- Email: [email protected]
Effect of vascular endothelial growth factor on RANK gene expression in osteoclast precursors and on osteoclastogenesis
Archives of Oral Biology (2006) 51, 596—602
www.intl.elsevierhealth.com/journals/arob
Effect of vascular endothelial growth factor on RANK gene expression in osteoclast precursors and on osteoclastogenesis Shaomian Yao, Dawen Liu, Fenghui Pan, Gary E. Wise * Department of Comparative Biomedical Sciences, School of Veterinary Medicine, Louisiana State University, Baton Rouge, LA 70803, United States Accepted 13 December 2005
KEYWORDS VEGF; CSF-1; RANK; Osteoclastogenesis; Tooth eruption
Summary Objective: The aim of this study was to determine if vascular endothelial growth factor (VEGF) can upregulate the gene expression of receptor activator of nuclear factor kappa B (RANK) in osteoclast precursors, as does CSF-1. A secondary aim was to determine if VEGF can promote osteoclastogenesis in vitro comparable to CSF-1. Design: Osteoclast precursors (mononuclear cells) were incubated with different concentrations of VEGF, CSF-1, or a combination of the two, and the gene expression of RANK was determined by RT-PCR. A TRAP assay also was conducted to determine their effect on osteoclastogenesis. An Alamar blue assay was done to analyse the effect of the molecules on proliferation of the osteoclast precursors. Results: VEGF upregulated RANK expression in osteoclast precursors as effectively as CSF-1. VEGF did not promote osteoclastogenesis, as did CSF-1. A combination of the two did. CSF-1 enhanced proliferation of the osteoclast precursors but VEGF did not. However, VEGF in combination with CSF-1 did increase proliferation. Conclusions: At the time of the secondary burst of osteoclastogenesis prior to tooth eruption, VEGF expression in the dental follicle is high but the expression of CSF-1 is low. This study demonstrates that VEGF can fully substitute for CSF-1 to upregulate the RANK expression in osteoclast precursors that is needed for osteoclastogenesis. However, VEGF alone neither can promote osteoclastogenesis nor stimulate proliferation of the osteoclast precursors in vitro. For proliferation and osteoclastogenesis, a low dose of CSF-1 in combination with VEGF is needed. # 2005 Elsevier Ltd. All rights reserved.
* Corresponding author. Tel.: +1 225 578 9889 (O); fax: +1 225 578 9895. E-mail address: [email protected] (G.E. Wise). 0003–9969/$ — see front matter # 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.archoralbio.2005.12.006
VEGF on RANK gene expression
597
Introduction
Materials and methods
Tooth eruption requires alveolar bone resorption in order for the tooth to exit its bony crypt (see review by1). In the first mandibular molar of the rat, the maximal number of osteoclasts seen on the surrounding alveolar bone is at day 3 postnatally2,3 followed by a subsequent decrease until approximately day 10 when there is a lesser increase of osteoclasts.3 The dental follicle appears to regulate the major burst of osteoclastogenesis by maximally expressing colony-stimulating factor-1 (CSF-1) at day 34. In the dental follicle, CSF-1 promotes osteoclastogenesis by both recruiting mononuclear cells (osteoclast precursors) to the follicle5 and by inhibiting osteoprotegerin expression in the follicle cells.6 Others have shown that CSF-1 also promotes growth and differentiation of mononuclear cells (e.g. see7,8) and that osteoclastogenesis (osteoclast formation) requires the presence of CSF1.9—11 One reason that osteoclastogenesis requires CSF1 is because CSF-1 up-regulates the gene expression of receptor activator of nuclear factor kappa B (RANK) in osteoclast precursors such that RANK is present on the surface of these cells.12 This enables cell-to-cell signalling to occur between the RANKL present on osteoblasts10 or on dental follicle cells13 and the RANK expressed on the osteoclast precursors (mononuclear cells). Such RANK-RANKL signalling activates the osteoclast precursors to form osteoclasts. As previously stated, CSF-1 is maximally expressed in the follicle at day 3 but at the time of the minor burst of osteoclastogenesis at day 10 its expression is greatly reduced.4 Thus, a substitute is needed for CSF-1 to promote this secondary round of osteoclast formation. A likely candidate for replacement of CSF-1 is vascular endothelial growth factor (VEGF), a gene that is maximally expressed in the dental follicle at days 9—11 postnatally.14 It has been reported that VEGF can substitute for CSF1 in vitro to promote osteoclastogenesis in the presence of RANKL15 and, in vivo, can recruit osteoclasts to the site of injection of VEGF in osteopetrotic mice.15,16 Thus, it was the objective of this study to determine at the molecular level if VEGF promotes osteoclastogenesis in the same manner as does CSF-1; i.e., does VEGF upregulate the gene expression of RANK in osteoclast precursors and can VEGF promote the proliferation of the precursors? A secondary aim was to determine if VEGF could substitute for CSF-1 to promote osteoclastogenesis in vitro, as well as determine if the two combined had an additive effect on osteoclast formation.
Purification of mononuclear cells Spleens from 1-month-old male rats were surgically isolated and washed with MEM-a medium containing antibiotics. Each spleen was cut into approximate 0.5 cm long pieces and pressed against a 60 mesh screen with a glass pestle to dissociate the cells. The cell suspension was collected in a centrifuge tube and pelleted by centrifugation for 5 min at 1000 rpm. The cell pellet was resuspended in 5 ml Red Blood Cell Lysing Buffer (Sigma—Aldrich Corp., St. Louis, MO, USA) for 5 min followed by centrifugation. Next, the cells were cultured overnight in MEM-a medium containing 10% heat inactivated FBS and 20 ng/ml human CSF-1 (PeproTech Inc., Rocky Hill, NJ, USA). Non-adherent cells were gently transferred to the top of 10 ml Ficoll Hypaque-76 in a centrifuge tube and then subjected to gradient centrifugation for 15 min at 1000 g. The cells in the medium at the top of gradient above the Ficoll were collected and cultured in MEM-a medium supplemented with 20 ng/ml CSF-1 for 2 days to allow the attachment of the mononuclear cells. Prior to the experiment, purified cells were cultured for 2 days in CSF-1 free medium. To establish bone marrow cell cultures, bone marrow was washed out from marrow of tibias with MEM-a medium and then treated with Red Blood Cell Lysing Buffer. The cells were collected by centrifugation and cultured in MEM-a medium without further purification.
