https://orcid.org/0000-0001-8396-307X
Abstract
Xanthohumol (XAN), a natural isoflavone from Humulus lupulus L., possesses biological activities on relieving oxidative stress and osteoporosis (OP). This study aimed to evaluate the antioxidative and osteoprotective effect of XAN on Aβ-injured osteoblasts, and explore its underlying mechanism.
Osteoblasts were pretreated with XAN followed by stimulation with Aβ1–42. Cell proliferation, ALP activity, bone mineralization and bone formation index were measured. Apoptosis and reactive oxygen species (ROS) were analysed with flow cytometer. PI3K inhibitor LY294002 or siRNA-Nrf2 was added and transfected in osteoblasts, to further confirm whether the pathway participated in the regulation of XAN-induced cytoprotection.
XAN markedly improved the proliferation, differentiation and mineralization of Aβ-injured osteoblasts. Additionally, XAN reduced cell apoptosis rate and ROS level, and increased the expression of p-AKT, Nrf2, NQO1, HO-1 and SOD-2. More importantly, LY294002 or siNrf2 abolished the beneficial effect of XAN on osteoblasts activity and decreased the PI3K expression and inhibited its downstream proteins, indicating XAN activated PI3K/AKT/Nrf2 pathway in Aβ-injured osteoblasts.
It was the first time to reveal the antioxidative and osteoprotective effect of XAN through regulating PI3K/AKT/Nrf2 pathway in Aβ-injured osteoblasts, which provides reference for the clinical application of XAN in the prevention and treatment of OP.
Introduction
Osteoporosis (OP) is featured by trabecular architecture deterioration and bone mass decline. Aging is a main pathogenic factors leading to OP, which leads to reactive oxygen species (ROS) release and induces oxidative damage, thus decreases osteoblastic bone formation.[1] Amyloid β-protein (Aβ), which has strong neurotoxic effect, is closely correlated with oxidative damage in several aging-related diseases.[2] Clinical studies have confirmed that OP is closely related to oxidative damage caused by Aβ deposition.[3, 4] Additional, the expression of Aβ was increased in osteoporotic bone tissue and ovariectomised rat bone tissue, and Aβ could inhibit osteoblast differentiation and activation.[5] More importantly, it was discovered that in APP/PS1 mutated transgenic mice, both brain and bone tissues showed Aβ deposition accompanied with peroxidation injury, and antioxidant components could improve the cognitive ability and bone loss caused by Aβ deposition and oxidative damage.[6] It is therefore speculated that antioxidants may prevent bone loss caused by Aβ deposition and oxidative damage.
Xanthohumol (XAN, PubChem CID: 639665, Supplementary Figure S1) is a kind of natural isoflavones from Humulus lupulus L., which has been reported to protect bone tissue, improve memory impairment, alleviating menstrual syndrome and relieving oxidative stress.[7] It is also widely used in Europe to relieve postmenopausal OP.[8] Previous studies discovered that XAN not only inhibited osteoclast-related gene expression in RAW264.7 osteoclasts[9] but also stimulated the osteoblastic differentiation.[10] Moreover, it was reported that XAN had antioxidant effect via up-regulating the expression of NQO1 and HO-1[11] and could inhibit oxidative damage through activating the Nrf2 signalling pathway.[12] Interestingly, our previous study also found that XAN could improve memory impairment of APP/PS1 mice via increasing superoxide dismutase level and reducing Aβ deposition.[13] These findings proved the bone-protective effect of XAN might be relevant to its effects on Aβ clearance and antioxidation. Nevertheless, the potential mechanism of action remains unclear.
It is common knowledge that oxidative damage will destroy cell function and disorder cell pathway.[14] The PI3K/AKT pathway is crucial in the process of cell survival, growth and differentiation[15] and is related to the clearance and deposition of Aβ.[16] Research has reported that inhibition of PI3K/AKT signalling pathway can bring about osteoblasts damage. Moreover, it is usually participated in the Nrf2-dependent transcription responding ROS injury.[17] Han’s research discovered that chlorogenic acid protected osteoblasts activity through regulating PI3K/AKT-mediated Nrf2 signalling and promoted bone formation,[18] indicating that Nrf2 might be a downstream signal target of PI3K/AKT with potential antioxidant effect. Nrf2 can also up-regulate the antioxidant enzymes including HO-1, NQO1 and superoxide dismutase 2 (SOD-2), and markedly reduce oxidative damage.[19] These findings proved that PI3K/AKT/Nrf2 pathway probably played a crucial role in relieving Aβ-stimulated oxidative damage, and maintaining bone homeostasis.
