Tyrosine‐protein phosphatase non‐receptor type 2 inhibits alveolar bone resorption in diabetic periodontitis via dephosphorylating CSF1 receptor

Abstract Tyrosine‐protein phosphatase non‐receptor type 2 (PTPN2) is an important protection factor for diabetes and periodontitis, but the underlying mechanism remains elusive. This study aimed to identify the substrate of PTPN2 in mediating beneficial effects of 25‐Hydroxyvitamin D3 (25(OH)2D3) on diabetic periodontitis. 25(OH)2D3 photo‐affinity probe was synthesized with the minimalist linker and its efficacy to inhibit alveolar bone loss, and inflammation was evaluated in diabetic periodontitis mice. The probe was used to pull down the lysates of primary gingival fibroblasts. We identified PTPN2 as a direct target of 25(OH)2D3, which effectively inhibited inflammation and bone resorption in diabetic periodontitis mice. In addition, we found that colony‐stimulating factor 1 receptor (CSF1R) rather than JAK/STAT was the substrate of PTPN2 to regulate bone resorption. PTPN2 direct interacted with CSF1R and dephosphorylated Tyr807 residue. In conclusion, PTPN2 dephosphorylates CSF1R at Y807 site and inhibits alveolar bone resorption in diabetic periodontitis mice. PTPN2 and CSF1R are potential targets for the therapy of diabetic periodontitis or other bone loss‐related diseases.

that 25(OH)2D 3 binds to vitamin D receptor (VDR) to mediates its biological effects, the detailed mechanism of action needs further investigations.
Tyrosine-protein phosphatase non-receptor type 2 is essential to bone homoeostasis and development because the number of osteoclast and bone resorption increase in PTPN2 knockout mice. 13 PTPN2 is a negative regulator of many signalling pathways by dephosphorylating the corresponding substrates, including non-receptor protein tyrosine kinases, receptor protein tyrosine kinases and the Src family kinases, either in the nucleus or in the cytoplasm. [14][15][16][17] Among the signalling molecules, receptor activator for nuclear factor-κ B Ligand (RANKL) and colony-stimulating factor-1 (CSF-1) play an important role in osteoclastogenesis. 18,19 CSF-1 binds to the unique receptor tyrosine kinase CSF1R on osteoclasts to activate signalling cascade and induce the expression of osteoclast lineage-related genes, including calcitonin receptor, 3-integrin 7, cathepsin K (CATK) and tartrateresistant acid phosphatase (TRAP). 20,21 Interestingly, PTPN2 could negatively regulate CSF1R signalling and mononuclear phagocyte development in hematopoiesis. 22 However, it remains unclear whether PTPN2 regulates CSF-1/CSF1R signalling in osteoclastogenesis.
Therefore, in present study, we aimed to investigate the crosstalk of PTPN2 and CSF-1/CSF1R signalling in the regulation of alveolar bone resorption in the condition of diabetic periodontitis.

| Animals
Four-week-old male C57BL/6 wild-type mice were purchased from the Experimental Animal Laboratory of Shandong University and maintained in the cage individually with an artificial light cycle at room temperature (23 ± 2℃) and humidity (55 ± 5%). Mice were given free access to food and tap water. All of the experiments were performed in accordance with Guide for the Care and Use of Laboratory Animals of the National Research Council (the USA, 2011) and approved by the IRB of Shandong University (No. GD201608).
At 6-week old, the mice consumed high-fat food (48 kcal% fat) for 7 days and fasted overnight. Then the mice received intraperitoneal injection of streptozotocin (STZ, Sigma, dissolved in citric acid buffer, pH 4.5) for five consecutive days at a dose of 40 mg/kg (bodyweight). After STZ injection the mice continued high-fat food.
To induce periodontitis, the mice were infected with Porphyromonas gingivalis W50 (ATCC: 53978) at the age of 8 weeks as described previously. 25 Control mice received 100 μL of PBS with

| Bodyweight and fasting blood glucose measurement
The body weight and fasting blood glucose of the mice in each group were monitored every 2 weeks. Briefly, after fasting for 10 hours, the tail veins of the mice were cut, and the blood was collected, and fasting blood glucose levels were measured using a glucometer (OneTouch Glucometer).

