The functional status and mechanism of increased VDR in GHS rats were investigated. Basal VDR and calbindins were increased in GHS rats. 1,25(OH)2D3 increased VDR and calbindins in controls but not GHS rats. VDR half-life was prolonged in GHS rats. This study supports the mechanism and functional status of elevated VDR in GHS rats.
Introduction: Genetic hypercalciuric stone-forming (GHS) rats form calcium kidney stones from hypercalciuria arising from increased intestinal calcium absorption and bone resorption and decreased renal calcium reabsorption. Normal serum 1,25-dihydroxyvitamin D3 ‘1,25(OH)2D3’ levels and increased vitamin D receptor (VDR) protein suggest that high rates of expression of vitamin D-responsive genes may mediate the hypercalciuria. The mechanism of elevated VDR and state of receptor function are not known.
Materials and Methods: GHS and non-stone-forming control (NC) male rats (mean, 249 g), fed a normal calcium diet, were injected intraperitoneally with 1,25(OH)2D3 (30 ng/100 g BW) or vehicle 24 h before cycloheximide (6 mg/100 g, IP) and were killed 0–8 h afterward. Duodenal VDR was measured by ELISA and Western blot, and duodenal and kidney calbindins (9 and 28 kDa) were measured by Western blots.
Results and Conclusions: Duodenal VDR protein by Western blot was increased 2-fold in GHS versus NC rats (633 ± 62 versus 388 ± 48 fmol/mg protein, n = 4, p < 0.02), and 1,25(OH)2D3 increased VDR and calbindins (9 and 28 kDa) further in NC but not GHS rats. Duodenal VDR half-life was prolonged in GHS rats (2.59 ± 0.2 versus 1.81 ± 0.2 h, p < 0.001). 1,25(OH)2D3 prolonged duodenal VDR half-life in NC rats to that of untreated GHS rats (2.59 ± 0.2 versus 2.83 ± 0.3 h, not significant). This study supports the hypothesis that prolongation of VDR half-life increases VDR tissue levels and mediates increased VDR-regulated genes that result in hypercalciuria through actions on vitamin D-regulated calcium transport in intestine, bone, and kidney.
GENETIC HYPERCALCIURIC STONE-FORMING (GHS) rats form calcium kidney stones caused by the hypercalciuria and supersaturation of the urine with respect to calcium salts.(1–6) Whereas the genetic basis for the phenotype remains unknown,(7) the source of the hypercalciuria(8) is caused by intestinal calcium hyperabsorption,(9,10) increased bone resorption,(11–13) and decreased tubule calcium reabsorption.(14) There is evidence that the alterations in calcium transport may be caused by a pathologic increase in vitamin D receptor (VDR) in intestine,(10) bone,(11) kidney,(10) and splenic monocytes (unpublished observations). A 2-fold increase in duodenal and kidney cortical homogenate capacity to specifically bind3H-1,25-dihydroxyvitamin D3 [1,25(OH)2D3] without a change in ligand binding affinity(10) strongly suggests that the 2-fold increase in VDR by Western blotting largely reflects a biologically active receptor. The cause of the elevated VDR is not known, but may involve the observed hyper-responsiveness of VDR gene expression in response to 1,25(OH)2D3 administration.(15) However, VDR protein levels are elevated at a time when serum 1,25(OH)2D3 levels are normal, not elevated,(9) and VDR mRNA levels have been reported to be normal or low by Northern blot.(10,15) Furthermore, previous studies have shown no significant differences in duodenal VDR mRNA sequence, transcription rate, or turnover rate between GHS and control rats.(15) The prolonged in vivo half-life of GHS rat intestinal VDR mRNA suggests that altered post-transcriptional regulation of VDR protein synthesis may play an important role in determining VDR levels.(15) Thus, there are two remaining possibilities that can explain the increased VDR protein level in GHS rats: an increase in half-life (stability) and/or an increase in protein translation (synthesis). This study was undertaken to determine whether the accumulation of GHS rat intestinal VDR is caused by protein stability, whether 1,25(OH)2D3 can alter the turnover of its receptor, and whether the synthesis of vitamin D-dependent gene products is regulated by the VDR in GHS rats.