Consequently, this study was to investigate antioxidative and osteoprotective effect of XAN on Aβ1–42 oligomer-injured osteoblasts, and probe into its potential mechanism via the PI3K/AKT/Nrf2 signalling pathway, which provides insights into the new molecular mechanism of xanthohumol against OP.
Materials and Methods
Chemicals and reagents
Xanthohumol (XAN, Lot No: D91020804, Purity ≥ 98%) was purchased from Liding Biotechnology Co., Ltd. (Shanghai, China). Amyloid β1–42 oligomer (Catalog No: 052487, Purity ≥ 98%) was purchased from GL Biochem (Shanghai, China). N-Acetyl-l-cysteine (NAC, Purity ≥ 99%) was purchased from Beyotime Institute of Biotechnology (Shanghai, China). SOD, CAT, ROS and ALP assay kits were purchased from Beyotime Institute of Biotechnology (Shanghai, China). LY294002 (PI3K inhibitor) was purchased from MedChemExpress (Shanghai, China). Antibodies against OPN (ab8448, 1:1000), COL-I (ab34710, 1:1000), PI3K (ab191606, 1:1000), Nrf2 (ab137550, 1:1000), HO-1 (ab13243, 1:1000), NQO1 (ab34173, 1:1000) and SOD-2 (ab13533, 1:2000) were purchased from Abcam. Antibodies against Bax (AF0120, 1:1000), Bcl-2 (AF6139, 1:1000) were purchased from Affinity Biosciences. Antibodies against AKT (9272S, 1:1000), Phospho-AKT (4060T, 1:1000) and GAPDH (2118, 1:2000) were purchased from Cell Signaling Technology.
Osteoblasts culture and treatment
New-born Wistar rats (License No. SYXK (Su) 2018-0006) were obtained from Experimental Animal Center of the Naval Medical University. Osteoblasts were isolated from the calvarias of newborn Wistar rats.[20] The osteoblasts were grown in Minimum Essential Medium α containing 10% FBS, 1% penicillin-streptomycin, and incubated at 37°C with humidified atmosphere containing 5% CO2. All animal experiments were approved by the Committee on Ethics of Medical Research Second Military Medical University on 2019/09/25 (No. 201930921).
According to experiment design, osteoblasts were incubated overnight and then pretreated with different concentration of XAN (1, 5 and 25 µM), NAC (2.5 mM) or LY294002 (20 µM) for 4 h. After that, the plate was taken out from the incubator, and 4 μL of 1 mM Aβ1–42 oligomer mother liquor was added to per well, resulting in a final concentration of 40 μM Aβ. The cells were incubated for another 44 h, and cell growth in the orifice without the treatment was used as the control group.
Small interfering RNA transfection
To silence Nrf2, 1 × 106 osteoblasts were transfected with siRNA-Nrf2 or siRNA-negative control (GenePharma Co., Ltd). SiRNA was transfected with Lipofectamine 3000 (Invitrogen) reagent for 24 h. After that, osteoblasts were treated with Aβ and XAN. The transfection efficiency has been testified by western blotting.
Osteoblastic proliferation and ALP activity assays
For cells proliferation assay, the cultured osteoblasts were added in 96-well culture plates (2 × 104 cells/well). After treated with Aβ and XAN for 48 h, the MTT assay was carried out for measuring cell proliferation.[21]
For ALP activity assay, the cultured osteoblasts were added in 24-well plates overnight (5 × 104 cells/well), and treated with XAN or NAC for another 5 days. Then, osteoblasts were cultured with XAN or NAC, which contained 40 μM Aβ for another 2 days. ALP activity was assayed using a commercial kit according to the instruction.
Bone mineralisation assay
The cultured osteoblasts were added in 24-well plates overnight (5 × 104 cells/well), and then cultured with osteogenic differentiation medium (10 nM dexamethasone, 50 μg/ml ascorbic acid and 10 mM β-glycerophosphate) for 18 d. Then, the cells were cultured in XAN or NAC, which contained 40 μM Aβ for another 2 days. Osteoblasts were fixed in ice-cold 4% paraformaldehyde for 10 min, and then dyed with 0.1% Alizarin Red solution at 37°C for 30 min. After washing, osteoblasts were completely dissolved in 10% cetylpyridinium chloride for 15 min, and then measured at 570 nm.