| Stereotaxic injection of virus
On the fourth day after the infection with Porphyromonas gingivalis, the mice were anesthetized by intraperitoneal injection of 5% chloral hydrate. 2 × 10 9 adeno-associated virus (AAV) shPTPN2 viral particles (Santa Cruz Biotechnology) was injected about 0.3-0.5 mm above the gingival margin of the maxillary molars on the right and left palatal aspects. The injection was repeated for seven consecutive days.

| Histological analysis
After the sacrifice, the mandibular jaws were gathered and boiled in water for 15 minutes to remove the soft tissues. The alveolar bone loss of the second molar was assessed by a stereomicroscope (×15) with an attached digital camera (Leica MZ FLIII, Germany). Next, the maxillary jaws were fixed by 4% paraformaldehyde for 24 hours and then decalcified by 10% EDTA for 25 days. The samples were dehydrated using 50%, 70%, 95% and 100% alcohol, embedded in paraffin and then cut into 4-μm sections for haematoxylin-eosin (H&E) staining.

| Cell isolation and culture
Primary gingival fibroblasts were isolated from mandibular tissues obtained from diabetic periodontitis mice. Briefly, the gingival tissue was rinsed six times in PBS to diminish microbial contaminants and then finely minced with scalpels in a 35-mm dish (Corning, Corning Incorporated) containing 2 mL of DMEM supplemented with 10% foetal bovine serum and 1% penicillin/streptomycin. The gingival fibroblasts were allowed to explant from the minced tissue in complete medium with changes every 3 days. The cells were seeded at a density 2 × 10 4 cells/cm 2 and cultured to 80% confluency for the following experiments.

| Western blot analysis
The cells or tissue were lysed in lysis buffer (150 mmol/L NaCl, 20 mmol/L Tris-HCl, 10% glycerol, 1% Triton X-100, 1 mmol/L Na 3 VO 4 , 0.1 mmol/L PMSF, pH 7.4) on ice for 20 minutes and centrifuged at 20 000 × g for 20 minutes. The supernatants were subjected to Western blot analysis using appropriate primary and secondary antibodies for xx.

| Pull-down analysis
The lysates of primary gingival cells were mixed with 50 μmol/L Biotin-N 3 and 25(OH)2D 3 -Probe or 25(OH)2D 3 . Then 2 mmol/L TCEP, 2 mmol/L CuSO 4 and 200 μmol/L TBTA were added to the lysates and incubated with streptavidin agarose (Invitrogen) at 4°C overnight. On the following day, the beads were washed six times with lysis buffer on ice after UV irradiation for 20 minutes, then the immunoprecipitations were eluted using low pH elution buffer at 4°C for 15 minutes. 1/20 volume of 1 mol/L Tris-HCl (pH 9.4) was used to neutralize the acid elution, and the eluted bead-bound proteins were separated by SDS-PAGE and detected by silver staining or Western blot analysis.

| Immunofluorescence
Primary gingival cells were grown to 50%-70% confluence and treated with 400 ng/mL 25(OH)2D 3 or DMSO for 24 hours. After washing with PBS, the cells were fixed for 10 minutes in freshly prepared 4% paraformaldehyde and then permeabilized for 10 minutes in 0.25% Triton X-100. The cells were blocked for 1 hour with 2 mg/ mL BSA at room temperature and then incubated with antibody for PTPN2-45kD or CSF1R at 4°C overnight. The cells were then incubated with appropriate secondary antibody for 1 hour at room temperature in the dark. Finally, the nuclei were stained by DAPI at 37°C for 30 minutes in the dark. The stained cells were observed using confocal microscope system (CSU10, Yokogawa Electric Co) with an inverted microscope (IX-71, Olympus Optical Co., Ltd) and a CoolSNAP-HQ camera (Roper Industries).

