Melatonin synthesis under calcium constraint in gilthead sea bream (Sparus auratus L.)

Brain or blood plasma melatonin was analysed as a measure for pineal melatonin production in sea bream. Access to calcium was limited by diluting the seawater to 2.5 per thousand and removing calcium from the diet or by prolonged feeding of vitamin D-deficient diet. Interactions/relations between melatonin and calcium balance and the hypercalcemic endocrines PTHrP and calcitriol were assessed. Restricting calcium availability in both water and diet had no effect on plasma melatonin, but when calcium was low in the water or absent from food, increased and decreased plasma melatonin was observed, respectively. Fish on a vitamin D-deficient diet (D- fish) showed decreased plasma calcitriol levels and remained normocalcemic. Decreased brain melatonin was found at all sampling times (10-22 weeks) in the D- fish compared to the controls. A positive correlation between plasma Ca2+ and plasma melatonin was found (R(2)=0.19; N=41; P <0.01) and brain melatonin was negatively correlated with plasma PTHrP (R(2)=0.78; N=4; P <0.05). The positive correlation between plasma levels of melatonin and Ca2+ provides evidence that melatonin synthesis is influenced by plasma Ca2+. The decreased melatonin production in the D- fish points to direct or indirect involvement of calcitriol in melatonin synthesis by the pineal organ in teleosts. The hypercalcemic factors PTHrP and calcitriol appeared to be negatively correlated with melatonin and this substantiates an involvement of melatonin in modulating the endocrine response to cope with hypocalcemia. It further points to the importance of Ca2+ in melatonin physiology.     

The vitamin D receptor and its ligand 1alpha,25-dihydroxyvitamin D3 in Atlantic salmon (Salmo salar)

Seaward migration of Salmo salar is preceded by preparatory physiological adaptations (parr-smolt transformation) to allow for a switch from freshwater (FW) to seawater (SW), which also means a switch in ambient calcium from hypocalcic (<1 mM Ca(2+)) to the plasma (~1.25 mM Ca(2+)) and to strongly hypercalcic (8-12 mM Ca(2+)). Uptake, storage (skeleton, scales) and excretion of calcium need careful regulation. In fish, the vitamin D endocrine system plays a rather enigmatic role in calcium physiology. Here, we give direct evidence for calcitriol involvement in SW migration. We report the full sequence of the nuclear vitamin D receptor (sVDR0) and two alternatively spliced variants resulting from intron retention (sVDR1 and sVDR2). In FW parr, SW adapting smolts, and in SW adults, plasma concentrations of 25(OH)D(3) and 24,25(OH)(2)D(3) did not change significantly. Plasma calcitriol concentrations were lowest in FW parr, doubled during smoltification and remained elevated in SW adults. Increased calcitriol coincided with a twofold decrease in sVDR mRNA levels in gill, intestine, and kidney of FW smolts and SW adults, when compared with parr. Clearly, there was a negative feedback and dynamic response of the vitamin D endocrine system during parr-smolt transformation. The onset of these dynamic changes in FW parr warrants a further search for the endocrines that initiate these changes. We speculate that the vitamin D system plays a crucial role in calcium and phosphorus handling in Atlantic salmon.     

Differential expression of PTHrP and its receptor in pituitary gland and gills in estradiol-treated gilthead sea bream (Sparus auratus, L.)

In the gilthead sea bream (Sparus auratus) 17β-estradiol (E₂) plays an important role in the synthesis of vitellogenin. During vitellogenesis, vitellogenin as a nutritional precursor protein is loaded with calcium, which requires elevated plasma calcium levels. This is accomplished via E₂-dependent processes. Reports have shown that hypercalcemic effects of E₂ are possibly mediated by another hypercalcemic factor, viz. parathyroid hormone related protein (PTHrP). To further investigate the possibility of PTHrP as a mediator of E₂-induced hypercalcemia, we investigated the local expression levels of the pthrp mRNA and of the gene coding for the PTHrP receptor, PTH1R (pth1r) in two tissues involved in the calcium regulation (gills, pituitary gland) of the sea bream treated with E₂. Compared to control, treatment with E₂ resulted in: significantly increased total calcium and plasma PTHrP levels (P < 0.01), a down-regulation of pthrp mRNA in the pituitary gland (P < 0.01), and up-regulation of expression levels for both pthrp and pth1r in the branchial system (P < 0.05). These findings provide direct evidence for a mediating role of PTHrP in E₂ induced hypercalcemia, and in addition support the idea for the presence of two independent systems, an endocrine pituitary PTHrP system and a peripheral paracrine branchial PTHrP system.     

