Targeting soluble epoxide hydrolase promotes osteogenic–angiogenic coupling via activating SLIT3/HIF‐1α signalling pathway

Abstract Type H vessels have recently been identified to modulate osteogenesis. Epoxyeicostrioleic acids (EETs) have an essential contribution to vascular homeostasis. However, whether increased EETs with soluble epoxide hydrolase (sEH) inhibitor TPPU enhance the coupling of angiogenesis and osteogenesis remains largely unknown. The effects of TPPU on cross‐talk between co‐cultured human umbilical vein endothelial cells (HUVECs) and human dental pulp stem cells (hDPSCs), and on long bone growth and calvarial defect repair in mice were investigated in vitro and in vivo. TPPU enhanced osteogenic differentiation of co‐cultured HUVECs and hDPSCs in vitro and increased type H vessels, and long bone growth and bone repair of calvarial defect. Mechanistically, TPPU promoted cell proliferation and angiogenesis, reclined cell apoptosis, and significantly increased CD31hiEMCNhi endothelial cells (ECs) and SLIT3 and HIF‐1α expression levels in co‐cultured HUVECs and hDPSCs. Knockdown of Slit3 in hDPSCs or Hif‐1α in HUVECs impaired the formation of CD31hiEMCNhi ECs and reversed TPPU‐induced osteogenesis. We defined a previously unidentified effect of TPPU coupling angiogenesis and osteogenesis. TPPU induced type H vessels by upregulating the expression of hDPSCs‐derived SLIT3, which resulted in the activation of ROBO1/YAP1/HIF‐1α signalling pathway in ECs. Targeting metabolic pathways of EETs represents a new strategy to couple osteogenesis and angiogenesis, sEH is a promising therapeutic target for bone regeneration and repair.

factors and provide a way to recruit inflammatory cells, fibroblasts and pre-osteoblasts/osteoclasts to the injured site. 4 The cross-talk between the endothelial cells (ECs) and surrounding cells maintains the molecular microenvironment for osteogenesis.
Type H vessels (CD31 hi EMCN hi ) have recently been identified to modulate osteogenesis. 5 They are mainly distributed near the epiphysis in the long bone and are densely surrounded by Runx2 (+) and Osterix (+) bone progenitor cells, which mediate the growth of the vascular system and provide pivotal signals for bone progenitor cells. 6,7 On the other hand, osteoclasts, osteoblasts and chondrocytes can secrete related factors to regulate EC proliferation and stability. 8,9 Several molecules have been found to mediate the development of type H vessels, which participate in the coupling of angiogenesis and osteogenesis, including HIF-1α, Notch, VEGF, and slit guidance ligand 3 (SLIT3). 5,10,11 Schnurri 3 (SHN3) can impair bone formation by inhibiting extracellular signal-regulated kinase (ERK) to reduce SLIT3 expression in osteoblasts and type H vessels, providing a possibility for activating the ERK-SLIT3 pathway of osteoblasts to promote bone formation. 9 Epoxyeicostrioleic acids (EETs), metabolites of arachidonic acid (AA), 12 have multiple effects, such as reducing inflammation, promoting angiogenesis, inhibiting osteoclast and adipogenic differentiation of stem cells. [13][14][15] Especially in the cardiovascular system, EETs have been shown to promote ECs proliferation and angiogenesis in various models. 16,17 Recent in vivo studies demonstrate that EETs significantly promote liver regeneration, compensatory growth of renal and lung, corneal neovascularization and retinal angiogenesis. [17][18][19] Our previous research revealed that EETs enhance the repair ability of irradiation-damaged salivary glands by improving angiogenesis and inhibiting cell apoptosis. 20 EETs can increase ERK phosphorylation levels and regulate EC function, such as proliferation and migration, through mitotic signalling cascade. 21 11, 12-EET can induce more robust tube formation by markedly increasing VEGF-A and bFGF expression in myocardial infarction. 22 EETs also encourage the generation of angiogenesis transcription factors through the HIF-1α pathway, and increase HIF-1α-DNAbinding activity to improve the resistance of myocytes to acute ischemia-reperfusion. [23][24][25][26][27] However, soluble epoxide hydrolase (sEH) can rapidly hydrolyze EETs to bio-inactive dihydroxyeicosatrienoic acids (DHETs) in vivo, so the half-life of EETs is short, which limits the pharmacological efficacy through EETs administration. 28 Thus, stabilizing endogenous EETs by sEH inhibitor (sEHi) becomes a candidate strategy. Studies have indicated that sEHi can significantly increase the concentration of EETs in organs, showing the potential of sEHi to be used in humans because it has no obvious toxicity. 29 TPPU, a potent and highly selective sEHi, shows higher efficiency in inhibiting EETs hydrolysis and stabilizing EETs levels. 30 Many of the effects of EETs described above are related to angiogenesis, and some of the pathways regulated by EETs also imply that they may be involved in osteogenesis. However, whether increasing EETs levels using TPPU can promote the formation of type H vessels and enhance bone regeneration remains to be determined.
In this study, we define a previously unidentified mechanism by which TPPU links type H vessels and bone formation by targeting sEH. The in vivo results showed that TPPU resulted in high bone mass and CD31 hi EMCN hi endothelium in mouse femurs and calvarial defect sites. We established an in vitro human umbilical vein endothelial cells (HUVECs) and human dental pulp stem cells (hDPSCs) co-culture system under osteogenic conditions to study the effect of TPPU on the cross-talk between them, indicating that TPPU enhanced cell proliferation, angiogenesis and osteogenic differentiation of co-cultured cells.
Mechanistically, TPPU induces a positive feedback loop by upregulating hDPSCs-derived SLIT3 to stabilize HIF-1α levels in ECs, thereby promoting the growth of type H vessels to link osteogenesis with angiogenesis. Therefore, our study reveals that TPPU promotes the coupling of osteogenesis and angiogenesis for bone formation by activating SLIT3/HIF-1α signalling pathway, which provided a novel therapeutic strategy for bone repair and regeneration through targeting sEH. For femurs bone assay, ten 3-week-old mice were divided into two groups randomly and gavaged with or without TPPU (3 mg/kg) every other day. After 2 weeks, the femurs from the euthanized mice were isolated and fixed for 48 h with 4% paraformaldehyde for micro-CT assay and histochemical staining.

