the biosafety of PTH  in vivo.

Histology evaluation
The knee joints were fixed in 10% neutral buffered formalin for one week after the photograph for gross observation, decalcified by 10% ethylene diamine tetra acetic acid (EDTA) (pH ¼ 7.4, 37 C, shaker, 80 rpm/min, 3 days), embedded in paraffin, and prepared 3.5 μm microsections for staining. The HE staining was applied for a general view while the Toluidine blue O (TB) staining and the Safranin O-Fast Green (SOFG) staining were performed to detect the glycosaminoglycan (GAG) contents of the articular cartilage. Moreover, the Alican blue (AB) staining and Masson trichrome (MT) staining were performed as well. All staining kits were purchased from Solarbio (Beijing, China). Representative changes were photographed by microscope (Leica, Germany) and evaluated by blinded observers according to Osteoarthritis Research Society International (OARSI) system (Table S2) [20]. Cartilage thickness was measured by width-calculating while the special-stained area was measured by an area-calculating algorithm with ImageJ (Version 5.0, USA).

Viability evaluation
The viability was assessed by cell counting kit-8 assay (CCK-8) (KeyGen Biotech, China) and the optical density (OD) was measured by Microplate Reader (Thermo Scientific Multiskan SkyHigh, USA) at 450 nm.

ECM anabolism evaluation
Quantitative reverse transcription polymerase chain reaction analysis (RT-qPCR) was performed to detect the relative mRNA expression of marked genes. The total mRNA extracted with Trizol (Life Technologies Co., USA) and cDNA was reversed transcript by TaqMan reagents (Takara, Japan). Primers were retrieved from National Center for Biotechnology Information (NCBI) bank and confirmed by the BLAST analysis ( Table 1). The relative mRNA expression of Aggrecan (AGC), Type II collagen (COLII), SRY-Box Transcription Factor 9 (SOX9), Matrix Metallopeptidase 13 (MMP13), Type X collagen (COLX), and ADAM Metallopeptidase with Thrombospondin Type 1 Motif 5 (ADAMTS5) were measured. GAPDH was used as endogenous control while the relative mRNA expression (fold control) was calculated with the 2 ÀΔΔCt method.

Statistical analysis
All experiments were performed in triplicates independently. Oneway analysis of variance (ANOVA) was performed among multiple comparisons within the WT mice groups or OPG À/À mice groups while two-way ANOVA was applied for multiple comparisons among all the mice groups. Two-tailed P < 0.05 were considered statistically significant. Values are labelled as *P < 0.05, **P < 0.01. Statistical analysis data were measured by R (Version 4.0.2, R Foundation for Statistical Computing, Vienna, Austria) or GraphPad Prism (Version 9.0, GraphPad Software, La Jolla, CA).

PTH (1-34) alleviates cartilage wear
All procedures followed the designed protocol and all mice survived well (Fig. 1a). The macroscopic of tibia plateaus were displayed in Fig. 1b while a much rougher surface with fibrillation was detected in DMM mice. According to the macroscopic of tibia plateaus in the PTH (1-34)treated mice, the surface of tibia plateaus seems with less wear or tear, which implied with less lesion of the articular cartilage. And therefore, the wear and tear were alleviated in the PTH (1-34)-treated mice (Fig. 1c). In addition, the macroscopic scoring implied that cartilage wear was alleviated by PTH  at the 12th week. The biosafety of PTH (1-34) is favorable because no significant differences were observed in major organs in all groups according to the historical evaluations (Fig. 2).

