Cav1.2 regulates osteogenesis of bone marrow‐derived mesenchymal stem cells via canonical Wnt pathway in age‐related osteoporosis

Abstract Aims Age‐related bone mass loss is one of the most prevalent diseases that afflict the elderly population. The decline in the osteogenic differentiation capacity of bone marrow‐derived mesenchymal stem cells (BMMSCs) is regarded as one of the central mediators. Voltage‐gated Ca2+ channels (VGCCs) play an important role in the regulation of various cell biological functions, and disruption of VGCCs is associated with several age‐related cellular characteristics and systemic symptoms. However, whether and how VGCCs cause the decreased osteogenic differentiation abilities of BMMSCs have not been fully elucidated. Methods Voltage‐gated Ca2+ channels related genes were screened, and the candidate gene was determined in several aging models. Functional role of determined channel on osteogenic differentiation of BMMSCs was investigated through gain and loss of function experiments. Molecular mechanism was explored, and intervention experiments in vivo and in vitro were performed. Results We found that Cav1.2 was downregulated in these aging models, and downregulation of Cav1.2 in Zmpste24−/− BMMSCs contributed to compromised osteogenic capacity. Mechanistically, Cav1.2 regulated the osteogenesis of BMMSCs through canonical Wnt/β‐catenin pathway. Moreover, upregulating the activity of Cav1.2 mitigated osteoporosis symptom in Zmpste24−/− mice. Conclusion Impaired osteogenic differentiation of Zmpste24−/− BMMSCs can be partly attributed to the decreased Cav1.2 expression, which leads to the inhibition of canonical Wnt pathway. Bay K8644 treatment could be an applicable approach for treating age‐related bone loss by ameliorating compromised osteogenic differentiation capacity through targeting Cav1.2 channel.


| INTRODUC TI ON
Aging is a progressive deterioration of physiological functions accompanied by bone loss, leading to bone fragility, a disease that is known as osteoporosis (Singh, Brennan, et al., 2016). With advancing age, the balance between bone formation and resorption is compromised, resulting in overall bone loss and structural damage. It has been confirmed that the impaired capacity of osteogenic differentiation of BMMSCs from the elderly people is a causative factor leading to age-related osteoporosis (Baker, Boyette, & Tuan, 2015).
However, the underlying mechanisms that account for impaired osteogenic differentiation of aging BMMSCs remain unknown.
Calcium channels, which are in charge of calcium transportation into and out of cells and organelles, play an important role in aging (Surmeier, 2007;Warnier et al., 2018). In terms of bone aging, numerous studies have indicated the relationship between disruption of calcium channels and age-related bone mass loss (Agacayak et al., 2014;Shimizu et al., 2012). There are several kinds of calcium channels, among which voltage-gated Ca 2+ channels (VGCCs) are important components. They are formed by α1, α2δ, β, and γ subunits and are transmembrane surface proteins indispensable for many physiological events including membrane depolarizes, cellular motility, gene expression, and differentiation (Zamponi, Striessnig, Koschak, & Dolphin, 2015). On the basis of the differences in α1 subunits, there are three families of VGCCs in mammals known as Ca v 1, Ca v 2, and Ca v 3 (Zamponi, 2016). Ca v 1, also termed as L-type calcium channel, can be further subdivided  (Silva et al., 2017;Zamponi, 2016). Previous study confirmed that age-associated downregulation of Ca v 3.1 was involved in increased production of amyloid beta peptide 1-42 in N2a cells that led to neurodegenerative diseases (Rice, Berchtold, Cotman, & Green, 2014). However, unlike other age-related diseases, the relationship between VGCCs and age-related bone loss remains a great dispute. Several studies showed that the activation of VGCCs played an important role in bone formation (Li, Duncan, Burr, Gattone, & Turner, 2003;Li, Duncan, Burr, & Turner, 2002;Noh, Park, Zheng, Ha, & Yim, 2011), while others reported that inhibition of VGCCs could promote bone formation and suppress bone resorption (Marie, 2010;Ritchie, Maercklein, & Fitzpatrick, 1994). Besides, whether and how VGCCs cause the decreased osteogenic differentiation of BMMSCs is poorly explored at the molecular level.
It has been widely acknowledged that canonical Wnt pathway plays a significant role in the regulation of BMMSCs osteogenic differentiation (Cook, Fellgett, Pownall, O'Shea, & Genever, 2014;Lin & Hankenson, 2011). The canonical Wnt pathway causes an accumulation of β-catenin in the cytoplasm and its eventual translocation into the nucleus to act as a transcriptional coactivator. In Wnt signaling, GSK3β, a negative regulator of the canonical Wnt signaling, forms a multimeric complex with APC, AXIN1, and β-catenin and leads to β-catenin degradation by targeting it for ubiquitination.
Phosphorylation of GSK3β can interfere with the complex and block β-catenin degradation. Numerous studies have confirmed that inhibition of Wnt/β-catenin antagonists, such as sclerostin, DKK1, and WIF-1, can emerge as a promising therapeutic approach in the treatment of osteoporosis, and some are even in Phase III studies (Chen et al., 2014;Ke, Richards, Li, & Ominsky, 2012). However, it remains unknown whether VGCCs could regulate osteogenic differentiation through canonical Wnt pathway.
Zmpste24 is a metalloproteinase that partakes in the maturation of lamin A. Zmpste24-deficient mice recapitulate many phenotypes observed in physiological aging, including reduced bone density and an increased risk of fracture (Ghosh & Zhou, 2014).
The availability of Zmpste24-deficient mice could provide a new model to aid the study of mechanistic events underlying age-related osteoporosis. In this study, we provide insights into the importance of VGCCs in regulating age-related bone mass loss and the mechanism that underlie VGCCs mediated bone mass loss in Zmpste24-deficient mice. We screened the expression of VGCCs

