Fig 2: Gel‐BG@ab enabled sustained release of anti‐Sema4D. (A) Schematic illustration of the procedure of Gel‐BG@ab synthesis. BG, mesoporous bioactive glass; BG@ab, BG loaded with anti‐sema4D antibody; mixture indicates BG@ab added to a solution of PPP. (B) XRD pattern of BG. (C) FTIR spectra of BG. (D) XPS spectra of BG. (E) TEM images of BG and BG@ab. Scale bars, 50 nm. (F) TEM images of BG@ab and the corresponding element maps (Si, Ca, P, S and N). Scale bars, 50 nm. (G) EDX spectrum of BG@ab. (H) Images of the gel (PPP) and Gel‐BG@ab at 37°C. (I) The cumulative anti‐sema4D antibody released from the 3% BG and 5% BG in the hydrogel from day 0 to day 7. (J) The hydrogel contained BG@ab (3% or 5%) degraded gradually in PBS from day 0 to day 21. (K) AOPI staining images of osteoblasts cultured on the gel. 20% gel indicated 20% PPP, and 3% BG@ab and 5% BG@ab indicated the weight ratio of BG in the gel. Scale bars, 100 μm. The viability of osteoblasts cultured on the gel was measured via CCK‐8 assay kits. ns indicates no significance. Data are mean ± SD.

Fig 3: Soluble Sema4D of activated γδ T cells inhibited osteoblast differentiation. (A) Representative scatter plots of γδ T cell proliferation were shown, and ZOL increased the proliferation of γδ T cells from the spleens at different concentrations for 3 days. n = 4, **p < 0.01, ns indicates no significance. (B) Purity of γδ T cells incubated with ZOL for 6 days. n = 13, **p < 0.01, ns indicates no significance. (C) Representative contour plots of Sema4D+ γδ T cells. (D) The Sema4D expression on expanded γδ T cells decreased significantly at 15 μM ZOL treatment on day 3 and day 6. n = 10, *p < 0.05, **p < 0.01, ns indicates no significance. (E) The sSema4D in γδ T cell supernatant increased significantly at 15 μM ZOL treatment on day 3 and day 6. n = 3, *p < 0.05, **p < 0.01, ns indicates no significance. (F) The mRNA level of Alpl, Runx2 and Osterix was measured after being cultured with γδ T cell supernatant from WT ZOL mice, and the mRNA levels declined in a dose‐dependent manner. n = 4, **p < 0.01. (G) ALP staining of osteoblasts after incubation with the γδ T cell supernatant for 7 days, and the supernatant of γδ T cells was collected from WT ZOL and TCRδ−/− ZOL mice treated with different concentrations of ZOL. ALP enzyme activity was also analysed on day 7 after incubation with γδ T cell supernatant. The supernatant from WT mice showed much stronger inhibition of osteoblast differentiation than that of TCRδ−/− mice. n = 4, **p < 0.01. ns indicates no significance. (H and I) The γδ T cell supernatant supplemented with anti‐Sema4D antibody (αSema4D) alleviated the inhibition of osteoblast differentiation; however, the Sema‐FC enhanced differentiation inhibition of the supernatant of TCRδ−/− mice. SN, supernatant of γδ T cells. n = 4, *p < 0.05, **p < 0.01. (J) The mRNA level of Alp was measured after incubation with the γδ T cell supernatant supplemented with anti‐Sema4D antibody (αSema4D) and/or Sema‐FC. n = 4, *p < 0.05, **p < 0.01. Data are mean ± SD.

Fig 4: Sema4D derived from γδ T cells. (A) KEGG analysis of the DEGs of WT ZOL and TCRδ−/− ZOL mice. (B) GSEA results of WT ZOL and TCRδ−/− ZOL mice showed lymphocyte‐mediated immunity, activation of immune response and bone maturation pathways. (C) The mRNA levels of Sema4d and Sema3g decreased, and Sema3a increased in the sockets of TCRδ−/− ZOL mice compared to WT ZOL mice. n = 4, *p < 0.05, **p < 0.01. (D) The level of sSema4D in the sockets was detected by ELISA. n = 4, **p < 0.01. (E) Colocalization staining of CD3 and Sema4D in the sockets at 7 days after tooth extraction in WT ZOL and TCRδ−/− ZOL mice. Scale bars, 50 μm. (F) The number of CD3+ Sema4D+ T cells in the sockets of TCRδ−/− ZOL mice was much less than that in WT ZOL mice. **p < 0.01. (G) Representative scatter plots of CD3+ Sema4D+ T cells were shown. (H) The Sema4D expression on CD3+ T cells of mucosa, blood and spleen in WT ZOL and TCRδ−/− ZOL mice was analysed on day 7. n = 4, *p < 0.05. Data are mean ± SD.

Fig 5: Autocrine MMP3 secretion induced mSema4D cleavage in γδ T cells. (A) KEGG analysis of DEGs in WT ZOL and TCRδ−/− ZOL mice was highly enriched in matrix metalloproteinase (MMP)‐related pathways. (B) The mRNA levels of Mmp3 and Adam17 in the extraction sockets of WT ZOL and TCRδ−/− ZOL mice. n = 4, *p < 0.05, ns indicates no significance. (C) The mRNA levels of Adam17, Adam10, Mmp2, Mmp3, Mmp9 and Mmp11 after γδ T cells were incubated with 15 μM ZOL for 3 and 6 days. Only the mRNA level of Mmp3 increased significantly and kept a high level after 6 days incubation. n = 6, *p < 0.05, **p < 0.01. (D) The protein level of MMP3 was much higher than MMP2 and MMP9 in the γδ T cell supernatant measured by ELISA. n = 4, **p < 0.01, ns indicates no significance. (E) The proportions of γδ T cells after treatment with GM6001 and TAPI‐2. GM6001 and TAPI‐2 are MMP inhibitors. T10 and T20 indicate 10 and 20 nM TAPI‐2, respectively; G50 and G100 indicate 50 and 100 nM GM6001, respectively. n = 6, ns indicates no significance. (F) The expression of Sema4D on γδ T cells after treatment with GM6001 and TAPI‐2. (G) Analysis of Sema4D expression in γδ T cells after treatment with GM6001 or TAPI‐2 on day 2 and 4. n = 6, *p < 0.05, **p < 0.01. (H) The level of sSema4D in the supernatant of γδ T cells after treatment with GM6001 or TAPI‐2. n = 3, *p < 0.05, **p < 0.01. (I) The level of sSema4D in the supernatant of γδ T cells after incubating with MMP2, MMP3 and MMP9 in synergy with GM6001. n = 4, *p < 0.05, **p < 0.01. Data are mean ± SD.