The cartilage-protective effect of PTH (1-34)
No degeneration of cartilage with similar smooth surface was demonstrated in sham mice while the worst situations were detected in DMM mice with cartilage surface irregularities, rough, crack, strip, or even defect according to the HE staining, the TB staining, and the SOFG staining (Fig. 3a-c). Particularly, thinner cartilage with fewer GAG contents were detected in OPGÀ/À mice and more aggressive KOA progression the OPGÀ/À-DMM mice. All DMM mice exhibited cartilage damage, but they were alleviated after the administration of PTH (1-34) differently in WT and OPG À/À mice. The OARSI score ranked highest in all the DMM mice but reduced significantly after the treatment (Fig. 3d). In addition, the thickness of cartilage in DMM mice was increased after the administration of PTH (1-34) as well (Fig. 3e). Similarly, significant loss of GAG contents was observed in DMM mice but increased after the treatment of PTH (1-34) (Fig. 3f). Moreover, the staining of Alican Blue and Masson implied similar results (Fig. S3). We further detected the expression of MMP13, COLX, and ADAMTS5 which are key factors for affecting the development of OA. The upregulated expression of MMP13, COLX, and ADAMTS5 were noticed in the DMM mice but they were downregulated after the treatment by PTH (1-34). Moreover, the expression of MMP13, COLX, and ADAMTS5 from the IHC study implied that the OA progression in both WT and OPG À/À mice could be delayed by PTH (1-34) (Fig. 4). Above all, the pathological change exhibited reducing OARSI score, restoring cartilage thickness, and increasing GAG contents after the administration of PTH  in both WT and OPG À/À DMM mice.

PTH (1-34) ameliorates the deterioration of the SCB
Representative images of Micro-CT were displayed in Fig. 5a. Both WT-DMM and OPG À/À -DMM mice had significantly lower BMD, Tb.N, Tb.Th, and BV/TV but higher SMI at the 8th and the 12th weeks ( Fig. 5b-f). The deteriorated microarchitecture was inversed and abnormal remodeling of the SCB was inhibited only in the WT-DMM þ

The effects of PTH (1-34) are associated with the PTH1R/OPG/ RANKL/RANK signaling
According to the IHC study in Fig. 6, the expression of PTH1R and RANK were up-regulated in both WT-DMM þ PTH (1-34) mice and OPG À/À -DMM þ PTH (1-34) mice significantly (Fig. 6a and e and Fig. 6d and h). Compared with the OPG À/À mice, the expression of OPG was only upregulated in the WT-DMM þ PTH (1-34) mice. Compared with the WT-DMM mice, the expression of RANKL was downregulated remarkably. Particularly, no significant differences of the OPG and RANKL were detected in the OPG À/À mice and they responded little to the treatment of PTH  in the OPG À/À -DMM þ PTH (1-34) mice ( Fig. 6b and f,  Fig. 6c and g). The IHC study implied that the therapeutic effect of PTH  was associated with the expression of PTH1R/OPG/RANKL/ RANK signaling.

The proliferation and the synthesis of ECM were promoted by PTH (1-34) in vitro
Similar dead cells (red dots) were observed according to the Live/ Dead staining (Fig. 7a and b), which showed no significant cytotoxicity in the current concentrations. Neoplastic cells labeled with the EdU (red) implied that the PTH (1-34) at 0.5 nM (130%), 1 nM (190%), and 3 nM (150%) would promote the proliferation (Fig. 7c and d). The MMP13 expression was down-regulated while COLII and SOX9 expression were up-regulated at protein level at the concentration of 1 nM (Fig. 7e-i). What's more, the expression of proteins including OPG/RANKL/RANK were checked. We found that only the RANK was expressed (Fig. S4) but no significant difference was detected after the stimulation of PTH . Above all, PTH (1-34) could increase the proliferation and enhance the ECM anabolism in the ATDC5 chondrocytes. The in vitro study implied that the chondro-protective effect of PTH (1-34) might be independent of OPG, which is in consistent with the in vivo study.

The catabolism of ECM and secretion of inflammatory cytokines were inhibited by PTH (1-34) in vitro
The viability of chondrocytes induced by IL-1β (10 ng/ml) might be due to the increased apoptosis but then it was increased after PTH (1-34) treated (Fig. 8a). Both the secretion of IL-6 and TNF-α were found to upregulated in the IL-1β-induced-ATDC5 chondrocytes but they were downregulated after the PTH (1-34) treated (Fig. 8b and c). The anabolic mRNA expression (AGC, COLII, and SOX9) would be up-regulated (Fig. 8d-f) while the ECM catabolism (MMP13, COLX, and ADAMTS5) decreased after the stimulation of PTH (1-34) (Fig. 8g-i). The adverse effect caused by IL-1β were significantly inversed after treatment by PTH (1-34).

