Vitamin D and the Kidney

Abstract

The kidney is essential for the maintenance of normal calcium and phosphorus homeostasis. Calcium and inorganic phosphorus are filtered at the glomerulus, and are reabsorbed from tubular segments by transporters and channels which are regulated by 1α,25-dihydroxyvitamin (1α,25(OH)₂D) and parathyroid hormone (PTH). The kidney is the major site of the synthesis of 1α,25(OH)₂D under physiologic conditions, and is one of the sites of 24,25-dihydroxyvitamin D (24,25(OH)₂D) synthesis. The activity of the 25(OH)D-1α-hydroxylase, the mixed function oxidase responsible for the synthesis of 1α,25(OH)₂D, is regulated by PTH, 1α,25(OH)₂D, fibroblast growth factor 23 (FGF23), inorganic phosphorus and other growth factors. Additionally, the vitamin D receptor which binds to, and mediates the activity of 1α,25(OH)₂D, is widely distributed in the kidney. Thus, the kidney by regulating multiple transport and synthetic processes is indispensible in the maintenance of mineral homeostasis in physiological states.

1. INTRODUCTION

The kidney has a unique function in mineral homeostasis. Both calcium and phosphorus are filtered, reabsorbed and excreted in the urine to a varying degree, generally in amounts that reflect the endogenous requirements of the two substances. The vitamin D-endocrine system plays a key role in controlling the renal excretion of both calcium and phosphorus. The reabsorption of calcium in the kidney is controlled by several factors, including 1α,25-dihydroxyvitamin D (1α,25(OH)₂D) [1–5]. The kidney is the major site of synthesis of 1α,25(OH)₂D, the active, hormonal form of vitamin D [6–8]. The renal 25-hydroxyvitamin D₃ 1α–hydroxylase (1α–hydroxylase) and 25-hydroxyvitamin D-24-hydroxylase (24-hydroxylase), and other 1α,25-dihydroxyvitamin D₃ and vitamin D analog metabolizing enzymes are expressed in kidney tissue [9–11]. The kidney expresses the vitamin D receptor (VDR) [11, 12]. The kidney expresses several 1α,25(OH)₂D-dependent proteins that are important in calcium reabsorption e.g. the plasma membrane calcium pump (PMCa) [13–15], the epithelial calcium channel (ECaC) [16], the sodium calcium exchanger [17], and the calbindins [13–15]. Finally, the kidney has an important role in the control of plasma phosphate, which following filtration in the glomerulus is reabsorbed in nephron segments at a rate influenced by many of the same hormones and factors involved in calcium regulation [18–23].

To understand how vitamin D influences the efficiency of calcium and phosphorus reabsorption, a brief review of normal calcium and phosphorus handling by the kidney follows.

2. CALCIUM AND PHOSPHORUS FILTRATION AND REABSORPTION IN THE KIDNEY

1.1 Calcium handling by the kidney

On account of protein-binding, slightly more than half of total plasma calcium (plasma concentration 10 mg/dL or 2.5 mM) is filtered at the glomerulus [24]. The concentration of calcium in the glomerular filtrate is the same as that of plasma ultra-filterable calcium [25–27]. Assuming a glomerular filtration rate of 140 L per day and an ultrafiltrable calcium of 5.5 mg/dL, about 8000 mg of calcium are filtered by the kidney in 24 hours; 98% of the filtered load of calcium is reabsorbed, resulting in an excreted calcium of about 150–200 mg/24 h [1, 24]. Fifty to sixty percent of the filtered load of calcium is reabsorbed in the proximal tubule [2, 3, 28] by sodium-dependent, para-cellular mechanisms. Inhibition of sodium-potassium ATPase activity by ouabain reduces the amount of calcium reabsorbed in the proximal tubule, as does the substitution of sodium with lithium [29]. A reduction of tubular sodium reabsorption by volume expansion inhibits proximal tubule calcium reabsorption, while an increase in sodium reabsorption associated with volume contraction, enhances proximal tubule calcium reabsorption [2, 29]. Importantly, calcium reabsorption in the proximal tubule is not altered by thiazide diuretics, hormones such as PTH or 1α,25(OH)₂D₃, or by hydrogen ions [2, 3, 28, 29]. Vitamin D-dependent proteins that play a role in trans-cellular calcium transport, such as the calbindins, ECaCs, and the PMCa pump are either not expressed in the proximal tubule, or are expressed in low amounts when compared with the amounts expressed in the distal tubule. Minimal amounts of calcium are absorbed in the descending loop and the thin ascending limb of the loop of Henle. About 20% of the filtered load of calcium is reabsorbed in the thick ascending limb of the loop of Henle; another 10–15% of filtered calcium is reabsorbed in the distal tubule with the remaining 5% being reabsorbed in the collecting duct [2, 3, 28, 29]. The movement of calcium in the distal nephron is energy dependent and occurs against a concentration gradient; furthermore, the tubular lumen is electro-negative and becomes progressively more so towards the end of the distal tubule [30, 31]. In the distal nephron, calcium reabsorption can be dissociated from sodium reabsorption by thiazide diuretics which inhibit sodium reabsorption but enhance calcium reabsorption [30, 31]. In contrast to the proximal tubule, where hydrogen ion have no effect on calcium reabsorption, in the distal nephron hydrogen ions inhibit calcium reabsorption.

