- Email: [email protected]
Vitamin D-dependent rickets type II: regulation of human osteocalcin gene expression in cells with defective vitamin D receptors by 1,25-dihydroxyvitamin D-3, retinoic acid, and triiodothyronine
BB
ELSEVIER
Biochimica et BiophysicaActa 1227 (1994) 195-199
Biochi~ic~a et BiophysicaA~ta
Vitamin D-dependent rickets type II: regulation of human osteocalcin gene expression in cells with defective vitamin D receptors by 1,25-dihydroxyvitamin D-3, retinoic acid, and triiodothyronine Eiji Takeda a,,, Ken-ichi Miyamoto a, Megumi Kubota a, Hisanori Minami a, Ichiro Yokota b, Takahiko Saijo b, Etsuo Naito b, Michinori Ito b, Yasuhiro Kuroda b a
Department of Clinical Nutrition, School of Medicine, University of Tokushima Kuramoto-cho 3, Tokushima 770, Japan b Department of Pediatrics, School of Medicine, University of Tokushima Kuramoto-cho 3, Tokushima 770, Japan
Received 2 March 1994; revised 8 July 1994
Abstract The vitamin D receptor (VDR) is a nuclear transcription factor which binds to the vitamin D response element (VDRE) of the human osteocalcin gene and regulates its expression. Humans with VDR gene mutations, ever among those with the same point mutation in their VDR gene, demonstrate clinical heterogeneity. In addition, in some patients with these mutations, rickets has not recurred following cessation of therapy during follow-up ranging from 6 to 24 years. While important, it is likely that the VDR protein is not the sole factor in the development of rickets. To try to understand these clinical findings, the complex formed between the VDRE and one or more proteins in the nuclear extracts of cultured skin fibroblasts treated with 1,25-dihydroxyvitamin D-3 (1,25(OH)eD3) , retinoic acid (RA), and/or triiodothyronine (T3) was investigated since such complexes are likely to precede the transcription of the VDR gene. Complex formation in the control cells with an intact VDR was increased by treatment with either 0.1 nM, 1 nM, 10 nM 1,25(OH)2D3, 100 nM RA, or 100 nM T3; however, combinations of these compounds did not produce an additive effect. In cells of affected patients, 1,25(OH)2D3, RA, or T3 increased complex formation, while no combination had an additive effect. These results indicate that 1,25(OH)2D3, RA, and T3 play a role in the regulation of bone remodeling through modulating the formation of protein complexes on the VDRE. Therefore, the clinical observations in patients with a VDR mutation might be explained at least in part by the overlapping control of osteocalcin expression by 1,25(OH)2D3, RA and T3. Keywords: Vitamin D-dependent rickets type II; Vitamin D receptor; Gene expression; Osteocalcin gene; (Human)
1. Introduction The vitamin D receptor (VDR) is a member of the steroid, thyroid and retinoic acid receptor gene family [1]. The requirement for the VDR in the activation of the osteocalcin promoter provides strong evidence that this protein mediates vitamin D activity. The DNA-binding domain of the VDR specifically recognizes the vitamin D response element (VDRE), which is located in the promoter region of the human osteocalcin gene and acts to regulate transcription [2,3]. The consensus sequence motif of the osteocalcin VDRE is thought to be GGGTGACT-
* Corresponding author. Fax: +81 886 337094. 0925-4439/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0925-4439(94)00060-3
C A C C G G G T G A A C G G G G G C A T r , although this remains somewhat controversial [4-6]. The VDR, thyroid hormone receptor (TR), and retinoic acid receptor (RAR) preferentially bind to and activate hormone response elements composed of half-sites arranged as direct repeats [5]. Specificity is conferred by both the half-site sequence and the number of nucleotides separating the two half-sites [2]. The VDR, TR and RAR must form heterodimers with the retinoid X receptor (RXR) in order to bind with high affinity to target D N A [7,8]. Vitamin D-dependent rickets type II (VDDR II) consists of a spectrum of intracellular VDR defects and is characterized by the early onset of severe rickets and associated alopecia. Without treatment, patients suffer inanition, severe skeletal deformity, recurrent respiratory infections, and death by 8 years of age. This disease shows marked
196
E. Takeda et al. /Biochimica et Biophysica Acta 1227 (1994) 195-199
heterogeneity in its clinical and biochemical presentation [9]. In previous studies, six Japanese patients have been diagnosed and successfully treated with 3 to 6 /xg/kg per day of lot(OH)D 3 or 50000 U of vitamin D 2 for 1 to 3 years [10-12]. In five of our patients, we found a unique G-to-A transition at position 140 in exon 3 of the VDR eDNA that results in the substitution of arginine for glutamine at residue 47 which is located within the zinc finger region [12,13]. However, the severity of alopecia, rickets and resistance to vitamin D in these patients was quite heterogeneous [12]. Furthermore, rickets did not recur in these patients, as evaluated by rentogenography and serum biochemical markers, after the cessation of therapy during 6 to 24 years of follow-up, despite their cells continuing to harbor the VDR mutation [11]. These clinical findings suggest that other mechanism(s) in addition to that of the vitamin D-VDR system may be involved. Therefore, the formation of complex between the VDRE of the human osteocalcin gene and one or more nuclear proteins in the nuclear extracts of control and patient fibroblasts, which were treated with 1,25-dihydroxyvitamin D 3 (1,25(OH)2D3), retinoic acid (RA) a n d / o r triiodothyronine (T3) was determined. These complexes may represent an early transcription step. Gel mobility-shift assays were employed for quantitative analysis.
