Assessing the osteoblast transcriptome in a model of enhanced bone formation due to constitutive G_s-G protein signaling in osteoblasts

Abstract

G protein–coupled receptor (GPCR) signaling in osteoblasts (OBs) is an important regulator of bone formation. We previously described a mouse model expressing Rs1, an engineered constitutively active G_s-coupled GPCR, under the control of the 2.3-kb Col I promoter. These mice showed a dramatic age-dependent increase in trabecular bone of femurs. Here, we further evaluated the effects of enhanced G_s signaling in OBs on intramembranous bone formation by examining calvariae of 1- and 9-week-old Col1(2.3)/Rs1 mice and characterized the in vivo gene expression specifically occurring in osteoblasts with activated G_s G protein–coupled receptor signaling, at the cellular level rather than in a whole bone. Rs1 calvariae displayed a dramatic increase in bone volume with partial loss of cortical structure. By immunohistochemistry, Osterix was detected in cells throughout the inter-trabecular space while Osteocalcin was expressed predominantly in cells along bone surfaces, suggesting the role of paracrine mediators secreted from OBs driven by 2.3-kb Col I promoter could influence early OB commitment, differentiation, and/or proliferation. Gene expression analysis of calvarial OBs revealed that genes affected by Rs1 signaling include those encoding proteins important for cell differentiation, cytokines and growth factors, angiogenesis, coagulation, and energy metabolism. The set of G_s-GPCRs and other GPCRs that may contribute to the observed skeletal phenotype and candidate paracrine mediators of the effect of G_s signaling in OBs were also determined. Our results identify novel detailed in vivo cellular changes of the anabolic response of the skeleton to G_s signaling in mature OBs.

Introduction

Osteoblasts (OBs) play a key role in control of skeletal homeostasis by influencing the initiation and extent of bone resorption and bone formation through complex mechanisms that are only partially understood. A number of factors contribute to this complexity. Fortunately, considerable progress has been made over the past few decades in defining the molecular events associated with the transition of committed osteoprogenitors to fully differentiated, post-proliferative OBs and osteocytes [1,2]. In particular, the roles of transcription factors, such as Runx2 and Osterix, in the progression of OB differentiation are well documented [3,4]. Much less clear are the roles of extrinsic factors, such as hormones, cytokines, and other elements of the skeletal microenvironment, in regulating OB commitment, differentiation, and function. Moreover, very little is known about how such extrinsic factors regulate OBs in different anatomic components.

One of the most important mechanisms for cellular response to extrinsic factors is through G protein signaling. G protein-coupled receptors (GPCRs) are highly relevant to OB differentiation and function, as evidenced by the anabolic skeletal response to parathyroid hormone (PTH) and prostaglandin E [5,6], two agents that act on GPCRs presented on OB lineage cells. Genetic studies in human and mice further support the key role played by G protein signaling [7–9]. We described a number of transgenic mouse models in which G protein coupled signaling has been manipulated in vivo in OBs that express the 2.3 kb-Col I promoter. These models include OB-specific expression of an engineered constitutively active G_s-coupled receptor, Rs1, and OB-specific expression of pertussis toxin (PTX) to block G_i signaling; both of which demonstrate an anabolic bone phenotype [10,11]. Mice expressing Rs1, showed an increase in bone accrual within the skull and in femur size, assessed by the whole body microCT analysis, with a dramatic age-dependent increase in trabecular bone with features resembling fibrous dysplasia. At 9 weeks of age, the male and female mutant mice showed dramatic increases (380%) in whole-body areal bone mineral density, as determined by dual-energy x-ray absorptiometry scanning. Histological assessment of femoral bones indicated that there was an increase in OB lineage cells, especially immature OBs, indicated by an expansion of cells expressing early OB markers, Runx2 and Osterix. Increased osteoclast activity was suggested by the large number of tartrate-resistant acid phosphatase (TRAP)-positive regions adjacent to the trabeculi within the lesions. Bone tissue quality; mineralization, composition, and maturity, of calvariae and femurs in Rs1 mice were also assessed by the complementary techniques of fourier transform infrared (FTIR) spectroscopy and synchrotron radiation micro-computed tomography (SRμCT) [12]. We demonstrated that mineral-to-matrix ratio and cross-link ratio were significantly lower in 6- and 15-week mutant bones. No differences in FTIR spectroscopic parameters were detected between the two anatomic sites despite the different bone-formation processes (endochondral vs. intramembranous). Tissue mineral density was also significantly lower in 3- and 9-week transgenic femoral diaphyses. The results indicate that continuous G_s activation in mature OBs lead to deposition of immature bone tissue with reduced mineralization.

The ability of Rs1 signaling in mature OBs to drive expansion of immature OBs could be mediated by paracrine factors, secreted from mature OBs that influence early OB commitment, differentiation, and/or proliferation. However, little is known about the in vivo cellular basis of the skeletal changes in response to enhanced G_s signaling in mature OBs.

In this study, we determined how G_s signaling in mature OBs affects bone formation by examining the Rs1 calvarial bone phenotype at 1 and 9 weeks of age and investigated the effect of enhanced OB G_s signaling on the OB transcriptome by examining alterations in gene expression in vivo in OBs from calvariae of 1-week-old Rs1 mice, compared to controls. The functionally related, differentially expressed genes, the G_s- and G_i-GPCRs, and candidate paracrine mediators that may contribute to the observed skeletal phenotype of the effect of G_s signaling in OBs were determined.

Methods

Transgenic mice

To examine the influence of G_s signaling in OBs, we used a mutated version of the G_s-coupled 5HT-4 serotonin receptor (Rs1) that has constitutive G_s signaling activity [13], together with the tetracycline-regulated system (Tet-off) to regulate the expression of Rs1 in vivo. The Col1(2.3)-tTA/TetO-Rs1 double transgenic mice (abbreviated Col1(2.3)/Rs1) were generated by crossing mice carrying the heterozygous TetO-Rs1 transgene with mice carrying the homozygous Col1(2.3)-tTA transgene. Endochondral bones of these mutant mice were well characterized as described [10,14,15].

