PDGF Receptor β Is a Potent Regulator of Mesenchymal Stromal Cell Function

Authors


  • The authors state that they have no conflicts of interest.

  • Published online on April 14, 2008

Abstract

Mesenchymal stromal cells (MSCs) in bone marrow are important for bone homeostasis. Although platelet-derived growth factor (PDGF) has been reported to be involved in osteogenic differentiation of MSCs, the role remains controversial and the network of PDGF signaling for MSCs has not been clarified. To clarify the underlying regulatory mechanism of MSC functions mediated by PDGF, we deleted the PDGF receptor (PDGFR)β gene by Cre-loxP strategy and examined the role of PDGF in osteogenic differentiation of MSCs and fracture repair. In cultured MSCs, the mRNA expression of PDGF-A, -B, -C, and -D as well as PDGFRα and β was detected. Depletion of PDGFRβ in MSCs decreased the mitogenic and migratory responses and enhanced osteogenic differentiation as evaluated by increased alkaline phosphatase (ALP) activity and mRNA levels of ALP, osteocalcin (OCN), bone morphogenetic protein (BMP) 2, Runx2, and osterix in quantitative RT-PCR. PDGF-BB, but not PDGF-AA, inhibited osteogenic differentiation accompanied by decreased ALP activity and mRNA levels, except for BMP2. These effects of PDGF-BB were eliminated by depletion of PDGFRβ in MSCs except that PDGF-BB still suppressed osterix expression in PDGFRβ-depleted MSCs. Depletion of PDGFRβ significantly increased the ratio of woven bone to callus after fracture. From the combined analyses of PDGF stimulation and specific PDGFRβ gene deletion, we showed that PDGFRβ signaling distinctively induces proliferative and migratory responses but strongly inhibits osteogenic differentiation of MSCs. The effects of PDGFRα on the osteogenic differentiation were very subtle. PDGFRβ could represent an important target for guided tissue regeneration or tissue engineering of bone.

INTRODUCTION

Bone homeostasis is balanced between bone formation by osteoblasts and bone absorption by osteoclasts to retain a constant bone volume and is altered dramatically in physiological and pathological conditions including metabolic diseases and injuries. Multipotent mesenchymal stromal cells (MSCs) reside in bone marrow and play key roles in bone homeostasis, because they actively proliferate, migrate, and undergo osteogenic, chondrogenic, or adipogenic differentiation in response to different stimuli.(1) The biology of MSCs is a focus of interest in clinical medicine for the treatment of various bone diseases. For example, recent cell-based bioengineering trials have shown the usefulness of MSCs for treatment of osteogenesis imperfecta, large bone defects, and steroid-induced osteonecrosis of the femoral head.(2–4) However, the regulatory mechanism of MSC differentiation remains to be elucidated.

Bone morphogenetic proteins (BMPs), which are members of the TGFβ superfamily, have well-characterized roles in the osteogenic differentiation process of MSCs, because they induce migration and gain of osteogenic phenotypes in these cells.(5,6) BMPs activate downstream signaling molecules, including Dlxs and Smads, and induce the expression of osteogenic genes, such as runt-related transcription factor (Runx) 2, osterix, alkaline phosphatase (ALP), and osteocalcin (OCN).(6) In addition, many different cytokines, growth factors, and their receptors have been identified in MSCs.(7) Among the growth factor receptors, membrane-bound tyrosine kinase-type receptors for growth factors such epidermal growth factor (EGF), fibroblast growth factor, and platelet-derived growth factor (PDGF) activate diverse signaling molecules such as mitogen-activated protein (MAP) kinases, signal transducer and activator of transcription (Stat) proteins, and Smads, and data are accumulating to show that these signaling molecules strongly affect the BMP signaling pathway that induces osteogenic differentiation of MSCs. The activation of each growth factor receptor has quite different effects on the proliferation, migration, and differentiation of MSCs, and it is indispensable to clarify the significance of individual growth factor signals to understand their roles in bone formation.

PDGFs are potent mitogens for mesenchymal cells and are involved in the wound healing processes of various organs by the induction of fibroplasia and angiogenesis.(8) PDGF-BB also strongly induces the proliferation and migration of MSCs.(9–11) PDGF ligands and receptors are upregulated during tissue remodeling in bone fractures, and PDGF-A and -B synthesized by osteoblasts, chondrocytes, and MSCs are assumed to recruit MSCs to lesion sites to accelerate the repair process.(12,13) In addition, PDGFs are considered to affect MSC differentiation, although a lot of controversy exists. PDGF-BB inhibited the expression of ALP, OCN, and type I collagen, marker proteins of mature osteoblasts, in mineralizing osteoblast cultures and inhibited the mineralization of pre-osteocytic cell line cells.(14,15) It was also reported that PDGF did not alter ALP activity and mineralization in human MSCs.(9,16) Imatinib-induced blockade of PDGFRβ showed mutually opposing effects on the osteogenic differentiation of MSCs in two different reports.(10,17) In vivo application of PDGF enhanced bone formation in three reports(18–20) but suppressed it in another report.(21)

