Synthesis and Biological Activity of Vitamin D, 3P-Sulfate ROLE OF VITAMIN D.3 SULFATES IN CALCIUM HOMEOSTASIS*

To determine the biological activity of vitamin D sulfates, we synthesized vitamin D3 3P-sulfate and tested its biological activity in vitamin D-deficient hy-pocalcemic rats. When vitamin D3 sulfate was admin- istered as a single oral dose of 208,000 or 416,000 pmol (100 pg or 200 pg), it increased active calcium transport in the duodenum and was also able to mobilize calcium from bone and soft tissue. Dose levels below this failed to elicit a response. Vitamin D3 itself was active at doses as low as 260 pmol when administered in this manner. In order to test the biological activity of vitamin D3 sulfate in various doses when administered chronically, we tested the biological activity of vitamin D3 sulfate after 5 days of oral dosing: vitamin D3 sulfate was active at doses of 52,000 pmol/day (25 pg), whereas vitamin D3 was active at doses of 65 to 260 pmol/day over a period of 5 days. When administered as a single intravenous dose, vitamin D3 sulfate exhibited no biological activity in doses as high as 52,000 pmol. Vitamin D3, however, was active at a dose of as low as 65 pmol. We conclude that vitamin D3 sulfate, a metabolite of vitamin D3 of heretofore unknown biological activity, is considerably less active than vitamin D3 itself. The role of vitamin DO sulfate’ in calcium homeostasis is unclear.

The role of vitamin DO sulfate' in calcium homeostasis is unclear. Vitamin DZ sulfate has been isolated from rabbit urine, rat liver homogenates, and chicken tissues (Higaki et al., 1965); human and cow's milk (Sahashi et al., 1967) may also contain vitamin D P sulfate. Additionally, "vitamin D sulfate" has been isolated from milk (Lakdawala and Widdowson, 1977), although precise chemical identification and biological potency testing was not performed. Vitamin Dz sulfate is biologically active in rats, and when vitamin DZ sulfate (10 IU or -250 ng) was administered for 28 days to rachitic rats, it was effective in maintaining growth, increasing the mineral ash content of bones, and healing rickets as assessed by x-ray analysis of the bones (Sahashi et al., 1969). The antirachitic activity of human milk due to free vitamin D or nonconjugated vitamin D metabolite activity has been estimated (Lakdawala and Widdowson, 1977) to be between 0.01 and 0.15 pg/dl, a value below the amount required to prevent rickets in a * This research was supported by Grants AM25409 and AM26808 and by a grant from the R. K. Mellon Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
* To whom all inquiries should be addressed.
' The abbreviations used are: vitamin D3 sulfate, 3p-sulfoxy-9,lOsecocholesta-5,6,10(19)-triene sodium salt; vitamin Dz sulfate, 3p-sulfoxy-9,10-secoergosta-5,7,10(19)-triene ammonium salt; HPLC, high performance liquid chromatography; NMR, nuclear magnetic resonance; IR, infrared. normal human infant. Therefore, it has been postulated that because breast-fed infants do not become rachitic, the antirachitic activity of milk is probably due to the biological activity of vitamin D sulfates. It has been hypothesized that the hydrolysis of vitamin D sulfate to free vitamin D could result in the formation of sufficient vitamin D to prevent rickets. Various workers have claimed that human milk contains between 1 to 2 pg/dl of putative vitamin D sulfate (Lakdawala and Widdowson, 1977).
Extensive data concerning the biological activity of vitamin D3 sulfate are not available. Sorgue and Miravet (1978) Boulch et al. (1979, a and b) reported that in the lactating and gestating female rat, vitamin D3 sulfate was hydrolyzed to free "vitamin D3." However, vitamin D3 was not identified by any method other than chromatographic mobility on LH-20 Sephadex. Marnay-Gulat et al. (1975) showed that vitamin D3 sulfoconjugate was inactive in chickens. Miravet et al. (1975) showed that vitamin Ds sulfate had some biological activity in young rats, although only two dose ranges were tested in small numbers of rats. T o clarify these ambiguities, we synthesized vitamin Ds 3P-sulfate in pure form (judged by ultraviolet spectroscopy, mass spectroscopy, elemental analysis, nuclear magnetic resonance spectroscopy, and liquid chromatography) and tested its biological activity in uiuo.

