HnRNPL inhibits the osteogenic differentiation of PDLCs stimulated by SrCl2 through repressing Setd2

Abstract Osteoporosis has been shown to intensify bone loss caused by periodontitis and both share common risk factors. One strategy utilized to manage the disease has been via the release of Sr ions by Strontium Ranelate having a direct effect on preventing osteoclast activation and promoting osteoblast differentiation. Previously we have developed and characterized porous Sr‐mesoporous bioactive glass (Sr‐MBG) scaffolds and demonstrated their ability to promote periodontal regeneration when compared to MBG alone. Our group further discovered a splicing factor, heterogeneous nuclear ribonucleoprotein L (hnRNPL), was drastically down‐regulated in periodontal ligament stem cells (PDLCs) stimulated by Sr through the activation of AKT pathway. Furthermore, hnRNPL restrained the osteogenic differentiation of PDLCs through down‐regulating H3K36me3‐specific methyltransferase Setd2. The goal of the present study was to investigate the mechanism of periodontal regeneration stimulated by Sr It was first found that the epigenetic mechanism of splicing factor hnRNPL participated in the osteogenesis processing of PDLCs stimulated by SrCl2. Meanwhile, the different role of hnRNPL and SET domain containing 2 (Setd2) may provide some implication of the treatment of periodontitis patients simultaneously suffering from osteoporosis.

effective and precision therapies are constantly sought as treatment options for the continuously rising number of patients with both periodontal disease and osteoporosis.
A variety of treatment options for periodontal bone defects healing have been proposed using novel tissue engineering scaffolds. 4 In recent years, mesoporous bioactive glass (MBG) has been utilized as a bioactive material with pore sizes ranging from 5 to 20 nm and a well-ordered mesoporous channel structure capable of promoting new bone formation in vivo. [5][6][7][8] Strontium Ranelate has been applied clinically to prevent osteoporosis and strontium has been shown to activate calcium sensing receptor (CaSR) and downstream protein phosphorylation to promote osteogenesis. 9 Recently our group developed a kind of MBG scaffold containing strontium and demonstrates its positive effective on promoting osteogenesis and periodontal regeneration in osteoporotic animal models. [10][11][12] Although strontium ions play an important role by inhibiting bone resorption and promoting bone formation through different signaling pathways, 13,14 the mechanism in the nucleus at the genetic level during periodontal regeneration stimulated by Sr ions remains unclear.
Alternative splicing is a kind of the regulatory mechanism in eukaryotic cells, whereby the pre-mRNA can be spliced into multiple mature mRNA due to the introns existing in the DNA, followed by translation to produce a variety of proteins. 15,16 HnRNPL is an important member in the family of ribonucleic protein splicing factors enriched in cell nucleus and plays an important role in many physiological and pathological processes. 17,18 HnRNPL contains four RRM domains, a proline-rich sequence separating RRMs 2 and 3 and a glycine-rich N-terminal domain. 19 This ubiquitous splicing-regulatory protein is critical for the oncogenesis such as osteosarcoma and the development of T cells. 20 It can also inhibit P53 to prevent DNA damage-induced apoptosis in embryonic stem cells. 21 Moreover, cluster analysis of genes regulated by hnRNPL showed genes containing hn-RNPL-regulated exons were associated with chromatin regulation, DNA modification and transcription, implied the epigenetic role in hnRNPL-related splicing process. 22 HnRNPL can interact with Set2 and is related to the retention of introns in eukaryotic cells. 22 We have previously shown the positive regulation of Setd2 in the process of osteogenesis stimulated by Sr. 23 However, whether hnRNPL can regulate the osteogenesis process is still lack of evidence.
In this study, we demonstrated that hnRNPL could play a role in the osteogenic differentiation of PDLCs stimulated by Sr by regulating Setd2. Following stimulation with Sr, the AKT pathway was activated and then repressed the expression of splicing factor hn-RNPL, ultimately promoted the osteogenic differentiation of PDLCs through Setd2 up-regulation.
The polyurethane sponges (20 ppi) were cleaned and completely immersed into this solution for 10 minutes, then transferred to a Petri dish to allow evaporation at room temperature for 12 hours. This procedure was repeated three times. Once the samples were completely dry, they were calcined at 700°C for 5 hours yielding the Sr-MBG scaffolds. MBG scaffolds without Sr (Sr/Ca/P/Si=0/15/5/80, molar ratio, named MBG) were prepared by the same method except for their Sr and Si contents.