The PDGF ligands and receptors that are important for MSCs have not yet been characterized. The PDGF family members PDGF-A and -B, and the newly discovered PDGF-C and -D, are assembled as disulfide-linked homo- or heterodimers and exert their activity by binding to and activating specific high-affinity cell surface receptors.(22,23) Two receptor tyrosine kinases, PDGFRα and β, that can form homo- and heterodimeric receptor complexes have been identified. PDGFRαα can be specifically activated by PDGF-AA, -AB, -CC, and -BB, whereas PDGFRαβ is activated by PDGF-AB, -BB, and -CC, and PDGFRββ is activated by PDGF-BB and -DD.

Two types of PDGF receptors (PDGFRs) were found to be expressed in MSCs(7) and elaborated similar but distinctive signals.(24) Although PDGFRβ is considered to be important for the migration response of MSCs based on the effects of neutralizing antibodies against PDGFRs,(25) the relevance of each PDGFR in MSCs has not been characterized. Recently, we established mutant mice in which selective depletion of PDGFRβ can be induced both in vivo and in vitro by use of the Cre-loxP system.(24,26) In this study, we aimed to clarify the underlying regulatory mechanism of MSC functions mediated by PDGF. We depleted PDGFRβ in MSCs from the above-described mutant mice and analyzed their proliferative, migratory, and differentiation responses.

MATERIALS AND METHODS

Mice

The mutant mice, in which exons 4–7 of the PDGFRβ gene were flanked by two loxP sequences (PDGFRβflox/flox),(24) were cross-bred to Cre transgenic mice (Cre-ER; CAGGCre-ER; Jackson Laboratories) systemically expressing a fusion protein consisting of Cre recombinase and a mutated form of the mouse estrogen receptor ligand-binding domain.(27) The resultant offspring with genotypes Cre-ER+/−/PDGFRβflox/flox and PDGFRβflox/flox were used for our experiments. The mice were maintained with free access to laboratory pellet chow and water and exposed to a 12-h light/12-h dark cycle. All animal procedures were performed according to the Institutional Animal Care and Use Committee Guidelines at University of Toyama under an approved protocol.

Isolation of MSCs

Murine MSCs were obtained from excised femurs, humeri, and tibias by flushing the shafts with DMEM (Nissui). The cells were cultured in DMEM containing 20% FBS (JRH Biosciences), 4 mM l-glutamine, and penicillin (100 U/ml)/streptomycin (100 ng/ml) in humidified 5% CO2/95% air at 37°C. Nonadherent cells were removed by washing at 4 days after plating, and adherent MSCs were obtained.

Induction of PDGFRβ gene deletion in cultured MSCs

MSCs obtained from Cre-ER+/−/PDGFRβflox/flox mice were treated with 100 nM 4-hydroxy-tamoxifen (Sigma-Aldrich) for 48 h(27) to obtain MSCs in which the PDGFRβ gene had been deleted (PDGFRβdel/del MSCs). MSCs obtained from PDGFRβflox/flox mice were treated identically and used as controls (PDGFRβflox/flox MSCs).

Proliferation assay

MSC proliferation was examined using a Cell Counting Kit-8 (Dojindo) according to the manufacturer's instructions. MSCs treated with 4-hydroxy-tamoxifen were replated at 3000 cells/well on 96-well plates (BD Falcon) and cultured in DMEM containing 20% FBS for 24 or 48 h. Next, the dye solution was added, and the cells were incubated for a further 2 h at 37°C. Formazan synthesis in each well was measured by the absorbance at 450 nm (Immuno Reader NJ-2100UV; Nippon Intermed). The experiment was repeated six times.

Wound scratch assay

PDGFRβflox/flox and PDGFRβdel/del MSCs were replated on separate type I collagen-coated 35-mm dishes (Iwaki). After reaching confluence, the MSCs were serum-starved for 24 h and scratched with a plastic pipette tip. Next, the cells were cultured in DMEM containing either 20% FBS or 0.4 nM PDGF-BB (Chemicon) for 40 h. Wound closure was observed under an inverted phase-contrast microscope (Axiovert 135; Carl Zeiss).

