BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS 244, 724–727 (1998)ARTICLE NO. RC988318

Evidence for a 1a,25-Dihydroxyvitamin D3 Receptor/Binding Protein in a Membrane Fraction Isolatedfrom a Chick Tibial Fracture-Healing Callus

Akira Kato,*,1 June E. Bishop,* and Anthony W. Norman*,†,2

*Department of Biochemistry and †Division of Biomedical Sciences, University of California, Riverside,Riverside, California 92521

transduction systems (10). Thus the VDRnuc is knownPrevious biological studies have implicated two vita- to be involved in regulation of gene transcription (11).

min D metabolites, 1a,25(OH)2-vitamin D3[1a,25(OH)2- In contrast, the VDRmem is postulated to be involved inD3] and 24R,25(OH)2-vitamin D3 [24R,25(OH)2D3] in the the initiation of rapid, nongenomic biologic responsesprocess of skeletal fracture-healing. While a nuclear that are associated with the opening of voltage-gatedreceptor for 1a,25(OH)2D3 is known to be present in

Ca2/ channels (6), chloride channels (12) in ROS 17/osteoblast and absent in osteoclast cell lines, no sys-2.8 cells, and in the chick intestine with the rapid stim-tematic study has been carried out on the callus tissueulation of Ca2/ transport, termed transcaltachia (13).which is formed during fracture-healing. The presentWith regard to a receptor for 24R,25(OH)2D3, althoughreport shows that a binding protein/receptor forone preliminary report appeared earlier (14), no con-1a,25(OH)2D3 resides both in a postnuclear membranevincing evidence has been subsequently presented de-fraction and in a high speed cytosol fraction of thescribing a nuclear receptor. Recently, preliminary evi-callus tissue obtained 10 days after imposition of adence for the existence of a 24R,25(OH)2D3 membranetibial fracture. The dissociation constant, KD, for

1a,25(OH)2D3 was 0.83 { 0.34 M and 0.66 { 0.38 nM re- receptor in the chick tibial fracture-healing callus hasspectively, for the membrane and cytosol fractions. Re- been described (15, 16). Also, the gene knock-out of thesults from a panel of steroid competition assays indi- 25(OH)D3-24-hydroxylase, a kidney enzyme requiredcate that both receptor/binding proteins greatly prefer for the production of 24R,25(OH)2D3, resulted in the1a-hydroxylated ligands as compared to 1a-deoxy or generation of a phenotype characterized by an abnor-24-hydroxylated ligands. The presence of 1a,25(OH)2D3 mality in the cartilage cell process of intramembra-receptors in the fracture-healing callus is consistent neous calcification (17).with the known biological effects of the metabolite on A useful model for studying the combined biologicalthe fracture-healing process. q 1998 Academic Press actions of 1a,25(OH)2D3 and 24R,25(OH)2D3 is the

chick tibial fracture-healing system (18). Imposition ofa fracture initiates a two-week complex ontogeny ofcellular and biochemical processes which are necessaryA host of recent studies on vitamin D3 suggests thatto fully repair the fracture. Both 1a,25(OH)2D3 (19; 20)many of its biological responses are mediated princi-and 24R,25(OH)2D3 (2) have been shown to generatepally by its daughter metabolite 1a,25(OH)2D3 (1), butspecific biological effects. The implication of the biologi-that in some selected tissues, including bone (2) and

cartilage (3; 4), it is believed to be the combined actions cal effects is that there are receptors present for bothboth of 1a,25(OH)2D3 and 24R,25(OH)2D3 that are nec- dihydroxylated metabolites.essary. The nuclear receptor, VDRnuc , for 1a,25(OH)2D3 The present study was carried out to investigate theis widely distributed in over 30 tissues (1); in addition possible existence of a 1a,25(OH)2D3 receptor/bindingpreliminary evidence has been presented for the exis- protein in chick fracture-healing callus cell mem-tence of a membrane receptor VDRmem in chick intesti- branes. Our results indicate that there exist in chicknal cells (5), ROS 17/2.8 osteoblast cells (6; 7), and in tibial callus, both a highly specific membrane bindingNB4 human leukemia cells (8; 9). The VDRnuc and moiety and a cytosolic binding protein for 1a,25-VDRmem are believed to be coupled to different signal (OH)2D3 that are both distinct from the 24R,25(OH)2D3

receptor/binding protein present in a membrane frac-tion of callus. Thus, in fracture healing, we propose

1 Dr. Kato is a Visiting Scientist from the Kureha Chemical Indus- the direct involvement of 1a,25(OH)2D3 generating itstry Company, Tokyo, Japan.biological functions in a manner distinct from the2 Dr. Norman is the corresponding author. Fax: (909) 787-4784;

E-mail: norman@ucrac1.ucr.ed. signal transduction pathway of 24R,25(OH)2D3. These

0006-291X/98 $25.00Copyright q 1998 by Academic PressAll rights of reproduction in any form reserved.