Oxidative stress determination
The cultured osteoblasts were inoculated into 24-well plates (5 × 104 cells/well). After treatment with XAN or NAC for 48 h, the medium was discarded, and the cell lysate was added into each well for 30 min. The lysate was centrifuged at 12 000 rpm for 10 min at 4°C. SOD and CAT activities were assayed using a commercial kit under the instructions.
Cells apoptosis and ROS measurement
The cultured osteoblasts were inoculated into 6-well plates (2 × 105 cells/well). Cells apoptosis rate was determined by the Annexin V-FITC Apoptosis Kit. Briefly, after treatment, osteoblasts were collected and centrifuged at 2000 rpm for 5 min. Then, 195 μL of Annexin V-FITC binding buffer, 5 μL of Annexin V-FITC and 10 μL of PI were added into osteoblasts, and then incubated at 25°C in dark conditions for 20 min.
For intracellular ROS measurement, osteoblasts were treated with reagents for 48 h, and then incubated with 5 μM DCFHDA for 30 min at room temperature. Finally, the intracellular ROS level was analysed with a flow cytometer according to the instruction of ROS Assay Kit.
Western blot
Osteoblasts were seeded in 6-well plates (2 × 105 cells/well). After treatment with reagents for 48 h, osteoblasts were lysed and centrifuged at 12 000 rpm for 10 min at 4°C. Cell protein lysates (20 µg) were separated on 10% sodium dodecyl sulfate-polyacrylamide gels, and electroblotted onto a polyvinylidene fluoride membrane. Membranes were blocked in 5% BSA at 37 °C for 1 h and incubated with OPN, COL-I, Bax, Bcl-2, AKT, p-AKT, Nrf2, NQO1, HO-1, SOD-2 and GAPDH antibodies overnight at 4°C. After washed with TBST for three times, the membranes were incubated with secondary antibodies at 37°C for 1 h. Membranes were treated by enhanced chemiluminescent (ECL) and photographing with Gel imaging system.
Statistical analysis
Data were expressed as means ± SD. All analyses were performed by one-way analysis ANOVA of variance using GraphPad Prism 5.0. P-values of ˂0.05 was considered statistically significant.
Results
Aβ 1–42 oligomer displayed cytotoxicity against osteoblasts
The osteoblasts proliferation was measured via MTT assay. The results showed that Aβ1–42 oligomer decreased the viability of osteoblasts significantly, and the inhibition rate was in a does-dependent manner and reached 53.43% at 40 μM. Therefore, 40 μM Aβ1–42 oligomer was used in subsequent experiments (Figure 1).
Effects of Aβ and XAN on the proliferation, differentiation, mineralization and the expression of bone formation-related proteins in osteoblasts. (A. B) Cell viability measured by MTT assay. (C) ALP activity assay. (D–F) The expression of OPN and COL-Ι by Western blotting. (G, H) The bone mineralisation nodule and extracellular matrix mineralisation level by Alizarin Red staining. The data are expressed as the mean ± SD, n = 4. #P < 0.05, ##P < 0.01 versus control group; *P < 0.05, **P < 0.01 versus model group.
XAN promotes bone formation of osteoblasts injured by Aβ1-42 oligomer
As shown in Figure 1B and C, 40 μM Aβ significantly decreased cell proliferation and ALP activity, while XAN and NAC markedly rescued the proliferation and ALP activity in Aβ-injured osteoblasts. OPN and COL-Ι play a crucial role in bone mineralization, formation and reconstruction. As shown in Figure 1D–F, XAN and NAC significantly up-regulated the content of OPN and COL-Ι in Aβ-injured osteoblasts. In addition, the results of bone mineralization analysis also showed that XAN could significantly reversed the inhibition of Aβ on the bone mineral nodules formation in osteoblasts (Figure 1G, H).
XAN inhibits apoptosis in Aβ1-42 oligomer-injured osteoblasts
The osteoblasts apoptosis rate was analyzed with a flow cytometer. As shown in Figure 2A, B, treatment with Aβ significantly increased the cell apoptosis rate, while XAN and NAC significantly reduced the apoptosis rate of osteoblasts. In addition, as shown in Figure 2C, E, the expression of antiapoptotic protein Bcl-2 was increased after XAN treatment, and XAN also reduced the protein content of Bax in a does-dependent manner. The results suggested XAN exhibited potent anti-apoptosis effects in Aβ-injured osteoblasts.
