Unoccupied 1,25-Dihydroxyvitamin D, Receptors NUCLEAR/CYTOSOL RATIO DEPENDS ON IONIC STRENGTH*

Previous failures to detect cytosol receptors for 1,25-dihydroxyvitamin D3 (1,25(OH)zD3) in low ionic strength buffers were reassessed, since these buffers are used routinely with other steroid hormones. In the present studies, crude nuclei or chromatin fractions contained 90% of the tissue unoccupied 1,25(OH)zD3 receptors when the intestinal mucosa of vitamin D- deficient chicks was homogenized in low salt buffer (TED; 10 mM Tris, 1.5 m~ EDTA, 1.0 m~ dithiothreitol). Significant numbers of these receptors (25 to 50%) were also present in these subcellular fractions when the tissue was homogenized in higher ionic strength (100 to 300 nm) buffers. This property was not simply the result of generalized sticking of acidic proteins to nuclear components, since the acidic 25(0H)D3-binding protein was not present in crude nuclear preparations. The unoccupied nature of the 1,25(OH)zD3 receptor sites was verified by their non-detection in nuclei of chicks treated with 200 units of 1,25(OH)2D3 2 h prior to killing. The co-identity of these 1,25(OH)zD3 binding sites with “cytosol” receptors was confirmed by their extracta- bility with high ionic strength buffers, by Scatchard analysis (Kd = 0.65 to 0.97 m), and by gel filtration (Sephacryl S-200) chromatography. The proportion of unoccupied 1,25(OH)zD3 receptors associated with nu- clear components varied inversely with the ionic strength of the buffers. Conversely, omission of sucrose from a buffer routinely used in such studies to stabilize nuclei had no effect on the cytosol/nuclear ratio. Unoccupied

Previous failures to detect cytosol receptors for 1,25-dihydroxyvitamin D3 (1,25(OH)zD3) in low ionic strength buffers were reassessed, since these buffers are used routinely with other steroid hormones. In the present studies, crude nuclei or chromatin fractions contained 90% of the tissue unoccupied 1,25(OH)zD3 receptors when the intestinal mucosa of vitamin Ddeficient chicks was homogenized in low salt buffer (TED; 10 m M Tris, 1.5 m~ EDTA, 1.0 m~ dithiothreitol). Significant numbers of these receptors (25 to 50%) were also present in these subcellular fractions when the tissue was homogenized in higher ionic strength (100 to 300 nm) buffers. This property was not simply the result of generalized sticking of acidic proteins to nuclear components, since the acidic 25(0H)D3-binding protein was not present in crude nuclear preparations. The unoccupied nature of the 1,25(OH)zD3 receptor sites was verified by their non-detection in nuclei of chicks treated with 200 units of 1,25(OH)2D3 2 h prior to killing. The co-identity of these 1,25(OH)zD3 binding sites with "cytosol" receptors was confirmed by their extractability with high ionic strength buffers, by Scatchard analysis (Kd = 0.65 to 0.97 m), and by gel filtration (Sephacryl S-200) chromatography. The proportion of unoccupied 1,25(OH)zD3 receptors associated with nuclear components varied inversely with the ionic strength of the buffers. Conversely, omission of sucrose from a buffer routinely used in such studies to stabilize nuclei had no effect on the cytosol/nuclear ratio. Unoccupied 1,25(OH)zD3 receptors were predominantly (61 to 924) associated with nuclear components after TED homogenization in all tissues studied: chick intestinal mucosa, parathyroid, kidney, and pancreas; rat intestinal mucosa, kidney, and testes; and an osteoblast-like mouse bone cell line. Although the subcellular localization of unoccupied 1,25(OH)zD3 receptors in vivo remains unresolved, the nuclear association in low salt buffer in vitro has many important biochemical and physiological ramifications.
1,25-Dihydroxyvitamin DB, an active metabolite of the secosteroid vitamin DB, is now generally considered to be a steroid hormone by all the classical criteria (1)(2)(3)(4)(5). Cytoplasmic receptors for this hormone have been described in multiple target * This work was supported by United States Public Health Service Grant AM-09012 and Training Grant AM-07310. This is Paper XXVII in a series, "Studies on the Mode of Action of Calciferol." 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.