2. Materials and methods 2.1. Preparation of nuclear extracts from cultured skin fibroblasts
Skin samples were obtained from all subjects by a punch biopsy of skin on the forearm. The cells were cultured at 37°C under 5% CO 2 in air in Eagle's minimum essential medium (MEM; Nissui, Tokyo), containing 10% fetal calf serum (M.A. Bioproducts, Walkersville, MD, USA), glutamine and sodium bicarbonate. Confluent fibroblasts were harvested by trypsinization 16 h following exposure to 1,25(OH)2D 3, RA, a n d / o r T3. The cells then were washed twice with ice-cold PBS. The washed cell pellet (108 cells) was resuspended in 1 ml of hypotonic buffer (10 mM Hepes (pH 7.9), 1.5 mM MgC12, 10 mM KC1, 1 mM Dq-T, 1 mM PMSF, 1 mM pepstatin A, 1 mM leupeptin) and incubated on ice for 15 min prior to the addition of 50 /xl of 10% Nonidet P-40 and centrifugation at 12000 rpm (20 min 4°C). The crude nuclear pellet was
A
C C G G G T G A A C
G G G G G C A T T
G G C C C A C T T G
CC
C C C G T A A T T C G A
Fig. 1. Alignment of the nucleotide sequence of the human osteocalcin vitamin D-responsive element. The osteocalcin VDRE is limited to 15 nucleotides closely juxtaposed to a distal functional AP-1 site [5]. The VDR, TR and RAIl response elements within the region are indicated by arrows.
ii~!!!!# ! i!¸
~i~iii~i,¸¸¸ ill !~ii~i!i ¸
!i i I i!iii!ii
1
2
Fig. 2. Inhibition of VDR-VDRE complex formation by an anti-VDR monoclonal antibody. Oligonucleotides corresponding to the VDRE of the human osteocalcin gene were end-labeled and incubated with nuclear extracts of 1,25(OH)2D3(10 nM)-treated fibroblasts in the absence (lane 1) or presence (lane 2) of 0.4 /zg of purified anti-VDR monoclonal antibody 9A7. A gel mobility-shift assay was carried out as described in Section 2. The arrow indicates the position of the VDR-VDRE complex.
resuspended in 5 0 / z l of nuclear extraction buffer (20 mM Hepes (pH 7.9), 0.2 mM EDTA, 1.5 mM MgC12, 0.4 M KCI, 25% glycerol, 1 mM D'I-T, 1 mM PMSF, 1 mM pepstatin A, 1 mM leupeptin), rotated in a cold room for 15 min, and centrifuged at 15 000 rpm (15 min at 4°C). The final extract was stored at - 8 0 ° C in small aliquots at a concentration of 20 mg of protein/ml. 2.2. Gel mobility-shift assay
Two complementary synthetic oligonucleotide strands (5'-CCGGGTGAACGGGGGCATF-3' and 5'-AGCTrAATGCCCCCGTfCACCCGGA-3') were annealed to form the double stranded human osteocalcin VDRE (Fig. 1). The synthetic VDRE was labeled with [32p]ATP and T4 polynucleotide kinase. For the protein-DNA binding assay, 20 /zg of nuclear extract protein were preincubated in binding buffer (10 mM Hepes (pH 7.9), 0.2 mM EDTA (pH 8.0), 1 /xg/ml poly(dI/dC), and 4% glycerol) on ice for 10 min. Subsequently, 2 ng of 32P-labeled-VDRE DNA (4 • 104 dpm) were added to the reaction mixture. After an additional incubation at 20°C for 30 min, the protein-DNA complexes were resolved by 4% polyacrylamide gel electrophoresis (150V, 4°C) in 1 × TGE (50 mM Tris, 0.38 M
E. Takedaet al./ Biochimica et BiophysicaActa 1227 (1994) 195-199
197
formation. Control and patient cells were indistiguishable. In contrast, when both control and patient cells were treated with 10 nM 1,25(OH)2D a, the amount of the VDR-VDRE complex was increased significantly in normal cells at KCI concentrations of 70 mM, 95 mM, and 120 mM. Complex formation did not change significantly in patient cells under these conditions (Fig. 3).
glycine, 2 mM EDTA). The dried gel was exposed with an intensifying screen at - 7 0 ° C with Kodak XAR film. The intensity of the complex was measured by laser densitometry.
3. Results 3.1. Identification of VDR-VDRE complex
3.3. Effect of1,25(OH)2D3, RA, a n d / o r T3 treatment on complex formation with the VDRE of the human osteocalcin gene
The protein-DNA complex was isolated in a gel mobility-shift assay when an extract of control cells treated with 10 nM 1,25-(OH)2D 3 was incubated with the labeled VDRE oligonucleotide (Fig. 2). In order to test whether this complex represented the VDR-VDRE complex, we used an anti-VDR monoclonal antibody known to interact immediately with the carboxyl DNA-binding domain of the VDR and to alter DNA binding [14]. This antibody specifically blocked the formation of the complex at very low concentrations.