| Ca v 1.2 is downregulated in aging BMMSCs with impaired osteogenic differentiation
Genotype identification of wild-type, Zmpste24+/-, and Zmpste24−/− mice was shown in Figure S1. Firstly, we used micro-CT to measure relative density of bone mineral from 3-month-old wild-type and Zmpste24-deficient mice, and micro-CT images showed reduced mineralization of Zmpste24deficient mice (Figure 1a). Bone mineral density (BMD), bone volume/total volume (BV/TV), and trabecular number (Tb.N) were also significantly decreased in Zmpste24-deficient mice as measured by micro-CT densitometry (Figure 1b showed that lower expressions of Ca v 1.2 and Ca v 2.2 in BMMSCs from Zmpste24-deficient mice compared to the wild-type mice ( Figure 1k). We also screened the expressions of VGCCs related genes in 3-month-old SAMR1 mice and SAMP6 mice. The results showed that Ca v 1.2 and osteogenesis were still downregulated in BMMSCs from 3-month-old SAMR1 mice compared with SAMP6 mice ( Figure S2a,b). Besides, we also explored expressions of VGCCs related genes in BMMSCs from young individuals compared with that from old individuals (donor information was listed in Table S1) and found changes of VGCCs associated genes, among which Ca v 1.2 was still downregulated ( Figure S3a,b).
Moreover, we also confirmed downregulated Ca v 1.2 expression of BMMSCs in natural aging model (Figure 1m). To further confirm the change of Ca v 1.2 during aging process, we also investigated the protein level of Ca v 1.2. Western blot analysis showed that Ca v 1.2 protein was also downregulated in Zmpste24−/− and natural aging mice (Figure 1l,n).

| Ca v 1.2 regulates osteogenic differentiation of aging BMMSCs
With the aim of investigating the potential role of Ca v 1.2 on im- We also investigated the effect of Ca v 1.2 on adipogenic differentiation as BMMSCs exhibit an age-dependent reduction in osteogenic differentiation with an increased tendency toward adipocyte differentiation . Our results also confirmed the increased adipogenic differentiation ability of Zmpste24−/− BMMSCs ( Figure S4a). To further explore the effect of Ca v 1.2 on adipogenic differentiation ability of Zmpste24−/− BMMSCs, we overexpressed Ca v 1.2 in Zmpste24−/− BMMSCs and no significant difference of adipogenic differentiation was found compared to Zmpste24−/− BMMSCs ( Figure S4b).