MATERIALS AND METHODS
Animals a n d Tissue Preparation: Chick Intestinal Mucosa-White Leghorn cockerels obtained on the hatch date from Pace/ Setter, Alto Loma, CA, were raised for 3 to 4 weeks on a standard rachitogenic diet (26). After decapitation, the duodenal loop was rapidly removed, stripped of contents, and washed at 4°C in 0.9% NaC1. All subsequent steps were performed a t 4OC. The mucosa was scraped from the serosa with a glass slide and the scraping was homogenized in the desired buffer (5 to 20% w/v) with 10 strokes in a glass-Teflon homogenizer. After a low speed spin (5000 X g, 10 min) of the homogenate, cytosol was prepared by centrifuging at 105,000 X g for 1 h. Nuclei or chromatin was prepared by washing the initial pellet three times in T E D or in TED with 0.54 Triton X-100, respectively, and both preparations were resuspended in TED for incubation. Chromatin preparations necessitated higher force spins: 10,ooO to 20,000 X g, 10 to 15 min. When desired, chromatin extracts were prepared by exposing the crude chromatin to STKM or KTED buffer (same volume as homogenization buffer) at 4°C for 45.min with frequent blending on a Vortex mixer. Then the residual pellet was recovered after centrifugation (5000 X g , 10 min) and the supernatant (extract) was cleaned of debris (105,000 X g , 1

6799
This is an Open Access article under the CC BY license.

Unoccupied 1,25(0H)& Receptors Associated
with Nuclei aliquots in polystyrene test tubes (12 x 75 mm) with 5 n~ 1,25(OH)L [26,27-'"H]D:j (9 Ci/mmol) in the presence or absence of 200-fold exces unlabeled 1,25(OH)zD:I a t 4°C for 90 min or 18 to 24 h. For assay of cytosol and soluble chromatin extracts, the incubation was terminated by the addition of 500 p1 of 50% v/v hydroxylapatite (Bio-Rad:Bio-Gel HTP) in T E D and further incubation a t 4OC for 15 min with frequent blending on a Vortex mixer. Although not strictly necessary, the addition of hydroxylapatite to chromatin pellets decreases the possibility of accidental pellet losses during the postincubation washes (25). In all cases, bound and free hormone were separated by washing the pellets three times with T E D plus 0.5% Triton at 5000 X g. Then radioactivity was extracted with 1.0 ml of 100% ethanol at 30°C for 30 min with blending on a Vortex mixer for scintillation counting in 5 ml of Amersham Searle's ACS. Specific Receptor Assay in Chick Kidney and Parathyroid Gland-The procedures outlined above for animal preparation and assay of unoccupied 1,25(OH)zD3 receptors were essentially unchanged for experiments with subcellular fractions of the chick kidney. However, in order to allow hypertrophy of the parathyroid glands, some chicks were maintained on the rachitogenic diet for 5 to 6 weeks prior to killing. No other modifications were necessary for assay of receptors in the parathyroid.
Receptor Assay in Tissues of Vitamin D-deficient Rats-Weanling male rats (Sprague-Dawley) were obtained from Hilltop Lab Animals, Inc., Chatsworth, CA, and were placed on a synthetic nonrachitogenic (0.47% calcium, 1.2% phosphate) vitamin D-deficient diet (5, 27) for 7 to 9 weeks. Following decapitation, small intestinal mucosa, kidney, and testes were removed and washed in 0.9% NaCI. AU tissues were resuspended in TED + 10 mM Na2MoOl (33% w/v) prior to homogenization. Cytosol and crude chromatin were prepared as described above; chromatin was resuspended in TEDMo for incubation. Ali  Table I summarizes the amounts of 1,25(OH)zD3 receptors detected in various cytosol preparations from the intestinal mucosa of vitamin D-deficient chicks, Very little, 1,25(OH)*D:r receptor is found in cytosols prepared in TED or T E D plus 20% glycerol; and STKM cytosol contains less receptor than KTED cytosol. Due to the relatively gentle incubation conditions (4"C, 90 min) and the presence of a saturating level of ligand, receptor instability or degradation might not account for the differences between TED and the more frequently utilized buffers KTED and STKM. Additionally, the effect of STKM seems unrelated to possible stabilization of receptor proteins by the sucrose, since the addition of glycerol to TED gave no substantial improvement in cytosol receptor detection.
