Vitamin D Metabolism and Calcium Absorption*
ANTHONY
W. NORMAN.
Ph.D.
Jli! r,il~lc,. c;~lllllIrnlcl
Vitamin D along with parathyroid hormone (PTH) and calcitonin (CT) are the three principal effecters of calcium and phosphorus homeostasis. The secosteroid, vitamin Ds, is subject to metabolic conversion to its biologically active form(s) 1,25-dihydroxyvitamin D3 [1,25(OHhD~] and 24,25-dihydroxyvitamin D3 [24,25(OHhD3] prior to initiation of its physiologic responses in the intestine and skeletal system. The production of 1,25(OHhD3 is stringently regulated by a variety of endocrine signals including PTH as well as the “calcium needs” of the organism. At the target intestine, 1,25-(OHhD3 stimulates the intestinal absorption of calcium via a mechanism analogous to that of other steroid hormones. Definitive biochemical evidence exists supporting the existence in the intestine of a highly specific protein receptor for 1,25(OH)zDs. After formation of the steroidreceptor complex, it migrates to the nucleus of the cell and stimulates messenger-RNA synthesis for proteins (including a calcium-binding protein) which are necessary for the generation of the biologic response. Current efforts to biochemically characterize vitamin Dmediated intestinal calcium transport include efforts to understand the role of calcium-binding protein in this process, as well as to identify.other protein components present either in the brush border or basal lateral membranes. In recent years an important concept has emerged regarding the role of vitamin D in calcium and phosphorus homeostasis. This is that vitamin D in reality is a steroidt hormone and that its mode of action is analogous to that of other classic steroid hormones [I]. It has further become evident that it is informative to analyze the role of vitamin D in calcium and phosphorus homeostasis from an endocrinologic perspective; as a consequence, conceptual advances have occurred which permit the integrations of the actions of vitamin D on calcium and phosphate metabolism with the more classic endocrinologic actions of parathyroid hormone and calcitonin on these ions Some of these relationships are summarized in Figure I.
* This article is dedicated to the memory of Professor Daniel V. Kimberg who played an indispensable role in identifying many of the intricate details of vitamin D mediated intcstina) calcium transport. I:r~~rn the Department of Biochemistry, lJnivcrsit! of California. Riverside. California. This \vclrk M’~S suppcrrted in part by IJSPHS Gr;lnts ,\hl-ll!113I:! and AM-14750. Requests for reprints ~hc1111tl1~: atklressetl to Dr. Anthnny W. Norman, I J(~J)~rrtmc:nt trf Riochemistry. LJniversity of Calil’omia. Riverside. California 925~ 1.
+ Chemically vitamin Ds is a seco-steroid. formally known as ~IlO)scco-cholesta5.7,10(19)-dien-9 p-01. Seco-steroids are those in which one of the rings has undergone fission by breakage of a carbon-carbon bond: in the instance of vitamin D this is the 9.10 carbon bond of ring B of 7-dehydrocholesterol. Vitamin D:I is the naturally occurring form of the vitamin and is normally derived by exposure to sunlight of the precursor 7-dehydrocholesterol which is present in the skin. The minimum daily requirement of vitamin D is 400 IU: 1.0 IIJ is equivalent tn 25 ng 1or 65 pmol.
December 1979
The American Journal of Medicine
Volume 67
989
VITAMIN
D METABOLISM
AND
CALCIUM
ABSORPTION-NORMAN
Fdthydrorhokrleral skin
1premlin
I
1 I
I
I Short Feedback Loop
Figure 1.
Flow chart of the vitamin D endocrine system.
Vitamin D Metabolism and Its Endocrine System. Figure 1 summarizes the metabolic pathway for processing the prohormone, vitamin D, into its hormonally active form(s). It also summarizes certain endocrinologic interrelationships existent between 1,25-dihydroxyvitamin DB [1,25(OH)zDs], &&dihydroxyvitamin D3 [24,25(OH]~Ds], parathyroid hormone [PTH] and mineral homeostasis. It is now agreed that the biologically active form of vitamin D, particularly in the intestine, is the steroid 1,25(OH)zDs [1,2]. The endocrine gland which produces the biologically active form(s) of vitamin D is the kidney [3,4]. After metabolic conversion of vitamin Ds into 25-hydroxyvitamin D3 [25(OH)2D3] by a liver mitochondrial enzyme [5], this circulating form of the seco-steroid serves as substrate for either the renal 25(0H)D-1-hydroxylase or the 25(OH)D-24hydroxylase. Both enzymes [6,7] are located in the mitochondrial fraction of the kidney cortex. In fact, the l-hydroxylase has been shown to be localized in the kidney of members of every vertebrate class from teleost through amphibia, reptila and aves to many mammals including primates [8]. It has been shown that the l-hydroxylase enzyme system is a classic mixed function steroid hydroxylase similar to the steroid hormone hydrotilases located in the adrenal cortical mitochondria. The l-
990
December 1979
The American Journal of Medicine
hydroxylase is a cytochrome P-45O-containing enzyme [7] which involves an adrenodoxin component and which incorporates molecular oxygen into the la-hydroxyl functionality [g]. Extensive evidence has been presented supporting the view that the steroid hormone 1,25(OH)zD3 is produced only in accord with strict physiologic signals dictated by the calcium “demand” of the organism; a bimodal mode of regulation has been suggested. On a time scale of minutes, changes in the ionic environment of the kidney mitochondria resulting from accumulation or release of calcium or inorganic phosphate may alter the enzymatic activity of the l-hydroxylase [7]. In addition, the peptide hormone, PTH, has been shown on a time scale of hours to be capable of stimulating the production of 1,25(OH)?Ds [7,10] possibly by stimulating the biosynthesis of the l-hydroxylase. Some evidence has also recently been put forward supporting the existence of a diurnal variation in the l-hydroxylase [ll]. Recent studies, utilizing primary cell cultures of chick kidney cells, have conclusively shown that the presence of 1,25(OH)2D3 induces the formation of the 24-hydroxylase enzyme system and the production of 24,25(OH)ZD3 [lo]. Thus, the kidney is clearly an endocrine gland capable of producing, in a physiologically