| In vivo animal experiments
The model of calvarial bone defect as previously described was applied to evaluate the therapeutic benefits of TPPU. 31,32 Briefly, a critical-sized (5 mm diameter) circular bone defect was made in the exposed frontal bone of mice under anaesthesia using a microsurgical drill. Plastically compressed collagen gels scaffolds (5 mm diameter) seeded with mice bone marrow mesenchymal stem cells (mBMSCs, 2 Â 10 6 cells/ml) were transplanted into calvarial defect sites. 31,33 Finally, the skin was sutured, and the animal was observed in accordance with predetermined postoperative guidelines. Skulls were harvested up to 8 weeks after transplantation and were performed micro-computed tomography (CT) and histological assays.

| Cell isolation and culture
This study for isolation and culture of hDPSCs was approved by the For co-culture, the HUVECs and hDPSCs were mixed in a 1:1 ratio and then inoculated into the well plates to construct a co-culture system with direct cell-cell contact between HUVECs and hDPSCs.
Male 5-week-old mice's femurs were used to extract mBMSCs, which were then cultivated in α-MEM medium (HyClone) containing 15% FBS as previously described. 34 For the in vivo investigation, passages 3-5 of mBMSCs were used.

| EdU staining assay
The cells were plated (1 Â 10 6 cells/ml) and incubated with EdU (EdU test kit Abcam) for 4 h. The cells were then permeabilized using permeabilization buffer, fixed with 4% paraformaldehyde, and treated with EdU Reaction Mix. Finally, a flow cytometer (BD Pharmingen) was applied to determine how many cells had incorporated EdU at Ex/Em = 491/520 nm.

| TUNEL staining
The cells seeded in Coverslips were treated with permeabilization buffer (0.25% Triton-X-100) for 5 min. Deparaffinization and permeabilization of the tissue were performed using 20 g/ml Proteinase K in PBS. The apoptotic cells were treated with a TUNEL assay kit (Apoptosis Detection Kit, Vazyme) complied with the manufacturer's instructions and our earlier study. 20 Finally, DAPI (Sigma-Aldrich) was used to mount the sections before fluorescence microscopy imaging (Olympus Corporation).

| Angiogenesis assay on Matrigel
The tube formation experiment was performed as reported before. 34 The cells were inoculated in 96-well plates (1 Â 10 4 per well) precoated with Matrigel (BD Biosciences) and cultured in presence or absence of TPPU (10 μM). The capillary-like structures were examined with a phase microscope.

