Functional Characterization and Purification of an Intracellular Vitamin D-binding Protein in Vitamin D-resistant New World Primate Cells
AMINO ACID SEQUENCE HOMOLOGY WITH PROTEINS IN THE HSP-70 FAMILY*

(Received for publication, July 12, 1996, and in revised form, November 22, 1996)

Mercedes A. Gacad , Hong Chen , Jonathan E. Arbelle , Thomas LeBon Dagger and John S. Adams §

From the Cedars-Sinai Burns and Allen Research Institute, UCLA School of Medicine, Los Angeles, California 90048, and the Dagger  Beckman Research Institute, City of Hope Hospital and Medical Center, Los Angeles, California 92714

ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES


ABSTRACT

Most genera of New World primates exhibit resistance to vitamin D. These monkeys harbor high circulating concentrations of the prohormone 25-hydroxyvitamin D and the active vitamin D hormone 1,25-dihydroxyvitamin D. Previous work from this laboratory indicated that resistance is associated with the overexpression of a 60-65-kDa intracellular protein that binds vitamin D metabolites competitively. In the current studies 25-[3H]hydroxyvitamin D3 (25-OHD3) was used as a competitive ligand to investigate the ability of a number of small lipid molecules to interact with this intracellular vitamin D-binding protein (IDBP) in post-nuclear extracts of a prototypical lymphoblast cell line from the common marmoset, a vitamin D-resistant New World primate. Only those vitamin D metabolites with a hydroxyl moiety in the C-25 position were bound by IDBP. Disruption of the C-25 hydroxyl obviated binding, whereas more proximal alterations in the vitamin D side chain did not. Modifications in the A-ring of 25-hydroxylated vitamin D metabolites, most specifically hydroxylation of C-1, diminished but did not abolish ligand binding. Of more than two dozen other small lipid molecules examined, only the C-19 17-hydroxysteroids, 17beta -estradiol and testosterone, and the C-21 steroid progesterone were found to be capable of binding specifically to IDBP. Using a combination of physical and serial chromatographic techniques, we enriched IDBP 25-OHD3 binding activity 17,588-fold in extracts of B95-8 cells. Two-dimensional sodium dodecyl sulfate-polyacrylamide gel electrophoresis of this purified fraction demonstrated a predominant 65-kDa molecular species with a pI ~ 4.5. Seven different peptide fragments were isolated from the 65-kDa protein, each possessing sequence similarity to the hsp-70 family of proteins. Ligand binding analyses confirmed that human inducibly expressed hsp-70-bound 25-OHD3 with approximately similar affinity (~10-7 M) as did purified IDBP. In summary, these results suggest a novel action for the hsp-70 family of proteins as intracellular vitamin D- and gonadal steroid hormone-binding molecules.


INTRODUCTION

With the notable exception of nocturnal primates in the genus Aotus (1), New World primates (NWP)1 exhibit vitamin D resistance (1, 2). This resistant state is characterized biochemically by high serum concentrations of the two major circulating vitamin D metabolites, 25-OHD and 1,25-(OH)2D (1-3), and clinically by rickets in rapidly growing adolescent animals deprived of adequate sunlight exposure (4). Levels of 25-OHD may be up to 10-fold and 1,25-(OH)2D up to 100-fold greater than those observed in Old World primates (OWP), including man. Unlike the majority of resistant states described for other steroid hormones and vitamin D in Homo sapiens (5, 6), resistance in NWP does not appear to be related to a mutation in the vitamin D receptor (VDR) protein (1, 7). Rather, the vitamin D-resistant state in NWP is associated with the apparent overexpression of an intracellular vitamin D-binding protein (IDBP) (1), which is distinct from members of the serum vitamin D-binding protein DBP/albumin families of proteins (8, 9). Unlike the cysteine-rich sterol/steroid-binding proteins in serum vitamin D-binding protein/albumin (9) and vitamin D/steroid receptor protein (10) families, which are confined prinicipally to the extracellular domain and nucleus, respectively, IDBP is a relatively abundant, cysteine-poor protein concentrated in, but not confined to, the cytoplasmic compartment of cells (11). However, like these other sterol/steroid-binding proteins (8, 9), IDBP has been shown preferentially to bind 25-hydroxylated vitamin D metabolites (12). In the current report we have greatly expanded the vitamin D metabolite-IDBP binding studies to define more clearly which structural modifications of the vitamin D molecule are important for either enhancing or reducing ligand binding to IDBP extracted from NWP cells. We also document the IDBP binding potential of a much more extensive panel of molecules in an attempt to define additional classes of small, lipid-soluble signaling molecules which may interact with IDBP. Although we were previously able to enrich vitamin D metabolite binding activity in extracts of NWP cells (13), attempts to purify IDBP to homogeneity were unsuccessful. Here we report the purification and structural identification of the NWP IDBP as a member of the hsp-70 family of proteins and demonstrate the sterol/steroid binding properties of hsp-70-like proteins.


MATERIALS AND METHODS

Vitamin D Sterols, Steroids, and Other Compounds

25-[3H]Hydroxyvitamin D3 (25-[3H]OHD3; specific activity, 181 Ci/mmol) and 17beta -[3H]estradiol (specific activity, 154 Ci/mmol) were purchased from Amersham Corp. The source of other compounds evaluated as potential ligands is provided in Table I; three vitamin D analogs and RU486 were provided as gifts. Inducibly expressed, recombinant human heat shock protein-70 (hsp-70) was purchased from StressGen, Victoria, B.C., Canada. All other buffer constituents were from Sigma. Chromatographic supports used in protein purification were purchased from Bio-Rad, Pharmacia Biotech Inc., MetaChem (Torrance, CA), and Millipore (Bedford, MA).

Table I.