Differentiation induction and staining

MSCs plated at 2.5 × 104 cells/well on 6-well plates (BD Falcon) were incubated in osteogenic differentiation medium (DF-B; Toyobo), and osteogenic differentiation was confirmed by ALP staining as previously described.(28) Other MSCs were incubated in chondrogenic differentiation medium (DF-C; Toyobo) for 4 wk according to a previously described pellet culture technique.(29) After harvesting, the cell pellets were fixed in 4% paraformaldehyde, embedded in paraffin, cut into 5-μm-thick sections, and stained with Safranin O to confirm chondrogenic differentiation. Finally, MSCs were cultured for 1 wk in adipogenic differentiation medium (PADM; Toyobo), and adipogenic differentiation was assessed by Oil Red O staining.

Table Table 1.. Primers Used for Quantitative RT-PCR
original image

ALP activity

MSCs were plated at 1 × 104 cells/well on 96-well plates and cultured with osteogenic differentiation medium with or without either 0.8 nM PDGF-AA (Chemicon) or 0.8 nM PDGF-BB. The culture medium was exchanged for fresh medium every 3 days. At 7 days after plating, the ALP activities in cell lysates were assayed by colorimetric measurement of the amount of p-nitrophenol hydrolyzed from p-nitrophenylphosphate at 405 nm using the above-mentioned Immuno Reader. Each ALP activity was normalized to the corresponding total protein concentration assessed using a BCA Protein Assay Kit (Pierce).

Quantitative RT-PCR

Total RNA was extracted from MSCs using Isogen (Nippon Gene), purified with an SV Total RNA Isolation System (Promega), and used as a template for cDNA synthesis by Multiscribe Reverse Transcriptase (PE Applied Biosystems). Primers were purchased from Takara or designed as indicated in Table 1. The reactions were performed in an ABI PRISM 7000 system using a QuantiTect SYBR Green PCR kit (Qiagen) according to the manufacturer's instructions. After validation assays for each gene primer set, the gene expression levels in each sample were determined by the comparative Ct method using the β-actin gene as an endogenous control. All reactions were performed in triplicate and repeated two to four times.

Fracture experiments

PDGFRβ was depleted in Cre-ER+/−/PDGFRβflox/flox mice by oral administration of tamoxifen (Sigma-Aldrich; 9 mg/40 g body weight, 5 consecutive days) at 4 wk of age. PDGFRβflox/flox mice were treated identically and used as controls. After a further 4 wk, the mice were subjected to the fracture experiment. Under pentobarbital anesthesia (0.5 mg/10 g body weight), the tibia was subjected to osteotomy at a site 4 mm peripheral to the tibial tuberosity using bone scissors, repositioned, and stabilized with an intramedullary nail. At 7 days after the operation, the animals were perfusion-fixed and immersion-fixed with 4% paraformaldehyde in 0.01 M PBS at 4°C overnight. After decalcification with 0.5 M ethylenediaminetetraacetic acid solution at 37°C for 72 h, the fractured bone was embedded in paraffin, sectioned in the sagittal direction, and mounted on silane-coated glass slides. The tissue sections were deparaffinized, rehydrated, and stained with hematoxylin-eosin (HE), Azan, and Safranin O to appreciate callus, woven bone, and cartilage, respectively. The area of woven bone and cartilage was measured in the posterior part of the bone where we observed consistent reparative tissue morphology. The area of these components was measured by a VH analyzer (VH-H1A5; Keyence).

Statistical analyses

Data were expressed as means ± SE. Statistical analyses were performed using Student's t-test. p < 0.05 were considered to indicate significance.

RESULTS

Differentiation induction in bone marrow–derived MSCs

Before initiating 4-hydroxy-tamoxifen treatment, we confirmed that plastic-adherent MSCs isolated from CreER+/−/PDGFRβflox/flox and PDGFRβflox/flox mice were capable of undergoing multiple differentiation processes. In addition to osteogenic differentiation, MSCs incubated in chondrogenic differentiation medium synthesized proteoglycans detected by Safranin O staining and expressed mRNAs for Sox9, type II collagen, aggrecan, and type X collagen, whereas MSCs incubated in adipogenic differentiation medium accumulated intracytoplasmic lipid droplets detected by Oil Red O staining and showed enhanced expression of peroxisome proliferative-activated receptor γ mRNA (data not shown).