1.2 Phosphate handling by the kidney

Inorganic phosphate in the serum is freely filtered at the glomerulus [18–20]. About 80% of filtered phosphorus is reabsorbed in the proximal tubule [18]. The amount of phosphorus reabsorbed in the proximal tubule is greatest in the first half of the proximal tubule with some further phosphorus reabsorption occurring in the pars recta [18]. Little or no phosphorus reabsorption occurs in the loop of Henle or the distal tubule. The reabsorption of phosphate is sodium-dependent and is mediated by sodium-phosphate co-transporters (Na-Pi IIa,SLC34A1; Na-Pi IIc, SLC34A3, and Pit-2, SLC20A2) [32, 33]. Na-Pi IIa activity is increased by a low phosphate diet and decreased by PTH.[34–37] The recently described phosphatonins, fibroblast growth factor-23 (FGF-23), MEPE and secreted frizzled related protein-4 (sFRP-4), inhibit sodium dependent phosphate transport [23, 38]. In opossum kidney (OK) cells, NaPi II is internalized from the cell membrane in response to FGF-23 and sFRP-4, similar to the effects of PTH [39, 40]. Additional factors involved in phosphorus reabsorption are noted in Table 1.

Table 1.

Factors that alter renal phosphate excretion

Increase	Decrease
1. High phosphate diet	1. Low-phosphate diet
2. Parathyroid hormone	2. Parathyroidectomy
3. Calcitonin	3. Thyroxine
4. Chronic vitamin D	4. Acute vitamin D
5. Glucagon	5. Insulin
6. Glucocorticoids	6. Growth hormone
7. Volume expansion	7. Volume contraction
8. Increased pCO2	8. Decreased pCO2
9. Chronic Acidosis
10. Starvation
11. Diuretics
12. “Phosphatonin”
FGF-23
sFRP-4

3. ROLE OF THE KIDNEY IN THE METABOLISM OF 25OH D

3.1. The 25-hydroxyvitamin D₃-1α-hydroxylase and the synthesis of 1α,25(OH)₂D₃

25-hydroxyvitamin D₃-1α-hydroxylase is a multi-component, cytochrome P-450 containing enzyme present in mitochondria of renal proximal tubular cells which transfers electrons from NADPH to the cytochrome P450, Cyp27B1 [7, 41–56]. The latter, using molecular oxygen, converts 25-hydroxyvitamin D₃ (25(OH)D₃) to 1α,25(OH)₂D₃ and water. Nephrectomy greatly decreases circulating 1α,25(OH)₂D₃ concentration in vivo except during pregnancy, granuloma-forming diseases, and lymphomas associated with the ectopic production of 1α,25(OH)₂D₃ [57–60]. While the kidney is the major site of 1α,25(OH)₂D₃ production, 25(OH)D₃-1α-hydroxylase activity has been found in vitro in several other cell types [49, 61–72]. In vitro, chick renal epithelial cells in culture, mammalian nephron segments and homogenates derived from avian and mammalian (mostly rodent) renal cells metabolize 25(OH)D₃ to 1α,25(OH)₂D₃ [73–77]. Proximal and distal tubular segments synthesize 1α,25(OH)₂D₃ [78, 79]. Zehnder et al have demonstrated the presence of 1α-hydroxylase mRNA and protein in the distal convoluted tubule, cortical collecting duct, thick ascending limb of the loop of Henle, and Bowman’s capsule [79]. Recent experiments in which the 25(OH)D₃-1α-hydroxylase cytochrome P450 gene (Cyp27B1) was deleted in mice, point to the central role of this enzyme in vitamin D metabolism [80]. Table 2 summarizes some of the key factors known to regulate the activity of this enzyme in vivo and in vitro. The major regulators appear to be PTH, inorganic phosphorus and 1α,25(OH)₂D₃ itself. Fibroblast growth factor 23 (FGF 23), and its binding protein Klotho that is produced mainly in the distal tubule, and which is essential in mediating FGF 23 activity, inhibit 25(OH)D₃-1α-hydroxylase activity and regulate the production of 1α,25(OH)₂D₃ [81–93]. Indeed, FGF 23 may mediate the effects of changing dietary phosphorus on 1α,25(OH)₂D₃ production. Regulators such as 1α,25(OH)₂D₃ which alter the expression of the Cyp27B1 gene have reciprocal effects on the expression of the Cyp24A1 gene [94] that are mediated via vitamin D-response elements in the promoters of the respective genes [95, 96]. In cell culture, the extra-renal 25(OH)D₃-1α-hydroxylase is regulated by nitric oxide and by activation of toll-like receptors [66–72].

