Regulator of G protein signaling protein 12 is required for mouse osteoblast differentiation through controlling calcium channel/Gαi-calcium oscillation-ERK signaling

Abstract

Bone homeostasis is intimately relied on the balance between osteoblasts (OBs) and osteoclasts (OCs). Our previous studies have revealed that regulator of G protein signaling protein 12 (Rgs12), the largest protein in the Rgs super family, is essential for osteoclastogenesis from hematopoietic cells and OC precursors. However, how Rgs12 regulates OB differentiation and function is still unknown. To understand that, we generated an OB-targeted Rgs12 conditional knockout (CKO) mice model by crossing Rgs12^fl/fl mice with Osterix (Osx)-Cre transgenic mice. We found that Rgs12 was highly expressed in both OB precursor cells (OPCs) and OBs of Wide Type (WT) mice, and gradually increased during OB differentiation, whereas Rgs12-CKO mice (Osx; Rgs12^fl/fl) exhibited a dramatic decrease in cancellous bone mass, evidenced by reduced percentage of trabecular bone volume (BV/TV, 48%), trabecular number (Tb.N, 64%), trabecular thickness (Tb.Th, 76%), bone formation rate (BFR, 59%) and increased trabecular space (Tb.Sp, 171%). Loss of Rgs12 in OPCs in vitro significantly inhibited OB differentiation and the expression of OB marker genes, resulted in suppression of OB maturation and mineralization. Further mechanism study showed, deletion of Rgs12 in OPCs significantly inhibited GTPase activity and impaired Ca²⁺ oscillations via restraints of major Ca²⁺ entry sources (extracellular Ca²⁺ influx and intracellular ER Ca²⁺ release), partially contributed by the blockage of L-type Ca²⁺ channel mediated Ca²⁺ influx. Downstream mediator ERK was also inactive in Rgs12-deficient OPCs, while application of pertussis toxin (PTX) or overexpression of Rgs12 could partially rescue the defective OB differentiation and function via activation of ERK. Our findings reveal that Rgs12 is an important regulator in OB differentiation and function, and highlight Rgs12 as a potential therapeutic target for bone disorders.

Introduction

Bone homeostasis is intimately relied on the balance between bone-forming osteoblasts (OBs) and bone-degrading osteoclasts (OCs) ^{(1, 2)}. Any dysfunctional bone formation relative to bone resorption leads to defective skeletal integrity ^{(3, 4)}. In contrast to anti-resorptive treatment for bone diseases due to excessive bone resorption, bone disorders such as skeletal dysplasias or osteogenic bone tumors (too much or disorganized bone formation), and chronic kidney disease (CKD) or aging (reduced bone formation) are more challenging to treat due to the limited number of reliable drugs promoting osteogenesis and activity ^(5–8). Therefore, uncovering potential therapeutic targets that regulate OB differentiation and function during skeletal development and remodeling remains important objectives for skeletal restoration ⁽⁷⁾.

OBs are mononucleated cells derived from mesenchymal stem cell (MSC) lineage in the bone marrow and are uniquely responsible for bone formation ^{(2, 9)}. The speed and effectiveness of OB precursor cells (OPCs) differentiating into mature OBs determines the rate of bone formation and is dependent on the expression of OB-specific transcription factors like runt related transcription factor 2 (Runx2) and osterix (Osx), followed by the expression of alkaline phosphatase (ALP) and secretion of extracellular matrix containing type1 collagen (Col1α1) and non-collagenous proteins, leading to the deposition and mineralization of bone matrix ^{(2, 5, 10, 11)}. Calcium (Ca²⁺) oscillations play an indispensible role during these processes through controlling gene expression and affecting OB differentiation and proliferation via Ca²⁺ signalling ^{(9, 12–14)}. The excitability of Ca²⁺ signalling is determined by the contribution of extracellular Ca²⁺ influx, the effect of inositol 1, 4, 5-trisphosphate (IP3) to release Ca²⁺ from the endoplasmic reticulum (ER) and the capacity of removing Ca²⁺ from the cytosol by sarco/endoplasmic reticulum Ca²⁺ ATPase pump (SERCA) and plasma membrane Ca²⁺ ATPase (PMCA) ^{(13, 15, 16)}. Ca²⁺ channels such as voltage-sensitive Ca²⁺ channels (VSCCs) are involved in these Ca²⁺ signalling activities ^(16–19). Some studies have showed that elevated intracellular Ca²⁺ by Ca²⁺ influx through the L-type Ca²⁺ channel (LTCC), one of the VSCCs, will eventually lead to the activation of the extracellular signal-related protein kinase (ERK) signaling which is critical for OB differentiation and survival ^{(11, 19–21)}. Nevertheless, it remains largely unknown which factor(s) trigger and maintain Ca²⁺ oscillations in OBs and how Ca²⁺ signalling regulates OB differentiation via Ca²⁺ channels.

Regulators of G-protein signalling (Rgs) 12 (Rgs12) is the largest protein in the Rgs family with its multi-domain architecture ^{(9, 22)}. It possesses a PDZ (PSD-95/Dlg/ZO-1) domain and a phosphotyrosine-binding (PTB) domain that are capable of binding the C terminus of G protein–coupled receptors (GPCRs) and interacts with GPCR chemokine receptors respectively, a conserved Rgs domain that is responsible for the GTPase accelerating protein (GAP) activity, a pair of Ras-binding domains (RBDs) that allows Rgs12 to integrate both heterotrimeric and monomeric G proteins signalling, and a GoLoco motif that carries guanine nucleotide dissociation inhibitor (GDI) activity toward Gαi subunits ^{(9, 22, 23)}. This multi-domain structure provides Rgs12 with the potential to modulate multiple signalling pathways ^{(9, 24, 25)}, including the signaling mediated by Gi-coupled GPCR, which has been reported to lead to osteopenia phenotype through the transgenic expression of a constitutively active Gi-coupled GPCR in OB ⁽²⁶⁾. Our previous studies have revealed that Rgs12 plays an essential role in osteoclastogenesis through the regulation of PLCγ-Ca²⁺ channel-Ca²⁺ oscillations-nuclear factor of activated T cells 2 (NFAT2) pathway ^(27–29). Loss of Rgs12 impaired Ca²⁺ oscillations and reduced NFAT2 expression in OC, leading to the failure in OC differentiation and function ^{(27, 29)}. However, the role of Rgs12 in OB differentiation and function, and how Ca²⁺ oscillations and signaling regulate these processes is poorly understood.