RANK gene expression To study the effect of VEGF on RANK gene expression, cells from above were incubated in MEM-a for 3 h containing one of the following: (1) human CSF-1 at 30 ng/ml, (2) VEGF (R&D Systems Inc., Minneapolis, MN, USA) at 10, 20, 30, or 40 ng/ml; and (3) CSF-1 at 30 ng/ml plus VEGF at 30 ng/ml. The controls were cultured in the same manner but without adding CSF-1 or VEGF. The cells then were collected into TRI REAGENT (Molecular Research Center, Cincinnati, OH, USA), and total RNA was extracted according to the manufacturer’s protocol and treated with DNase I to remove any contaminating DNA. For RT-PCR studies, the RNA concentration was measured at A260. Twenty microlitres of cDNA then was generated from 1 to 2 mg total RNA of each sample. Gene expression was determined by semiquantitative PCR. Each PCR was prepared by mixing 2 ml of cDNA with buffer, dNTP, RANK gene-specific primers, and Taq DNA polymerase at a total volume of 25 ml. The forward RANK specific primer was 50 TCTCAGATGTCTTTTCGTCAACAG-30 and the reverse
598 was 50 -AGCCACCACTACCA CAGAGATG-30 .17 The PCR was conducted for 30 cycles of denaturing at 94 8C for 45 s, annealing at 60 8C for 1 min and extension at 72 8C for 1 min. PCR for actin used as an endogenous control was carried out in parallel under the same conditions except that amplification was 23 cycles. The PCR product was electrophoresed in an agarose gel stained with SYBR green I. The intensity of RANK and actin amplification bands was measured by a Kodak-1D image system. The RANK gene expression value was calculated using the ratio of RANK/ actin. Any given experiment was repeated at least three times.
Real-time PCR For real-time RT-PCR studies, 2 ml of cDNA was mixed with Taqman PCR Master Mix (Applied Biosystems, Foster City, CA, USA), 200 nm FAM-BHQ probe and 400 nm RANK forward primer (50 -CACCCAGGGTAGGCAACT-30 ) and reverse primer (50 -CACCCAGGGTAGGCAACT-30 ) in a 25 ml reaction volume. The bactin gene was amplified in parallel as an internal control using forward primer (50 -GCAGACAAACAATACGCCATTCCC-30 ) and reverse primer (50 -CTAAGACCAACCGTGAAAAGAT-30 ). The reaction was performed with the initial stage at 95 8C for 10 min, followed by 40 cycles at 95 8C for 15 s and 60 8C for 1 min. A CT (threshold cycle) value was generated for each sample, and the comparative CT method was used to calculate the relative gene expression (RGE) per manufacturer’s instructions (Applied Biosystems) by the formula 2DDC . T
Tartrate resistant acid phosphatase (TRAP) assay and staining For osteoclastogenesis studies, cells purified from spleen were incubated in medium containing 30 ng/ ml RANKL (PeproTech, Inc.) with one of the following: (1) CSF-1 at 1, 5, or 30 ng/ml; (2) VEGF at 10, 20, or 30 ng/ml; and (3) CSF-1 at 5 ng/ml plus VEGF at 10, or 20 ng/ml. After 8—10 days of culture with a medium change every 4 days, cells in the flasks were lysed for the TRAP assay and cells on coverslips were fixed for TRAP staining. For the TRAP assay, the cells were lysed with whole cell extraction buffer containing protease inhibitors. Cell debris was removed by a 15 min centrifugation at 13,000 g. The protein concentration was measured with the BCA method (Pierce Biotechnology, Inc. Rockford, IL, USA). The TRAP assay was conducted by mixing 50 mg protein, 0.25 ml tartrate buffer and 0.25 ml phosphatase substrate solution followed by incubation at 37 8C. After 30 min of incubation, the reaction was
S. Yao et al. stopped by adding 2.5 ml 0.1N NaOH. The colour of each reaction was measured at 405 nm. For TRAP staining, cells on coverslips were fixed and stained using the Sigma leukocyte acid phosphatase kit (Sigma—Aldrich Corp.) according to the protocol provided. The numbers of TRAP-positive cells in 1 mm2 grids were counted under a light microscope.
Immunostaining for CD11b Purified mononuclear cells were grown on coverslips and fixed in cold acetone for 2 min. After incubation in 3% normal goat serum, 0.3% Triton X-100 and 1% BSA in PBS for 1 h at room temperature, the coverslips were incubated with mouse anti-rat CD11b/c monoclonal antibody (BD Biosciences Pharmingen, San Diego, CA, USA) at 1:150 dilution overnight at 4 8C. Next, the coverslips were incubated with the biotinylated goat anti-mouse secondary antibody and strepavidin-horseradish peroxidase at room temperature for 30 min. The horseradish peroxidase then was detected with DAB, and the cells were counterstained with hematoxylin QS. For controls, the primary antiserum was omitted or preimmune serum was substituted.
Cell proliferation analysis Purified mononuclear cells were grown in MEM-a containing either CSF-1 at 1, 5, 10, 30 ng/ml or VEGF at 10, 20, 40 ng/ml, as well as combinations of CSF-1 at 5 ng/ml and VEGF at 10 or 20 ng/ml. The controls were cultured in medium without CSF-1 and VEGF. Cell proliferation was tested after 5 days of culture using Alamar blue; i.e., the cultures were placed in medium containing 10% Alamar blue (Biosource International, Camarillo, CA, USA). After 7 h of incubation, 100 ml of the medium was transferred to the wells of a 96-well plate and the OD was measured at wavelengths of 570 and 595 nm. The reduction of Alamar blue was calculated with a formula provided by the manufacturer. The greater the percentage reduction of Alamar blue reflects the greater the cell proliferation.
Data analysis All experiments were performed using a randomised block design with three to five replications. Statistical analysis was conducted with analysis of variance to determine the treatment effects. Difference of the means was separated using least significant difference or student t-tests at a significance level of P 0.05. The data were reported as means standard deviations.
VEGF on RANK gene expression
Figure 1 Effect of VEGF on upregulation of RANK gene expression in mononuclear cells purified from rat spleen (A and B) and bone marrow cultures (C), as determined by RT-PCR (A and B) and real-time RT-PCR (C). Note that RANK gene expression is significantly enhanced (P 0.05) by treatment of VEGF 10 or 20 ng/ml and CSF-1 at 30 ng/ml (A and B, lanes 2—4). A combination of VEGF and CSF-1 has no significant additive effect on RANK expression (A, lane 7), but it does enhance expression over the controls. Realtime RT-PCR indicated that VEGF up-regulates RANK expression at concentrations of 10—30 ng/ml (C).