Volume
67
SWEAT
DIETARY P
FECAL
P
URINE
Figure 2. Schematic model of calcium and phosphorus metabolism in an adult man having a calcium intake of 1.O g/day and a phosphorus intake of 0.8 g/day. All numerical values are shown in milligrams per day. All entries relating to phosphate are calculated as phosphorus. Reproduced from Norman [ 1 l]
regulated
manner, appropriate amounts of 1,25(OH)zI& and 24,25(OH)-D3. There is both physiologic and biochemical evidence of a “short” feedback loop between the kidney and the parathyroid gland wherein the dihydroxylated vitamin D metabolites modulate PTH secretion. It has long been known that release of PTH is governed by a “long” loop feedback of ionized calcium concentration in the blood. Recently, it has been shown that direct infusion into the parathyroid gland of both 1,25(OH)~Da and 24,25(OH)ZDa can inhibit the secretion of PTH [ll]. Also, there is direct biochemical evidence for the existence of both cytosol and nuclear receptors for 1,25(OH)2D3 in the parathyroid gland [a]. Thus, the hormonally active form(s) of vitamin D can apparently modulate the secretion of PTH; this is similar to the action of the classic steroids such as the glucocorticoids or estrogens which “feedback” directly on the hypothalamus and pituitary to inhibit secretion of ACTH and follicle-stimulating hormone.
The mode of action of 1,25(0H)zDa in the target organ, intestinal mucosa, to stimulate calcium absorption has been conclusively shown to be analogous to that of other steroid hormones [1,2,~!]. Definitive biochemical evidence supports the existence of a two-step process; first the steroid hormone 1,25(OH)zDs associates with a 3.7~3 cytosol receptor (60,000 daltons) which then migrates to the intestinal nuclear chromatin fraction where there are specific acceptor sites. As a consequence of the presence of the steroid-receptor complex in the nucleus, there ensues a stimulation of template activity [12] including the biosynthesis of a messenger RNA for a calcium-binding protein [IS]. Dynamics and Requirements of Calcium and Phosphate Metabolism. Figure 2 is a schematic diagram indicating the handling of calcium and phosphorus by normal man. Calcium and phosphorus are both absorbed into the body primarily in the duodenum and j ejunum. In addition to the calcium ingested in the diet (for normal man, the dietary requirements range from
December 1979 The American Journal of Medicine
Volume 67
991
VITAMIN
D METABOLISM
AND
CALCIUM
ABSORPTION-NORMAN
Figure 3. Flow chart analysis of the process of transfer of calcium and phosphorus from the lumen of the intestine across the intestinal cell to the blood. Reproduced from Norman [I].
400 to 1,200mg/day),
approximately 600 to 700 mg are added to the intestinal contents via intestinal secretions. Thus, of the approximate total of 1,600 to 1,700 mg of calcium present in the lumen of the intestine, shown in the model in Figure 2, approximately 700 mg are reabsorbed, or absorbed into the blood stream, leaving the remainder of 900 to 1,000 mg to be excreted in the feces. After calcium has entered the extracellular pool, it is in constant flux or exchange with the calcium already present in the extra- and intracellular fluids and in certain compartments of the bone and the glomerular filtrate. The entire extracellular pool of calcium turns over between 40 and 50 times a day. On a daily basis, the glomerulus will filter some 10,000 mg of calcium; but the renal tubular reabsorption of this ion is so efficient that under normal circumstances only between 100 and 150 mg will appear in the urine. In the event of hypercalcemia, the urinary excretion of calcium will increase in a compensatory fashion; however, it rarely exceeds a value of 400 to 600 mg/day. The renal tubular reabsorption of calcium is stimulated by the PTH, and possibly by vitamin D or one of its metabolites. Also, it should be noted that an additional excretory route of calcium is the loss of a small amount (30 to 100 mg per day) of calcium into the sweat through the skin. The term “readily exchangeable bone pool” is not
992
December 1979 The American Journal of Medicine
definable explicitly in anatomic terms. The readily exchangeable bone pool simply represents that fraction of calcium in bone which is in rapid equilibrium with calcium in the plasma, where rapid is meant to imply a capability of exchanging on a minute-by-minute basis. Of the total bone calcium of 1,000 to 1,200 g in a 70 kg man, approximately 0.4 per cent, or 4,000 mg, is present in the readily exchangeable bone pool compartment. The dynamics of phosphate metabolism are not particularly different from that of calcium. Under normal circumstances, approximately 70 per cent of the phosphate present in the diet is absorbed. Absorption of phosphate is interrelated in a complex fashion with the presence of calcium and can be stimulated by a low calcium diet as well as by the administration of vitamin D or its metabolites. The intestinal absorption of phosphate is inhibited by a high dietary calcium level. Phosphate present in the body is also partitioned between three major pools: the kidney ultrafiltrate, the intracellular compartments present in the various soft tissues and the readily exchangeable fraction of bone. As indicated in Figure 2, the major excretory route for phosphate is by the kidney. The glomerulus can filter some 6,000 to 10,000mg of phosphorus a day. A normal man given a diet of 900 mg of phosphorus will excrete approximately 600 mg/day in the urine.