In control fibroblasts, treatment with 0.1 nM, 1 nM and 10 nM 1,25(OH)2D 3 increased the formation of the VDRVDRE complex to 322%, 235%, and 148% of untreated cells, respectively (Fig. 4A lanes 1, 2, 3, 4). Exposure of the cells to 100 nM RA or 100 nM T3 also increased complex formation to 123% and 306% of the basal level, respectively (Fig. 4B and C, lane 1). However, both RA and T3 in the presence of 0.1 nM, 1 nM or 10 nM 1,25(OH)2D 3 did not produce an additive effect (Fig. 4A, B and C, lanes 2, 3, 4). In contrast, complex formation in untreated patient fibroblasts was almost absent compared with that in normal cells. Complex formation in mutant cells was slightly increased by the treatment with 1 nM and 10 nM
3.2. Effect of KC1 concentrations on VDR-VDRE complex formation Untreated fibroblasts from control subjects and patients with VDDR II revealed only small amounts of complex
(^) Control subject
(B) Patient
KCi(mM)
20
70
95 120 145 170 195
1__. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
No t r e a t i c n t
20
70
95 120 145 170 195
..-I I
d
10nM 1,25(011)2D~ t r o a t l o n t
Fig. 3. Effect of KC1 concentrations on VDR-VDREcomplex formation. 32P-labeled VDRE were incubated with nuclear extracts prepared from normal cells (A) and patient cells (B) with or without 1,25(OH)2D3(10 nM) treatment at different KC1concentrations. Gel mobility-shiftassays were employedas described in Section 2. Arrows indicate the position of the VDR-VDREcomplexes.
198
E. Takeda et al. /Biochimica et Biophysica Acta 1227 (1994) 195-199
!ii ¸¸ iiiiiiii!iiiiii
(B) RA lOOnM
(C) T3
UW
lOOnM
----
I| lane 1.25(OII)2D3(nM)
1
2
3
4
5
fi
?
0
O.l
l
10
0
1
10
I
I
Control
t
I
Patient
Fig. 4. Effects of 1,25(OH)2D 3, RA and/or T3 on receptor-VDRE complex formation. Nuclear extracts were prepared from normal cells (lane 1, 2, 3, 4), and patient cells (lanes 5, 6, 7) and were treated with different concentrations of 1,25(OH)2D 3 (A), 100 nM RA and 1,25(OH)2D 3 (B), or 100 nM T3 and 1,25(OH)2D 3 (C), respectively. These extracts were incubated with 32P-labeled VDRE. Specific binding of proteins to the VDRE probe was assayed by a gel mobility-shift assay as described in Section 2. Arrows indicate the position of the VDR-VDRE complexes.
1,25(OH)eD 3 (Fig. 4A, lanes 5, 6, 7). When the cells were treated with 100 nM RA or 100 nM T3 in the absence of 1,25(OH)2D3, complex formation was definitely increased. However, an additive effect on complex formation was not seen when 100 nM RA or 100 nM T3 were used in the presence of 1 nM or 10 nM 1,25(OH)2D 3 (Fig. 4B and C, lanes 5, 6, 7).
4. Discussion
Osteocalcin is synthesized in osteoblasts and represents the major non-collagenous matrix constituent of bone. This protein is a useful indicator of bone formation and turnover, and indeed the levels of osteocalcin correlate well with
histomorphometric parameters of bone formation in metabolic bone diseases [15]. The osteocalcin gene illustrates the biological and molecular parameters of vitamin D-mediated gene expression. Since the DNA-binding of VDR is salt-dependent, the optimal KC1 concentration for VDR-VDRE complex formation was investigated to compare its level in both control and patient cells (Fig. 3). An increase in VDR-VDRE complex formation by 1,25(OH)2D 3 was demonstrated in control fibroblasts in this study, presumably reflecting the up-regulation of the VDR system by 1,25(OH)2D 3 as previously reported [16,17]. In contrast, only a slight increase in VDR-VDRE complex formation was observed by using nuclear extracts of patient cells treated with 1,25(OH) 2D 3. This finding agrees with previous reports that a point mutation introduced into the VDR cDNA that encodes the DNA-binding domain of VDR leads to the synthesis of proteins of normal size and ability to bind 1,25(OH)2D 3. HoweVer, protein is unable to interact with nonspecific DNA [13,18]. Therefore, the gel mobility-shift assay using VDRE of the human osteocalcin gene is useful as a diagnostic tool for VDDR II. Our previous clinical studies have suggested that treatment with a sufficient amount of vitamin D derivatives during the active phase of this disease is essential and that bone deformities do not appear after cessation of therapy [11]. In addition, other patients whose VDR defect was located in the vitamin D binding domain have show spontaneous improvement of rachitic bone changes at ages 6 to 7 years [19]. Previous reports have suggested that a variety of circulating hormones, growth factors, and other agents regulate the level of osteocalcin expression in osteoblasts [20]. Consensus DNA- binding sequences for another steroid hormone, retinoic acid, and for the nuclear protooncogene encoding the Fos and Jun proteins [21] are similar to the VDR-binding domain of the osteocalcin gene VDRE. Other results also support the ability of retinoic acid [4] and the Fos-Jun complex [22] to modulate osteocalcin gene transcription and suggest multiple regulatory activities acting at the VDRE. Such activities within the VDRE in a broader context reflect the contribution of multiple promoter-binding factors competing for or complexing with. VDR, TR, and RAR specifically activate response elements containing A / G G G T G A direct repeats with three-, four-, and five-nucleotide spacings, respectively [2]. Furthermore, VDRE in the human osteocalcin gene confers responsiveness to RA and T3. An increase in the complex formation by treatment with either RA or T3 was also demonstrated in both control and patient cells in this study. However, there was not a clear additive effect in the presence of 1,25(OH)2D 3. The abrogation of the effect of RA and T3 in cells treated with 1,25(OH)2D 3 might be explained by the trapping of RXR to VDR. Our results are in agreement with the lack of modulation by 9-cis-retinoic acid of the response to 1,25(OH)2D 3 or its derivatives in COS-7 and ROS 17/2.8 cells [23].
