DPSCs isolation and culture

The human DPSCs experiments involved in this study was approved by the Ethical Committee of Beijing Stomatological Hospital, Capital Medical University (Ethical Review NO. CMUSH-IRB-KJ-PJ-2022-22). Human premolars or third molars were obtained after obtaining informed consent. The DPSCs were obtained as described previously [25]. Use passages 3–5 of DPSCs in subsequent experiments.

HUVECs culture

HUVECs (Cat No. HUVEC-20001), and complete medium of human umbilical vein endothelial cells (HUVECs) (Cat No. HUVEC-90011), and 0.25% tryptic digestion solution (Cat. TEDTA-10001) were purchased from Cyagen Biosciences, Inc. (Guangzhou, China). Passages 5–6 of HUVECs were used in our experiments.

Plasmid construction and viral infection

Plasmid construction and viral transfection were performed as previously described [26]. Following standard methods, the plasmids were constructed and the structures were identified using restriction enzymes and sequence. The hemagglutinin (HA)‐tag and human full-length NQO1 cDNA were constructed via a whole-gene synthesis method, and the HA‐tag and NQO1 sequence were cloned into NheI/NotI-digested pCDH-CMV-MCS-EF1a-puro (HA-NQO1). Empty pCDH-CMV-MCS-EF1a-puro vectors was used as control (Vector). The plasmids were packaged as lentivirus. GenePharma (Suzhou, China) was responsible for synthesizing NQO1 shRNA (NQO1sh) and control shRNA (Consh). The DPSCs were then transfected by lentiviruses by using polybrene (6 μg/mL; Sigma‐Aldrich, St. Louis, MO) for 12 h. The transfected DPSCs were cultured by adding 2 μg/mL puromycin after 72 h. The control group provided a reference point for the results of the experiment to better evaluate the influence of experimental conditions on the results of the experiment.

Co-culture of DPSCs and HUVECs

An Angiogenesis assay in vitro was performed as previously described [27]. Wells were coated with 50 μL Matrigel (Cat No. 356234, BD Biosciences) in a 96-well plate. 100 μL cell suspension of DPSCs and HUVECs (1 × 104 cells, respectively) was added into each well after gelatinization. After co-culture for 3 h, we obtained images with a microscope (Olympus, Japan). ImageJ software was used to quantify the tubules formed. For the sake of methodological rigor, there were three parallel wells in each group.

Scratch migration assay

The scratch migration assay was performed as described previously [26]. Changes in the width of the void area relative to that at 0 h (relative width) were evaluated in each group. In the experimental design, triplicate wells were established for both the experimental and control groups to ensure the reliability and reproducibility of the results. Within these triplicate wells, nine random fields of view were selected for microscopic examination to minimize sampling bias.

Transwell chemotaxis assay

Transwell chemotaxis assays were performed as described previously [26]. After 24 and 48 h, the number of migrated cells was calculated from the micrographs. Triplicate assay conditions were implemented for both the experimental group and the control group. Within each replicate, a systematic random sampling approach was employed to select six discrete fields of view for microscopic analysis.

Angiogenesis assay in vivo

All animal experiments were performed after approval from the Animal Care and Use Committee of the Beijing Stomatological Hospital Animal Laboratory (KQYY-202212-001). Ten nude mice (female, 8-week-old, nu/nu) were obtained from the Beijing Stomatological Hospital Animal Laboratory (Beijing, China). Nude mice were anesthetized using sodium pentobarbital. 150 μL matrigel containing HUVECs and DPSCs (1 × 106 cells, respectively) was transplanted into the nude mice subcutaneously. Each nude mouse received two implants. Samples were obtained 2 weeks after transplantation. Our work has been reported in accordance with ARRIVE Guide 2.0.

Alkaline phosphatase (ALP) activity assay and Alizarin red staining (ARS)

DPSCs were induced by using osteogenic induction medium (100 µM/ml ascorbic acid, 2 mM β‐glycerophosphate, 1.8 mM KH2PO4 and 10 nM dexamethasone). Using an ALP activity kit (Sigma-Aldrich) to detect ALP activity after 3 and 5 days, and Alizarin Red staining (Sigma-Aldrich) was performed after 2 weeks, as described previously [26]. Each experimental cohort was complemented with three parallel wells.