Results In cultured spleen mononuclear cells (osteoclast precursors), VEGF upregulates RANK gene expression as does CSF-1 (Fig. 1). Maximal effect of VEGF on RANK gene expression was at concentrations of 10—20 ng/ml, which induced a similar level of expression as CSF-1 at 30 ng/ml (Fig. 1). CSF-1 and VEGF combined did not show a significant additive effect on RANK expression at the concentrations tested (Fig. 1A and B). Similar results were observed in bone marrow cells except that concentrations of 10, 20 or 30 ng/ml of VEGF significantly elevated the RANK expression above the control as seen by real-time RT/PCR (Fig. 1C). Untreated bone marrow cells expressed high endogenous levels of RANK (Fig. 1C), which probably masked some of the treatment effect of CSF-1 and VEGF. The purity of the cultures of putative spleen mononuclear cells (monocytes) was analysed by immunostaining for CD11b, a monocyte osteoclast
599 progenitor marker.18 As seen in Fig. 2D, almost all of the purified cells stain for CD11b whereas only a portion of the cells in a non-purified spleen cell population stain (Fig. 2B). In addition to the ability of VEGF to up-regulate RANK gene expression, as does CSF-1, its ability to promote osteoclastogenesis in the presence of RANKL was examined. TRAP staining and counting of cells showed that CSF-1 was significantly more effective than VEGF in inducing TRAP-positive mononuclear cells and osteoclasts but that a combination of CSF-1 and VEGF was the most efficacious for induction of TRAP-positive cells (Table 1, Fig. 3). Quantitative TRAP assays indicated that CSF-1 at 30 ng/ml induced a significantly higher degree of osteoclastogenesis than in the untreated control but low concentrations of CSF-1 at 1 or 5 ng/ml had no significant effect on osteoclastogenesis (Fig. 4A). In contrast, VEGF did not induce a higher degree of osteoclastogenesis than in the controls at all concentrations tested (Fig. 4A). In the experiments combining CSF-1 and VEGF, however, the combination of CSF-1 (5 ng/ml) and VEGF (10 ng/ml) promoted osteoclastogenesis that was significantly higher than either CSF-1 or VEGF alone (Fig. 4B), as also seen by the cell counts (Table 1). Qualitatively, it appeared that mononuclear cells grew more vigorously in cultures containing CSF-1 than in cultures containing no CSF-1. Thus, cell proliferation was analysed using an Alamar blue reduction assay. The results indicated that VEGF did not enhance the growth of mononuclear cells at the concentrations tested (Fig. 5). In contrast, CSF-1 at concentrations of 5 ng/ml or greater resulted in a significantly higher proliferation of the cells than in the untreated controls (Fig. 5). This proliferation increase also occurred when VEGF at 10 or 20 ng/ml was added to the cultures containing CSF-1 at 5 ng/ml (Fig. 5).
Discussion The role of the dental follicle in promoting the major burst of osteoclastogenesis in the alveolar bone of the first mandibular molar of the rat, as well as the mouse, has been reasonably well established (see review by1). However, many of the molecules that are expressed maximally at day 3 in the rat, or at day 5 in the mouse,19 are expressed at a much lower level at day 10, the time of the minor burst of osteoclastogenesis. This includes CSF-1,4 monocyte chemotactic protein-1 (MCP-1),5 and interleukin-1a.20 Recently, however, we have shown that VEGF is maximally expressed in the dental follicle at days 9—1114 and
600
S. Yao et al.
Figure 2 Immunostaining of spleen cell populations with CD11b, a monocyte osteoclast progenitor lineage marker. Both stained (closed arrow), and non-stained (open arrow) cells were seen in the unpurified cell population, as well as some stained osteoclasts (OC) (B). Almost all the cells in the purified population were stained (D). Controls for immunostaining show absence of staining when preimmune serum is substituted for primary antibody in unpurified (A) and purified (C) populations.
others have shown that VEGF can substitute for CSF-1 to promote osteoclastogenesis in vitro.15 This study suggests that the molecular basis for the ability of VEGF to substitute for CSF-1 is that VEGF can upregulate RANK expression in osteoclast precursors just as does CSF-1 (Fig. 1). Moreover, a concentration of VEGF of either 10 or 20 ng/ml enhances RANK expression as much as does a 30 ng/ml concentration of CSF-1 (Fig. 1). This is the first report of the ability of VEGF to upregulate RANK expression in osteoclast precursors but it is indirectly corroborated by a recent study showing that VEGF enhances RANK expression in human endothelial cells.21 As stated earlier, CSF-1 has many roles in osteoclastogenesis ranging from recruiting osteoclast precursors to enhancing RANK expression. Thus, we wanted to determine if VEGF can promote osteoclastogenesis in vitro at the same level as does CSF1. As seen in Fig. 4, it does not. Unlike CSF-1, VEGF appears not to promote the proliferation of the osteoclast precursors (mononuclear cells), as seen in the Alamar blue reduction assay (Fig. 5). Because
CSF-1 supports the growth and differentiation of osteoclast precursors,7,8 the CSF-1 treated cultures would have an increased number of precursor cells and osteoclasts formed (both TRAP-positive) than in the VEGF treated cultures, as is indeed seen in the TRAP-staining counts (Table 1). This lack of effect of VEGF on proliferation may explain why any osteoclasts induced in vitro by VEGF are smaller in size than those induced by CSF-1.15 Table 1 VEGF and CSF-1 effect on the in vitro generation of TRAP+ cellsa Treatment (ng/ml)
Mononuclear cells b
Osteoclast b
Total
Control VEGF10 CSF-1 5 VEGF10 + CSF-1 5
0A 6.4 A 30.3 B 45.0 C
0A 0.6 A 10.3 B 15.8 C
0 7.0 40.6 60.8
a
Based on four separate counts of 1 mm2 grids. Numbers followed by different letters within a row indicate a significant difference at P 0.05 with the LSD test. b
VEGF on RANK gene expression
601
Figure 3 TRAP staining showing that osteoclastogenesis occurs in cultures of mononuclear cells purified from rat spleen following CSF-1 (A) with increased osteoclastogenesis in presence of CSF-1 and VEGF (B).
Because osteoclasts are formed by a fusion of precursor cells, the number of precursors available for fusion at the site of formation of a given osteoclast would be less in the presence of VEGF as compared to CSF-1. It is important to note that a combination of VEGF and a low concentration of CSF-1 does enhance osteoclastogenesis more than either VEGF alone or CSF-1 alone (Fig. 4B). This is of significance with regard to the minor burst of osteoclastogenesis that occurs in vivo around day 10 prior to eruption. At that time, the CSF-1 level is low in the dental
Figure 4 Effect of VEGF on osteoclastogenesis using purified mononuclear cells from spleen as measured by TRAP assay. Note that TRAP was not enhanced by VEGF or CSF-1 at the low concentrations but that CSF-1 significantly enhanced osteoclastogenesis (P 0.05) at a higher concentration of 30 ng/ml (A). A combination of low doses of VEGF and CSF-1 did significantly enhance osteoclastogenesis (P 0.05) above the controls, as well as above VEGF or CSF-1 alone (B).