TABLE 2.

Effect of increased level or activity of various factors on 1α,25(OH)₂D₃ concentration or 1α-hydroxylase activity

Factor	Animals	Humans	Ref.
Parathyroid hormone	↑	↑	[27, 75, 291–299]
Serum inorganic phosphorus	↓	↓	[127, 300–302]
1α,25(OH)2D3	↓	↓	[291, 303]
Calcium (direct)	?	↓	[304, 305]
Calcitonin	↑,↓,0	↑	[24, 75, 291, 292, 306, 307]
Hydrogen ion	↓	0	[293, 308, 309]
Sex steroids	↑	↑	[126, 310]
Prolactin	↑	0	[311–313]
Growth hormone and insulin-like growth factor-1	↑	↑,↓,0	[225, 305, 314–319]
Glucocorticoids	↓,0	↑,↓,0	[183, 320–323]
Thyroid hormone	?	↓*	[324–326]
Fibroblast growth factor 23/Klotho axis	↓	?	[81–93]
Frizzled related protein 4	↓	?	[23]
Pregnancy	↑	↑*	[327, 328]

↑ Stimulation or increase; ↓, suppression or decrease; 0, no effect; ?, effect not known.

^*Effects may be secondary to changes in calcium, phosphorus or parathyroid hormone. (With permission, modified from Kumar R[94])

3.2 Modeling Cyp27B1 and Computational Docking Studies

Given its critical role in the synthesis of 1α,25(OH)₂D₃, it is important to determine how the Cyp27B1 cytochrome P450 binds 25(OH)D₃. Two energy-minimized homology models for Cyp27B1 were created by manual rebuilding from Cyp24A1 (open-cavity form, PDB ID 3K9V) [97] and Cyp2R1 (closed, PDB 3C6G [98] crystal structures, guided by I-TASSER and PHYRE2 modeling [99–101]. Figure 1A shows putative Cyp27B1 residues of the substrate binding cavity. The “yellow-carbon” Cyp27B1 residues differ from Cyp24A1’s homologous amino acids.

Figure 1.

A. Stereo view of the Cyp27B1 model’s substrate cavity residues above the heme cofactor. B. Highest-scoring 25(OH)D₃ dockings in the Cyp27B1 homology model did not seem plausible, though possible binding modes, see C, were among the top 10 results. D. In contrast, top 3 and many lower scoring dockings of 1α,25(OH)₂D₃ inside the Cyp24A1 crystal structure seemed credible.

Cyp27B1 models, the Cyp24A1 and Cyp2R1 structures and all ligand atom coordinates used were parameterized by Autodock Tools (http://autodock.scripps.edu/resources/adt), with heme Fe charge explored from +0.2 to 1.0, both with & without a partially-charged oxygen bound to the Fe atom. Control Autodock Vina computations [102], chosen for high binding mode accuracy, correctly oriented & positioned Chaps within 0.59 Å RMSD of that in the Cyp24A1 crystal (chain A) [97] and 25(OH)D₃ within 0.99 Å RMSD of the 25(OH)D₃ observed in the Cyp2R1 crystal [103]. The search box covered all molecular extents, with num_mode and exhaustiveness at 500. Predictions concerning the plausibility of the binding modes, and signal-noise level of calculations, were based on computed free energy of binding, clustering of binding modes, proximity to key residues identified in literature, and distances between the known hydroxylated carbon atoms and heme Fe.