MATERIALS AND METHODS
Animals and protocol
Male Sprague-Dawley rats were used as controls, and male GHS rats born into the GHS colony created by the breeding of spontaneously hypercalciuric male and female rats were fed normal rat chow and tap water ad libitum. Animals were fasted overnight and randomly assigned to intraperitoneal injection of either cycloheximide 6 mg/100 g body weight (BW) in saline or 0.1 ml saline (control). The dose of cycloheximide was selected because it causes maximal inhibition of protein synthesis in rats.(16) Groups of rats (n = 4/experimental group) were killed at baseline (0 time) and 2, 4, 8, and 10 h after cycloheximide injection. In other experiments, rats were injected intraperitoneally with 1,25(OH)2D3 24 before death at doses ranging from 10 to 200 ng/100 g BW. Rats were placed under deep ether anesthesia and were exsanguinated through the abdominal aorta. The experimental protocol was approved by the University of Chicago Animal Care and Use Committee.
Immediately after exsanguination, the abdominal cavity was opened, and 12 cm of proximal duodenum was removed and placed in ice-cold PBS. The segments were rinsed twice in PBS and opened by a longitudinal incision. The mucosal epithelial layer was scraped from the underlying muscle coats, placed in SDS containing Laemelli buffer (1% SDS, 50 mM Tris pH 6.8, 10% glycerol, 1.0% 2-mercaptoethanol, 0.005% bromophenol blue), and boiled for 5 minutes. The samples were flash-frozen at −70°C for subsequent analysis of protein content and VDR and calbindin proteins by Western blotting.
VDR by Western blotting
Protein separation by Western blotting was performed using a modification of techniques described previously for semiquantification of VDR.(10) Twenty micrograms of mucosal aliquots in Laemelli buffer was loaded onto gels, and a standard curve was created by the addition of purified recombinant human VDR (rhVDR) at concentrations ranging from12.5 to 100 pg. VDR standards (Calbiochem) and VDR in tissue samples were resolved on 10% SDS-PAGE at 80 V and transferred onto polyvinylidene difluoride membranes (PVDF; Immobilon-P; Millipore) overnight by electroblotting at 10 μA. Completeness of protein transfer was determined by Coomassie blue staining. Nonspecific binding of immunoglobulins was minimized by incubation of the membranes in Tris-buffered saline containing Tween-20 (TBS-T; 20 mM Tris base, 137 mM NaCl, 0.5% Tween-20) in the presence of 5% nonfat milk powder for 2 h at room temperature. Anti-chick VDR monoclonal antibody (clone 9A7γ)(17) to VDR was added to membranes at 1:2500 dilution in 1% bovine serum albumin (BSA) and 3 mM NaN3 in TBS-T and incubated overnight. Membranes were washed three times with TBS-T (once with TBS-T with 5% NaCl) and incubated with a horseradish peroxidase (HRP)-rabbit anti-rat secondary antibody (JCN Biomedicals) at 1:2000 dilution for 2 h. After washing with TBS-T, membranes were exposed to the chemiluminescent agent LumiGlo (Kirkgaard and Perry Laboratories) for 1 minute. Membranes were exposed to Kodak X-Omat AR film for 2–30 minutes, and the intensities of the protein bands were quantified by scanner (AMBIS; Scanalytics). A representative duodenal VDR Western blot is shown in Fig. 1. A less intense band at 54 kDa observed in some but not all Western blots could be abolished by additional washes and likely reflects immunoglobulin accumulation in tissues. Tissue samples were blinded to the technician and were therefore loaded in a randomized order. The assignment code was revealed only after all samples were analyzed.
Calbindin (9 kDa) by Western blotting
Calbindin (9 kDa) standards (Sigma) and 30 μg of duodenal mucosal protein were resolved by 16.5% SDS-PAGE and transferred to PVDF membranes by electroblotting (50 μA) overnight with 10% 10× Tris-glycine buffer (pH 7.4) and 20% methanol. Membranes were blocked with TBS-T and 5% milk powder for 2 h. Membranes were washed with TBS-T and incubated in the presence of a 1:10,000 dilution of polyclonal rabbit anti-rat antibody raised against 9-kDa calbindin and 5% milk powder in TBS-T for 1 h. Membranes were washed three times with TBS-T (once with TBS-T with 5% NaCl) and incubated with a 1:10,000 dilution of HRP-goat rat anti-rabbit secondary antibody (Gibco-BRL) in TBS-T and 5% milk powder for 1 h. Membranes were washed and subjected to autoradiography as described above for the VDR Western blot. A representative Western blot of 9-kDa calbindin is shown in Fig. 2A.