| Animals and surgical procedures
Establishment of the osteoporotic rat models were performed as previous described 12,24,25   Two months after the osteoporosis model was established, periodontal fenestration defects (standardized with 2.8 mm in length, 1.4 mm in height and 0.5 mm in depth) were created as previous described. 12 Briefly, under general anesthesia, rats were subjected to bilateral extra-oral incision at the base of the mandible. The buccal mandibular bone overlying the first molar roots was removed to create a defect (2.8 mm in length, 1.4 mm in height and <0.5 mm in depth) using a size-4 round bur. The procedure was performed under an operating microscope to avoid perforation of intraoral mucosa.
The roots of the first molar were carefully denuded of their periodontal ligament, overlying cementum, and superficial dentin. The height was standardized to the width of the round bur (diameter 1.4 mm) and extended longitudinally to either side. Then, two scaffolds were gently implanted respectively into the bilateral periodontal fenestration defects (MBG in the left and Sr-MBG in the right defect). Following implantation of scaffolds, the muscle and the skin were repositioned and sutured separately. Postoperatively, penicillin (40,000 IU/mL, 1 mL/kg) was injected intramuscularly for 3 days.
Four weeks after surgery, the rats were sacrificed by sodium pentobarbital, following a protocol approved by the Institutional Animal Care and Use Committee (IACUC).

| Immuno-histochemical staining
The immune-histochemical staining was performed as previous described. 23 Periodontal bone regeneration samples were immersed in 10% EDTA changed twice weekly for 3 weeks until decalcifica-

| Viral transfection
Viral packaging and transfection was performed as previously described. 26 The target sequences for shRNA1 were: hnRNPL of humans:

| Alizarin red staining and quantification
Alizarin red staining was performed to determine the presence of extracellular matrix mineralization in vitro. PDLCs were fixed in 4% formaldehyde for 15 minutes after culture for 21 days and stained with 0.01% alizarin red solution (pH 4.2) at room temperature for 1 hours. For quantifying the relative amount of calcium, the stained nodes were dissolved using 200 μl of 1% cetylpyridinium chloride (Sigma, USA) per well at room temperature for 4 hours and then the OD value was measured at 560 nm.

| ALP staining and quantification
PDLCs were fixed in 4% formaldehyde for 15 minutes after being cultured for 7, 14 and 21 days and stained with ALP staining kit (Beyotime Biotechnology, China) at 37°C for 30 minutes.

| Protein extraction and western blotting
Protein extraction and western blotting was performed as previously described. 26 PDLCs were starved for 24 hours, then stimulated by SrCl 2 , collected with PBS, mixed with loading buffer and then heated at 95°C for 10 minutes for total protein denaturation. Nuclear and cytoplasmic proteins were extracted by Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime Biotechnology, China). The protein contents were determined by using a bicinchoninic acid (BCA) assay kit (Thermo Scientific).
Proteins were separated by SDS-PAGE then transferred to nitrocellulose membranes. The membranes were blocked using 5% non-fat milk or 5% BSA for 60 minutes at room temperature.

| Statistical analysis
GraphPad Prism software (GraphPad, San Diego, CA) was used for the statistical analyses. Two-way ANOVA and Students′ t test was used to assess differences between the control and test groups. P-values less than 0.05 were considered statistically significant.