Irrespective of whether the Cyp27B1 model was derived from Cyp24A1 (open cavity) or Cyp2R1 (closed) structures, the 25(OH)D₃ docking modes deemed most plausible were not among those top ranked by Autodock Vina’s binding-energy scoring function. Compare the highest-rank 25(OH)D₃ pose, −9.2 kcal/mol, in Figure 1B to the pose in Figure 1C (9^th in rank) that we find somewhat promising at −8.1 kcal/mol. Allowing cavity residues of Cyp27B1 to be flexible during computations did not improve the likely validity of top-scored poses. Interestingly, we found top binding modes for the product of Cyp27B1 metabolism, 1α,25(OH)₂D₃, more believable as the molecules were oriented with A-rings closest to Cyp27B1’s heme Fe, though not in positions able to explain catalysis (best score −9.5 kcal/mol, not shown). Our top poses of 1α,25(OH)₂D₃ docked to the Cyp24A1 crystal structure, with best at −9.0 kcal/mol in Figure 1D, are acceptable predictions that appear nearly equivalent to a pose published by Rhieu et al. (Fig10 in [103]). Template-bias prevented our two homology models of Cyp27B1 from being useful prediction tools for understanding ligand-binding. Crystal structures of Cyp27B1 will be required to definitively establish its structure.

3.3 The 25(OH)D₃-24-hydroxylase and the synthesis of 24,25-dihydroxyvitamin D₃

The 25(OH)D₃-24-hydroxylase enzyme is a multi-component cytochrome P450 enzyme expressed in the kidney as well as many other tissues.[104–125]. The renal 24-hydroxylase is regulated by calcium and phosphorus such that elevated or normal calcium levels induce the 24-hydroxylase activity, whereas, low calcium levels inhibit it [116–118, 126]. Similarly, elevated serum phosphorus concentrations increase 24-hydroxylase activity and low phosphorus diets decrease the enzyme activity [127]. 24-hydroxylase enzyme activity and mRNA expression in the kidney is up-regulated by 1α,25(OH)₂D₃ [128–134]. The effect of 1α,25(OH)₂D₃ is blunted in vitro and in vivo by parathyroid hormone [128–131]. We used antibodies against the 24-hydroxylase cytochrome P-450 to examine the distribution of the enzyme in the human kidney, and found high concentrations of the cytochrome P-450 in distal tubular cells and lower amounts in the proximal tubule [11]. 24-Hydroxylase activity, however, is expressed primarily in the proximal tubule but is also present in more distal segments [135, 136]. It is responsible for the conversion of 25(OH)D₃ and 1α,25(OH)₂D₃ to 24,25-dihydroxyvitamin D₃ (24,25(OH)₂D₃) and 1α,24,25-trihydroxyvitamin D₃, respectively. There is conflicting evidence as to whether 25(OH)D₃ or 1α,25(OH)₂D₃ is the preferred substrate for 24-hydroxylase [137]. Some have reported 1α,25(OH)₂D₃ is the preferred substrate with a K_m approximately 10-fold lower than that for 25(OH)D₃ [138, 139], while others have found K_m values substantially lower for 25(OH)D₃ [140]. It has been suggested that 24,25(OH)₂D₃ has unique properties and actions in cartilage and bone [124, 125, 141, 142], but others have not confirmed these observations [143–149].

3.4 The Structure of the 25(OH)D₃-Cyp24A1

Recently, the structure of the Cyp24A1 in association with CHAPS was reported [97]. The crystal structure shows an open cleft leading to the active-site heme prosthetic group. The entrance to the cleft is flanked by conserved hydrophobic residues on helices A' and G'. The determinants of adrenodoxin recognition are conserved residues from helices K, K'', and L.

3.5 Other vitamin D metabolizing enzymes present in the kidney

The kidney is also capable of transforming 25(OH)D₃ to several other compounds listed in Table 3 [150–170]. The specific physiological roles of these various metabolites are not known with certainty. Several polar metabolites of 1α,25(OH)₂D₃ are formed in the liver, including calcitroic acid and the glucuronide and sulfate conjugates of the hormone; these and small amounts of unchanged dihydroxylated and tri-hydroxylated metabolites of vitamin D are excreted in the urine [171–185]. Many of the transformations that occur with 25(OH)D₃ also occur in the case of 1α,25(OH)₂D₃.

Table 3.