| Ca v 1.2 upregulates osteogenic differentiation of BMMSCs via Wnt/βcatenin pathway
Given that canonical Wnt/β-catenin signaling pathway plays an essential role in regulating osteogenic differentiation of BMMSCs, we speculated that Ca v 1.2 regulated osteogenic differentiation via canonical Wnt/β-catenin signaling. In order to confirm our hypothesis, we first detected canonical Wnt/β-catenin pathway in wild-type and Zmpste24−/− BMMSCs to determine whether the canonical Wnt/β-catenin signaling is disrupted in Zmpste24−/− BMMSCs. To investigate whether Ca v 1.2 regulated osteogenic differentiation through Wnt/β-catenin signaling, we used siRNA to knockdown . We also investigated the other two signal pathways, PKC-ERK1/2 (Muchekehu & Harvey, 2008;Soletti et al., 2010) and CAMKII (Liu et al., 2018;Wei, Wang, Wang, & Bai, 2017), as they can also be regulated by calcium channel. We found that the expression of Ca v 1.2 has no significant effect on these two signal pathways after inhibiting Ca v 1.2 in wild-type BMMSCs ( Figure S6a) and overexpressing Ca v 1.2 in Zmpste24−/− BMMSCs ( Figure S6b), which further verified that Ca v 1.2-mediated osteogenic differentiation was through Wnt/β-catenin pathway.

| Bay K8644 treatment activates Ca v 1.2 and rescues the impaired osteogenic differentiation of ZMPSTe24−/− BMMSCs
The notion that reduced Ca v 1.2 expression contributes to compromised osteogenic differentiation of Zmpste24−/− BMMSCs suggested that it might be possible to rescue the differentiation through activating channels pharmacologically. Considering that Ca v 1.2 belongs to L-type calcium channel, we investigated this possibility by using the potent L-type Ca 2+ channel agonist Bay K8644. Firstly, we determined the optimal concentration of Bay K8644 by measuring cell proliferation, intracellular calcium current, and calcium concen-