Calbindin (28 kDa) by Western blotting
Calbindin (28 kDa) standards and 2.5 μg of kidney protein were resolved by 10% SDS-PAGE and transferred to PVDF membranes by electroblotting (50 μA) overnight with 10% 10× Tris-glycine buffer (pH 7.4) and 15% methanol. Membranes were blocked with TBS-T and 5% milk powder for 2 h. Membranes were washed with TBS-T and incubated in the presence of a 1:10,000 dilution of monoclonal mouse anti-chick antibody to 28-kDa calbindin (Sigma Chemical) and 5% milk powder in TBS-T for 1 h. Membranes were washed three times with TBS-T (once with TBS-T with 5% NaCl) and incubated with a 1:10,000 dilution of HRP-rabbit anti-mouse secondary antibody (Calbiochem) in TBS-T and 5% milk powder for 1 h. Membranes were washed and subjected to autoradiography as described above for the VDR Western blot. A representative Western blot of 28-kDa calbindin is shown in Fig. 2B.
Duplicate aliquots of mucosal samples were subjected to the Bradford protein assay.(18)
Duodenal extracts were stored at −70°C and were assayed in a fashion that blinded the identity of the samples to those conducting the assay. The assay was performed as previously described(19) and involves VDR binding between monoclonal antibodies IVG8C11 and B-VD2F12, which are specific for two epitopes of the VDR. Quantification occurs by detection of alkaline phosphatase that is bound through avidin to one antibody. Absorbance was read at 410 nm and compared with a standard curve created from purified recombinant human VDR. Samples were assayed in triplicate.
In a preliminary experiment, 24 GHS rats were injected with 6 mg/100 g BW cycloheximide and were killed 0, 1, 2, 4, 6, and 8 h afterward. Duodenal mucosa was prepared, and each sample was assayed for VDR by the ELISA or a standard saturation binding assay.(10) In the saturation binding assay, specific binding sites were calculated based on the binding of3H-1,25(OH)2D3.(10) As maximal cytosolic fraction VDR binding of3H-1,25(OH)2D3 was reached at 3.0 nM, VDR in each sample from all time-points was determined at a single concentration of3H-1,25(OH)2D3 (3.0 nM). The calculated half-life of VDR was 2.2 h compared with 2.59 h by ELISA (Table 1). The half-life of VDR did not differ when the saturation binding assay was expressed as fmol VDR per milligram protein or fentomol per microgram tissue DNA. Based on this good agreement between the ELISA and the saturation binding assay, subsequent half-life experiments were performed using the VDR ELISA.
Table Table 1. In Vivo Duodenal VDR Half-Life
Data were analyzed using Minitab statistical software (State College, PA, USA). Differences between group means were analyzed by paired t-test when two groups were compared. Multiple group comparisons were conducted using ANOVA and Tukey's HSD procedure. The VDR half-life was determined using a logarithmic conversion to a first-order function and variance calculated by the method of Armitage and Berry.(20) Group means differing by p < 0.05 were considered statistically significant. Data are expressed as mean ± SE.
Duodenal VDR levels and response to 1,25(OH)2D3
Under basal conditions, duodenal mucosa from GHS rats contained VDR that was two to four times greater than in duodenum from normal control (NC) rats (example from representative experiment: 633 ± 62 versus 388 ± 48 fmol/mg protein, n = 4, p < 0.02). Twenty-four hours after a single dose of 1,25(OH)2D3 (200 ng/100 g BW), VDR was increased in NC and GHS rats (Fig. 3), whereas lower doses of 1,25(OH)2D3 did not raise VDR levels in either NC or GHS rats. The lower doses of 1,25(OH)2D3 tested (10 and 30 ng/100 g BW) did not increase duodenal VDR in NC or GHS rats by 72 h (data not shown).
Kidney VDR levels
Under baseline conditions, renal cortical VDR levels were greater in GHS rats (Figs. 4A and 4B). Figure 4A shows that a single intraperitoneal dose of 1,25(OH)2D3 (200 ng/100 g BW) increased VDR in both NC and GHS rats by 24 h. Figure 4B shows that a lower dose of 1,25(OH)2D3 (30 ng/100 g BW) increased VDR in NC rats by 24, 48, and 72 h and in GHS rat kidney by 72 h.
In vivo VDR half-life in duodenum
Cycloheximide administration caused a time-dependent decline in duodenal VDR protein in GHS rats. The disappearance of GHS VDR was slower than that observed in NC duodenum (Fig. 5A). By 2 h after cycloheximide administration, GHS VDR had declined to 50% of baseline and continued to decline to about 15% of baseline by 8 h. Using the time-points in a second-order regression model, the in vivo half-life of duodenal VDR was calculated at 2.59 h (Table 1). Compared with VDR half-life in NC rats, the VDR half-life in GHS rats was prolonged by 43%. The administration of a small dose of 1,25(OH)2D3 (30 ng/100 g BW, IP) 24 h before cycloheximide administration prolonged the half-life in NC rats to a level that is not different from that in GHS rats under basal conditions (Fig. 5B). The same dose of 1,25(OH)2D3 in GHS rats tended to shorten the half-life toward that of the 1,25(OH)2D3-stimulated NC rats, so that 1,25(OH)2D3 abolished the basal differences between the VDR half-lives of the two groups (Table 1).