Identity, 25-OHD3 binding index, and source of potential IDBP ligands tested


Compounds RBIb Source

Vitamin D-related
  7-Dehydrocholesterol 0 Sigma
  Vitamin D3 0 Sigma
  Vitamin D2 0 Hoffman-LaRoche
  Dihydrotachysterol 0 Sigma
  1-OHD3 0 Hoffman-LaRoche
  25-OHD3 100 Hoffman-LaRoche
  25-OHD2 100 Hoffman-LaRoche
  1,25-(OH)2D3 33 Hoffman-LaRoche
  1,25-(OH)2D2 35 Hoffman-LaRoche
  24,25-(OH)2D3 72 Hoffman-LaRoche
  25,26-(OH)2D3 70 Hoffman-LaRoche
  1,24,25-(OH)3D3 36 Hoffman-LaRoche
  Calcipotriol (MC903) 0 Leo Pharmaceutical
  EB1089 0 Leo Pharmaceutical
  KH1060 0 Leo Pharmaceutical
Steroids
  Pregnenolone 0 Sigma
  Progesterone 19 Sigma
  Hydrocortisone 0 Sigma
  Cortisone 0 Sigma
  Corticosterone (B) 0 Sigma
  Deoxycorticosterone (DOC) 0 Sigma
  Aldosterone 0 Sigma
  DHEA 0 Sigma
  DHEAS 0 Sigma
  Dexamethasone 0 Sigma
  Androstendione 0 Sigma
  Androstendiol 0 Sigma
  Testosterone 30 Sigma
  5-Dihydroxytestosterone 0 Sigma
  17beta -Estradiol 33 Sigma
  Tamoxifen 0 Sigma
  RU486 0 Roussel-U.C.L.A.F.
Other
  Cholesterol 0 Sigma
  25-Hydroxycholesterol 0 Sigma
  Arachidonic acid 0 Biomol
  PGE2 0 Biomol
  PGF2alpha 0 Biomol
  All-trans-Retinoic acid 0 Sigma
  9-cis-Retinoic acid 0 Biomol
  Triiodothyronine 0 Sigma

a DHEA and DHEAS, dehydroepiandrosterone and sulfate form, respectively; PGr, prostaglandin.
b RBI (relative binding index) is the relative ability of the maximally effective concentration of the compound in question to displace competitively 25-[3H]OHD3 from IDBP present in the unfractionated, 100,000 × g supernatant extract of B95-8 cells; of the compounds examined 25-OHD3 was the most effective and was assigned a score of 100.

Cell Culture

The B-lymphoblastoid cell line B95-8 was obtained from the American Type Culture Collection; this cell line was established by Epstein-Barr virus transformation of blood leukocytes from the vitamin D-resistant NWP, Callithrix jacchus (common marmoset). The cell line was maintained in RPMI 1640 medium (Irvine Scientific, Irvine, CA) with 10% fetal calf serum (Gemini BioProducts, Calabasas, CA), 100 units/ml penicillin, 100 µg/ml streptomycin, 2 mM L-glutamine (all supplements from Life Technologies, Inc.) in an atmosphere of 95% air, 5% CO2.

Cell Extracts

Confluent cultures of B95-8 cells were harvested, pelleted, and washed twice in ice-cold phosphate-buffered saline (20 mM Na2HPO4 and 150 mM NaCl, pH 7.2). Cell pellets were resuspended in ETD buffer (1 mM EDTA, 10 mM Tris-HCl (pH 7.4), 5 mM dithiothreitol) containing 1 mM phenylmethylsulfonyl fluoride and homogenized by Polytron on ice in five 15-s bursts. Nuclei, with associated nuclear steroid receptor proteins, were pelleted at 4,000 × g for 30 min at 4 °C. The nuclear pellet was discarded, and the supernatant was subjected to further centrifugation at 100,000 × g for 1 h at 4 °C. The supernatant was either aliquoted and stored at -70 °C for future study, used in competitive ligand binding analyses, or subjected to further purification.

Chromatographic Enrichment of 25-[3H]OHD3 Specific Binding Activity in B95-8 Cell Extracts

The 100,000 × g supernatant of the B95-8 cell extract was used in either competitive ligand binding analyses (see below) or as starting material for chromatographic enrichment of 25-[3H]OHD3 binding activity. Eleven different types of chromatographic matrices in open column, FPLC, and HPLC formats were evaluated to identify the optimal conditions for purification. The final sequence of column chromatography was: (i) ion exchange FPLC on BPS-DE cellulose (MetaChem) through an inclining linear NaCl gradient; (ii) hydrophobic interaction FPLC on phenyl-Sepharose (Pharmacia) through a declining NaCl stepwise gradient; and (iii) HPLC over HTP-hydroxyapatite (Bio-Rad) through a declining Na2HPO4 gradient. Protein was assayed according to the method of Bradford (14). The approximate molecular mass of 25-OHD3-binding moieties resolved with hydroxyapatite column chromatography was determined with size exclusion HPLC. Two pooled peaks containing specific 25-[3H]OHD3 binding activity were microconcentrated and desalted through a Centricon concentrator with a nominal molecular weight cut-off of 30,000 (Amicon, Inc., Beverly, MA), resuspended in mobile buffer of ETD, 0.5 M NaCl (pH 7.3) and injected onto a Beckman TSK 4000PW size exclusion column (7.5 × 300 mm, Fullerton, CA) at a flow rate of 0.5 ml/min. 1.0-ml fractions were collected for reanalysis of specific hormone binding.

Ligand Binding Analyses

3H-Ligand binding was measured in crude B95-8 100,000 × g supernatants, in chromatographically enriched fractions, and in solutions of recombinant human inducible hsp-70 (0.05-50 µg/ml) by competitive protein binding assay (15). Crude material and post-anion exchange and post-hydrophobic interaction chromatography fractions were adjusted with NaCl-containing ETD buffer (pH 8.0) to achieve a final salt concentration of 0.5 M NaCl for assay; specific binding of 25-[3H]OHD3 to IDBP was found to be equivalent or superior to specific binding in the traditional assay buffer of KETD (ETD containing 0.3 M KCl (pH 7.4)) (16). Briefly, crude or post-FPLC fractions were incubated in ETD, 0.5 M NaCl overnight at 4 °C with 4 nM 25-[3H]OHD3 in the presence or absence of 1-100 nM unlabeled competitive ligand. Protein-bound 25-[3H]OHD3 was separated from unbound sterol by incubation with dextran-coated charcoal. Specifically bound 25-OHD3 was determined by subtracting the mean of duplicate determinations of binding in the presence of 100 nM radioinert competitive ligand (nonspecific binding) from the mean of triplicate determinations of binding in the absence of added competitive ligand (total binding). Specific binding of 17beta -[3H]estradiol as well as 25-[3H]OHD3 to proteins was also determined in individual fractions eluting from hydroxyapatite chromatography; the binding assay conditions were the same as those stated above except that specific binding in post-hydroxyapatite fractions was determined in ETD, 0.3 M Na2HPO4 buffer (pH 6.8).