| Wound healing assay
Cells were added into six-well plates (2 Â 10 5 cells/well), and a linear scratch was made with a tip after the cells reached 90% confluence.
Then, the cells were incubated in a serum-deprived medium with or without TPPU (10 μM). Image-Pro Plus software was used to calculate cell migration rate (Media Cybernetics).
When the cells reached 80% confluence, the media was transferred to the osteogenic medium that contains 10% FBS, 0.1 μM dexamethasone (Solarbio), 50 μg/ml L-ascorbic acid and 10 mM β-glycerophosphate disodium in α-MEM culture medium with or without TPPU. 34 Cells were stained with a fast blue solution after being fixed in 10% neutral formalin buffer to detect ALP activity (Sigma-Aldrich). Following a wash, the samples were treated with an alkaline buffer solution and phosphatase substrate from Sigma-Aldrich. The ALP activity was read at the absorbance of 405 nm. For Alizarin red S (ARS) quant assay, the cells after ARS staining was destained with cetylpyridinium conchloride (CPC; Sigma-Aldrich; 10%) and the supernatant was read at an absorbance of 570 nm.

| Cell transfection
Transfection of cells was done in accordance with the Genechem manual. The cells were added into six-well plates (approximately 5 Â 10 4 cells per well) until they reached 50% confluence. Using a recombinant lentivirus, the sh-Slit3 plasmid was transfected into hDPSCs and the sh-Hif-1α plasmid into hHUVECs. The cells were cultured in the media with free FBS and screened with the media containing 0.1% puromycin to remove the untransfected cells.
HIF-1α and SLIT3 levels were measured by western blotting (WB) and RT-qPCR. The siRNA target sequence was shown in Table S2.

| Statistical analysis
At least three times each of the tests were repeated. GraphPad Prism 9 was employed for Student-Newman-Keuls tests and one-way ANOVA. Data were considered statistically significant at p < 0.05 and showed the means ± SEM of replicate measurements.

| RESULTS
3.1 | TPPU enhances the osteogenic differentiation potential and induces CD31 hi EMCN hi endothelium formation in co-cultured HUVECs and hDPSCs TPPU plays considerable effects on angiogenesis and tissue regeneration by specifically inhibiting sEH to increase endogenous EETs. To further investigate the role of TPPU in coupled angiogenesis and osteogenesis, a novel co-cultured hDPSCs and HUVECs system with direct cell-cell contact was constructed to mimic in vivo cross-talk between blood vessels and surrounding stem cells. DPSCs are now considered to be a type of mesenchymal stem cells (MSCs) and demonstrate higher clonogenic and proliferative potential than BMSCs.
Compared with other MSCs, such as BMSCs, DPSCs are very easily isolated from extracted teeth by low invasive surgery without any ethical issues. Previous studies have confirmed that TPPU can inhibit the adipogenic differentiation of MSCs through increased endogenous EETs, so we first tested the osteogenic differentiation effect of TPPU on hDPSCs and found that TPPU had no obvious promotion for the osteogenesis differentiation of hDPSCs ( Figure 1A). Next, we examined the roles of TPPU on osteogenic differentiation of co-cultured cells. Consistent with previous studies, co-cultured hDPSCs with HUVECs exhibited higher osteogenic capacity than hDPSCs alone under osteogenic conditions. Interestingly, TPPU further enhanced the osteogenic differentiation ability of co-cultured HUVECs and hDPSCs. Alkaline phosphatase (ALP) staining showed that TPPU further induced a significantly higher ALP activity in co-cultured cells.
Meanwhile, the Alp and Runx2 expression was markedly upregulated in TPPU-treated co-cultured cells as compared to that of vehicle ( Figure 1B). WB results also confirmed that the level of ALP and Runx2 proteins was significantly upregulated in TPPU-treated cocultured cells following 7 days of osteogenic induction ( Figure 1C). Subsequently, a substantially higher level of mineralization was detected by Alizarin red (AR) staining ( Figure 1D) and a strengthen expression of Ocn mRNA and protein levels ( Figure 1E,F) in cocultured cells after 21 days of osteogenic induction. The above results suggested that TPPU did not directly affect the osteogenic differentiation of MSCs alone, implying that TPPU might enhance the osteogenic differentiation of MSCs through acting on ECs to regulate the cross-talk between HUVECs and hDPSCs.
Because TPPU is known to enhance angiogenesis in vivo and in vitro with upregulating endogenous EETs, we next questioned whether TPPU enhanced the osteogenic differentiation potential of  Figure S1A) and resulted in higher levels of migration in co-cultured HUVECs and hDPSCs relative to a vehicle control ( Figure 1G). TPPU also displayed an enhanced ability to induce tube formation of HUVECs alone, greatly enhance tube formation of co-culture cells, but had almost no effect on tube formation of hDPSCs alone relative to vehicle control ( Figure 1H). Therefore, we suspected that TPPU mainly acted on HUVECs to promote the cross-talk between HUVECs and hDPSCs, thus increasing the coupling of angiogenesis and osteogenesis in the co-culture system. Then, we examined the expression of type H vessel-specific markers CD31 and EMCN. TPPU modestly increased the mRNA level of CD31 and Emcn in HUVECs. Notably, co-culture of HUVECs and hDPSCs resulted in higher levels of CD31 and Emcn expression, and TPPU treatment further dramatically increased the expression of CD31 (over 10-fold) and Emcn (over 5-fold) in the co-cultured cells relative to vehicle control ( Figure 1I). Thus, these results provide proof of the principle that targeting sEH with TPPU contributes to CD31 hi EMCN hi ECs by enhancing the reciprocal interactions with HUVECs and hDPSCs to induce type H vessel-related factors.