724

AID BBRC 8318 / 694d$$$321 03-11-98 10:59:05 bbrcg AP: BBRC

Vol. 244, No. 3, 1998 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

TABLE 1results suggest that both 1a,25(OH)2D3 and 24R,25-(OH)2D3 may activate signal transduction pathways re- Saturation Analysis of 1a,25(OH)2D3 Binding in Chick

Tibial Fracture-Healing Callus Subcellular Fractionslated to biological responses involved in the fracture-healing process.

Fraction KD [nM] Bmax [fmol/mg]

MATERIALS AND METHODS Nuclear not saturable not saturableMembrane 0.83 { 0.35 35.5 { 5.28

Animals. Male White Leghorn chicks were obtained at hatching Cytosol 0.66 { 0.37 9.8 { 1.40(Hyline International, Lakeview, CA) and were fed a vitamin D3-replete diet (2000 IU vitamin D3/kg diet, 0.6% Ca, 0.4% P) (21) contin- Note. Saturation analysis was performed as described in the Mate-uously. When chicks were three weeks old, a tibial fracture was rial and Methods section. The KD and Bmax were determined by aimposed as described below [approved by University of California one-site binding hyperbola analysis with a commercially availableChancellor’s Committee on Laboratory Animal Care]. The chick was computer software (GraphPad, San Diego, CA). Values representanesthetized with Metofane (methoyxflurene: Pitman-Moore, IL) in- mean { SEM from at least 3 separate experiments.halation and a skin incision (Ç1 cm) was made over the right tibia.After separation of overlying muscle tissues, a hole (Ç2 mm in diame-ter) was drilled with a dental burr in the diaphysis. The skin was

fractions were carried out; Table 1 summarizes the re-then closed, sutured, and a complete fracture was made by applyinglight manual pressure. No fixation of the fracture was made. Ten sults. The binding of [3H]-1a,25(OH)2D3 observed in thedays after imposition of the fracture, the callus tissue was dissected nuclear fraction was not saturable. Saturable bindingand utilized for various studies (15). was observed in the callus membrane and the cytosol

Callus tissue preparation and fractionation. Immediately after fractions where the KD/Bmax values were 0.83 { 0.35the dissection of the callus tissue, 1:10 homogenates in TED (10 mM nM/35.8 { 5.28 fmol/mg and 0.66 { 0.38 nM/9.8 { 1.4Tris, 1.5 nM EDTA, 1 mM dithiothreitol at pH 7.4) were prepared

fmol/mg, respectively. One major concern relative tousing a Polytron homogenizer (Brinkmann Instruments, Westburg,the saturable 1a,25(OH)2D3 binding moiety observedNY) briefly. The homogenate was centrifuged at 500 1 g for 10 min

to obtain a nuclear pellet. This nuclear pellet was washed twice with in the cytosol fraction is the possible contaminationTED and then extracted with 0.3 M KCI/TED by homogenizing to by the ubiquitous plasma vitamin D binding proteinrelease chromatin-bound nuclear receptors and further centrifuged (DBP). Ten days after imposition of the tibial fracture,at 203,900 1 g for 30 min. The resulting supernatant was used as

the callus is highly vascularized, and inevitably therea source of extracted nuclear proteins. The post-nuclear supernatantis the potential for the presence of significant amountswas centrifuged at 203,900 1 g for 30 min. The resulting pellet was

washed twice with TED by homogenizing and then centrifuging at of DBP, particularly in the cytosol fraction. Saturation203,900 1 g for 30 min. After the washes, the final postnuclear pellet analysis of 1a,25(OH)2D3 with authentic DBP was car-was resuspended in TED and used as a source of nonnuclear cellular ried out for comparison using 1:100 diluted chick serummembranes. The first 203,900 1 g supernatant from the postnuclear

as a DBP source. The binding activity of 1a,25(OH)2D3supernatant was used as a cytosol fraction. All preparation proce-for the DBP at this dilution was quite low (total bind-dures were conducted at 47C.ing/non specific bindingõ1.4). Even when higher chickSaturation binding analysis of callus cell fractions. Saturationserum concentrations were evaluated, the weak bind-binding analysis was carried out over the range of 0.0625-5 nM of