Volume 67
Intestinal Calcium and Phosphorus Absorption. Any detailed consideration of the process of intestinal calcium and phosphorus absorption must take into account the unique morphology of the intestinal columnar epithelial cells. These cells arc highly differentiated in design for their special role in various absorptive processes. The reader is referred to references [1,14,15] for a detailed discussion of these issues. It is the columnar cpithelial cells which serve as the principal site of vitamin D metabolite action vis-a-vis intestinal calcium and phosphorus absorption. It is possible to identify in a flow chart the various steps that one may anticipate should, or could, be associatcd with the multisteps and multicompartmental problem of intestinal calcium and phosphorus absorption (see Figure 3). The problem reduces to the simple transfer of the ion in qnestion from the lumen of the intestine across the intervening membranes to the plasma. Kinetically, one may follow this as the disappearance of ion from the lumen or appearance of ion in the plasma. IIowever, consideration of the various intervening membranes and compartments that one can postulate should be involved, suggests that a simple straightforward kinetic analysis may not expose the sophistication of the details of the over-all process. There are basically three steps involved with the transport process-(l) uptake across the extracellular membrane of the intestinal cell, (2) transport across the cell and 13) efflux out of the cell across another cell membrane. One must be concerned with the state of the ion as it is transporter!. keeping in mind that calcium may be chelated by an endogcnous substance such as citrate and that phosphate is capable of forming covalent bonds with organic molecules. Also, the charge on the phosphate ion may be of importance since Borle [16] has shown that rcLnal.membranes,are severalfold more permeable to divalent rather than monovalent inorganic phosphate. The first step concerns the possible interaction of these substances with the negatively charged glycocalyx mucopolysaccharide with its exposed sulfate groups. Then at some point, interaction with the microvillar membrane should occur. In this respect a pertinent question is whether passive diffusion. carrier-mediated transport or active transport across this membrane is the means of ion translocation. Related to this is the fact that the electric potential of the inside of the cell is negative with respect to the exterior and thus, for calcium, is an attractive force, whereas for phosphate it would be a repulsive force. Similar considerations in a reverse manner would apply to the exit step at the basal side of the cell. Inside the cell a multitude of interactions may occur. A l)ertincnt question is what role the mitochondrial and microsomal membranes may have in the transcellular migration of phosphate and calcium? The movement of the substances through the membrane at the basal side of the cell and the possible intervention of a sodium/
potassium activated ATPase, have been studied by a number of investigators. Having exited from the cell, the ions in question then must move through interstitial space until they finally encounter the endothelial cells of the blood capillaries. In terms of the potential regulators or mediators of intestinal absorption of these substances, e.g., vitamin D and its metabolitcs, it is apparent that there are numerous possible sites on which they may exert putative specific actions tn facilitate the translocation of calcium and phosphorus into the body. It has also been suggested by Diamond and Wright [ 171that a significant fraction of materials absorbed by the intestine may be transported via diffusional pathways between cells rather than through cells. Administration of vitamin D to rachitic animals mcdiatcs a two and a half to three and a half-fold increase in intestinal calcium transport [ 18,191with the increase occurring only after a time lag of 24 to 48 hours [zo]. The present concept of the mechanism of action of vitamin D involves at least two major time-dependent processes which can partially account for this time lag: (1) the time required for the two-step metabolism ol’vitamin D to 1.25~dihydroxyvitamin D: 12) the time required for interaction of l.~5(OHj2D~~ with its intestinal cytoplasmic and nuclear receptor [?I] resulting in de novo svnthesis or induction of proteins, including calcium-dinding protein [x!], alkaline phosphatase [23] and an ATPase activated by calcium [24.25]. The activity of all these proteins has been shown to increase in intestinal mucosal cells after administration of vitamin D or 1,25-dihydroxycholecalciferol. The important point, however, is that the appearance and loss of these proteins could be dependent upon the ratr! of tllrnover of the mncosal cells themselves. The following questions can be asked concerning the effect of vitamin D on intestinal calcium transport: (1) Does vitamin D affect the physiology of the intestine? (2) Is such a change a prerequisite for the increase of intestinal calcium transport? (3) Is the tllrnover time of the intestinal epithelial cells a determinant factor in the gradual increase or the decrease of calcium transport after a single dose of calciferol? Another pertinent fact is that the increase of intestinal calcium transport and the increase of brush border enzJ,rne activities are known to be maximal by 35 to 40 hollrs after the administration of a single dose of vitamin D or in only 12 to 20 hours when l,2!jOHJ2D:3 is given to chicks. At this time the villus size and rate of migration in the vitamin D or metabolite-treated birds would not be significant11 different from that of the deficient birds. Thus, the result obtained in the chick indicates that the primary action of vitamin D-increased calcium transport-is not integrally dependent upon a change at the cellular lcvcl of the increase in cell size, number or migration rate. Clearly, vitamin D and its metabolitcs have effects on the villus size and migration rate; however. these ap-
December 1979
The American Journal of Medicine
Volume 67
993
VITAlblIN
D METABOLISM
AND
CALCXI Ih,l ADSORI”I’ION--NOR~lAN