CYP27A1 expression in gilthead sea bream (Sparus auratus, L.): effects of calcitriol and parathyroid hormone-related protein

Little is known about vitamin D metabolism in fishes. Several reports have shown hydroxylase activities in various organs to produce vitamin D metabolites, but the enzymes involved have not been isolated or characterized. We isolated and characterized a renal mitochondrial hydroxylase, CYP27A1, that governs vitamin D metabolism in gilthead sea bream, Sparus auratus. The enzyme is highly expressed in kidney and to a far lesser extent in liver. When treated with 25-hydroxy vitamin D or calcitriol, the kidney responded differentially and time dependently with CYP27A1 mRNA expression levels. This response substantiates a role for CYP27A1 in fish vitamin D metabolism. This notion is strengthened by upregulation of CYP27A1 in sea bream treated with parathyroid hormone-related protein (PTHrP), and suggests an original role for PTHrP in calcitriol-regulated processes n fish similar to the role of PTH in mammalian vitamin D-dependent processes.     

Calcitropic hormones, platelet calcium, and blood pressure in essential hypertension

Plasma ionized calcium, platelet cytosolic calcium (using the fura-2 method in gel-filtered platelets), parathyroid hormone (both the intact hormone and a midmolecule portion), calcitriol, and calcidiol were measured in 19 untreated male patients with essential hypertension and 19 age-matched normotensive male research subjects. Mean levels of platelet cytosolic calcium, parathyroid hormone, calcitriol, and calcidiol were all significantly higher, whereas plasma ionized calcium was significantly lower, in the hypertensive group compared with the normotensive group. Both platelet cytosolic calcium and intact parathyroid hormone were positively correlated with mean arterial pressure (r = 0.58, p less than 0.001; r = 0.54, p less than 0.001, respectively), whereas plasma ionized calcium was inversely correlated with mean arterial pressure (r = -0.60, p less than 0.001) in the combined group of all study subjects. All three of these correlations were significant in the hypertensive group alone but not in the normotensive group alone. When analyzed with plasma ionized calcium, body mass index, serum calcitriol, and calcidiol in a multivariable regression model, the significance of the partial regressions of platelet cytosolic calcium and parathyroid hormone with mean arterial pressure persisted. Intact parathyroid hormone was positively correlated to platelet cytosolic calcium (r = 0.43, p less than 0.01) and plasma ionized calcium was inversely correlated to platelet cytosolic calcium (r = -0.44, p less than 0.01). These results confirm previous reports of disturbances of calcium metabolism in essential hypertension and suggest that the elevated platelet cytosolic calcium observed in essential hypertension may be linked to one or more of these alterations of calcium metabolism.     

Depression of calcium pump activity in renal cortex of vitamin D-deficient rats with secondary hyperparathyroidism

To examine the hormonal regulation of the ATP-dependent Ca2+ pump in the kidneys, the ATP-dependent Ca2+ uptake by the basolateral membrane vesicles in the renal cortex was measured using radioactive calcium (45Ca2+) in rats with vitamin D deficiency or rats undergoing thyroparathyroidectomy. The Vmax of the Ca2+ pump activity was increased not only by administering calcitriol, but also by normalizing the serum calcium level in vitamin D-deficient rats. PTH suppressed the Ca2+ pump activity in normocalcemic vitamin D-deficient rats. Thyroparathyroidectomy did not affect the Ca2+ pump activity in the kidneys of normal rats. It was concluded that the ATP-dependent Ca2+ pump activity was depressed by secondary hyperparathyroidism in vitamin D-deficient rats.     

Modulation of chick intestinal and renal calbindin gene expression by dietary vitamin D3, 1,25-dihydroxyvitamin D3, calcium and phosphorus

