An activator of the cAMP/PKA/CREB pathway promotes osteogenesis from human mesenchymal stem cells^†

Chemical synthesis

All reactions sensitive to air and/or moisture were carried out in oven- or flame-dried apparatus, usually a 3-neck round-bottom flask equipped with a rubber septum inlet, under a dry nitrogen or argon atmosphere using magnetic stirring. ¹H spectra were recorded at 300 MHz on a Bruker 300 NMR spectrometer. Chemical shifts (days) are reported in parts per million (ppm) with reference to tetramethylsilane or the solvent. HPLC analyses were obtained on an Waters 2695 separation module using the following conditions: INNO C18 HPLC column (5.0 µm, 250 mm L × 4.6 mm ID), 25°C column temperature, 1.0 ml/min flow rate, Waters 2996 photodiode array detector (210–600 nm), linear mobile phase gradient of 15–95% B over 20 min, holding 5 min at 95% B (mobile phase A: 0.1% TFA in water; mobile phase B: acetonitrile). Flash chromatography was carried out using silica gel 60 (230–400 mesh, Merck, Whitehouse Station, NJ). Reagent-grade chemicals were purchased from Sigma–Aldrich (St. Louis, MO) and used as received unless otherwise specified. Purity of the final compounds was >98% by HPLC. Synthetic procedures of CW008 are described in Supplementary Materials and Methods.

Cell culture

Human bone marrow-derived mesenchymal stem cells (hBMSCs) were purchased from Lonza (Walkersville, MD) and maintained in Dulbecco's Modified Eagle Medium (DMEM) (Lonza) supplemented with 20% fetal bovine serum (FBS, Gibco BRL, Grand Island, NY), 100 units/ml penicillin and 0.1 mg/ml streptomycin (Gibco BRL). For osteogenesis, hBMSCs were plated 10⁴ cells/well in 96-well plates. After 2 days, hBMSCs were cultured under osteogenesis induction medium (OIM, 20% FBS, 10 mM β–glycerophosphate (USB Corp., Cleveland, OH), 50 µM ascorbate-2-phosphate (Sigma–Aldrich) and 100 nM dexamethasone (Sigma–Aldrich) in DMEM). OIM was changed every 3 days. Cells were grown at 37°C in a humidified atmosphere containing 5% CO₂. hBMSCs for all experiments were used between cell passages 6–10.

Estimation of alkaline phosphatase (ALP) activity

Following culture, cells were lysed using protein lysis buffer (50 mM Tris–HCl (pH 7.5) and 0.1% TX-100). Cellular alkaline phosphatase activity was assayed colorimetrically by incubating cell lysates with the substrate p-nitrophenylphosphate (Sigma–Aldrich) in assay buffer (4 mM MgCl₂ (pH 10.5) and 200 mM 2-amino-2-methyl-1-phosphate) at 37°C for 15 min. Absorbance was measured at 405 nm.

Alizarin red S staining

Mineral accumulation of differentiated osteoblasts was observed and photographed by using a Zeiss Axiovert 135 microscope plus Olympus DP71 CCD camera (Olympus Corporation, Japan).

To confirm the mineral deposition by differentiated osteoblasts, cells were washed twice with phosphate buffered saline (PBS), fixed in 4% paraformaldehyde for 20 min, washed twice with distilled water and then stained with a 1% alizarin red solution (Sigma–Aldrich) for 20 min. Cells were then washed three times with distilled water and examined for the presence of calcium deposits, which were identified by the presence of a red color. Images were obtained via scanning.

RNA extraction and real-time quantitative RT-PCR

Total RNA was extracted from hBMSCs using TRIzol reagent (Invitrogen, Grand Island, NY). cDNA was reverse-transcribed from 2 µg total cellular RNA using oligo (dT) primers and Moloney murine leukemia virus reverse transcriptase (Promega, Madison, WI). Quantitative real-time PCR was performed using SYBR Green I Master mix (Roche, Indianapolis, IN) with a Roche LightCycler 480 Real-Time PCR system. PCR conditions consisted of a 10 min hot start at 95°C followed by 45 cycles of 15 sec at 95°C, 10 sec at 60°C and 30 sec at 72°C. Following primers were used: human alkaline phosphatase (ALP) 5′-taacatcagggacattgacg-3′ (sense), 5′-tgcttgtatctcggtttgaa-3′ (antisense); human runt-related transcription factor 2 (Runx2) 5′-cagcctgcagcccggcaaaa-3′ (sense), 5′-cgcaaccggggcactgcag-3′ (antisense); human bone morphogenic protein-2 (BMP2) 5′- ggcatcctctccacaaaaga-3′ (sense), 5′- gtggcagtaaaaggcgtgat-3′ (antisense); human inhibitor of DNA binding 2 (ID2) 5′- cgtcatcgactacatcttggac-3′ (sense), 5′-acacagtgctttgctgtcattt-3′ (antisense); human leptin (LEP) 5′-caaaaagtccaagatgacacca-3′ (sense), 5′- tgttggtagactgccagtgtct-3′ (antisense); the reference gene, human ribosomal protein large P0 (RPLP0) 5′-ggaatgtgggctttgtgttc-3′ (sense), 5′-tgcccctggagattttagtg-3′ (antisense). The expression levels of the mRNAs were normalized to the expression level of RPLP0 and compared.