To examine the tissue distribution of unoccupied 1,25(OH)2D:j receptors more fully, intestinal mucosa of vitamin D-deficient chicks was lightly mixed in TED and then half the tissue pool was homogenized in either TED or KTED. Cytosol and crude nuclei were prepared and 1,25(oH)~D:1 receptors were assessed by incubating with ["'H]1,25(OH2D:j as described under "Materials and Methods." Surprisingly, a  The above observations could have resulted simply from a generalized sticking of acidic cytoplasmic components to nuclear contents (28,29). To test this possibility, the relative contamination of nuclear components by the vitamin D-binding globulin was assessed. This acidic protein (30), which binds 25(OH)D3 with high affinity, is prominent in plasma and is also present in many cytosol preparations (31)(32)(33)(34).
When r3H]25(OH)Ds binding was measured in mucosal cytosol and chromatin prepared in TED, only 12% of the specific 25(OH)D3 binding was observed in the chromatin fraction (Fig. 1B). These data suggested that the nuclear association of the unoccupied 1,25(OH)2D3 receptors results from a specific biochemical property of the receptor molecules.
Unoccupied Nature of the Nuclear 1,25(oH)~D3 Binding Sites in Vitamin D-deficient Chicks-Although the chicks used in these experiments were fed only a rachitogenic diet (26) for 3 to 4 weeks prior to killing, the presence of endogenous metabolites bound to the receptors could have explained the relatively high affinity for chromatin. In order to test this possibility, rachitic chicks were untreated or were injected subcutaneously with 500 units of 1,25(OH)2Ds 2 h prior to killing. TED-chromatin was prepared and was incubated with 5 nM [3H]1,25(OH)2D3 f 1 PM unlabeled 1,25(OH)2Do at 4°C for 2 or 24 h. As shown in Fig. 2 The identity of the 1,25(OH)~D3 binding sites associated with TED-chromatin was explored by gel filtration chromatography. STKM extracts of TED-chromatin or KTED extracts of TED-chromatin were incubated with 5 nM ["HI-1,25(OH)~D3 at 4°C for 2 h. T h e n 5 0 0~1 aliquots of the extracts or other tissue preparations were applied to an S200 (Pharmacia) column (0.5 X 85 cm) in the cold room. Samples were eluted with the equilibration buffer (KTED) at a flow rate of 8 ml/h. Importantly, the ["H]1,25(OH)2Ds elution patterns in the chromatin extracts were similar to those of the control cytosol preparations (Fig. 4).  (39,40) and triiodothyronine receptors (41,42). However, the situation is not as clear for the buffer STKM, where the ionic concentration is only 80 mM in spite of the high osmolarity. We therefore questioned whether specific chemical components of this buffer solution might cause the observed receptor distribution through some property other than ionic strength. The most likely single components for such effects were the divalent cation Mg", which exerts multiple effects on DNA and nuclei (37), and sucrose which may preserve nuclear integrity (43), thus also preserving the true in vivo receptor distribution. In separate experiments mucosal scrapings of vitamin D-deficient chicks were mixed in TED or in TKM prior to complete homogenization of aliquots in TED, TED * 5 mM MgC12, and KTED or in STKM and TKM, respectively. The addition of 5 mM MgC12 to TED did not result in substantial receptor solubilization into the cytosol. Additionally, the exclusion of sucrose from STKM had no effect on the cytoso1:chromatin distribution of the unoccupied 1,25(OH)2Ds receptors (not shown). were identical with those of Fig. 4, except that the tissue fractions to application to the column.