Confirmation of PDGFRβ depletion by quantitative RT-PCR

For functional analyses, we added 4-hydroxy-tamoxifen to the culture media of MSCs with two different genotypes and induced specific deletion of the PDGFRβ gene. The efficiency of PDGFRβ gene deletion was evaluated by quantitative RT-PCR using primers that amplified cDNA sequences targeted by Cre recombinase (Table 1). After 4-hydroxy-tamoxifen treatment, the relative amount of amplifiable PDGFRβ mRNA in MSCs from Cre-ER+/−/PDGFRβflox/flox mice was far lower than that in control PDGFRβflox/flox MSCs, and these cells were designated PDGFRβ-deleted MSCs (PDGFRβdel/del MSCs, Fig. 1; PDGFRβdel/del cells, 1.00 × 10−3 ± 2.00 × 10−4; PDGFRβflox/flox cells, 1.05 × 10−2 ± 8.00 × 10−4; p < 0.001). We also examined the expression levels of PDGFRα and all PDGF ligands. The expression levels of PDGF-A and -C were much higher than those of PDGF-B and -D. All mRNAs, except for PDGFRβ, were at similar levels between the two types of MSCs, indicating that deletion of the PDGFRβ gene did not affect the expression levels of the other genes examined.

Figure Figure 1.

mRNA levels of PDGF ligands and receptors in PDGFRβflox/flox (horizontally striped bars) and PDGFRβdel/del (open bars) MSCs. Quantitative RT-PCR for PDGF-A, -B, -C, and -D as well as PDGFRα and β was performed using RNA extracted from the MSCs on day 1 after treatment with 100 nM 4-hydroxy-tamoxifen for 48 h. The expression level of each gene was calculated by the comparative Ct method and expressed as the fold amount relative to the corresponding expression level of β-actin. Values are means ± SE.

Effects of PDGFRβ depletion on cell proliferation and wound scratch assays

We evaluated the effects of PDGFRβ depletion on MSC proliferation based on formazan formation in DMEM containing 20% FBS (Fig. 2). Proliferation was decreased in PDGFRβdel/del MSCs compared with PDGFRβflox/flox MSCs at 24 h (PDGFRβdel/del cells, 0.035 ± 0.003; PDGFRβflox/flox cells, 0.087 ± 0.008; p < 0.001) and 48 h (PDGFRβdel/del cells, 0.107 ± 0.006; PDGFRβflox/flox cells, 0.262 ± 0.021; p < 0.001). We estimated the effects of PDGF-BB and PDGFRβ on the combined migratory and proliferative response of MSCs to close a wound after scratching (Fig. 3). In control PDGFRβflox/flox cells cultured in DMEM containing 20% FBS, the wound was closed within 40 h after scratching. Similarly, substantial numbers of PDGFRβflox/flox cells had filled the scratched area after stimulation with 0.4 nM PDGF-BB. On the other hand, wound closure was mostly suppressed in PDGFRβdel/del cells after culture in DMEM containing 20% FBS or 0.4 nM PDGF-BB.

Figure Figure 2.

Effects of PDGFRβ gene deletion on MSC proliferation. The proliferation activities were analyzed by a formazan assay. MSCs were cultured on 96-well plates for 24 and 48 h, and the dye solution was added to the culture medium at 2 h before analysis. The absorbances of formazan in PDGFRβflox/flox (horizontally striped bars) and PDGFRβdel/del (open bars) cells were measured at 450 nm. Values are means ± SE.

Figure Figure 3.

Effects of PDGFRβ gene deletion on wound scratch assays. PDGFRβflox/flox and PDGFRβdel/del MSCs were plated on type I collagen-coated 35-mm dishes. After serum starvation for 24 h, the cells were scraped with a plastic pipette tip and incubated in DMEM containing 20% FBS or 0.4 nM PDGF-BB for 40 h. Large black dots were placed on the bottom surface of each culture dish to specify the site of observation. Representative images from three separate experiments are shown.

Effects of PDGFRβ depletion and PDGF stimulation on ALP activity

To evaluate the effects of PDGF and PDGFRβ on osteogenic differentiation, ALP staining was performed on PDGFRβflox/flox and PDGFRβdel/del MSCs cultured in osteogenic differentiation medium for the indicated periods (Fig. 4A). In PDGFRβflox/flox cells, ALP staining became visible on day 7 and was increased on day 14. In PDGFRβdel/del MSCs, diffuse ALP staining was observed on day 1 and increased thereafter. When compared, the staining in PDGFRβdel/del MSCs was stronger than that in PDGFRβflox/flox MSCs throughout the experiment.

Figure Figure 4.

Effects of PDGFRβ depletion and PDGF stimulation on ALP activity. (A) ALP staining was performed on PDGFRβflox/flox and PDGFRβdel/del MSCs cultured in osteogenic differentiation medium (DF-B) at the indicated times. (B) ALP staining was performed on PDGFRβflox/flox and PDGFRβdel/del MSCs cultured in DF-B alone (left), DF-B + 0.8 nM PDGF-AA (middle), and DF-B + 0.8 nM PDGF-BB (right) for 7 days. (C) ALP staining was performed on PDGFRβflox/flox and PDGFRβdel/del MSCs cultured in DF-B alone (open bars), DF-B + 0.8 nM PDGF-AA (dotted bars), and DF-B + 0.8 M PDGF-BB (oblique bars) for 7 days. Values are means ± SE. All experiments were performed three times.