Discussion
In the present study, the administration of PTH (1-34) exhibits a potential effect on preventing KOA progression by ameliorating the degeneration and wear of cartilage, restoring the thickness of cartilage, retaining the content of GAG, and affecting the expression of PTH1R/ OPG/RANKL/RANK signaling in vivo. What's more, PTH (1-34) exhibits a  KOA is a progressive disease but the currently available treatments fail to alleviate the deterioration effectively [22]. Poor muscle function and physical performance are strong predictors relevant to symptoms clinically while the control of risk factors is essential for delaying the progression of KOA [23]. Strategies to develop disease-modifyingtreatments and newly established KOA medications have attracted much attention, which would reduce the costs and benefit patients. PTH  was used to treat OP, improve healing of the bone fracture widely [24], and emerge as a promising agent for KOA treatment. Therefore, PTH (1-34) might be a medication for KOA but the potential mechanism of its pharmacological effects remains to be explored.
PTH  was demonstrated to be effective in preventing the degeneration of cartilage. The wear of the cartilage in all the DMM mice was ameliorated after the treatment of PTH . Furthermore, the reduced cartilage thickness and the loss of GAG contents were observed in the DMM mice in the 4th, 8th, and 12th weeks, which implied the progression of KOA evolved with time. The chondro-regenerative effect after the administration of PTH (1-34) is correlated with the regeneration, an increase in the cartilage thickness, and enhancement of the GAG. ness, and (f) GAG contents were displayed as well. Each column represents the mean AE SD (n ¼ 6). *, P < 0.05; **, P < 0.01 compared with the WT-sham group. #, P < 0.05; ##, P < 0.01 compared with the WT-DMM group. △, P < 0.05; △△, P < 0.01 compared with the OPG À/À -sham group. ☆, P < 0.05; ☆☆, P < 0.01 compared with the OPG À/À -DMM group. Figure 4. Expression of MMP13, COLX, and ADAMTS5 from the study joints. The expression of (a) MMP13, (b) COLX, and (c) ADAMTS5 could be detected in cartilage. The DMM mice with higher expression but the DMM þ PTH (1-34) mice with lower expression, which implied that PTH  could alleviate the cartilage damage in both WT and OPG À/À mice. (d) to (f) The relative expression of MMP13, COLX, and ADAMTS5. Each column represents the mean AE SD (n ¼ 6). *, P < 0.05, **, P < 0.01, significant difference compared with the sham group.
The articular cartilage lacks intrinsic repair capacity, moreover, the restoration of damaged tissue was hard to duplicate its original structure and composition [25]. PTH (1-34) exhibits cartilage protection and chondrocyte regeneration, which was in line with a previous study [26]. Possible mechanisms might be the inhibition of hypertrophic differentiation or the reduction of apoptosis in chondrocytes [27]. The HE staining from primary organs and no significant difference was found, which was neither specific nor sensitive compared with blood chemistry examination but supported the biosafety of PTH  in mice [28]. No malignancy pathology was observed varied from the previous report, which might be possible due to the species-specific [29]. OPG is a member of the TNF receptor superfamily in maintaining the SCB and articular cartilage.
Intra-articular injection of PTH (1-34) alleviated KOA by directly protecting cartilage [30] rather than affecting the SCB [31]. Liu et al. proved that OPG played an important role in maintaining homeostasis of articular cartilage in the femoral head [32]. The deficiency of the OPG caused thinner cartilage and extensive remodeling of the SCB, which was approved in our current study. The supplementation of the PTH  improved the thickness and GAG of cartilage as well as the enhanced synthesis of ECM in both WT and OPG À/À mice, which implied that the therapeutic roles of PTH (1-34) might depend on OPG.