Unoccupied 1,25(OH)zD~ Receptors Associated with Nuclei
To pursue these questions further, equal aliquots of mocosal scrapings were homogenized in STKM, TKM, 10 mM TKM, TED, and KTED after initial mixing in TED. As shown in Fig. 7 , the number of chromatin-associated receptors decreased as an inverse of the ionic concentration of the homogenization buffer. This relationship is presented more accu- The close relationship between the numbers of extracted receptors and the ionic strength of the buffer coupled with the specific lack of effect by sucrose suggests that low ionic strength buffers may more accurately represent the location of unoccupied 1,25(0H)*D3 receptors in uiuo.

Lack of 1,25(0H)2D3 Receptor Binding Activity in Highly Purified Brush Border Membranes-Other
investigators have demonstrated that crude chromatin prepared by methods similar to those used herein may be contaminated with brush border membrane components (44,45). In order to differentiate between hypothetical 1,25(OH)2D3 binding components from this subcellular organelle and classical hormone receptors more specifically associated with nuclei, ["HI-1,25(OH)2D3 binding was assessed in highly purified brush border membranes (46)  hydroxylapatite assay was continued to completion. Under these assay conditions, the specific [:'H]1,25(OH)2D3 binding detectable in brush border membranes was negligible, even when the membranes were incubated at much higher tissue concentrations than those used for the chromatin preparations herein. Note that this experiment does not necessarily disprove the presence of specific ['H]1,25(OH)2Ds binding components in brush border membranes; however, it does ensure that these sites, if present, would not be detected by the hydroxylapatite assay as defined herein.
Cytosol and Nuclear Distribution of Unoccupied 1,25-(OH),Ds Receptors in Target Tissues of Chicks a n d Rats-The question of whether the nuclear association of unoccupied 1,25(OH)2D3 receptors is unique to chick intestinal mucosa is important in establishing the physiological and biochemical relevance of this observation. Therefore, the cytosol and nuclear distribution of 1,25(OH)*D3 receptors in TED buffer was assessed in parathyroid gland and kidney of rachitic chicks, in intestinal mucosa, kidney, and testes of vitamin D-deficient rats, and in an established osteoblast-like cell line from mouse calvaria (49). As shown in Table 11, in all these tissues the majority (81.7 % 0.4% S.E.) of the unoccupied 1,25(OH)2D~ receptors were associated with the chromatin preparation upon homogenization in TED. Thus, the nuclear association of unoccupied 1,25(OH)2D3 receptors results from an intrinsic property of the receptors or of the vitamin D hormone system, rather than a property of chick intestinal mucosa preparations.

DISCUSSION
These data clearly demonstrate that unoccupied 1,25-(OH)*Ds receptors are associated with nuclear and chromatin fractions of target tissues homogenized in low salt buffers similar to those buffers employed in studying other steroid hormone receptors (23, 24). Indeed, the relative proportion of the unoccupied 1,25(OH)2D3 receptors in the chromatin decreases with the calculated ionic strength of the homogenization buffer (Fig. 7 ) . These results raise the question of whether previously reported cytosol 1,25(OH)2D3 receptors represent instead receptors closely associated with nuclear components which are solubilized at the ionic strengths (100 to 300 mM) of the buffers originally used for tissue preparation. That the presence of unoccupied 1,25(OH)2D3 receptors in nuclei is not unique to chick intestinal mucosa (Table 11) emphasizes the fact that this phenomenon may represent an inherent property of the 1,25(OH)*D.3 receptor, or its endocrine system, or both, and is not a peculiarity of the behavior of mucosal tissue per se.