follicle4 but the VEGF level is elevated.14 Thus, it is likely that the CSF-1 could stimulate a moderate increase in the proliferation of the osteoclast precursors and the VEGF would upregulate RANK expression in them, both effects leading to a minor burst of osteoclastogenesis. It is probable that the lesser amount of osteoclastogenesis seen at day 10 than at day 3 is because the lesser amount of CSF-1 at day 10 would result in fewer mononuclear cells being stimulated to divide. Thus, at day 10 there would be a smaller pool of osteoclast precursors available for osteoclast formation. Finally, the work of Niida et al.15 suggested that VEGF could fully replace CSF-1 to promote osteoclastogenesis in vitro. A possible explanation for this is that mouse bone marrow cells were used for their in vitro assay. In our study, we found that the endogenous levels of RANK in bone marrow cultures were higher than in the spleen cell cultures. Thus, in osteoclastogenesis assays, despite no CSF-1 being added, the endogenous level of RANK in the bone marrow cultures may be sufficient with the VEGF
Figure 5 Comparison of the effects of CSF-1 and VEGF on mononuclear cell proliferation as determined by Alamar blue reduction. Note that CSF-1 at concentrations of 5 ng/ml and higher significantly enhanced proliferation above the untreated controls (P 0.05). VEGF alone did not enhance proliferation but when combined with low doses of CSF-1 did significantly enhance proliferation (P 0.05) over the mononuclear cell control.
602 added to induce osteoclastogenesis. It should also be noted that their analysis of osteoclastogenesis was qualitative; i.e., they simply noted that some osteoclasts were present if VEGF were used instead of CSF-1. We also noted a few osteoclasts in the presence of VEGF only, but not as many as are seen with CSF-1 treatment (Table 1). In conclusion, this study shows that VEGF can fully substitute for CSF-1 in upregulating RANK gene expression in osteoclast precursors but that VEGF by itself does not promote proliferation of the osteoclast precursors and osteoclastogenesis. For proliferation and osteoclastogenesis, a small amount of CSF-1 is required such that the VEGF can work in concert with CSF-1 to enhance these processes.
Acknowledgements Supported by NIH-R01 grant DE08911-13 to GEW. The authors thank Ms. Cindy Daigle for processing this manuscript.
References 1. Wise GE, Frazier-Bowers S, D’Souza RN. Cellular, molecular, and genetic determinants of tooth eruption. Crit Rev Oral Biol Med 2002;13:323—34. 2. Wise GE, Fan W. Changes in the tartrate-resistant acid phosphatase cell population in dental follicles and bony crypts of rat molars during tooth eruption. J Dent Res 1989;68:150—6. 3. Cielinski MJ, Jolie M, Wise GE, Ando DG, Marks Jr SC.. Colonystimulating factor-1 (CSF-1) is a potent stimulator of tooth eruption in the rat. In: Davidovitch Z, editor. The biological mechanisms of tooth eruption, resorption and replacement by implants. Birmingham, AL: EBSCO Media; 1994. p. 429—36. 4. Wise GE, Lin F, Zhao L. Transcription and translation of CSF-1 in the dental follicle. J Dent Res 1995;74:1551—7. 5. Que BG, Wise GE. Colony-stimulating factor-1 and monocyte chemotactic protein-1 chemotaxis for monocytes in the rat dental follicle. Arch Oral Biol 1997;42:855—60. 6. Wise GE, Lumpkin SJ, Huang H, Zhang Q. Osteoprotegerin and osteoclast differentiation factor in tooth eruption. J Dent Res 2000;79:1937—42. 7. Stanley ER, Guilbert LJ, Tushinski R, Bartelmez S. CSF: a mononuclear phagocyte lineage-specific hematopoietic growth factor. J Cell Biochem 1983;21:151—9.
S. Yao et al. 8. Tanaka S, Takahashi N, Udagawa N, Tamura T, Akatsu T, Stanley ER, et al. Macrophage colony-stimulating factor is indispensable for both proliferation and differentiation of osteoclast progenitors. J Clin Invest 1993;91:257—63. 9. Kodama H, Nose M, Niida S, Yamasaki A. Essential role of macrophage colony-stimulating factor in the osteoclast differentiation supported by stromal cells. J Exp Med 1991; 173:1291—4. 10. Yasuda H, Shima N, Nakagawa N, Yamaguchi K, Kinosaki M, Mochizuki SI, et al. Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc Natl Acad Sci USA 1998;95:3597—602. 11. Yasuda H, Shima N, Nakagawa N, Yamaguchi K, Kinosaki M, Goto M, et al. A novel molecular mechanism modulating osteoclast differentiation and function. Bone 1999;25: 109—13. 12. Arai F, Miyamoto T, Ohneda O, Inada T, Sudo T, Brasel K, et al. Commitment and differentiation of osteoclast precursor cells by the sequential expression of c-Fms and receptor activator of nuclear factor kB (RANK) receptors. J Exp Med 1999;190: 1741—54. 13. Yao S, Ring S, Henk WG, Wise GE. In vivo expression of RANKL in the rat dental follicle as determined by laser capture microdissection. Arch Oral Biol 2004;49:451—6. 14. Wise GE, Yao S. Expression of vascular endothelial growth factor in the dental follicle. Crit Rev Eukaryot Gene Exp 2003;13:173—80. 15. Niida S, Kaku M, Amano H, Yoshida H, Kataoka H, Nishikawa S, et al. Vascular endothelial growth factor can substitute for macrophage colony-stimulating factor in the support of osteoclastic bone resorption. J Exp Med 1999;190:293—8. 16. Kaku M, Niida S, Kawata T, Maeda N, Tanne K. Dose- and timedependent changes in osteoclast induction after a single injection of vascular endothelial growth factor in osteopetrotic mice. Biomed Res 2000;21:67—72. 17. Guillonneau C, Louvet C, Renaudin K, Heslan JM, Heslan M, Tesson L, et al. The role of TNF-related activation-induced cytokine-receptor activating NF-kappa B interaction in acute allograft rejection and CD401L-independent chronic allograf refection. J Immunol 2004;172:1619—29. 18. Husheem M, Nyman JKE, Va ¨¨ ara ¨niemi J, Vaananen HK, Hentunen TA. Characterization of circulating human osteoclast progenitors: development of in vitro resorption assay. Calcif Tissue Int 2005;76:222—30. 19. Wise GE, Huang H, Que BG. Gene expression of potential tooth eruption molecules in the dental follicle of the mouse. Eur J Oral Sci 1999;107:482—6. 20. Wise GE, Lin F, Zhao L. Immunolocalization of interleukin-1a in rat mandibular molars and its enhancement after in vivo injection of epidermal growth factor. Cell Tiss Res 1995; 280:21—6. 21. Min J-K, Kim Y-M, Kim Y-M, Kim I-C, Gho YS, Kang I-J, et al. Vascular endothelial growth factor up-regulates expression of receptor activator of NF-kB (RANK) in endothelial cells. J Biol Chem 2003;278:39548—57.