parently are not absolute prerequisites for stimulation of intestinal calcium transport. Calcium-Binding Protein. One of the most outstanding and important contributions in the vitamin D field has been the discovery of intestinal calcium binding protein. Wasserman and Taylor [26] first reported the existence of this vitamin D-dependent protein in 1966. Since that time Wasserman, Taylor, Corradino and associates have published an impressive series of papers and reviews describing various aspects of this subject [27-291. It has been conclusively shown that the presence of calcium-binding protein in the intestinal mucosa is totally dependent upon the prior administration of vitamin Ds or one of its biologically active metabolites, e.g., 25(OH)Ds, 1,25(OHJzD3 or 24,25(OH)2D3. No calciumbinding protein is present in the intestine of a vitamin D-deficient chick or rat. The de novo synthesis of calcium-binding protein in response to vitamin D has been established by noting that its appearance was virtually blocked by the prior administration of actinomycin D [30]. Experiments by MacGregor et al. [31] who studied the inhibitory effect of actinomycin D on the incorporation of tritiated amino acids into calcium-binding protein, confirmed this result. A major advance was made in 1975 by the combined efforts of Cohn’s and Wasserman’s laboratories: together they established the primary amino acid sequence of bovine intestinal calcium-binding protein [32]. It was found to consist of a single amino acid chain with its N-terminal lysine blocked by an N-ace@ group. A comparison of the sequence of intestinal calciumbinding protein to the “test” sequence of muscle proteins [33] suggested to Huang et al. [32] that the bovine calcium-binding protein may contain one or two regions of homology (denoted as E-F hands*] which account for its calcium-binding property. A particularly useful system for a study of the molecular biology of calcium-binding protein has been the organ culture of intestinal segments obtained from 20-day chick embryos [34]. The embryonic chick intestine is devoid of calcium-binding protein; it only appears on the second or third day after hatching. It is possible, however, to put embryonic chick in organ culture and add vitamin D3 or a metabolite or related analog and induce the synthesis of calcium-binding protein. This induction of calcium-binding protein in organ culture has been shown to be blocked by the presence of actinomycin D or a-amanitine [35,36]. As little as 0.5nM (5 X lo-‘” M) 1,25(OH)zDS is capable of initiating the synthesis of calcium-binding protein whereas a 10,000 times higher concentration of the parent vitamin D3 was required for an equivalent effect. The relative effec-
* The E-F hand nomenclature refers to a region of a protein chain consisting of an alpha helix followed by a loop followed by another alpha helix. It has been proposed that calcium is bound to the loop region of the E-F structure [33].
994
December 1979 The American Journal of Medicine
of steroids was found to be 1,25(OH)-D3 > l(OH)DZ > 25(OH)Da > 24,25(OH)ZD3. This dramatically emphasizes the biologic potency of 1,25(OH]zD~ and that it is the primary mediator or vitamin D activator in the intestine. An exceedingly important point of concern has been to identify the cellular localization of calcium-binding protein. Holdsworth [25] has advanced the view that since intestinal calcium-binding protein is found in the high-speed supernatant after cell breakage via homogenization, this is the cellular site of action of calcium-binding protein. However, this is hardly conclusive proof, since if calcium-binding protein were only loosely attached, perhaps via ionic protein interactions to a cellular membrane, then it is entirely possible that calcium-binding protein would be released into the cytosol upon homogenization. In contrast, Taylor and Wasserman [37] found, using immunofluorescent antibody techniques, that although calcium-binding protein was highly concentrated inside goblet cells, it was outside the columnar epithelial cells immediately adjacent to the brush border region. As yet, there is no resolution of the question of the subcellular localization of calcium-binding protein. It has not been possible to date to provide a detailed description at the molecular level of the role of calcium-binding protein in intestinal calcium absorption. In this respect there are a number of obvious problems. (1) Calcium-binding protein has an extremely high affinity for calcium, Ka = 2 X 10” M-l. It is not known how calcium can be rapidly and efficiently removed or released frdm calcium-binding protein. Does calciumbinding protein function dynamically or statically as an ion exchanger? (2) Calcium transport and the effect of vitamin D and/or 1,25(OH)zDx therein is not so simple as to be explained merely by invoking the presence of only calcium-binding protein. (3) There may be and probably should be other vitamin D-dependent components; some evidence for their existence is already available. In this regard, Kowarski and Schacter [38] have reported the presence of a particulate, vitamin D-dependent calcium-binding activity in rat intestine which had a molecular weight of 0.5 to 1.0 X 10”. It is likely that important advances will be made in this area in the near future. It is only fair to indicate, though, that not all workers support the view that intestinal calcium absorption obligatorily involves calcium-binding protein. Lawson and co-workers [39-411 have studied in vitamin D-deficient chicks given 1,25(OH)zD3 the chronology of onset of stimulated intestinal calcium absorption and the appearance of de novo calcium-binding protein. They found that although calcium absorption was definitely stimulated in 1 to 2 hours, messenger RNA for calcium-binding protein was not increased until after 4 hours, and the calcium-binding protein could not be detected until after 5 or 6 hours. Clearly, further work tiveness
Volume
67
TABLE I
Tissue Distribution of Chick Intestinal Calcium-binding Protein (CaBP)
Tissue Jejunum Duodenum Ileum Kidney Pancreas Bone (tibia) Adrenal Lungs Esophagus Liver Skeletal muscle Myocardium Thyroid Hypothalamus Parathyroid glands Testes Cerebral cortex Serum
-D (ng CaBPlmg protein) 79f 36j, 31 f 490 f 21 f 4*1 18 f 0.75 f 5.5 If: 1.2 f 2.0 f 1.1 f 17.6 i 275 f 25 f 16f3 248 f 0
18 14 5 13 7 1.3 0.7 1 0.3 0.2 0.15 5 34 4 18
No. of Animals
to (ng CaBPlmg protein)
No of Animals
6 6 6 4 4 6 3” 5 5 5 4 6 3’ 4 4t 311
32,000 f 5,900 25,000 * 4,400 10,000 f 1,400 3,100 f 370 1,500 f 200 110 f 22 46 f 3 21 f8 11 f3 3 f 0.6 4.3 f 1.5 1.7 f 0.5 11.3 f 2 340 f 14 31.3 * 3 15.4 f 4 261 f 22 49 ?C8 ng/ml (1.1 ng/mg protein)
6 6 6 5 6 6 4’ 6 5 5 5 4 3’ 3 39 5 2 6
4 6
-__
-tD vs -D (p value)
ANTHONY
W. NORMAN.
Ph.D.