In vivo osteoporosis model

Female 8-week-old mice were ovariectomized (OVX) under pentobarbital anesthesia and divided into five groups; normal, sham-operated control, OVX-vehicle, OVX-CW008 (15 mg/kg), and OVX-alendronate (5 mg/kg) (Alexander et al., 2001). Eight mice per group were examined. At 4 weeks after the surgery, the mice were subcutaneously injected with vehicle, CW008 or alendronate once a day. After 4 weeks, the mice were euthanized and the major organs were weighed. Bone formation was analyzed with a three-dimensional microcomputed (µCT) system (TOMONIX V90, RAY Co, Korea) as previously described (Ofek et al., 2006; Smoum et al., 2010). Animal study protocols were approved by the Institutional Animal Care and Use Committee at Pohang University of Science and Technology.

Immunoblotting

For immunoblotting, whole-cell lysates were prepared in lysis buffer (50 mM Tris–HCl (pH 7.4), 150 mM NaCl, 1% TX-100, 1 mM EDTA, 1 mM EGTA, 50 mM NaF, 1 mM Na₃VO₄, 1 mM PMSF, and protease inhibitor cocktail (Sigma–Aldrich)). Lysates were then centrifuged at 14,000 rpm for 15 min (4°C). A total of 20 µg of protein was separated by **SDS-PAGE using 10% gels and transferred to nitrocellulose membranes. The membranes were blocked with 5% nonfat dry milk in Tween 20/Tris-buffered saline. For detection, the membranes were incubated with primary antibodies overnight at 4°C. Horseradish peroxidase-conjugated secondary antibodies were incubated with the membrane for 1 h. Signals were visualized by chemiluminescence (ECL system, Amersham Biosciences, Piscataway, NJ).

Plasmids and luciferase activity assay

To assess luciferase activity, a pCRE-luciferase reporter plasmid, which contains the SV40 30 intron and a polyadenylation signal, was purchased from Agilent Technologies (Santa Clara, CA). The pRL-SV40 construct, which was used Renilla luciferase expression vector, was purchased from Promega. hBMSCs were seeded into 24-well plates and cultured for 24 h before transfection. The cells were transfected with a DNA mixture containing the pCRE-luciferase reporter plasmid and the internal control plasmid pRL-SV-40 (25 ng) using the Lipofectamine transfection reagent (Invitrogen) according to the manufacturer's instructions. Forty-eight hours after transfection, the cells were treated with either forskolin or test compound and incubated for 6 h. The luciferase activity of the cell lysates was measured using the Dual-Luciferase® Reporter Assay System according to the manufacturer's instructions (Promega). Relative luciferase activity was normalized for transfection efficiency using the corresponding Renilla luciferase activity.

cAMP immunoassay

cAMP concentrations in hBMSCs were analyzed using the cAMP kit from R&D Systems (Minneapolis, MN) according to the manufacturer's instructions. In brief, hBMSCs (10⁶ cells) were treated with vehicle, CW008 or forskolin for 6 h in osteogenic differentiation condition, and the cell lysates were used for the assay.

Detection of leptin secretion

Leptin concentrations in the conditioned medium were measured using the MILLIPLEX Human Bone Panel according to the manufacturer's instructions (Millipore, Billerica, MA).

General reagents

Phospho-CREB and CREB antibodies were purchased from Cell Signaling Technology (Beverly, MA). Anti-mouse and anti-rabbit horseradish peroxidase-conjugated secondary antibodies were obtained from KPL (Gaithersburg, MD).

Forskolin and H-89 were purchased from Calbiochem (San Diego, CA) and leptin was purchased from R&D Systems.

Statistical analysis

The data were analyzed using Student's t-tests. P < 0.05 and P < 0.01 were considered to be significant.