www.intl.elsevierhealth.com/journals/arob
Effect of vascular endothelial growth factor on RANK gene expression in osteoclast precursors and on osteoclastogenesis Shaomian Yao, Dawen Liu, Fenghui Pan, Gary E. Wise * Department of Comparative Biomedical Sciences, School of Veterinary Medicine, Louisiana State University, Baton Rouge, LA 70803, United States Accepted 13 December 2005
KEYWORDS VEGF; CSF-1; RANK; Osteoclastogenesis; Tooth eruption
Summary Objective: The aim of this study was to determine if vascular endothelial growth factor (VEGF) can upregulate the gene expression of receptor activator of nuclear factor kappa B (RANK) in osteoclast precursors, as does CSF-1. A secondary aim was to determine if VEGF can promote osteoclastogenesis in vitro comparable to CSF-1. Design: Osteoclast precursors (mononuclear cells) were incubated with different concentrations of VEGF, CSF-1, or a combination of the two, and the gene expression of RANK was determined by RT-PCR. A TRAP assay also was conducted to determine their effect on osteoclastogenesis. An Alamar blue assay was done to analyse the effect of the molecules on proliferation of the osteoclast precursors. Results: VEGF upregulated RANK expression in osteoclast precursors as effectively as CSF-1. VEGF did not promote osteoclastogenesis, as did CSF-1. A combination of the two did. CSF-1 enhanced proliferation of the osteoclast precursors but VEGF did not. However, VEGF in combination with CSF-1 did increase proliferation. Conclusions: At the time of the secondary burst of osteoclastogenesis prior to tooth eruption, VEGF expression in the dental follicle is high but the expression of CSF-1 is low. This study demonstrates that VEGF can fully substitute for CSF-1 to upregulate the RANK expression in osteoclast precursors that is needed for osteoclastogenesis. However, VEGF alone neither can promote osteoclastogenesis nor stimulate proliferation of the osteoclast precursors in vitro. For proliferation and osteoclastogenesis, a low dose of CSF-1 in combination with VEGF is needed. # 2005 Elsevier Ltd. All rights reserved.
* Corresponding author. Tel.: +1 225 578 9889 (O); fax: +1 225 578 9895. E-mail address: [email protected] (G.E. Wise). 0003–9969/$ — see front matter # 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.archoralbio.2005.12.006
VEGF on RANK gene expression
597
Introduction
Materials and methods
Tooth eruption requires alveolar bone resorption in order for the tooth to exit its bony crypt (see review by1). In the first mandibular molar of the rat, the maximal number of osteoclasts seen on the surrounding alveolar bone is at day 3 postnatally2,3 followed by a subsequent decrease until approximately day 10 when there is a lesser increase of osteoclasts.3 The dental follicle appears to regulate the major burst of osteoclastogenesis by maximally expressing colony-stimulating factor-1 (CSF-1) at day 34. In the dental follicle, CSF-1 promotes osteoclastogenesis by both recruiting mononuclear cells (osteoclast precursors) to the follicle5 and by inhibiting osteoprotegerin expression in the follicle cells.6 Others have shown that CSF-1 also promotes growth and differentiation of mononuclear cells (e.g. see7,8) and that osteoclastogenesis (osteoclast formation) requires the presence of CSF1.9—11 One reason that osteoclastogenesis requires CSF1 is because CSF-1 up-regulates the gene expression of receptor activator of nuclear factor kappa B (RANK) in osteoclast precursors such that RANK is present on the surface of these cells.12 This enables cell-to-cell signalling to occur between the RANKL present on osteoblasts10 or on dental follicle cells13 and the RANK expressed on the osteoclast precursors (mononuclear cells). Such RANK-RANKL signalling activates the osteoclast precursors to form osteoclasts. As previously stated, CSF-1 is maximally expressed in the follicle at day 3 but at the time of the minor burst of osteoclastogenesis at day 10 its expression is greatly reduced.4 Thus, a substitute is needed for CSF-1 to promote this secondary round of osteoclast formation. A likely candidate for replacement of CSF-1 is vascular endothelial growth factor (VEGF), a gene that is maximally expressed in the dental follicle at days 9—11 postnatally.14 It has been reported that VEGF can substitute for CSF1 in vitro to promote osteoclastogenesis in the presence of RANKL15 and, in vivo, can recruit osteoclasts to the site of injection of VEGF in osteopetrotic mice.15,16 Thus, it was the objective of this study to determine at the molecular level if VEGF promotes osteoclastogenesis in the same manner as does CSF-1; i.e., does VEGF upregulate the gene expression of RANK in osteoclast precursors and can VEGF promote the proliferation of the precursors? A secondary aim was to determine if VEGF could substitute for CSF-1 to promote osteoclastogenesis in vitro, as well as determine if the two combined had an additive effect on osteoclast formation.
Purification of mononuclear cells Spleens from 1-month-old male rats were surgically isolated and washed with MEM-a medium containing antibiotics. Each spleen was cut into approximate 0.5 cm long pieces and pressed against a 60 mesh screen with a glass pestle to dissociate the cells. The cell suspension was collected in a centrifuge tube and pelleted by centrifugation for 5 min at 1000 rpm. The cell pellet was resuspended in 5 ml Red Blood Cell Lysing Buffer (Sigma—Aldrich Corp., St. Louis, MO, USA) for 5 min followed by centrifugation. Next, the cells were cultured overnight in MEM-a medium containing 10% heat inactivated FBS and 20 ng/ml human CSF-1 (PeproTech Inc., Rocky Hill, NJ, USA). Non-adherent cells were gently transferred to the top of 10 ml Ficoll Hypaque-76 in a centrifuge tube and then subjected to gradient centrifugation for 15 min at 1000 g. The cells in the medium at the top of gradient above the Ficoll were collected and cultured in MEM-a medium supplemented with 20 ng/ml CSF-1 for 2 days to allow the attachment of the mononuclear cells. Prior to the experiment, purified cells were cultured for 2 days in CSF-1 free medium. To establish bone marrow cell cultures, bone marrow was washed out from marrow of tibias with MEM-a medium and then treated with Red Blood Cell Lysing Buffer. The cells were collected by centrifugation and cultured in MEM-a medium without further purification.