Effects of XAN on the apoptosis and oxidative stress level in Aβ-injured osteoblasts. (A, B) Cell apoptosis by flow cytometer. (C–E) The expression of Bcl-2 and Bax by western blotting. The data are expressed as the mean ± SD, n = 4. #P < 0.05, ##P < 0.01 versus control group; *P < 0.05, **P < 0.01 versus model group.
XAN inhibits oxidative stress in Aβ1-42 oligomer-injured osteoblasts
As shown in Figure 3A, B, Aβ markedly improved the ROS production in osteoblasts, while XAN markedly reduced the ROS level in a dose-dependent manner. Additional, Aβ markedly decreased the expression of CAT and SOD in osteoblasts. After treatment, XAN (5, 25 µM) significantly improved the SOD level, and XAN (1, 5, 25 µM) up-regulated the CAT level in a does-dependent manner (Figure 3C, D).
Effect of XAN on the oxidative stress level in Aβ-injured osteoblasts. (A, B) Intracellular ROS level by flow cytometer. (C–D) SOD and CAT activities assay by commercial kits. The data are expressed as the mean ± SD, n = 4. #P < 0.05, ##P < 0.01 versus control group; *P < 0.05, **P < 0.01 versus model group.
XAN stimulates PI3K/AKT/Nrf2 pathway in Aβ1-42 oligomer-injured osteoblasts
To evaluate the effect of XAN on AKT/Nrf2 pathway, the proteins expression of Nrf2, p-AKT, HO-1, NQO1 and SOD-2 were assayed by Western blotting. Compared with control group, Aβ markedly reduced the expression of Nrf2, p-AKT, HO-1, NQO1 and SOD-2. After treatment, XAN significantly reversed the decreased expression of Nrf2, p-AKT, HO-1, NQO1 and SOD-2 in Aβ-injured osteoblasts, indicating that XAN could stimulated PI3K/AKT/Nrf2 pathway in Aβ1-42 oligomer-injured osteoblasts (Figure 4).
Effect of XAN on the proteins expression of AKT/Nrf2 pathway in Aβ-injured osteoblasts. (A, B) p-AKT/AKT; (C–D) Nrf2; (E-H) HO-1, NQO1 and SOD-2 by western blotting. The data are expressed as the mean ± SD, n = 3. #P < 0.05, ##P < 0.01 versus control group; *P < 0.05, **P < 0.01 versus model group.
The PI3K inhibitor LY294002 counteracts the beneficial effects of XAN in Aβ-injured osteoblasts
To further demonstrate whether PI3K/AKT pathway participated in the regulation of XAN-induced cytoprotection on Aβ-injured osteoblasts, PI3K inhibitor LY294002 (20 μM) were added. The results showed LY294002 counteracted the beneficial effect of XAN against Aβ-induced decline of cell viability (Figure 5A), ALP activity (Figure 5B), and Aβ-induced high levels of apoptosis (Figure 5C) and ROS (Figure 5D). In addition, compared with the XAN group, LY294002 with or without XAN significantly decreased PI3K expression and its downstream proteins p-AKT, Nrf2, HO-1, NQO1 and SOD-2 (Figure 5E).
Effects of XAN and the selective inhibitor LY294002 on Aβ-induced cytotoxicity in osteoblasts. (A) Cell proliferation measured by MTT assay. (B) ALP activity assay. (C) Cell apoptosis by flow cytometer. (D) Intracellular ROS level by flow cytometer. (E) Relative expression of PI3K, p-AKT/AKT, Nrf2, HO-1, NQO1 and SOD-2 by Western wlotting. The data are expressed as the mean ± SD, n = 3. *P < 0.05, **P < 0.01 versus Aβ group; #P < 0.05, ##P < 0.01 versus Aβ+XAN group.
