Although vitamin D was discovered less than 100 years ago, its deficiency diseases (rickets, and its adult-onset counterpart, osteomalacia) were clearly recognized by physicians such as Daniel Whistler (1645) in the Netherlands and Francis Glisson (1650) in England much earlier (1, 2). Rickets is characterized by a “bone matrix which is insufficiently mineralized or calcified,” usually as a result of vitamin D deficiency. Rickets results in a deformed and misshaped skeleton, particularly bending or bowing of the long bones and enlargement of the epiphyses of the joints of the rib cage, arms, legs, and neck. These defects result in painful breathing and difficulties holding up the head, and in women they even give rise to problems with childbirth in later life, if rachitic deformities of the pelvis persist. By the turn of the twentieth century, rickets had reached almost epidemic proportions, particularly in the industrialized cities of Northern Europe, where the excessive degree of air pollution that blocked ultraviolet (UV) exposure and the sweatshop labor practices that kept workers indoors during the daylight hours combined to prevent skin synthesis of vitamin D.

During the nineteenth and early twentieth centuries, the following physicians and scientists investigated the cause of rickets and made key observations that led to the discovery of vitamin D: Sniadecki (1822), who noted that rickets was common in city-dwelling children but not rural dwellers (3); Palm (1890), who noted the importance of latitude in the incidence of rickets in China and concluded that sunshine is the main etiologic factor in rickets (4); Percival (1789), who discussed the medicinal uses of cod liver oil in treating rickets (5); Raczynski (1913), Huldschinsky (1919), and later Hess and Unger (1922), who showed that laboratory animals and children with rickets were cured when exposed to sunlight or mercury (UV) lamps (6, 7, 8); Mellanby (1919), who fed a semisynthetic oatmeal diet to dogs to induce rickets, which was reversible by administering cod liver oil (9); McCollum et al (1922), who demonstrated that the antirachitic factor in cod liver oil was distinct from vitamin A and named it vitamin D (10); Hess and Weinstock (1924) and Steenbock and Black (1924), who were able to reconcile the apparently disparate concepts of sunlight and dietary factors by showing that irradiation of certain food (e.g., plant oils or yeast) increased vitamin D activity (11, 12) (this plant-derived antirachitic factor is now known as vitamin D₂); and Windaus et al (1936), who received the Nobel Prize in 1928 principally for elucidating the structure of sterols, including vitamin D (13).

As already pointed out, vitamin D can be derived from either endogenous synthesis in the skin or from the diet. In the strictest use of the term, vitamin D is therefore not a vitamin, at least during the summer months. Consequently, it should be referred to as a prohormone. Dietary sources become critical during the winter months (above 43 degrees north between October and April) when the zenith angle of the sun is such that UVB light does not penetrate the atmosphere and synthesis of vitamin D₃ in the skin is insignificant. Unfortunately, dietary sources of vitamin D are few, and most foodstuffs are devoid of vitamin D. The only significant sources of vitamin D (D₂ or D₃) are animal liver, fatty fish (e.g., salmon, halibut, cod), egg yolks, and fish oils. Unfortified cow’s milk is not a rich source. Because human milk is an extremely poor source of vitamin D, breast-fed infants require a vitamin D supplement. Most grains, lean meat, vegetables, and fruits are virtually devoid of measurable amounts of vitamin D. Although vitamin D₂ can be derived from the plant sterol ergosterol, the likelihood that this provitamin D₂ will be UV irradiated naturally seems low, even though some cultures sun-dry vegetables. The artificial irradiation of shitake mushrooms, a rich source of ergosterol, has increased the possibility of finding vitamin D₂ in the diet.