RANK gene expression To study the effect of VEGF on RANK gene expression, cells from above were incubated in MEM-a for 3 h containing one of the following: (1) human CSF-1 at 30 ng/ml, (2) VEGF (R&D Systems Inc., Minneapolis, MN, USA) at 10, 20, 30, or 40 ng/ml; and (3) CSF-1 at 30 ng/ml plus VEGF at 30 ng/ml. The controls were cultured in the same manner but without adding CSF-1 or VEGF. The cells then were collected into TRI REAGENT (Molecular Research Center, Cincinnati, OH, USA), and total RNA was extracted according to the manufacturer’s protocol and treated with DNase I to remove any contaminating DNA. For RT-PCR studies, the RNA concentration was measured at A260. Twenty microlitres of cDNA then was generated from 1 to 2 mg total RNA of each sample. Gene expression was determined by semiquantitative PCR. Each PCR was prepared by mixing 2 ml of cDNA with buffer, dNTP, RANK gene-specific primers, and Taq DNA polymerase at a total volume of 25 ml. The forward RANK specific primer was 50 TCTCAGATGTCTTTTCGTCAACAG-30 and the reverse
598 was 50 -AGCCACCACTACCA CAGAGATG-30 .17 The PCR was conducted for 30 cycles of denaturing at 94 8C for 45 s, annealing at 60 8C for 1 min and extension at 72 8C for 1 min. PCR for actin used as an endogenous control was carried out in parallel under the same conditions except that amplification was 23 cycles. The PCR product was electrophoresed in an agarose gel stained with SYBR green I. The intensity of RANK and actin amplification bands was measured by a Kodak-1D image system. The RANK gene expression value was calculated using the ratio of RANK/ actin. Any given experiment was repeated at least three times.
Real-time PCR For real-time RT-PCR studies, 2 ml of cDNA was mixed with Taqman PCR Master Mix (Applied Biosystems, Foster City, CA, USA), 200 nm FAM-BHQ probe and 400 nm RANK forward primer (50 -CACCCAGGGTAGGCAACT-30 ) and reverse primer (50 -CACCCAGGGTAGGCAACT-30 ) in a 25 ml reaction volume. The bactin gene was amplified in parallel as an internal control using forward primer (50 -GCAGACAAACAATACGCCATTCCC-30 ) and reverse primer (50 -CTAAGACCAACCGTGAAAAGAT-30 ). The reaction was performed with the initial stage at 95 8C for 10 min, followed by 40 cycles at 95 8C for 15 s and 60 8C for 1 min. A CT (threshold cycle) value was generated for each sample, and the comparative CT method was used to calculate the relative gene expression (RGE) per manufacturer’s instructions (Applied Biosystems) by the formula 2DDC . T
Tartrate resistant acid phosphatase (TRAP) assay and staining For osteoclastogenesis studies, cells purified from spleen were incubated in medium containing 30 ng/ ml RANKL (PeproTech, Inc.) with one of the following: (1) CSF-1 at 1, 5, or 30 ng/ml; (2) VEGF at 10, 20, or 30 ng/ml; and (3) CSF-1 at 5 ng/ml plus VEGF at 10, or 20 ng/ml. After 8—10 days of culture with a medium change every 4 days, cells in the flasks were lysed for the TRAP assay and cells on coverslips were fixed for TRAP staining. For the TRAP assay, the cells were lysed with whole cell extraction buffer containing protease inhibitors. Cell debris was removed by a 15 min centrifugation at 13,000 g. The protein concentration was measured with the BCA method (Pierce Biotechnology, Inc. Rockford, IL, USA). The TRAP assay was conducted by mixing 50 mg protein, 0.25 ml tartrate buffer and 0.25 ml phosphatase substrate solution followed by incubation at 37 8C. After 30 min of incubation, the reaction was
S. Yao et al. stopped by adding 2.5 ml 0.1N NaOH. The colour of each reaction was measured at 405 nm. For TRAP staining, cells on coverslips were fixed and stained using the Sigma leukocyte acid phosphatase kit (Sigma—Aldrich Corp.) according to the protocol provided. The numbers of TRAP-positive cells in 1 mm2 grids were counted under a light microscope.
Immunostaining for CD11b Purified mononuclear cells were grown on coverslips and fixed in cold acetone for 2 min. After incubation in 3% normal goat serum, 0.3% Triton X-100 and 1% BSA in PBS for 1 h at room temperature, the coverslips were incubated with mouse anti-rat CD11b/c monoclonal antibody (BD Biosciences Pharmingen, San Diego, CA, USA) at 1:150 dilution overnight at 4 8C. Next, the coverslips were incubated with the biotinylated goat anti-mouse secondary antibody and strepavidin-horseradish peroxidase at room temperature for 30 min. The horseradish peroxidase then was detected with DAB, and the cells were counterstained with hematoxylin QS. For controls, the primary antiserum was omitted or preimmune serum was substituted.
Cell proliferation analysis Purified mononuclear cells were grown in MEM-a containing either CSF-1 at 1, 5, 10, 30 ng/ml or VEGF at 10, 20, 40 ng/ml, as well as combinations of CSF-1 at 5 ng/ml and VEGF at 10 or 20 ng/ml. The controls were cultured in medium without CSF-1 and VEGF. Cell proliferation was tested after 5 days of culture using Alamar blue; i.e., the cultures were placed in medium containing 10% Alamar blue (Biosource International, Camarillo, CA, USA). After 7 h of incubation, 100 ml of the medium was transferred to the wells of a 96-well plate and the OD was measured at wavelengths of 570 and 595 nm. The reduction of Alamar blue was calculated with a formula provided by the manufacturer. The greater the percentage reduction of Alamar blue reflects the greater the cell proliferation.
Data analysis All experiments were performed using a randomised block design with three to five replications. Statistical analysis was conducted with analysis of variance to determine the treatment effects. Difference of the means was separated using least significant difference or student t-tests at a significance level of P 0.05. The data were reported as means standard deviations.
VEGF on RANK gene expression
Figure 1 Effect of VEGF on upregulation of RANK gene expression in mononuclear cells purified from rat spleen (A and B) and bone marrow cultures (C), as determined by RT-PCR (A and B) and real-time RT-PCR (C). Note that RANK gene expression is significantly enhanced (P 0.05) by treatment of VEGF 10 or 20 ng/ml and CSF-1 at 30 ng/ml (A and B, lanes 2—4). A combination of VEGF and CSF-1 has no significant additive effect on RANK expression (A, lane 7), but it does enhance expression over the controls. Realtime RT-PCR indicated that VEGF up-regulates RANK expression at concentrations of 10—30 ng/ml (C).