In this study, we generated OB-specific knockout of Rgs12 in mice and found that Rgs12 influences cancellous bone mass, leading to an osteopenia phenotype, and targeted deletion of Rgs12 in OPCs impaired OB differentiation and function through the disruption of Ca²⁺ channel/ Gαi-Ca²⁺ oscillations-ERK activation. Further, the defective OB differentiation and function could partially be rescued by the application of pertussis toxin (PTX), a Gi specific inhibitor, or overexpression of Rgs12 in Rgs12 deficient OPCs. We concluded that Rgs12 is required for OB differentiation and function, and highlight Rgs12 as a potential therapeutic target for bone disorders.

Materials and Methods

Detail of all reagents and antibodies are listed in supplemental materials and methods

Generation of Rgs12 conditional knockout (CKO) mice

Methodologies for generation of homozygous Rgs12^fl/fl and Osx-Cre mice on C57Bl/6 backgrounds have been previously described ^{(27, 30)}. To specifically inactivate Rgs12 in the OB lineage, we crossed Rgs12^fl/fl mice with Osx-Cre transgenic mice to generate Osx; Rgs12^fl/+ progeny, which were used for subsequent mating to produce homozygous Osx; Rgs12^fl/fl mice. Because the phenotype in Rgs12^fl/fl mice were indistinguishable from that in Osx-Cre mice after the age of 12 weeks, Rgs12^fl/fl mice from the same litter of Osx; Rgs12^fl/fl were used as control ^(31–34). All mice were maintained in C57BL/6J background and housed maximally at 5 per cage under a standard 12-hour light/12-hour dark cycle condition and free access to water and rodent diet. Once reached 12 weeks old, qualified mice (average body shape and bodyweight) were randomized in each group according to their genotyping (see supplemental materials and methods). Both genders were used to include potential variation related to gender. All experiments using mice were performed following protocols and approved by the Institutional Animal Care and Use Committee (IACUC) of the State University of New York (Buffalo) and the University of Pennsylvania.

Primary OPCs culture, virus infection and differentiation

Primary OPCs were isolated from the calvarial bone of Rgs12^fl/fl mice at postnatal day 4 (4 pups of either gender) according to a serial digestion method as previously performed ⁽³⁰⁾. In brief, calvarial bone was dissected, fragmented and subjected to sequential digestions in collagenase type I and 0.25% trypsin. Cells from this digestion were plated in complete culture medium (α-MEM, 10% FBS, penicillin and streptomycin) at 37°C in a humidified 5% CO₂ atmosphere. Upon reaching 90% of confluence, cells were trypsinized, reseeded, and cultured for further experiments. OPCs from Rgs12^fl/fl mice were infected with adenovirus (Ad-Cre, Ad-Null or Ad-GFP). Ad-Cre infection yielded an approximate 85% deletion of Rgs12 in Rgs12^fl/fl OPCs, which were labeled as Rgs12^d/d. Ad-Null or Ad-GFP treated cells were used as control and marked as Rgs12^f/f and Ad-Null treated cells were specially used in Ca²⁺ imaging experiments. For osteogenic differentiation, OPCs were treated with osteogenic medium (OS medium), composed of complete culture medium with addition of β-glycerophosphate, ascorbic acid and dexamethasone. For experiments evaluating Gα signalling, PTX (100 ng/ml, Thermo Fisher Scientific) was also added to the OS medium ^{(35, 36)}.

MTS assay

OPCs after adenovirus infection were seeded at 5000 cells per well in 96-well plate and performed according to manufacturer’s instructions (CellTiter 96 AQueous kit, Promega, Madison, WI, USA).

Gene transfer

Plasmid pCMV-Flag-Rgs12 vector was constructed by inserting the full length mouse Rgs12 cDNA into the plasmid of p3xFlag-myc-CMV-26 (Sigma). For transfection of 5 x 10⁵ cells, 2 μg DNA mixed with 6 μl Fugene HD transfection reagent (Promega) was used. After adding serum-free medium for 4 h, cells were changed into 2% FBS medium and incubated overnight, and replenished with complete culture medium containing G418 (400 μg/ml, Gibco) for 2 weeks.

ALP activity assay

ALP activity was measured at day 7 of osteogenic differentiation as previously described ⁽³⁰⁾. Briefly, the cells were washed twice with PBS then harvested with harvest buffer (see supplemental materials and methods), and sonicated at low power, followed by centrifugation and assay buffer adding, then, incubated at 37ºC until color change. NaOH solution was added to stop the reaction, and recorded the time. The optical density of ALP was measured at 405nm in a 96-well plate using microplate reader.

Measurement of bone nodule formation

Extracellular matrix calcium deposits were captured by Von Kossa staining and Alizarin red staining at day 14, day 21 or day 28 of osteogenesis as described previously ⁽³⁰⁾. In brief, cells were fixed in 4% (w/v) paraformaldehyde (Sigma) and stained with 5% silver nitrate (Sigma) solution then exposed under ultraviolet (UV) light, or stained with 40 mM Alizarin red S (Sigma) solution (pH 4.4) at room temperature avoid light. Stained cells were scanned for image acquisition. To quantify Alizarin Red staining, 10% (w/v) cetylpyridinium chloride (Sigma) in 10 mM sodium phosphate (pH 7.0) was used to detain cells and measured using microplate reader (see supplemental materials and methods).

Quantitative real-time PCR (qPCR)

QPCR was carried out as described previously, with slight modifications ^{(30, 37)} (see supplemental materials and methods). In brief, total RNA was extracted and quantified. cDNA was synthesized from 2 μg of total RNA, and PCR amplifications were performed in triplicate with SYBR Green qPCR Master Mix (Bimake) on CFX96 real-time PCR machine (Bio-Rad) using the following steps: 95°C for 5 min, 40 cycles of 95°C for 15 s, and 60°C for 30 s. Gene expression were calculated using 2^−ΔΔCT method and normalized to housekeeping gene GAPDH. Primer sequences were listed in supplementary Table 1.

Western blot (WB)

WB was performed as described previously ^{(27, 30, 37)}. Briefly, cells and tissues were homogenized in RIPA buffer (see supplemental materials and methods) and protein concentrations were measured using BCA protein assay (Pierce). A total of 20 μg of protein was subjected to 10% SDS-PAGE gels, then transferred to a PVDF membrane (Millipore), followed by 30 min 5% milk blocking and overnight primary antibodies incubation at 4°C. After horseradish peroxidase–conjugated secondary antibodies incubation, visualization was performed using enhanced chemiluminescence (ECL) detection kit (Bio-Rad) on a ChemiDoc™ touch imaging system (Bio-Rad) and analyzed using the Image J software (National Institutes of Health, Bethesda, MD). Primary antibodies directed against Rgs12 (1:500, Abcam), Col1 (1:1000, Abcam), OCN (1:500, Abcam), phospho-p44/42 ERK (1:1000, CST), p44/42 ERK (1:1000, CST), β-actin (1:3000, Santa Cruz) and GAPDH (1:2000, Genscript) were used. GAPDH or β-actin was used as the internal control.

Bone Micro-CT analysis

A quantitative analysis of the femur morphology and microarchitecture was performed with the Micro-CT system (USDA Grand Forks Human Nutrition Research Center, Grand Forks, ND, USA). Femurs from 12 weeks old Rgs12^fl/fl and Osx; Rgs12^fl/fl mice of either gender (n=3 per gender, n=6 in total of each group) were fixed with 4% (w/v) paraformaldehyde for 24 h, rinsed with PBS, scanned and reconstituted as three-dimensional (3D) images. The trabecular and cortical bone architecture were assessed at the distal femoral metaphysis and midshaft. About 150 slices (1.5 mm) of bone were evaluated to determine percentage of bone volume (BV/TV, %), trabecular number (Tb.N, mm⁻¹), trabecular thickness (Tb.Th, mm) and trabecular spacing (Tb.Sp, mm) as previously descried ^{(27, 28, 30)}.

Bone histomorphometric analysis

To measure dynamic bone formation, mice were intraperitoneally injected with calcein (25 mg/kg) at postnatal day 90 and day 96. Mice were euthanized under CO₂ condition and harvested 2 days after last injection. Tibias were fixed in 4% (w/v) paraformaldehyde and longitudinal sections (8 μm) of proximal tibias were cut (see supplemental materials and methods). Bone formation rate per bone surface (BFR/BS, μm³/um⁻² per day), mineral apposition rate (MAR, μm per day), OB surface per bone surface (Ob.S/BS, %), OB number per bone perimeter (N.Ob/B.Pm, mm⁻¹) and OC number per bone perimeter (N.Oc/B.Pm, mm⁻¹) were measured.

Bone histological analysis

Tibia from 12 weeks old mice were fixed in 4% (w/v) paraformaldehyde for 24 h and subsequently decalcified at 4°C in 10% EDTA for 3-4 weeks. The samples were dehydrated in gradients of ethanol, followed by 100% xylene, and then embedded in paraffin. After deparaffinization, sectioned slides in 5 μm were stained with haemotoxylin and eosin (H&E) and imaged using a Leica inverted microscope (DMI6000B, Leica, Germany)

IF staining

IF staining was performed based on protocols described previously with slight modifications ^{(27, 30, 38)}. In brief, sectioned slides prepared above were permeabilized with 0.05% Triton X-100, blocked in 10% goat serum (Vector Laboratories, Burlingame, CA, USA) and incubated with anti-Rgs12 antibody (1:500, Abcam) at 4°C overnight. Primary antibody was coupled with Alexa Fluor 488-conjugated anti-Chicken secondary antibody (1:500, Abcam), followed by counterstained for nuclei using DAPI. Slides were mounted and images were acquired using the Leica fluorescent microscope (DMI6000B, Leica, Germany).

Ca²⁺ imaging

Measurements of Ca²⁺ oscillation were performed as previously described with slight modification ^{(27, 39)}. OPCs after Ad-Null or Ad-Cre infection were reseeded at a density of 2 × 10⁵ cells/ 35 mm-dish and induced with OS medium for 3 days. Then, cells were loaded with 5 μM fluo-4 AM (Thermo Fisher Scientific) for 50 min in complete culture medium containing 0.05% pluronic F127 (Sigma), washed and incubated in Hank’s Balanced Salt Solution (HBSS). Change of intracellular Ca²⁺ concentration was accessed by a Leica fluorescent microscope (DMI6000B, Leica, Germany) with excitation wavelength at 488 nm and emission wavelength at 505–530 nm. Signals were recorded simultaneously at 5-s intervals for 30 min, and fluorescence intensity was analyzed using Leica Application Suite X (Leica, Germany) software and plotted over time.