Jli! r,il~lc,. c;~lllllIrnlcl
Vitamin D along with parathyroid hormone (PTH) and calcitonin (CT) are the three principal effecters of calcium and phosphorus homeostasis. The secosteroid, vitamin Ds, is subject to metabolic conversion to its biologically active form(s) 1,25-dihydroxyvitamin D3 [1,25(OHhD~] and 24,25-dihydroxyvitamin D3 [24,25(OHhD3] prior to initiation of its physiologic responses in the intestine and skeletal system. The production of 1,25(OHhD3 is stringently regulated by a variety of endocrine signals including PTH as well as the “calcium needs” of the organism. At the target intestine, 1,25-(OHhD3 stimulates the intestinal absorption of calcium via a mechanism analogous to that of other steroid hormones. Definitive biochemical evidence exists supporting the existence in the intestine of a highly specific protein receptor for 1,25(OH)zDs. After formation of the steroidreceptor complex, it migrates to the nucleus of the cell and stimulates messenger-RNA synthesis for proteins (including a calcium-binding protein) which are necessary for the generation of the biologic response. Current efforts to biochemically characterize vitamin Dmediated intestinal calcium transport include efforts to understand the role of calcium-binding protein in this process, as well as to identify.other protein components present either in the brush border or basal lateral membranes. In recent years an important concept has emerged regarding the role of vitamin D in calcium and phosphorus homeostasis. This is that vitamin D in reality is a steroidt hormone and that its mode of action is analogous to that of other classic steroid hormones [I]. It has further become evident that it is informative to analyze the role of vitamin D in calcium and phosphorus homeostasis from an endocrinologic perspective; as a consequence, conceptual advances have occurred which permit the integrations of the actions of vitamin D on calcium and phosphate metabolism with the more classic endocrinologic actions of parathyroid hormone and calcitonin on these ions Some of these relationships are summarized in Figure I.
* This article is dedicated to the memory of Professor Daniel V. Kimberg who played an indispensable role in identifying many of the intricate details of vitamin D mediated intcstina) calcium transport. I:r~~rn the Department of Biochemistry, lJnivcrsit! of California. Riverside. California. This \vclrk M’~S suppcrrted in part by IJSPHS Gr;lnts ,\hl-ll!113I:! and AM-14750. Requests for reprints ~hc1111tl1~: atklressetl to Dr. Anthnny W. Norman, I J(~J)~rrtmc:nt trf Riochemistry. LJniversity of Calil’omia. Riverside. California 925~ 1.
+ Chemically vitamin Ds is a seco-steroid. formally known as ~IlO)scco-cholesta5.7,10(19)-dien-9 p-01. Seco-steroids are those in which one of the rings has undergone fission by breakage of a carbon-carbon bond: in the instance of vitamin D this is the 9.10 carbon bond of ring B of 7-dehydrocholesterol. Vitamin D:I is the naturally occurring form of the vitamin and is normally derived by exposure to sunlight of the precursor 7-dehydrocholesterol which is present in the skin. The minimum daily requirement of vitamin D is 400 IU: 1.0 IIJ is equivalent tn 25 ng 1or 65 pmol.
December 1979
The American Journal of Medicine
Volume 67
989
VITAMIN
D METABOLISM
AND
CALCIUM
ABSORPTION-NORMAN
Fdthydrorhokrleral skin
1premlin
I
1 I
I
I Short Feedback Loop
Figure 1.
Flow chart of the vitamin D endocrine system.
Vitamin D Metabolism and Its Endocrine System. Figure 1 summarizes the metabolic pathway for processing the prohormone, vitamin D, into its hormonally active form(s). It also summarizes certain endocrinologic interrelationships existent between 1,25-dihydroxyvitamin DB [1,25(OH)zDs], &&dihydroxyvitamin D3 [24,25(OH]~Ds], parathyroid hormone [PTH] and mineral homeostasis. It is now agreed that the biologically active form of vitamin D, particularly in the intestine, is the steroid 1,25(OH)zDs [1,2]. The endocrine gland which produces the biologically active form(s) of vitamin D is the kidney [3,4]. After metabolic conversion of vitamin Ds into 25-hydroxyvitamin D3 [25(OH)2D3] by a liver mitochondrial enzyme [5], this circulating form of the seco-steroid serves as substrate for either the renal 25(0H)D-1-hydroxylase or the 25(OH)D-24hydroxylase. Both enzymes [6,7] are located in the mitochondrial fraction of the kidney cortex. In fact, the l-hydroxylase has been shown to be localized in the kidney of members of every vertebrate class from teleost through amphibia, reptila and aves to many mammals including primates [8]. It has been shown that the l-hydroxylase enzyme system is a classic mixed function steroid hydroxylase similar to the steroid hormone hydrotilases located in the adrenal cortical mitochondria. The l-
990
December 1979
The American Journal of Medicine
hydroxylase is a cytochrome P-45O-containing enzyme [7] which involves an adrenodoxin component and which incorporates molecular oxygen into the la-hydroxyl functionality [g]. Extensive evidence has been presented supporting the view that the steroid hormone 1,25(OH)zD3 is produced only in accord with strict physiologic signals dictated by the calcium “demand” of the organism; a bimodal mode of regulation has been suggested. On a time scale of minutes, changes in the ionic environment of the kidney mitochondria resulting from accumulation or release of calcium or inorganic phosphate may alter the enzymatic activity of the l-hydroxylase [7]. In addition, the peptide hormone, PTH, has been shown on a time scale of hours to be capable of stimulating the production of 1,25(OH)?Ds [7,10] possibly by stimulating the biosynthesis of the l-hydroxylase. Some evidence has also recently been put forward supporting the existence of a diurnal variation in the l-hydroxylase [ll]. Recent studies, utilizing primary cell cultures of chick kidney cells, have conclusively shown that the presence of 1,25(OH)2D3 induces the formation of the 24-hydroxylase enzyme system and the production of 24,25(OH)ZD3 [lo]. Thus, the kidney is clearly an endocrine gland capable of producing, in a physiologically