Duodenal and kidney 9-kDa calbindin
Under basal conditions, 9-kDa calbindin was greater in GHS rat duodenum (Figs. 6A and 6B). Figure 6A shows that 24 h after 1,25(OH)2D3 administration (10, 30, and 200 ng/100 g BW), 9-kDa calbindin levels were increased in duodenum from NC rats. In contrast, the same doses of 1,25(OH)2D3 did not increase 9-kDa calbindin levels in GHS rat duodenum above the high baseline values. Figure 6B shows that by 72 h after 1,25(OH)2D3 administration (30 ng/100 g BW), NC 9-kDa calbindin levels were increased above baseline, whereas the same dose did not increase GHS duodenal 9-kDa calbindin levels by 72 h.
Calbindin (9 kDa) was detected in kidney cortical tissue from both NC and GHS rats, with 5-fold higher levels in GHS rats (Fig. 7). Doses of 1,25(OH)2D3 from 10 to 200 ng/100 g BW caused no further increase in 9-kDa calbindin in GHS rat kidney. However, the highest dose of 1,25(OH)2D3 (200 ng/100 g BW) did increase 9-kDa calbindin in NC kidney cortical tissue by 24 h (Fig. 7).
Kidney 28-kDa calbindin
Baseline 28-kDa calbindin levels were higher in GHS rat kidney (Fig. 8). A single dose of 1,25(OH)2D3 (30 ng/100 g BW) increased GHS rat 28-kDa calbindin by 48 h and was sustained for 72 h (Fig. 8), whereas NC rat kidney 28-kDa calbindin did not increase after doses of 1,25(OH)2D3 of up to 200 ng/100 g BW (data not shown).
High VDR levels in GHS rat intestine, bone, and kidney seem to mediate the increased intestinal calcium transport,(9,10) enhanced bone resorption,(11,12) and decreased tubule calcium reabsorption(14) that create and sustain hypercalciuria in GHS rats. This study shows that, in the GHS rat duodenum, elevated baseline VDR levels are caused in part by a prolongation of the half-life of the receptor. Because in vivo half-life was measured after cycloheximide administration, new VDR synthesis was excluded as the cause of the increase in VDR protein. Duodenal VDR protein half-life in NC rats could be prolonged to the level found in GHS rats by the administration of a small dose of 1,25(OH)2D3 (30 ng/100 g BW). However, this same dose of 1,25(OH)2D3 caused no further alteration in stability of the receptor in GHS rats, suggesting that receptor stability may have an upper limit of regulation. Decreased VDR degradation may be responsible for the increased VDR content in other GHS tissues including kidney and bone,(10,11) but VDR turnover in these tissues was not assessed directly in this study.
In a previous study, DNA sequencing of duodenal VDR cDNA failed to reveal any difference between GHS and Sprague-Dawley control rats,(15) suggesting that mechanisms other than an exon mutation of the VDR gene must be involved in generating the high levels of tissue VDR. Using semiquantitative Northern blotting, Li et al.(10) and Yao et al.(15) reported duodenal VDR mRNA levels to be normal to low in GHS rat duodenum, and Yao et al.(15) found elevated VDR in GHS rat kidney cortex.(15) These two studies strongly suggest that the elevated VDR protein in GHS rats may not be caused by an increase in VDR gene expression.
However, the regulation of VDR gene expression may be altered in GHS rats as suggested by the study of Yao et al,(15) in which administration of 1,25(OH)2D3 at a small dose (30 ng/100 g BW) that did not increase duodenal VDR gene expression in wildtype control rats markedly increased VDR mRNA levels in GHS rat duodenum. Furthermore, Yao et al. also found that the same small dose of 1,25(OH)2D3 increased duodenal VDR mRNA levels in GHS rats through prolongation of the half-life of the VDR mRNA.(15) In this study, this same low dose of 1,25(OH)2D3 increased in vivo GHS duodenal VDR protein half-life, but increased kidney VDR only 2- and 3.5-fold in GHS and NC rats, respectively, and only after 72 h. This dose did not increase GHS or control duodenal VDR protein levels. Thus, the available studies suggest that GHS rats are characterized by a hyper-responsiveness to small doses of 1,25(OH)2D3 at the level of gene expression and turnover of mRNA and protein but not at the protein translational level.