| Intraperitoneal injection of Bay K8644 mitigates osteoporosis symptom in Zmpste24−/− mice
To investigate whether intraperitoneal injection of Bay K8644 could alleviate osteoporosis symptom of Zmpste24−/− mice, 1 mg/kg Bay K8644 was administrated to 3-month-old Zmpste24−/− mice through intraperitoneal injection. The scheme for intraperitoneal injection was shown in Figure 5a. The results showed that this dosage treatment could improve intracellular calcium current and calcium concentration of Zmpste24−/− BMMSCs after 5 days stimulation (Figure 5b,c). We also set the groups of DMSO-treated Zmpste24−/− and DMSO-treated wild-type mice. The treatment was initiated when male mice at 3 months of age and was given every 3 days.
After treating for two months, mice were sacrificed for Micro-CT analysis or processed to perform calcein labeling assay. As shown in Figure 5d-g, BMD, BV/TV, and Tb.N were significantly increased in Bay K8644-treated Zmpste24−/− mice compared to DMSO-treated Zmpste24−/− mice. In addition, Bay K8644 treatment also significantly enhanced bone formation as carried out by double-calcein F I G U R E 1 Ca v 1.2 is downregulated in aging BMMSCs with impaired osteogenic differentiation. (a) Bone masses of 3-month-old wildtype and Zmpste24-deficient mice were tested by micro-CT (n = 3), and the "interesting zone" was highlighted. (b-d) Bone mineral density (BMD), bone volume/total volume (BV/TV), and trabecular number (Tb.N) of 3-month-old wild-type and Zmpste24-deficient mice were analyzed with the micview software. (e) Bone formations of 3-month-old wild-type and Zmpste24−/− mice were examined by double-calcein labeling (n = 3). Scale bar, 25 um. (f) Mineral apposition rate (MAR) was analyzed by Image J under fluorescence microscope. (g) Expressions of osteogenic-related genes of ALP, Runx2, and OCN in wild-type and Zmpste24-deficient mice were detected by qRT-PCR after osteogenic induction for 5 days (n = 3). The results were normalized to GAPDH. (h) Expressions of osteogenic-related proteins of ALP, Runx2, and OCN in wild-type and Zmpste24-deficient mice were detected by Western blot after osteogenic induction for 7 days (n = 3). GAPDH was used as an internal control. (i) Mineralized nodules of BMMSCs from wild-type and Zmpste24-deficient mice were assayed by alizarin red staining after osteogenic induction for 14 days and quantified with a spectrophotometer after dissolving with cetylpyridinium chloride (n = 3). (j) Alizarin red staining was performed to detect mineralized nodules of BMMSCs from 3-month-old to 18-month-old normal mice after osteogenic induction for 14 days and quantified with a spectrophotometer after dissolving with cetylpyridinium chloride (n = 3). (k) Expressions of voltage-gated Ca 2+ channels (VGCCs) related genes in 3-month-old wild-type and Zmpste24-deficient mice were explored by qRT-PCR (n = 7). The results were normalized to GAPDH. (l) The protein level of Ca v 1.2 in 3-month-old wild-type and Zmpste24-deficient mice was explored by Western blot (n = 7). (m) Expression of Ca v 1.2 in 3-month-old young mice and 18-month-old natural aging mice was explored by qRT-PCR (n = 7). The results were normalized to GAPDH. (n) The protein level of Ca v 1.2 in 3-month-old young mice and 18month-old natural aging mice was explored by Western blot (n = 7). Scale bar of alizarin red, 50 um. Data are shown as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, which was determined by unpaired two-tailed Student's t test labeling (Figure 5h,i). However, we also found that BMD, BV/TV, and Tb.N were still lower compared with DMSO-treated wild-type group (Figure 5d-g), which indicated that Bay K8644 could only partly ameliorate osteoporosis symptom. Consistent with such observations, double-calcein labeling also implied partial recovery of osteoporosis features after Bay K8644 treatment (Figure 5h,i).
F I G U R E 2 Ca v 1.2 regulates osteogenic differentiation of aging BMMSCs. (a) Wild-type BMMSCs were transfected with scramble siRNA or Ca v 1.2 siRNA, and the transfection efficiency was tested by qRT-PCR after transfection for 48 hr (n = 3). (b) Expressions of osteogenic-related genes of ALP, Runx2, and OCN in wild-type BMMSCs transfected with scramble siRNA or Ca v 1.2 siRNA were detected by qRT-PCR after osteogenic induction for 5 days (n = 3). (c) Expressions of osteogenic-related proteins of ALP, Runx2, and OCN in wild-type BMMSCs transfected with scramble siRNA or Ca v 1.2 siRNA were detected by Western blot after osteogenic induction for 7 days (n = 3). (d) Alizarin red staining of wild-type BMMSCs transfected with scramble siRNA or Ca v 1.2 siRNA were performed after osteogenic induction for 14 days (n = 3). (e) Wild-type and Zmpste24−/− BMMSCs were transfected with control overexpression vector or Ca v 1.2 plasmid, and the transfection efficiency was tested by qRT-PCR after transfection for 48 hr (n = 3). (f) Wild-type and Zmpste24−/− BMMSCs were transfected with control overexpression vector or Ca v 1.2 plasmid, and the expressions of osteogenic-related genes of ALP, Runx2, and OCN were detected by qRT-PCR after osteogenic induction for 5 days (n = 3). (g) Wild-type and Zmpste24−/− BMMSCs were transfected with control overexpression vector or Ca v 1.2 plasmid, and the expressions of osteogenic-related proteins of ALP, Runx2, and OCN were detected by Western blot after osteogenic induction for 7 days (n = 3). (h) Wild-type and Zmpste24−/− BMMSCs were transfected with control overexpression vector or Ca v 1.2 plasmid, and alizarin red staining was performed after osteogenic induction for 14 days (n = 3). The expression levels of the target genes and proteins were normalized to GAPDH. Scale bar, 50 um. Data are shown as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, which was determined by paired two-tailed Student's t test