Food fortification with either synthetic vitamin D₂ or, later, vitamin D₃ was pioneered and patented in the United States in the 1930s by Steenbock. His concept was to fortify staples such as breakfast cereals, milk, and margarine with vitamin D, initially in the form of irradiated ergosterol, to ensure delivery of this scarce nutrient and also to overcome the seasonal variability in potency found in natural sources (e.g., fish oils). This public health initiative to fortify certain foods with vitamin D eradicated rickets in the United States and virtually everywhere else in the world where it was introduced. In Quebec, Canada, which was the last province or state in North America to introduce vitamin D fortification, statistics showed a dramatic decline in the annual incidence of rickets cases at a downtown Montreal hospital (Ste Justine pour les Enfants) from 130 per 1000 to
virtually 0 per 1000 over an 8-year span between 1968 and 1976; this period coincided with provincial legislation making it mandatory for dairies to fortify milk (15). A few countries resisted vitamin D fortification programs, not because they believed that they would be unsuccessful in the eradication of rickets but because of financial questions about whether government or industry would pay for the programs or out of concern about vitamin D intoxication in the neonatal period—fears that were proved largely unfounded.

The dietary reference intakes (DRIs) were reviewed in a report published by a panel of the Institute of Medicine (IOM) at the National Academy of Sciences in Washington, DC (16). The chosen parameters for reporting optimal vitamin D intakes are currently adequate intake (AI) for the newborn to 1-year life stage and recommended dietary allowance (RDA) for all other life stages. Excessive intake is defined as tolerable upper intake level (UL) for all age groups; and the DRI values for the various life stage groups are provided in Table 18.1.

Vitamin D₃ is synthesized from the sterol 7-dehydrocholesterol by a process involving UVB light in the wavelengths 290 to 315 nm (17) (Fig. 18.2). UV irradiation opens the 9,10-bond of the provitamin to give a previtamin D₃ intermediate in the upper layers of the skin before it is isomerized nonenzymatically by heat to give vitamin D₃ in the lower layers. Transport of vitamin D₃ is carried by a specific plasma protein, vitamin D-binding protein (DBP), from skin to storage tissues or to the liver for the first step of activation. The D vitamins can also be derived from the diet, both as vitamin D₃ and vitamin D₂. Transport to storage depots or liver in the dietary case is on chylomicrons, although some evidence indicates that transfer from chylomicrons to DBP may also occur during transit. Vitamin D from skin or dietary sources does not circulate for long in the bloodstream, but instead is immediately taken up within hours by adipose tissue or liver for storage or activation (18).

Ultimately, vitamin D₃ undergoes its first step of activation (Fig. 18.3), namely, 25-hydroxylation in the liver. Over the years, some controversy has existed over whether 25-hydroxylation of vitamin D₃ is carried out by one enzyme or two and whether this cytochrome P-450-based enzyme is found in mitochondrial or microsomal fractions of liver (18). Biochemical research has established that one human mitochondrial enzyme (CYP27A1) and several microsomal cytochrome P-450 enzymes (including CYP2R1, CYP3A4 and CYP2J3) are able to carry out the 25-hydroxylation of vitamin D₂ or vitamin D₃, or both (see Fig. 18.3) (reviewed in 19). The physiologic relevance of one of these enzymes, CYP2R1, is particularly pertinent because of a single report of a human mutation at Leu99Pro within the CYP2R1 gene in an individual with rickets (20); the enzyme is a 1α-OH-D₂-25-hydroxylase with a high affinity for its vitamin D₂ substrate (21). Investigations have provided a crystal structure of CYP2R1 with several of the known vitamin D substrates bound in the active site (22). Furthermore, a genome-wide association study of the genetic determinants of serum 25-hydroxyvitamin D concentrations (23) concluded that the chromosomal locus for CYP2R1 (11p15) showed the second strongest association of only a handful of sites in the whole genome; the others were DBP (or Gc), CYP24A1, and 7-dehydrocholesterol reductase (DHR7). Notably, variants of the other 25-hydroxylases, such as CYP27A1 or CYP3A4, were not identified as associated with serum 25-OH-D concentrations. This finding suggests that CYP2R1 is the most physiologically-relevant 25-hydroxylase.

The product of the 25-hydroxylation step, 25-OH-D₃, is the major circulating form of vitamin D₃ and in humans is present in plasma at concentrations in the range 10 to 80 ng/mL (25 to 200 nmol/L) (24). The main reason for the extended plasma half-life of 25-OH-D₃ is its strong affinity for DBP, and the DBP-XO mouse shows accelerated rates of clearance and low 25-OH-D₃ levels (25). Serum levels of 25-OH-D₃ therefore represent a measure of the vitamin D status of the animal in vivo.

The circulating metabolite, 25-OH-D₃, is converted to the active form of vitamin D known as calcitriol or 1α,25-(OH)₂D₃. The second step of activation, 1α-hydroxylation,
occurs primarily in the kidney (18). The synthesis of circulating 1α,25-(OH)₂D₃ in the normal, nonpregnant mammal appears to be the exclusive domain of the kidney. A specific mechanism appears to involve the cell surface receptors megalin and cubilin to provide uptake of the substrate, in the form of the 25-OH-D/DBP complex, by renal proximal cells (see Fig. 18.3) (26). Megalin knockout mice show reduced vitamin D metabolite levels and vitamin D deficiency. Additional evidence for renal synthesis of circulating calcitriol stems from clinical medicine. Patients with chronic kidney disease exhibit reduced 1α,25-(OH)₂D₃ levels and frank rickets or osteomalacia resulting from a deficiency of 1α,25-(OH)₂D₃ that is caused by lack of renal-1α-hydroxylase, a situation that can be reversed by 1α,25-(OH)₂D₃ hormone replacement therapy (27). The cytochrome P-450 enzyme, CYP27B1, representing the 1α-hydroxylase enzyme, was cloned virtually simultaneously from several species including rat, human, and mouse
(28, 29, 30, 31). Investigators had known for some time that the kidney mitochondrial 1α-hydroxylase enzyme comprises three proteins—a cytochrome P-450, a ferredoxin, and a ferredoxin reductase for activity—and is strongly downregulated by 1α,25-(OH)₂D₃ and up-regulated by PTH as part of the calcium homeostatic loop (32). Investigators showed that a similar phosphate homeostatic loop also exists involving FGF-23, seemingly the long-postulated “phosphatonin” hormone that down-regulates CYP27B1 enzyme activity presumably at the transcriptional level (33). The promoter for the CYP27B1 gene appears to contain the necessary regulatory elements (cyclic AMP response elements [CREs] and negative VDREs) necessary to explain the observed physiologic regulations of PTH and 1α,25-(OH)₂D₃, respectively, at the transcriptional level. Whether additional elements exist to explain the action of FGF-23 through the klotho receptor remains to be addressed (33). Human CYP27B1 gene mutations result in vitamin D dependency rickets type 1 (VDDR-I) (34), a disease state first proposed, in 1973, to result from a genetic defect of the 1α-hydroxylase enzyme (35). Two independent groups generated the analogous mouse CYP27B1 knockout model by deletion of the cyp27B1 gene, and the 1α,25-(OH)₂D₃-deficient model revealed further insights into the regulation of the gene by different stimuli and subtleties in the roles of 1α,25-(OH)₂D₃ (36, 37).

The final piece of the vitamin D metabolic machinery has been elucidation of the catabolism of 25-OH-D₃ and 1,25-(OH)₂D₃ in the body. This involves another cytochrome P-450 enzyme named CYP24A1, originally known as the 25-OH-D-24-hydroxylase. CYP24A1 inactivates both 25-OH-D₃ and 1α,25-(OH)₂D₃ by subjecting these metabolites to 24-hydroxylation and thus giving rise initially to 24R,25-(OH)₂D₃ and 1α,24,25-(OH)₃D₃, respectively (38, 39) (see Fig. 18.3).

Currently, some speculation exists that 24-hydroxylated metabolites may play a unique biologic role in bone fracture repair, but most of the evidence favors the concept that 24-hydroxylation is primarily an inactivating step. The enzyme 24-hydroxylates both 25-OH-D₃ and 1α,25-(OH)₂D₃, the latter with a 10-fold higher efficiency (40, 41). In the complete absence of DBP, however, this substrate discrimination is less evident. Because the circulating level of 25-OH-D₃ is approximately 1000 times higher than that of 1α,25-(OH)₂D₃, the clearance of products of 25-OH-D₃ by CYP24A1 (e.g., 24R,25-(OH)₂D₃) is readily evident in the bloodstream. The enzyme, particularly the renal form, appears to be expressed at high constitutive levels in the normal animal and may be involved in the inactivation and clearance of excess 25-OH-D₃ from the circulation.