Gel Electrophoresis and Amino Acid Sequencing

Post-hydroxyapatite fractions enriched for specific binding of 25-OHD3 were desalted, microconcentrated, and electrophoresed through an 11% discontinuous SDS-polyacrylamide electrophoresis gel as described by Laemmli (17) for 1 h at 200 volts. Visualization of resolved proteins was accomplished with silver staining (PhastGel, Pharmacia). The molecular weight of 25-OHD3-binding proteins was also estimated by specific "in-gel" binding of radiolabeled 25-[3H]OHD3. Microconcentrated, post-hydroxyapatite retentate was loaded onto adjacent lanes and subjected to SDS-polyacrylamide gel electrophoresis as described above. Following electrophoresis lanes of the gel were individually sliced, washed five times, each for 5 min, in ETD buffer at 4 °C, and then placed in a hybridization bag with 2.4 nM 25-[3H]OHD3 in ETD, 0.5 M NaCl buffer in the presence or absence of 100 nM 25-OHD3. After overnight incubation at 4 °C on a rocker platform, the gels were washed several times in 25-OHD3 takedown buffer (18) containing 123 mM sodium barbital, 123 mM sodium acetate (pH 8.6), and 1 mM sodium azide without dextran or charcoal. The gel lanes were treated with Fluoro-Hance (RPI, Mt. Prospect, IL), dried under vacuum at 60 °C for 1 h, and cut horizontally into 5-mm slices. Each gel slice was placed into scintillation vials containing Cytoscint (ICN, Irvine, CA) and radioactivity measured.

To resolve further the protein species present in microconcentrated, fractionated post-hydroxyapatite retentates, aliquots of retentate were also subjected to two-dimensional electrophoresis as described by O'Farrell (19). Aliquots were loaded onto a pre-cast, pH gradient (pH 3-10) slab gel (Novex, San Diego, CA) and electrophoresed for 1 h at 100 volts, a 2nd h at 200 volts, and finally at 500 volts for 30 min, separating proteins on the basis of their isoelectric point. The lane containing the resolved proteins was fused onto an 8.5% SDS-polyacrylamide preparative gel and electrophoresed for 80 min at 200 volts, separating proteins according to their molecular mass. Resolved proteins were either transblotted (20) onto polyvinylidene difluoride (Bio-Rad) or subjected to in-gel protease digestion (21) prior to amino acid sequence analysis of the peptides. With respect to the former, the polyvinylidene difluoride membrane was rinsed in distilled water three times each for 5 min. The membrane was stained for 5 min in 0.025% Coomassie Blue R-250 (Bio-Rad) in 40:60 methanol:water, destained for 15 min in 50% methanol, and allowed to dry at 23 °C. The bands of interest were excised from the membrane and the protein extracted for analysis of amino acid composition (22), including cysteine content (23), prior to amino acid sequencing. Because the amino terminus of the dominant protein of interest was blocked, generation of a panel of internal peptides was also achieved by incubating Coomassie-stained gel slices with 1.0 µg of trypsin, 50 µg of protein in 200 mM NH4HCO3 buffer. After overnight incubation at 37 °C, the enzymatic reaction was neutralized and peptides extracted from the gel with the addition of 60% acetonitrile in 0.1% trifluoroacetic acid. Extracted peptides or peptide fragments were purified by reverse-phase HPLC on a C18 silica column (21 × 250 mm, Vydac, Hesperia, CA) through a linear 0-80% acetonitrile gradient with an initial running condition of 100% 10 mM trifluoroacetic acid. Peptides were resolved over a 2-h period at a flow rate of 0.2 ml/min. Tryptic masses were screened for purity and integrity by matrix-assisted laser desorption/ionization mass spectrometry (24). Selected peptides were subjected to Edman sequencing (25), and the free phenylthiocarbamyl-labeled amino acids were identified with a Porton 2090 Sequencer (Beckman).

Statistical Analysis

When appropriate, experimental values for ligand binding under varying conditions were compared using Student's t test for unpaired samples.


RESULTS

Binding of Vitamin D Metabolites and Analogs

The ability of vitamin D3 and various of its metabolites to displace 25-[3H]OHD3 competitively in unfractionated 100,000 × g supernatant extracts of B95-8 cells is shown in Fig. 1. When incubated with extract at a 100 nM concentration, the naturally occurring metabolites 25-OHD3, 24,25-(OH)2D3, and 25,26-(OH)2D3 were roughly equivalent in their ability to inhibit 2 nM 25-[3H]OHD3 binding competitively. The increased polarity, because of the presence of a C-1 alpha -hydroxyl group, in 1,25-(OH)2D3 and 1,24,25-(OH)3D3, significantly decreased effective competitive binding; the C-1 position is the only naturally occurring site of metabolic hydroxylation in the A-ring of vitamin D3 (26). Two synthetic compounds that possess a C-1 alpha -hydroxyl group (1-OHD3) or a pseudo-C-1 alpha -hydroxyl group (dihydrotachysterol), but lack hydroxyl substitutions in the molecular side chain, were also incapable of displacing 25-[3H]OHD3. Vitamin D3 that is not modified in the side chain and lacks a C-1 alpha -hydroxyl was similarly ineffective as a competitive ligand. 25-[3H]OHD3 displacement by serial dilution of those vitamin D3 metabolites shown to be effective competitive inhibitors of 25-[3H]OHD3 binding is shown in Fig. 1B. Assuming an 80% displacement of bound 25-[3H]OHD3 as the maximal inhibition of binding, the ED50 for binding could be determined for only three of the five metabolites, 25-OHD3, 25,26-(OH)2D3, and 24,25-(OH)2D3; 1,25-(OH)2D3 and 1,24,25-(OH)3D3 displaced 25-OHD3 by only 20% at 10-7 M with no apparent displacement at lower concentrations. The ED50 for 25-OHD3, 25,26-(OH)2D3, and 24,25-(OH)2D3 was 5 × 10-10, 5 × 10-9, and 5 × 10-8 M, respectively.