E. Takeda et al. / Biochimica et Biophysica Acta 1227 (1994) 195-199
In s u m m a r y , the present study demonstrates that 1,25(OH)2D3, RA, and T3 regulate b o n e r e m o d e l i n g b y m o d u l a t i n g osteocalcin levels. This might partly explain both the clinical heterogeneity of V D D R II seen in patients with the same m u t a t i o n in the V D R gene and the absence of rickets after the cessation of treatment.
Acknowledgements This w o r k was supported by G r a n t - i n - A i d for Scientific Research ((C): 06670800) from the Ministry of Education, Science and Culture of Japan.
References [1] Evans, R.M. (1988) Science 240, 889-895. [2] Umesono, K., Murakami, K., Thompson, C.C. and Evans, R.M. (1991) Cell 65, 1255-1266. [3] Kerner, S.A., Scott, R.A. and Pike, J.W.(1989) Proc. Natl. Acad. Sci. USA 86, 4455-4459. [4] Schule, R., Umesono, K., Mangelsdorf, D.J., Bolado, J., Pike, J.W. and Evans, R.M. (1990) Cell 61, 497-504. [5] Ozono, K., Liao, J., Scott, R.A., Kerner, S.A. and Pike, J.W. (1990) J. Biol. Chem. 265, 21881-21888. [6] Schrader, M., Bendik, I., Becker-Andre, M. and Carlberg, C. (1993) J. Biol. Chem. 268, 17830-17836. [7] Yu, V.C., Delsert, C., Andersen, B., Holloway, J.M., Devary, O.V., Naar, A.M., Kim, S.Y., Boutin, J-M., Glass, C. and Rosenfeld, M.G. (1991) Cell 67, 1251-1266. [8] Leid, M., Kastner, P., Lyons, R., Nakshatri, H., Saunders, M., Zacharewski, T., Chen, J-Y., Staub, A., Gamier, J-M., Mader, S. and Chambon, P. (1992) Cell 68, 377-395.
199
[9] Marx, S.J., Liberman, U.A., Eil, C., Defrange, D.E., and Bliziotes, M.M. (1985) In Vitamin D: Chemical, Biochemical and Clinical Update (Norman, A.W., Schaefer, K., Grigoleit, H.-G. and Herrath, D.V., eds.), pp. 107-116, Walter de Gruyter, Berlin. [10] Takeda, E., Kuroda, Y., Saijo, T., Naito, E., Kobashi, H., Yokota, I. and Miyao, M. (1987) Pediatrics 80, 97-101. [11] Takeda, E., Kuroda, Y., Kawakami, I., Hashimoto, T., Miyao, M. and Arase, S. (1989) Eur. J. Pediatr. 149, 54-57. [12] Saijo, T., Takeda, E., Ito, M., Naito, E., Yokota, I., Matsuda, J. and Kuroda, Y. (1992) Clin. Pediatr. Endocrinol. 1, 15-19. [13] Saijo, T., Ito, M., Takeda, E., Huq, A.H.M.M., Naito, E., Yokota, I., Matsuda, J. and Kuroda, Y. (1991) Am. J. Hum. Genet. 49, 668-673 [14] Pike, J.W. (1985) In Vitamin D: Chemical, Biochemical and Clinical Update (Norman, A.J., Schaefer, K., Grigoleit, H.-G. and Herrath, D.V., eds.), pp. 97-106, Walter de Gruyter, Berlin. [15] Gundberg, C.M., Lian, J.B., Gallop, P.M. and Steinberg, J.J. (1983) J. Clin. Endocrinol. Metab. 57, 1221-1225. [16] Pike, J.W. and Sleator, N.M. (1985) Biochem. Biophys. Res. Commun. 131, 378-385. [17] Costa, E.M., Hirst, M.A. and Feldman, D. (1985) Endocrinology 117, 2203-2210. [18] Sone, T., Scott, R.A., Hughes, M.R., Malloy, P.J., Feldman, D., O'Malley, B.W. and Pike, J.W. (1989) J. Biol. Chem. 264, 2023020234. [19] Hochberg, Z., Benderli, A., Levy, J., Vardi, P., Weisman, Y., Chen, T. and Feldman, D. (1984) Am. J. Med. 77, 805-811. [20] Noda, M., Yoon, K. and Rodan, G.A. (1988) J. Biol. Chem. 263, 18574-18577. [21] Lian, J., Stewart, C., Puchacz, E., Mackowiak, S., Shalhoub, V., Collart, D., Zambetti, G. and Stein, G. (1989) Proc. Natl. Acad. Sci. USA 86, 1143-1147. [22] Owen, T.A., Bortell, R., Yocum, S.A., Smock, S.L., Zhang, M., Abate, C., Shalhoub, V., Aronin, N., Wright. K i . , van Wijnen, A.J., Stein, J.L., Curran, T., Lian, J.B. and Stein, G.S. (1990) Proc. Natl. Acad. Sci. USA 87, 9990-9994. [23] Ferrara, J., McCuaig, K., Hendy, G.N., Uskokovic, M. and White, J.H. (1994) J. Biol. Chem. 28, 2971-2981.