[3H]-1a,25(OH)2D3 (98.0 Ci/mmol) in the presence or absence of 200- ing activity was not improved (data not shown). There-fold excess of nonradioactive 1a,25(OH)2D3 (22). fore, we conclude that the saturable 1a,25(OH)2D3The relative ability of an analog to compete with [3H]-1a,25- binder in the callus cytosol fractions is not attributable(OH)2D3 was measured by determination of the analog’s Relative

to DBP.Competitive Index (RCI) value (23). RCI studies were performed aspreviously reported; in this assay, increasing amounts of nonradioac- We have recently reported the presence of a 24R,25-tive 1a,25(OH)2D3, 24R,25(OH)2D3, 24S,25(OH)2D3 and 25(OH)D3 (OH)2D3 receptor/binding protein in the chick calluswere incubated with a saturating concentration of [3H]-1a,25(OH)2D3 membrane fraction (15); therefore, we compared the(1 nM). The reciprocal values of the percentage of maximally bound binding of [3H]-1a,25(OH)2D3 and [3H]-24R,25(OH)2D3analogs were calculated and plotted as a function of the relative

in this fraction. Presented in Fig. 1, saturation analysisconcentration of the analog and [3H]-1a,25(OH)2D3. The slopes ofthe analog binding plots were then normalized to a standard curve showed a higher binding affinity of 1a,25(OH)2D3 (KDobtained with nonradioactive 1a,25(OH)2D3 where the RCI value for Å 0.9 nM) than of 24R,25(OH)2D3 (KD Å 15.3 nM).1a,25(OH)2D3 was set, by definition, to 100%. Therefore, we also conclude the binding proteins for

Vitamin D metabolites. [3H]-1a,25(OH)2D3 was obtained from these two vitamin D metabolites are, indeed, different.Amersham (Arlington Heights, IL). 25(OH)D3 and 1a,25(OH)2D3 We conducted a further analysis of the ligand speci-were kindly provided by Dr. M. R. Uskokovic (Hoffmann-La Roche,

ficity of the 1a,25(OH)2D3 receptor/binding protein inNutley, NJ). [3H]-24R,25(OH)2D3, 24R,25(OH)2D3 and 24S,25-the chick fracture-healing callus system, testing struc-(OH)2D3 were generous gifts from Dr. N. Taniguchi (Kureha Chemi-

cal Industry, Tokyo, Japan). turally related vitamin D metabolites for their abilityto compete with the binding of [3H]-1a,25(OH)2D3 bydetermination of their Relative Competitive IndexRESULTS(RCI). The results are summarized in Table 2. Thebinding in both callus membrane and cytosol fractionsIn order to investigate the existence of physiologi-

cally relevant binding, saturation analyses for all callus was selective for seco-steroids possessing a 1a-hydroxyl

725

AID BBRC 8318 / 694d$$$322 03-11-98 10:59:05 bbrcg AP: BBRC

Vol. 244, No. 3, 1998 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 1. Saturation analysis with [3H]-1a,25(OH)2D3 and [3H]-24R,25(OH)2D3 of a membrane fraction obtained from chick fracture-healing callus. The saturation analysis was performed on the chickcallus membrane over the concentration range of 0.0625-5 nM of[3H]-1a,25(OH)2D3 and 1-50 nM of [3H]-24R,25(OH)2D3, respectively.

group. Also, the natural metabolite 24R,25(OH)2D3 wasnot an effective competitor for [3H]-1a,25(OH)2D3.