Fig. 1. Competitive displacement of 25-[3H]OHD3 from NWP IDBP by vitamin D3 sterols. Panel A shows the ability of C-25-, C-1-, and C-25-hydroxylated vitamin D3 metabolites (100 nM) to compete with 4 nM 25-[3H]OHD3 for specific binding to IDBP in B95-8 NWP cell extract. Metabolites (1alpha -OHD3), analogs (dihydrotachysterol), and prohormones (vitamin D3) lacking the C-25 hydroxyl group did not compete for ligand binding. Panel B shows the displacement of 4 nM 25-[3H]OHD3 from IDBP through increasing concentrations of competitive 25-hydroxylated vitamin D3 metabolites. All data are expressed as the percent maximal binding of 4 nM 25-[3H]OHD3 in the absence of added competitor; each data point is the mean ± S.D. of at least three replicates.
[View Larger Version of this Image (31K GIF file)]


Vitamin D2, a Delta 5,7-diene steroid synthesized primarily in plants, differs structurally from vitamin D3 in the side chain (Fig. 2A). In contrast to vitamin D3 and its metabolites, vitamin D2 and its metabolites possess a Delta 22 (C-22-C-23 double bond) and a C-24 methyl group. 25-OHD2 and 1,25-(OH)2D2 were bound equivalently to 25-OHD3 and 1,25-(OH)2D3 in unfractionated extracts (Fig. 2B). In comparison with modifications in the proximal portion of the side chain that did not alter sterol binding, structural changes in the terminal portion of the side chain did alter IDBP binding; for example, none of the C-1-hydroxylated, so-called "nonhypercalcemic" analogs MC903, EB1089, and KH1060 (Fig. 2A) was able to displace 25-[3H]OHD3 binding in unfractionated B95-8 extracts (Fig. 2B).


Fig. 2. Failure of the C-25-modified, hypercalcemia-resistant vitamin D analogs to bind to NWP IDBP. Panel A shows the side chain structure of naturally occurring 1,25-dihydroxyvitamin D3 and 1,25-dihydroxyvitamin D2 as well as that of three synthetic analogs (see Table I). The vitamin D2 metabolites are characterized by the presence of a C-24 methyl group and a C-22-C23 double bond, whereas the non-hypercalcemia-causing vitamin D analogs all have structural alterations in the terminal aspect of the side chain. Panel B shows that modification of the vitamin D side chain in the C-25 position abolishes competitive 25-[3H]OHD3 binding, whereas more proximal alteration of the side chain, as occurs in the vitamin D2 metabolites, is without effect on ligand binding to IDBP. All data are expressed as the percent maximal binding of 4 nM 25-[3H]OHD3 in the absence of added competitor. Each data point is the mean ± S.D. of at least three replicates.
[View Larger Version of this Image (37K GIF file)]


Binding of Non-vitamin D Steroids and Bioactive Lipids

As depicted in Table I we also assessed the binding potential of a number of steroid precursor molecules, naturally occurring steroids, and two clinically useful steroid analogs, tamoxifen and RU486. With the exception of progesterone, 17beta -estradiol, and testosterone (12), none of these compounds was able to compete with 25-OHD3 for binding in 100,000 × g supernatant extracts of B95-8 cells. This includes 25-hydroxycholesterol, a molecule that bears the C-3 and C-25-hydroxyl groups but lacks the Delta 5,7-diene structure of the preferred secosteroid. Interestingly, none of the six adrenal/gonadal steroid precursor molecules, glucocorticoids, mineralocorticoids, or the major intracellular metabolite of testosterone, 5-dihydrotestosterone, was a capable ligand. The same lack of binding capacity was demonstrated with bioactive lipids in the arachidonic acid cascade as well as lipid molecules that bind to other members of the steroid receptor superfamily of proteins.

IDBP Purification

Physicochemical Measures

Limited preliminary data using 25-[3H]OHD3 as displaceable ligand indicated that there was a single class of binding moieties in the extracts of NWP cells which bound 25-OHD3 with relatively low affinity but high capacity (12). These observations did not exclude the possibility that there exists in unfractionated extracts of NWP cells more than a single protein with 25-OHD3 binding potential, which may also have the potential to bind steroids other than 25-hydroxylated vitamin D metabolites. Previous work from this laboratory (13) suggested that a 60-65-kDa protein(s) was responsible for most if not all of the specific 25-[3H]OHD3 binding in crude extracts of NWP cells. In an attempt to elucidate the nature of this protein, we subjected B95-8 cell extracts to a process of serial enrichment by physical and chemical means. Earlier studies (13) demonstrated that IDBP was partitioned in the NWP cell; under "low salt" extraction conditions, when the VDR was associated with the nuclear subfraction of the aqueous cell extract, IDBP was concentrated in the post-nuclear 4,000 × g supernatant of the cell extract. This salt-dependent partitioning led to a 3.4-fold increase in specific 25-[3H]OHD3 binding in post-nuclear extracts compared with specific 25-OHD3 uptake by whole cells (Table II, centrifugation). Subjecting the post-nuclear supernatant to ultracentrifugation resulted in an additional 53% increase in specific binding and increased overall total protein recovery by 15-20% in subsequent column chromatographic purification of B95-8 cell extracts. This was presumably the result of removing interfering lipid present in the "low speed" post-nuclear extract of these cells.

Table II.

IDBP enrichment by cell fractionation (centrifugation) and chemical means (chromatography)


Differential centrifugation
Serial column chromatography
Cell fraction Specific binding activity Fraction Specific binding activity

nmol 25-OHD3/mg protein fold enrichment
Whole cells 0.017 Cell extract 1
Post-4,000 × g sup 0.057 Post-DEAE 14
Post-100,000 × g sup 0.087 Post-phenyl-HIC 63
Post-hydroxyapatite 17,588

Serial Chromatography

The cumulative enrichment of specific 25-[3H]OHD3 binding activity through chromatography is depicted under "Chromatography" in Table II. Based on the relative stability and capacity of unfractionated 100,000 × g supernatant extract to bind 25-[3H]OHD3 at pH 8.0 (data not shown), elution of IDBP from a relatively "weak" anion exchange resin was achieved between 0.7 and 1.0 M NaCl. Fractions constituting the peak of sterol binding activity were pooled, applied to a phenyl-Sepharose FPLC column in a high ionic strength buffer, and eluted through a declining NaCl gradient in the absence of added detergent; phenyl-substituted resins provided the best recovery of specific binding activity. Although a substantial increase in specific 25-[3H]OHD3 binding activity was not obtained with hydrophobic interaction chromatography, it was critically important in separating the homodimer of glyceraldehyde-3-phosphate dehydrogenase from IDBP (27). Furthermore, chromatographic separation of glyceraldehyde-3-phosphate dehydrogenase and other proteins from IDBP on the basis of their hydrophobicity dramatically amplified the enrichment capability of hydroxyapatite FPLC for the protein(s) of interest.