In this study, we utilized green fluorescent protein (GFP)-based reporters to identify Rs1 expressing cells for investigating OB transcriptome analysis. We co-expressed a histone-GFP marker in OBs in vivo alone (abbreviated Col1(2.3)/GFP) or with Rs1 in triple transgenic mice (abbreviated Col1(2.3)/GFP/Rs1). This was accomplished by generating mice harboring a TetO-histone-GFP gene (mouse line Tg (TetO-HIST1H2BJ/GFP) 47Efu/J, Jackson Laboratory) with a TetO-Rs1 gene. Double transgenic mice were then crossed with mice homozygous for the 2.3 Col1-tTA genes to examine the skeletal effects of G_s signaling in mature OBs. Both transgenes driven by TetO were expressed in OBs only when the 2.3 Col1-tTA transgene was present. In this approach, 25% of the progeny were controls (express only histone-GFP) and 25% were experimental (express both histone-GFP and Rs1). All animals were maintained in the FVB/N background. Mice were on regular chow (LabDiet 5053; PMI Nutrition, St. Louise, MO) without doxycycline administration to allow transgene expression in experimental mice since conception. All protocols were approved by the Institutional Animal Care and Use Committee of the San Francisco Veterans Affairs Medical Center.

Calvarial bone histology

Calvariae collected for hematoxylin/eosin (H&E) staining and immunohistochemistry were fixed in 10% neutral buffered formalin (NBF; Fisher Scientific, Pittsburgh, PA) for 24 hours and stored at 4°C in 70% ethanol. Bones were decalcified in 10% EDTA at 4°C for 2–3 days before paraffin embedding, sectioning, and H&E staining. Assessment of calvarial bone was performed on 10-μm-thick sections at the midpoint of the parietal part.

Bone immunohistochemistry

Osteocalcin and Osterix were detected on formalin-fixed, paraffin-embedded sections (10 μm thickness) with primary antibodies from Takara Bio (Cat#M173) and AbCam (Cat#ab22552) as described [10]. Since Rs1 bears an N-terminal FLAG tag, the detection of Rs1 expressing OBs was done on alcohol-fixed, paraffin-embedded bone sections with a M1 FLAG primary antibody (F3040; Sigma-Aldrich) in M.O.M. diluents (dilute protein concentrate stock solution 1:13.5 in Tris-buffered saline). Detection of CD31, the endothelial cell marker, was performed using an Anti-Ig HRP detection kit (BD Pharmingen) following the manufacturer’s protocol.

Static and dynamic histomorphometry

Calvariae from 1- and 9-week-old mice were isolated at the time of euthanasia and fixed in 10% NBF for 1–2 days before embedding in methyl methacrylate and sectioning with Jung 2065 and 2165 microtomes (Leica, Bannockburn, IL). Sectioned bones were processed for Von Kossa staining as described [16]. For dynamic histomorphometry, 9-week-old mice were injected with 20 mg/kg of calcein (Sigma-Aldrich) 21 and 7 days before euthanasia and with 15 mg/kg of demeclocycline (Sigma-Aldrich) 2 days before euthanasia. Mice were euthanized at 9 weeks, and calvariae were isolated, fixed in 10% NBF, and stored in 70% ethanol. The undecalcified bone samples were embedded in methyl methacrylate. Histological assessment of calvarial bone was performed on 10-μm-thick sections at the midpoint of the parietal part (Fig. 1A). Mosaic-tiled images were acquired at ×20 magnification with a Zeiss Axioplan Imager M1 microscope (Carl Zeiss MicroImaging, Thornwood, NY) fitted with a motorized stage. The tiled images were stitched and converted to a single image with the Axiovision software (Carl Zeiss MicroImaging) before histomorphometry analyses.

Figure 1.

(A) Gross calvarial specimens from 1- and 9-week-old mice. White box represents the region of interest for histological assessment of calvariae. (B) H&E and VK images of 1-and 9-week-old Col1(2.3)/GFP/Rs1 mice compared to controls, Col1(2.3)/GFP mice. Mutant calvariae had a dramatic increase in trabecular bone volume with a distorted cortical structure. (C) Fluorescence imaging of calvariae showed a disorganized mineralization in the 9-week-old mutant calvariae. Red arrow represents the first (calcein; green) and second labels (demeclocycline; orange). (Scale bar, 50μm) Ca, central area; Pa, peripheral area; tb, trabeculi; bm, bone marrow

Separation of cell populations

Calvariae with OB-specific expression of histone-GFP either alone or together with Rs1 under control of the 2.3kb-Col I promoter were isolated from 1-week-old transgenic mice, pooled and divided into three groups according to their genotype (Col1(2.3) and Col1(2.3)/Rs1; Col1(2.3)/GFP; or Col1(2.3)/GFP/Rs1). After removal of the calvarial sutures, calvarial tissues were washed once with PBS and then subjected to four sequential, 30-minute digestions in an enzymatic mixture (1.5 U/ml collagenase P (Roche), 0.05% trypsin and 0.25 mM EDTA in PBS) at 37°C on a rocking platform (0.5–1 ml of mixture per calvaria). Cell fractions 2–4 were collected, pooled and resuspended in Dulbecco’s modified Eagle’s medium (DMEM; Life Technology) containing 10% fetal calf serum (FCS) and centrifuged. Then, cells were resuspended in 0.5–1 ml of 2% FCS in PBS and filtered through a 70-μm cell strainer. Individual cell populations were sorted based on GFP expression and fluorescent activated cell sorting (FACS) (BD FACSAria; BD Biosciences) with a 100-μm nozzle at the Cell Sorting Core at the San Francisco VA Medical Center. Cells from non-GFP-expressing mice (Col1(2.3) plus Col1(2.3)/Rs1) were used as controls to preset the sorting gate. Sorted cells were collected into 30% FCS in DMEM. A fraction of the sorted cell population (5×10³) was used for FACS reanalysis by the same gating strategy in the sorted procedure to assure the maximum purity. Cell suspensions were kept cold during the entire sorting process to minimize changes in gene expression.