Volume
67
SWEAT
DIETARY P
FECAL
P
URINE
Figure 2. Schematic model of calcium and phosphorus metabolism in an adult man having a calcium intake of 1.O g/day and a phosphorus intake of 0.8 g/day. All numerical values are shown in milligrams per day. All entries relating to phosphate are calculated as phosphorus. Reproduced from Norman [ 1 l]
regulated
manner, appropriate amounts of 1,25(OH)zI& and 24,25(OH)-D3. There is both physiologic and biochemical evidence of a “short” feedback loop between the kidney and the parathyroid gland wherein the dihydroxylated vitamin D metabolites modulate PTH secretion. It has long been known that release of PTH is governed by a “long” loop feedback of ionized calcium concentration in the blood. Recently, it has been shown that direct infusion into the parathyroid gland of both 1,25(OH)~Da and 24,25(OH)ZDa can inhibit the secretion of PTH [ll]. Also, there is direct biochemical evidence for the existence of both cytosol and nuclear receptors for 1,25(OH)2D3 in the parathyroid gland [a]. Thus, the hormonally active form(s) of vitamin D can apparently modulate the secretion of PTH; this is similar to the action of the classic steroids such as the glucocorticoids or estrogens which “feedback” directly on the hypothalamus and pituitary to inhibit secretion of ACTH and follicle-stimulating hormone.
The mode of action of 1,25(0H)zDa in the target organ, intestinal mucosa, to stimulate calcium absorption has been conclusively shown to be analogous to that of other steroid hormones [1,2,~!]. Definitive biochemical evidence supports the existence of a two-step process; first the steroid hormone 1,25(OH)zDs associates with a 3.7~3 cytosol receptor (60,000 daltons) which then migrates to the intestinal nuclear chromatin fraction where there are specific acceptor sites. As a consequence of the presence of the steroid-receptor complex in the nucleus, there ensues a stimulation of template activity [12] including the biosynthesis of a messenger RNA for a calcium-binding protein [IS]. Dynamics and Requirements of Calcium and Phosphate Metabolism. Figure 2 is a schematic diagram indicating the handling of calcium and phosphorus by normal man. Calcium and phosphorus are both absorbed into the body primarily in the duodenum and j ejunum. In addition to the calcium ingested in the diet (for normal man, the dietary requirements range from
December 1979 The American Journal of Medicine
Volume 67
991
VITAMIN
D METABOLISM
AND
CALCIUM
ABSORPTION-NORMAN
Figure 3. Flow chart analysis of the process of transfer of calcium and phosphorus from the lumen of the intestine across the intestinal cell to the blood. Reproduced from Norman [I].
400 to 1,200mg/day),
approximately 600 to 700 mg are added to the intestinal contents via intestinal secretions. Thus, of the approximate total of 1,600 to 1,700 mg of calcium present in the lumen of the intestine, shown in the model in Figure 2, approximately 700 mg are reabsorbed, or absorbed into the blood stream, leaving the remainder of 900 to 1,000 mg to be excreted in the feces. After calcium has entered the extracellular pool, it is in constant flux or exchange with the calcium already present in the extra- and intracellular fluids and in certain compartments of the bone and the glomerular filtrate. The entire extracellular pool of calcium turns over between 40 and 50 times a day. On a daily basis, the glomerulus will filter some 10,000 mg of calcium; but the renal tubular reabsorption of this ion is so efficient that under normal circumstances only between 100 and 150 mg will appear in the urine. In the event of hypercalcemia, the urinary excretion of calcium will increase in a compensatory fashion; however, it rarely exceeds a value of 400 to 600 mg/day. The renal tubular reabsorption of calcium is stimulated by the PTH, and possibly by vitamin D or one of its metabolites. Also, it should be noted that an additional excretory route of calcium is the loss of a small amount (30 to 100 mg per day) of calcium into the sweat through the skin. The term “readily exchangeable bone pool” is not
992
December 1979 The American Journal of Medicine
definable explicitly in anatomic terms. The readily exchangeable bone pool simply represents that fraction of calcium in bone which is in rapid equilibrium with calcium in the plasma, where rapid is meant to imply a capability of exchanging on a minute-by-minute basis. Of the total bone calcium of 1,000 to 1,200 g in a 70 kg man, approximately 0.4 per cent, or 4,000 mg, is present in the readily exchangeable bone pool compartment. The dynamics of phosphate metabolism are not particularly different from that of calcium. Under normal circumstances, approximately 70 per cent of the phosphate present in the diet is absorbed. Absorption of phosphate is interrelated in a complex fashion with the presence of calcium and can be stimulated by a low calcium diet as well as by the administration of vitamin D or its metabolites. The intestinal absorption of phosphate is inhibited by a high dietary calcium level. Phosphate present in the body is also partitioned between three major pools: the kidney ultrafiltrate, the intracellular compartments present in the various soft tissues and the readily exchangeable fraction of bone. As indicated in Figure 2, the major excretory route for phosphate is by the kidney. The glomerulus can filter some 6,000 to 10,000mg of phosphorus a day. A normal man given a diet of 900 mg of phosphorus will excrete approximately 600 mg/day in the urine.