There were two major peaks of specific 25-[3H]OHD3 binding activity in the eluent from hydroxyapatite chromatography (Fig. 3A), one eluting at ~50 mM (peak I) and a smaller one eluting at ~100 mM (peak II) Na2HPO4. Specific sterol binding was noted at no other position of the gradient. To assess the apparent molecular mass of these binding moieties an aliquot from the pooled fractions constituting peak I and peak II was subjected to gel filtration HPLC under nondenaturing conditions (Fig. 3B). Both peaks from hydroxyapatite chromatography exhibited specific 25-[3H]OHD3 binding activity in the range of 60-65 kDa. This was confirmed by in-gel labeling with 25-[3H]OHD3.


Fig. 3. Binding of 25-[3H]25-OHD3 by IDBP purified from B95-8 NWP cell extracts. Panel A demonstrates the gradient elution of proteins with specific 25-[3H]OHD3 binding activity from hydroxyapatite HPLC. Two peaks of specific binding activity eluted at 50 and 75 mM salt, referred to as peak I and peak II, respectively. Specific 25-[3H]OHD3 binding activity was similar in both peaks (inset). Peak I and II from hydroxyapatite chromatography were committed to gel filtration HPLC under nondenaturing conditions (panel B). In both, a single peak of specific 25-[3H]OHD3 binding activity eluted from the column with an apparent molecular mass in the range of 65 kDa. Elution of standards blue dextran (BD), bovine serum albumin (BSA, 66 kDa), and carbonic anhydrase (CA, 29 kDa) under identical chromatographic conditions is depicted.
[View Larger Version of this Image (29K GIF file)]


Considering that gonadal steroids were also bound in post-nuclear (12) as well as in 100,000 × g supernatants of the post-nuclear extract of B95-8 cells (Table I), we examined the possibility that one or both of the 25-OHD3 binding peaks from hydroxyapatite chromatography were also responsible for gonadal steroid binding observed in the unfractionated extracts of B95-8 cells. Aliquots, matched for protein concentration, from peak I and peak II were subjected to ligand binding analysis with tracer quantities of 25-[3H]OHD3 or 17beta -[3H]estradiol in the presence and absence of 100 nM radioinert 25-OHD3 and 17beta -estradiol. Specific binding activity for 25-OHD3 exceeded that of 17beta -estradiol by 21-fold (8.3 versus 0.2 nmol/mg of protein) and 17-fold (6.7 versus 0.2 nmol/mg of protein) in peak I and peak II, respectively. These data indicated that although both peak I and peak II harbored 25-OHD3 and 17beta -estradiol binding capacity, they were relatively enriched for specific 25-[3H]OHD3 binding activity compared with unpurified NWP cell extracts (see Table I).

Gel Electrophoresis and Amino Acid Sequence Analysis

Two-dimensional polyacrylamide gel electrophoresis of the 25-[3H]OHD3 binding moieties in peak I from hydroxyapatite chromatography (see Fig. 3A) identified at least three distinct proteins in the molecular mass range of interest (60-70 kDa). The dominant, 65-kDa, Coomassie-stained spot displayed a pI ~ 4.5. Amino acid composition of the eluted protein(s) revealed the presence of very few cysteine residues (4% the total residues), once again distinguishing this protein(s) from the cysteine-rich proteins in steroid/sterol/thyronine receptor superfamily (10) and in the albumin family of proteins which includes the circulating vitamin D-binding protein (8, 9). Because the amino terminus of the most abundant protein in the isolate was blocked, we committed the remaining sample to in-gel proteolytic cleavage (21) prior to amino acid sequence analysis of the proteolytic products. Reverse-phase HPLC of the resultant peptides resulted in the separation and reproducible resolution of seven mass peaks belonging to the parent protein. The amino acid sequence of each was determined after Edman degradation of the peptide fragments. Sequence analyses identified major tryptic peptides ranging in length from 10 to 15 residues and possessing 89, 83, 75, 70, 60, 55, and 27% amino acid sequence identity to human inducibly expressed hsp-70 (28) (Fig. 4A). Two of the three tryptic peptides with most sequence homology to hsp-70 were located within the amino-terminal domain of hsp-70 (residues 39-128), that region of the hsp-70 molecule known to be most conserved among different mammalian species, whereas the least homologous fragment appeared to reside in the more variable, COOH-terminal region of hsp-70 (residues 538-552).


Fig. 4. Amino acid sequence and functional similarity of NWP IDBP, and human inducibly expressed hsp-70. Panel A demonstrates sequence similarity among seven different tryptic peptides of IDBP and recombinant inducibly expressed hsp-70 of human origin. Sequence identity of IDBP peptides with hsp-70 diminished considerably in the COOH terminus of the hsp-70 molecule. Panel B compares the displacement of 4 nM 25-[3H]OHD3 from post-hydroxyapatite-purified IDBP and recombinant human, inducibly expressed hsp-70 (both at 0.01 µg/tube) by increasing concentrations of radioinert 25-hydroxyvitamin D3. Each point is the mean of triplicate assessments of total binding; total 25-[3H]OHD3 binding by hsp-70 in the absence of added sterol is the mean ± S.D. of six replicate determinations. The concentration of 25-OHD3 required to achieve maximal displacement of 25-[3H]OHD3 was the same (100 nM) for IDBP and hsp-70.
[View Larger Version of this Image (38K GIF file)]


If, as the sequence data suggest, IDBP is structurally related to the hsp-70 family of proteins and the ligand (sterol) binding domain of IDBP is preserved in hsp-70s, then recombinant human, inducibly expressed hsp-70 should be able to bind 25-[3H]OHD3 specifically. This was the case. Inducibly expressed hsp-70 exhibited specific 25-[3H]OHD3 binding activity (Fig. 4B). The concentration of radioinert 25-OHD3 required to achieve half-maximal displacement of 25-[3H]OHD3 from both post-hydroxyapatite-purified IDBP and from inducibly expressed hsp-70 was approximately the same, 5-10 nM.