To measure intracellular Ca²⁺ concentration change during extracellular Ca²⁺ influx or release from ER in OPCs, Cells were washed and incubated a in Ca²⁺ free isotonic solution (ISO) (see supplemental materials and methods) after fluo-4 AM loading. Change of intracellular Ca²⁺ concentration was accessed and recorded for 45 min as mentioned above. Once the cells reached a stable baseline level of cytosolic Ca²⁺, the Ca²⁺ free ISO was exchanged for a Ca²⁺ free ISO containing 1μM TG (Sigma) to trigger Ca²⁺ release from the ER, or a normal ISO containing 1.3 mM CaCl₂ to induce extracellular Ca²⁺ influx, or a normal ISO containing 10μM Bay K8644 (Sigma) to induce L-type Ca²⁺ channel amplified influx ^{(19, 39)}. Changes of intracellular Ca²⁺ intensity was normalized by the initial Ca²⁺ level of cell itself.

GTPase assay

GTPase activity was performed according to manufacturer’s instructions (602-0120, Innova Biosciences, San Diego, CA, USA) ⁽⁴⁰⁾. Colorimetric measurements were read at the wavelength of 635 nm and GTPase activities were assessed on the basis of inorganic complex solutions.

Statistical Analysis

Animal number in each group was calculated to detect a 50% difference at alpha = 0.05 and power = 0.8 using Graphpad StatMate 2.0 software (GraphPad Software, San Diego, USA). All data are presented as mean ± standard deviation (SD) (n >3) and analyzed by using unpaired, two-tailed Student’s t-test for the comparison between two groups or one-way ANOVA followed by Sidak’s multiple comparison test for grouped samples. p<0.05 were considered to be significant. Graphpad Prism 6.0 software (GraphPad Software, San Diego, USA) was used to perform statistical analysis.

Results

Rgs12 was expressed in murine OBs and gradually increased throughout osteogenesis

To evaluate Rgs12 expression is different tissues, samples from brain, lung, kidney and bone were isolated from WT mice. WB showed Rgs12 protein highly expressed in brain and lung, with a lower expression level in bone (Fig. 1A). Further, to monitor Rgs12 expression pattern during osteogenesis, OPCs derived from calvarial bone of WT mice were stimulated with OS medium for 0, 3, 7 and 14 days. The mRNA level of Rgs12 was gradually increased and a significant increase occurred after day 7 of osteogenesis (Fig. 1B). The result was further confirmed at the protein level (Fig. 1C).

Fig. 1. Rgs12 is expressed in murine bone tissues and osteoblasts (OBs), with expression increases during osteogenesis.

(A) WB showed Rgs12 protein expression in brain, lung, kidney and bone tissue of WT mice. **p <0.01 versus (vs) brain, n=4. (B) qPCR results showed Rgs12 mRNA expressed in primary OB precursors (OPCs) derived from WT calvarial bone, and increased with osteogenic medium (OS medium) stimulation days. (C) qPCR results were confirmed at protein level. All quantitative data were normalized to GAPDH and presented as mean ± SD, **p <0.01, ***p <0.001 vs day 0, n=4.

Targeted deletion of Rgs12 in OPCs reduced cancellous bone mass with decreased OB number

To better understand the role of Rgs12 during osteogenesis in vivo, we generated Rgs12 CKO mice (Osx; Rgs12^fl/fl mice) described previously. Osx; Rgs12^fl/fl mice were born with expected Mendelian ratios. Deletion of Rgs12 was examined using genomic DNA from calvarial bone (Supplemental Fig. 1A). Low levels of Rgs12 mRNA (Supplemental Fig. 1B) and significantly diminished expression of Rgs12 protein were found in calvarial bone of Osx; Rgs12^fl/fl mice (Supplemental Fig. 1C). In addition, expression of Rgs12 was significantly reduced in femur sample of Osx; Rgs12^fl/fl mice (Supplemental Fig. 1D), indicating a successful generation of knockout model.

To assess whether deletion of Rgs12 affects bone physiology, femurs of 12 weeks old mice were evaluated by Micro-CT analysis. With a marked reduction in cancellous bone mass (Fig. 2A), the percentage of BV/TV, Tb.N, and Tb.Th in Osx; Rgs12^fl/fl mice were 48%, 64% and 76% of those in Rgs12^fl/fl mice, whereas Tb.Sp increased to 171% (Fig. 2B), indicating an osteopenia phenotype. H&E staining on tibias also exhibited a decreased trabecular bone in Osx; Rgs12^fl/fl mice (Fig. 2C). Dynamic histomorphometric analysis of 12 weeks old Osx; Rgs12^fl/fl mice further demonstrated a significant reduction in BFR/BS, MAR, Ob.S/BS and N.Ob/B.Pm (Fig. 2D, E), and no change was found in N.Oc/B.Pm between Osx; Rgs12^fl/fl mice and control littermates (Fig. 2E). These data suggested an osteopenia phenotype caused by impaired OB differentiation after Rgs12 deletion.

Fig. 2. Deletion of Rgs12 in OPCs (OSX; Rgs12^fl/fl) causes significant osteopenia with reduced OB formation.

(A) Micro-CT analysis of the femurs from 12 weeks old Rgs12^fl/fl and OSX; Rgs12^fl/fl mice (upper panel, axial view of the metaphyseal region; lower panel, longitudinal view). An apparent reduce in the cancellous bone mass was observed in OSX; Rgs12^fl/fl mice. (B) Quantitative analysis of the percentage of bone volume (BV/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th) and trabecular spacing (Tb.Sp) of samples shown in (A). *p <0.05, ***p <0.001 vs Rgs12^fl/fl mice, n=6. (C) H&E staining of the proximal tibia metaphyseal regions of 12 weeks old mice. Lower panel: higher magnification. Scale bars, 500μm (upper) or 50μm (lower). (D) Dynamic histomorphometry of tibia after double calcein labeling. (E) Quantitative analysis of bone formation rate per bone surface (BFR/BS), mineral apposition rate (MAR), OB surface per bone surface (Ob.S/BS), OB number per bone perimeter (N.Ob/B.Pm) and OC number per bone perimeter (N.Oc/B.Pm) from double calcein labeling. *p <0.05, ***p <0.001 vs Rgs12^fl/fl mice, n=6. Quantitative data were presented as mean ± SD.

Deletion of Rgs12 impairs OB differentiation

To further determine whether the decreased bone mass in Osx; Rgs12^fl/fl mice resulted from impaired OBs, we investigated the role of Rgs12 in OB differentiation and function in vitro. OPCs were isolated from Rgs12^fl/fl mice and infected with different adenovirus as previously described. QPCR result showed Ad-Cre infection yielded an approximate 85% reduction of Rgs12 in Rgs12^fl/fl OPCs, while Ad-GFP and Ad-Null infection had no reduced effects of Rgs12 in OPCs (Fig. 3A). Deletion of Rgs12 in OPCs apparently impaired OB differentiation, evidenced by lower expression level of OB differentiation markers including ALP, Col1α1, OCN and Runx2 after 7 days of osteogenic induction (Fig. 3B). These impairments were also confirmed by reduced expression of extracellular matrix proteins, Col1 and OCN (Fig. 3C). Blockage of OPCs differentiation and mineralization were further detected by exhibiting less ALP activity (Fig. 3D) and reduced bone nodule formation (Fig. 3D, E) in Rgs12^d/d OBs. All differentiation-related experiments were performed under cell confluency condition to eliminate the effect of proliferation on cell differentiation (Supplemental Fig. 2).

Fig. 3. Deletion of Rgs12 impairs OB differentiation.

(A) Rgs12 mRNA level in OPCs after different adenovirus infection, normalized to GAPDH (n=6, triplicates per group). ***p <0.001 vs Ad-GFP; ^###p <0.001 vs Ad-Null; NS, not statistical significant between groups. (B) qPCR analysis of OB marker genes ALP, collagen type I alpha 1 (Col1α1), osteocalcin (OCN) and Runx2 in OPCs after 7 days of OS medium culture. OPCs from Rgs12^fl/fl mice that infected with Ad-GFP were marked as Rgs12^f/f, while infected with Ad-Cre were labelled as Rgs12^d/d. ***p <0.001 vs control, n=4, triplicates per group. (C) WB analysis of extracellular matrix proteins expression at day 7 of osteogenic induction. **p <0.01, ***p <0.001 vs control, n=4. (D) ALP activity and Alizarin Red absorption of OPCs at indicated days of osteogenic induction. **p < 0.01, ***p < 0.001 vs Rgs12^f/f cells within groups; ^###p < 0.001 vs Rgs12^f/f cells at day 14 of osteogenic induction; n=4, triplicates per group. (E) Representative images of Von Kossa and Alizarin Red staining (triplicates per group). Quantitative data were presented as mean ± SD.

Deletion of Rgs12 inhibited GTPase activity and Ca²⁺ oscillations, blocking ERK activation in OPCs, and application of PTX partially rescued defective OB differentiation and function.

Studies show that signaling mediated by Gi-coupled GPCR served as a negative regulator of OB differentiation ^{(9, 26, 41)}, while Rgs12 acts as an antagonist of this signaling via enhancing GTPase activity towards Gαi subunit ^{(9, 22)}. Therefore, we compared the differences of GTPase activities between Rgs12^f/f and Rgs12^d/d OPCs and our result showed that Pi released in Rgs12^d/d OPCs was 53% of that in Rgs12^f/f OPCs (Fig. 4A), indicating a possible enhanced Gαi-mediated signaling in Rgs12^d/d OPCs. Accumulating studies show that Gαi-mediated signaling can crosstalk with Ca²⁺ signaling ^(42–44), which also plays a critical role during OB differentiation and function ^{(9, 22)}, and our previous study demonstrated that Rgs12 regulated Ca²⁺ oscillations during OC differentiation ⁽²⁷⁾. Thus, we hypothesized Rgs12 is also required for Ca²⁺ oscillations during OB differentiation. By monitoring intracellular Ca²⁺ change in OPCs after OS medium stimulation for 3 days, we found a sustained frequency of Ca²⁺ oscillations at approximately two minutes per interval in the Rgs12^f/f cells (Fig. 4B, left panel), nevertheless, in Rgs12^d/d cells, Ca²⁺ oscillations were barely detectable (Fig. 4B, right panel). To further confirm the involvement of Rgs12 in these two pathways, phosphorylated ERK (p-ERK), the downstream mediator of both pathways, was examined ^{(20–22, 45)}. Our results showed the expression level of p-ERK was significantly lower in Rgs12^d/d cells after OS medium stimulation (Fig. 4C).

Fig. 4. Deletion of Rgs12 inhibits GTPase activity, Ca²⁺ oscillations and sequentially blocks ERK activation; and PTX treatment partially rescued defective OB differentiation and function.

(A) GTPase assay showed inorganic phosphate (Pi) release level in OPCs. ***p <0.001 vs Rgs12^f/f cells; n=4. (B) Analysis of Ca²⁺ oscillations in OPCs after 3 days of OS medium stimulation. Cells were traced, calculated, and plotted over time using different colors (n=10 cells of each group). (C) WB showed phosphorylated ERK (p-ERK) expression in OPCs. Cells were starved for 6h prior to OS medium stimulation, p-ERK were normalized to total ERK. **p <0.01, ***p <0.001 vs 0 min; ^#p <0.05, ^###p < 0.001 vs same time point in Rgs12^f/f cells; n=4. (D) ALP activity of OPCs with or without PTX treatment at day 7 of osteogenic induction. (E) Alizarin Red absorption and staining of OPCs with or without PTX treatment at day 28 of osteogenic induction. *p <0.05, **p <0.01, ***p <0.001 vs Rgs12^f/f cells within groups; NS, not statistical significant or ^#p <0.05 vs same cells without PTX treatment; n=4. (F) WB analysis of p-ERK in OPCs with or without PTX treatment. Cells in PTX group were pretreated with PTX (100ng/ml) overnight in complete culture medium, followed by 6h starvation and then OS medium stimulation combine PTX for 15 min. NC and OS group were only stimulated with OS medium for 0 or 15 min. *p <0.05, **p <0.01, ***p <0.001 vs NC within groups; ^#p <0.05 vs same subgroup in Rgs12^f/f; ^&p <0.05, ^&&p <0.01 vs OS within groups; n=4. Quantitative data were presented as mean ± SD.

PTX was applied to purposely feedback the role of Rgs12 in Gαi-mediated signaling during osteogenesis. As expected, application of PTX displayed a partial rescue of ALP activity and bone nodule formation in Rgs12^d/d OBs (Fig. 4D, E). Further results showed p-ERK was significantly upregulated in PTX-treated Rgs12^d/d cells (Fig. 4F). These findings supported the regulator role of Rgs12 in OB differentiation and function through Gαi-mediated signaling.

Deletion of Rgs12 restrained the entry sources of cytosolic free Ca²⁺ in OPCs

Given that Ca²⁺ oscillations arise from periodic Ca²⁺ influx and repetitive release and reuptake of free Ca²⁺ from intracellular stores ^{(46, 47)}, we investigated whether Rgs12 deletion affects Ca²⁺ entry from these two main sources. First, we determined the influence of Rgs12 on Ca²⁺ influx. The fluorescence intensity increased gradually in Rgs12^f/f cells after normal ISO application, whereas elevating amplitude in Rgs12^d/d cells was dramatically decreased (Fig. 5A, B; Supplemental Video 1 and 2). Significant differences between Rgs12^f/f and Rgs12^d/d cells occurred after 3 min of ISO application, and reached maximum after 13:10 min (Fig. 5C), indicating a restraint on Ca²⁺ entry to cytosol from extracellular environment. Next, we determined whether deletion of Rgs12 also affects Ca²⁺ release from ER. To achieve that, cells were treated with 1μM TG to inhibit SERCA and unmask an endogenous leak for Ca^{2+ (39, 46)}. Similar to Ca²⁺ influx, elevating amplitude of ER Ca²⁺ release in Rgs12^d/d cells were dramatically decreased (Fig. 5D, E; Supplemental Video 3 and 4), and significant differences were captured after 15 s of TG application, and reached maximum level after 90 s (Fig. 5F). Collectively, these data suggested deletion of Rgs12 restrained both extracellular and intracellular entry sources of cytosolic free Ca²⁺, leading to inhibition on Ca²⁺ oscillations in OPCs.

Fig. 5. Deletion of Rgs12 restrains Ca²⁺ entry from extracellular environment and intracellular ER store to the cytosol.

(A), (B) and (C): Ca²⁺ influx in Rgs12^f/f or Rgs12^d/d OPCs. (A) Representative images of intracellular Ca²⁺ in OPCs before (0 min) and after isotonic solution (ISO, started from ISO 0 min) stimulation. OPCs were incubated in Ca²⁺ free ISO to reach a stable baseline level of cytosolic Ca²⁺, and then changed into normal ISO, which contains 1.3 mM CaCl₂ to induce extracellular Ca²⁺ influx. Upper panel, Rgs12^f/f OPCs; lower panel, Rgs12^d/d OPCs. (B) Overlay of traces from OPCs. Arrow indicated the time point when extracellular Ca²⁺ presented. Data was expressed as a fold change of fluo-4 intensity, accomplished by normalized each cell’s fluorescence intensity to its initial intensity (n=39 cells per group). (C) Quantitative analysis of fluo-4 intensity at indicated time point. *p <0.05, ***p <0.001 vs Rgs12^f/f cells, n=39 cells per group. (D), (E) and (F): ER Ca²⁺ release in Rgs12^f/f or Rgs12^d/d OPCs. (D) Representative images of intracellular Ca²⁺ in OPCs before (0 min) and after stimulation of 1μM thapsigargin (TG, started from TG 0 min). OPCs were incubated in Ca²⁺ free ISO as previous, and treated with TG in Ca²⁺ free ISO to trigger Ca²⁺ release form the ER. Upper panel, Rgs12^f/f OPCs; lower panel, Rgs12^d/d OPCs. (E) Overlay of traces from OPCs. Arrow indicated the time point of TG application (n=47 cells per group). (F) Quantitative analysis of fluo-4 intensity at indicated time point. *p <0.05, ***p <0.001 vs Rgs12^f/f cells, n=47 cells per group. Scale bars, 50μm. Quantitative data were presented as mean ± SD.

Deletion of Rgs12 attenuated the BayK8644-amplified Ca²⁺ influx

LTCCs, via Ca²⁺ influx, play key role in regulating intracellular Ca²⁺ homeostasis of OB ^{(19, 20)}. Thus, we use BayK8644, an agonist of LTCCs, to determine whether LTCCs partly contribute to the restrain of Ca²⁺ influx after Rgs12 deletion ⁽⁴⁸⁾. First, Rgs12^f/f OPCs were treated with or without 10μM BayK8644 in normal ISO, to ensure an amplified Ca²⁺ influx induced by BayK8644. Our results showed the fluorescence intensity of BayK8644-treated cells increased immediately after BayK8644 application (Fig. 6A, B; Supplemental Video 5 and 6) and reached significant differences after 30 s of application (Fig. 6C), demonstrating an amplified Ca²⁺ influx via LTCCs. Next, we compared the effect of BayK8644 on Rgs12^f/f and Rgs12^d/d cells, and no obvious increasing signals were detected in Rgs12^d/d cells after a limited and transient fluorescence peak, induced by BayK8644 application (Fig. 6D, E; Supplemental Video 7 and 8). Significant differences between Rgs12^f/f and Rgs12^d/d cells were even captured at logarithmic phase of fluorescence signals (Fig. 6F). These results confirmed the involvement of LTCCs in restraint of Ca²⁺ influx caused by Rgs12 deletion.

Fig. 6. Deletion of Rgs12 attenuates the BayK8644-amplified Ca²⁺ influx through L-type Ca²⁺ channel.

(A), (B) and (C): Effect of BayK8644 on intracellular Ca²⁺ in Rgs12^f/f OPCs. (A) Representative images of intracellular Ca²⁺ in Rgs12^f/f OPCs in the presence or absence of 10μM BayK8644 stimulation. OPCs were incubated in Ca²⁺ free ISO as previous described, and treated with BayK8644 in ISO (lower panel) or ISO alone (upper panel). (B) Overlay of traces from OPCs. Arrow indicated the time point when different solutions applied (S, started from S 0 min). Data was expressed as a fold change as previous described (n=28 cells per group). (C) Quantitative analysis of fluo-4 intensity at indicated time point. *p <0.05, ***p <0.001 vs cells without BayK8644, n=28 cells per group. (D), (E) and (F): Effect of BayK8644 on Rgs12^f/f or Rgs12^d/d OPCs. (D) Representative images of intracellular Ca²⁺ for OPCs before (0 min) and after stimulation of 10μM BayK8644 (BayK, started from BayK 0 min). OPCs were incubated in Ca²⁺ free ISO as previous, and treated with BayK8644 in ISO. Upper panel, Rgs12^f/f OPCs; lower panel, Rgs12^d/d OPCs. (E) Overlay of traces from OPCs. Arrow indicated the time point of BayK8644 application. Data was expressed as a fold change (n=28 cells per group). (F) Quantitative analysis of fluo-4 intensity at indicated time point. *p <0.05, ***p <0.001 vs Rgs12^f/f cells, n=28 cells per group. Scale bars, 50μm. Quantitative data were presented as mean ± SD.

Ectopic expression of Rgs12 partially rescued defective OB differentiation in Rgs12^d/d cells

To test potential therapeutic effect of Rgs12 on defective OB, we ectopically expressed Rgs12 in Rgs12^d/d OPCs. Successful Rgs12 overexpression was confirmed by WB (Fig. 7A). Overexpression of Rgs12 in Rgs12^fl/fl OPCs exhibited an enhanced effect on OB differentiation, while differentiation in Rgs12^d/d OBs was also partially rescued, with a significant improvement in both ALP activity and bone nodule formation (Fig. 7B, C), supporting the regulator role of Rgs12 during osteogenesis (Fig. 7D).

Fig. 7. Ectopic expression of Rgs12 partially rescues defective OB differentiation and function.

(A) WB showed Rgs12 protein expression in Rgs12^f/f and Rgs12^d/d OPCs transfected with or without pCMV-Flag-Rgs12 plasmid, normalized to GAPDH. (B) ALP activity of OPCs at day 7 of osteogenic induction. (C) Alizarin Red staining and absorption of OPCs at day 21 of osteogenic induction. *p <0.05, **p <0.01 vs Rgs12^f/f cells within groups; ^&p <0.05, ^&&p <0.01 or ^#p <0.05 vs same cells without Rgs12 overexpression; n=3, triplicates per group. Quantitative data were presented as mean ± SD. (D) Proposed role of Rgs12-regulated Gαi and Ca²⁺ signaling during OB differentiation.

Discussion

Rgs proteins are broadly involved in bone development and remodeling via their influences on OC and OB differentiation and function individually or mutually ^{(9, 22, 49)}. Our previous study demonstrated that Rgs12 promotes osteoclastogenesis during bone remodeling ⁽²⁷⁾. Here, we provide the first evidence to demonstrate a close dependency of osteogenesis on the expression of Rgs12, and to reveal the mechanism underlying how Rgs12, through both Ca²⁺ channel-Ca²⁺ oscillation-ERK signaling and traditional Gαi-ERK signaling, promotes OB differentiation and function during osteogenesis. The facilitating role of Rgs12 in bone formation was defined by obvious osteopenia phenotypes in Rgs12 conditional knockout mice (Osx; Rgs12^fl/fl) which specially inactivated Rgs12 on OB lineage (Fig. 2), and also supported by OPCs which underwent in vitro Rgs12 deletion, exhibited a striking inability to differentiate into mature and functional OBs (Fig. 3). Additionally, ectopic expression of Rgs12 in Rgs12 defective OPCs (Rgs12 ^d/d) demonstrated a partial restoration of differentiation ability and surprisingly showed enhanced differentiation ability when overexpressed in normal OPCs (Rgs12 ^f/f) (Fig. 7B, C). Therefore, our findings indicate that Rgs12 plays an indispensable role during osteogenesis.