ELSEVIER
Biochimica et BiophysicaActa 1227 (1994) 195-199
Biochi~ic~a et BiophysicaA~ta
Vitamin D-dependent rickets type II: regulation of human osteocalcin gene expression in cells with defective vitamin D receptors by 1,25-dihydroxyvitamin D-3, retinoic acid, and triiodothyronine Eiji Takeda a,,, Ken-ichi Miyamoto a, Megumi Kubota a, Hisanori Minami a, Ichiro Yokota b, Takahiko Saijo b, Etsuo Naito b, Michinori Ito b, Yasuhiro Kuroda b a
Department of Clinical Nutrition, School of Medicine, University of Tokushima Kuramoto-cho 3, Tokushima 770, Japan b Department of Pediatrics, School of Medicine, University of Tokushima Kuramoto-cho 3, Tokushima 770, Japan
Received 2 March 1994; revised 8 July 1994
Abstract The vitamin D receptor (VDR) is a nuclear transcription factor which binds to the vitamin D response element (VDRE) of the human osteocalcin gene and regulates its expression. Humans with VDR gene mutations, ever among those with the same point mutation in their VDR gene, demonstrate clinical heterogeneity. In addition, in some patients with these mutations, rickets has not recurred following cessation of therapy during follow-up ranging from 6 to 24 years. While important, it is likely that the VDR protein is not the sole factor in the development of rickets. To try to understand these clinical findings, the complex formed between the VDRE and one or more proteins in the nuclear extracts of cultured skin fibroblasts treated with 1,25-dihydroxyvitamin D-3 (1,25(OH)eD3) , retinoic acid (RA), and/or triiodothyronine (T3) was investigated since such complexes are likely to precede the transcription of the VDR gene. Complex formation in the control cells with an intact VDR was increased by treatment with either 0.1 nM, 1 nM, 10 nM 1,25(OH)2D3, 100 nM RA, or 100 nM T3; however, combinations of these compounds did not produce an additive effect. In cells of affected patients, 1,25(OH)2D3, RA, or T3 increased complex formation, while no combination had an additive effect. These results indicate that 1,25(OH)2D3, RA, and T3 play a role in the regulation of bone remodeling through modulating the formation of protein complexes on the VDRE. Therefore, the clinical observations in patients with a VDR mutation might be explained at least in part by the overlapping control of osteocalcin expression by 1,25(OH)2D3, RA and T3. Keywords: Vitamin D-dependent rickets type II; Vitamin D receptor; Gene expression; Osteocalcin gene; (Human)
1. Introduction The vitamin D receptor (VDR) is a member of the steroid, thyroid and retinoic acid receptor gene family [1]. The requirement for the VDR in the activation of the osteocalcin promoter provides strong evidence that this protein mediates vitamin D activity. The DNA-binding domain of the VDR specifically recognizes the vitamin D response element (VDRE), which is located in the promoter region of the human osteocalcin gene and acts to regulate transcription [2,3]. The consensus sequence motif of the osteocalcin VDRE is thought to be GGGTGACT-
* Corresponding author. Fax: +81 886 337094. 0925-4439/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0925-4439(94)00060-3
C A C C G G G T G A A C G G G G G C A T r , although this remains somewhat controversial [4-6]. The VDR, thyroid hormone receptor (TR), and retinoic acid receptor (RAR) preferentially bind to and activate hormone response elements composed of half-sites arranged as direct repeats [5]. Specificity is conferred by both the half-site sequence and the number of nucleotides separating the two half-sites [2]. The VDR, TR and RAR must form heterodimers with the retinoid X receptor (RXR) in order to bind with high affinity to target D N A [7,8]. Vitamin D-dependent rickets type II (VDDR II) consists of a spectrum of intracellular VDR defects and is characterized by the early onset of severe rickets and associated alopecia. Without treatment, patients suffer inanition, severe skeletal deformity, recurrent respiratory infections, and death by 8 years of age. This disease shows marked
196
E. Takeda et al. /Biochimica et Biophysica Acta 1227 (1994) 195-199
heterogeneity in its clinical and biochemical presentation [9]. In previous studies, six Japanese patients have been diagnosed and successfully treated with 3 to 6 /xg/kg per day of lot(OH)D 3 or 50000 U of vitamin D 2 for 1 to 3 years [10-12]. In five of our patients, we found a unique G-to-A transition at position 140 in exon 3 of the VDR eDNA that results in the substitution of arginine for glutamine at residue 47 which is located within the zinc finger region [12,13]. However, the severity of alopecia, rickets and resistance to vitamin D in these patients was quite heterogeneous [12]. Furthermore, rickets did not recur in these patients, as evaluated by rentogenography and serum biochemical markers, after the cessation of therapy during 6 to 24 years of follow-up, despite their cells continuing to harbor the VDR mutation [11]. These clinical findings suggest that other mechanism(s) in addition to that of the vitamin D-VDR system may be involved. Therefore, the formation of complex between the VDRE of the human osteocalcin gene and one or more nuclear proteins in the nuclear extracts of control and patient fibroblasts, which were treated with 1,25-dihydroxyvitamin D 3 (1,25(OH)2D3), retinoic acid (RA) a n d / o r triiodothyronine (T3) was determined. These complexes may represent an early transcription step. Gel mobility-shift assays were employed for quantitative analysis.
2. Materials and methods 2.1. Preparation of nuclear extracts from cultured skin fibroblasts
Skin samples were obtained from all subjects by a punch biopsy of skin on the forearm. The cells were cultured at 37°C under 5% CO 2 in air in Eagle's minimum essential medium (MEM; Nissui, Tokyo), containing 10% fetal calf serum (M.A. Bioproducts, Walkersville, MD, USA), glutamine and sodium bicarbonate. Confluent fibroblasts were harvested by trypsinization 16 h following exposure to 1,25(OH)2D 3, RA, a n d / o r T3. The cells then were washed twice with ice-cold PBS. The washed cell pellet (108 cells) was resuspended in 1 ml of hypotonic buffer (10 mM Hepes (pH 7.9), 1.5 mM MgC12, 10 mM KC1, 1 mM Dq-T, 1 mM PMSF, 1 mM pepstatin A, 1 mM leupeptin) and incubated on ice for 15 min prior to the addition of 50 /xl of 10% Nonidet P-40 and centrifugation at 12000 rpm (20 min 4°C). The crude nuclear pellet was
A
C C G G G T G A A C
G G G G G C A T T
G G C C C A C T T G
CC
C C C G T A A T T C G A
Fig. 1. Alignment of the nucleotide sequence of the human osteocalcin vitamin D-responsive element. The osteocalcin VDRE is limited to 15 nucleotides closely juxtaposed to a distal functional AP-1 site [5]. The VDR, TR and RAIl response elements within the region are indicated by arrows.
ii~!!!!# ! i!¸
~i~iii~i,¸¸¸ ill !~ii~i!i ¸
!i i I i!iii!ii
1
2
Fig. 2. Inhibition of VDR-VDRE complex formation by an anti-VDR monoclonal antibody. Oligonucleotides corresponding to the VDRE of the human osteocalcin gene were end-labeled and incubated with nuclear extracts of 1,25(OH)2D3(10 nM)-treated fibroblasts in the absence (lane 1) or presence (lane 2) of 0.4 /zg of purified anti-VDR monoclonal antibody 9A7. A gel mobility-shift assay was carried out as described in Section 2. The arrow indicates the position of the VDR-VDRE complex.
resuspended in 5 0 / z l of nuclear extraction buffer (20 mM Hepes (pH 7.9), 0.2 mM EDTA, 1.5 mM MgC12, 0.4 M KCI, 25% glycerol, 1 mM D'I-T, 1 mM PMSF, 1 mM pepstatin A, 1 mM leupeptin), rotated in a cold room for 15 min, and centrifuged at 15 000 rpm (15 min at 4°C). The final extract was stored at - 8 0 ° C in small aliquots at a concentration of 20 mg of protein/ml. 2.2. Gel mobility-shift assay
Two complementary synthetic oligonucleotide strands (5'-CCGGGTGAACGGGGGCATF-3' and 5'-AGCTrAATGCCCCCGTfCACCCGGA-3') were annealed to form the double stranded human osteocalcin VDRE (Fig. 1). The synthetic VDRE was labeled with [32p]ATP and T4 polynucleotide kinase. For the protein-DNA binding assay, 20 /zg of nuclear extract protein were preincubated in binding buffer (10 mM Hepes (pH 7.9), 0.2 mM EDTA (pH 8.0), 1 /xg/ml poly(dI/dC), and 4% glycerol) on ice for 10 min. Subsequently, 2 ng of 32P-labeled-VDRE DNA (4 • 104 dpm) were added to the reaction mixture. After an additional incubation at 20°C for 30 min, the protein-DNA complexes were resolved by 4% polyacrylamide gel electrophoresis (150V, 4°C) in 1 × TGE (50 mM Tris, 0.38 M
E. Takedaet al./ Biochimica et BiophysicaActa 1227 (1994) 195-199
197
formation. Control and patient cells were indistiguishable. In contrast, when both control and patient cells were treated with 10 nM 1,25(OH)2D a, the amount of the VDR-VDRE complex was increased significantly in normal cells at KCI concentrations of 70 mM, 95 mM, and 120 mM. Complex formation did not change significantly in patient cells under these conditions (Fig. 3).
glycine, 2 mM EDTA). The dried gel was exposed with an intensifying screen at - 7 0 ° C with Kodak XAR film. The intensity of the complex was measured by laser densitometry.
3. Results 3.1. Identification of VDR-VDRE complex
3.3. Effect of1,25(OH)2D3, RA, a n d / o r T3 treatment on complex formation with the VDRE of the human osteocalcin gene
The protein-DNA complex was isolated in a gel mobility-shift assay when an extract of control cells treated with 10 nM 1,25-(OH)2D 3 was incubated with the labeled VDRE oligonucleotide (Fig. 2). In order to test whether this complex represented the VDR-VDRE complex, we used an anti-VDR monoclonal antibody known to interact immediately with the carboxyl DNA-binding domain of the VDR and to alter DNA binding [14]. This antibody specifically blocked the formation of the complex at very low concentrations.
In control fibroblasts, treatment with 0.1 nM, 1 nM and 10 nM 1,25(OH)2D 3 increased the formation of the VDRVDRE complex to 322%, 235%, and 148% of untreated cells, respectively (Fig. 4A lanes 1, 2, 3, 4). Exposure of the cells to 100 nM RA or 100 nM T3 also increased complex formation to 123% and 306% of the basal level, respectively (Fig. 4B and C, lane 1). However, both RA and T3 in the presence of 0.1 nM, 1 nM or 10 nM 1,25(OH)2D 3 did not produce an additive effect (Fig. 4A, B and C, lanes 2, 3, 4). In contrast, complex formation in untreated patient fibroblasts was almost absent compared with that in normal cells. Complex formation in mutant cells was slightly increased by the treatment with 1 nM and 10 nM
3.2. Effect of KC1 concentrations on VDR-VDRE complex formation Untreated fibroblasts from control subjects and patients with VDDR II revealed only small amounts of complex
(^) Control subject
(B) Patient
KCi(mM)
20
70
95 120 145 170 195
1__. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
No t r e a t i c n t
20
70
95 120 145 170 195
..-I I
d
10nM 1,25(011)2D~ t r o a t l o n t
Fig. 3. Effect of KC1 concentrations on VDR-VDREcomplex formation. 32P-labeled VDRE were incubated with nuclear extracts prepared from normal cells (A) and patient cells (B) with or without 1,25(OH)2D3(10 nM) treatment at different KC1concentrations. Gel mobility-shiftassays were employedas described in Section 2. Arrows indicate the position of the VDR-VDREcomplexes.
198
E. Takeda et al. /Biochimica et Biophysica Acta 1227 (1994) 195-199
!ii ¸¸ iiiiiiii!iiiiii
(B) RA lOOnM
(C) T3
UW
lOOnM
----
I| lane 1.25(OII)2D3(nM)
1
2
3
4
5
fi
?
0
O.l
l
10
0
1
10
I
I
Control
t
I
Patient
Fig. 4. Effects of 1,25(OH)2D 3, RA and/or T3 on receptor-VDRE complex formation. Nuclear extracts were prepared from normal cells (lane 1, 2, 3, 4), and patient cells (lanes 5, 6, 7) and were treated with different concentrations of 1,25(OH)2D 3 (A), 100 nM RA and 1,25(OH)2D 3 (B), or 100 nM T3 and 1,25(OH)2D 3 (C), respectively. These extracts were incubated with 32P-labeled VDRE. Specific binding of proteins to the VDRE probe was assayed by a gel mobility-shift assay as described in Section 2. Arrows indicate the position of the VDR-VDRE complexes.
1,25(OH)eD 3 (Fig. 4A, lanes 5, 6, 7). When the cells were treated with 100 nM RA or 100 nM T3 in the absence of 1,25(OH)2D3, complex formation was definitely increased. However, an additive effect on complex formation was not seen when 100 nM RA or 100 nM T3 were used in the presence of 1 nM or 10 nM 1,25(OH)2D 3 (Fig. 4B and C, lanes 5, 6, 7).
4. Discussion
Osteocalcin is synthesized in osteoblasts and represents the major non-collagenous matrix constituent of bone. This protein is a useful indicator of bone formation and turnover, and indeed the levels of osteocalcin correlate well with
histomorphometric parameters of bone formation in metabolic bone diseases [15]. The osteocalcin gene illustrates the biological and molecular parameters of vitamin D-mediated gene expression. Since the DNA-binding of VDR is salt-dependent, the optimal KC1 concentration for VDR-VDRE complex formation was investigated to compare its level in both control and patient cells (Fig. 3). An increase in VDR-VDRE complex formation by 1,25(OH)2D 3 was demonstrated in control fibroblasts in this study, presumably reflecting the up-regulation of the VDR system by 1,25(OH)2D 3 as previously reported [16,17]. In contrast, only a slight increase in VDR-VDRE complex formation was observed by using nuclear extracts of patient cells treated with 1,25(OH) 2D 3. This finding agrees with previous reports that a point mutation introduced into the VDR cDNA that encodes the DNA-binding domain of VDR leads to the synthesis of proteins of normal size and ability to bind 1,25(OH)2D 3. HoweVer, protein is unable to interact with nonspecific DNA [13,18]. Therefore, the gel mobility-shift assay using VDRE of the human osteocalcin gene is useful as a diagnostic tool for VDDR II. Our previous clinical studies have suggested that treatment with a sufficient amount of vitamin D derivatives during the active phase of this disease is essential and that bone deformities do not appear after cessation of therapy [11]. In addition, other patients whose VDR defect was located in the vitamin D binding domain have show spontaneous improvement of rachitic bone changes at ages 6 to 7 years [19]. Previous reports have suggested that a variety of circulating hormones, growth factors, and other agents regulate the level of osteocalcin expression in osteoblasts [20]. Consensus DNA- binding sequences for another steroid hormone, retinoic acid, and for the nuclear protooncogene encoding the Fos and Jun proteins [21] are similar to the VDR-binding domain of the osteocalcin gene VDRE. Other results also support the ability of retinoic acid [4] and the Fos-Jun complex [22] to modulate osteocalcin gene transcription and suggest multiple regulatory activities acting at the VDRE. Such activities within the VDRE in a broader context reflect the contribution of multiple promoter-binding factors competing for or complexing with. VDR, TR, and RAR specifically activate response elements containing A / G G G T G A direct repeats with three-, four-, and five-nucleotide spacings, respectively [2]. Furthermore, VDRE in the human osteocalcin gene confers responsiveness to RA and T3. An increase in the complex formation by treatment with either RA or T3 was also demonstrated in both control and patient cells in this study. However, there was not a clear additive effect in the presence of 1,25(OH)2D 3. The abrogation of the effect of RA and T3 in cells treated with 1,25(OH)2D 3 might be explained by the trapping of RXR to VDR. Our results are in agreement with the lack of modulation by 9-cis-retinoic acid of the response to 1,25(OH)2D 3 or its derivatives in COS-7 and ROS 17/2.8 cells [23].
E. Takeda et al. / Biochimica et Biophysica Acta 1227 (1994) 195-199
In s u m m a r y , the present study demonstrates that 1,25(OH)2D3, RA, and T3 regulate b o n e r e m o d e l i n g b y m o d u l a t i n g osteocalcin levels. This might partly explain both the clinical heterogeneity of V D D R II seen in patients with the same m u t a t i o n in the V D R gene and the absence of rickets after the cessation of treatment.
Acknowledgements This w o r k was supported by G r a n t - i n - A i d for Scientific Research ((C): 06670800) from the Ministry of Education, Science and Culture of Japan.
References [1] Evans, R.M. (1988) Science 240, 889-895. [2] Umesono, K., Murakami, K., Thompson, C.C. and Evans, R.M. (1991) Cell 65, 1255-1266. [3] Kerner, S.A., Scott, R.A. and Pike, J.W.(1989) Proc. Natl. Acad. Sci. USA 86, 4455-4459. [4] Schule, R., Umesono, K., Mangelsdorf, D.J., Bolado, J., Pike, J.W. and Evans, R.M. (1990) Cell 61, 497-504. [5] Ozono, K., Liao, J., Scott, R.A., Kerner, S.A. and Pike, J.W. (1990) J. Biol. Chem. 265, 21881-21888. [6] Schrader, M., Bendik, I., Becker-Andre, M. and Carlberg, C. (1993) J. Biol. Chem. 268, 17830-17836. [7] Yu, V.C., Delsert, C., Andersen, B., Holloway, J.M., Devary, O.V., Naar, A.M., Kim, S.Y., Boutin, J-M., Glass, C. and Rosenfeld, M.G. (1991) Cell 67, 1251-1266. [8] Leid, M., Kastner, P., Lyons, R., Nakshatri, H., Saunders, M., Zacharewski, T., Chen, J-Y., Staub, A., Gamier, J-M., Mader, S. and Chambon, P. (1992) Cell 68, 377-395.
199
[9] Marx, S.J., Liberman, U.A., Eil, C., Defrange, D.E., and Bliziotes, M.M. (1985) In Vitamin D: Chemical, Biochemical and Clinical Update (Norman, A.W., Schaefer, K., Grigoleit, H.-G. and Herrath, D.V., eds.), pp. 107-116, Walter de Gruyter, Berlin. [10] Takeda, E., Kuroda, Y., Saijo, T., Naito, E., Kobashi, H., Yokota, I. and Miyao, M. (1987) Pediatrics 80, 97-101. [11] Takeda, E., Kuroda, Y., Kawakami, I., Hashimoto, T., Miyao, M. and Arase, S. (1989) Eur. J. Pediatr. 149, 54-57. [12] Saijo, T., Takeda, E., Ito, M., Naito, E., Yokota, I., Matsuda, J. and Kuroda, Y. (1992) Clin. Pediatr. Endocrinol. 1, 15-19. [13] Saijo, T., Ito, M., Takeda, E., Huq, A.H.M.M., Naito, E., Yokota, I., Matsuda, J. and Kuroda, Y. (1991) Am. J. Hum. Genet. 49, 668-673 [14] Pike, J.W. (1985) In Vitamin D: Chemical, Biochemical and Clinical Update (Norman, A.J., Schaefer, K., Grigoleit, H.-G. and Herrath, D.V., eds.), pp. 97-106, Walter de Gruyter, Berlin. [15] Gundberg, C.M., Lian, J.B., Gallop, P.M. and Steinberg, J.J. (1983) J. Clin. Endocrinol. Metab. 57, 1221-1225. [16] Pike, J.W. and Sleator, N.M. (1985) Biochem. Biophys. Res. Commun. 131, 378-385. [17] Costa, E.M., Hirst, M.A. and Feldman, D. (1985) Endocrinology 117, 2203-2210. [18] Sone, T., Scott, R.A., Hughes, M.R., Malloy, P.J., Feldman, D., O'Malley, B.W. and Pike, J.W. (1989) J. Biol. Chem. 264, 2023020234. [19] Hochberg, Z., Benderli, A., Levy, J., Vardi, P., Weisman, Y., Chen, T. and Feldman, D. (1984) Am. J. Med. 77, 805-811. [20] Noda, M., Yoon, K. and Rodan, G.A. (1988) J. Biol. Chem. 263, 18574-18577. [21] Lian, J., Stewart, C., Puchacz, E., Mackowiak, S., Shalhoub, V., Collart, D., Zambetti, G. and Stein, G. (1989) Proc. Natl. Acad. Sci. USA 86, 1143-1147. [22] Owen, T.A., Bortell, R., Yocum, S.A., Smock, S.L., Zhang, M., Abate, C., Shalhoub, V., Aronin, N., Wright. K i . , van Wijnen, A.J., Stein, J.L., Curran, T., Lian, J.B. and Stein, G.S. (1990) Proc. Natl. Acad. Sci. USA 87, 9990-9994. [23] Ferrara, J., McCuaig, K., Hendy, G.N., Uskokovic, M. and White, J.H. (1994) J. Biol. Chem. 28, 2971-2981.