Next, we added either 0.8 nM PDGF-AA or PDGFBB to the osteogenic differentiation medium and examined their effects on ALP staining (Fig. 4B). After 7 days of culture, PDGF-AA in the culture medium had no effect on the ALP staining in either PDGFRβflox/flox or PDGFRβdel/del MSCs. In contrast, PDGF-BB strongly suppressed the ALP staining in PDGFRβflox/flox cells but did not affect that in PDGFRβdel/del cells. The ALP activities were measured by colorimetry (Fig. 4C). The ALP activity was strongly increased in PDGFRβdel/del cells compared with PDGFRβflox/flox MSCs cultured in osteogenic differentiation medium for 7 days. PDGF-BB, but not PDGF-AA, significantly suppressed the ALP activity in PDGFRβflox/flox MSCs (DF-B alone, 12.7 ± 1.26 U/mg protein; DF-B + PDGF-AA, 11.4 ± 0.837 U/mg protein; DF-B + PDGF-BB, 5.40 ± 0.513 U/mg protein; p < 0.05 for DF-B alone and DF-B + PDGF-AA versus DF-B + PDGF-BB). Neither PDGF-AA nor PDGF-BB affected the high level of ALP activity detected in PDGFRβdel/del MSCs (DF-B alone, 23.1 ± 1.84 U/mg protein; DF-B + PDGF-AA, 23.1 ± 2.14 U/mg protein; DF-B + PDGF-BB, 19.9 ± 2.26 U/mg protein).

Effects of PDGFRβ depletion on the ALP and OCN mRNA levels

We evaluated the mRNA levels of ALP and OCN in MSCs cultured in osteogenic differentiation medium for the indicated times (Fig. 5). The mRNA levels of both ALP and OCN were significantly increased in PDGFRβdel/del cells compared with PDGFRβflox/flox cells up to day 7 (PDGFRβflox/flox cells versus PDGFRβdel/del cells; ALP: day 1, 1.000 ± 0.262 versus 2.254 ± 0.414; p < 0.05; day 3, 0.182 ± 0.053 versus 1.694 ± 0.509; p < 0.001; day 7, 3.062 ± 0.781 versus 16.03 ± 5.691; p < 0.01; OCN: day 1, 1.000 ± 0.232 versus 3.232 ± 0.474; p < 0.01; day 3, 0.587 ± 0.328 versus 8.436 ± 4.246; p < 0.05; day 7, 4.093 ± 0.715 versus 8.256 ± 1.074; p < 0.001). On day 14, the mRNA levels of both ALP and OCN were higher in PDGFRβdel/del cells than in PDGFRβflox/flox cells, but the differences between the cells were not significant.

Figure Figure 5.

Effects of PDGFRβ gene deletion on the ALP and OCN mRNA levels in MSCs. Quantitative RT-PCR for ALP and OCN was performed using PDGFRβflox/flox (horizontally striped bars) and PDGFRβdel/del (open bars) MSCs at the indicated times after osteogenic differentiation. The mRNA levels in each sample were normalized by the corresponding level of β-actin mRNA. The expression levels in PDGFRβflox/flox cells on day 1 after osteogenic differentiation were set as 1, and the relative expression levels were calculated. Values are means ± SE.

Expression of PDGFRα and β during osteogenic differentiation

We evaluated the mRNA expression levels of PDGFRα and β in MSCs cultured in osteogenic differentiation medium for 7 days (Fig. 6). When treated with PDGF-BB, PDGFRα expression tended to decrease in PDGFRβflox/flox MSCs, but the difference was not significant (PDGFRβflox/flox cells, 1.000 ± 0.183; PDGFRβflox/flox cells + PDGF-BB, 0.545 ± 0.237). In PDGFRβdel/del MSCs, PDGFRα was highly expressed compared with PDGFRβflox/flox MSCs (PDGFRβdel/del cells, 1.578 ± 0.140; PDGFRβflox/flox cells, 1.000 ± 0.183; p < 0.05) and slightly decreased by PDGF-BB stimulation (PDGFRβdel/del cells, 1.578 ± 0.140; PDGFRβdel/del cells + PDGF-BB, 1.189 ± 0.188). PDGFRβ expression in PDGFRβflox/flox MSCs was decreased by PDGF-BB stimulation (PDGFRβflox/flox cells, 1.000 ± 0.026; PDGFRβflox/flox cells + PDGF-BB, 0.609 ± 0.070; p < 0.05).

Figure Figure 6.