DISCUSSION

The existence of steroid hormone-resistant states in monkeys from the New World was first described by Chrousos and his colleagues (29). They reported high circulating levels of adrenal (29-31) and gonadal (32) steroids in NWP from several different genera. Vitamin D resistance in NWP was first noted by Shinki et al. (2) and confirmed by other investigators (33-35). NWP have adapted well to their resistance to gonadal and adrenal steroids (36). They do not suffer from adrenal insufficiency or infertility, presumably because of their ability to augment adrenal and gonadal steroid production from endogenous cholesterol substrates to meet physiological demands. Neither do they experience clinical vitamin D deficiency when residing in their normal habitat, the canopy of the equatorial rain forests of South and Central America.2 It is only when these primates are imported and maintained in artificial environments at elevated latitudes that they develop the clinical syndrome of vitamin D deficiency; in fact, exposure of primates with clinical rickets or osteomalacia to an artificial source of ultraviolet radiation will dramatically increase circulating 25-OHD and 1,25-(OH)2D concentrations and resolve their metabolic bone disease (4).

A number of theories have been put forward to explain the steroid- and vitamin D-resistant state in NWP. Some investigators have suggested that alteration in the structural character, quantity, and/or avidity of circulating steroid-binding proteins for ligand may contribute to the high circulating concentrations of hormone (37). However, we (33) and others (38) have not been able to discern a functional difference among the circulating steroid-binding proteins in OWP and NWP. Brandon et al. (39) suggested that steroid hormone resistance in NWP, as in humans, may result from changes (conserved mutations) in the receptor proteins for these ligands; these investigators have shown decreased receptor-ligand and receptor-DNA binding activity in NWP cells compared with OWP cells (32). However, specific function-altering mutations (i.e. coding for a premature stop codon) in the NWP receptor proteins, in either the ligand or DNA binding domains, have not been described (39). In fact, our experimental results with the VDR in NWP cells suggest that it is normal in all respects (1, 13). Furthermore, in evolutionary terms, the odds of independent, coincident, and conserved resistance-causing mutational events taking place in several of the steroid/sterol receptor genes in just the last 50 million years, since the occurrence of Old and New World continental drift (40), are very slim. More likely is the possibility that an alteration in expression of normally expressed gene products has been adopted to selective advantage in the suborder Platyrrhini (NWP).

Moore et al. (41) have suggested that glucocorticoid and mineralocorticoid resistance in NWP results in part from an alteration in the regulation of one such molecule, an enzyme, 11beta -hydroxysteroid dehydrogenase, which catalyzes cortisol conversion to the biologically inactive metabolite cortisone. Decreased activity of this enzyme will promote accumulation of cortisol in the serum as is observed in most NWP species in which it has been examined (42). However, structural and/or functional alterations in a single enzyme cannot explain the fact that a variety of different steroid/sterol molecules are affected in NWP, most synthesized and metabolized by different enzymes in different tissues. Furthermore, these investigators recently cloned and expressed the NWP 11beta -hydroxysteroid dehydrogenase in COS cells (41) and found its activity to be not different enough from OWP enzyme to explain the 50-fold increase in serum cortisol levels observed in NWP. These data suggest that although alteration in the NWP 11beta -hydroxysteroid dehydrogenase may provide a partial explanation for the adrenal steroid-resistant state, it cannot account for either the magnitude of resistance or apparent diversity of steroid/sterol ligands to which these animals are resistant.

We have shown that there is at least one other protein that is overexpressed in cultured dermal fibroblasts and B-lymphocytes from a variety of genera of NWP and associated with the state of cellular vitamin D (1) and gonadal steroid resistance (12). Crude extracts of NWP cells containing this protein, which we have termed the intracellular vitamin D-binding protein, competitively bind 25-hydroxylated vitamin D sterols and gonadal steroid hormones. Because of its relatively high capacity for ligand and by virtue of its presence in the cytoplasmic compartment of the cell (13), we have hypothesized that IDBP can intercept hormones en route to their respective receptor proteins. The purpose of the current investigation was to test this hypothesis by gaining further insight into IDBP from a functional and structural standpoint.

We theorized that the C-1-hydroxylated metabolites of vitamin D3, which bind most avidly to the VDR, would be most avidly bound by IDBP. This was not the case (Fig. 1). The C-25-hydroxylated metabolites of vitamin D3, 25-OHD3, 24,25-(OH)2D3, and 25,26-(OH)2D3, were the most tightly bound of the naturally occurring vitamin D3 metabolites tested. In fact, hydroxyl substitution in the C-1 position of the A-ring of vitamin D3 reduced, not increased, the affinity of the ligand for IDBP. We considered the possibility that IDBP, or a related protein, may be the long sought intracellular receptor protein for some of those vitamin analogs that have potent immunoinhibitory action in vitro and in vivo without causing hypercalcemia in the host (26). Three such synthetic compounds, which harbor structural alterations in the terminal aspects of the vitamin D side chain (Fig. 2) did not bind to IDBP. On the contrary, more proximal alteration of the vitamin D side chain, as occurs naturally in the vitamin D2 molecules synthesized by plants, did not alter binding as long the C-25 hydroxyl moiety was present. This observation argues against the theory set forth by Marx and colleagues (43) to explain vitamin D resistance in NWP. These investigators postulated that there was preferential utilization of vitamin D3 (cholecalciferol) metabolites over vitamin D2 (ergocalciferol) metabolites in NWP cells.

The fact that 1,25-(OH)2D3 was apparently not the vitamin D ligand of choice for IDBP coupled with the finding that the structurally dissimilar gonadal steroid hormones 17beta -estradiol, testosterone, and progesterone were also capable of specifically displacing 25-OHD3 from binding protein(s) in NWP cell extracts (12), led us to speculate that (i) we had not yet identified the preferred ligand for IDBP and (ii) there was likely to be more than a single ligand-binding species in our crude cell extracts. In an attempt to address the former we took a "candidate ligand" approach (Table I). Aside from confirmation of the fact that IDBP bound the three previously mentioned sex steroids, albeit with a relative binding index less than or equal to 25-hydroxylated vitamin D metabolites, there was no evidence for IDBP being able to bind any of the other 25 candidate ligands chosen for investigation. This list of candidates included adrenal and gonadal steroid precursor molecules, metabolites and analogs, as well as other known bioactive lipid molecules. The relatively uninformative nature of this search for a potential candidate ligand led us to concentrate efforts on purification and structural identification of IDBP(s). Serial chromatographic purification of extracts of NWP cells showed IDBP to be a 65-kDa protein (Fig. 3). In contrast to previously evaluated crude whole cell extract (12), the material derived from serial chromatography in the current studies exhibited much higher specific 25-OHD3 binding activity (Table II), a substantial decrement in specific 17beta -estradiol binding, and complete loss of specific testosterone and progesterone binding (data not shown). Since we purified our extracts selectively on the basis of 25-OHD3 binding, these results indicate that there are distinct binding proteins or IDBP isoforms that bind gonadal steroids in NWP cells.