| Statistical analysis
Data were expressed as mean ± SD and analysed using SPSS 12.0 statistical analysis package (SPSS Inc). The comparison was analysed using t test or ANOVA P < .05 was accepted as significant difference.

| CSF1R is the substrate of PTPN2
Since PTPN2 inhibits inflammation by negatively regulating classic inflammatory JAK/STAT pathway, we examined whether anti-inflam-  exhibit the same protective effect on bone resorption as 25(OH)2D 3 , implying that some unidentified PTPN2 substrates may be responsible for the effects of 25(OH)2D 3 on bone resorption ( Figure 4D).
To identify the substrates of PTPN2-45kDa in the presence of 25(OH)2D 3 , we transfected primary gingival fibroblasts with either PTPN2-WT or PTPN2-45D182A construct and then stimulated the cells F I G U R E 3 25(OH)2D 3 induces PTPN2 phosphatase activity. A, Primary gingival fibroblast lysates were incubated with Biotin or 25(OH)2D 3 -probe and then exposed to UV, after pull-down the precipitates were resolved in SDS-PAGE and stained by Coomassie Blue. The indicated bands were examined by mass spectrometry and Western blot analysis. B, AAV vector encoding PTPN2 shRNA was injected into the gingival tissue of diabetic periodontitis mice, which were then treated with 25(OH)2D 3 and alveolar bone loss was evaluated. C, Primary gingival fibroblasts were treated with 25(OH)2D 3 , and PTPN2 protein expression level was analysed by Western blot. D, Primary gingival fibroblasts were treated with 25(OH)2D 3 , stained by PTPN2 antibody (green), and then counterstained with DAPI (blue). E-F, Primary gingival fibroblasts were treated with 25(OH)2D 3 , and phosphatase activity of PTPN2-45kDa was measured. Data are mean ± SD (n = 8). **P < .01, ***P < .001 compared to vehicle group. # P < .05, ##P < .01, ###P < .001 compared to IFN-γ group with 25(OH)2D 3 . The bands only appeared with PTPN2-45D1892A but not PTPN2-WT were cut down, and CSF1R was identified ( Figure 5A).
GFP-tagged PTPN2-45D182A appeared in the nuclei in basic condition, after 25(OH)2D 3 stimulation it appeared in the cytoplasm and was co-localized with CSF1R ( Figure 5D). To determine whether PTPN2 could directly dephosphorylate CSF1R, we detected the phosphorylation status of CSF1R by altering PTPN2 expression level. PTPN2 overexpression or 25(OH)2D 3 stimulation led to significantly decreased and shortened phosphorylation of CSF1R, while knock-down of PTPN2 led to strong and constitutive phosphorylation of CSF1R ( Figure 5E). Taken together, these results confirm that PTPN2 directly dephosphorylates CSF1R.

| Tyrosine 807 of CSF1R is the main site for the dephosphorylation by PTPN2
Next, we explored the specific tyrosine sites on CSF1R that are the ( Figure 6F). Collectively, these results indicate that Y807 at CSF1R is the main site being dephosphorylated by PTPN2 in response to 25(OH)2D 3 .