| TPPU increases CD31 hi EMCN hi ECs via upregulating SLIT3 levels in vitro
Given that EETs are known to increase ERK activity, and ERK has been shown to upregulate osteoblast-derived SLIT3 to increase CD31 hi EMCN hi endothelium. We hypothesized that TPPU could upregulate osteoblast-derived SLIT3 to enhance the coupled angiogenesis and osteogenesis. We detected the factors known closely related to type H vessels in co-cultured HUVECs and hDPSCs. The mRNA levels of Vegf, Hif-1α and Slit3 were greatly increased in cocultured cells treated with TPPU as well as protein levels ( Figure 2A To further explore the role of TPPU on SLIT3 coupling angiogenesis and osteogenesis in the co-culture cells, we performed shRNAmediated Slit3 knockdown in hDPSCs ( Figure S2A). ALP staining showed that Slit3 knockdown in hDPSCs significantly impaired the osteogenic differentiation ability of co-cultured cells, and TPPU could not reverse the impaired osteogenic ability ( Figure 2E). Knockdown of Slit3 in hDPSCs resulted in decreased mRNA levels of type H vesselsassociated CD31 and Emcn, osteogenic differentiation-associated Alp and Runx2, and TPPU could not restore their expression ( Figure 2F).
Similarly, AR staining showed that knockdown of Slit3 in hDPSCs resulted in decreased mineralization activity in co-cultured cells after 21 days of osteogenic induction ( Figure 2G). Therefore, the above results suggested that SLIT3 is indispensable for the osteogenic differentiation of stem cells, and SLIT3 is also necessary for TPPU to participate in osteogenesis and angiogenesis.
We then investigated the signalling pathways downstream of SLIT3 to detect the effect of TPPU on cross-talk between hDPSCs and HUVECs under osteogenic conditions. TPPU increased the expression of Robo1 and Yap1 in co-cultured cells, which was notably decreased after Slit3 knockdown in hDPSCs, and TPPU could no longer rescue their expression ( Figure 3A). The protein levels of HIF-1α, a wellknown factor for type H vessels, were considerably reduced in the co-  Figure 3C,E). As expected, knockdown of Hif-1α in HUVECs also caused a substantial decrease of type H vascular-related genes (CD31 and Emcn) and osteogenesis-related genes (Runx2 and Alp) after 7 days of osteogenic induction, and TPPU did not reverse the expression of these genes ( Figure 3D), further confirmed that TPPU mainly induced an increase of HIF-1α expression in ECs. Moreover, previous studies have proved that HIF-1α could directly bind to the promoters of SLIT3. However, we did not observe evidence that knockdown of Hif-1α in HUVECs resulted in a dramatic decrease in Slit3 ( Figure S3). Taken together, these results further suggested that TPPU further enhanced HIF-1α expression in ECs mainly by upregulating hDPSCs-derived SLIT3, thereby promoting the coupling of osteogenesis and angiogenesis.