Effects of PDGFRβ gene deletion and PDGF-BB stimulation on the mRNA levels of PDGFRs, BMP2, Runx2, and osterix. MSCs were cultured in osteogenic differentiation medium (DF-B; horizontally striped and open bars) or DF-B + 0.8 nM PDGF-BB (dotted and oblique bars) for 7 days. Quantitative RT-PCR for PDGFRα and β, BMP2, Runx2, and osterix was performed using PDGFRβflox/flox (horizontally striped and dotted bars) and PDGFRβdel/del (open and oblique bars) MSCs. The mRNA levels in each sample were normalized by the corresponding level of β-actin mRNA. The expression levels in PDGFRβflox/flox cells cultured in DF-B were set as 1, and the relative expression levels were calculated. Values are means ± SE.

Induction of osteogenic transcription factors in PDGFRβflox/flox and PDGFRβdel/del MSCs

We further studied the mRNA expression levels of BMP2 and the osteogenic transcription factors Runx2 and osterix in MSCs cultured in osteogenic differentiation medium with or without PDGF-BB for 7 days (Fig. 6). Expression of BMP2 was significantly higher in PDGFRβdel/del cells than in PDGFRβflox/flox cells (PDGFRβdel/del cells, 2.380 ± 0.187; PDGFRβflox/flox cells, 1.000 ± 0.259; p < 0.05). Expression of BMP2 in both cell types remained unchanged after PDGF-BB stimulation. Expression of Runx2 was significantly higher in PDGFRβdel/del cells than in PDGFRβflox/flox cells (PDGFRβdel/del cells, 1.463 ± 0.096; PDGFRβflox/flox cells, 1.000 ± 0.109; p < 0.05). PDGF-BB significantly decreased Runx2 expression in PDGFRβflox/flox MSCs (DF-B, 1.000 ± 0.109; DF-B + PDGF-BB, 0.649 ± 0.063; p < 0.05), but did not change Runx2 expression in PDGFRβdel/del MSCs (DF-B, 1.463 ± 0.096; DF-B + PDGF-BB, 1.364 ± 0.089). Expression of osterix was significantly higher in PDGFRβdel/del cells than in PDGFRβflox/flox cells (PDGFRβdel/del cells, 2.500 ± 0.114; PDGFRβflox/flox cells, 1.000 ± 0.200; p < 0.05) PDGF-BB significantly decreased the expression of osterix in both types of MSCs (PDGFRβflox/flox cells: DF-B alone, 1.000 ± 0.200; DF-B + PDGF-BB, 0.134 ± 0.048; p < 0.05; PDGFRβdel/del cells: DF-B alone, 2.500 ± 0.114; DF-B + PDGF-BB, 1.228 ± 0.113; p < 0.05).

In vivo analysis of mesenchymal cell responses to fracture

After administration of tamoxifen, Cre-ER+/−/PDGFRβflox/flox and PDGFRβflox/flox mice grew up in good health without any apparent abnormalities, including the skeletal system. Histological evaluation of the in vivo fracture model was performed at 7 days after the fracture. In PDGFRβflox/flox mice, the fracture showed soft callus formation characterized by proliferation of connective tissues, including cartilage (Fig. 7A). On the other hand, the reparative tissues in PDGFRβdel/del mice formed a large amount of woven bone around the fracture site with lesser amounts of cartilage, resembling a phase of hard callus formation characterized by the formation of woven bone by either intramembranous or endochondral ossification (Fig. 7B). The callus area tended to be smaller in PDGFRβdel/del mice than in PDGFRβflox/flox mice (PDGFRβflox/flox mice, 2.719 ± 0.526 mm2; PDGFRβdel/del mice, 2.346 ± 0.387 mm2; Fig. 7C). The ratio of woven bone to callus was significantly higher in PDGFRβdel/del mice than in PDGFRβflox/flox mice (PDGFRβdel/del mice, 0.409 ± 0.033; PDGFRβflox/flox mice, 0.238 ± 0.074; p < 0.05; Fig. 7D). On the other hand, the ratio of cartilage to callus tended to be higher in PDGFRβflox/flox mice than in PDGFRβdel/del mice (PDGFRβflox/flox mice, 0.192 ± 0.080; PDGFRβdel/del mice, 0.097 ± 0.025), although the difference was not significant (Fig. 7E).

Figure Figure 7.

Effects of PDGFRβ gene deletion on fracture healing. The fractures were made by open osteotomy and fixed with an intramedullary nail. H&E-stained sections of 7-day-old bone fractures in control (A) and PDGFRβ knockout (B) mice are shown. Cartilage and woven bone are indicated by arrows and arrowheads within the regenerated tissue, respectively. The scale bars represent 1 mm. The reparative areas around the fractures (fracture callus) (C), the ratio of woven bone to callus (D), and the ratio of cartilage to callus (E) were measured using digital images and a morphometric analyzer. Values are means ± SE.