Volume 67
Intestinal Calcium and Phosphorus Absorption. Any detailed consideration of the process of intestinal calcium and phosphorus absorption must take into account the unique morphology of the intestinal columnar epithelial cells. These cells arc highly differentiated in design for their special role in various absorptive processes. The reader is referred to references [1,14,15] for a detailed discussion of these issues. It is the columnar cpithelial cells which serve as the principal site of vitamin D metabolite action vis-a-vis intestinal calcium and phosphorus absorption. It is possible to identify in a flow chart the various steps that one may anticipate should, or could, be associatcd with the multisteps and multicompartmental problem of intestinal calcium and phosphorus absorption (see Figure 3). The problem reduces to the simple transfer of the ion in qnestion from the lumen of the intestine across the intervening membranes to the plasma. Kinetically, one may follow this as the disappearance of ion from the lumen or appearance of ion in the plasma. IIowever, consideration of the various intervening membranes and compartments that one can postulate should be involved, suggests that a simple straightforward kinetic analysis may not expose the sophistication of the details of the over-all process. There are basically three steps involved with the transport process-(l) uptake across the extracellular membrane of the intestinal cell, (2) transport across the cell and 13) efflux out of the cell across another cell membrane. One must be concerned with the state of the ion as it is transporter!. keeping in mind that calcium may be chelated by an endogcnous substance such as citrate and that phosphate is capable of forming covalent bonds with organic molecules. Also, the charge on the phosphate ion may be of importance since Borle [16] has shown that rcLnal.membranes,are severalfold more permeable to divalent rather than monovalent inorganic phosphate. The first step concerns the possible interaction of these substances with the negatively charged glycocalyx mucopolysaccharide with its exposed sulfate groups. Then at some point, interaction with the microvillar membrane should occur. In this respect a pertinent question is whether passive diffusion. carrier-mediated transport or active transport across this membrane is the means of ion translocation. Related to this is the fact that the electric potential of the inside of the cell is negative with respect to the exterior and thus, for calcium, is an attractive force, whereas for phosphate it would be a repulsive force. Similar considerations in a reverse manner would apply to the exit step at the basal side of the cell. Inside the cell a multitude of interactions may occur. A l)ertincnt question is what role the mitochondrial and microsomal membranes may have in the transcellular migration of phosphate and calcium? The movement of the substances through the membrane at the basal side of the cell and the possible intervention of a sodium/
potassium activated ATPase, have been studied by a number of investigators. Having exited from the cell, the ions in question then must move through interstitial space until they finally encounter the endothelial cells of the blood capillaries. In terms of the potential regulators or mediators of intestinal absorption of these substances, e.g., vitamin D and its metabolitcs, it is apparent that there are numerous possible sites on which they may exert putative specific actions tn facilitate the translocation of calcium and phosphorus into the body. It has also been suggested by Diamond and Wright [ 171that a significant fraction of materials absorbed by the intestine may be transported via diffusional pathways between cells rather than through cells. Administration of vitamin D to rachitic animals mcdiatcs a two and a half to three and a half-fold increase in intestinal calcium transport [ 18,191with the increase occurring only after a time lag of 24 to 48 hours [zo]. The present concept of the mechanism of action of vitamin D involves at least two major time-dependent processes which can partially account for this time lag: (1) the time required for the two-step metabolism ol’vitamin D to 1.25~dihydroxyvitamin D: 12) the time required for interaction of l.~5(OHj2D~~ with its intestinal cytoplasmic and nuclear receptor [?I] resulting in de novo svnthesis or induction of proteins, including calcium-dinding protein [x!], alkaline phosphatase [23] and an ATPase activated by calcium [24.25]. The activity of all these proteins has been shown to increase in intestinal mucosal cells after administration of vitamin D or 1,25-dihydroxycholecalciferol. The important point, however, is that the appearance and loss of these proteins could be dependent upon the ratr! of tllrnover of the mncosal cells themselves. The following questions can be asked concerning the effect of vitamin D on intestinal calcium transport: (1) Does vitamin D affect the physiology of the intestine? (2) Is such a change a prerequisite for the increase of intestinal calcium transport? (3) Is the tllrnover time of the intestinal epithelial cells a determinant factor in the gradual increase or the decrease of calcium transport after a single dose of calciferol? Another pertinent fact is that the increase of intestinal calcium transport and the increase of brush border enzJ,rne activities are known to be maximal by 35 to 40 hollrs after the administration of a single dose of vitamin D or in only 12 to 20 hours when l,2!jOHJ2D:3 is given to chicks. At this time the villus size and rate of migration in the vitamin D or metabolite-treated birds would not be significant11 different from that of the deficient birds. Thus, the result obtained in the chick indicates that the primary action of vitamin D-increased calcium transport-is not integrally dependent upon a change at the cellular lcvcl of the increase in cell size, number or migration rate. Clearly, vitamin D and its metabolitcs have effects on the villus size and migration rate; however. these ap-
December 1979
The American Journal of Medicine
Volume 67
993
VITAlblIN
D METABOLISM
AND
CALCXI Ih,l ADSORI”I’ION--NOR~lAN
