Thirty patients who underwent anterior cervical decompression surgery participated in this study. Fifteen patients were identified in this cohort as having OPLL, while the remaining fifteen were diagnosed with cervical disk herniation without presenting with OPLL. Seven patients exhibited the segmental type, six exhibited the local type, and two presented with the mixed type of OPLL. Proximal longitudinal ligament tissue samples were procured and preserved promptly in liquid nitrogen. For each participant, written informed consent was obtained, and the research was approved by the Ethics Committee of Changzheng Hospital in Shanghai, with an ethical approval number of 2021MS13.

We isolated primary ligament fibroblasts from collected posterior longitudinal ligament tissues using a protocol adapted from prior research [14]. After removing tissues from non-calcified areas, they were minced, cleansed, and incubated in Dulbecco’s Modified Eagle’s Medium (DMEM) (HyClone, SH30022.LS, South Logan, UT, USA) enriched with 10% fetal bovine serum (FBS) (Thermo Fisher Scientific, A5669801, Waltham, MA, USA) at 37°C. To induce osteogenesis, 10 nM dexamethasone (Sigma-Aldrich, D4902-25MG, St. Louis, MO, USA), 25 μg/mL ascorbic acid (Sigma-Aldrich, A8960-5G, St. Louis, MO, USA), and 10 mM β-glycerophosphate (Sigma-Aldrich, G9422, St. Louis, MO, USA) were added to the culture medium. The cultures were then incubated for 14 days. All utilized cells in this study were free from mycoplasma contamination.

RNA was isolated from cells or tissues using TRIzol reagent (Invitrogen, 15596026CN, Carlsbad, CA, USA) 48 h after transfection. The RNA was transcribed into complementary DNA (cDNA) using a Reverse Transcription Kit (Qiagen, RT31-020, Hilden, North Rhine-Westphalia, Germany). The Green Premix Ex Taq II (TaKaRa, RR820Q, Osaka, Japan) was used for quantitative PCR on Step One Plus Real-Time PCR System (Applied Biosystems, 4376600, Foster City, CA, USA). The 2^−ΔΔCt method was used to quantify gene expression levels, with normalization to GAPDH. The primer sequences of RNAs were shown as follows: human DAPK2: Forward: 5′-TGCAGCCAAGTTCATCAAGAAGCG-3′, Reverse: 5′-ACACTAGCTCAAGGATGAGCACCA-3′; mice DAPK2: Forward: 5′-TCCTGGATGGGGTGAACTAC-3′, Reverse: 5′-CAGCTTGATGTGTGGAA-3′; human ALP: Forward: 5′-GCCTGGATCTCATCAGTATTTGG-3′, Reverse: 5′-GTTCAGTGCGGTTCCAGACAT-3′; human COL1A1: Forward: 5′-GGGTCTAGACATGTTCAGCTTTGTG-3′, Reverse: 5′-ACCCTTAGGCCATTGTGTATGC-3′; human OSX: Forward: 5′-CTCTCTGCTTGAGGAAGAAG-3′, Reverse: 5′- GTCCATTGGTGCTTGAGAAG-3′; human OCN: Forward: 5′-GGCGCTACCTGTATCAATGG-3′, Reverse: 5′-GTGGTCAGCCAACTCGTCA-3′; human RUNX2: Forward: 5′-CCGGGAATGATGAGAACTA-3′, Reverse: 5′-ACCGTCCACTGTCACTTT-3′; human LC3: Forward: 5′-GAAGTTCAGCCACCTGCCAC-3′, Reverse 5′-TCTGAGGTGGAGGGTCAGTC-3′; human p62: Forward: 5′-GTACCAGGACAGCGAGAGGAA-3′, Reverse: 5′-CCCATGTTGCACGCCAAAC-3′; human Beclin-1: Forward 5′-ATACTGTTCTGGGGGTTTGCG-3′, Reverse 5′-GTCTCTCCTTTTTCCACCTCTTC-3′; human GAPDH: Forward: 5′-AGAAGGTGGTGAAGCAGGCATC-3′, Reverse: 5′-CGAAGGTGGAAGAGTGGGAGTTG-3′.

The cellular components were disrupted using radioimmunoprecipitation assay (RIPA) lysis buffer (Beyotime, P0013B, Shanghai, China) to obtain the entire protein content 48 h after transfection. Equal quantities of protein were subjected to separation using a 10% SDS-PAGE gel. Afterward, the proteins that had been separated were moved onto polyvinylidene fluoride (PVDF) membranes (Beyotime, P0021S-1L, Shanghai, China). Subsequently, the membranes were blocked and subjected to incubation with primary antibodies including ALP (Abcam, 1/1000, ab307726, Cambridge, MA, USA), COL1A1 (Abcam, 1/1000, ab138492, Cambridge, MA, USA), OSX (Abcam, 1/1000, ab209484, Cambridge, MA, USA), OCN (Abcam, 1/1000, ab133612, Cambridge, MA, USA), RUNX2 (Abcam, 1/1000, ab236639, Cambridge, MA, USA), LC3 (Abcam, 1/2000, ab192890, Cambridge, MA, USA), Beclin 1 (Abcam, 1/2000, ab207612, Cambridge, MA, USA), p62 (Abcam, 1/10000, ab109012, Cambridge, MA, USA), p-Raptor (Cell Signaling Technology, 1/2000, #2083, Massachusetts, USA), Raptor (Abcam, 1/1000, ab40768, Cambridge, MA, USA), p-Thr389 (Abcam, 1/500, ab60948, Cambridge, MA, USA), p70S6K (Abcam, 1/5000, ab32529, Cambridge, MA, USA), p-Thr46 (Abcam, 1/1000, ab278686, Cambridge, MA, USA), 4E-BP1 (Abcam, 1/2000, ab32024, Cambridge, MA, USA), p-mTOR (Abcam, 1/1000, ab109268, Cambridge, USA), mTOR (Abcam, 1/10000, ab134903, Cambridge, MA, USA), ULK1 (Abcam, 1/10000, ab177472, Cambridge, MA, USA), DAPK2 (Invitrogen, 1/1000, MA5-25084, Carlsbad, CA, USA), GAPDH (Abcam, 1/10000, ab181602, Cambridge, MA, USA) overnight at a temperature of 4°C. Afterward, the specimens were subjected to incubation with the secondary antibody (Abcam, 1/2000, ab172730, Cambridge, MA, USA) for a duration of one hour, followed by detection using ECL reagents (Sigma-Aldrich, B8522-1EA, St. Louis, MO, USA). The quantification of band density was performed using ImageJ software (National Institutes of Health, version 23.0, Bethesda, MD, USA).

Short hairpin RNA targeting DAPK2 (sh-DAPK2) and sh-negative control (NC), as well as pcDNA3.1 and pcDNA3.1-DAPK2, were obtained from Gene-Pharma (Shanghai, China). The sequence of sh-DAPK2#1 and sh-DAPK2#2 used in the study were GGAAACGGCUCACAAUCCA and GGAAUUUGUUGCUCCAGAA. Ligament fibroblasts were transfected with sh-DAPK2, sh-NC, pcDNA3.1, or pcDNA3.1-DAPK2 using Lipofectamine 3000 (Invitrogen, L3000150, Carlsbad, CA, USA) as per the guidelines provided by the manufacturer. Following 48 h of transfection, cells were subjected to selection with puromycin to isolate stable transfectants. The selection medium was refreshed every three days, and the selection process was continued for a minimum of three weeks to ensure the elimination of non-transfected cells.

The fibroblasts of the ligament (1 × 10⁵ cells) were cultured on 6-well plates, subsequently treated with 4% paraformaldehyde for fixation, and permeabilized using 0.1% Triton X-100. Then, the cells were blocked in 1% bovine serum albumin (BSA, Sigma-Aldrich, B2064, St. Louis, MO, USA) at room temperature for 30 min. Afterward, the cells were subjected to primary antibody incubation, wherein anti-Vimentin (Abcam, 1/1000, ab16700, Cambridge, MA, USA), anti-DAPK2 (Invitrogen, 1/1000, MA5-25084, Carlsbad, CA, USA), and anti-LC3 (Abcam, 1/1000, ab192890, Cambridge, MA, USA) antibodies were employed and kept at a temperature of 4°C overnight. Subsequently, the cells were subjected to culturing with a secondary antibody Abcam, 1/2000, ab172730, Cambridge, MA, USA) for a duration of one hour at room temperature. The nuclei were stained using DAPI (Sigma-Aldrich, D9542, St. Louis, MO, USA). The observation was conducted using a fluorescence microscope manufactured by Olympus IX70 in Japan.

Ligament fibroblasts (5 × 10³ cells) were incubated in a 12-well plate with osteogenic induction medium for two weeks. The cells were treated with 95% ethanol for 30 min to fix them, followed by staining with 0.1% ARS (ScienCell, Catalog #0223, San Diego, CA, USA) for an additional 20 min.

Ligament fibroblasts were gathered and subjected to centrifugation. Next, they were rinsed with PBS, fixed with 2.5% glutaraldehyde, postfixed with 1% osmic acid, and dehydrated with an acetone gradient. A microtome (Leica Biosystems, HI1220, Nussloch, Germany) was utilized to make the sections. Later, they were double-dyed with uranyl acetate and lead citrate, followed by observation by TEM (Hitachi, HT7800 series, Tokyo, Japan).

Sixteen 4-week-old BALB/c homozygous nude mice were acquired from Shanghai Slack Laboratory Animal Co., Ltd. (Shanghai, China) and housed in a specific pathogen-free (SPF) facility with a regulated 12-h light/dark cycle at a constant temperature of 24°C. The animal experiments were conducted by Cyagen (Suzhou, China) Biotechnology Co., Ltd., Suzhou, China. The study protocols were sanctioned by the Institutional Animal Care and Use Committee of Cyagen (Suzhou, China) Biotechnology Co., Ltd., with ethical approval number IACUC-2109025. Fibroblasts, which are prevalent in connective tissues throughout the human body, including skin, have shown capabilities for bone generation both naturally and under pathological conditions, as well as in laboratory-induced osteogenic differentiation [15,16]. Human ligament fibroblasts were genetically modified with sh-DAPK2 and selected using puromycin over a three-week period. The mice then received daily intraperitoneal injections of 4 mg/kg rapamycin (LC Laboratories, R-5000, Woburn, MA, USA) for three weeks, following established guidelines [17]. After a two-week osteogenic induction in culture, the cells were combined with Bio-Oss collagen scaffolds measuring 7 mm × 5 mm × 2 mm (Geistlich, GEWO GmbH, Baden-Baden, Germany) and incubated for two days. The cell-scaffold constructs were then implanted subcutaneously on the dorsal side of the mice. Six weeks post-implantation, the implants were harvested, fixed in 4% paraformaldehyde, and examined using a Quantum FX microCT scanner (PerkinElmer, Shelton, CT, USA). Image analysis and three-dimensional reconstructions were performed using Inveon Research Workplace software (Siemens Healthcare GmbH, Erlangen, Germany), focusing on the designated regions within the scaffolds to assess the bone volume/tissue volume (BV/TV) ratio and bone mineral density (BMD) from the micro-CT data.

The scaffold specimens were decalcified by 10% EDTA for 30 days. After that, specimens were dehydrated, embedded in paraffin, and cut into 5 μm sections. Next, they were dyed with hematoxylin and eosin. A microscope (Olympus IX70, Okaya, Japan) was employed for observation.

The sections were subjected to deparaffinization and rehydration with xylene and graded ethanol. Antigen retrieval was conducted with citrate buffer. Endogenous peroxidase was blocked with 3% H₂O₂. After blocking nonspecific antigen binding with 5% bovine serum albumin (BSA, Sigma-Aldrich, B2064, St. Louis, MO, USA), sections were cultured with the primary antibody against COL1A1 (Abcam, 1/1500, ab138492, Cambridge, MA, USA) or DAPK2 (Invitrogen, 1/50, MA5-25084, Carlsbad, CA, USA) overnight at 4°C, followed by culturing with secondary antibodies (Abcam, 1/2000, ab172730, Cambridge, MA, USA) for 2 h at room temperature. Subsequently, they underwent the process of dyeing using DAB (Solarbio, DA1010, Beijing, China) and were subsequently counterstained with hematoxylin. Ultimately, an Olympus IX70 microscope (Okaya, Japan) was employed for observation.

Statistical analysis was implemented using GraphPad Prism 8 software (GraphPad, San Diego, CA, USA). The group difference was analyzed with one-way ANOVA and Student’s t test. The results are described as the mean ± SD from three individual repeats. p < 0.05 was considered significant.

Initially, we assessed the levels of DAPK2 expression in ligament samples obtained from both individuals with OPLL and individuals without OPLL. The RT-qPCR analyses exhibited that the expression of DAPK2 was considerably greater in tissues obtained from patients with OPLL in comparison to the control samples (Fig. 1A). Subsequently, fibroblasts from the primary ligaments were isolated from these tissues. Upon microscopic examination, these cells displayed a morphology resembling fibroblasts, with a spindle-shaped structure (Fig. 1B). Vimentin, commonly used as a fibroblast marker due to its presence in fibroblasts as well as in endothelial and lymphoid cells, was also assessed [18]. Immunofluorescence staining demonstrated both elevated expression and co-localization of Vimentin and DAPK2 in the fibroblasts from the OPLL group (Fig. 1C–E). Further RT-qPCR analysis confirmed the pronounced upregulation of DAPK2 in the ligament fibroblasts from OPLL patients (Fig. 1F).

We investigated the function of DAPK2 in the ossification process within primary ligament fibroblasts. Initially, DAPK2 expression in fibroblasts from OPLL patients was downregulated using sh-DAPK2 transfection. Both RT-qPCR and Western blot analysis confirmed a substantial decline in DAPK2 mRNA and protein expressions post-transfection (Fig. 2A,B). Subsequently, using Alizarin Red S staining, calcium accumulation was evaluated. This analysis indicated a significant reduction in osteogenic differentiation following DAPK2 suppression (Fig. 2C). Additionally, the levels of ossification markers such as alkaline phosphatase (ALP), collagen type I alpha 1 chain (COL1A1), osterix (OSX), osteocalcin (OCN), and runt-related transcription factor 2 (RUNX2) were evaluated through RT-qPCR and Western blot, showing a lessening in both mRNA and protein levels due to sh-DAPK2 transfection, thereby supporting DAPK2’s role in enhancing ossification (Fig. 2D–E).

Conversely, overexpression of DAPK2 was achieved by transfecting ligament fibroblasts from OPLL patients with pcDNA3.1-DAPK2. Consequently, there was a substantial rise in both mRNA and protein levels of DAPK2, as evidenced by RT-qPCR and Western blot analyses (Fig. A1A,B). Further, ARS staining confirmed that DAPK2 overexpression markedly enhanced osteogenic differentiation (Fig. A1C). Additional tests demonstrated an increase in the levels of ALP, COL1A1, OSX, OCN, and RUNX2, indicating an upregulation of osteogenic markers following DAPK2 overexpression (Fig. A1D,E).

Considering the pivotal role of autophagy in OPLL pathogenesis [13], we investigated the influence of DAPK2 on autophagy. LC3, a widely recognized autophagy marker, was analyzed using fluorescence staining. The findings we obtained demonstrated that DAPK2 knockdown resulted in a substantial decrease in LC3 fluorescence intensity, suggesting DAPK2 knockdown decreased autophagy (Fig. 3A). Further analysis of autophagy-related proteins showed that DAPK2 depletion led to a lower LC3-II/I ratio and decreased Beclin1 levels, alongside an increase in p62 accumulation, indicating DAPK2 knockdown blocked autophagy (Fig. 3B). This was validated by RT-qPCR results, which showed that DAPK2 deletion down-regulated LC3 and Beclin1 transcripts and up-regulated p62 mRNA (Fig. 3C). Collectively, these findings implied that DAPK2 positively regulated autophagic flux in primary ligament fibroblasts.

The mTORC1 pathway is a key regulator of autophagy and its inhibition is a recognized inducer of autophagy [19]. To determine whether DAPK2 plays a promoting role in autophagy involves the regulation of mTORC1, we assessed the impact of DAPK2 inhibition on the activity of the mTORC1 complex. Western blot analysis revealed that DAPK2 silencing resulted in decreased phosphorylation of Raptor and unc-51 like kinase 1 (ULK1), coupled with enhanced phosphorylation of mTOR and its downstream targets p70S6K, and 4E-binding protein 1 (4E-BP1), which were hallmarks of increased mTORC1 activity (Fig. 4). These changes indicated that inhibition of mTORC1 was alleviated after DAPK2 knockdown, which was consistent with the reduction in autophagy. Hence, we concluded that DAPK2 promoted autophagy in ligament fibroblasts, at least in part, through inhibiting mTORC1 activity.

To further analyze the effect of DAPK2 in vivo, we carried out a heterotopic bone formation experiment on mice. Ligament fibroblasts of human stably silenced by DAPK2 were cocultured with Bio-Oss collagen scaffolds for two days. Next, mice were subjected to subcutaneous implantation with the scaffolds on the back for six weeks. Following this, the animals were euthanized eight weeks later, and micro computed tomography (micro-CT) scans were utilized to determine the bone mineral density (BMD) and bone volume/tissue volume (BV/TV). We revealed that the BV/TV rate and BMD were markedly reduced in mice implanted with sh-DAPK2 (Fig. 5A,B). Then, HE staining results illustrated that the formation of lamellar bone tissues was reduced in mice of the DAPK2 knockdown group in comparison to the control group (Fig. 5C). Moreover, IHC assays showed that the numbers of COL1A1-positive cells and DAPK2-positive cells were reduced by DAPK2 silencing (Fig. 5D). Thus, we confirmed that DAPK2 could facilitate ossification of ligament fibroblasts in vivo.

Moreover, we carried out rescue experiments on mice to verify whether mTORC1 inhibitor promotes bone formation in conjunction with DAPK2 knockdown. It was discovered that the reduced BV/TV rate and BMD in mice implanted with sh-DAPK2#1 were reversed after rapamycin treatment (Fig. A2A,B). Then, HE staining results illustrated that the reduce formation of lamellar bone tissues in mice of the DAPK2 knockdown group was offset after rapamycin treatment (Fig. A2C). Furthermore, IHC assays showed that the reduced numbers of COL1A1-positive cells and DAPK2-positive cells caused by DAPK2 silencing were enhanced after rapamycin treatment (Fig. A2D).