DISCUSSIONFIG. 2. Relative Competitive Index (RCI) values of vitamin D

metabolites competing with [3H]-1a,25(OH)2D3 for binding to theWe describe, for the first time, evidence for two puta-chick callus membrane and DBP. Figure 2A: RCI profile of callustive receptors for 1a,25(OH)2D3 in the chick tibial frac-membrane. Figure 2B: RCI profile of callus cytosol. At 1 nM [3H]-ture healing callus membrane and cytosol fractions.1a,25(OH)2D3, the RCI values for 24R,25(OH)2D3, 24S,25(OH)2D3,The KD of the callus membrane and cytosol putative and 25(OH)D3 were obtained where the value for 1a,25(OH)2D3 was

receptors for 1a,25(OH)2D3 were in the subnanomolar set, by definition, at 100% (30).range (Table 1) and the values are consistent with theKD for the chick intestinal nuclear and basal lateral

binding site (RCI Å 2.0). Thus, this present study indi-membrane receptors for 1a,25(OH)2D3 (5).cates that the putative receptors for 24R,25(OH)2D3Further, competition analysis revealed a high ligandand 1a,25(OH)2D3 are different or, at least, recognizingspecificity for the 1a,25(OH)2D3 binding (Fig. 2). There-different binding sites. This may explain the distinctfore, it is suggested that both the receptor/binding pro-biological function for both hormonally active vitaminteins for 1a,25(OH)2D3 are different from that for theD metabolites that generate their biological responsespreviously reported callus membrane receptor forin bone system and cartilage in a direct fashion via24R,25(OH)2D3 (15). If both metabolites were bound toligand specific signal transduction pathways.the same ligand binding domain, 1a,25(OH)2D3 with

The biological responses of the parent vitamin D areits higher binding affinity would compete well (RCI úbelieved to be generated principally by 1a,25(OH)2D3100) to the binding site for 24R,25(OH)2D3. However,and, in some specialized circumstances, by the com-the competition analysis in our previous report exhib-bined actions of 1a,25(OH)2D3 and 24R,25(OH)2D3 (24).ited low affinity of 1a,25(OH)2D3 for the 24R,25(OH)2D3 1a,25(OH)2D3 is known to initiate biological responsesvia interaction with two receptor systems which in-clude the classical nuclear receptor (VDRnuc) that regu-

TABLE 2 lates gene transcription (25), and a membrane receptorRelative Competitive Index (RCI) of Vitamin D Analogs for (VDRmem) which initiates rapid responses, includingBinding to Chick Tibia Fracture-Healing Callus Fractions transcaltachia (5), and the opening of voltage-gated cal-

cium (6) and chloride channels (12). The ligand speci-Relative competitive index ficities of the VDRnuc and VDRmem receptors for the con-

formationally flexible 1a,25(OH)2D3 have been shownCallus membrane Callus cytosolto be different (26). In contrast, the only receptor for

1a,25(OH)2D3 100 100 24R,25(OH)2D3 that has been clearly described is pres-24R,25(OH)2D3 1.3 { 0.22 0.67 { 0.34 ent in the chick tibial fracture-healing callus mem-24S,25(OH)2D3 0.22 { 0.13 0.39 { 0.23 brane fraction; no nuclear receptor for 24R,25(OH)2D325(OH)D3 2.3 { 0.3 1.0 { 0.5 has yet been described.

726

AID BBRC 8318 / 694d$$$322 03-11-98 10:59:05 bbrcg AP: BBRC

Vol. 244, No. 3, 1998 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

7. Baran, D. T., Ray, R., Sorensen, A. M., Honeyman, T., and Hol-While nuclear receptors for 1a,25(OH)2D3 have beenick, M. F. (1994) J. Cell. Biochem. 56, 510–517.previously reported to be present in osteoblasts (27)

8. Bhatia, M., Kirkland, J. B., and Meckling-Gill, K. A. (1995) J.but absent in osteoclasts (28), no previous reports have Biol. Chem. 270, 15962–15965.utilized callus tissue derived from actively healing frac- 9. Song, X.-D., and Norman, A. W. (1997) Endocrinology (in press).tures. In the present study, we did not find clear evi- 10. Norman, A. W. (1997) In Vitamin D (D. M. Feldman, F. H. Glor-dence for the presence of a VDRnuc in the callus tissue; ieux, and J. W. Pike, Eds.), pp. 233–256. Academic Press, San

Diego, CA.however, we did observe a receptor/binding protein11. Minghetti, P. P., and Norman, A. W. (1988) FASEB J. 2, 3043–with a specificity for 1a,25(OH)2D3 present in the

3053.203,900 1 g supernatant obtained after removal of the12. Zanello, L. P., and Norman, A. W. (1997) J. Biol. Chem. 272,nuclear pellet. It is likely that this represents a solubi-

22617–22622.lized form of the VDRnuc . It is known that the distribu- 13. Nemere, I., and Norman, A. W. (1987) J. Bone Miner. Res. 2,tion of the VDRnuc between the nuclear and cytosolic 167–169.fractions of the cell is dependent upon the ambient ionic 14. Somjen, D., Somjen, G. J., Weisman, T., Harell, A., and Bind-strength (29). erman, I. (1981) Isr. J. Med. Sci. 17, 1192.

15. Seo, E.-G., Kato, A., and Norman, A. W. (1996) Biochem. Bio-In summary, we present in this report, evidence for aphys. Res. Commun. 225, 203–208.putative 1a,25(OH)2D3 receptor in both the chick callus

16. Seo, E.-G., 24R,25-Dihydroxyvitamin D3: A vitamin D3 metabo-membrane and cytosol fractions which may generatelite essential for both normal bone development and the healingbiological effects via pathways distinct from those of process of a fracture. [Ph.D. dissertation]. 1–253. 1996. Univer-

the callus membrane receptor for 24R,25(OH)2D3. Cer- sity of California-Riverside.tainly further characterization and purification of this 17. St-Arnaud, R., Arabian, A., and Glorieux, F. H. (1996) J. Bone

Miner. Res. 11, S126. [Abstract]callus protein(s) will allow a greater insight into the18. Seo, E.-G., and Norman, A. W. (1997) J. Bone Miner. Res. 12,biological role of 1a,25(OH)2D3 in the fracture-healing

598–606.callus system.19. Lidor, C., Dekel, S., Meyer, M. S., Blaugrund, E., Hallel, T., and

Edelstein, S. (1990) J. Bone Joint Surgery 72B(1), 137–140.20. Klaus, G., Merke, J., Eing, H., Hugel, U., Milde, P., Reichel, H.,ACKNOWLEDGMENTS

Ritz, E., and Mehls, O. (1991) Calcif. Tissue Int. 49, 340–348.21. Norman, A. W., and Wong, R. G. (1972) J. Nutr. 102, 1709–1718.Portions of this work were supported by grants from the USPHS22. Wecksler, W. R., and Norman, A. W. (1980) J. Biol. Chem. 255,(DK09012-32) and the Biomedical Research Laboratories of the

3571–3574.Kureha Chemical Industry Co., Tokyo, Japan (to AWN).23. Wecksler, W. R., and Norman, A. W. (1980) in Methods in Enzy-

mology: Vitamins and Co-Enzymes, Vol 67 pp. 494–500, Aca-demic Press, New York, NY.REFERENCES

24. Norman, A. W., Roth, J., and Orci, L. (1982) Endocr. Rev. 3, 331–366.1. Bouillon, R., Okamura, W. H., and Norman, A. W. (1995) Endocr.

25. Hannah, S. S., and Norman, A. W. (1994) Nutr. Reviews 52, 376–Rev. 16, 200–257.382.2. Lidor, C., Atkin, I., Ornoy, A., Dekel, S., and Edelstein, S. (1987)

26. Okamura, W. H., Midland, M. M., Hammond, M. W., Rahman,J. Bone Miner. Res. 2, 91–98.N. A., Dormanen, M. C., Nemere, I., and Norman, A. W. (1995)

3. Schwartz, Z., Schlader, D. L., Swain, L. D., and Boyan, B. D. J. Steroid Biochem. Mol. Biol. 53, 603–613.(1988) Endocrinology 123, 2878–2884. 27. Walters, M. R., Rosen, D. M., Norman, A. W., and Luben, R. A.

4. Schwartz, Z., Brooks, B., Swain, L., Del Toro, F., Norman, A. W., (1982) J. Biol. Chem. 257, 7481–7484.and Boyan, B. D. (1992) Endocrinology 130, 2495–2504. 28. Merke, J., Klaus, G., Hugel, U., Waldherr, R., and Ritz, E. (1986)

5. Nemere, I., Dormanen, M. C., Hammond, M. W., Okamura, J. Clin. Invest. 77, 312–314.W. H., and Norman, A. W. (1994) J. Biol. Chem. 269, 23750– 29. Walters, M. R., Hunziker, W., and Norman, A. W. (1980) J. Biol.23756. Chem. 255, 6799–6805.

30. Siebert, P. D., Ohnuma, N., and Norman, A. W. (1979) Biochem.6. Caffrey, J. M., and Farach-Carson, M. C. (1989) J. Biol. Chem.264, 20265–20274. Biophys. Res. Commun. 91, 827–834.

727

AID BBRC 8318 / 694d$$$322 03-11-98 10:59:05 bbrcg AP: BBRC