| TPPU enhances bone formation, bone repair and increases type H vessels in vivo
To further investigate whether TPPU contributes to bone growth in vivo, we administered TPPU by oral gavage every other day to three-week-old C57BL/6 mice for 2 weeks to determine whether TPPU contributes to bone formation. Then, we examined the long bones and found that the freshly isolated femurs looked brighter red and larger in size in TPPU-treated mice, implying that TPPU seems to have an effect on angiogenesis and bone growth ( Figure 4A). We next F I G U R E 2 TPPU upregulates SLIT3 levels to promote osteogenesis in vitro. (A) RT-qPCR showing the relative mRNA expression of Slit3, Hif-1α and Vegf in co-cultured cells. (B) Western blotting assay for protein levels of SLIT3, HIF-1α and VEGF in co-cultured cells. (C) Representative images of immunofluorescence staining of co-cultured HUVECs and hDPSCs after 7 days of osteogenic culture for SLIT3 (red) and HUVECs (green) and quantification of SLIT3 expression. Scale bar, 50 μm. (D) Representative images of immunofluorescence staining of HIF-1α (red) and HUVECs (green) in the co-cultured HUVECs and hDPSCs treated with or without TPPU after 7 days of osteogenic culturing and quantitative analysis of HIF-1α expression. Scale bar, 50 μm. (E) Representative images of ALP staining and ALP activity assay of co-cultured Slit3-knockdown hDPSCs and HUVECs after 7 days of osteogenesis induction. (F) RT-qPCR analysis of Alp, Runx2, CD31 and Emcn in co-cultured Slit3-knockdown hDPSCs and HUVECs after 7 days of osteogenesis induction. (G) Representative images of alizarin red staining and relative mineralization assay co-cultured Slit3-knockdown hDPSCs and HUVECs after 21 days of osteogenesis induction. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. hDPSC, human dental pulp stem cell; HUVEC, human umbilical vein endothelial cell. evaluated whether TPPU promoting long bone growth was associated with cell proliferation and apoptosis. Cell proliferation is known to be promoted by the co-culture of MSCs and ECs. CCK8 showed that TPPU further enhanced the proliferation of co-cultured cells ( Figure S4A), and EdU labelling showed a similar increase compared with vehicle ( Figure S4B). The apoptotic cells were significantly less in TPPU-treated co-cultured cells by TUNEL staining ( Figure S4C). Consistent with the in vitro data, in vivo immunohistochemistry (IHC) and IF analysis showed that TPPU significantly increased Ki67-positive cells ( Figure S4D) and decreased TUNEL-positive cells in the femoral metaphysis of C57/BL6 mice ( Figure S4E). Our results demonstrated that TPPU might be involved in bone growth by coordinating cell proliferation and apoptosis.
Micro-CT analysis further confirmed that TPPU resulted in a higher bone mass and density ( Figure 4B). In addition, we also found that increased blood vessels and collagen by HE and Masson's trichrome staining in TPPU-treated mice femur compared with vehicle mice ( Figure 4C,D). EMCN is known to be highly enriched in kidney and lung, and we found that TPPU could promote CD31 hi EMCN hi ECs in the kidneys and lungs of C57/BL6 ( Figure S5A,B). Given that type H vessels modulate bone sculpting and remodelling, we performed immunofluorescence (IF) colocation staining to measure the abundance of CD31 hi EMCN hi vessels in the epiphysis. Notably, TPPU treatment considerably increased type H vessels with CD31 hi EMCN hi endothelium beneath the growth plate in the femur ( Figure 4E). Consistent with in vitro results, TPPU induced substantially greater levels of SLIT3 and HIF-1α expression near the growth plate of femur.
Immunofluorescence analysis confirmed that most of the SLIT3-positive cells were adjacent to EMCN-expressing cells ( Figure 4F, a few co-localized with EMCN), and an increase of HIF-1α expression co-localized with EMCN in the femoral metaphysis ( Figure 4G).
To evaluate whether TPPU has the same effect on bone repair, the critical-size calvarial defects in mice were applied to test the F I G U R E 3 TPPU activates HIF-1α to enhance type H vessels by increasing osteoblast-derived SLIT3. (A) RT-qPCR analysis of Robo1 and Yap1 in co-cultured Slit3-knockdown hDPSCs and HUVECs after 7 days of osteogenesis induction treated with or without TPPU. (B) Western blotting assay and quantification of HIF-1α expression of co-cultured Slit3-knockdown hDPSCs and HUVECs after 7 days of osteogenesis induction treated with or without TPPU. (C) Representative images of ALP staining and ALP activity assay of co-cultured Hif-1α knockdown HUVECs and hDPSCs treated with or without TPPU after 7 days of osteogenesis induction. (D) RT-qPCR analysis of mRNA expression of Alp, Runx2, CD31 and Emcn in co-cultured Hif-1α knockdown HUVECs and hDPSCs treated with or without TPPUs after 7 days osteogenic induction. (E) Representative images of alizarin red staining and relative mineralization assay of co-cultured Hif-1α knockdown HUVECs and hDPSCs treated with or without TPPUs after 21 days of osteogenesis induction. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. hDPSC, human dental pulp stem cell; HUVEC, human umbilical vein endothelial cell.
F I G U R E 4 Legend on next page. therapeutic effects of TPPU on bone regeneration. The bone repair efficiency inside the defect regions 8 weeks after mBMSCs implantation was assessed using radiographic and histological methods. The significantly more regenerated bone tissue in the TPPU-treated mice than in the control mice, according to the rebuilt morphology of the defect locations ( Figure 5A). Masson's trichrome staining in calvarial defects showed thicker regenerated tissue and more collagen fibres in TPPU-treated mice ( Figure 5B). Immunofluorescent staining showed that TPPU increased the number of positive cells with high expression for CD31 and EMCN ( Figure 5C), SLIT3 ( Figure 5D) and HIF-1α ( Figure 5E) in defect area. Therefore, the findings demonstrated that oral TPPU administration induced type H vessels which enhanced bone growth and bone repair.