| SrCl 2 in the concentration of 1 mmol/L promotes PDLCs osteogenic differentiation without influencing proliferation
We then investigated the mechanism of osteoblastic differentiation stimulated by Sr in vitro. To determine the optimal concentration of Sr, PDLCs were cultured in osteogenic differentiation media with SrCl 2 at various concentrations ranging from 0 to 3 mmol/L.
The results showed that the ALP activity in 0.01, 0.1 and 3 mmol/L groups were decreased after 7 days of induction (Figure 2A,C).
After 14 days of induction, the expression levels of ALP in the three groups were also declined ( Figure 2E) while the ALP activityand the expression levels of osteogenic markers such as ALP, OCN and BSP in 1 mmol/L group were all increased and highest among all groups ( Figure 2C,E). It was also observed that the influence of SrCl 2 to the calcification ability of PDLCs was dose-dependent when the concentration was less than 1 mmol/L. However, if the concentration was 3 mmol/L, it showed a negative effect on the calcification ability of PDLCs ( Figure 2B,D). Then we suspected if this effect was due to the proliferation of PDLCs, whereas the results showed no effect of the concentration of SrCl 2 on the proliferation of PDLCs ( Figure 2F).

| SrCl 2 promotes PDLCs osteogenic differentiation through AKT pathway
Strontium was shown to activate calcium sensing receptor (CaSR) and downstream protein phosphorylation and to promote osteogenesis. 9 AKT is one of the most important protein kinases downstream of CaSR. Furthermore, AKT pathway was involved in Sr induced osteogenesis and angiogenesis. 27

| HnRNPL represses PDLCs osteogenic differentiation stimulated by Sr
Since hnRNPL was expressed in low levels in the Sr-MBG group in vivo, we then examined the expression of hnRNPL in PDLCs.
Western blot results showed hnRNPL was significantly decreased from 2 hours after SrCl 2 stimulation both in total and nuclear protein ( Figure 3A,C). With the pretreatment of AKT inhibitor MK2206, Sr failed to repress the expression of hnRNPL in PDLCs ( Figure 3D), indicating that the repression of Sr to hnRNPL was AKT-associated.
To further investigate the role of hnRNPL in the process of PDLCs osteogenic differentiation stimulated by Sr, we silenced hnRNPL in PDLCs with two shRNAs ( Figure 4A). It was found that mineralization, ALP activity and the expression levels of osteogenic-related markers ALP, Runx2 and OCN were all increased after interfering the expression of hnRNPL with or without Sr stimulation ( Figure 4B-H).

| HnRNPL can inhibit the expression of Setd2 to suppress PDLCs osteogenic differentiation stimulated by Sr
Setd2-mediated H3K36me3 has also previously been shown to be involved in the regulation of alternative splicing, 28 but whether it is associated with the splicing process mediated by hnRNPL is remained unknown. Immunohistochemistry results showed that the  Figure 1G-O). It was revealed that the expression of Setd2 was significantly increased from 2 hours after SrCl 2 stimulation (time when hnRNPL decreased).
When pre-treated with the AKT inhibitor MK2206, the expression of Setd2 failed to be promoted by Sr in PDLCs ( Figure 3D). More importantly, when hnRNPL was silenced, the expression of Setd2 was significantly increased in PDLCs, with the highest expression observed following Sr stimulation ( Figure 4I). These results implied that the AKT pathway promoted Setd2 expression through repressing downstream hnRNPL.
However, whether Setd2 played a positive role similar to the AKT pathway remained unknown. We therefore investigated the role of Setd2 during Sr-induced osteogenic differentiation of PDLCs by Setd2 knock-down or over-expression ( Figure 5A,B). The results showed that mineralization, ALP activity and the expression of osteogenic-related markers ALP, Runx2 and OCN were decreased in the knock-down group in PDLCs even when stimulated with SrCl 2 (Figure 5C-G). In contrast, over-expression of Setd2 in PDLCs showed that SrCl 2 can further enhance osteogenic differentiation ( Figure 5C-G).