CW008 stimulates osteogenesis of hBMSCs

To identify a small molecule that promotes osteoblast differentiation, we performed a stem cell-based screen of a chemical library of containing 192 kinase scaffold compounds. As a primary screening, we selected a small compound 1C2 that increased alkaline phosphatase (ALP) activity during early osteogenic differentiation. After then, more compounds were synthesized based on the hit compound of primary screening and we performed second screening, that is to identify small compounds to stimulate mineral accumulation in the extracellular matrix. The differentiation of hBMSCs into mature osteoblasts generally occurred after 7–12 days in the presence of osteogenic differentiation conditions in our system and osteoblast differentiation was confirmed by alizarin red S staining to detect the mineral accumulation and calcium deposits. As a result, we could select an active small compound, CW008, to promote differentiation of hBMSCs into osteoblasts (Fig. 1A and Supplementary Fig. 1). Treatment of hBMSCs with CW008 enhanced ALP activity at 2 and 4 days after osteogenesis induction (Fig. 1B). Under osteogenic differentiation conditions, CW008 induced dramatic mineralization of osteoblasts compared with vehicle treatment (Fig. 1C). Mineralization of differentiated osteoblasts was detected by alizarin red S staining at 12 days after osteogenesis induction (Fig. 1D). CW008 increased calcium deposits in the extracellular matrix in a dose-dependent manner. Remarkably, CW008 also induced the expression of osteogenic marker genes, including alkaline phosphatase (ALP) and Runx2, during hBMSC osteogenesis (Fig. 1E). These results show that identified small compound CW008 is a potent activator of osteogenesis. In contrast to the osteogenic effect of CW008, CW008 suppressed the adipocyte differentiation of hBMSCs in a dose-dependent manner (Supplementary Fig. 2). In response to CW008 treatment, accumulation of intracellular lipid droplets, a characteristic of mature adipocytes, was reduced and it was detected by oil red O staining.

CW008 enhances bone formation in vivo osteoporosis model

Next, we examined the effect of CW008 in ovariectomized (OVX) mice, which are widely used as a model of estrogen deficiency and bone loss (Jerome and Peterson, 2001; Yamane et al., 2009; Smoum et al., 2010). Mice were sham-operated or ovariectomized and left for 4 weeks to allow for the occurrence of bone loss. Trabecular bones exhibited marked bone loss after ovariectomy compared with those from sham-operated mice. Following bone loss, vehicle, CW008 (15 mg/kg), or alendronate (ALN, 5 mg/kg) was administrated daily for 4 weeks. Alendronate, which is currently used osteoporosis drug (Baron et al., 2011), was used as a control to compare the activity of CW008. Microcomputed tomography (µCT) analyses revealed that mice treated with CW008 showed a marked increase in bone mass compared with vehicle-treated mice (Fig. 2A). Similarly, the trabecular bone volume density (BV/TV) in the distal femoral metaphysis was increased by CW008 treatment to a degree comparable to that achieved by alendronate treatment (Fig. 2B). In addition, trabecular thickness (Tb.Th) and cortical thickness (Col.Th) in CW008-treated OVX mice were higher than those in vehicle-treated OVX mice (Fig. 2C,D). These results demonstrate that active compound CW008 increases bone formation in osteoporosis treatment model.

CW008 activates the cAMP/PKA/CREB signaling pathway

To define the molecular mechanism underlying CW008-stimulated osteogenic activity, we examined several known signaling pathways that have been implicated in osteogenic differentiation, including the cAMP/PKA, the Wnts/β-catenin and the MAPK pathway. hBMSCs were treated with vehicle or CW008 (1 µM) in osteogenic differentiation medium (Fig. 3A). During the first 12 h of differentiation, CW008 increased CREB phosphorylation compared with vehicle treatment. The level of CREB phosphorylation then returned to baseline at 24 h after treatment. The ability of CW008 to activate CREB phosphorylation was compared to that of the inactive compound CW 138 (1 µM) or forskolin, a PKA activator. Forskolin also induced CREB phosphorylation to a degree similar to CW008 treatment (Fig. 3B).

We next examined the effect of CW008 on CRE-dependent transcription via a CRE-directed luciferase reporter gene assay (Fig. 3C). hBMSCs were transfected with a CRE-Luciferase construct and treated with vehicle, CW008 (1 µM) or forskolin (1 µM) for 6 h in osteogenic differentiation medium. In response to CW008 treatment, CRE-luciferase activity was increased approximately 2.5-fold compared with cells treated with vehicle. Forskolin was used as a positive control.

To determine the correlation between CW008-mediated CRE-dependent transcription and CREB phosphorylation and intracellular increases in cAMP, we examined whether CW008 (1 µM) induced cAMP production in hBMSCs. We found that after 6 h of incubation, CW008 caused a slight increase in cAMP production. This increase in cAMP production was comparable to that caused by forskolin (1 µM) treatment (Fig. 3D).