Synthetic oligonucleotide probes complementary to chick calbindin-28 kDa-mRNA were used to study the latter's regulation and relationship to calbindin in the chick. The effects of vitamin D3 sources and dietary alteration on the genomic expression were characterized by Northern blot and solution hybridization. Intestinal calbindin and its mRNA were almost absent in vitamin D-deficient chicks and were not affected by dietary alteration. Renal calbindin and its mRNA were lower in the vitamin D-deficient than in vitamin D3- or 1,25-dihydroxyvitamin D3 (1,25(OH)2D3)-fed chicks. In the same animal, renal calbindin mRNA and calbindin were higher than intestinal. In vitamin D3-fed chicks, dietary calcium (Ca) or phosphorus (P) restriction induced, and high dietary Ca inhibited, intestinal calbindin and its mRNA synthesis. In the same chicks, dietary P restriction induced renal calbindin mRNA and calbindin synthesis. In 1,25-(OH)2D3-fed chicks, dietary P restriction induced and high dietary Ca inhibited the synthesis of intestinal and renal calbindin. The results suggest that: (a) most of the changes in renal and intestinal calbindin could be attributed to the changes in the mRNA; (b) the adaptation to dietary Ca and P alterations requires vitamin D metabolites; (c) high dietary Ca affects intestinal and renal calbindin-mRNA and calbindin via mechanisms independent of kidney 1-hydroxylase; and (d) plasma Ca and renal calbindin or its mRNA tend to change together in vitamin D-deficient or vitamin D3-fed, but not in 1,25(OH)2D3-fed chicks.     

Vitamin D and pancreatic islet function. I. Time course for changes in insulin secretion and content during vitamin D deprivation and repletion

After weaning, rats were given free access to a control or vitamin D-deprived diet for 2 to 5 weeks. In the vitamin D deficient rats, plasma concentrations of 25-(OH)D3, 1,25-(OH)2D3,24,25-(OH)2D3 calcium, glucose and insulin were all decreased. After an overnight fast, the plasma insulin concentration was also decreased even when the plasma glucose concentration was not significantly affected. The food intake and body growth was also impaired in vitamin D-deficient rats. Administration of vitamin D3 in oil for 3 to 6 days to vitamin D-deficient rats increased the plasma concentration of vitamin D metabolites, calcium and insulin, but not that of glucose, and stimulated food intake and body growth to a larger extent than in rats treated by oil alone. Vitamin D deprivation decreased and vitamin D treatment increased the insulin content of the whole pancreas or isolated islets and the secretory response of the islets to D-glucose. The changes in insulin release remained significant when the hormonal output was expressed relative to the insulin content of the islets. These findings confirm that vitamin D deficiency causes alterations of pancreatic B-cell function. Moreover, the time course for changes in biological variables during vitamin D deprivation and treatment suggests that such an alteration cannot be solely accounted for by concomitant abnormalities in either plasma calcium or glucose concentrations.     

Vitamin D deficiency in rats with normal serum calcium concentrations.

Rats were raised after weaning on a vitamin D-deficient diet which used whole wheat and casein as the major protein source. For at least the first year of life, plasma calcium concentrations of these rats were the same as those of vitamin D-replete rats, and the rate of growth was normal for at least 6 months. The following evidence establishes the vitamin D deficiency of the rats (both male and female) on this diet: (i) plasma levels of 1,25-dihydroxycholecalciferol (1,25-dihydroxyvitamin D3) became undetectable after 6 weeks on the diet; (ii) by 4 months of age, the epiphyseal growth plates of the tibia were significantly enlarged and disorganized; (iii) when subjected to fracture in a dynamic torsion machine, the femur showed marked weakening as indicated by stress analysis; (iv) isolated kidney cells from the deficient rats showed a 3-fold increase in 25-hydroxyvitamin D 1-hydroxylase activity. When mother rats were placed on the vitamin D-deficient diet during lactation, plasma calcium values in the pups decreased and remained low throughout life and there was a stunted body growth pattern. It is concluded that hypocalcemia is not a necessary manifestation of vitamin D deficiency, that the onset of vitamin D deficiency during neonatal life influences the calcium homeostatic system, and that the normocalcemic, vitamin D-deficient animal provides an experimental model in which the effects of vitamin D deficiency can be studied independently of hypocalcemia.     

THE EFFECT OF TRANSDERMAL OESTROGEN ON BONE, CALCIUM-REGULATING HORMONES AND LIVER IN POSTMENOPAUSAL WOMEN

Transdermal oestradiol, 100 micrograms/d, was used to treat 11 women suffering from postmenopausal symptoms. After 3 weeks therapy there was a significant rise in the plasma oestradiol into the premenopausal range and a significant fall in plasma FSH level and symptom score. Bone resorption, assessed by urinary excretion of calcium and hydroxyproline, decreased significantly while plasma alkaline phosphatase activity remained constant. There was a significant fall in plasma calcium and phosphate but the plasma concentrations of PTH, calcitonin and calcitriol and the urinary excretion of cAMP were unchanged. Plasma levels of vitamin D binding protein, albumin and globulin were unaltered, and blood pressure did not rise. These effects were similar to those found in postmenopausal women with oral ethinyloestradiol, 30 micrograms/d, (Selby et al., 1985), apart from those on plasma vitamin D binding protein, total calcitriol, albumin, globulin, tubular reabsorption of phosphate and blood pressure, changes which probably arise from a direct action of oral oestrogen on the liver.     