| D ISCUSS I ON
Periodontal tissue regeneration has been a major problem for clinicians for a long time, of which periodontal bone regeneration is the most important part. Periodontal tissues include gingiva, cementum, periodontium and alveolar bone. Osteoporosis is a disease characterized by reduced bone density, causes more complex treatment and poor prognosis of periodontal defects. 29 The molecular mechanisms underlying this bone condition are related to estrogen (G) Expression level of osteogenic genes, Runx2, ALP, OCN of PDLCs after Setd2 knockdown or over-expression and induction towards osteogenic differentiation with or without SrCl 2 stimulation. *P < 0.5; **P < 0.01; ***P < 0.001 deficiency and overproduction of some cytokines. 30 Strontium Ranelate is believed to be effective in osteoporosis treatment, and many studies have confirmed that strontium ions can prevent osteoporosis by regulating osteoblasts and osteoclasts. In the treatment of periodontal tissue destruction, there is a need to discover key factors of bone metabolism as potential targets for the treatment and management osteoporosis.
Mesoporous bioactive glass (MBG) with well-ordered mesoporous channels has been shown to facilitate osteogenesis and bioactivity in vivo. [5][6][7][8]31 In addition, the ions and degradation products of MBG have been reported to effectively promote the expression of osteogenic genes and proteins. Various groups have now combined biomaterials with strontium (Sr) ions in order to improve their biological activity and osteogenic potential. 14,32 Our research group recently demonstrated that the MBG scaffolds combined with Sr, efficiently promoted periodontal regeneration in rat periodontal fenestration defects even in severely osteoporotic animals. 10 In the present study, immunohistochemistry staining of Runx2 was performed to detect the early osteogenic differentiation ability of PDLCs around these biomaterials. The results showed that the larger proportion of

Runx2-positive cells was detected in the Sr-MBG group and Masson
staining also showed more bone formation in the Sr-MBG group.
Strontium was previously shown to activate calcium sensing receptor (CaSR) and downstream protein phosphorylation to promote osteogenesis 9 . Activated CaSR can promote the activation of phosphatidyl inositol -3-hydroxylase (PI3K), sequentially catalyse substrate diphosphoinositol to produce triphosphoinositol.
Triphosphoinositol can following activate AKT through promoting phosphoinositol depending protein kinase (PDK) activation. 33 To investigate the mechanism of the effect of Sr on PDLCs, we examined one of the classical protein phosphorylation pathway, AKT pathway. As expected, Sr promoted PDLCs osteogenic differentiation through activation of AKT pathway ( Figure 6). AKT is a kind of serine/threonine protease containing PH domain, which can combine with phosphatidyl inositol triphosphate (PIP3) on the inner side of cell membrane. After activated by PDK, AKT leaves the cell membrane and enters the nucleus, then promotes the expression of downstream genes. 34 Therefore, we tested the expression levels of  35,36 The change of hnRNPL expression after drug treatment in osteosarcoma cells indicates that hnRNPL was involved in the physiological activity of bone-related cells, 37 and the alternative splicing mediated by hnRNPL is associated with protein phosphorylation. 38 We therefore investigated the expression and effect of hnRNPL in PDLCs during the osteogenic differentiation with Sr stimulation. The results showed a negative role of hnRNPL in this process and also suppressed by AKT pathway.
Recently one study showed histone H3K36 methylation affected alternative splicing in plants with H3K36me3 being involved in the regulation of alternative splicing. 28 The methylation of H3K36 was also proposed to be associated with osteogenesis. 39,40 We then de- Setd2 by increasing the production and recruitment of inhibiting or repressor proteins to suppress the expression of Setd2.
In conclusion, we provide rational describing the role of hnRNPL and Setd2 during periodontal regeneration. It is therefore possible to target hnRNPL and Setd2 as potential novel therapeutic targets for the treatment of periodontal tissue destruction most notably in osteoporotic patients.

ACK N OWLED G EM ENTS
This work was supported by the funds of the National Natural

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest.