THE JOURNAL OF BIOLOGICAL CHEMISTRY Q 1993 by The American Society for Biochemiatry and Molecular Biology, Inc

Vol. 268, No. 19, Issue of July 5, pp. 13811-13819,1993 Printed in U.S.A.

Structure-Function Studies of 1,25-Dihydroxyvitamin D3 and the Vitamin D Endocrine System 1,25-DIHYDROXY-PENTADEUTERIO-PREVITAMIN D3 (AS A 6-s-CIS ANALOG) STIMULATES NONGENOMIC BUT NOT GENOMIC BIOLOGICAL RESPONSES*

(Received for publication, November 2, 1992)

Anthony W. Norman$#ll, William H. Okamurall, Mary C. Farach-Carson**, Katrien Allewaert$$, Dimitri Branisteanu$$, Ilka Nemere$, K. Raman Muralidharan11 , and Roger Bouillon$$ From the $Department of Biochemistry, the $Division of Biomedical Sciences, and the IlDepartment of Chemistry, University of California, Riverside, California 92521, the **Department of Biological Chemistry, University of Texas Dental Branch, Houston, Texas 77225, and the $$Laboratorium voor Experimentele Geneeskunde en Endocrinologie, Katholieke Universiteit Leuuen, Gasthuisberg, B-3000, Leuven, Belgium

The hormone 1,25-dihydroxyvitamin D3 (1,225- (OH)2D3) generates biological responses via both ge- nomic and nongenomic mechanisms. This article ad- dresses activity differences between the 6-s-trans (ex- tended) and the 6-s-cis (steroid-like) conformation of 1,25-(OH)2D3 to initiate these responses. Because of facile interconversion of the 6-s-trans and 6-s-cis con- formers of 1,25-(OH)2D3, kinetically competent amounts of both conformers exist to interact with any potential receptors for 1,25-(OH)zD3. We have chemi- cally synthesized 1,25-(OH)2-9,14,19,19,19-penta- deuterio-pre-D3 (1,25-(OH)z-da-pre-D3), an analog of the 6-s-cis conformation of 1,25-(OH)zD3. We found that l,25-(OH)2-d6-pre-D3 and 1,25-(OH)2D3 were equivalently active in two nongenomic systems (tran- scaltachia as measured in the perfused chick intestine and “Ca” uptake through voltage-gated Ca2+ channels in ROS 17/2.8 cells). 1,25-(OH)z-d6-pre-D3 was signif- icantly less active both in binding in vitro to the plasma vitamin D-binding protein (7%) and to the chick (10%) and pig (4%) intestinal nuclear 1,25-(OH)2D3 receptors generating genomic biological responses in vivo (in- duction of plasma levels of osteocalcin, < 5%) or in cultured cells (inhibition of HL-60 cell differentiation, < 5%; inhibition of MG-63 proliferation, < 2%; and induction of osteocalcin, < 2%). These results suggest that the genomic and nongenomic responses are me- diated by separate receptors. Further, the 6-s-cis form (steroid-like conformation) of the natural hormone, 1,25-(OH)zDs, may be selectively responsible for its nongenomic function(s).

The secosteroid’ vitamin DB is responsible for a wide array ~~ ~~~

* This work was supported in part by United States Public Health Service Grants DK-09012-28 (to A. W. N.), DK-16,595 (to W. H. O.), and DE-10,318-01 (to M. C. F.-C.) and by the Belgian Fonds voor Geneeskundig Wetenschappelijk Onderzoek FGWO 3.0044.89 (to R. B.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

ll Prepared while on sabbatical leave at Katholieke Universiteit- Leuven September through December, 1992. To whom all correspond- ence should be addressed: Dept. of Biochemistry, University of Cali- fornia, Riverside, CA 92521.

‘ Secosteroids are by definition compounds in which one of the cyclopentanoperhydrophenanthrene rings of the steroid ring struc- ture is broken. In the case of vitamin D, the 9-10 carbon-carbon bond of the B ring is broken generating a seco-B steroid. The official IUPAC name for vitamin D3 is 9,10-secocholesta-5,7,10(19)-trien-3B- 01.

of biological responses in higher animals, including mainte- nance of calcium homeostasis, immunomodulation, and se- lected cell differentiation (1, 2). However, the parent vitamin D3 is biologically inert, and it is only as a consequence of its metabolism to 1,25-(OH)zD32 and other metabolites that its biological effects are achieved.

It is well established that 1,25-(OH)zD3 generates many biological responses as a consequence of its interaction with nuclear receptors so as to regulate gene transcription (3-5). Indeed, the nuclear receptor for 1,25-(OH)zD3 belongs to the same superfamily of proteins which includes receptors for the steroid hormones, retinoic acid, and thyroxine (6). However not all 1,25-(OH)zD3 responses are mediated by genome acti- vation; there is clear evidence that 1,25-(OH)zD3 is able to generate biological responses by nongenomic pathways. These include the rapid stimulation of intestinal Ca2+ transport, termed transcaltachia (7-9), which involves the rapid opening of Ca2+ channels (lo), rapid effects on phospholipid metabo- lism in the intestine ( l l ) , liver (12), parathyroid cells (13), and kidney (14), and an opening of voltage-gated Ca2+ chan- nels in rat osteosarcoma cells (15, 16). Current evidence suggests that different forms of the 1,25-(OH)zD3 receptor are involved in the signal transduction processes associated with genomic and nongenomic biological responses (17, 18).

The A ring, triene, and side chain of vitamin DB and all of its metabolites are, in comparison with steroid hormones, unusually conformationally mobile (19,20), and it is pertinent in a structure-function sense whether different conformers of 1,25-(OH)zD3 differ in their ability to mediate biological re- sponses, i.e. differ in their ability to interact with the various species of 1,25-(OH)& receptors. In this study, we focus on the conjugated triene system characteristic of vitamin D. Vitamin D secosteroids can undergo rotation about the 6-7 carbon-carbon single bond to create two distinct conformers, the 6-s-cis (steroid-like conformation) and the 6-s-trans (ex- tended steroid conformation); see Fig. 1. In addition, vitamin D and 1,25-(OH)zD3 are in a slower thermal equilibrium with their double bond shifted previtamin D forms, so that all biological systems may be exposed to not only the pair of 6- s-cis and 6-s-trans conformers, hut also to finite concentra-

The abbreviations used are 1,25-(OH)2D3, 1,25-dihydroxyvitamin D3; 1,25-(OH)2-pre-D3, 1,25-dihydroxyprevitamin D, (analog BC); D,, vitamin D3; pre-Ds, previtamin D3; 1,25-dihydroxy-9,14,19,19,19-pen- tadeuterio-previtamin D,, 1,25-(OH)z-&-pre-D3 (analog HF); 1,25- (OH)*-&-&, 1,25-dihydroxy-9,14,19,19,19-pentadeuterio-vitamin D3 (analog HG); DBP, vitamin D-binding protein; h, human; GBSS, Grey’s balanced salt solution; RCI, relative competitive index.

13811

13812 Structure-Function Studies of 1,25-(OH)2-previtamin D3 tions of the more slowly produced 1,25-(OH),-pre-D3 species.

We report here the extensive biological evaluation in both genomic and nongenomic systems of two pentadeuteriated forms of 1,25-(0H)2D3, namely the 1,25-(OH),-&-pre-D3 and 1,25-(0H)z-&-D3. As was shown by kinetic investigations of these substrates, the presence of the deuterium atoms on the l,25-(OH),-&-pre-D3 suppresses its rearrangement from the previtamin to the vitamin form (21). Thus, 1,25-(OH),-&- pre-D3 can function as an analog only of the 6-s-cis form of 1,25-(OH)z&. The results indicate that two nongenomic sys- tems can respond as effectively to 1,25-(OH),-&-pre-D3 as 1,25-(OH)~&, whereas all genomic systems tested discrimi- nate markedly against the 1,25-(OH)z-&-pre-D3 species, sug- gesting that the ligand binding domain of the 1,25-(OH)2D3 genomic receptors may be fundamentally different from the nongenomic receptor.

EXPERIMENTAL PROCEDURES

Chemicals

(OH),D, was the kind gift of Dr. Milan Uskokovic (Hoffmann La “CaCl, was obtained from Du Pont-New England Nuclear. 1,25-

Panel A

___c

7-Dehydrocholesteroi

- H O ’

Vitamin D,

(6-s-trans)

Previtamin D,

Vitamin D,

(6-s-cis)

Panel B

( extended ) 6-s-trans or

conformation

6-s-cis or

conformation steroid-like

FIG. 1. Panel A, metabolic scheme for production of vitamin D3. The provitamin, 7-dehydrocholesterol, present in the skin is con- verted by ultraviolet irradiation into the secosteroid vitamin Da. Previtamin D3 is in thermal equilibrium with vitamin Dg; the conver- sion involves a [1,7]-sigmatropic shift, i.e. the intramolecular migra- tion of a hydrogen from carbon-19 to carbon-9. The resulting product vitamin D3 is a conformationally mobile molecule with respect to the orientation of the A ring in relation to the C/D ring structure. As indicated in the bottom line of this panel, the seco-B ring can assume one of two conformations as a consequence of rotation about the carbon 6-7 single bond in the 6-s-cis orientation the A ring is related to the C/D rings as in the conventional steroid orientation, referred to here as the “steroid-like conformation” and when the conformation is in the 6-s-trans orientation, the A ring is present in an “extended conformation.” Panel B, the hormonally active form of vitamin D3, 1,25-(OH)2D3, also has free rotation about the single bond between carbon-6 and carbon-7; accordingly it too can assume in solution the steroid-like conformation (6-s-cis) or the extended conformation (6- s-trans) orientation.

Roche, Nutley, NJ). (MethyL3H]Thymidine (2 Ci/mmol) was pur- chased from Amersham Corp. Cell culture media were purchased from GIBCO. Penicillin and streptomycin were from Boehringer (Mannheim, Germany). 4-nitro blue tetrazolium was obtained from Sigma. Human plasma vitamin D-binding protein (DBP) was pre- pared by affinity chromatography as described previously (22, 23).

Chemical Synthesis of 1,25-(~H~2-9,14,19,19,19-d6-pre-D3 (Analog F) Analogs HF and 1,25-(OH)p-&-Da (analog HG) were synthesized

according to the method of Curtin and Okamura (21). When 1,25- (OH)~-&-pre-Da, which had been stored at -60 ‘C for about 1 year with occasional warming to ambient temperatures for withdrawal of samples for biological evaluation, the sample analyzed to be comprised of 4.4% of the vitamin and 95.6% of the previtamin form of analog HF. A similar sample of 1,25-(0H)z-&-pre-D3 maintained at the same temperature without occasional warming or a freshly synthesized sample was, by comparison, found to be comprised of 1.2% of the vitamin form and 98.8% of the preform. The composition determi- nations were carried out by analytical high performance liquid chro- matography on a normal phase column (Whatman Partisil column using 90% ethyl acetate, 10% hexane as solvent; 5 ml/min flow rate) using a Waters photodiode array detector. The electronically inte- grated peak area was the average of two values, one obtained at 260 nm and the other at 266 nm (1,25-(OH),-&-D3D E = 17,200; 1,25- (OH)z-Q-pre-D3 E = 7,200). A separate comparison using cut and weigh integrated peak areas was used as a cross-check, and the overall agreement was estimated to be k 0.7%. The retention times were as follows: 1,25-(OH)2-&-D3, about 18 min; 1,25-(OH),-&-pre-D3, 24.5 min. These retention times are essentially identical to those of the undeuteriated forms of these two secosteroids.

Animals and Cells Riuerside-White Leghorn cockerels (Lakeview Farms, Lakeview,

CA) were obtained on the day of hatch and maintained on a vitamin D-supplemented diet (1.0% calcium and 1.0% phosphorus; 0. H. Kruse Grain and Milling, Ontario, CA) for 5-6 weeks to prepare normal vitamin D3-replete chicks. All experiments employing animals were approved by the University of California-Riverside Chancellor’s Committee on Animals in Research.

Leuuen-The human promyelocytic leukemia cell line (HL-60) and the MG-63 cells were obtained from the American Type Culture Collection (Rockville, MD). One-day-old RIR chicks were housed in a windowless room and raised on a vitamin D-replete diet for 1 week followed by a vitamin D-deficient diet (Hope Farms, Woerden, The Netherlands) for the next 5 weeks. After a total of 6 weeks, they were divided into groups and received a single intramuscular injection of 400 ng of 1,25-(0H),D3 or analogs HF or HG solubilized in 10:10:80 v/v/v ethanol, Tween 80, NaCl, 0.9%. Blood was obtained from the wing vein at 2, 4, 6, 9, 12, 16, 24, and 36 h after the time of administration of the secosteroid. Serum osteocalcin was measured by radioimmunoassay using specific anti-chick antisera raised against these chick proteins. Mice, strain NMRI, were fed a normal diet (Hope Farms) for 40-60 days. They received a daily subcutaneous dose of 1,25-(OH)2D3 or analogs HF and HG for 7 days. Serum Ca2+ was determined via atomic absorption spectrophotometry and serum osteocalcin levels via radioimmunoassay. The pig intestinal mucosa was obtained from a normal 20-kg pig under Ketalar anesthesia. The mucosa was scraped and stored at -80 “C until time of preparation of the 1,25-(OH),D3 nuclear receptor (see below).

Houston-The ROS 17/2.8 cells (kindly provided originally by Dr. Gideon Rodan, Merck, Sharp and Dohme, West Point, PA) were cultured in Dulbecco’s modified Eagle’s medium:Ham’s F-12 medium 1:l containing 10% fetal calf serum (GIBCO-BRL). The medium was supplemented with 1.1 mM CaCl, as described (24). For “Ca2+ uptake experiments, cells were seeded at a density of 30,000 cells/ml into 3.5-cm dishes and grown to approximately 50% confluence.

Calcium Uptake Assays ROS 17/2.8 cells were assayed for Ca2+ uptake using procedures

described previously (15). Assays were standardized to 1 min, which preliminary experiments demonstrated to be within the interval of linear uptake. Culture medium was aspirated and the cells washed with room temperature Hanks’ buffered saline and then incubated 1 min with either “resting buffer” (containing, in mM concentration, 132 NaCl, 5 KCl, 1.3 MgClz, 1.2 CaC12, 10 glucose, and 25 Tris-HC1, pH 7.4) or “stimulating buffer” (containing, in mM concentration, 5 NaCl, 132 KCl, 1.3 MgCl,, 1.2 CaCl,, 10 glucose, and 25 Tris-HC1,

Structure-Function Studies of 1,25-(OH)z-previtamin D3 13813

pH 7.4). Both uptake solutions contained 12.5 pCi/ml "Can+ (Du Pont-New England Nuclear) and the concentrations of vitamin D agonists indicated in Fig. 2. Uptake was terminated by aspiration of the labeling solution, followed by three washes with ice-cold resting buffer. Cell-associated "Ca2+ was extracted by a 2-h incubation with 0.5 M NaOH and measured by liquid scintillation counting. It was found that "Ca2+ uptake by monolayer cultures of ROS 17/2.8 cells was density-dependent. Maximal uptake rates were consistently found for cultures that were between 50 and 80% confluent.

Intestinal T!a2+ Transport (Transcaltachia) Measurements of 46Ca2+ transport were carried out in perfused

chick duodena as described previously (7-9). Normal vitamin D- replete chicks weighing approximately 500 g were anesthetized with Chloropent (Fort Dodge, IA; 0.3 m1/100 g), and the duodenal loop was surgically exposed. Blood vessels branching off from the celiac artery were ligated before cannulation of the celiac artery itself and simultaneous initiation of vascular perfusion. The duodenal loop was then excised and, after cannulation of the celiac vein, placed between layers of saline-moistened cheesecloth at 24 "C. The arterial perfusion was initiated during cannulation with modified Grey's balanced salt solution (CBSS) modified to contain 0.9 mM CaC12 and oxygenated with 95% 0 2 and 5% COZ at a flow rate of 2 ml/min. An auxiliary pump was used for the introduction of vehicle (ethanol) or test substances plus albumin (0.125% w/v final concentration) to the vascular perfusate at a rate of 0.25 ml/min. The intestinal lumen was then flushed and filled with GBSS containing 46CaC1 (5 pCi/ml) but without bicarbonate or glucose. A basal transport rate was established by perfusion with control medium for 20 min after the lumen was filled with "Ca". The tissue was then exposed to 1,25-(OH)zD3 or 1,25-(0H),-&-pre-D3 or reexposed to control medium for an addi- tional 40 min. The vascular perfusate was collected at 2-min intervals during the last 10 min of the basal and during the entire treatment period. Duplicate 100-p1 aliquots were taken for determination of the '%a2+ levels by liquid scintillation spectrometry. The results are expressed as the ratio of the "Ca2' appearing in the 40-min test period over the average initial basal transport period.

Ligand Binding Studies The relative ability of each analog to compete with [3H]1,25-

(OH)2D3 for binding to the chick intestinal nuclear receptor for 1,25- (OH),D3 was carried out under in vitro conditions according to our standard procedures (25,26). In this assay, increasing concentrations of nonradioactive 1,25-(OH)z& or the test analog are incubated with a fixed saturating amount of [3H]1,25-(OH)zD3 and chick intestinal nuclear extract obtained from vitamin D-deficient chicks; the recip- rocal of the percentage of maximal binding of [3H]1,25-(OH)2D3 was then calculated and plotted as a function of the relative concentration of the analog and [3H]1,25-(OH)2D3. As shown in Fig. 6B, such plots give linear curves characteristic for each analog, the slopes of which are equal to the analog's competitive index value (25). The competi- tive index value for each analog is then normalized to a standard curve obtained with nonradioactive 1,25-(0H),D3 as the competing steroid and placed on a linear scale of relative competitive index (RCI), where the RCI of 1,25-(0H),D3 by definition is 100. Binding to the 1,25-(OH),D3 receptor was determined in mucosa obtained from a vitamin D-replete pig. Frozen (-80 "C) duodenal mucosa was sonicated in 4 volumes of buffer (0.5 M Tris-HC1, 0.5 M KCl, 5 mM dithiothreitol, 10 mM NazMo04, 1.5 mM EDTA, pH 7.5). The high speed supernatant was then incubated with 0.2 nM [3H]l,25-(OH)2D3 and increasing concentrations of nonradioactive 1,25-(OH),D3 or its analogs in a final volume of 0.3 ml overnight a t 25 "C followed by 5 min at 4 "C. Phase separation was then obtained by the addition of cold dextran-coated charcoal.

Binding of the 1,25-(OH)2D3 and its analogs to hDBP was per- formed at 4 'C essentially as described previously (27). [3H]1,25- (OH)ZDB and 1,25-(OH)~D3 or its analogs were added in 5 p1 of ethanol into glass tubes and incubated with hDBP (0.18 p ~ ) in a final volume of 1 ml (0.01 M Tris-HC1, 0.154 M NaC1, pH 7.4) for 4 h at 4 "C. Phase separation was then obtained by the addition of 0.5 ml of cold dextran-coated charcoal.

Culture Conditions for HL-60 and MG-63 Cells HL-60 cells were seeded at 1.2 X 10' cells/ml, and 1,25-(OH)zD3 or

its analogs were added in ethanol (final concentration < 0.2%) in RPMI 1640 medium supplemented with 10% heat-inactivated fetal calf serum (GIBCO), 100 units/ml penicillin, and 100 pg/ml strepto-

mycin (Boehringer). After 4 days of culture in a humidified atmos- phere of 5% COz in air at 37 "C the dishes were shaken to loosen any adherent cells, and all cells were then assayed for differentiation by nitro blue tetrazolium reduction assay and for proliferation by [3H] thymidine incorporation. The MG-63 cells were seeded at 5 X lo3 cells/ml in 96-well flat bottomed culture plates (Falcon, Becton Dickinson, NJ) in a volume of 200 pl of Dulbecco's modified Eagle's medium containing 2% of heat-inactivated charcoal-treated fetal calf serum, and 1,25-(OH),D3 or its analogs were added in ethanol (final concentration < 0.2%). After 72 h of culture in a humidified atmos- phere of 5% CO, in air at 37 'C the inhibition of proliferation by [3H] thymidine incorporation and measurement in the medium of osteo- calcin concentration using a homologous human radioimmunoassay were performed (27).

Nitro Blue Tetrazolium Reduction Assay

Superoxide production was assayed by nitro blue tetrazolium- reducing activity as described previously (27). HL-60 cells at 1.0 X lo6 cells/ml were mixed with an equal volume of freshly prepared solution of phorbol 12-myristate 13-acetate (200 ng/ml) and nitro blue tetrazolium (2 mg/ml) and incubated for 30 min at 37 "C. The percentage of cells containing black formazan deposits was deter- mined using a hemacytometer.

Statistics Statistical evaluation of the data was performed by Student's t test

for unpaired observations.

RESULTS

This article presents a comparison of the biological profile of the two deuteriated analogs, 1,25-(OH)z-&-pre-D3 and 1,25- (OH)z-&-Ds i n relation t o 1,25-(0H)*D3. The structures of the analogs are given in Fig. 2. The 1,25-(OH)z-&-pre-D3, analog HF, is kinetically suppressed (21) in its previtamin form (because of a primary deuterium kinetic isotope effect)

1,25(OH),Dl (Compound C)

warm

HO

6-s-trans conformation

At 37 T , pure 1,25(OH),-pre-D,

a half life of 13.4 hour: isomerizes to 1,25(OH) D with

warm

1,25(OH)2-d,-D, (Compound HG) HO

HO.\AOH 5% at equilibrium

6-s-trans conformation

At 37 'C, pure 1,25(OH),-d5-pre-D, isomerizes to 1,25(OH)2-d,-Dl with a half life of 81.0 hours

FIG. 2. Equilibrium relationship between the previtamin D and vitamin D forms. Since the [1,7]-sigmatropic shift which is involved in the intramolecular conversion of previtamin D to vitamin D secosteroids is reversible, a11 vitamin D secosteroids are therefore in equilibrium with their partner previtamin form; the rate of inter- change between the two partner structures (which involves the move- ment of a proton from carbon-19 to carbon-9) is dependent upon the temperature. Thus the equilibrium ratio of 1,25-(OH)~D~:1,25-(0H),- pre-Do is 94:6 at 37 "C. In the case of the pentadeuteriated analogs, the intramolecular shift of the deuterium atom occurs more slowly at any given temperature because of a primary deuterium kinetic isotope effect. Thus the rate of interchange is slowed but the final equilibrium ratio for the two 1,25-(OH)2-&-D3:1,25-(OH)z-&-pre-D3 is the same within experimental error. The Riverside vitamin D compound code names (one or two letters) are given in parentheses; for example, the natural hormone 1,25-(OH),D3 is also referred to as compound C.

13814 Structure-Function Studies of 1,25-(OH)z-previtamin D3

and thus can function as an analog only of the 6-s-cis form of 1,25-(OH)2D3 (Fig. 1).

Nongenomic Actions-Fig. 3A presents an evaluation of the relative ability of 1,25-(OH)zD3 and 1,25-(OH)z-&-pre-D3 to stimulate the nongenomic biological response of transcalta- chia. Vascular perfusion with the physiological concentration of 60 pM 1,25-(OH)2-&-pre-D3 for 34 min yielded a 4.5-fold increase in 45Ca2+ transport over control levels. The stimula- tory effect of both secosteroids on 45Ca2+ becomes significant within 2-8 min as observed previously (7-9). Fig. 3B shows the dose-response relationship for each secosteroid in terms of its ability to stimulate transcaltachia. The analog 1,25- (OH)2-&-pre-D3 was able to stimulate transcaltachia signifi- cantly at a dose of 10 phi, and the maximal response was

A

0 1 ,25(OH&D,

1 .25(OHk Q-pre-D3

I I I I 1 I I I 1 1 1 1 I I I I 1 ,

0 2 4 6 8 10 12 14 16 18 20 22 24 28 28 30 32 34

TIME (rnin.)

B tt.

f f .*

10 25 Bo 200 400 pM 25 60 130 650 1 3 W 6500 pM

1 ,25(OHk-4-pre-D, 1 .25(OH),D,

FIG. 3. Panel A, effect of 1,25-(OH)2-&-pre-D3 and 1,25-(OH)& on the appearance of ‘Ta2+ in the venous effluent of perfused duodena from vitamin D-replete chicks. Each duodenum, filled with 45Ca*+ (5 pCi/ml) in GBSS was vascularly perfused (25 “C) for the first 20 min with control medium (GBSS containing 0.125% bovine serum albu- min and 0.05 p1 of ethanol/ml) and then at time zero either 60 pM 1,25-(OH)2-&-pre-D3, 60 p~ 1,25-(OH)z-&-pre-D3, or control me- dium. Values are the mean f S.E. for five duodena within each experimental group. 0, 1,25-(OH)zD3; A, 1,25-(OH)Z-&-pre-D3; 0, control. Panel E , dose-response analysis of 1,25-(OH)2D3 and 1,25- (OH)2-&-pre-D3 in stimulating transcaltachia in the perfused duo- dena. The experimental conditions were as described in panel A. Values are the mean f S.E. at 30 min for three to five duodena perfused with each concentration of agonist. *p < 0.02; **p < 0.01; ***p < 0.005 with respect to duodena not exposed to agonist.

attained at 60 PM secosteroid. As had been noted before for all secosteroids that have been evaluated with regard to their transcaltachic activity, the dose-response for 1,25-(OH)2-&- pre-D3 is biphasic (7-9). 1,25-(OH)2D3 is active at the low concentration of 25 pM, and the maximum stimulation is achieved over the range of 60-650 PM 1,25-(OH)2D3 (Fig. 3B). Also the typical biphasic dose response is apparent. It is apparent from Fig. 3B that there is no significant difference in the potency of analog 1,25-(OH)2-&-pre-D3, which is locked in the pre-form with that of 1,25-(OH)2D3, which suggests that the signal transducing agent (receptor?) for transcalta- chia is capable of responding to a secosteroid in the 6-s-cis conformation.

Fig. 4 presents an evaluation of the ability of 1,25-(OH)2- &-pre-D3 to stimulate 45Ca2+ uptake into ROS 17/2.8 cells. As originally described by Caffrey and Farach-Carson (15), this response occurs as a consequence of the ability of 1,25- (OH)2D3 or its analogs to open dihydropyridine-sensitive Ca2+ channels via a nongenomic mechanism. The concentration range of 1-10 x lo-’ M 1,25-(OH)+&-pre-D3 produced a maximum uptake of 45Ca2+ within 1 min of the application of the secosteroids. Previous studies have established that this is the range of maximum response to 1,25-(OH)2D3 (15, 17).

The voltage-gated Ca2+ channel can either be opened by exposure to appropriate agonists (vitamin D analogs or the dihydropyridine BAY K-8644 (15)) or by depolarization of the cell membrane, which is achieved by the 132 mM external KC1 (stimulating buffer; see “Experimental Procedures”). Such stimulated uptake of 45Ca2+ in the presence of depolar- izing extracellular solutions is characteristic of cells express- ing voltage-gated Ca2+ channels, and the level of this stimu- lation is directly related to the concentration of the Ca2+ channels on the cell surface. Thus, the maximum influx of “Ca2+ which can be achieved (see inset of Fig. 4) occurs in the presence of high external K+. The level of 45Ca2+ uptake occurring in low K+ represents the basal uptake, which is a reflection of the Ca2+ permeability of the resting membrane. In growth phase ROS 17/2.8 cells the maximum ratio of (stimulated)/(resting), S/R is approximately 2.4-fold, which

’t FIG. 4. Evaluation of 45Caa+ uptake in osteosarcoma cells

stimulated by 1,26-(OH)zDs or 1,2S-(OH)a-ds-pre-Ds. The left inset displays the 45Ca2+ uptake properties of the cells in resting buffer ( R ) , stimulating buffer (S), and when exposed to the optimal concen- tration of 1,25-(OH)*D3, designated as C. The ROS 17/2.8 cells were assayed for 45Ca2+ as described under “Experimental Procedures.” In

the resting buffer. 1,25-(OH)zD3 stimulated 45Ca2+ uptake with a all cases, additions of the secosteroids were made to cells exposed to

maximum response occurring between 1.0 and 5.0 nM (see Ref. 16). Data points represent the mean of triplicate measurements f S.D.

Structure-Function Studies of l,25-(OH)2-previtamin D3 13815

represents 100% Ca2+ channel opening. Both the analog 1,25- (OH)2-&-pre-D3 and 1,25-(OH)zD3 achieve a %fold stimula- tion of 4sCaz+ uptake over that which occurs in the low K+ environment. Again there is no significant difference in the biological activity of 1,25-(OH)z-&-pre-D3, which is locked in the previtamin form as compared with the 1,25-(OH)zD3, which has a facile interchange between the steroid-like con- formation and the extended steroid conformation.

Ligand Binding Studies-In Fig. 5 is presented the deter- mination of the RCI for binding to the intestinal nuclear 1,25- (OH)zD3 receptor from both the chick and pig, as determined under in vitro conditions. The calculated RCI for the chick and pig, respectively, is as follows: 1,25-(OH)zD3 (RCI = 100% and loo%), 1,25-(OH)z-&-D3 (RCI = 92% and 67%), and for 1,25-(OH)z-&-pre-D3 (RCI = 10% and 4%). Thus there is no effect of species difference in binding of these three secoster- oids to the intestinal nuclear receptor. The steroid that is kinetically repressed in the previtamin form, the 1,25-(OH)2- &-pre-D3, has a reduction in its RCI from 90-100% to 12- 14%, which suggests that the nuclear 1,25-(OH)zD3 receptor can discriminate between the previtamin 6-s-cis and the vi- tamin D form, which may exist either as the 6-s-cis or 6-s- trans forms, with the latter predominating.

The principal carrier of vitamin D secosteroids in the blood compartment is the plasma DBP. This protein has a binding domain that tightly binds its ligand with a Kd of 5 x lo-’ M and 5 x lo-’ M for 25-(OH)D3 and 1,25-(OH)& respectively (28); thus the affinity of any ligand for DBP will effectively determine its “free” concentration in the plasma and perhaps influence its relative availability to target cells. In Fig. 6 is presented the competition curves of 1,25-(OH)& 1,25- (OH)z-&-D3, and 1,25-(OH)z-&-pre-D3, respectively, with [3H]1,25-(OH)2D3 for binding to human DBP as determined under in vitro conditions. The calculated RCI for 1,25- (OH)ZD~ is loo%, 95% for 1,25-(OH),-&-D3, and 7% for 1,25-

FIG. 5. Panel A , affinity of the 1,25- (OH)ZD3 nuclear receptor from pig intes- tinal mucosa for 1,25-(OH),D3 (e), 1,25- (OH)2-&-pre-D3 (A), and 1,25-(OH),-&- DB (A). For details see “Experimental Procedures.” Data presented are mean & S.D. of three experiments. Panel B, de- termination of the RCI for the chick intestinal receptor and the pig receptor for 1,25-(OH),D3 (analogc), 1,25-(OH),- &-pre-DB (analog HF), and 1,25-(OH),- &-Da (analog HG). The RCI values are indicated in the figure and represent the mean and where available & S.E. (n = 5) and were calculated as described un- der “Experimental Procedures.”

(OH)z-&-pre-D3. This result suggests that the DBP ligand binding domain has a preference for the extended (6-s-trans conformation) steroid conformation of vitamin D3 secoster- oids as compared with the steroidal (6-s-cis) conformation. Thus freezing 1,2E~-(OH)~-&-pre-D3 in the previtamin form reduces the RCI from 100 to 7%. This reduced affinity for DBP means effectively that there will be a higher free con- centration of 1,25-(OH)z-&-pre-D3 as compared with 1,25- (OH)2D3 or 1,25-(OH)~-&-D3.

Genomic Actions-Figs. 7-10 report an evaluation of the biological efficacy of the previtamin form of the pentadeuter- iated analog of 1,25-(OH)zD3 under in uiuo conditions as well as in cultured cells. Fig. 7 reports the levels of serum osteo- calcin which are achieved after a single intramuscular injec- tion of vitamin D-deficient chick with 400 ng of either 1,25- (OH)zD3, 1,25-(OH)~-&-D3, or 1,25-(0H)&-pre-D3. 1,25- (OH)zD3 has been shown to induce via interaction with a nuclear 1,25-(OH)zD3 receptor present in bone osteoblast cells the de nouo biosynthesis of osteocalcin; small amounts of the osteocalcin are released into the blood (as a consequence of bone remodeling) where it may be conveniently determined via a radioimmunoassay (27). It is apparent that the 1,25- (OH)z-&-pre-D3, when administered as a single dose under in vivo conditions, has little ability to interact effectively with the nuclear 1,25-(OH)zD3 receptor to induce osteocalcin. In contrast, both 1,25-(OH)zD3 and 1,25-(OH)2-d6-D3 caused a significant increase in the plasma levels of osteocalcin. Al- though identical doses of 1,25-(OH)zD3 and 1,25-(OH)z-&-D3 were administered, treatment with the deuteriated analog resulted in a consistently higher induction of the plasma osteocalcin levels. This conceivably could be a result of a slowed rate of catabolism or metabolic clearance of the deu- teriated analog in comparison with the proteo form of 1,25- (OHIzD3.

Fig. 8 reports the levels of serum Ca2+ and osteocalcin

- HG 92+7 HF l&6

I I I

1 2 3 [analog] / [3H-l. 25 (OH) 2D,l

CONCENTRATION (M)

13816 Structure-Function Studies of 1,25-(OH)2-previtarnin D3

0‘ , I 1 O 1 O 10.0 10.8 1 0 ’ 10

CONCENTRATION (M)

FIG. 6. Affinity of purified human DBP for l,!&i-(OH)d& (0), 1,25-(OH)z-d6-pre-D3 (A), and 1,25-(OH)z-da-D3 (A). For details see “Experimental Procedures.” Data presented are mean f S.D. (n = 2).

I - 1 200

200 200 ~~

0 0 0 5 10 15 20 25 30 35 40

Hours

FIG. 7. Effect of 1,26-(OH)zD3 and its analogs on the serum concentration of osteocalcin after a single 400-ng intramus- cular injection in vitamin D-deficient chicks. e, 1,25-(0H)&; A, 1,25-(OH)p&-pre-D3; A, 1,25-(OH)*-&-D3. For details see “Exper- imental Procedures.” Values represent mean f S.E. (eight birds/ group). The experiment was repeated a second time with equivalent results.

achieved after 1 week of daily treatment with doses of 1,25- (OH)2Da or the analogs 1,25-(OH)z-&-pre-D3 and 1,25-(OH)2- &-D3. Both 1,25-(OH)&3 and 1,25-(OH)z-&-D3 were virtually equipotent in regard to elevation of serum Ca2+ and osteocal- cin. In contrast 1,25-(OH)z-&-pre-D3 was only able to elevate serum Ca2’ at a dose (10 pg/kg/day) that was 20 times higher than the dose of 1,25-(OH)zD3 and 1,25-(OH)z-&-D3 ( 5 pg/ kg/day), which achieved a significant elevation of these pa- rameters above the base line. An even higher dose of 1,25- (OH)*-&-pre-D3 (50 wg/kg/day) was required to elevate the osteocalcin levels significantly above that of the control group. Since a single dose of l,25-(OH)2-&-pre-D3 did not elevate serum CaZ+ in chicks (Fig. 7), it is likely that the elevation of serum Ca” and osteocalcin which resulted in the mice has occurred as a consequence of the slow thermal conversion, in vivo, of 1,25-(OH)z-&-pre-D3 to 1,25-(OH)z-&-D3.

1,25-(OH)2D3 and its two analogs were evaluated in MG-63 cells with respect to inhibition of proliferation as assessed by [3H]thymidine incorporation (Fig. 9A) and induction of hu- man osteocalcin (Fig. 8B). Both of these responses are me- diated by the nuclear 1,25-(OH)zD3 receptor. In both assays,

0 .05 .1 .2 .5 .05 .1 .2 .5 .5 2 10 50 pg/kg/day

1 .25(0W2D3 1 ,25(0H)2-diD3 1 .25(OH);d5-pre-D.

300 l B

z 2 100/ cn w

300

250

lo 200

150

100

50

0 .05 .1 .2 .5 .05 .1 .2 .5 .5 2 10 50pg/kg/day

1 25(OH)2D3 1 .25(OH)2-d;D3 1 ,25(0H);d5-pre-D.

FIG. 8. Effect of chronic administration to mice of 1.26- (OI&DS and its analogs 1,25-(OH)z-ds-pre-D3 and 1.26- (OH)z-da-Ds on serum Ca” (panel A) and osteocalcin (panel B ) . The mice were administered the indicated daily intraperitoneal dose of the indicated secosteroid for 7 days. Values represent the mean f S.E. (six mice/group). *p < 0.01 in comparison with the control group.

the lr25-(OH)2-&-pre-D3 had only approximately 1% of the activity of 1,25-(OH)zD3 or 1,25-(OH)z-d5-D3. In addition, the 1,25-(OH)z-&-D3 was indistinguishable from 1,25-(OH)zD3 in its ability to induce osteocalcin and displayed approximately 90% of the activity of 1 , 2 f ~ - ( o H ) ~ D ~ in terms of its ability to inhibit cell proliferation. These results are consistent with the interpretation that the previtamin form of 1,25-(OH)zD3 is not able to efficiently interact with the nuclear 1,25-(OH)& receptor.

The differentiation of HL-60 cells was markedly enhanced by the presence of 1,25-(OH)zD3 or 1,25-(OH)z-ds-D3 (Fig. 10). In contrast the 1,25-(OH)z-&-pre-D3 displayed only 1- 4% of the potency of 1,25-(OH)& which again suggests that the HL-60 nuclear 1,25-(OH)& receptor does not efficiently bind this ligand.

DISCUSSION

In this report we are principally comparing the biological profile of 1,25-(OH)2-&-pre-D3 with that of the pair of rapidly

Structure-Function Studies of 1,25-(OH)2-previtamin D3 13817

- 100

- so

- 60

- 40

- 20

5 B l 400 i a

g 200 c

/ / I 1 o-sx 70. 1;. 1 0 7

lo

CONCENTRATION (M)

FIG. 9. Effect of ~ , ~ F P ( O H ) ~ D ~ and its analogs in MG-63 cells on the inhibition of proliferation (paneZA) and induction of osteocalcin (panel B ) . 0, 1,25-(OH)&; A, 1,25-(OH)2-Q-pre- DS; A, 1,25-(OH)~-&-D3. For details see “Experimental Procedures.” The data presented are from a representative experiment; a total of three experiments was conducted.

interconverting 6-s conformers of 1,25-(OH)zD3 (Figs. 1 and 2). Our results demonstrate that two nongenomic biological systems are fully responsive to the 1,25-(OH)2-&-pre-D3 an- alog. Both the process of transcaltachia as studied in the isolated perfused chick duodenum (Fig. 3) (7, 8) and the process of Caz+ channel opening in the rat osteogenic sarcoma cell line, ROS 17/2.8 cells (Fig. 4) (15), respond with equiva- lent potency to both 1,25-(OH)z-&-pre-D3 and 1,25-(OH)~D3. Thus the responsiveness of the signal transduction process for these two nongenomic systems occurs in two species, the rat and chick, and in two different vitamin D target organs, the intestine and bone. In both systems there is evidence that the biological response involves the opening of voltage-sensi- tive Caz+ channels that are located in the outer cell membrane (15, 18). It has been postulated for both systems that the signal transduction process that results in the opening of the Ca2+ channel involves a putative membrane receptor for 1,25- (OH)2D3. There is preliminary evidence in the transcaltachic system that tritiated 1,25-(OH)zD3 binds specifically, in a saturable manner, to an intestinal basal lateral membrane system that may be a component of the transcaltachia signal transduction process (31). Further, in both the ROS 17/2.8 cell system and in the perfused intestinal transcaltachic sys- tem, an evaluation has been made of a series of analogs with differing structural modifications of the reference compound,

CONCENTRATION (M)

FIG. 10. Effect of 1,25-(OH)~Ds and its analogs on differ- entiation of HL-60 cells. The differentiating effect was evaluated by nitro blue tetrazolium (NBT) reduction. 0, 1,%-(OH)zD3; A, 1,25- (OH)a-Q-pre-D3; and A, 1,25-(OH)~-&-D3. For details see “Experi- mental Procedures.” The data presented are from a representative experiment that was repeated twice.

1,25-(OH)&. Evidence was obtained for two classes of ana- logs: those that bind effectively to the 1,25-(OH)*D3 nuclear receptor but which are ineffective at opening Caz+ channels, and those analogs that are effective in stimulating the opening of Ca2+ channels but which bind poorly to the 1,25-(OH)& nuclear receptor (17, 18). These results have been interpreted as suggesting the existence of two forms of the 1,25-(OH)2D3 receptor: one present in the nucleus/cytosol concerned with genomic responses, and a second species present in the plasma membranes of some cells, which are involved in some fashion with nongenomic biological responses related to 1,25- (OH).& Further support for two receptors for 1,25-(OH)& has come from studies with l@,25-(OH)zD3; this analog binds very poorly to the nuclear 1,25-(OH)~D3 receptor (RCI = 0.01), and although it is devoid of agonist activity in tran- scaltachia, it has been found that lfi,25-(OH)zD3 is a potent antagonist of 1,25-(OH)2D3 stimulated transcaltachia?

In these studies we have employed the 1,25-(OH)z-&-pre- D3 as an analog of the 6-s-cis conformer of 1,25-(OH)2D3. But it should be recognized that there are three important struc- tural differences between 1,25-(OH)z-&-pre-D3 and 1,225- (OH)&. (a) In 1,25-(OH)z-&-pre-D3, the triene system is 6 5-10, 6-7, 8-9 rather than the 10-19, 5-6, 7-8 triene system present in 1,25-(0H)&,. ( b ) In 1,25-(OH)zD3 carbon-19 is present as a methylene, whereas in 1,25-(OH)z-&-pre-D3 it is present as a methyl group. (c) The 1,25-(OH)2-&-pre-D3 has five deuterium atoms in place of protons. To some extent, the impact of the presence of the five deuterium atoms can be evaluated by comparing the biologic profile of l,%-(OH)z-&- D3 with 1,25-(OH)&. As shown by the collective results in Figs. 5-10 1,25-(OH)2-&-D3 and 1,25-(OH)zD3 are biologically equivalent.

The principal carrier of vitamin D secosteroids throughout the body is the plasma DBP. This protein transports the vitamin forms of vitamin D3, 25-(OH)D3, 1,25-(OH)2D3r and 24R,25-(OH)2D3 throughout the plasma compartment (21, 22). To date no detailed study of the ability of DBP to

A. W. Norman, I. Nemere, K. R. Muralidharan, and W. H. Okamura, submitted for publication.

13818 Structure-Function Studies of 1,25-(OH)2-previtamin D3

transport the previtamin form of 1,25-(OH)zD3 has been reported. Further, it is of importance to learn whether the ligand binding domain of DBP prefers the steroid-like con- formation (6-s-cis conformer) or the extended steroid confor- mation (6-s-trans conformer). When the data of Fig. 5 are utilized to calculate the RCI, the results indicate that 1,25- (OH)z-&-pre-D3 has an RCI that is only 7% of 1,25-(OH)&, whose RCI is 100%. Thus we tentatively conclude that the ligand binding domain of DBP does not prefer the 6-s-cis steroid-like conformation over that of the 6-s-trans extended steroid conformation of 1,25-(OH)zD3. In related studies a similar result with respect to poor binding to DBP has been noted for a 19-nor analog of 1,25-(OH)zD3, namely 1,25- (OH)2-19-nor-previtamin D3; the RCI of this analog was about 5.* It seems likely that the same relationship would be oper- ative for the other vitamin D metabolites and parent vitamin D3 which are transported by DBP.

Biological Evaluation in Vivo and i n Whole Cells-The ability of the analog l,25-(OH)+&-pre-D3 to interact with the nuclear 1,25-(OH)2D3 receptor and support genomic responses was extensively evaluated. Fig. 6 indicates that both the pig and chick intestinal 1,25-(OH)zD3 nuclear receptors discrim- inate against the previtamin form of the secosteroid. The RCI of 1,25-(OH)z-&-pre-D3 for binding to the chick intestinal receptor was 10% and for the pig intestinal receptor was 4%. Thus we may conclude that the nuclear 1,25-(OH)zD3 recep- tor’s ligand binding domain favors the 6-s-trans conformer (extended steroid conformation) over the 6-s-cis (steroid-like conformation); however, it would be desirable to study an analog of 1,25-(OH)zD3 which is “locked in the 6-s-trans conformation. Further we found that the presence of the five deuterium atoms on 1,25-(OH)zD3 as in 1,25-(OH)z-&-D3 did not significantly alter the RCI (89.5% for chick and 67% for the pig).

The relative inability of the nuclear 1,25-(OH)zD3 receptor to effectively bind the 6-s-cis (steroid-like conformation) of 1,25-(OH)zD3, as studied by the analog, 1,25-(OH)Z-&-pre-D3, implies that the latter compound would not be an efficient mediator of genomic responses. Figs. 7-10 report the evalua- tion of 1,25-(OH)z-&-pre-D3, 1,25-(OH)z-&-D3, and 1,25- (OH),D3 in four systems that all generate biological effects via a nuclear receptor-mediated regulation of gene transcrip- tion. Fig. 7 reports the in vivo measurement of serum osteo- calcin levels, after a single dose of 1,25-(OH)z-&-pre-D3 to chicks. Fig. 8 reports the results of daily administration of 1,25-(OH)z-&-pre-D3 on serum Ca2+ and osteocalcin levels in mice. Fig. 9 reports, using MG-63 cells in cell culture, the inhibition of cell proliferation and the induction of osteocal- cin. Fig. 10 reports, using HL-60 cells in culture, the inhibition of cell proliferation. In each system there is extensive evidence implicating the functional involvement of the nuclear 1,25- (OH)2D3 receptor (27,29,30). The relative inability of analog 1,25-(OH)z-&-pre-D3 (less than 2% for osteocalcin induction i n viuo, Fig. 7; less than 1% for inhibition of proliferation in MG-63 cells, Fig. 9A; approximately 2% for osteocalcin in- duction in MG-63 cells, Fig 9B; and approximately 10% in promoting differentiation of HL-60 cells), in comparison with 1,25-(OH)2D3 and 1,25-(OH)z-&-D3, to effect the appropriate generation of a genomic response is consistent with the data of Fig. 6 reporting its low RCI for the nuclear receptor.

In addition, the very low ability of 1,25-(OH)z-&-pre-D3, when administered daily for 7 days to mice (Fig. 8) to elevate serum Ca2+ and osteocalcin levels, supports the conclusion

‘ R. Bouillon, L. A. Sarandeses, K. Allewaert, J. Zhao, J. L. Mas- carenas, A. Mourino, s. Vrielynck, P. DeClercq, and M. Vandewalle, submitted for publication.

that 1,25-(OH)z-&-pre-D3 per se has a low ability to activate these genome-dependent responses. These results also imply that the previtamin form, 1,25-(OH)&,-pre-D3, is not subject to some metabolic transformation that converts it to the vitamin form, i.e. 1,25-(OH)2-&-D3. The time duration of each of these assays was 40 h (chicks, in Fig. 7); 7 days (mice, in Fig. 8), and 96 h (MG-63 cells, in Fig. 9; HL-60 cell, in Fig. lo), and if the vitamin form, 1,25-(OH)z-&-D3 had been generated there would have been ample time for appearance of detectable manifestations of the various genomic responses. In fact the results from the chronic dosing of mice with 1,25- (OH)z-&-pre-D3 (Fig. 8) suggest that the previtamin form is subject to metabolic clearance before it has had an opportu- nity to isomerize thermally into the biologically active 1,25- (OH)z-&-D3.

This article addresses the issue of whether there are differ- ences in activity of the 6-s-trans (extended steroid confor- mation) and the 6-s-cis (steroid-like conformation) of 1,25- (OH)2D3 (see Fig. 1B) with respect to the initiation of genomic and nongenomic biological responses. The interconversion between the 6-s-trans and 6-s-cis forms occurs in solution millions of times per s at room temperature; as yet the true equilibrium ratio of the 6-s-trans and 6-s-cis conformers has not been rigorously determined. As indicated in the Introduc- tion, because of the facile interconversion of the 6-s-trans and 6-s-cis conformers of 1,25-(OH)zD3, there will be kinetically competent amounts of both conformers present to interact with any potential receptor for 1,25-(OH)zD3.

It has been possible through the chemical synthesis of the pentadeuteriated forms of the previtamin and vitamin forms of 1,25-(OH)zD3 to effectively kinetically suppress the 1,25- (0H).&,-pre-D3 into its vitamin form since the 1,2f~(OH)~- &-pre-D3 only isomerizes to 1,25-(OH)z-&-D3 at 37 “C with a half-life of 81 h (23); see Fig. 2. This contrasts with the 13.4- h half-life of interconversion of 1,25-(OHh-pre-D3 into 1,25- (OH)ZD3, at 37 “C. Given the more rapid interconversion of 1,25-(OH)2-pre-D3 into 1,25-(oH)zD3, it has not proven fea- sible to evaluate differences in intrinsic biologic activity of these two species since many of the i n vivo assays that involve genomic responses require 8-16 h at a minimum. The presence of the deuterium atoms on carbon-9 and carbon-19 effectively slows the [ 1,7] -sigmatropic shift which is integrally associated with the previtamin-vitamin interconversion; Curtin and Okamura (21) have reported the existence of a significant primary kinetic isotope effect for this process of K h l K d of about 6 at 37 “C.