Bone metabolism during pregnancy

During pregnancy, mineral concentrations, of calcium and phosphorus in particular, are maintained at a high level in fetal blood so that the developing skeleton may accrete adequate mineral content. The placenta actively transports minerals for this purpose. Maternal intestinal absorption increases in order to meet the fetal demand for calcium, which is only partly dependent on calcitriol. Mineral regulation is essentially dependent on parathyroid hormone (PTH) and PTH-related protein (PTHrP). The calcium-sensing receptor (CaSR) regulates PTH and PTHrP production. If calcium intake is insufficient, the maternal skeleton will undergo resorption due to PTHrP. After birth, a switch from fetal to neonatal homeostasis occurs through increase in PTH and calcitriol, and developmental adaptation of the kidneys and intestines with bone turnover contributing additional mineral to the circulation. Calcium absorption becomes progressively active and dependent on calcitriol. The postnatal skeleton can transiently present with osteoposis but adequate mineral diet usually allows full restoration. Cases of primary osteoporosis must be identified. Loss of trabecular mineral content occurs during lactation in order to provide calcium to the newborn. This programmed bone loss is dependent on a “brain-breast-bone” circuit. The physiological bone resorption during reproduction does not normally cause fractures or persistent osteoporosis. Women who experience fracture are likely to have other causes of bone loss.     

Effect of 1,25-dihydroxyvitamin D3 on insulin secretion: direct or mediated?

In order to clarify the role of vitamin D (D) in regulating insulin secretion, we studied the effect of long term (10 days) and short term (3 days) supplementation with D and/or calcium (Ca) on insulin secretion from the isolated, perfused pancreas of D- and Ca-deficient rats. The influence of the nutritional state induced by D deficiency was also evaluated. The long term supplementation of either D, Ca, or both restored the body weight and improved insulin secretion induced by high glucose concentration to the same extent; thus, no significant difference in insulin secretion was found between the D-only-supplemented group and the Ca-only-supplemented group. When the insulin secretion was compared in D-deplete vs. D-replete rats given the same amount of Ca, insulin secretion was significantly higher in D-replete animals, although plasma Ca levels were also higher. In short term experiments, insulin release was significantly augmented to a similar extent in D- or Ca-replete rats as compared with D- and Ca-deficient rats, despite no significant change in body weight. In a separate experiment, the pancreas from D-deficient rats was perfused with or without 1,25-dihydroxyvitamin D3 [1,25-OH)2D3] to observe its acute effect on insulin release. The perfusion with 1,25-(OH)2D3 did not affect insulin release. This result suggests that impaired insulin secretion in D-deficient rats is caused by a decrease in Ca in the body fluid and possibly by the lack of D effect on the pancreas.     

Effects of parathyroid hormone on the accumulation of cyclic AMP in bone of vitamin D-deficient rats.

The mechanism of skeletal refractoriness to parathyroid hormone (PTH) in vitamine D-deficient animals was studied in terms of the adenylate cyclase-cyclic AMP system in rat calvaria. In vitamin D-deficient, thyroparathyroidectomized rats, plasma calcium concentration was not elevated by iv administration of PTH, while responsiveness to the hormone was recovered within 24 h after a single dose (2.5 mug) of vitamin D3. In spite of the remarkable dependency of PTH on vitamin D for mobilization of calcium from bone, PTH stimulated adenylate cyclase activity in particulate bone cell fractions in vitro. PTH also enhanced the levels of cyclic AMP in the skeletal tissues of vitamin D-deficient rats in vivo and in vitro to an extent similar to those found in rats given 2.5 mug of D3. Administration of theophylline or dibutyryl cyclic AMP to the vitamin D-deficient rats did not cause any significant hypercalcemic effects, while these drugs enhanced plasma calcium concentration significantly in the rats given vitamin D3. These data strongly indicate that the cause of the skeletal refractoriness to PTH in vitamin D-deficient animals is not a defective activation of adenylate cyclase, but must be related to a later step or steps in the biochemical events leading to bone cell activation.     