Results In cultured spleen mononuclear cells (osteoclast precursors), VEGF upregulates RANK gene expression as does CSF-1 (Fig. 1). Maximal effect of VEGF on RANK gene expression was at concentrations of 10—20 ng/ml, which induced a similar level of expression as CSF-1 at 30 ng/ml (Fig. 1). CSF-1 and VEGF combined did not show a significant additive effect on RANK expression at the concentrations tested (Fig. 1A and B). Similar results were observed in bone marrow cells except that concentrations of 10, 20 or 30 ng/ml of VEGF significantly elevated the RANK expression above the control as seen by real-time RT/PCR (Fig. 1C). Untreated bone marrow cells expressed high endogenous levels of RANK (Fig. 1C), which probably masked some of the treatment effect of CSF-1 and VEGF. The purity of the cultures of putative spleen mononuclear cells (monocytes) was analysed by immunostaining for CD11b, a monocyte osteoclast
599 progenitor marker.18 As seen in Fig. 2D, almost all of the purified cells stain for CD11b whereas only a portion of the cells in a non-purified spleen cell population stain (Fig. 2B). In addition to the ability of VEGF to up-regulate RANK gene expression, as does CSF-1, its ability to promote osteoclastogenesis in the presence of RANKL was examined. TRAP staining and counting of cells showed that CSF-1 was significantly more effective than VEGF in inducing TRAP-positive mononuclear cells and osteoclasts but that a combination of CSF-1 and VEGF was the most efficacious for induction of TRAP-positive cells (Table 1, Fig. 3). Quantitative TRAP assays indicated that CSF-1 at 30 ng/ml induced a significantly higher degree of osteoclastogenesis than in the untreated control but low concentrations of CSF-1 at 1 or 5 ng/ml had no significant effect on osteoclastogenesis (Fig. 4A). In contrast, VEGF did not induce a higher degree of osteoclastogenesis than in the controls at all concentrations tested (Fig. 4A). In the experiments combining CSF-1 and VEGF, however, the combination of CSF-1 (5 ng/ml) and VEGF (10 ng/ml) promoted osteoclastogenesis that was significantly higher than either CSF-1 or VEGF alone (Fig. 4B), as also seen by the cell counts (Table 1). Qualitatively, it appeared that mononuclear cells grew more vigorously in cultures containing CSF-1 than in cultures containing no CSF-1. Thus, cell proliferation was analysed using an Alamar blue reduction assay. The results indicated that VEGF did not enhance the growth of mononuclear cells at the concentrations tested (Fig. 5). In contrast, CSF-1 at concentrations of 5 ng/ml or greater resulted in a significantly higher proliferation of the cells than in the untreated controls (Fig. 5). This proliferation increase also occurred when VEGF at 10 or 20 ng/ml was added to the cultures containing CSF-1 at 5 ng/ml (Fig. 5).
Discussion The role of the dental follicle in promoting the major burst of osteoclastogenesis in the alveolar bone of the first mandibular molar of the rat, as well as the mouse, has been reasonably well established (see review by1). However, many of the molecules that are expressed maximally at day 3 in the rat, or at day 5 in the mouse,19 are expressed at a much lower level at day 10, the time of the minor burst of osteoclastogenesis. This includes CSF-1,4 monocyte chemotactic protein-1 (MCP-1),5 and interleukin-1a.20 Recently, however, we have shown that VEGF is maximally expressed in the dental follicle at days 9—1114 and
600
S. Yao et al.
Figure 2 Immunostaining of spleen cell populations with CD11b, a monocyte osteoclast progenitor lineage marker. Both stained (closed arrow), and non-stained (open arrow) cells were seen in the unpurified cell population, as well as some stained osteoclasts (OC) (B). Almost all the cells in the purified population were stained (D). Controls for immunostaining show absence of staining when preimmune serum is substituted for primary antibody in unpurified (A) and purified (C) populations.
others have shown that VEGF can substitute for CSF-1 to promote osteoclastogenesis in vitro.15 This study suggests that the molecular basis for the ability of VEGF to substitute for CSF-1 is that VEGF can upregulate RANK expression in osteoclast precursors just as does CSF-1 (Fig. 1). Moreover, a concentration of VEGF of either 10 or 20 ng/ml enhances RANK expression as much as does a 30 ng/ml concentration of CSF-1 (Fig. 1). This is the first report of the ability of VEGF to upregulate RANK expression in osteoclast precursors but it is indirectly corroborated by a recent study showing that VEGF enhances RANK expression in human endothelial cells.21 As stated earlier, CSF-1 has many roles in osteoclastogenesis ranging from recruiting osteoclast precursors to enhancing RANK expression. Thus, we wanted to determine if VEGF can promote osteoclastogenesis in vitro at the same level as does CSF1. As seen in Fig. 4, it does not. Unlike CSF-1, VEGF appears not to promote the proliferation of the osteoclast precursors (mononuclear cells), as seen in the Alamar blue reduction assay (Fig. 5). Because
CSF-1 supports the growth and differentiation of osteoclast precursors,7,8 the CSF-1 treated cultures would have an increased number of precursor cells and osteoclasts formed (both TRAP-positive) than in the VEGF treated cultures, as is indeed seen in the TRAP-staining counts (Table 1). This lack of effect of VEGF on proliferation may explain why any osteoclasts induced in vitro by VEGF are smaller in size than those induced by CSF-1.15 Table 1 VEGF and CSF-1 effect on the in vitro generation of TRAP+ cellsa Treatment (ng/ml)
Mononuclear cells b
Osteoclast b
Total
Control VEGF10 CSF-1 5 VEGF10 + CSF-1 5
0A 6.4 A 30.3 B 45.0 C
0A 0.6 A 10.3 B 15.8 C
0 7.0 40.6 60.8
a
Based on four separate counts of 1 mm2 grids. Numbers followed by different letters within a row indicate a significant difference at P 0.05 with the LSD test. b
VEGF on RANK gene expression
601
Figure 3 TRAP staining showing that osteoclastogenesis occurs in cultures of mononuclear cells purified from rat spleen following CSF-1 (A) with increased osteoclastogenesis in presence of CSF-1 and VEGF (B).
Because osteoclasts are formed by a fusion of precursor cells, the number of precursors available for fusion at the site of formation of a given osteoclast would be less in the presence of VEGF as compared to CSF-1. It is important to note that a combination of VEGF and a low concentration of CSF-1 does enhance osteoclastogenesis more than either VEGF alone or CSF-1 alone (Fig. 4B). This is of significance with regard to the minor burst of osteoclastogenesis that occurs in vivo around day 10 prior to eruption. At that time, the CSF-1 level is low in the dental
Figure 4 Effect of VEGF on osteoclastogenesis using purified mononuclear cells from spleen as measured by TRAP assay. Note that TRAP was not enhanced by VEGF or CSF-1 at the low concentrations but that CSF-1 significantly enhanced osteoclastogenesis (P 0.05) at a higher concentration of 30 ng/ml (A). A combination of low doses of VEGF and CSF-1 did significantly enhance osteoclastogenesis (P 0.05) above the controls, as well as above VEGF or CSF-1 alone (B).