We were surprised by the amino acid sequence data, which clearly demonstrated that the IDBP isolated here was a member of the hsp-70 family of proteins (Fig. 4); unlike some of the other molecular chaperones, including the immunophilins (44) and with the exception of sulfoglycolipids (45), hsp-70 and related proteins are not known to bind small lipophylic ligands specifically. Although they can be associated with multimeric cytoplasmic protein complexes that include steroid hormone receptors (46), members of the hsp-70 family of proteins have not been recognized previously for their potential to bind specific sterol/steroid hormones. Moreover, the capacity of IDBP to bind 25-hydroxylated vitamin D sterols is apparently not confined to hsp-70-related proteins in NWP cells. A functional homolog of IDBP can be identified in OWP cells including those of human origin (11). Whether or not this homolog is the same as or closely related to inducibly expressed hsp-70 (Fig. 4) remains to be seen. Preliminary data indicate that constitutively expressed hsp-70 (hsc-70), like inducibly expressed hsp-70 and purified IDBP, is able to bind 25-hydroxylated vitamin D metabolites specifically.3

Why might IDBP be overexpressed in a constitutive fashion in NWP cells? We favor the hypothesis that its overexpression is a remnant of the NWP ancient response to exogenous vitamin D intoxication by ingestion of 25-hydroxylated vitamin D metabolite-containing plants in the genus Solanum (47, 48). In this theory IDBP buffers the target cell from a plant-derived vitamin D sterol that may itself bind or be metabolized to a compound that binds to the VDR, initiating the calcemic response in the host. Even today, Solanum species can be found in regions of South America which coincide with the point of radiation of ancestral primates across the continent (48) and have been reported (49) to cause life-threatening hypercalcemia in dairy cattle grazing on these plants. We have observed recently that extracts of these plants contain a vitamin D metabolite, distinct from 25-OHD and 1,25-(OH)2D, which binds avidly to purified IDBP and hsp-70.4 Given the decided preference of IDBP for 25-hydroxylated vitamin D metabolites (Fig. 1), it is also possible that IDBP functions as an intracellular concentrator of substrate for the mitochondrial vitamin D-1-hydroxylase in animals with a dramatically increased need for the vitamin D hormone, 1,25-(OH)2D. Proteins in the hsp-70 family have been recognized as key components of the mitochondrial import machinery; unfortunately, none of the seven tryptic peptides we have so far isolated is from the extreme amino terminus of hsp-70 where one would anticipate the presence of a mitochondrial targeting sequence (50). This alternative hypothesis would suggest, of course, that IDBP overexpression NWP cells is an adaptive response to resistance caused by another protein. We have recently identified such a candidate, a nuclear binding protein that interacts in a dominant-negative fashion with the vitamin D response element (7). Whether IDBP prevents or promotes the 1,25-(OH)2D-VDR interaction awaits characterization of the full-length IDBP and its transient overexpression in OWP cells.


FOOTNOTES

*   This work was supported by National Institutes of Health Grants AR-37399 (to J. S. A.) and DK-07682 (to H. C. and J. E. A.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Dedicated to the memory of Dr. Bayard Catherwood whose advice in protein purification was essential.


§   To whom correspondence should be addressed: B131, Cedars-Sinai Medical Center, Los Angeles, CA 90048. Fax: 310-967-0119.
1   The abbreviations used are: NWP, New World primate(s); 25-OHD, 23-hydroxyvitamin D; 1,25-(OH)2D, 1,25-dihydroxyvitamin D; OWP, Old World primate(s); VDR, vitamin D receptor; IDBP, intracellular vitamin D-binding protein; 25-[3H]OHD3, 25-[3H]hydroxyvitamin D3; FPLC, fast protein liquid chromatography; HPLC, high performance liquid chromatography.
2   A. Baker, personal communication.
3   M. A. Gacad and J. S. Adams, unpublished results.
4   M. A. Gacad, G. Lee, and J. S. Adams, unpublished results.

ACKNOWLEDGEMENTS

We thank Audree Fowler and Kay McCollough of the UCLA microsequencing facility for technical support.