RNA isolation and microarray analysis

Total RNA from FACS-sorted cell populations (GFP-expressing cells from Col1(2.3)/GFP and Col1(2.3)/GFP/Rs1 group) were isolated immediately after sorting and purified using the Arcturus PicoPure RNA isolation kit (Applied Biosystems, Carlsbad, CA), followed by DNase treatment with the RNase-Free DNase Set (Qiagen, Valencia, CA), according to manufacturer’s instructions. The quantity and quality of total RNA were assessed by NanoDrop ND-1000 Spectrophotometer and Agilent 2100 Bioanalyzer (Agilent Technologies). The 28S/18S ratios of the RNA were in the range of 1.8–2.1, and the RNA Integrity Numbers were in the range of 8.8–10. Reverse transcription and amplification of isolated RNA into cDNA were performed using the NuGEN FFPE WTA kit (NuGEN, San Carlos, CA). The integrity of resultant cDNA was assessed using the Agilent 2100 Bioanalyzer and individual samples were further processed and hybridized to Affymetrix Mouse Gene 1.0 ST arrays (Affymetrix, Cleveland, OH) before scanning, according to the protocol in WT Sense Target Labeling Assay Manual from Affymetrix (Version 4; FS450_0007) at the UCSF Gladstone Genomics Core Facility. Expression analysis was performed on three separate OB preparations of RNA from each genotype. Data were normalized using Guanine Cytosine Robust Multi-Array Analysis (GC-RMA). Bayesian statistical analysis was carried out using Linear Models of Microarrays (LIMMA) [17] to identify statistically significant differentially expressed genes between Col1(2.3)/GFP and Col1(2.3)/GFP/Rs1. Moderated t-statistics and the associated p-values were calculated and p-values less than 0.01 were considered to be statistically significant. Comparison groups were annotated with statistically significant Gene Ontology (GO) term overrepresentation using the GO-Elite software packages [18]. Microarray data have been submitted to the Gene Expression Omnibus (NCBI, NIH) database (series accession number GSE57104; http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?token=idwzccomrjellqz&acc=GSE57104).

Quantitative real-time PCR

For gene expression analysis by quantitative real-time PCR (qPCR), we compared selected genes that were differentially expressed in Col1(2.3)/GFP/Rs1 to Col1(2.3)/GFP by microarray analysis. Gene expression was quantified by SYBR Green I-based qPCR utilizing SYBR Green PCR Master Mix (part#4309155, Applied Biosystems). Primers were synthesized by Elim Biopharmaceuticals, Inc (Supplemental Table 1). qPCR was carried out in ABI Prism 7300 real-time thermocycler (Applied Biosystems). The results were analyzed using the Sequence Detection System software (SDS) supplied with the thermocycler. All reactions were performed in triplicates from the different experiments, and the expression of target genes was displayed normalized to GAPDH. Expressed GPCRs were identified using the Taqman® real time Mouse GPCR Array (part#4378703, Applied Biosystems), according to the manufacturer’s instructions. Array plates were run on the ABI Prism 7900HT system, and the data were analyzed using the SDS 2.3 and RQ Manager 1.2 software provided by Applied Biosystems. Quantitative analysis of GPCR expression, including housekeeping gene (18S), was expressed as cycle threshold (Ct) values. Average Ct values <30 on a TaqMan Array considered positive reactions reflecting detectable cDNA target copies in the sample. So we selected a cutoff of Ct= 30. Data shown were from the average of three independent samples.

Bone marrow stromal cells (BMSCs) cultures

Primary mouse BMSCs were isolated from the femurs and tibiae of 10–12-week old WT mice. The cells were collected in primary culture medium (PCM) consisting of ∞-Modification of Eagle’s medium, 10% fetal bovine serum, 100U/ml penicillin, 100μg/ml streptomycin, and 0.25μg/ml Fungizone (Gibco Cell Culture), and plated onto 10-cm cell culture dishes at a density of 7×10⁶ cells/dish. Cells were incubated at 37°C with 5% CO2 and maintained undisturbed for five days to allow for cell attachment. After that, PCM was removed along with all non-adherent cells and replaced with secondary osteogenic differentiation medium (SDM) (PCM containing 50μg/ml ascorbic acid and 3mM β-glycerolphosphate) to initiate OB differentiation. Thereafter, SDM was replaced every two or three days. Human PTH [1–34] (pPTH; #H-4835, Bachem) and forskolin (#F6886, Sigma) were prepared by dissolving in acetic acid (4.07μM final concentration) and ethanol (0.82% final concentration), respectively. Cells were exposed to 10⁻⁷M hPTH or forskolin at final concentration of 0.05 mM at day 28 for 24 hours. Total RNA from BMSCs cultures was isolated using PureLink Micro-to-Midi total RNA purification system (Cat#12183-018, Ambion) and further purified using RNeasy Mini Kit (Cat#74104, Qiagen). Reverse transcription was carried out using TaqMan Reverse Transcription Reagents (Part#N808-0234, Applied Biosystems). Fibroblast growth factor 9 (Fgf9) gene expression was quantified as described in qPCR section. The results of the experiments were confirmed by repeating the experiment three times.

Mineralization

BMSCs cultures were performed in the absence or presence of murine Fgf9 (SRP4057, Sigma-Aldrich). BMSCs were isolated from WT mice, collected in PCM and plated onto 10-cm cell culture dishes at a density of 7×10⁶ cells/dish. After incubating BMSCs in PCM for five days, the medium was removed along with all non-adherent cells and SCM was replaced and changed every two or three days. 5 ng/ml murine Fgf9 was added to the culture system from day 0 to day 20. To assess mineralization, two percent silver nitrate solution (Sigma-Aldrich) was added to cell culture dishes on day 20 for Von Kossa (VK) staining and UV-crosslinked for 10 minutes. Stained cultures were scanned and quantified using Improvision Openlab software version 5.0.2. The results of the experiment were confirmed by repeating the experiment three times.

Statistical analysis

All qPCR data were evaluated using a two-tailed Student’s t tests assuming equal variance. Data were presented as the mean ± SD. Statistical significance was taken as p<0.05.

Results

Bone histomorphometry

Mice expressing Rs1 displayed notable asymmetric thickening of the calvariae starting at one week of age, and the observed phenotypes became more severe with age (Fig. 1A). By histological assessment, Rs1 calvariae had a dramatic increase in trabecular bone volume with a distorted cortical structure. High-magnification images of 9-week-old Rs1 calvariae revealed a large number of cells with stromal-type morphology between trabeculae with many appearing stacked on and near the trabecular surfaces. Bone marrow elements appeared to be scattered in small areas between the trabecular bones (Fig. 1B). Bone formation rates and patterns in 9-week-old samples were examined by dynamic histomorphometry. However, we could not quantify bone formation rate in the mutant mice due to the disorganized nature of bone formation, which displayed a diffuse labeling pattern consistent with the rapid bone formation (Fig. 1C).

Bone immunohistochemistry

Osterix expression was readily detected by immunohistochemistry in cells throughout the intertrabecular space in Rs1 calvariae, but not in controls (Fig. 2A). OBs that attached to cortical bone surfaces in control calvariae might be fully mature and express only low levels of Osterix. Osteocalcin was expressed predominantly in cells along bone surfaces in mutant and control mice (Fig. 2B). Immunohistochemistry with an antibody against the FLAG tag on Rs1 identified a significant number of cells within the Rs1 calvariae located along bone surfaces at the same location as Osteocalcin-expressing cells (Fig. 2C). They represented maturing OBs, based on the known expression patterns of the 2.3-kb Col I promoter fragment [19] and on the abundant Osteocalcin in the bone lesions. Endothelial cells, identified by CD31 expression, were broadly distributed within the expanded calvarial tissue of Col1(2.3)/Rs1 mice (Fig. 2D). The immunohistochemistry results indicated that many of the uniform cells in the mutant bone lesions are in the OB lineage with an increase in normal-looking blood vessel structures.

Figure 2.

High-magnification immunohistochemistry of decalcified Col1(2.3)/GFP/Rs1 (central area) and control (whole area) calvariae from 9-week-old mice. (A) Osterix immunohistochemistry on formalin-fixed samples, demonstrating significant nuclear staining (brown; red arrowheads) by cells throughout the Col1(2.3)/GFP/Rs1 bony lesion mostly between the trabeculi. (B) Osteocalcin immunohistochemistry on formalin-fixed samples, demonstrating that cytoplasmic (brown; blue arrowhead) and extracellular (brown; red arrowhead) osteocalcin are present in the bone lesions, confirming the presence of mature osteoblasts. (C) FLAG immunohistochemistry on ethanol-fixed samples showed cytoplasmic staining of Rs1 expression only in Col1(2.3)/GFP/Rs1 calvarial sections (brown; red arrowhead). (D) CD31 immunohistochemistry on formalin-fixed samples, demonstrating diffuse positivity of blood vessels and sinusoids in the mutant clavaria (brown; red arrowheads). (Scale bar 100μm) tb, tabeculi

Isolation of GFP-expressing cell population

Cell populations obtained by enzymatic digestion of calvariae from 1-week-old double and triple transgenic GFP reporter (Col1(2.3)/GFP/Rs1 mice vs. Col1(2.3)/GFP mice) were FACS sorted for GFP expression to separate two cell populations (GFP-positive and GFP-negative cells) in each genotype. GFP-negative cells, representing various non-OB cell lineages mixed with osteoprogenitor and immature OBs, were clearly separated from the GFP-positive cell populations. Approximately 10% of the isolated cells are typically GFP positive. Sorted populations were 97–99% pure by FACS reanalysis (Fig. 3A).

Figure 3.

(A) FACS analysis of GFP expression of cells obtained by enzymatic digestion of neonatal calvariae isolated from transgenic mice. Cell populations were reanalyzed for purity. (B) qPCR analysis of osteoblast and osteoclast marker genes with RNA isolated from GFP negative population and GFP positive population from 3 Rs1 expression OBs and 3 controls after FACS sorting. Data were normalized to GAPDH RNA expression in the same sample. *P<0.01 (C) qPCR analysis with RNA isolated from GFP positive population from 3 Rs1 expression OBs and 3 controls after FACS sorting. Data were normalized to GAPDH RNA expression in the same sample. Rs1 expression was detected only in Rs1 group.

Identifying GPCRs expressed on OBs and genes differentially expressed in Rs1 mice

We compared the gene expression profiles in mature OBs from transgenic mice and controls from three biological replicates of each genotype. The identifiable cell populations were processed through RNA extraction, in vitro transcription, cDNA labeling and array hybridization. The Affymetrix Mouse Gene 1.0 ST Arrays were used to determine differences in gene expression. Each array contains a total of 26,166 coding transcripts. Significantly regulated genes were clustered using the GO Tree Machine, and functional analysis was done with GO-Elite software. We identified 358 differentially expressed transcripts if p ≤ 0.01 was set (moderate t-statistics). Small magnitudes of differential expression were detected, ranging from 0.26-fold to 4.63-fold changes in expression. Applying a threshold of a greater than 1.5-fold change in expression level revealed a filtered list of 133 genes, 114 of which were associated with functional annotations in the GO-Elite software (52 upregulated genes and 62 downregulated genes, Table 1).

Table 1.

Genes differentially expressed in Rs1 expressing OBs compared to controls

Gene symbol	Gene name	Mean ratio (Rs1/control)	p-value
Growth factor
Figf	c-fos induced growth factor	2.08	1.6×10−4
Fgf9	fibroblast growth factor 9	0.63	4.0×10−5
Transcription factor and regulator
Dennd4a	DENN/MADD domain containing 4A	1.99	1.1×10−3
Ldb2	LIM domain binding 2	1.79	6.6×10−3
Hmgb3	high mobility group box 3	1.72	8.3×10−3
Mlx	MAX-like protein X	1.61	9.0 ×10−5
Asxl2	additional sex combs like 2 (Drosophila)	0.65	4.1×10−3
Supt7l	suppressor of Ty 7 (S. cerevisiae)-like	0.63	1.6×10−3
Tceal5	transcription elongation factor A (SII)-like 5	0.62	9.2×10−3
Ddx1	DEAD (Asp-Glu-Ala-Asp) box polypeptide 1	0.58	4.0×10−5
Nfxl1	nuclear transcription factor, X-box binding-like 1	0.56	9.6×10−4
Zbtb20	zinc finger and BTB domain containing 20	0.53	4.1×10−3
Samd9l	sterile alpha motif domain containing 9-like	0.44	5.1×10−3
Bex1	brain expressed gene 1	0.37	8.5×10−4
Cell cycle regulator
Fbxo5	F-box protein 5	1.97	8.4×10−3
Tubb3	tubulin, beta 3	1.93	1.3×10−3
Cdc2a	cell division cycle 2 homolog A (S. pombe)	1.84	1.0×10−3
Lsm10	U7 snRNP-specific Sm-like protein LSM10	1.69	5.0×10−3
Bub1b	budding uninhibited by benzimidazoles 1 homolog, beta (S. cerevisiae)	1.65	6.8×10−3
Casc5	cancer susceptibility candidate 5	1.57	7.4×10−3
Prc1	protein regulator of cytokinesis 1	1.56	3.2×10−3
Anln	anillin, actin binding protein	1.55	5.1×10−3
Elovl1	elongation of very long chain fatty acids (FEN1/Elo2, SUR4/Elo3, yeast)-like 1	1.54	3.5×10−3
Ndc80	NDC80 homolog, kinetochore complex component (S. cerevisiae)	1.53	7.5×10−3
Cdkn2b	cyclin-dependent kinase inhibitor 2B (p15, inhibits CDK4)	1.53	3.5×10−3
Cdca2	cell division cycle associated 2	1.51	1.3×10−3
Cspp1	centrosome and spindle pole associated protein 1	0.26	3.0×10−3
Regulation of apoptosis
Bard1	BRCA1 associated RING domain 1	1.78	5.3×10−3
Trp53inp1	transformation related protein 53 inducible nuclear protein 1	0.61
Casp4	caspase 4, apoptosis-related cysteine peptidase	0.48	6.8×10−3
Trim13	tripartite motif-containing 13	0.44	9.9×10−3
Clec2d	C-type lectin domain family 2, member	0.43	7.8×10−3
Casp1	caspase 1	0.26	1.6×10−4
Enzyme activity
Car8	carbonic anhydrase 8	1.85	8.2×10−3
Slpi	Secretory leukocyte peptidase inhibitor	1.81	4.4×10−3
Adat2	adenosine deaminase, tRNA-specific 2, TAD2 homolog (S. cerevisiae)	1.64	1.4×10−3
Xdh	xanthine dehydrogenase	1.62	9.3×10−3
Sae1	SUMO1 activating enzyme subunit 1	1.55	1.1×10−3
Got1	glutamate oxaloacetate transaminase 1, soluble	1.52	5.0×10−5
Ctdspl2	CTD (carboxy-terminal domain, RNA polymerase II, polypeptide A) small phosphatase like 2	0.66	3.8×10−3
Dub2a	deubiquitinating enzyme 2a	0.64	4.4×10−3
Car2	carbonic anhydrase 2	0.62	8.7×10−3
Gstm1	glutathione S-transferase, mu 1	0.60	8.0×10−3
Ppp1r12b	protein phosphatase 1, regulatory (inhibitor) subunit 12B	0.48	5.9×10−3
Lipid and carbohydrate metabolic process
Lipi	Lipase, member I	4.63	9.0×10−5
Idi1	isopentenyl-diphosphate delta isomerase	2.21	1.6×10−3
Fabp3	fatty acid binding protein 3, muscle and heart	1.69	2.2×10−4
Me1	malic enzyme 1, NADP(+)-dependent, cytosolic	1.60	7.7×10−3
Gpr116	G protein-coupled receptor 116	0.49	8.3×10−3
Transporter
Slc35a1	solute carrier family 35 (CMP-sialic acid transporter), member 1	1.83	1.8×10−3
Efha1	EF hand domain family A1	1.67	1.5×10−3
Arfgef2	ADP-ribosylation factor guanine nucleotide	0.66	1.4×10−3
Trappc2l	trafficking protein particle complex 2-like	0.63	2.3×10−3
Mfsd7b	major facilitator superfamily domain containing 7B	0.53	2.2×10−3
Atp6v1b2	ATPase, H+ transporting, lysosomal V1 subunit B2	0.53	1.0×10−3
Atp6v0d2	ATPase, H+ transporting, lysosomal V0 subunit D2	0.26	6.7×10−3
Kinase activity
Prkrip1	Prkr interacting protein 1 (IL11 inducible)	1.66	3.6×10−3
Lmnb1	lamin B1	1.54	1.8×10−4
Fgd4	FYVE, RhoGEF and PH domain containing 4	0.62	4.2×10−3
Adrbk2	adrenergic receptor kinase, beta 2	0.61	3.5×10−3
Pip4k2a	phosphatidylinositol-5-phosphate 4-kinase, type II, alpha	0.61	2.4×10−3
Ptpn3	protein tyrosine phosphatase, non-receptor type 3	0.44	4.2×10−3
Protein binding
Mrpl17	mitochondrial ribosomal protein L17	2.27	5.5×10−3
Iqcb1	IQ calmodulin-binding motif containing1	0.62	2.9×10−3
Ppfia2	protein tyrosine phosphatase, receptor type, f polypeptide (PTPRF), interacting protein (liprin), alpha 2	0.59	3.2×10−3
Cytoskeleton
Dync2li1	dynein cytoplasmic 2 light intermediate chain 1	1.71	1.0×10−3
Plekhh2	pleckstrin homology domain containing, family H (with MyTH4 domain) member2	0.64	3.0×10−3
Syne2	synaptic nuclear envelope 2	0.30	5.0×10−5
Solute carrier
Slc38a4	solute carrier family 38, member 4	1.91	3.3×10−3
Slc25a19	solute carrier family 25 (mitochondrial thiamine pyrophosphate carrier), member 19	0.58	5.7×10−3
Slc5a3	solute carrier family 5 (inositol transporters), member 3	0.50	2.4×10−4
Immune system
Cd1d1	CD1d1 antigen	1.89	4.2×10−3
Srgn	Serglycin	0.48	3.9×10−3
Organogenesis
Rbp4	Retinol binding protein 4, plasma	1.69	4.7×10−4
Lmbr1	limb region 1	0.53	7.9×10−3
Aard	alanine and arginine rich domain containing protein	0.51	4.0×10−5
Nucleoside/nucleotide metabolic process
Mtap	methylthioadenosine phosphorylase	1.53	3.0×10−3
Gpatch8	G patch domain containing 8	0.62	2.7×10−3
Rbm47	RNA binding motif protein 47	0.60	6.9×10−3
Gimap6	GTPase, IMAP family member 6	0.57	8.4×10−3
Rab30	RAB30, member RAS oncogene family	0.57	3.2×10−3
Gtpbp10	GTP-binding protein 10 (putative)	0.45	3.1×10−3
Membrane protein
Tmem43	transmembrane protein 43	1.62	6.8×10−3
Tm4sf1	transmembrane 4 superfamily member 1	1.61	2.0×10−5
Neto2	neuropilin (NRP) and tolloid (TLL)-like 2	0.66	9.5 ×10−4
Tm4sf19	transmembrane 4 L six family member19	0.64	6.1×10−3
Tspan13	tetraspanin 13	0.63	9.4×10−3
Fam171b	family with sequence similarity 171, member B	0.59	2.8×10−3
Tspan7	tetraspanin 7	0.53	7.6×10−3
Coagulation cascades
Thbd	Thrombomodulin	0.57	1.7×10−4
Erythrocyte homeostasis
Fech	Ferrochelatase	0.61	6.1×10−3
Ossification
Ostn	Osteocrin	0.31	3.7×10−3
Signal transduction
Rras2	related RAS viral (r-ras) oncogene homolog 2	2.05	8.1×10−4
Srgap3	SLIT-ROBO Rho GTPase activating protein 3	0.65	4.8×10−4
Angiogenesis
Mfge8	Milk fat globule-EGF factor 8 protein	1.50	4.2×10−3
Cell adhesion molecule
Pcdhb17	protocadherin beta 17	1.68	3.8×10−3
Ion channel
Kcnj15	potassium inwardly-rectifying channel, subfamily J, member 15	1.98	7.7×10−4
Kcnk1	potassium channel, subfamily K, member 1	1.53	2.2×10−4
Wdfy2	WD repeat and FYVE domain containing2	0.65	5.2×10−3
Zfp788	zinc finger protein 788	0.63	5.2×10−3
Phf20l1	PHD finger protein 20-like 1	0.63	1.9×10−3
Zfp763	zinc finger protein 763	0.55	1.5×10−3
Matrix protein
Ecm1	extracellular matrix protein 1	1.61	1.9×10−3
Chad	Chondroadherin	0.65	9.0×10−3
Cytokine secretion
Irf1	interferon regulatory factor 1	0.60	3.2×10−3
Lcp2	lymphocyte cytosolic protein 2	0.47	8.1×10−3
Other
Lgals2	lectin, galactose-binding, soluble 2	3.31	9.0 ×10−5
Fv1	Friend virus susceptibility 1	2.11	1.0 ×10−4
Fam84a	family with sequence similarity 84, member A	1.80	1.2×10−3
Mpv17	Mpv17 transgene, kidney disease mutant	1.72	7.7×10−3
Thada	thyroid adenoma associated	1.63	2.4×10−4
Myg1	melanocyte proliferating gene 1	1.55	3.5×10−3
Prosc	proline synthetase co-transcribed	0.62	8.1×10−3
Hmcn1	hemicentin 1	0.36	5.0×10−3

^*Values are mean ratio of the Rs1 vs. control among the 6 samples, P<0.01 in microarray data

The largest group of functionally related, differentially expressed genes was those involved in cell cycle and transcriptional regulation. Others included genes associated with apoptosis, enzyme activity, transporter and membrane protein. We observe no changes in genes associated with skeletal signaling pathways (i.e., Wnt/β catenin signaling, BMP, or IGF signaling pathways). Because our transgenic mouse model supported the concept of non-cell-autonomous effects of OB G_s signaling, we sought to identify candidate paracrine mediators by determining the effect of Rs1 on the expression level of genes encoding secreted proteins. We identified 13 regulated genes, from those in Table 1, that encode secreted proteins.

Furthermore, we identified a set of GPCRs expressed on mature OBs. G_s signaling activity in the unperturbed mature OB is determined by the aggregate activity of this complement of receptors, which also determine the sensory inputs to which G_s signaling activity can respond. No differences in GPCRs expression were observed in the two groups. These results are summarized in Table 2.

Table 2.

G_s- and G_i-coupled GPCRs expressed in FACS-sorted osteoblasts ranked by abundance (validated by Taqman real-time GPCR Array at Ct number below 30)

Gene symbol	Gene name	Ct number (18S=11)	Class	Endogenous ligand(s)	Coupling
S1pr1	sphingosine-1-phosphate receptor 1	17	A	sphingosine-1-phosphate, sphingosylphosphorylcholine	Gi/Go
S1pr2	sphingosine-1-phosphate receptor 2	20	A	sphingosine-1-phosphate, sphingosylphosphorylcholine	Gs, Gq/G11, G12/G13
Lpar1	lysophosphatidic acid receptor 1	20	A	lysophosphatidic acid	Gi/Go, Gq/G11, G12/G13
Lpar6	lysophosphatidic acid receptor 6	20	A	lysophosphatidic acid	Gs, Gi/Go, G12/G13
Ccr1	chemokine (C-C motif) receptor 1	20	A	CCL15, CCL23	Gi/Go
Ptger4	prostaglandin E receptor 4 (subtype EP4)	20	A	PGE2, PGF2α, PGI2, PGD2	Gs/Gi/Go
Gpr30	G-protein coupled receptor 30	21	A	17β-estradiol	Gi/Go, Gs
Agtr1a	angiotensin II receptor, type 1a	21	A	angiotensin I, angiotensin II, angiotensin III	Gi/Go, Gq/G11
Pth1r	parathyroid hormone 1 receptor	22	B	Parathyroid hormone, parathyroid hormone related peptide	Gs, Gq/G11
F2rl2	coagulation factor II (thrombin) receptor-like 2	22	A	thrombin	Gi/Go, Gq/G11, G12/G13
Gpr1 33	G protein-coupled receptor 133	22	Adhesion	orphan	Gs
Aplnr	Apelin receptor	23	A	Apelin-13, apelin-17, apelin-36, [Pyr1]apelin-13	Gi/Go
Lpar4	lysophosphatidic acid receptor 4	24	A	lysophosphatidic acid	Gs, Gi/Go, Gq/G11, G12/G13
Gabbr1	gamma-aminobutyric acid (GABA) B receptor, 1	25	C	GABA	Gi/Go
Lpar3	lysophosphatidic acid receptor 3	26	A	lysophosphatidic acid	Gi/Go, Gq/G11
Gpr34	G protein-coupled receptor 34	26	A	lysophosphatidylserine	Gi/Go
Gpr68	G protein-coupled receptor 68	26	A	protons	Gi/Go, Gq/G11
Cxcr4	chemokine (C-X-C motif) receptor 4	27	A	stroma cell-derived factor 1	Gi/Go
Ptger2	prostaglandin E receptor 2 (subtype EP2)	27	A	PGE2, PGF2α, PGI2, PGD2	Gs
Cmklr1	chemokine-like receptor 1	27	A	resolving E1, chemerin	Gi/Go
Ccr8	chemokine (C-C motif) receptor 8	27	A	CCL1, CCL8	Gi/Go
Gpbar1	G protein-coupled bile acid receptor 1	28	A	lithocholic acid, deoxycholic acid, chenodeoxycholic acid, cholic acid	Gs
S1pr3	sphingosine-1-phosphate receptor 3	29	A	sphingosine-1-phosphate, sphingosylphosphorylcholine	Gi/Go, Gq/G11, G12/G13
Sstr2	somatostatin receptor 2	29	A	SRIF-14, SRIF-28, CST-17	Gi/Go, Gq/G11
Adra2a	adrenergic receptor, alpha 2a	29	A	adrenaline, noradrenaline	Gi/Go, Gs
Gpr84	G protein-coupled receptor 84	29	A	medium-chain-length fatty acid	Gi/Go
Cxcr6	chemokine (C-X-C motif) receptor 6	29	A	CXCL16	Gi/Go

Validation of microarray results by real-time qPCR

qPCR was performed to verify the effectiveness of the sorting process and validated the microarray results. We analyzed the mRNA levels of the transgene and OB markers in GFP positive cells compared to GFP-negative cells from control mice. As expected, early OB marker genes, Runx2 and Osterix, were highly expressed in GFP-negative population, whereas mature OB marker genes, Col1a1 and Osteocalcin, were abundant in GFP-positive cells (Fig. 3B). Rs1 mRNA expression was detected only in OBs from Rs1mice, and its expression was absent in GFP-positive population from control mice (Fig. 3C). These findings suggest that the sorted GFP positive population represents mainly differentiating OBs that express the tetO-responsive GFP and Rs1 genes. We were able to validate by qPCR that 10 of 13 regulated genes that encode secreted proteins were differentially expressed, and the direction of the changes was consistent with the microarray results (Table 3). The difference in the fold change between the two techniques likely reflects difference in sensitivity of the methods.

Table 3.

Validating expression of gene-coded secreted proteins changes in microarrays and qPCR

Gene symbol	Gene name	Fold change
		Microarray#	qPCR
Lipi	Lipase, member I	4.6	8.3*
Figf	c-fos induced growth factor	2.1	15.7*
Slpi	Secretory leukocyte peptidase inhibitor	1.8	7.0*
Rbp4	Retinol binding protein 4, plasma	1.7	1.7*
Ecm1	Extracellular matrix protein 1	1.6	2.0
Mfge8	Milk fat globule-EGF factor 8 protein	1.5	8.1*
Chad	Chondroadherin	0.7	0.3*
Comp	Cartilage oligomeric matrix protein	0.7	0.6
Fgf9	Fibroblast growth factor 9	0.6	0.8*
Thbd	Thrombomodulin	0.6	0.06*
Car2	Carbonic anhydrase 2	0.6	0.3*
Srgn	Serglycin	0.5	0.5
Ostn	Osteocrin	0.3	0.05*

The values in microarray and qPCR analysis are presented as fold changes between three Rs1 and three control mice.

^#P<0.01 in microarray and

^*P≤0.05 in qPCR

Effects of PTH and forskolin treatment on Fgf9 expression in BMSC cultures

We used primary mouse OBs culture to assess whether Fgf9 is a direct target of G_s signaling. We found that exposure of BMSCs with PTH or with forskolin, to raise cyclic AMP levels, markedly down-regulated the expression of Fgf9 compared to controls (−64% and −72% respectively, p<0.05) (Fig. 4A), suggesting that Fgf9 expression maybe directly regulated by G_s signaling in OBs.

Figure 4.

(A) Effects of PTH and forskolin treatment on Fgf9 expression in primary BMSC cultures. Data were normalized to GAPDH RNA expression in the same sample and expressed as percent changes compared to controls. (B) Effect of continuous treatment of BMSCs with murine Fgf9 assessed by VK staining. VK-positive mineralized nodules were stained black. (C) Exposure of Fgf9 from Day 0 to day 20 increased VK staining area. Total colony area positive for VK staining was quantified as area (mm²). The results of these assays were confirmed by repeating the experiment three times. All data are shown as mean ± SD. *: p<0.05.

Effect of treatment of BMSCs with mouse FGF9

The effect of Fgf9 on osteogenesis was assessed in cell culture system. When added Fgf9 directly to mouse BMSCs grown under osteogenic conditions, Fgf9 dramatically suppressed (by>80%) the formation of mineralized nodules, assessed by VK staining, compared to controls (Fig. 4B).

Discussion

We previously reported that strong basal G_s signaling by the Rs1 RASSL expressed in OBs driven by the 2.3-kb Col1 promoter fragment resulted in a dramatic age-dependent increase in endochondral bone formation [10,14,15]. In this study, we investigated in detail the effects of increased G_s signaling in mature OBs on intramembranous bone formation in neonatal and growing Rs1 transgenic mice. The phenotype of Rs1 calvariae resembled that seen in Rs1 femoral bones, indicating that G_s signaling in mature OBs is sufficient to initiate dramatic and similar skeletal responses in both endochrondral and intramembranous bones. The histological analyses demonstrated that Rs1 expression produces an expansion of osteoprogenitor cells that do not express Rs1, supporting the notion that the effect of Rs1 signaling in mature OBs alters bone formation by regulating the expression of factors that influence the early commitment, proliferation, and/or differentiation of osteoprogenitors.

To determine the cellular response in vivo by which mature OBs regulate bone formation in response to G_s signaling, we utilized a microarray approach to examine Rs1-induced alterations in the OB transcriptome. The approach was used to identify an approximate snapshot of the OB transcriptome at the time of sacrifice by isolating the OB population that expresses Rs1 by GFP labeling, without the use of cell culture. We compared gene expression between control OBs and OBs-expressing Rs1. Successful isolation of mature OBs population was confirmed by abundant expression of differentiating OB marker genes, such as Osteocalcin and alkaline phosphatase, in GFP-positive cells. In addition, we compared our data to the findings of Paic et al. [20] who utilized dual GFP receptor mice in which OBs were identified by expression of GFP (cyan) driven by 2.3 kb of the Col1a promoter (Col2.3 cyan). We found a good correlation of OB marker genes and genes associated with OB differentiation from our control OBs and their Col2.3cyan positive OBs (r = 0.74) (Supplemental Fig. 1). These results further validate the efficacy of our procedure for isolating mature OBs.

Interestingly, we found that the magnitude of changes seen in Rs1-expressing OBs was relatively small despite the striking skeletal phenotype. Since the mice were about 1 week old and still in a phase of rapid skeletal growth, the anabolic program in OBs was presumably highly active in wild-type animals, and that might obscured the effects of additional anabolic signaling in response to G_s activation. It is also possible that compensatory signals from the other cells in the bone environment might have attenuated the response of mature OBs to chronic G_s signaling. However, microarray analysis revealed that genes involved in cell cycle and transcriptional regulation were the most changed in mature OBs in response to enhanced G_s signaling.

We then focused on identifying candidate paracrine mediators that were differentially expressed in mature OBs expressing Rs1. Of 13 such regulated genes, 10 were validated by qPCR although the magnitude of changes did not always quantitatively correspond to the microarray results. Interestingly, seven genes (Lipi, Rbp4, Chad, Mfge8, Fgf9, Car2 and Thbd) had at least half site cyclic AMP responsive elements (CREs) within 5 kb of the transcriptional start site, and two of them (Fgf9 and Thbd) had predicted functional CREs in their promoters, as reported by Zhang et al. [21] This implies that these validated genes are potential downstream transcriptional targets of G_s signaling.