DISCUSSION

Several growth factors, including PDGF, comprise the signaling network required to regulate bone formation. This study analyzed the functional relevance of PDGFRβ by specific gene deletion in bone marrow–derived MSCs. The obtained results indicated that PDGFRβ plays a major role in the induction of MSC proliferation and migration and that PDGFRβ, but not PDGFRα, strongly suppresses osteogenic differentiation.

To understand the significance of PDGF signaling in MSCs, we analyzed the expression levels of PDGF ligands and receptors in these cells. RT-PCR detected all four ligands, namely PDGF-A, -B, -C, and -D, at relatively high levels for PDGF-A and -C and low levels for -B and -D. PDGFRα and β transcripts were detected in these cells and were also expressed in MSCs cultured for 7 days in osteogenic differentiation medium. Deletion of the PDGFRβ gene was clearly detected by markedly decreased amplification by RT-PCR, and this deletion did not affect the expression levels of the other transcripts.

PDGF is a major mitogen of mesenchymal cells and also strongly induces the mitogenic and migratory response of MSCs.(30,31) In this study, we showed that the mitogenic activity was, to a large extent, decreased in PDGFRβdel/del MSCs compared with PDGFRβflox/flox MSCs after culture in DMEM supplemented with 20% FBS for 24 and 48 h. Closure of wounds scratched into confluent monolayers of cultured MSCs was strongly suppressed in PDGFRβdel/del MSCs compared with PDGFRβflox/flox MSCs in DMEM supplemented with either 20% FBS or 0.4 nM PDGF-BB. A previous study on osteogenic MC3T3-E1 cells stimulated by PDGF-BB showed that the mitogenic response was dependent on extracellular signal–regulated kinase (Erk) and c-jun N-terminal kinase (JNK), whereas the migratory response was dependent on p38 MAP kinase and JNK.(31) We previously observed that Erk and JNK, but not p38, were substantially activated downstream of PDGFR-β using dermal fibroblasts in which PDGFRβ was selectively depleted.(24) The data obtained in this study indicated that PDGFRβ, but not PDGFRα, is the major pathway for mediating FBS-induced proliferation and FBS- or PDGF-induced migration in MSCs and that JNK and Erk may be involved as important signaling molecules.

To study the role of PDGFRβ in osteogenesis, we evaluated the expression levels of the osteogenic differentiation markers ALP and OCN in PDGFRβ-deleted MSCs during the process of osteogenic differentiation in cell culture. ALP staining was consistently enhanced in PDGFRβ-deleted MSCs throughout the experimental period. Consistent with this, the mRNA expression levels of ALP and OCN were elevated in PDGFRβ-deleted MSCs throughout the experimental period. A previous study found that the ALP and OCN mRNA expression levels were decreased in osteoblastic cells after addition of PDGF-BB to the osteogenic differentiation medium.(14) These two lines of evidence suggest that PDGF signaling is suppressive for the induction of osteogenic differentiation of MSCs. Furthermore, these data suggested that PDGFRβ, but not PDGFRα, may be the main mediator of PDGF-BB–induced suppression of osteogenic differentiation.

In addition, we differentially activated two PDGFRs and analyzed the relevance of each receptor to osteogenic differentiation. PDGF-BB administered to the osteogenic differentiation medium suppressed ALP activity in PDGFRβflox/flox cells but did not affect that in PDGFRβdel/del cells, in which ALP activity was detected by both histochemical and biochemical colorimetric measurements. However, PDGF-AA administration did not affect the ALP activities in both genotypes of MSCs, even though PDGFRα expression was detected in these cells. We conclude that PDGFRβ, but not PDGFRα, is important for suppressing the osteogenic differentiation of MSCs.

BMP2 plays important roles in the differentiation of bone marrow–derived precursor cells into osteogenic lineages.(32) BMP2 induces the expression of Runx2 and osterix, which are both master osteogenic transcription factors, to stimulate the osteoblast marker genes of type I collagen and fibronectin at early stages, and ALP and OCN at later stages, of differentiation.(6) PDGFRα was increased in PDGFRβdel/del MSCs compared with PDGFRβflox/flox MSCs, and PDGFRβ expression remained at a low level in PDGFRβdel/del MSCs at 7 days after osteogenic differentiation induction. Under this condition, the expression levels of BMP2, Runx2, and osterix in MSCs were apparently increased in PDGFRβdel/del MSCs compared with PDGFR-βflox/flox MSCs. PDGF-BB stimulation of PDGFRα on PDGFRβdel/del cells had no effect on the expression levels of BMP2 and Runx2 except for osterix. We conclude that depletion of PDGFRβ in MSCs plays a major role in increasing the expression levels of BMP2, Runx2, and osterix and affects the osteogenic differentiation of MSCs.