Nrf2 knockdown counteracts the beneficial effects of XAN in Aβ-injured osteoblasts
To additionally assess the role of Nrf2 in XAN protecting osteoblasts against Aβ-injured bone loss and oxidative stress, Nrf2 was silenced in osteoblasts through transfecting with siRNA-Nrf2. The results showed that Nrf2 knockdown largely counteracted the positive effect of XAN against Aβ-induced decline of cell viability (Figure 6A), ALP activity (Figure 6B) and Aβ-induced severe apoptosis (Figure 6C) and ROS release (Figure 6D). In addition, compared with the XAN group, siNrf2 with or without XAN both significantly decreased the protein expression of Nrf2 and the downstream proteins HO-1, NQO1 and SOD-2 (Figure 6E). These results indicated that XAN could promote bone formation and attenuate oxidative stress in Aβ-injured osteoblasts via regulating PI3K/AKT/Nrf2 signalling pathway (Figure 7).
Effects of XAN on Aβ-induced cytotoxicity in Nrf2-knockdown osteoblasts. (A) Cell proliferation measured by MTT assay. (B) ALP activity assay. (C) Cell apoptosis by flow cytometer. (D) Intracellular ROS level by flow cytometer. (E) Relative expression of Nrf2, HO-1, NQO1 and SOD-2 by Western blotting. The data are expressed as the mean ± SD, n = 3. *P < 0.05, **P < 0.01 versus Aβ group; #P < 0.05, ##P < 0.01 versus Aβ+XAN group.
A proposed signalling pathway involved in XAN against Aβ-induced oxidative damage and apoptosis.
Discussion
OP is a metabolic bone disease, and caused by various factors including aging and oxidative stress. Accumulating evidence suggest that Aβ deposition brings into play an vital role in the pathogenesis of OP. Aβ deposition induces oxidative damage, and eventually leads to bone loss and memory dysfunction.[22] Thus, it is urgent to develop drugs with antioxidative activities to treat OP. Osteoblasts play a crucial role in bone maintenance and regeneration, and reactive oxygen species induced by Aβ deposition can significantly inhibit osteoblastic proliferation and differentiation.[23] This study firstly revealed that XAN possessed osteoprotective effect against oxidative damage via regulating PI3K/AKT/Nrf2 pathway in Aβ-injured osteoblasts.
Oxidative stress, which usually generates excess ROS,[24] brings to play a crucial role in bone metabolisms via inhibiting osteoblastic bone formation, eventually leading to bone loss and OP. Aβ-induced oxidative stress is an important pathological feature of osteoblast apoptosis in OP, and the deposition of Aβ can induce the decrease of osteoblast activity. When Aβ accumulates to a certain extent, it exhibits a detrimental effect and elicits further oxidative stress.[25] Accumulating evidence suggest that Aβ deposition can cause oxidative damage, further leading to bone loss.[26] This study evaluated the effects of Aβ1-42 oligomer on osteoblasts, and found Aβ1-42 oligomer reduced osteoblast activity in a dose-dependent manner in vitro, and significantly reduced ALP activity and bone mineralization, suggesting that Aβ-injured osteoblasts might be a good cell model to induce exogenous Aβ in bone remodelling. Additional, Aβ could also increase the intracellular ROS level, and decrease the antioxidative enzymes expression in osteoblasts, successfully established an oxidative stress condition. XAN, a natural isoflavone, has antioxidant activity and can rescue osteoblasts injures from oxidative damage under various conditions.[27] It was discovered that XAN could ameliorate memory impairment and Aβ deposition in APP/PS1 mice.[13] As same of this study, XAN raised the proliferation and ALP activity, reversed the apoptosis rate, and improved the expression of COL-Ι, OPN and bone matrix mineralization in osteoblasts injured by Aβ1-42 oligomer. Also, XAN reduced ROS level, and improved SOD and CAT activities in Aβ-injured osteoblasts, suggesting that XAN could promote osteogenesis of osteoblasts in oxidative stress conditions stimulated by Aβ deposition.