PTH (1-34) exhibited a significantly protective effect on SCB. KOA is characterized by cartilage degeneration while the deterioration and abnormal remodeling of SCB affect the KOA progression as well [33]. The   micro-architecture of the SCB was clear and distinguishable in the WT mice, which was noticed to be markedly different in the OPG À/À mice. PTH (1-34) benefited the OPG À/À mice in preventing the degeneration of cartilage rather than the alleviation of SCB, which implied that PTH (1-34) might be associated with the PTH1R/OPG/RANK/RANKL signaling. Moreover, the therapeutic eccect of PTH  in ameliorating the deterioration of SCB might be independent on OPG.
The OPG/RANKL/RANK pathway was identified as a critical regulator of SCB remodeling and relevant gene deficiencies, or mutations will result in abnormal bone metabolism [34]. Remodeling of osteoclasts and osteoblasts was repeated continuously in bone. Osteoblasts mediate OPG to prevent the resorption of bone while the deficiency of OPG generates the onset of bone loss or even OP [35]. The historical evaluation indicated that damage of the cartilage degenerated over time, which was in line with that the deterioration of SCB being closely associated with OA progression [36]. Moreover, the deterioration of the SCB would be a risk factor for KOA progression but the amelioration of the SCB might provide a valuable basis for the treatment of KOA [5]. The OPG À/À mice with serious abnormal situations of the bone and the SCB had no response to the treatment of PTH . According to our investigation, PTH (1-34) might be dependent on OPG to alleviate the deterioration of SCB and inhibit the KOA progression. Thus, pathological change in SCB might serve as a mechanism in the OPG À/À mice. Moreover, a conditional knockout approach would be very useful to dissect tissue specific roles of OPG in bone and cartilage.
Consistent with the in vivo study, strategies to inhibit hypertrophic maturation of chondrocytes and enhance the synthesis of ECM represent potential new therapeutic modalities, an excellent chondro-protective effect had been approved in vitro studies. The chondrocytes are responsible for the maintenance of the articular cartilage [37], which would . Each column represents Each column represents the mean AE SD from 3 repeated and independent experiments. *, P < 0.05, **, P < 0.01. provide a valid strategy for the prevention and treatment of KOA [38]. Previous investigations implied that PTH (1-34) could enhance the proliferation of chondrocytes [39], which was replenished in our current study. PTH (1-34) increased proliferation and enhance ECM anabolism of chondrocytes by upregulating COLII and SOX9 but downregulating MMP13 at the protein level. Degeneration of the chondrocytes was observed after the stimulation with IL-1β, which might be associated with oxidative stress and apoptosis. Proinflammatory parameters are molecular characteristics of KOA and are responsible for ECM catabolism and cartilage degradation [40]. In particular, the anabolic genes are essential for the integrity of the ECM while the catabolic genes are involved in cartilage degeneration, which was consistent with the relevant protein expression. All adverse effects of chondrocytes caused by IL-1β were reversed after the stimulation by PTH (1-34), which possessed the desirable chondro-protective potential for inflammatory-induced chondrocytes degeneration. Moreover, the expression of OPG/RANKL/RANK were checked via WB and only the RANK expression was detected with no significant differences. The in vitro studies implied that the chondro-protective of PTH (1-34) might be independent on the OPG, which was in consist with the in vitro study.

Conclusion
In summary, we conclude that the PTH (1-34) effectively prevents KOA progression by alleviating degeneration of cartilage and ameliorating abnormal remodeling of SCB. The cartilage and SCB were alleviated in the WT mice, but the effect of cartilage-protection observed well in the OPG À/À mice. This implies that the PTH (1-34) may play excellent therapeutic role depends on the OPG in the SCB but not depend on the OPG in the cartilage. Moreover, the chondro-regenerative and chondroprotective characteristic features were well confirmed in vitro. The current study provides evidence that the PTH (1-34) might be a promising and potential medication for KOA.