The metabolism of 25OHD₃ by the kidney

25OHD3		24R, 25(OH)2D3		24-keto-25(OH)D3
		25S, 26(OH)2D3		25(OH)D3-lactone
		23S, 25(OH)2D3		23-keto-25(OH)D3

4. EFFECTS OF VITAMIN D, 25(OH)D₃ AND 1α,25(OH)₂D₃ ON THE RENAL HANDLING OF CALCIUM AND PHOSPHORUS

In vitamin D deficiency urine calcium concentrations are low, whereas, with vitamin D excess or intoxication hypercalciuria is frequently present [1]. The excretion of calcium and phosphorus in the urine, however, also reflects the decreased or increased calcium absorption in the intestine, the presence of hypo- or hyper-calcemia, and the presence of diminished or elevated concentrations of circulating PTH. Short term infusions of vitamin D₃, 25(OH)D₃ and 1α,25(OH)₂D₃ decrease phosphate, calcium and sodium clearance relative to the clearance of inulin in parathyroidectomzed dogs [186–189]. We performed similar studies examining the effects of 25(OH)D₃ on renal bicarbonate and phosphate reabsorption [190]. Unlike these earlier studies, we observed that phosphate and bicarbonate reabsorption increased only in intact animals but not in parathyroidectomized animals suggesting the need for PTH. Our observations are similar to those of others [191]. Yamamoto et al. examined the effects of 1α,25(OH)₂D₃ in vitamin D-deficient rats, vitamin D-deficient rats supplemented with dietary calcium to normalize plasma calcium and PTH levels, and vitamin D-replete rats following thyro-parathyroidectomy and the infusion of graded amounts of calcium [192]. Urinary calcium excretion was lower in vitamin D-replete rats than in vitamin D-deficient rats suggesting that vitamin D increased the efficiency of renal calcium reabsorption in the absence of PTH. In a second group of experiments, rats treated in the manner noted above, were TPTX and infused with PTH. The results of this experiment show that a lower dose of PTH is needed to exhibit a comparable effect on renal calcium reabsorption in vitamin D-replete rats when compared to vitamin D-deficient rats. This finding may be explained by in vitro studies of distal convoluted tubule cells in which 1α,25(OH)₂D₃ increased PTH/PTHrp receptor mRNA levels [193]. Micropuncture experiments have demonstrated that 25(OH)D₃ exerts its anti-phosphaturic and hypocalciuric effects in the distal tubule [194, 195]. These effects occur shortly after the administration of 25(OH)D₃ and are therefore independent of conversion of 25(OH)D₃ to 1α,25(OH)₂D₃. 1α,25(OH)₂D₃ enhances the reabsorption of calcium in distal tubular segments of the nephron [196]. These studies are consistent with the localization of many vitamin D responsive proteins exclusively in the distal nephron. In vitro studies show that vitamin D deficiency is associated with decreased calcium uptake in luminal and basolateral membranes derived from distal nephron segments [197]. Cultured rabbit connecting tubule cells show an increase in the transport of calcium when treated with 1α,25(OH)₂D₃ [198]. ECaC1 (TRPV5, present in the apical membrane of the distal convoluted tubule) and ECaC2 (TRPV6, localized to the principal cells of the cortical and medullary collecting ducts) protein and mRNA are diminished in vitamin D deficient rats and increased by 1α,25(OH)₂D₃ [199–201]. 1α,25(OH)₂D₃ increases mRNA and protein expression of the PMCa pump which is involved with active transport of calcium through the basolateral membrane of distal tubule cells [202].

A synthesis of the experimental results suggest that 1α,25(OH)₂D₃ has effects on the distal tubular reabsorption of calcium through several mechanisms (Figure 2). Expression of proteins responsible for distal tubule calcium uptake, intracellular trafficking, and basolateral transport of calcium are responsive to 1α,25(OH)₂D₃.

Figure 2.

Integrated model of active Ca²⁺ reabsorption in the distal part of the nephron. Apical entry of Ca²⁺ is facilitated by ECaC. Ca²⁺ then binds to calbindin-D_28K, and this complex diffuses through the cytosol to the basolateral membrane, where Ca²⁺ is extruded by a Na⁺/Ca²⁺ exchanger and a plasma membrane Ca²⁺-ATPase. The individually controlled steps in the activation process of the rate limiting Ca²⁺ entry channel include 1α,25(OH)₂D₃-mediated transcriptional and translational activation, shuttling to the apical membrane, and subsequent activation of apically located channels by ambient Ca²⁺ concentration, direct phosphorylation and/or accessory proteins. (With permission from Hoenderop, et al [268])

5. DISTRIBUTION AND REGULATION OF VITAMIN D-DEPENDENT PROTEINS IN THE KIDNEY

Several vitamin D-dependent proteins are expressed in the kidney and many of these play a role in calcium and phosphate transport. The VDR, calbindins, PMCa pump, and the ECaCs (TRPV5 and TRPV6) are all found in renal tubule cells and act co-ordinately in the regulation of calcium transport in the nephron [5, 11, 14, 15, 28, 193, 199, 201, 203–206]. Additional vitamin D responsive proteins involved in renal calcium and phosphate transport are listed in Table 4.

Table 4.