The high levels of VDR in GHS rat duodenum and kidney measured by Western blot and ELISA assays in this study are in good agreement with previous estimates of VDR using saturation binding kinetics(10) and Western blotting using the 9A7 monoclonal antibody.(10,19) In this study, the predominant species of VDR migrated as a 48-kDa protein in duodenal mucosal and kidney cortical tissue in both GHS and NC rats.
Calbindin (9 kDa) and 28-kDa calbindin proteins were measured to test the functionality of the high VDR in GHS rat duodenum and kidney. Calbindins (9 and 28 kDa) are products of unique 1,25(OH)2D3-regulated genes,(21,22) and both are thought to play important roles in calcium transport across intestinal (9-kDa calbindin)(22) and renal tubule (9- and 28-kDa calbindin)(22–24) epithelia. In this study, 9-kDa calbindin levels were higher in GHS rat duodenum and likely participate in the 4- to 5-fold increase in calcium active transport in GHS rats.(9) The kinetics of transcellular calcium transport includes an initial calcium influx across the brush border membrane followed by movement across the cell and extrusion across the plasma basolateral membrane.(25) 1,25(OH)2D3 administration stimulates 9-kDa calbindin gene expression before the first detectable increase in calcium absorption,(25) suggesting that 9-kDa calbindin may play a role in the early events of calcium transport across the intestinal epithelial cell. 1,25(OH)2D3 stimulates 9- and 28-kDa calbindin gene expression and increases protein levels in intestine and kidney, respectively.(21–24) In a previous study, GHS rat duodenal 9-kDa calbindin mRNA levels were low at baseline and rose about 6-fold after 1,25(OH)2D3 administration.(15)
Renal 28-kDa calbindin mRNA levels were not different in GHS and NC control rats, and 1,25(OH)2D3 stimulated 28-kDa calbindin gene expression about 14-fold in GHS rats, whereas gene expression in kidney from NC rats did not change.(15) Thus, in GHS rats, 9- and 28-kDa calbindin gene expressions are sensitive to small doses of 1,25(OH)2D3. In this study, we tested, in GHS, rats whether the hypersensitivity of VDR gene expression to 1,25(OH)2D3 would result in marked accumulation of duodenal and renal cortical 9-kDa calbindin and renal 28-kDa calbindin protein levels. 1,25(OH)2D3 failed to increase duodenal and renal cortical 9-kDa calbindin protein levels in GHS rats and caused a limited increase in these proteins in NC rats only at the highest dose used (200 ng/100 g BW). 1,25(OH)2D3 increased 28-kDa calbindin levels in GHS but not NC rat kidney. Like the responses observed for VDR, the limited response of the 9- and 28-kDa calbindin vitamin D-sensitive genes to 1,25(OH)2D3 may be caused by either a maximal production of the 9- and 28-kDa calbindin proteins under basal conditions in GHS rats or that a single dose of 1,25(OH)2D3 was sufficient to stimulate 9- and 28-kDa calbindin gene expression(15) but insufficient to sustain an increase in calbindin protein levels.
Calbindin (28 kDa) is the dominant calcium-binding protein in mammalian kidney(23,24) and has been implicated in tubule calcium transport because of its localization to the distal convoluted tubule,(26,27) which is the site of significant calcium reabsorption and/or secretion.(28) Calbindin (9 kDa) is found in rat neonatal kidney(21,22) and may persist in the kidney, albeit at low levels.(29) In addition, evidence from 28-kDa calbindin knockout mice suggests that 28-kDa calbindin may protect against hypercalciuria.(30) A function of 9-kDa calbindin in renal tubule calcium transport is supported by its localization along the more distal portions of the nephron that are involved in calcium transport. Renal 28-kDa calbindin also co-localizes along the nephron with VDR and the parathyroid hormone (PTH)/PTH-related peptide (PTHrP) receptor(31,32) and the Ca-sensing receptor (CaR),(32) which are thought to be involved in renal calcium transport.(31) The potential roles of 28- and 9-kDa calbindin and the CaR in the reduced renal tubule calcium transport in GHS rats were not directly investigated in this study.
In conclusion, GHS rats have increased protein VDR levels in both duodenum and kidney through both increased gene expression and stabilization of the receptor protein. The elevated 9-kDa calbindin levels in GHS rat duodenum and the increased 9- and 28-kDa calbindin levels in GHS rat renal cortex strongly suggest that the high VDR levels are functional and mediate the increases in calbindin gene expression and protein production as part of the pathophysiology of the hypercalciuria.
The authors thank Vrishali Tembe for technical assistance. This research was supported by National Institutes of Health Grant PO1 DK56788.