The observations reported here concerning the subcellular distribution of unoccupied 1,25(OH)2D3 receptors in vitro are consistent with reports from several other laboratories. In studies in rat intestinal mucosa (8,9) and in mouse bone cells (16,17) the inclusion of 0.3 M KC1 in buffers proved necessary for detecting cytosol 1,25(OH)2D3 receptors. Additionally, the number of 1,25(OH)zD3 receptors detected in cytosol from the shell gland of rachitic chicks (58) improved markedly when KTED was used instead of STKM? Other reports of cytosolic 1,25(OH)2D,7 receptors in multiple target tissues (7-17) have employed either STKM or KTED for homogenization. In a preliminary series of experiments, Lawson and Wilson previously observed unoccupied nuclear binding sites for 1,25(OH)2D:3 in uitro, even in purified nuclei (20). Importantly, due both to the increased receptor stability and to the simplicity of the Scatchard characteristics, the system described herein is more suitable for assessing the significance of these apparent nuclear binding sites.
The described distribution of the unoccupied 1,25(OHLD3 receptors upon homogenization in T E D could result from several phenomena: (a) nuclear or chromatin localization in uiuo; ( b ) nuclear proximity in vivo leading to nuclear association in uitro; ( c ) ionic charges on the receptor molecule resulting in adsorption to oppositely charged ionic species in nuclei or chromatin upon nuclear damage in uitro. Thus the described results may not reflect the true in vivo localization of unoccupied 1,25(OH)2D3 receptors. However, all of these possible explanations would provide important information on either the physiological or biochemical properties of the 1,25(OH)?D3 receptor. Unfortunately, the extreme susceptibility of the receptor distribution patterns to ionic strength (Fig. 7) will make resolution of these mechanisms difficult. An additional complication is the possibility that under some conditions EDTA may enhance the nuclear association of these receptors. Thus, nuclear and chromatin purification techniques, which subject nuclei to ionic strength changes, osmotic effects, or EDTA, may give 1,25(OH)*Ds receptor distribution data which are virtually uninterpretable. Therefore, resolution of the actual mechanisms of the phenomena described herein wiU be accomplished most effectively when methods become available for observing the receptor molecule in situ in the absence of ligand.
Nevertheless, in several respects, the present results are reminiscent of the well established concept that unoccupied W. A. Coty, personal communication.
receptors for triiodothyronine (T3) are located in nuclei and are indeed intrinsic non-histone chromosomal proteins (41,42,591. These T3 receptors are located in nuclei and chromatin in low salt buffer and both occupied and unoccupied receptors can be extracted by buffer containing 0.4 M KC1 (41,42). Importantly, De Groot and co-workers observed that STKM solubilizes the unoccupied T3 receptors from nuclei (39). Additionally, the rebinding of these acidic Ts receptors to chromatin is nonsaturable and is not tissue specific (41,42). Similar properties have been observed for the 1,25(OHhD3 receptors.2 Although it is presently unclear whether these similarities in receptor characteristics are physiologically meaningful, the parallels are certainly provocative. In summary, in low salt buffers unoccupied 1,25(OH)2D3 receptors are associated with nuclear and chromatin fractions of target tissues. Whether this phenomenon represents a true nuclear localization of these unoccupied receptors in vivo or an unusually high affinity of cytoplasmic 1,25(OH)2D3 receptors for nuclei or chromatin in uitro has not been established. However, operationally this characteristic presents many advantages. For example, 1,25(0H)*D3 receptor quantitation is best accomplished in TED-chromatin (25). Also, receptor visualization by sucrose density gradient ultracentrifugation and gel filtration chromatography is easier in chromatin extracts because of the virtual absence of the 25(0H)D3-binding component. Additionally, competition analyses, where the vitamin D-binding protein also competes with the receptor for analog binding, can now be easily and unambiguously reassessed in the absence of this 25(OH)D3 binding protein by the TED-chromatin assay. The absence of other cytoplasmic proteins renders the chromatin-associated receptors less susceptible to degradation.' Perhaps most importantly, the increased receptor numbers and the purity of the chromatin preparation may help identify 1,25(OH)'Ds receptors in target tissues where they have been previously overlooked (ex. testes, Table  11). Thus, regardless of the physiological significance of the nuclear association of unoccupied 1,25(OH)?Ds receptors, this attribute provides an important vehicle for further biochemical investigations.