| D ISCUSS I ON
In this study, we investigated the role of VGCCs played in regulating age-related bone mass loss and their potential mechanism. We found the bone mass of Zmpste24−/− mice was reduced and the osteo- Mechanistically, we elucidated that Ca v 1.2 regulated osteogenic differentiation ability by targeting Wnt/β-catenin signaling. We also illuminated that Bay K8644 can be an effective strategy for treating age-related osteoporosis in virtue of activating Ca v 1.2 channel and downstream Wnt/β-catenin signaling and promoting osteogenic differentiation of BMMSCs.
Aging is a gradual deterioration of physical abilities, and osteoporosis is one of its critical features. The molecular basis of osteoporosis during aging remains ambiguous. Compromised osteogenic differentiation abilities of BMMSCs have been demonstrated to be one of key factors contributing to age-related osteoporosis. Zmpste24-deficient mice recapitulate many phenotypes observed in physiological aging, and our study also showed the defective bone phenotype of Zmpste24−/− mice and decreased osteogenic differentiation of Zmpste24−/− BMMSCs, which was consistent with previous report (Rivas, Li, Akter, Henderson, & Duque, 2009). It has been borne out by numerous studies that aberrant signal pathways are involved in age-related bone mass loss, including Notch (Roforth et al., 2014) and Wnt (Garcia-Velazquez & Arias, 2017). These signaling pathways are intricate with a mixed network. Targeting the upstream regulatory mechanism may yield more effective results in treatment of age-related bone mass loss. VGCCs are the crucial influence factor of signal pathways and cell function. They encompass several channels with distinct functions (Striessnig, 2016). For example, Ca v 1.1 can mediate the process of excitation-contraction coupling (Wu et al., 2016), and Ca v 1.3 is associated with pace-making in dopaminergic neurons (Singh, Verma, et al., 2016). Abnormity of these channels can lead to a variety of disorders (Zamponi et al., 2015). Numerous studies have confirmed the engagement of VGCCs in age-related disease, including Parkinson disease (Ca v 1.3) (Zamponi et al., 2015), ataxia (Ca v 2.1) (Dorgans et al., 2017), Alzheimer's disease (Ca v 3.1) (Rice et al., 2014). Despite this, whether and how VGCCs exert influence on osteoporosis remains controversial. Several studies reported that the activation of VGCCs could promote bone formation (Li et al., 2003(Li et al., , 2002Noh et al., 2011), while others found that inhibition of VGCCs could suppress osteogenesis (Marie, 2010;Ritchie et al., 1994). In this study, we found decreased Ca v 1.2 expressions in several bone aging models, implying that Ca v 1.2 is an integral part of intricate mechanisms underlying age-related osteoporosis.

mias, and neuropsychiatric diseases. Previous study has shown that
Ca v 1.2 can promote osteoblast differentiation and bone formation and rescue estrogen deficiency-induced osteoporosis (Cao et al., 2017). However, whether Ca v 1.2 is involved in age-related osteoporosis remains unknown. In our study, we confirmed Ca v 1.2 is involved in regulating BMMSCs osteogenic differentiation through gain and loss of function experiments.
Next, we investigated how Ca v 1.2 participates in regulating defective osteogenic differentiation of aging BMMSCs. It has been widely accepted that canonical Wnt pathway is implicated in the regulation of BMMSCs osteogenic differentiation (Cook et al., 2014;Lin & Hankenson, 2011), and several studies have confirmed the link between calcium channel and Wnt/β-catenin signaling (Liu et al., 2014;Sagredo et al., 2018). For example, a previous study showed that Wnt/β-catenin signaling could be regulated by transient receptor potential cation channel subfamily M member 4 (TRPM4) (Sagredo et al., 2018). Thus, we wondered whether Ca v 1.2 regulated defective osteogenic differentiation of aging BMMSCs through Wnt/β-catenin signaling. In our study, we observed that canonical Wnt pathway was decreased in Zmpste24−/− BMMSCs and improving this signal pathway could ameliorate impaired osteogenic differentiation of aging BMMSCs. These results suggested that impaired osteogenic differentiation of BMMSCs was, to some degree, due to decreased canonical Wnt pathway. We also overexpressed Ca v 1.2 in Zmpste24−/− BMMSCs in the context of siRNA-mediated β-catenin silencing followed by analysis of the osteogenic differentiation, and the results further confirmed the involvement of canonical Wnt pathway for Ca v 1.2-mediated osteogenic differentiation.
Since Ca v 1.2 belongs to L-type calcium channel and Bay K8644 is regarded as a potent L-type Ca 2+ channel agonist, we We further demonstrated the feasibility of Bay K8644 to treat age-related bone loss. However, despite the results that osteogenic differentiation deficiency and bone mass loss being improved upon receiving Bay K8644 stimulation, this method had limited effects as results from micro-CT and double-calcein labeling only showed partial recovery of osteoporosis features after Bay K8644 treatment. We ascribed it to the complex etiology of age-related bone mass loss as it is a disease involving multiple factors, of which defective osteogenic differentiation is just one element (Fossett, Khan, Pastides, & Adesida, 2012;Singh, Brennan, et al., 2016;Wang et al., 2014). In terms of Zmpste24-deficient mice, age-related features mainly result from devastation of nuclear envelope and accumulation of progerin (Lee et al., 2016). Therefore, looking for another upstream mechanism for age-related bone mass loss will be the focus of our future research on treating age-related bone mass loss.
In summary, we indicate that defective osteogenic differentiation of Zmpste24−/− BMMSCs can be partly attributed to the inhibition of canonical Wnt pathway which is partly owing to decreased Ca v 1.2 expression, and Bay K8644 treatment could be an applicable approach for treating age-related bone loss by ameliorating compromised osteogenic differentiation capacity through targeting Ca v 1.2 channel and canonical Wnt pathway. Taken together, we propose a new mechanistic explanation underlying age-related bone loss and propose a feasible strategy via Ca v 1.2 and Wnt/β-catenin pathway activation. Hong Kong, which have been described previously (Pendas et al., 2002), and the wild-type mice were obtained from the breeding of Zmpste24+/-mice. Male littermates were used all through the study.