It has been proved that the activation of PI3K/AKT/Nrf2 pathway is beneficial to relieving oxidative damage under varieties oxidative stress conditions. Studies have found that XAN decreased oxidative stress mediated damages via stimulating PI3K/AKT/Nrf2 pathway in PC12 cells.[12] Specifically, PI3K is a vital signalling molecule for lots of cellular activities, which can catalyse the phosphorylation of D3 hydroxyl of phosphatidylinositol, and come into being phosphatidylinositol-3,4,5-trisphosphate, which further binds to the intracellular pleckstrin homology domain of AKT.[28] Next, the AKT stimulation can regulate the proliferation, differentiation and apoptosis of osteoblasts.[29] Also, AKT can activate the expression of downstream protein Nrf2 through its own phosphorylation,[30] thereby promoting the downstream antioxidant enzymes HO-1, NQO1 and SOD expression, and protecting osteoblast from oxidative stress induced injury. As a major transcriptional regulator, Nrf2 participates in the transcription of various cellular stress responses and antioxidant genes regulation. Meanwhile, Nrf2 is involved in regulating bone metabolism and can positively regulate Runx2 expression.[31] In this work, XAN activated the expression of Nrf2 and p-AKT, as well as up-regulated the antioxidant proteins including HO-1, NQO1 and SOD-2 in osteoblasts injured by Aβ1-42 oligomer. More importantly, treatment with PI3K/AKT blocker LY294002 and Nrf2 knockdown counteracted the beneficial effects of XAN in Aβ-treated osteoblasts, as evidenced of the decrease of cell proliferation and differentiation, increase of apoptosis and ROS level, and the inhibition of PI3K, p-AKT, Nrf2, HO-1, NQO1, SOD-2 expression. These results indicated that XAN relieving oxidative damage and promoting bone formation major involved in PI3K/AKT/Nrf2 signalling pathway.
In addition, Aβ deposition increased the ROS level in the early phase, and XAN reduced ROS release in osteoblasts after 48 h treatment. Persistent Aβ deposition leads to continuous oxidative stress and further brought out the inhibition of PI3K, AKT and AKT-mediated Nrf2, ultimately damaging the osteoblasts function. Also, it was confirmed that concomitant inhibition of PI3K/AKT pathway could increase apoptosis rate. Activation of PI3K can induce AKT phosphorylation, which signals to various downstream substrates, such as Nrf2, GSK-3, FOXO and mTOR, and modulates the cell apoptosis.[32] Among the downstream substrates of PI3K/AKT pathway, Nrf2 is an vital transcription factor for regulating antioxidant proteins, which can up-regulate antioxidant protection and reduce the cells apoptosis.[33] Nrf2 is activated under the low levels of oxidative stress and involved in the regulation of apoptotic outcome. Under basal conditions, Keap1 interacts with Nrf2 that mediates its degradation through the ubiquitin–proteasome pathway.[34] Oxidative stress-induced modification of cysteine thiols in Keap1 and Nrf2 molecules alters their structures, promoting the Nrf2 release from the complex with Keap1. Unbound Nrf2 translocates into the nucleus where it binds to the ARE in the upstream promoter region, and initiates transcriptional activation of Nrf2-target antioxidant and detoxification genes that will reduce harmful pro-apoptotic effects of free radicals.[35] In this study, XAN could activate the PI3K and AKT, and proceed to signal the downstream substrate Nrf2 in Aβ-injured osteoblasts. The activated Nrf2 further up-regulated the expression of its downstream antioxidant enzymes, finally inhibiting cell apoptosis and promoting osteoblasts activity. Altogether, these further demonstrate that PI3K/AKT/Nrf2 pathway is participated in the ameliorative effects of XAN on Aβ-induced oxidative damage and bone loss.
Conclusion
On the whole, it was demonstrated for the first time that XAN manifested effective improvement of oxidative stress and bone loss in Aβ-injured osteoblasts through promoting the cell proliferation, differentiation and mineralization, and inhibiting the apoptosis and ROS level. In addition, it was revealed that XAN exerted these effects through regulating PI3K/AKT/Nrf2 signalling pathway. Given the outstanding antioxidant and antiosteoporosis properties of XAN, it may prove to be a promising candidate for the prevention and treatment of OP.
Author Contributions
Tianshuang Xia: Writing-original draft, Investigation. Xiaoyan Liu: methodology, data curation. Nani Wang: investigation. Yiping Jiang: data curation. Huanhuan Bai: methodology. Weifan Xu: investigation, data curation. Kunmiao Feng: methodology. Ting Han: data curation. Hailiang Xin: conceptualization, methodology, writing-original draft. All authors have read and agreed with the author statement of the manuscript.
Funding
This work was supported by the National Natural Science Foundation of China (82174079, 82004015, U1603283), Project of Science and Technology Commission of Shanghai Municipality (21S21902600).
Conflict of Interest
The authors declare that they have no competing interests.
References