MATERIALS AND METHODS
All ultraviolet absorption spectra were taken in ethanol on a Beckman model 25 (Beckman Instruments, Palo Alto, CA) recording spectrophotometer. All mass spectra were recorded on Hewlett Packard spectrometer model 5985a (Hewlett Packard, Avondale, PA) equipped with a direct insertion probe. All NMR spectra were recorded on an IBM 80 MHz NMR spectrometer. Infrared spectra were recorded on a Perkin-Elmer 327 infrared spectrometer. Serum calcium was determined with an atomic absorption spectrometer Perkin-Elmer model 303 (Perkin-Elmer, Nonvalk, CT) using 0.1% Lac13 as diluent. Serum phosphorus was measured by the colorimetric assay (Chen et al., 1956). Active calcium transport in the duodenum was measured by the everted gut sac technique (Martin and DeLuca, 1969).
Waters LC 204 liquid chromatograph (Waters Associates, Milford, High performance liquid chromatography was performed using a MA) equipped with a variable UV detector and a gradient mixer. All melting points were uncorrected. Elemental analysis was performed by Galbraith Laboratories, Knoxville, TN.
Synthesis of 3~-Sulfoxyy-9,10-Secocholesta-5,7,10[19]-triene Sodium Salt-Pulverized sulfamic acid (20 mg, 0.2 mmol) was suspended in freshly distilled dry pyridine (2 ml) and vitamin DB (38.4 mg, 0.1 mmol) was added to the suspension (Higaki et al., 1965). The mixture was vigorously stirred at 90-95 "C (bath temperature) for I h. Once the reaction was complete (as judged by thin layer chromatography), the pyridine-sulfamic acid mixture was filtered and the solvent was removed on a rotary evaporator. The residue was stirred with ether (3 ml) and was kept at 0 "C overnight. The solid obtained in this manner was chromatographed on a Silica Gel 60 H column, with an eluent of 100% ethyl acetate to 308 methanol/ethyl acetate. Vitamin Da sulfate (yield 85%) ammonium salt obtained as colorless crystals (m.p. 105-107 "C).
The vitamin DB sulfate ammonium salt (100 mg) was dissolved in pyridine (2 ml) and a 15% aqueous solution of sodium hydroxide (3 m l ) was added (Joseph et al., 1966). The reaction mixture was stirred at room temperature for 3 h. The organic phase was separated and extracted with ether (3 X 5 ml). The combined organic layers were dried over anhydrous sodium sulfate. Solvent was evaporated to obtain a foamy solid (yield 80%) which was further purified on Silica Gel 60 H (eluant: ethyl acetate to 30% methanol in ethyl acetate).
The vitamin DJ sulfate sodium salt thus obtained was further purified by HPLC on a reverse phase, Cla pBondapak column using a linear gradient from 35% acetonitrile in water to 100% acetonitrile solvent system. Flow rate was 4 ml/min and the gradient was developed over a period of 15 min. The vitamin D:j sulfate eluted at 4.75 min, and a minor amount (less than 3%) of vitamin D:I was eluted at 16.31 min. The sulfoconjugate fraction was collected and solvent was removed to obtain colorless pure compound (m.p. 126-128 "C). Ultraviolet spectroscopy showed Xmax at 265 nm and X,,, at 228 nm. IR   Animals-Fifty-to sixty-g, weanling, albino, male rats were obtained from the Holtzman Co. (Holtzman Co., Madison, WI). They were maintained in individual overhanging wire cages in UV free light and were fed ad libitum a vitamin Da-deficient, 0.02% calcium, 0.3% phosphorus diet. After 4 weeks on this diet, the animals were used for the experiments described below.
Experimental Design-The doses of vitamin DS sulfate were purified by high performance liquid chromatography immediately before use. No free vitamin D? was present in this preparation. In the first experiment, a group of rats received varying doses of vitamin D:% sulfate dissolved in 33% aqueous ethanol (65 to 52,000 pmol) intravenously. Blood (300 to 500 pl) was obtained prior to the administration of the dose. Another group of animals received the corresponding doses of vitamin D,7 in ethanol. Twenty-four h later the animals were killed; blood was collected for measurement of calcium and phosphorus. The duodenum was removed and used to measure active calcium transport.
In a second experiment, we tested the efficacy of vitamin DS sulfate in mobilizing bone calcium and increasing calcium absorption in the duodenum, when administered orally as a single dose. A group of rats received varying doses of vitamin Ds sulfate (260 to 416,000 pmol) orally in 33% aqueous ethanol. Another group of animals received corresponding doses of vitamin D:I dissolved in ethanol orally. The doses were administered by instillation directly into the stomach.
In a third experiment designed to test the biological activity of the vitamin D:, sulfate when administered chronically, a group of rats received varying doses of vitamin D3 sulfate orally (65 to 52,000 pmol) every day for 5 days. For purposes of comparison, another group of rats received vitamin Dn in the same dose for the same period of time.
There were six to nine animals in a group at each dose tested. Control groups had 12 to 15 rats/group. The values reported are the results mean f S.E. of each such group of animals.
Measurement of Serum 25-Hydroxyvitamin D:3-This was measured using a modified method with high pressure liquid chromatography (Eisman et al., 1977). After collecting the appropriate fractions from the high performance liquid chromatography column, all Samples that contained low or undetectable (by UV) 25-hydroxyvitamin D:j concentrations were assayed by a competitive binding assay.
Statistical Analysis-This was performed using the Student's t test. calcium. Vitamin D:+ sulfate is inactive in this respect at doses as high as 52,000 pmol. In Fig. 2A are shown the results of the active calcium transport in the duodenum following a single oral dose of vitamin D3 or vitamin DS sulfate. Vitamin D;> is active at a dose of 260 pmol, whereas vitamin D3 sulfate exhibits activity at 208,000 pmol and higher. In Fig. 2B are shown the results of the change in serum calcium following a single oral dose of vitamin Da or vitamin D3 sulfate. Vitamin D3 sulfate is active at doses greater than 208,000 pmol. When vitamin D:$ sulfate or vitamin D:, were administered orally, once per day, for a period of 5 days, there was an increase in active calcium transport a t doses of 52,000 pmol or higher in the case of vitamin DS sulfate; vitamin D3 itself was active at a dose of 65 pmol/day (Fig. 3A). Vitamin Ds sulfate mobilized bone/soft tissue calcium at doses of 104,000 pmol when administered chronically; vitamin D, 3 was active at doses as low as 260 pmol (Fig, 3B). The results of increments in weight and serum phosphorus following oral dosing with vitamin DB or vitamin Ds sulfate are shown in Table I. It is clear that serum phosphorus and weight increases at the highest doses of D g sulfate used, whereas vitamin D, is effective at low doses.

Analysis of vitamin
The results of serum 25-hydroxyvitamin Dg levels at the ordinate. There were six to nine rats at each dose. * p < 0.02, ** p <   as an oral dose are shown in Table 11. It is clear that 25hydroxyvitamin D3 is present in the serum of rats administered vitamin D3 at a dose of 1,040 pmol or more. At a dose of 52,000 and 104,000 pmol, vitamin Ds sulfate exhibits no significant increment in serum 25-hydroxyvitamin D3 levels.

DISCUSSION
Conjugates of vitamin D or its analogs are known to occur in rat bile , and more recently, we have demonstrated that one of these can be reutiIized . We report here the biological activity of another conjugate, vitamin Da sulfate, in young rats. When administered intravenously, vitamin D3 sulfate is not active at the doses tested. This could be due to rapid clearance by renal tubular transport processes, Alternatively, the sulfate could be inactive at the doses we administered. When administered as a single oral dose, vitamin Ds sulfate is active at doses in excess of 100,000 pmol. When administered over a period of 5 days, vitamin DX sulfate is active a t doses of 52,000 pmol/day. Therefore, chronic oral dosing is an effective method for eliciting a response to the sulfate. Our data do not allow us to determine whether vitamin D3 sulfate is active as an intact moiety, or only after hydrolysis to vitamin Da. At the highest doses of vitamin D3 sulfate administered, there were small but statistically significant changes in serum 25-hydroxyvitamin D3 levels. Lower doses of vitamin DJ (65 to 520 pmol/rat/day for 5 days) showed biological activity but failed to raise the serum 25-dihydroxyvitamin D3 levels to a significant extent when assessed by our method. Experiments with radiolabeled Ds administered to rachitic animals show the presence of radiolabeled 25-hydroxyvitamin Ds in blood, suggesting that our methods may be too insensitive to detect small changes in serum 25-hydroxyvitamin Dn. Earlier work by Sahashi et al. (1969) demonstrated that vitamin D2 sulfate was a potent antirachitic agent, when administered in a dose of 10 IU (-250 ng) for a period of 28 days. Further, it appeared that vitamin D2 sulfate was nearly as active as vitamin D2 itself. Marnay-Gulat et al. (1975) showed that chronic oral administration of vitamin DS sulfoconjugate to vitamin D-deficient chickens had little effect on growth, serum calcium levels, or healing of rickets. Miravet et al. (1975) showed that vitamin Ds sulfoconjugate was active in young vitamin D-deficient rats administered this analog of vitamin D. In these experiments, the sulfate of vitamin Da was tested at two dose levels only. The duodenal calcium transport ratios (mucosal/serosal) were extremely low even in the vitamin D3-or 1,25-dihydroxyvitamin D3-treated group, raising the possibility of technical problems in that particular experiment. Similarly "bone calcium mobilization" was tested a t one dose level only in rats raised on a normal calcium diet (0.47% calcium). We feel that the presence of normal calcium concentrations in the diet makes it difficult to state, unequivocally, the source of calcium that results in the serum calcium increments. We cannot account for the differences in the activity observed by us relative to other investigators.
In any event, our results using biological tests of vitamin D3 function lead us to believe that young rats can utilize vitamin DB sulfate when it is administered chronically via the oral route at doses of 52,000 pmol or higher. Whether vitamin D3 sulfate acts without being hydrolyzed to vitamin D:% or acts after hydrolysis to free DO is not very clear. We conclude that the vitamin Ds sulfate is biologically active only in high doses when administered orally to young rats. It is possible that vitamin D sulfate could be utilized by infants as a source of vitamin D. Whether or not the amounts present in milk can be utilized efficiently by the human infant remains unclear.