| Animal
Genotype was identified via real-time polymerase chain reaction, and the primers are listed in Table S2.

| Osteogenic differentiation
Before osteogenic induction, 2 × 10 5 BMMSCs were seeded on 12well culture dishes and cultured in the growth medium until the cells reached 80%-90% confluence as defined by area covered. BMMSCs were then cultured under osteogenic culture conditions containing 100 μg/ml ascorbic acid (Sigma), 5 mmol/l β-glycerophosphate (Sigma), and 10 nmol/l dexamethasone (Sigma). After osteogenic induction for 5 days, quantitative RT-PCR (qRT-PCR) was performed F I G U R E 3 Ca v 1.2 upregulates osteogenic differentiation of BMMSCs via Wnt/β-catenin pathway. (a) Expressions of GSK3β, p-GSK3β, β-catenin, and active-β-catenin in wild-type and Zmpste24−/− BMMSCs were detected by Western blot analysis (n = 7). (b) Wild-type BMMSCs were transfected with scramble siRNA or Ca v 1.2 siRNA, and the protein expression levels of GSK3β, p-GSK3β, β-catenin, and active-β-catenin were confirmed by Western blot analysis after 72 hr (n = 3). (c) Zmpste24−/− BMMSCs were transfected with control overexpression vector or Ca v 1.2 plasmid, and the protein expression levels of GSK3β, p-GSK3β, β-catenin, and active-β-catenin were confirmed by Western blot analysis after 72 hr (n = 3). (d) Wnt target genes of cyclin D1 and c-myc were explored in wild-type and Zmpste24−/− BMMSCs by qRT-PCR (n = 7). (e) Wild-type BMMSCs were transfected with scramble siRNA or Ca v 1.2 siRNA, and Wnt target genes of cyclin D1 and c-myc were explored by qRT-PCR (n = 3). (f) Zmpste24−/− BMMSCs were transfected with control overexpression vector or Ca v 1.2 plasmid, and Wnt target genes of cyclin D1 and c-myc were explored by qRT-PCR (n = 3). (g) 48 hr after transfection with control overexpression vector or Ca v 1.2 plasmid in progerin-overexpressed 293T cells, TOP/FOP, and Renilla luciferase plasmid were transfected, and TOP/FOP reporter assay was performed to detect the activity of β-catenin after 24 hr (n = 3). (h) Alizarin red staining of Zmpste24−/− and 5 mM LiCl-treated Zmpste24−/− BMMSCs was performed to detect mineralized nodules after osteogenic induction for 14 days (n = 3). (i) Expressions of osteogenic-related proteins of ALP, Runx2, and OCN in Zmpste24−/− BMMSCs from control group, Ca v 1.2overexpressed and Ca v 1.2-overexpressed in the context of β-catenin siRNA were detected by Western blot after osteogenic induction for 7 days (n = 3), and each group contains equal volume of transfection agent. (j) Alizarin red staining of Zmpste24−/− BMMSCs from control group, Ca v 1.2-overexpressed and Ca v 1.2-overexpressed in the context of β-catenin siRNA were performed after osteogenic induction for 14 days (n = 3), and each group contains equal volume of transfection agent. The expression levels of the target proteins were normalized to GAPDH. Scale bar, 50 um. Data are shown as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, which was determined by paired two-tailed Student's t test F I G U R E 4 Bay K8644 treatment activates Ca v 1.2 and rescues the impaired osteogenic differentiation of Zmpste24−/− BMMSCs. (a) After treated by DMSO or 10 -6 , 10 -7 , and 10 -8 M Bay K8644 for 3 days, the