Vitamin D responsive proteins in the kidney

Vitamin D receptor

24-hydroxylase

Plasma membrane calcium pump

Epithelial calcium channels, TRPV5 and TRPV6

Calbindin D28K

Calbindin D9K

Calcium sensing receptor

PTH/PTHrp receptor

Sodium-phosphate co-transporter type 2

5.1 1α,25(OH)₂D₃ receptor (VDR) in the kidney

The VDR mediates many, if not all, of the effects of 1α,25(OH)₂D₃ in diverse organs [207–215]. The distribution of the VDR in the kidney has been assessed using a variety of techniques [135, 216–220]. A full discussion regarding its distribution is given in a separate contribution in this volume. Suffice it to say that the VDR is found in varying amounts in different species in proximal and distal tubules [11, 135, 217, 218, 221]. In the human kidney, we found that the VDR was present abundantly in the distal tubule and to a lesser extent in the proximal tubule. Cells expressing calbindin-D_28K also express the calcium pump and the epithelial calcium channel [199, 222]. Interestingly, not all cells in the distal tubule expressed the receptor. Acid secreting cells do not express the VDR in significant amounts. Taken together, the results are consistent with the notion that the VDR is present in significant amounts in the distal tubule where it regulates the amount and the activity of several vitamin D-dependent proteins such as the plasma membrane calcium pump, epithelial calcium channel, and calbindin D_28K. Although present in lesser amounts, the proximal tubule also expresses the VDR where it regulates the activity of 1α-hydroxylase and 24-hydroxylase. The VDR is detected in the developing rat kidney in vivo as early as day 15 post-coitum (p.c.) and in the cultured metanephros [12]. Significant amounts of the VDR are found in the metanephric mesenchyme and the ureteric bud. As the kidney matures, the VDR is observed in S-shaped and comma-shaped bodies and in the developing glomerulus, specifically in the parietal and visceral endothelial cells. VDR staining in the latter cells persists in the adult kidney as well. Similar patterns of VDR expression were found in the developing mouse kidney in vivo and in mouse metanephric cultures.

The VDR is regulated by several factors in diverse tissues [223]. Concentrations of the VDR in the kidney and parathyroid glands are mainly regulated by 1α,25(OH)₂D₃, PTH and dietary calcium. Exogenous administration of 1α,25(OH)₂D₃ in rats increases VDR levels in duodenal and renal tissues [224, 225]. When endogenous 1α,25(OH)₂D₃ concentrations are increased by adapting an animal to a low calcium diet (Table 5), VDR concentrations in the duodenum and kidney do not increase. The difference appears to be due to increases in the levels of PTH elicited by the low calcium diet and decreased PTH following 1α,25(OH)₂D₃ administration. Differences in Cyp24A1 expression (decreased in the presence of a low calcium diet, and increased following the administration of 1α,25(OH)₂D₃) might also contribute to this difference in VDR expression. PTH has been shown to down-regulate VDR in osteosarcoma cells as well as block up-regulation of VDR in rats infused with both PTH and 1α,25(OH)₂D₃ [226]. These results demonstrate the opposing effects of PTH and 1α,25(OH)₂D₃ on VDR expression in the kidney.

TABLE 5.

Effect of dietary calcium on unoccupied VDR content in rat duodenum and kidney

		Day
	Diet	2	7	14	21
VDR
Duodenum	1% calcium	341±26	197±17	202±17	259±26
	0.02% calcium	365±27	226±28	221±16	267±28
Kidney	1% calcium	ND	163±11	165±9	124±8
	0.02% calcium	ND	120±4*	131±10*	77±3*
Calcium (mg/dL)	1% calcium	9.98±0.18	9.82±0.08	9.16±0.16	9.52±0.41
	0.02% calcium	9.45±0.12	9.15±0.17*	9.09±0.14	8.70±0.29*
Phosphorus (mg/dL)	1% calcium	8.40±0.17	8.88±0.22	8.73±0.17	7.92±0.29
	0.02% calcium	8.15±0.22	8.65±0.13	8.37±0.28	8.41±0.23
1α, 25-(OH)2D3 (pg/mL)	1% calcium	153±11	113±13	139±9	160±32
	0.02% calcium	180±20	392±45**	682±44**	829±59**

Values are mean ± SEM.

Unoccupied vitamin D receptor content expressed as fmols [³H]1α,25(OH)₂D₃ bound per mg cytosol protein.

^*P<0.05;

^**P<0.001.

ND = not done.