We identify Gαi-ERK signaling as one of the pathway that Rgs12 exerts its regulatory role during osteogenesis. Rgs12 is a GTPase-activating protein that enhances the intrinsic GTPase activity of Gαi subunit, thereby inactivate downstream signal transduction ^{(9, 22, 23)}. It was previously noted that hyperactivation of the Gαi-mediated signalling served as a negative regulator of OB differentiation through inhibition of AC, and eventually lead to a reduction in trabecular bone formation ^{(26, 41)}, while the opposite, Gαs-AC-cAMP signalling is required for OB differentiation ^{(50, 51)}. In such a way, it would be rational to postulate that the deletion of Rgs12 shrinks GTPase activity towards Gαi subunit and subsequently eases the inhibitory effect on Gαi-mediated signalling, leading to an osteopenia phenotype. We found that Pi release reduced approximately 53% in Rgs12^d/d OPCs compare with control OPCs, indicating a dramatic shrink of GTPase activity towards Gαi subunit after Rgs12 deletion (Fig. 4A). Several studies showed that ERK activation serves as downstream effector of Gαi-mediated signalling via membrane recruitment of rap1GAPII and reduction of GTP-bound Rap1 ^{(41, 51–53)}, and is essential for OB differentiation and function ^{(11, 45, 54)}. Our results are in agreement with these findings and further showed a significantly lower expression of p-ERK in Rgs12^d/d cells induced by OS medium (Fig. 4C), suggesting an inactivation of ERK in OPCs after Rgs12 deletion. We also applied PTX on Rgs12^d/d cells to identify whether the effects followed by Rgs12 deletion could be reverse. Our results exhibit partial rescued effects on OB differentiation and maturation, and to some extent, were similar to the outcomes after ectopic expression of Rgs12 in defective OPCs (Fig. 4D, E, F), further confirming the involvement of Gαi-mediated ERK signaling pathway in Rgs12 regulated OB differentiation and function.

We also identify Ca²⁺ channel-Ca²⁺ oscillation-ERK signaling as another critical pathway that contributes to the regulatory role of Rgs12 during osteogenesis (Fig. 7D). Ca²⁺, as an intracellular second messenger, impacts nearly every aspect of biological processes through enormous versatility of Ca²⁺ signaling ^{(17, 55, 56)}. Components of Ca²⁺ signaling, including Ca²⁺ influx from extracellular environment through Ca²⁺ channel and Ca²⁺ release from intracellular Ca²⁺ store, regulate the frequency and amplitude of Ca²⁺ oscillations ^{(13, 22, 57)}, which is well known to play an critical role in promoting osteogenesis ^{(9, 14, 58)}. In our work, we found that Ca²⁺ oscillations were barely detectable in Rgs12^d/d OPCs (Fig. 4B), indicating an impairment on Ca²⁺ oscillations caused by Rgs12 deletion. We further determined whether this impairment on Ca²⁺ oscillations was attributed to the impact on major Ca²⁺ entry sources after Rgs12 deletion. Surprisingly, our results suggested that both extracellular Ca²⁺ influx and Ca²⁺ release from ER store were restrained after Rgs12 deletion in OPCs (Fig. 5). Several studies had conducted that these two major Ca²⁺ sources of Ca²⁺ oscillations played fundamental roles in regulating OB proliferation, differentiation and function individually ^{(20, 22, 46, 47)}. Therefore, our work establishes the role of Rgs12 as a key regulator that affects Ca²⁺ oscillations during osteogenesis. Moreover, Ca²⁺ enters cytosol through a variety of Ca²⁺ channels, among them, LTCCs have been listed as critical regulators of intracellular Ca²⁺ homeostasis in OB and are closely associated with the regulation of OB functions ^{(19, 47, 48)}. Our current data corroborates with those ideas and further suggested that LTCCs involve in the restraint of Ca²⁺ influx caused by Rgs12 deletion. Our results clearly showed steadily increased fluorescence signals in control OPCs after BayK8644 treatment, indicating a successful induction of amplified Ca²⁺ influx via LTCCs (Fig. 6A, B, C). Based on this, we further found that the long-lasting increasing signals of cytosolic free Ca²⁺ in control cells were unable to be detected in Rgs12^d/d cells, indicating a negative impact on LTCCs function after Rgs12 deletion (Fig. 6D, E, F). ERK also serves as downstream effectors of Ca²⁺ signaling, and the frequencies of Ca²⁺ oscillations was reported to influence Ca²⁺-mediated activation of Ras and signaling through the ERK/mitogen-activated protein kinase (MAPK) cascade ^{(21, 59, 60)}. The inactivation of ERK after Rgs12 deletion in OPCs that showed previously could also due to the decreased Ca²⁺ signaling originating from LTCCs (Fig. 4C, 7D).

A crosstalk between Gαi and Ca²⁺ signaling has long been reported ^(42–44), as both these pathways could share same terminal signals from cAMP signaling through Ca²⁺ sensor and Ca²⁺ oscillations. It was noted that Ca²⁺ sensor plays dual role on cAMP signaling, exerting through intracellular Ca²⁺ and PTX-sensitive Gαi pathways ⁽⁴²⁾. On the other hand, it is well known that Gαi subunits dampen the cAMP production via inhibiting AC, whereas Gαq subunits increase intracellular Ca²⁺ via phospholipase C (PLC) activation ^{(45, 61)}. Emerging evidence demonstrated that Gαi3 mutants with defective GTP binding and displayed lower GTPase activity, can block the Gαq-PLC-Ca²⁺ signaling pathway by forming an unproductive complex ^{(52, 62)}. In our work, we also found lower GTPase activity towards Gαi after Rgs12 deletion.

An interesting phenomenon from our present and previous studies is that Rgs12 deletion impaired Ca²⁺ oscillations in both OBs and OCs, and RANKL–Ca²⁺ oscillations–NFAT2 signaling pathway had been blocked in Rgs12 defective OCs ⁽²⁷⁾. A further study with more focus on bone remodeling using global Rgs12-knock out model is therefore suggested, and the deletion effects on Ca²⁺ channels and their subunits. Taken together, our results reveal a crucial role of Rgs12 in OB differentiation and bone formation via controlling Ca²⁺ channel/Gαi –Ca²⁺ oscillation-ERK activity, and highlight Rgs12 as a potential therapeutic target for skeletal abnormalities.