proliferation of Zmpste24−/− BMMSCs was detected by MTT (n = 3). (b) After treated by DMSO or 10 -6 , 10 -7 , and 10 -8 M Bay K8644 for 3 days, intracellular calcium current of Zmpste24−/− BMMSCs was explored by laser confocal microscopy (n = 3). (c) After treated by DMSO or 10 -6 , 10 -7 , and 10 -8 M Bay K8644 for 3 days, intracellular calcium concentration of Zmpste24−/− BMMSCs was explored by flow cytometry (n = 3). (d) After treated by 10 -7 M Bay K8644 or DMSO for 3 days, the protein expression levels of GSK3β, p-GSK3β, β-catenin, and active-β-catenin in Zmpste24−/− BMMSCs were confirmed by Western blot analysis (n = 3). (e) After treated by 10 -7 M Bay K8644 or DMSO for 3 days, Wnt target genes of cyclin D1 and c-myc were explored by qRT-PCR (n = 3). (f) 48 hr after stimulated by DMSO or Bay K8644 in progerin-overexpressed 293T cells, TOP/FOP, and Renilla luciferase plasmid were transfected, and TOP/FOP reporter assay was performed to detect the activity of β-catenin after 24 hr (n = 3). (g) After osteogenic induction for 7 days, expressions of osteogenic-related proteins of ALP, Runx2, and OCN in Zmpste24−/− BMMSCs of DMSO-treated, Bay K8644-treated, and Bay K8644-treated in the context of Ca v 1.2 siRNA were explored (n = 3). (h) After osteogenic induction for 14 days, mineralized nodule formations in Zmpste24−/− BMMSCs of DMSO-treated, Bay K8644-treated and Bay K8644-treated in the context of Ca v 1.2 siRNA were explored (n = 3). The expression levels of the target genes and proteins were normalized to GAPDH. Scale bar, 50 um. Data are shown as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, which was determined by paired two-tailed Student's t test for the expressions of alkaline phosphatase (ALP), runt-related transcription factor 2 (Runx2), and osteocalcin (OCN). Osteogenesis related proteins of ALP, Runx2, and OCN were assayed by Western blot after osteogenic induction for 7 days. Alizarin red staining was used to assess calcium deposit on day 14, and 10% cetylpyridinium chloride was added for quantitative analysis. The absorbance values were measured at 562 nm.
For the control groups of siRNA or plasmid transfection experiments, scramble siRNA (RiboBio) or control overexpression vector pcDNA6/V5-His (Invitrogen) was included. Lipo2000 (Invitrogen) was used as a transfection reagent according to the manufacturer's instructions. After transfection, the culture medium was substituted by normal culture medium and cells were harvested at 48 hr for RNA and 72 hr for protein extraction. For detection of the osteogenic differentiation capacity, transfection medium was removed in the next day and replaced by osteogenic induction medium.

| Quantitative RT-PCR
Total cellular RNA was extracted from cells using the TRIzol reagent™ (Invitrogen) according to the manufacturer's instructions and cDNA was synthesized in a 20 μl reaction volume (Takara). The qRT-PCR reactions were performed in a total volume of 10 μl using the SYBR Premix Ex Taq™II kit (Takara) and then detected on the CFX96 Real-Time System (Bio-Rad). Fold changes of mRNA were calculated by the 2 −ΔΔC t method after normalization to the expression of GAPDH.
The primer set sequences used for this study are listed in Table S3.