Modified from Goff JP, et al[224]

In addition to the VDR, several vitamin D-dependent proteins are expressed in the kidney. The pattern of regulation of these proteins is of great interest, in as much as it casts light on the different mechanisms by which calcium is transported in the kidney. Those proteins involved in calcium transport include: calbindin D_28K and calbindin D_9K, the plasma membrane calcium pump, the epithelial calcium channel, and the calcium sensing receptor. Additionally, although not involved in the transport of calcium, the 24-hydroxylase and the sodium-phosphate type 2 co-transporter (NaPi 2) are also expressed in the kidney. All of these proteins appear to be regulated by 1α,25(OH)₂D₃.

5.2 Calbindins D and the kidney

The two forms of calbindin-D, namely, calbindin-D_28K and calbindin-D_9K are variably distributed in different tissues [227–244]. The proteins are classical E-F-hand proteins. Calbindin-D_28K has 6-EF hands and the calbindin-D_9K has 2 such motifs [245–264]. Calbindin D_28K binds 4 moles of calcium per mole of protein and calbindin D_9K two moles of calcium per mole of protein [227]. In the mouse kidney, both forms are present and are regulated by vitamin D. In other species, only calbindin-D_28K is expressed in the kidney. Both proteins undergo conformational change upon binding to calcium [254, 256, 265]. We recently solved the structure of Ca²⁺-loaded calbindin-D_28K using NMR spectroscopy [266]. The protein is comprised of a single, globular fold consisting of six distinct EF-hand subdomains, which coordinate Ca²⁺ in loops on EF1, EF3, EF4 and EF5. Ran-binding protein M, myo-inositol monophosphatase, and procaspase-3-derived peptides interact with the protein on a surface comprised of alpha5 (EF3), alpha8 (EF4) and the EF2–EF3 and EF4–EF5 loops. The Ca²⁺-dependent conformational change is probably what allows the protein to act as a modulator of effector proteins.

Calbindin D_28K and calbindin-D_9K are 1α,25(OH)₂D₃-regulated proteins [258–264, 267]. Their expression is reduced in the kidneys of 1α–hydroxylase knock-out mice [268]. Calbindin D_28K and calbindin-D_9K expression is normalized by treatment with 1α,25(OH)₂D₃, however, only calbindin D_28K expression is increased when 1α–hydroxylase knock-out mice are fed a high calcium diet. In a mouse VDR-KO model, renal calbindin D_9K expression is nearly abolished whereas calbindin D_28K expression returns to control values as the animals age [269]. Calbindin-D_9K induction by 1α,25(OH)₂D₃ in vitro is absent in VDR null cells [270]. In vitro, PTH has a synergistic effect with 1α,25(OH)₂D₃ on calbindin D_9K expression [270]. In contrast, calbindin D_28K expression is enhanced by PTH infusion without elevations in 1α,25(OH)₂D₃ concentrations [271]. Preincubation with calbindin D_28K increases calcium uptake in luminal membranes while calbindin D_9K preincubation increases calcium uptake in basolateral membranes [272, 273]. Calbindin D_28K knock-out mice fed a high calcium (1%) diet have an elevated urinary calcium/creatinine ratio; curiously, no differences in serum calcium or PTH are noted compared to wild-type littermates [274, 275]. The elevated urinary calcium/creatinine ratio in calbindin D_28K knock-out mice is not apparent after fasting [274]. This is consistent with the finding that fractional excretion of calcium is not altered in calbindin D_28K knock-out mice fed a normal calcium diet [276]. The effects of calbindin D_28K on renal calcium conservation in mice are modest, perhaps because of compensatory increases in calbindin D_9K.

5.3. The plasma membrane calcium pump

In the adult human kidney, epitopes for the calcium-magnesium ATPase (calcium pump) are expressed in the basolateral membrane of distal tubular cells [14, 15, 28, 203, 204]. Similar patterns of expression were apparent in the rat [203] and rabbit kidney [277]. Not all cells of the distal tubule stained positively for the calcium pump; cells expressing acid secreting, carbonic anhydrase positive cells do not express the plasma membrane calcium pump, whereas the other cells of the distal tubule express the PMCa. Calbindin D_28K is co-expressed with the PMCa pump. Our studies have been confirmed by others [277]. The PMCa pump is widely distributed in a large number of other calcium transporting tissues many of which display vitamin D-dependent calcium transport [13, 278–283].