The level of BMP2 transcripts was consistently higher in PDGFRβdel/del MSCs, and neither PDGFRα and β activation in PDGFRβflox/flox MSCs nor PDGFRα activation in PDGFRβdel/del MSCs by PDGF-BB affected the expression of BMP2. This indicated that the BMP2 level was dependent on the expression of PDGFRβ in our experiment. It is possible that BMP2 was fully suppressed by the low level of PDGFRβ activation induced by the ligands derived from the FBS supplemented into the medium or derived from the MSCs in this study.

BMP2 is a key molecule for the induction of osterix and Runx2. However, PDGF-BB treatment of MSCs further suppressed the expression levels of osterix and Runx2 without affecting BMP2 expression. BMP2 phosphorylates serine residues located in the C terminus of Smad1 to induce its nuclear translocation, thereby leading to the induction of osteogenic gene expression.(6,33) In contrast, the tyrosine kinase receptors activated by EGF and hepatocyte growth factor inhibit BMP2-induced nuclear translocation of Smad1 through phosphorylation of the region linking its inhibitory and effector domains.(34) When we stimulated MSCs with PDGF-BB, Stat3 phosphorylation was only detected in PDGFRβflox/flox MSCs and not in PDGFRβdel/del MSCs (data not shown). PDGFRβ activates Stat3 and Smad1 in cultured mesangial cells.(35) A possible decrease in the nuclear translocation of PDGF-activated Smad1 may be the mechanism of additional suppression for osterix and Runx2. It is also noteworthy that PDGF-BB stimulation decreased the expression of osterix, but not Runx2, in PDGFRβdel/del MSCs. These two transcription factors are differentially regulated, and PDGFRα may also be involved in the regulation of osteogenic differentiation, although its effects seem to be more subtle than those of PDGFRβ.

This study showed that the roles of PDGFRα and PDGFRβ clearly differ during osteogenic differentiation. A previous comparative proteome analysis between PDGF- and EGF-induced phosphorylated proteins showed that Akt was the major pathway involved in mediating PDGF-specific inhibition of osteogenic differentiation.(9) Similar to Stat activation, PDGFRβ was shown to strongly activate Erk and Akt using PDGFRβdel/del dermal fibroblasts.(24) These three molecules, which are inhibitory for osteogenic differentiation of MSCs, may be responsible for the specific regulatory role of PDGFRβ in MSC differentiation.

The effects of PDGFRβ on the process of bone formation could be bidirectional. The stimulatory effects on MSC proliferation and migration should be beneficial for the recruitment of osteoblast precursors to sites requiring bone formation. On the other hand, PDGFRβ strongly suppressed the differentiation process of MSCs. Therefore, we examined the histological findings of bone fracture in mutant mice with PDGFRβ depletion induced by orally administered tamoxifen. At 7 days after the fracture, the tissue response tended to be smaller, with a significant increase of woven bone content in the mutant mice in the posterior part of the bone. These findings may indicate less recruitment and premature osteogenic differentiation of MSCs caused by PDGFRβ depletion, and this interpretation seems to be compatible with the findings obtained in the in vitro experiments. Overall, it is indispensable to consider the spatiotemporal regulation of PDGF signal transduction depending on the specific circumstances in regions where bone formation is required, when the functional significance and clinical use of PDGFs and their receptors are considered for in vivo bone formation.

In summary, we studied the role of PDGFRβ in the osteogenic differentiation process of primary cultured MSCs, in which PDGFRβ expression was specifically suppressed by gene manipulation for the first time. Depletion of PDGFRβ decreased the mitogenic and migratory responses and enhanced the expression of osteogenic marker proteins and transcription factors in MSCs. It was clearly shown that the two types of PDGFRs were involved in different ways in the differentiation and that PDGFRβ was very important for mediating PDGF signaling to regulate osteogenic differentiation of MSCs. PDGFRβ could represent an important target for guided tissue regeneration or tissue engineering of bone.

Acknowledgements

The authors thank Yo-ichi Kurashige, Takako Matsushima, Sayaka Kobayashi (University of Toyama, Japan), and Taro Uyama (Department of Reproductive Biology, National Institute for Child Health and Development, Japan) for excellent technical assistance and Tetsuya Taga (Department of Cell Fate Modulation, Institute of Molecular Embryology and Genetics, Kumamoto University, Japan) for valuable scientific discussion. This study was supported in part by Grants-in-Aid for Scientific Research (17590338, 16390114, and 16500239) from the Ministry of Education.

Ancillary