| Western blot analysis
Cells were washed with PBS twice and lysed in RIPA lysis buffer

| TOP/FOP flash reporter assay
1 × 10 5 293T cells were seeded into 24-well culture dishes and grown to 80%-90% confluence followed by serum starvation for 2 hr. To detect the activity of β-catenin after Ca v 1.2 overexpression, cells were transfected with progerin (OBiO Technology) and pcDNA6/V5-His plasmids in the control groups and progerin and Ca v 1.2 plasmids in the experimental groups. To detect the activity of β-catenin after 10 −7 M Bay K8644 treatment, cells were transfected with progerin plasmid and stimulated with DMSO in the control groups and progerin plasmid and stimulated with Bay K8644 in the experimental groups. After 48 hr, the cells were co-transfected with 0.25 μg TOP (Millipore) containing the TCF/LEF consensus sequence or 0.25 μg FOP (Millipore) containing mutated TCF binding sites flash luciferase reporter vector and 0.05ug Renilla luciferase plasmid pRL-SV40 (Promega). Lipo2000 was used as a transfection reagent. 24 hr later, the luciferase assay was performed to detect the activity of β-catenin by using dual-luciferase reporter assay kit (Promega, #E1910). GloMax 20/20 (Promega) was used to detect Firefly or Renilla luciferase activity.

| MTT assay
For the detection of cell proliferation, 1 × 10 4 Zmpste24−/− BMMSCs were seeded on ninety-six-well culture dishes. One day later (day 1), the culture medium was moved, and growth medium containing DMSO, 10 −6 M, 10 −7 M, and 10 −8 M Bay K8644 were added to the wells, respectively. MTT assay was performed on day 4. The process was as followed: First, 20 μl MTT (5 mg/ml in PBS) was added to the well; second, after culturing for 4 hr, the medium was carefully moved and 150 μl DMSO was added; and third, after vibrating for 10 min, the quantitative analysis was performed with 490 nm absorbance values.

| Intracellular calcium concentration analysis
For the detection of intracellular calcium concentration, BMMSCs were incubated with 5 μM Fluo-3/AM dye (Invitrogen, Life Technology) for 60 min at 37°C, followed by washing with PBS for three times. After digestion by trypsin, 2 × 10 5 cells were counted with PBS washing for two times. Then, cells were subjected to flow cytometric analysis (Beckman Coulter). Data are presented as fluorescence intensity.

| Calcium imaging
For the calcium imaging, BMMSCs were incubated with 5 μM Fluo-3/AM dye for 30 min at 37°C, followed by washing with calibrated EGTA/Ca 2+ solutions for three times. Images were collected every 4 s at 2 Hz with excitation at 488 nm and emission at 530 nm by confocal laser microscopy (Zeiss, Oberkochen FV1000). 30 mM KCl was added to the culture medium at 60 s. Fluo-3 fluorescence intensity reflected intracellular Ca 2+ level, which was described previously (Merritt, McCarthy, Davies, & Moores, 1990). The formula of Fluo-3 fluorescence intensity increase ratio is (F-F 0 )/F 0 . F means the fluorescence value detected and F 0 represents the minimum fluorescence value.

| Micro-CT analysis
To assess bone mass in mice, micro-CT (eXplore Locus SP, GE Healthcare) was used to scan femora. After sacrifice, bones were isolated and fixed in 4% paraformaldehyde overnight. The distal femoral metaphysis was scanned at a voltage of 80 kV with a current of 80 μA, and the voxel-size (mm) was (0.015898 × 0.015898 × 0.01 5898). The region of the trabecular bone for analyzing was defined from 0.1 to 1.5 mm away from the epiphyseal growth plate. The trabecular bone from each selected slice was segmented for three-dimension reconstruction. Parameters of BMD, BV/TV, and Tb.N were analyzed with the Micview software.

| Calcein labeling assay
Mice were intraperitoneal injected with 20 mg/kg calcein (Sigma) at day 14 and day 2 before sacrifice. The femora were isolated and fixed in 80% ethanol at 4°C for 48 hr. After using graded ethanol to dehydrate, the femora were embedded in methyl methacrylate and longitudinal cut into 30 μm thick sections using a hard tissue slicing machine (SP1600, Leica). Bone dynamic histomorphometric analyses for mineral apposition rate (MAR) were performed according to the standardized nomenclature for bone histomorphometry under fluorescence microscopy (Leica DM 6000B, German).

| Statistical analyses
All experiments were repeated at least three times, and data are presented as mean ± SD. Comparisons between two groups were performed using paired or unpaired two-tailed Student's t test.