| DISCUSSION
The relationship between blood vessels and osteoblasts affects how bone metabolism is balanced. Through certain vascular shapes and routes, angiogenesis and osteogenesis are connected. Blood vessels act as structural models for the creation of new bone and are the primary regulators of bone regeneration. 26,27 Early revascularization is essential for bone repair and tissue engineering regeneration.
Research in recent years has suggested that it is conducive to bone development and stability by promoting or maintaining type H vessels coupling osteogenesis. 35 Matrix metalloproteinase-9, HIF-1α and VEGF released by ECs from type H vessels promote cartilage absorption to enhance longitudinal bone growth. 27 Our in vivo study demonstrated that increased EETs levels by targeting sEH with TPPU Other studies have demonstrated that EETs inhibit osteoclastogenesis through modulation of multiple pathways both upstream and downstream of RANKL signalling. 42 Kusumbe et al. 5  interaction between these two kinds of cells, and SLIT3 and EMCN play an important role in this process. Our results also suggest that knockdown of Slit3 in hDPSCs significantly impairs TPPU-enhanced the coupling of osteogenesis and angiogenesis. However, TPPU has no obvious change on the expression of SLIT3 in ECs, so we did not separately detect the effect of SLIT3 knockdown on angiogenesis in ECs, which will be our further research work.
The ability of bone marrow-derived ECs to form tubes and the phosphorylation of the hippo pathway signalling intermediary YAP1 show that Robo1 knockdown affects how these cells respond to SLIT3. 9 According to previous research, YAP1 is essential for ECs migration and tube formation. 49,50 Our study found that TPPU could upregulate the expression of ROBO1 and YAP1 in the co-culture cells, but when SLIT3 was inhibited, the effects of TPPU on ROBO1 and YAP1 were blocked. When Slit3 was knocked down, HIF-1α upregulated by TPPU was significantly inhibited. HIF-1α is a heterodimer transcription factor that regulates both normal and abnormal angiogenesis and governs how cells react to oxygen fluctuations. 51,52 Many studies have proved that HIF-1α knockdown in ECs could reduce the ability of angiogenesis through a variety of signal pathways. Endothelial HIF-1α has been shown to be a significant activator of H-type vessels development in the metaphysis. Osteoprogenitors were significantly decreased after EC-specific ablation of HIF-1α, which was followed by a decrease in trabecular bone formation. 5,53 In the nucleus, YAP1 activation directly interacts with HIF-1α and maintains HIF-1α stability to stimulate the transcription of genes associated to angiogenesis. 54 Our results also confirm that HIF-1α is mainly induced by the MSC-derived SLIT3 upregulated by TPPU, while HIF-1α has little effect on SLIT3. Given that TPPU has little effect on the osteogenic differentiation of hDPSCs alone, suggesting that the cross-talk between ECs and MSCs plays a critical role, the detailed mechanism of TPPU regulating SLIT3 expression in stem cells through ERK requires further investigation. Furthermore, systemic administration demonstrates the effect of TPPU in coupling angiogenesis and osteogenesis. It is worth looking forward to improving the drug delivery method for local application in the future. 55,56 Thus, we prove that TPPU promotes type H vessels formation and enhances bone formation, and provides new evidence linking TPPU to the coupling of osteogenesis and angiogenesis via SLIT3-ROBO1/YAP-HIF-1α signalling axis, highlighting that inhibiting sEH to enhance the coupled angiogenesis and osteogenesis can be considered a potential strategy for bone repair and regeneration.