Osteogenesis assay in vivo

10 nude mice (female, 10‐week‐old, nu/nu) were purchased from the Beijing Stomatological Hospital Animal Laboratory (Beijing, China). None of the animals received any medical or surgical treatments. 2.0 × 106 cells were mixed with 20 mg HA/tricalcium phosphate (Engineering Research Center for Biomaterials, Chengdu, China), incubated at 37 °C for 3 h, and transplanted into the nude mice. After eight weeks, the subcutaneous grafts were removed from the back and the following experiments were performed. Our work has been reported in accordance with ARRIVE Guide 2.0.

Hematoxylin and eosin (H&E) staining

The harvested grafts were stained with H&E as described previously [27]. The number of formed vessels or mineralized tissues was analyzed. The slices were selected randomly and then generating a randomization sequence using a simple random sampling method.

Immunofluorescence staining

Immunofluorescence staining was performed as described previously [28]. Anti-CD31 primary antibody (Cat No. 11265-1-AP; Proteintech, Rosemont, IL, USA) was used for immunofluorescence staining. A fluorescence microscope (Olympus, Tokyo, Japan) was used to observe and photograph the stained samples.

Masson staining

Masson’s trichrome staining was performed as described previously [29]. The newly formed osteoid were observed under a microscope (Olympus, Japan).

Immunohistochemical staining

Immunohistochemical staining was performed as described previously [26]. The primary antibodies used were anti-dentin sialophosphoprotein (DSPP; Cat No. bs‐10316R; Bioss). The stained samples were observed and photographed using a microscope (Olympus, Tokyo, Japan).

Reverse transcriptase‐polymerase chain reaction (RT-PCR) and real‐time RT-PCR

Total RNA extraction, cDNA synthesis, and real-time RT-PCR were performed as described previously [30]. Each reaction was conducted in triplicate to ensure the reliability and consistency of the results. The primer sequences used in this study are listed in Additional file 1 (Table S1).

Western blot

Western blotting was performed as previously described [27]. The primary antibodies used were anti-NQO1 (Cat No. 11451-1-AP, Peintech, Rosemont, IL, USA), anti-DSPP (Cat No. bs-10316R, Bioss, Beijing, China), anti-Vinculin(Cat No. 26520-1-AP, Peintech, Rosemont, IL, USA), anti-HA (Cat No.51064-2-AP, Peintech, Rosemont, IL, USA), anti-JNK(Cat No. 9258; Cell Signaling Technology), anti-Phospho-JNK (Cat No. 80024-1-RR, Peintech, Rosemont, IL, USA), anti-Erk1/2 (Cat No.11257-1-AP, Peintech, Rosemont, IL, USA), anti-Phospho-Erk1/2 (Cat No.80031-1-RR, Peintech, Rosemont, IL, USA), anti-p38 (Cat No.14064-1-AP, Peintech, Rosemont, IL, USA), anti-Phospho-p38(Cat No.28796-1-AP, Peintech, Rosemont, IL, USA), SIRT1 Rabbit pAb(Cat No.A11267, Abclonal, Wuhan, China) and GAPDH (Cat No.60004-1-Ig, Peintech, Rosemont, IL, USA).

Oxygen consumption rate (OCR)

The experimental method is as described previously [31]. The DPSCs (6 × 104/well) were seeded onto XF-24 plates and the experiments were repeated three times for each group. The mitochondrial stress assay parameters were calculated by using the OCR curves. Each experimental cohort was complemented with five parallel wells.

Mitochondrial membrane potential (MMP) assay

The MMP of the DPSCs was determined using the JC-1 method (Solarbio, Beijing, China). Red fluorescence indicates a higher MMP and green fluorescence indicates a lower MMP. CCCP, an OXPHOS decoupling agent, was used as a positive control to induce MMP reduction in DPSCs. Each group was provided with three parallel wells.

Statistical analysis

All statistical analyses were performed using the SPSS 22 statistical software (SPSS Inc., Chicago, IL, USA). Student’s t-test or one-way ANOVA was used to calculate the data. Differences were considered statistically significant at P < 0.05.