Products of some of the validated genes have been reported to have roles in bone cell differentiation and function. For instance, carbonic anhydrase 2 (Car2) is essential for bone resorption and osteoclast differentiation [22,23]. Osteocrine (Ostn) and chrondroadherin (Chad) are reported to be important for bone development. Ostn is expressed in OBs, muscle and fat cells, and OB-specific Ostn-overexpressing mice display elongated long bones from enhanced proliferation and differentiation of growth plate chrondrocytes [24]. Chad is highly expressed in cartilaginous tissues with lower levels in bone and tendon [25,26]. Chad null mice displayed widening of the epiphyseal growth plate with possible impaired of hypertrophic differentiation of chondrocytes [27]. Ostn and Chad expression were significantly downregulated in our model of increase G_s signaling in OB. Nonetheless, we did not observe any appreciable chrondrocyte phenotypes of long bones in vivo. The contribution of Ostn and Chad to the bone changes in Rs1 model has to be further investigated.

Chronic inflammatory processes are often associated with bone loss [28,29]. We found upregulation of secretory leukocyte peptidase inhibitor (Slpi) expression in our transgenic OBs. Slpi downregulates the synthesis of tumor necrosis factors alpha suggesting that it may act as a proinflammation mediator [30]. Recently, Sharma et al. [31] reported downregulation of this gene in axial spondyloarthropathy, an effect that might be important role in the pathogenesis of this disease that often responds markedly to tumor necrosis factors inhibition. Slpi might participate in paracrine mediators effecting osteoclasts. The coagulation system influences bone mineralization [32]. We identified thrombomodulin (Thbd), a gene involved in blood coagulation, to be significantly downregulated by G_s signaling in OBs. Delvaeye et al. [33] showed that Thbd binds to C3b and factor H and negatively regulates complement. It also accelerates the inactivation of anaphylatoxins C3a and C5a and provides protection against complement activation. It remains to establish the role of coagulation system in bone environment in our mouse model. The microarray analysis has also highlighted previously unknown significant effects of G_s signaling in OBs on molecular effectors related to glucose and lipid metabolism, Lipi and Rbp4. Interplay between bone metabolism and glucose or fat metabolism have been extensively addressed [34–36]. Condition that alters genes related to energy metabolism may influence skeletal homeostasis.

At least three validated genes involved in cell growth and angiogenesis were significantly altered in Rs1 expressing OBs: c-fos induced growth factor (Figf; also known as VEGF-D) and milk fat globule-EGF factor 8 protein (Mfge8) were upregulated, and fibroblast growth factor 9 (Fgf9) was downregulated. Angiogenesis and bone formation are coupled during skeletal development and fracture healing [37–39]. Besides playing a role in vascular and lymphatic endothelial cells homeostasis, VEGF-D has been implicated in OB differentiation [40]. During mouse development VEGF-D expression was detected in the OBs of the growth plate, periosteal layer of the developing vertebral column and in the dental mesenchyme close to the enamel epithelium where dentin and enamel matrices are deposited, suggesting that VEGF-D might play a role in intramembranous and long-bone formation [40,41]. VEGF-D/FGFR-3 signaling is a strong candidate to participate in the downstream effects of OB G_s signaling on bone.

Among secreted factors differentially expressed in Rs1 mutants, none has previously been reported to be related to OB-specific G_s signaling. Fgf9 is particularly enticing because it has been identified as having a role during skeletal development and fracture healing [42]. In addition, Fgf9 expression was also significantly altered in OB-specific blockade of G_i-coupled signaling (PTX model) (Data not shown). We demonstrated that Fgf9 expression was markedly down-regulated in the BMSCs treated with forskolin or PTH to raise cyclic AMP levels. Moreover, Continuous treatment of BMSCs with Fgf9 diminished OB mineralization. Taken together, these results support the possibility that alteration in Fgf9 expression induced by G protein signaling in OBs contribute to the changes of bone mass in the Rs1 model. Further studies assessing the bone phenotypes in OB-specific Fgf9 KO mice might needed to clarify the issue.

The balance of G_s- and G_i-GPCR signaling is an important regulator of bone formation. The role of G_i signaling in OBs may be, at least in part, due to antagonized increase in cyclic AMP levels in response to G_s coupled signaling. The microarray results allowed us to assess the relative expression of GPCRs in mature OBs. We identified 27 G_s-and/or G_i-coupled GPCRs whose expressions on OBs were validated by Taqman real-time GPCR Array at Ct number below 30. Many of these receptors have known or suspected roles in OB and osteocyte differentiation and function. Not surprisingly, parathyroid hormone 1 receptor (Pth1r) was highly expressed in this OB population. For many of the identified OB GPCRs, their role in regulating normal bone development and skeletal function is not known. Definitive assessment of the role of specific GPCRs in mature OBs will require evaluation of the effects of targeted deletion.

Several groups investigated gene expression profiles by microarray analysis in OB cell lines [43] and in whole-bone samples [44,45] in response to exogenous parathyroid hormone. This is the first study to date examining alterations in gene expression in vivo in enhanced OB G_s signaling. We found only a small number of differentially expressed genes that overlap with their results (i.e., Car2, Fech, Ecm1, Mfge8, Slpi, Tm4sf1, Car8), consistent with the notion that there are different OB transcriptome effects, depending on the timing and source of G_s activation.

In conclusion, our study provides a useful tool to evaluate the in vivo gene expression specifically occurring in OBs with activated G_s-GPCR signaling, at the cellular level rather than in a whole bone. The study revealed that cell cycle and transcriptional regulation were the most highly enriched pathways. Complete set of GPCRs expressed in OBs and new candidate paracrine mediators in response to chronic G_s activation in OBs were also indentified. All these regulated genes could be used for further functional studies. Biological validation of the role of these mediators may lead to new therapeutic targets for enhancing bone formation in osteoporosis and for reversing bone disease associated with excess G_s signaling (fibrous dysplasia, hyperparathyroidism).

Highlights.

• OB expression of an engineered G_s-coupled receptor dramatically increases bone mass.

• We investigated the alterations in gene expression in vivo in enhanced OB G_s signaling.

• Genes in cell cycle and transcription were increased in enhanced OB G_s signaling.

• GPCRs and paracrine mediators of the effect of G_s signaling in OBs were determined.

Abbreviations

GPCR

G protein–coupled receptor

OBs

osteoblasts

TRAP

tartrate-resistant acid phosphatase

FTIR

fourier transform infrared

SRμCT

synchrotron radiation micro-computed tomography

GFP

green fluorescent protein

qPCR

quantitative real-time polymerase chain reaction

Ct

cycle threshold

FACS

fluorescent activated cell sorting

PCM

primary culture medium

SDM

secondary osteogenic differentiation medium

BMSCs

bone marrow stromal cells

VK

Von Kossa

CRE

cyclic AMP responsive elements