REFERENCES

  1. Gacad, M. A., and Adams, J. S. (1991) J. Clin. Invest. 87, 996-1001 [Medline] [Order article via Infotrieve]
  2. Shinki, T., Shiina, Y., Takahashi, N., Tanioka, Y., Koizumi, H., and Suda, T. (1983) Biochem. Biophys. Res. Commun. 114, 452-457 [Medline] [Order article via Infotrieve]
  3. Adams, J. S., Gacad, M. A., Baker, A. J., Gonzales, B., and Rude, R. K. (1985) Am. J. Primatol. 9, 219-224
  4. Gacad, M. A., and Adams, J. S. (1992) Am. J. Primatol. 28, 263-270
  5. Okuda, K.-I., Usui, E., and Ohyama, Y. (1995) J. Lipid Res. 36, 1641-1652 [Medline] [Order article via Infotrieve]
  6. Malloy, P. J., Hochberg, Z., Pike, J. W., and Feldman, D. (1989) J. Clin. Endocrinol. & Metab. 68, 263-269 [Abstract]
  7. Arbelle, J. E., Chen, H., Gacad, M. A., Allegretto, E. A., Pike, J. W., and Adams, J. S. (1996) Endocrinology 137, 786-789 [Abstract]
  8. Bouillon, R., Van Baelen, H., and DeMoor, P. (1976) Biochem. J. 159, 463-466 [Medline] [Order article via Infotrieve]
  9. Cooke, N. E., and Haddad, J. G. (1989) Endocr. Rev. 10, 294-307 [Medline] [Order article via Infotrieve]
  10. Tsai, M. J., and O'Malley, B. W. O. (1994) Annu. Rev. Biochem. 63, 451-486 [CrossRef][Medline] [Order article via Infotrieve]
  11. Gacad, M. A., and Adams, J. S. (1990) J. Bone Miner. Res. 5, 198S (abstr.)
  12. Gacad, M. A., and Adams, J. S. (1992) Endocrinology 131, 2581-2587 [Abstract]
  13. Gacad, M. A., and Adams, J. S. (1993) J. Bone Miner. Res. 8, 27-34 [Medline] [Order article via Infotrieve]
  14. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254 [CrossRef][Medline] [Order article via Infotrieve]
  15. Adams, J. S., and Gacad, M. A. (1988) J. Clin. Endocrinol. & Metab. 66, 224-229 [Abstract]
  16. Gacad, M. A., and Adams, J. S. (1990) Clin. Res. 38, 120 (abstr.)
  17. Laemmli, U. K. (1970) Nature 227, 680-685 [Medline] [Order article via Infotrieve]
  18. Adams, J. S. (1984) J. Steroid Biochem. 20, 857-862 [Medline] [Order article via Infotrieve]
  19. O'Farrell, P. H. (1975) J. Biol. Chem. 250, 4007-4021 [Abstract]
  20. Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acad. Sci. U. S. A. 76, 4350-4354 [Abstract]
  21. Rosenfield, J., Capdevielle, J., Guillemot, J. C., and Ferrara, P. (1992) Anal. Biochem. 203, 173-179 [Medline] [Order article via Infotrieve]
  22. Moos, M., Jr., Nguyen, N. Y., and Liu, T.-Y. (1988) J. Biol. Chem. 263, 6005-6008 [Abstract/Free Full Text]
  23. Rao, U. S., and Scarborough, G. A. (1990) J. Biol. Chem. 265, 7227-7235 [Abstract/Free Full Text]
  24. Beavis, R. C., Chaudhary, T., and Chait, B. T. (1992) Org. Mass Spectrom. 27, 156-158
  25. Hunkapiller, M. W., Granlund-Moyer, K., and Whiteley, N. W. (1986) in Methods of Protein Microcharacterization (Shively, J. E., ed), pp. 223-247, Humana, Clifton, NJ
  26. Bouillon, R., Okamura, W. H., and Norman, A. W. (1995) Endocr. Rev. 16, 200-257 [Medline] [Order article via Infotrieve]
  27. Kocha, T., Fukuda, T., Isobe, T., and Okuyama, T. (1989) Biochim. Biophys. Acta 991, 56-61 [Medline] [Order article via Infotrieve]
  28. Leung, T. K., Rajendran, M. Y., Monfries, C., Hall, C., and Lim, L. (1990) Biochem. J. 267, 125-132 [Medline] [Order article via Infotrieve]
  29. Chrousos, G. P., Loriaux, D. L., Brandon, D., Schull, J., Renquist, D., Hogan, W., Tomita, M., and Lipsett, M. B. (1984) Endocrinology 115, 25-32 [Abstract]
  30. Brandon, D. D., Markwick, A. J., Chrousos, G. P., and Loriaux, D. L. (1989) Cancer Res. 49, 2203-2213
  31. Tomita, M., Brandon, D. D., Chrousos, G. P., Vingerhoeds, A. C. M., Foster, C. M., Fowler, D., Loriaux, D. L., and Lipsett, M. B. (1986) J. Clin. Endocrinol. & Metab. 62, 1145-1154 [Abstract]
  32. Chrousos, G. P., Brandon, D., Renquist, D. M., Tomita, M., Johnson, E., Loriaux, D. L., and Lipsett, M. B. (1984) J. Clin. Endocrinol. & Metab. 58, 516-520 [Abstract]
  33. Adams, J. S., Gacad, M. A., Baker, A. J., Keuhn, G., and Rude, R. K. (1985) Endocrinology 116, 2523-2527 [Abstract]
  34. Liberman, U. A., deGrange, D., and Marx, S. J. (1985) FEBS Lett. 182, 385-389 [CrossRef][Medline] [Order article via Infotrieve]
  35. Takahashi, N., Suda, S., Shinki, T., Horiuchi, N., Shuna, Y., Tanioka, Y., Koizumi, H., and Suda, T. (1985) Biochem. J. 227, 555-563 [Medline] [Order article via Infotrieve]
  36. Coe, C. L., Smith, E. R., and Levine, S. (1985) in Handbook of Squirrel Monkey Research (Rosenblum, L. A., and Coe, C. L., eds), pp. 191-220, Plenum Press, New York
  37. Klosterman, L. L., Murai, J. T., and Siiteri, P. K. (1986) Endocrinology 118, 424-434 [Abstract]
  38. Hammond, G. L., Smith, C. L., Lahteenmaki, P., Grolla, A., WarmelsRodenhiser, S., Hodgert, H., Murai, J. T., and Siiteri, P. K. (1994) Endocrinology 134, 891-898 [Abstract]
  39. Brandon, D. D., Markwick, A. J., Flores, M., Dixon, K., Albertson, B. D., and Loriaux, D. L. (1991) Mol. Endocrinol. 7, 89-96
  40. Philbeam, D. (1984) Sci. Am. 250, 84-96 [Medline] [Order article via Infotrieve]
  41. Moore, C. C. D., Mellon, S. H., Murai, J., Siiteri, P. K., and Miller, W. L. (1993) Endocrinology 133, 368-375 [Abstract]
  42. Hammami, M. M., and Siiteri, P. K. (1991) J. Clin. Endocrinol. & Metab. 73, 326-334 [Abstract]
  43. Marx, S. J., Jones, G., Weinstein, R. S., Chrousos, G. P., and Renquist, D. M. (1989) J. Clin. Endocrinol. & Metab. 69, 1282-1290 [Abstract]
  44. Clardy, J. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, 56-61 [Abstract]
  45. Boulanger, J., Faulds, D., Eddy, E. M., and Lingwood, C. A. (1995) J. Cell. Physiol. 165, 7-17 [Medline] [Order article via Infotrieve]
  46. Peattie, D. A., Harding, M. W., Fleming, M. A., DeCenzo, M. T., Lippke, J. A., Livingston, D. J., and Benasutti, M. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 10974-10978 [Abstract]
  47. Wasserman, R. H., Henion, J. D., Haussler, M. R., and McCain, T. A. (1976) Science 194, 853-855 [Medline] [Order article via Infotrieve]
  48. Boland, R. L. (1986) Nutr. Rev. 44, 1-8
  49. Skliar, M. I., Boland, R. L., Mourino, A., and Tojo, G. (1992) J. Steroid Biochem. Mol. Biol. 43, 677-682 [Medline] [Order article via Infotrieve]
  50. Bhattacharyya, T., Karnezis, A. N., Murphy, S. P., Hoang, T., Freeman, B. C., Phillips, B., and Morimoto, R. I. (1995) J. Biol. Chem. 270, 1705-1710 [Abstract/Free Full Text]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.