follicle4 but the VEGF level is elevated.14 Thus, it is likely that the CSF-1 could stimulate a moderate increase in the proliferation of the osteoclast precursors and the VEGF would upregulate RANK expression in them, both effects leading to a minor burst of osteoclastogenesis. It is probable that the lesser amount of osteoclastogenesis seen at day 10 than at day 3 is because the lesser amount of CSF-1 at day 10 would result in fewer mononuclear cells being stimulated to divide. Thus, at day 10 there would be a smaller pool of osteoclast precursors available for osteoclast formation. Finally, the work of Niida et al.15 suggested that VEGF could fully replace CSF-1 to promote osteoclastogenesis in vitro. A possible explanation for this is that mouse bone marrow cells were used for their in vitro assay. In our study, we found that the endogenous levels of RANK in bone marrow cultures were higher than in the spleen cell cultures. Thus, in osteoclastogenesis assays, despite no CSF-1 being added, the endogenous level of RANK in the bone marrow cultures may be sufficient with the VEGF
Figure 5 Comparison of the effects of CSF-1 and VEGF on mononuclear cell proliferation as determined by Alamar blue reduction. Note that CSF-1 at concentrations of 5 ng/ml and higher significantly enhanced proliferation above the untreated controls (P 0.05). VEGF alone did not enhance proliferation but when combined with low doses of CSF-1 did significantly enhance proliferation (P 0.05) over the mononuclear cell control.
602 added to induce osteoclastogenesis. It should also be noted that their analysis of osteoclastogenesis was qualitative; i.e., they simply noted that some osteoclasts were present if VEGF were used instead of CSF-1. We also noted a few osteoclasts in the presence of VEGF only, but not as many as are seen with CSF-1 treatment (Table 1). In conclusion, this study shows that VEGF can fully substitute for CSF-1 in upregulating RANK gene expression in osteoclast precursors but that VEGF by itself does not promote proliferation of the osteoclast precursors and osteoclastogenesis. For proliferation and osteoclastogenesis, a small amount of CSF-1 is required such that the VEGF can work in concert with CSF-1 to enhance these processes.
Acknowledgements Supported by NIH-R01 grant DE08911-13 to GEW. The authors thank Ms. Cindy Daigle for processing this manuscript.
References 1. Wise GE, Frazier-Bowers S, D’Souza RN. Cellular, molecular, and genetic determinants of tooth eruption. Crit Rev Oral Biol Med 2002;13:323—34. 2. Wise GE, Fan W. Changes in the tartrate-resistant acid phosphatase cell population in dental follicles and bony crypts of rat molars during tooth eruption. J Dent Res 1989;68:150—6. 3. Cielinski MJ, Jolie M, Wise GE, Ando DG, Marks Jr SC.. Colonystimulating factor-1 (CSF-1) is a potent stimulator of tooth eruption in the rat. In: Davidovitch Z, editor. The biological mechanisms of tooth eruption, resorption and replacement by implants. Birmingham, AL: EBSCO Media; 1994. p. 429—36. 4. Wise GE, Lin F, Zhao L. Transcription and translation of CSF-1 in the dental follicle. J Dent Res 1995;74:1551—7. 5. Que BG, Wise GE. Colony-stimulating factor-1 and monocyte chemotactic protein-1 chemotaxis for monocytes in the rat dental follicle. Arch Oral Biol 1997;42:855—60. 6. Wise GE, Lumpkin SJ, Huang H, Zhang Q. Osteoprotegerin and osteoclast differentiation factor in tooth eruption. J Dent Res 2000;79:1937—42. 7. Stanley ER, Guilbert LJ, Tushinski R, Bartelmez S. CSF: a mononuclear phagocyte lineage-specific hematopoietic growth factor. J Cell Biochem 1983;21:151—9.
S. Yao et al. 8. Tanaka S, Takahashi N, Udagawa N, Tamura T, Akatsu T, Stanley ER, et al. Macrophage colony-stimulating factor is indispensable for both proliferation and differentiation of osteoclast progenitors. J Clin Invest 1993;91:257—63. 9. Kodama H, Nose M, Niida S, Yamasaki A. Essential role of macrophage colony-stimulating factor in the osteoclast differentiation supported by stromal cells. J Exp Med 1991; 173:1291—4. 10. Yasuda H, Shima N, Nakagawa N, Yamaguchi K, Kinosaki M, Mochizuki SI, et al. Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc Natl Acad Sci USA 1998;95:3597—602. 11. Yasuda H, Shima N, Nakagawa N, Yamaguchi K, Kinosaki M, Goto M, et al. A novel molecular mechanism modulating osteoclast differentiation and function. Bone 1999;25: 109—13. 12. Arai F, Miyamoto T, Ohneda O, Inada T, Sudo T, Brasel K, et al. Commitment and differentiation of osteoclast precursor cells by the sequential expression of c-Fms and receptor activator of nuclear factor kB (RANK) receptors. J Exp Med 1999;190: 1741—54. 13. Yao S, Ring S, Henk WG, Wise GE. In vivo expression of RANKL in the rat dental follicle as determined by laser capture microdissection. Arch Oral Biol 2004;49:451—6. 14. Wise GE, Yao S. Expression of vascular endothelial growth factor in the dental follicle. Crit Rev Eukaryot Gene Exp 2003;13:173—80. 15. Niida S, Kaku M, Amano H, Yoshida H, Kataoka H, Nishikawa S, et al. Vascular endothelial growth factor can substitute for macrophage colony-stimulating factor in the support of osteoclastic bone resorption. J Exp Med 1999;190:293—8. 16. Kaku M, Niida S, Kawata T, Maeda N, Tanne K. Dose- and timedependent changes in osteoclast induction after a single injection of vascular endothelial growth factor in osteopetrotic mice. Biomed Res 2000;21:67—72. 17. Guillonneau C, Louvet C, Renaudin K, Heslan JM, Heslan M, Tesson L, et al. The role of TNF-related activation-induced cytokine-receptor activating NF-kappa B interaction in acute allograft rejection and CD401L-independent chronic allograf refection. J Immunol 2004;172:1619—29. 18. Husheem M, Nyman JKE, Va ¨¨ ara ¨niemi J, Vaananen HK, Hentunen TA. Characterization of circulating human osteoclast progenitors: development of in vitro resorption assay. Calcif Tissue Int 2005;76:222—30. 19. Wise GE, Huang H, Que BG. Gene expression of potential tooth eruption molecules in the dental follicle of the mouse. Eur J Oral Sci 1999;107:482—6. 20. Wise GE, Lin F, Zhao L. Immunolocalization of interleukin-1a in rat mandibular molars and its enhancement after in vivo injection of epidermal growth factor. Cell Tiss Res 1995; 280:21—6. 21. Min J-K, Kim Y-M, Kim Y-M, Kim I-C, Gho YS, Kang I-J, et al. Vascular endothelial growth factor up-regulates expression of receptor activator of NF-kB (RANK) in endothelial cells. J Biol Chem 2003;278:39548—57.