parently are not absolute prerequisites for stimulation of intestinal calcium transport. Calcium-Binding Protein. One of the most outstanding and important contributions in the vitamin D field has been the discovery of intestinal calcium binding protein. Wasserman and Taylor [26] first reported the existence of this vitamin D-dependent protein in 1966. Since that time Wasserman, Taylor, Corradino and associates have published an impressive series of papers and reviews describing various aspects of this subject [27-291. It has been conclusively shown that the presence of calcium-binding protein in the intestinal mucosa is totally dependent upon the prior administration of vitamin Ds or one of its biologically active metabolites, e.g., 25(OH)Ds, 1,25(OHJzD3 or 24,25(OH)2D3. No calciumbinding protein is present in the intestine of a vitamin D-deficient chick or rat. The de novo synthesis of calcium-binding protein in response to vitamin D has been established by noting that its appearance was virtually blocked by the prior administration of actinomycin D [30]. Experiments by MacGregor et al. [31] who studied the inhibitory effect of actinomycin D on the incorporation of tritiated amino acids into calcium-binding protein, confirmed this result. A major advance was made in 1975 by the combined efforts of Cohn’s and Wasserman’s laboratories: together they established the primary amino acid sequence of bovine intestinal calcium-binding protein [32]. It was found to consist of a single amino acid chain with its N-terminal lysine blocked by an N-ace@ group. A comparison of the sequence of intestinal calciumbinding protein to the “test” sequence of muscle proteins [33] suggested to Huang et al. [32] that the bovine calcium-binding protein may contain one or two regions of homology (denoted as E-F hands*] which account for its calcium-binding property. A particularly useful system for a study of the molecular biology of calcium-binding protein has been the organ culture of intestinal segments obtained from 20-day chick embryos [34]. The embryonic chick intestine is devoid of calcium-binding protein; it only appears on the second or third day after hatching. It is possible, however, to put embryonic chick in organ culture and add vitamin D3 or a metabolite or related analog and induce the synthesis of calcium-binding protein. This induction of calcium-binding protein in organ culture has been shown to be blocked by the presence of actinomycin D or a-amanitine [35,36]. As little as 0.5nM (5 X lo-‘” M) 1,25(OH)zDS is capable of initiating the synthesis of calcium-binding protein whereas a 10,000 times higher concentration of the parent vitamin D3 was required for an equivalent effect. The relative effec-
* The E-F hand nomenclature refers to a region of a protein chain consisting of an alpha helix followed by a loop followed by another alpha helix. It has been proposed that calcium is bound to the loop region of the E-F structure [33].
994
December 1979 The American Journal of Medicine
of steroids was found to be 1,25(OH)-D3 > l(OH)DZ > 25(OH)Da > 24,25(OH)ZD3. This dramatically emphasizes the biologic potency of 1,25(OH]zD~ and that it is the primary mediator or vitamin D activator in the intestine. An exceedingly important point of concern has been to identify the cellular localization of calcium-binding protein. Holdsworth [25] has advanced the view that since intestinal calcium-binding protein is found in the high-speed supernatant after cell breakage via homogenization, this is the cellular site of action of calcium-binding protein. However, this is hardly conclusive proof, since if calcium-binding protein were only loosely attached, perhaps via ionic protein interactions to a cellular membrane, then it is entirely possible that calcium-binding protein would be released into the cytosol upon homogenization. In contrast, Taylor and Wasserman [37] found, using immunofluorescent antibody techniques, that although calcium-binding protein was highly concentrated inside goblet cells, it was outside the columnar epithelial cells immediately adjacent to the brush border region. As yet, there is no resolution of the question of the subcellular localization of calcium-binding protein. It has not been possible to date to provide a detailed description at the molecular level of the role of calcium-binding protein in intestinal calcium absorption. In this respect there are a number of obvious problems. (1) Calcium-binding protein has an extremely high affinity for calcium, Ka = 2 X 10” M-l. It is not known how calcium can be rapidly and efficiently removed or released frdm calcium-binding protein. Does calciumbinding protein function dynamically or statically as an ion exchanger? (2) Calcium transport and the effect of vitamin D and/or 1,25(OH)zDx therein is not so simple as to be explained merely by invoking the presence of only calcium-binding protein. (3) There may be and probably should be other vitamin D-dependent components; some evidence for their existence is already available. In this regard, Kowarski and Schacter [38] have reported the presence of a particulate, vitamin D-dependent calcium-binding activity in rat intestine which had a molecular weight of 0.5 to 1.0 X 10”. It is likely that important advances will be made in this area in the near future. It is only fair to indicate, though, that not all workers support the view that intestinal calcium absorption obligatorily involves calcium-binding protein. Lawson and co-workers [39-411 have studied in vitamin D-deficient chicks given 1,25(OH)zD3 the chronology of onset of stimulated intestinal calcium absorption and the appearance of de novo calcium-binding protein. They found that although calcium absorption was definitely stimulated in 1 to 2 hours, messenger RNA for calcium-binding protein was not increased until after 4 hours, and the calcium-binding protein could not be detected until after 5 or 6 hours. Clearly, further work tiveness
Volume
67
TABLE I
Tissue Distribution of Chick Intestinal Calcium-binding Protein (CaBP)
Tissue Jejunum Duodenum Ileum Kidney Pancreas Bone (tibia) Adrenal Lungs Esophagus Liver Skeletal muscle Myocardium Thyroid Hypothalamus Parathyroid glands Testes Cerebral cortex Serum
-D (ng CaBPlmg protein) 79f 36j, 31 f 490 f 21 f 4*1 18 f 0.75 f 5.5 If: 1.2 f 2.0 f 1.1 f 17.6 i 275 f 25 f 16f3 248 f 0
18 14 5 13 7 1.3 0.7 1 0.3 0.2 0.15 5 34 4 18
No. of Animals
to (ng CaBPlmg protein)
No of Animals
6 6 6 4 4 6 3” 5 5 5 4 6 3’ 4 4t 311
32,000 f 5,900 25,000 * 4,400 10,000 f 1,400 3,100 f 370 1,500 f 200 110 f 22 46 f 3 21 f8 11 f3 3 f 0.6 4.3 f 1.5 1.7 f 0.5 11.3 f 2 340 f 14 31.3 * 3 15.4 f 4 261 f 22 49 ?C8 ng/ml (1.1 ng/mg protein)
6 6 6 5 6 6 4’ 6 5 5 5 4 3’ 3 39 5 2 6
4 6
-__
-tD vs -D (p value)