| D ISCUSS I ON
In this study, we employed a variety of approaches to elucidate novel mechanism of the action of 25(OH)2D 3 on periodontitis and made important findings as follows: first, we synthesized photo-affinity probe of 25(OH)2D 3 , which effectively reduced alveolar bone loss and inhibited inflammatory cytokines secretion in mice with diabetic periodontitis. Second, we identified PTPN2-45kD as a target of 25(OH)2D 3 and found that PTPN2 mediated the effects of 25(OH)2D 3 on alveolar bone loss not through JAK/STAT pathway.
Third, we demonstrated that PTPN2 directly dephosphorylated its substrate CSF1R on Y807 site.
Clinical and epidemiological data suggest that people with diabetes are susceptible to periodontitis, which indicates that diabetes is a considerable risk factor for periodontitis. 27-29 25(OH)2D 3 has been proven to reduce inflammation and bone resorption in periodontitis, but the molecular mechanisms need further investigation.
Using photo-affinity probe of small molecules to explore target protein is by far the most commonly used method to reveal the mechanism of action. In this study, we successfully synthesized We performed affinity pull-down analysis and identified that 25(OH)2D 3 could directly bind PTPN2. There exist two forms of PTPN2: a nuclear 45-kDa form (PTPN2-45kDa) and an endoplasmic reticulum targeted 48-kDa form (PTPN2-48kDa). In response to external stimuli, PTPN2-45kDa translocates from the nuclei to the cytoplasm and plasma membrane, where it can dephosphorylate its substrates and regulate related signalling pathways. 30 We found that PTPN2 expression increased after stimulation by 25(OH)2D 3 , and PTPN2-45kD quickly translocated into the cytoplasm. In addition, PTPN2-45kDa phosphatase activity increased significantly after treatment with 25(OH)2D 3 . Therefore, 25(OH)2D 3 causes nuclear exit of PTPN2-45kDa and increases its enzymatic activity.
Tyrosine-protein phosphatase non-receptor type 2 regulates various cellular signalling pathways and biological processes by dephosphorylating its physiological substrates. JAK/STAT signalling pathway is one of known inflammation-related pathways that is negatively regulated by PTPN2. 30  F I G U R E 4 25(OH)2D 3 inhibits the phosphorylation of STAT1, STAT3 and p38, but STAT inhibitor has no effect on bone resorption. A-C, Primary gingival fibroblasts were incubated with IFN-γ and/or 25(OH)2D 3 , and the phosphorylation of STAT1, STAT3 and p38 was detected by Western blot. D, Periodontal tissue and the mandibular molar of diabetic periodontitis mice were stained with H&E and then observed by the stereomicroscope. Red arrows represent the epithelial spikes. Yellow-dotted areas indicate horizontal alveolar bone loss. Data are mean ± SD, n = 8. *P < .05, ** P < .01, ***P < .001 compared to without IFN-γ and 25(OH)2D 3 group or vehicle group; ##P < .01, ### P < .001 compared to 25(OH)2D 3 group mutant interacted with CSF1R in the presence of the stimulus of 25(OH)2D 3 . Tyrosine 544, 559 and 807 on CSF1R are required for the regulation of macrophage proliferation. 23 To identify which tyrosine residues on CSF1R are the targeting sites of PTPN2, we detected the phosphorylation status of these residues by using the particular phosphorylation antibodies and knock-down of PTPN2.
Our results indicate that Y807 residue of CSF1R is the main site to be dephosphorylated by PTPN2 in response to 25(OH)2D 3 .
Interestingly, CSF1R was reported to regulate osteoclast formation and function, and thereby is involved in bone resorption. 23 Our findings suggest that PTPN2 might protect bone resorption by recognizing CSF1R and dephosphorylating CSF1R at Y807 site. Nevertheless, further mechanistic studies are needed to reveal the role of Y807 dephosphorylation in the regulation of bone resorption.
In conclusion, to our knowledge this is the first study to demonstrate that PTPN2 dephosphorylates CSF1R at Y807 site and inhibits alveolar bone resorption in diabetic periodontitis mice. Our results suggest that PTPN2 and CSF1R are potential targets for the therapy of diabetic periodontitis or other bone lossrelated diseases.

CO N FLI C T O F I NTE R E S T
All authors declare no conflict of interest.
F I G U R E 6 Y807 is the main site on CSF1R dephosphorylated by PTPN2.

AUTH O R CO NTR I B UTI O N S
XX designed the study, DZ, YJ, DS, ZZ, CZ, LD performed the experiments and analysed the data. All authors participated in writing the manuscript and approved the manuscript.

DATA AVA I L A B I L I T Y
Data are available upon request.