We have reported previously (20) that the family of vitamin D compounds as a consequence of being a secosteroid, displays considerable conformational mobility of the A ring, such that there is a dynamic interchange between two chair conformers. This suggested that the ligand binding domain of receptors for 1,25-(OH)zD3 might prefer one chair conformation that optimized the orientation of the l-a hydroxyl (19). AS a consequence of our present studies, it is apparent that there may be additional important biological implications relating to the rotational flexibility around the 6-7 carbon-carbon single bond. The rotational flexibility around the 6-7 carbon- carbon bond has the consequence of generating a population distribution of conformers present in solution, all potentially able to interact with target organ receptors and the plasma transport protein, DBP. This contrasts markedly with the other steroid hormones, e.g. estrogen, progesterone, and glu- cocorticoids, which essentially have only a single conforma- tion available to interact with their cognate plasma transport proteins and target organ receptors.

On the basis of the data presented in this article there

Structure-Function Studies of 1,25-(OH)2-previtamin D3 13819

appear to be two classes of 1,25-(OH)2D3 receptors: (a) those involving nongenomic responses that can utilize as agonists both the previtamin D and vitamin D forms of 1,25-(OH),D3, and ( b ) those involving genomic responses that cannot re- spond to the previtamin form. Our data do not allow us to determine whether the nongenomic class of receptors re- sponds only to the 6-s-cis (steroid-like) conformation, but this is certainly a possibility. Thus it might prove feasible to synthesize analogs chemically which were agonists or antag- onists for only one class of 1,25-(OH)2D3 receptors. Further examination of this interesting possibility is dependent upon the chemical synthesis of appropriate vitamin, as opposed to previtamin, analogs that are locked in both the 6-s-cis or 6-s- trans conformations. We are currently conducting further studies to explore these various possibilities. Collectively these studies demonstrate the complexity of the vitamin D endocrine system with respect to the structure-function rela- tionships operative with regard to the ligand binding domain of the 1,25-(OH)zD3 receptors.

Acknowledgments-We thank Dr. Thomas Thysseril for skillful perfusion of the chick intestine in the transcaltachia assay, June E. Bishop for multiple contributions to this project, and Dr. Michael L. Curtin for providing starting materials for the chemical synthesis of 1,25-(0H)z-&-pre-D3.

REFERENCES 1. Reichel. H., Koeffler. H. P.. and Norman. A. W. (1989) N. End. J. Med.

320,'980-991 2. Norman, A. W. (1992) J. Cell. Biochem. 4 9 , l - 3 3. Haussler. M. R.. and Norman. A. W. (1969) Proc

62,155-162 ' ,~ I

4. Pike, J. W. (1991)Annu. Reu. Nutr. 11 , 189-216 5. Lowe, K. E., Maiyar, A. C., and Norman, A. W.

Gene EXD. 2.65-109

. Natl. Acad.

(1992) Crit.

Sei. U. S. A.

Reu. Euhnr.

6. Minghetti,'P. P., and Norman, A. W. (1988) FASEB J. 2,3043-3053

7. Nemere, I., Yoshimoto, Y., and Norman, A. W. (1984) Endocrinology 115 ,

8. Yoshimoto, Y., Nemere, I., and Norman, A. W. (1986) Endocrinology 118 ,

9. Yoshimoto, Y., and Norman, A. W. (1986) J. Steroid Biochem. 2 5 , 905-

10. deBoland, A. R., Nemere, I., and Norman, A. W. (1990) Biochem. Biuphys.

11. Lieberherr, M., Grosse, B., Duchambon, P., and Driieke, T. (1989) J. Biol.

12. Baran, D. T., Sorensen, A. M., Honeyman, T. W., Ray, R., and Holick, M.

13. Bourdeau. A,. Atmani. F.. Grosse. B.. and Lieberherr. M. (1990) Endocri-

1476-1483

2300-2304

909

Res. Commun. 166,217-222

Chem. 2 6 4 , 20403-20406

F. (1989) FEBS Lett. 2 5 9 , 205-208

14. Tsutsumi, M., Alvarez, U., Avioli, L. V., and Hruska, K. A. (1985) Am. J.

15. Ca!<Fx_y: J. M., and Farach-Carson, M. C. (1989) J. Bid. Chern. 264,20265-

nology i27,2738-2743

Physiol. 2 4 9 , F117-F123

16. Civitelli, R., Kim, Y. S., Gunsten, S. L., Fujimori, A., Huskey, M., Avioli,

17. Farach-Carson. M. C.. Sereeev. I.. and Norman. A. W. (1991) Endocrinolozv

Z W l 4

L. V., and Hruska, K. A. (1990) Endocrinology 127 , 2253-2262

129,1876-1884 . . , . .

18. Zhou, L.-X., Nemere, I., and Norman, A. W. (1992) J. Bone Min. Res. 7, Ari-i-AG2

19. Okamura, W. H., Norman, A. W., and Wing, R. M. (1974) Proc. Natl. Acad.

20. Wing, R. M., Okamura, W. H., Pirio, M. R., Sine, S. M., and Norman, A.

.-. ."" Sci. U. S. A . 71,4194-4197

W. (1974) ScLence 196.939-941 21. Curtin, M.-L., andOkamLra, W.H. (1991) J. Am. Chem. Soc. 113 , 6958-

6966 22. VanBaelen, H., and Bouillon, R. (1986) in Binding Proteins of Steroid

Hormones (Forest, M. G., and Pugeat, M., eds) pp. 69-83, John Libbey

23. Vanham, G., VanBaelen, H., Tan, B. K., and Bouillon, R. (1988) J. Steroid Eurotext, Paris

Biochem. 29,381-386 24. Ma'eska, R J , and Rodan, G. A. (1982) J. Biol. Chem. 2 5 7 , 3362-3365 25. Sieiert, P. D.; Ohnuma, N., and Norman, A. W. (1979) Biochem. Biophys.

Res. Commun. 91,827-834 26. Wecksler, W. R., and Norman, A. W. (1980) Methods Enzyrnol. 6 7 , 494-

500 27. Bouillon, R., Allewaert, K., van Leeuwen, J. P. T. M., Tan, B.-K., Xiang,

D. Z., De Clercq, P., Vandewalle, M., Pols, H. A. P., Bos, M. P., Van

28. Bouillon, R., Van Baelen, H., and De Moor, P. (1980) J. Steroid Biochem. Baelen, H., and Birkenhager, J. C. (1992) J. Biol. Chem. 267,3044-3051

29. Zhou. J.. Norman. A. W.. Lubbert. M.. Collins. E. D.. Ilskokovic. M. R.. 13, 1029-1034

and K'oeffler, H: P. (1989) Blood~74,'82-93 30. Koeffler, H. P., Bishop, J. E., Reichel, H., Singer, F., Nagler, A,, Tobler,

A,, Walka, M., and Norman, A. W. (1990) Mol. Cell. Endocrinul. 70, 1-

I~ I ~ ~~~ ~~ ~~ , ~~I

1 1 31. Nemere, I., and Norman, A. W. (1992) Mol. Biol. Cell3, (suppl.) 297 (abstr.)