As we observed that CW008 activated the cAMP/PKA/CREB signaling pathway, we examined whether CW008 regulated the expression of CRE-target genes. hBMSCs were treated with vehicle, CW008 (1 µM) or forskolin (1 µM) for 3 days in osteogenic differentiation medium. CW008 increased the expression of the known CRE-target genes BMP2 and ID2 compared to vehicle treatment (Fig. 3E). Forskolin also enhanced BMP2 and ID2 expression. Significantly, CW008-treated cells expressed higher levels of ID2 than forskolin-treated cells. We also examined changes in leptin (LEP) expression. Leptin is an adipocyte-derived hormone that has been known as an inhibitor of bone formation in vivo (Ducy et al., 2000; Elefteriou et al., 2004). PKA activity is known to regulate leptin secretion (Szkudelski et al., 2005), therefore, the effect of CW008 (1 µM) on LEP expression was examined (Fig. 3E). CW008-stimulated hBMSCs displayed reduced LEP expression compared to vehicle-treated cells. Forskolin also inhibited LEP expression. Significantly, LEP gene expression was almost undetectable following CW008 treatment. These results suggest that CW008 activates cAMP/PKA/CREB signaling pathway and regulates gene expressions of BMP2, ID2, and LEP.

Inhibition of PKA blocks CW008-stimulated osteogenesis

To determine whether CW008 exerts its osteogenic activity via the cAMP/PKA/CREB signaling pathway, we first confirmed that forskolin enhanced osteogenesis of hBMSCs (Fig. 4A).

Next, we examined the effect of H-89, which is an inhibitor of PKA activity, on CW008-mediated osteogenesis. In the presence of H-89 (1–10 µM), CW008-induced CREB phosphorylation was abolished and it was confirmed by immunoblotting (Fig. 4B). Additionally, H-89 treatment suppressed CW008-stimulated osteogenesis. In osteogenic differentiation condition, hBMSCs were treated with CW008 with or without H-89 for 7 days. Co-treatment with H-89 resulted in decreased mineral accumulation; in contrast to cells treated with only CW008, only baseline levels of differentiation were observed in H-89 co-treatment (Fig. 4C). Moreover, the increase in osteoblast marker expression by CW008 was abolished by H-89 (1 µM) co-treatment (Fig. 4D). Furthermore, CW008-mediated upregulation of BMP2 and ID2 expression was reduced in the presence of H-89. Conversely, CW008-mediated downregulation of LEP expression was increased in response to H-89 (Fig. 4E). Therefore, these results suggest that CW008-stimulated osteogenesis is mediated by the cAMP/PKA/CREB signaling pathway and that CW008-mediated upregulation of BMP2 and ID2 expression and downregulation of LEP expression are controlled via the cAMP/PKA/CREB signaling pathway.

CW008 promotes hBMSCs osteogenesis by suppressing leptin secretion

To further investigate the key mediator of CW008-stimulated osteogenesis, we focused on the suppression of LEP expression that was observed with CW008 treatment. Leptin secretion in conditioned medium (CM) from CW008-treated cells during osteogenic differentiation was markedly suppressed. In contrast to the CM from vehicle-treated cells, leptin was almost undetectable in the CM from CW008-treated cells at 3 and 6 days after osteogenesis induction (Fig. 5A). To confirm that PKA activation suppresses leptin secretion, H-89 (1 µM) was added to CW008-treated (1 µM) or forskolin-treated (1 µM) cells under osteogenic differentiation condition. After 6 days, the leptin concentration in the CM was examined (Fig. 5B). H-89 co-treatment remarkably abolished the suppressive effects of CW008 and forskolin on leptin secretion and restored leptin level in CM. Additionally, H-89 treatment increased leptin secretion in vehicle-treated cells.

To determine whether CW008-stimulated osteogenesis is mediated by the suppression of leptin secretion, hBMSCs were treated with CW008 in the presence or absence of leptin under osteogenic differentiation condition (Fig. 5C). After 7 days, osteogenic differentiation was confirmed by alizarin red S staining. Interestingly, leptin co-treatment reduced the CW008-mediated mineralization of osteoblasts in a dose-dependent manner. Furthermore, osteogenic marker expression activated by CW008 was also reduced in the presence of leptin (Fig. 5D). These results suggest that leptin plays a negative role in CW008-stimulated osteogenesis.

On the other hand, leptin has been known to promote hydrolysis of intracellular cAMP via phosphodiesterase (PDE) activation (Sahu, 2004; Hsu et al., 2006). We hypothesized that leptin might inhibit the osteogenesis activated by CW008 through stimulation of cAMP degradation. To demonstrate the hypothesis, 3-isobutyl-1-methylxanthine (IBMX, PDE inhibitor) was co-treated with CW008 and leptin in osteogenic differentiation condition and we could observe that IBMX recovered osteogenic activity of CW008 suppressed by leptin (Supplementary Fig. 3). These results show that activation of cAMP/PKA/CREB signaling pathway is essential and, at least part of, suppression of leptin secretion contribute to the osteogenic activity of CW008.