The PMCa is regulated by 1α,25(OH)₂D₃. In MDBK (bovine distal tubule) cells, 1α,25(OH)₂D₃ treatment increases PMCa pump mRNA and protein [202]. Vitamin D deficiency decreases PMCa pump activity in the distal tubule of the kidney [272, 273]. We have shown that the PMCa pump mRNA and protein expression in the intestine increase after the administration of 1α,25(OH)₂D₃ [279, 284, 285]. The increase occurs within 3–6 hours and is dose-dependent. Dietary calcium and phosphorus depletion also increase the amount of PMCa pump expressed in the intestine. Thus, in the intestine and in intestinal cell basolateral membranes, 1α,25(OH)₂D₃ increases the synthesis of the PMCa pump. The mechanism by which up-regulation of PMCa pump activity occurs in the kidney and intestine is uncertain. 1α,25(OH)₂D₃-mediated increases in PMCa pump synthesis or stimulation by calbindin D_9k or calbindin D_28k are possibilities [272, 273]. There is also evidence that stimulation of the calcium sensing receptor (CaSR) decreases calcium absorption by inhibiting PMCa pump activity [286].

5.4. The epithelial calcium channel/vanilloid-receptor related transient receptor potential channels 5 and 6

Hoenderop et al described epithelial calcium channels (ECaCs) which are expressed in the apical membrane of the distal tubule and principal cells of the collecting duct and are distinct from previously described calcium channels [16, 200, 222, 287, 288]. There are at least two members in this family of calcium channels, ECaC-1/TRPV5 and ECaC-2/TRPV6 [288]. ECaC1/TRPV5 expression is limited to the kidney while ECaC2/TRPV6 is expressed in several other tissues [287–290]. The ECaCs/TRPV channels have six putative transmembrane spanning domains, including a pore-forming hydrophobic region between transmembrane domains 5 and 6 [16]. Several putative vitamin D response elements (VDRE) have been identified within the promoter region of the human TRPV5 channel. Hoenderop et al also demonstrated that ECaC1/TRPV5 mRNA and protein levels are increased to near control levels after vitamin D rescue in rats fed a vitamin D deficient diet [199]. In a study using 1α-hydroxylase knock-out mice, a greater than 50% reduction in ECaC-1/TRPV5 expression was found compared to control mice [268]. Additionally, renal ECaC-1/TRPV5 expression in 1α-hydroxylase KO mice is normalized after treatment with 1α,25(OH)₂D₃ [268]. Similar findings were seen when examining calbindin D_28K expression which co-localizes to the same distal tubule cells as ECaC-1/TRPV5 [222, 268]. However, ECaC-1/TRPV5 is not regulated through vitamin D effects on calbindin D_28K. Calbindin D_28K KO mice and cyclosporine A induced down-regulation of calbindin D_28K has no effect on ECaC-1/TRPV5 expression [274]. Others have suggested that calcium also regulates ECaC-1 expression. Quantitative PRC techniques showed reduced expression of ECaC-1 in VDR KO mice compared to control mice. When fed high calcium diets, VDR KO mice had normalization of ECaC-1 concentrations [290].

V. CONCLUSION

The kidney plays a vital role in the conservation of calcium and phosphorus. Besides being the site of synthesis of 1α,25(OH)₂D₃, the kidney responds to the hormone by increasing the efficiency of calcium and phosphorus reabsorption. Elements of the calcium transport systems including calbindin D_28K, calbindin D_9K, the epithelial calcium channel, and the plasma membrane calcium pump all localize to the distal portion of the nephron and are regulated directly or indirectly by 1α,25(OH)₂D₃.

HIGHLIGHTS.

• The kidney is essential for the appropriate regulation of calcium and phosphorus homeostasis.

• It is the site of synthesis of the active, hormonal form of vitamin D, 1α,25(OH)₂D, as well as, the site of degradation of the hormone.

• Many proteins involved in the transport of calcium are regulated by 1α,25(OH)₂D.

Table 6.

Distribution of plasma membrane calcium pump in transporting epithelia as assessed by immunohistochemistry

Tissue	Source	Cell Type	Location in Cell	Reference
Kidney	Rat, human	Distal convoluted tubule, principal cell	Basolateral	[14, 15, 28, 203, 204]
Intestine	Rat, chick	Absorptive cell	Basolateral	[278, 279]
Trophoblast	Rat, human	Syncytiotrophoblast	Basal	[280]
Choroid plexus	Cat Human	Choroid plexus Secretory cell	Apical	[281]
Shell gland	Chick	Principal cell	Apical	[282]
Bone	Human	Osteoblast	Not vectorially oriented	[13]
Bone	Chick	Osteoclast	Not vectorially oriented	[283]

ABBREVIATIONS

1α,25(OH)₂D

1α,25-dihydroxyvitamin D

ECaC

epithelial calcium channel

PMCa

plasma membrane calcium pump

VDR

vitamin D receptor