Conversely, the extrarenal 24-hydroxylase appears to be primarily involved in target cell destruction of 1α,25-(OH)₂D₃ (42). Indeed, CYP24A1 has now been shown to be expressed fairly ubiquitously, in particular wherever the VDR is expressed. CYP24A1 enzyme activity has been demonstrated in various cell lines representing specific vitamin D target organs (intestine, CaCo₂ cells; osteosarcoma, UMR-106 cells; kidney, LLC-PK1 cells; keratinocyte, HPK1A and HPK1A-ras). Researchers have shown that 24-hydroxylation is the first step in the C-24 oxidation pathway, a five-step, vitamin D-inducible, ketoconazolesensitive pathway that changes the hydroxylated vitamins D molecules to water-soluble truncated products such as the biliary form, calcitroic acid (43, 44).

Although with human, rat, and mouse CYP24A1, the formation of calcitroic acid is the main pathway, other CYP24A1 isoforms, in particular the guinea pig and opossum analogs, predominantly carry out a 23-hydroxylation pathway to a 26,23-lactone product (45). The biologic value for the synthesis of the 26,23-lactone remains unknown, but the switch to the 23-hydroxylation pathway can be effected by a single Ala326Gly mutation in human CYP24A1 (45). In most biologic assays, the intermediates and truncated products of these 23- and 24-hydroxylation pathways possess lower or negligible biologic activity. Furthermore, many of these compounds have little or no affinity for DBP, thus making their survival in plasma tenuous at best. Polymerase chain reaction (PCR) studies of CYP24A1 led to the detection of CYP24A1 mRNA in a wide range of tissues, thereby corroborating the earlier studies reporting widespread 24-hydroxylase enzyme activity in most, if not all, calcitriol target cells.

Additional studies showed that mRNA transcripts for CYP24A1 are virtually undetectable in naive target cells, not exposed to 1α,25-(OH)₂D₃, but are dramatically induced by a VDR-mediated mechanism within hours of exposure to 1α,25-(OH)₂D₃ (46). In fact, the promoters of both human and rat CYP24A1 genes possess a double VDRE that has been shown to mediate the calcitriol-dependent induction of CYP24 enzyme in both species. It is therefore attractive to propose that not only is 24-hydroxylation an important step in the inactivation of excess 25-OH-D₃ in the circulation but also it is involved in the inactivation of 1α,25-(OH)₂D₃ inside target cells. As such, one can hypothesize that C-24 oxidation is a target cell attenuation or desensitization process that constitutes a molecular switch to turn off calcitriol responses inside target cells (47).

Loss-of-function mutations of human CYP24A1 result in the condition known as idiopathic infantile hypercalcemia (IIH) (48) characterized by hypercalcemia, hypercalciuria, nephrolithiasis and nephrocalcinosis which supports a counterregulatory role for CYP24A1 in vitamin D metabolism. The CYP24A1-knockout mouse exhibits a similar hypercalcemic phenotype and in 50% of animals is severe enough to cause death at around weaning. Conversely, surviving animals have unexplained changes in bone morphology involving excess unmineralized osteoid that could suggest an alternative role for 24-hydroxylase in bone mineralization, although double knockouts lacking both CYP24A1 and VDR do not exhibit this bone phenotype (49). Surviving CYP24A1 null animals have been shown to possess a much reduced ability to clear a bolus dose of
[1β-³H]1α,25-(OH)₂D₃ from their circulation as compared with normal, wild-type littermates (50). Thus, not much evidence exists for efficient non-CYP24A1-mediated backup excretory systems for calcitriol catabolism.

Calcitroic acid, the final water-soluble biliary product of 1α,25-(OH)₂D₃ catabolism, is probably not synthesized in liver because C-24 oxidation does not occur in hepatoma cells and presumably must therefore be transferred from target cells to liver by some plasma carrier. Although calcitroic acid has been found in various tissues in vivo (51), details of its transfer to bile have not been elucidated. Some emerging evidence indicates that high concentrations of vitamin D compounds can be metabolized by the general purpose liver cytochrome P-450 enzyme, CYP3A4, which is induced in intestine by calcitriol (52, 53, 54).