Vitamin D3 improves impaired glucose tolerance and insulin secretion in the vitamin D-deficient rat in vivo

It has previously been shown in this laboratory that vitamin D3 is essential for normal insulin secretion from the perfused rat pancreas. In this present study, the influence of vitamin D status on insulin secretion in vivo was investigated. Intravenous glucose tolerance tests were performed on conscious vitamin D-deficient rats (-D), vitamin D-replete rats fed ad libitum (+D AL), and vitamin D-replete rats pair fed to the D-deficient animals (+D PF). Vitamin D deficiency, easily recognizable by low daily dietary intake and depressed plasma calcium levels, was found to impair plasma glucose clearance as characterized by an elevated KG value (representing a function of the area beneath the tolerance curve). KG values for the +D AL, +D PF, and -D groups were 504 +/- 15, 480 +/- 46, and 641 +/- 28, respectively. The increase in KG corresponded to a significant reduction in glucose-mediated insulin secretion as compared to the +D animals. This difference appeared not to be related to the increase caloric intake associated with vitamin D repletion, since +D rats which had been pair fed to the -D animals also exhibited restored plasma insulin levels in response to glucose. Plasma phosphorus concentrations were comparable in all three groups, and thus this parameter is also unlikely to be a contributory factor in the observed phenomenon. Additional experiments were conducted to evaluate the involvement of hypocalcemia in the observed impaired glucose tolerance. Normalization of plasma calcium levels (from 4.8 mg/100 ml to 9.6/100 ml) of the -D rats, by dietary calcium and phosphorus manipulation, failed to improve glucose clearance (KG for -D normocalcemic rats = 639 +/- 61) or insulin secretion. These results support the concept that vitamin D plays a physiological role in insulin secretion, acting, at least in part, independently of nutritional factors and the prevailing plasma calcium and phosphorus concentrations.     

Altered hepatic function in vitamin D-deprived rats

Rats were rendered vitamin D-deficient by housing them in a room free of ultraviolet light and maintaining them for 20 weeks on a diet devoid of only vitamin D. The vitamin D-deficiency state was confirmed by the undetectable levels of circulating vitamin D metabolites, severe hypocalcaemia and significantly reduced intestinal calcium transport. Liver function and protein metabolism in these rats were assessed by bromosulphthalein (BSP) clearance, liver histology, plasma transaminases and alkaline phosphatase, and 14C-labelled amino acid incorporation into liver and plasma proteins. Subtle alterations in hepatic function, as manifested by delayed BSP clearance, elevated levels of plasma transaminases and alkaline phosphatase, were noticed. Liver histology revealed changes consistent with periportal necrosis. Synthesis of liver and plasma proteins were reduced by 26-34% (P less than 0.01), without affecting the circulating levels of plasma proteins, suggesting reduced protein turnover in vitamin D-deprived rats. The results strongly suggest the direct/indirect involvement of vitamin D in mediating the altered liver function.     

Regulation of the murine renal vitamin D receptor by 1,25-dihydroxyvitamin D3 and calcium

Renal vitamin D receptor (VDR) is required for 1,25-dihydroxyvitamin D3-[1,25(OH)2D3]-induced renal reabsorption of calcium and for 1,25(OH)2D3-induced 1,25(OH)2D3 24-hydroxylase. The long-term effect of vitamin D and dietary calcium on the expression of renal VDR was examined in the nonobese diabetic mouse. Vitamin D-deficient and vitamin D-replete mice were maintained on diets containing 0.02%, 0.25%, 0.47%, and 1.20% calcium with or without 50 ng of 1,25(OH)2D3 per day. Vitamin D-replete mice on a 1.20% calcium diet had renal VDR levels of 165 fmol/mg protein. Calcium restriction caused renal VDR levels to decrease to <30 fmol/mg protein in vitamin D-deficient mice and to approximately 80 fmol/mg protein in vitamin D-replete mice. When dietary calcium was present, 50 ng of 1,25(OH)2D3 elevated the VDR levels 2- to 10-fold, depending on vitamin D status and the level of calcium. In the absence of either vitamin D or calcium, the VDR mRNA was expressed at a basal level. 1,25(OH)2D3 supplementation caused relative VDR mRNA to increase 8- to 10-fold in the vitamin D-deficient mouse when dietary calcium was available. This increase was completely absent in the calcium-restricted mice. This in vivo study demonstrates that 1,25(OH)2D3 and calcium are both required for renal VDR mRNA expression above a basal level, furthering our understanding of the complex regulation of renal VDR by 1,25(OH)2D3 and calcium.     

Deficiency of vitamin D metabolites directly stimulates renal 25-hydroxyvitamin D3-1-hydroxylase activity in rats

Renal 25-hydroxyvitamin D3-1-hydroxylase (1-hydroxylase) enzyme activity in rats is known to be increased by parathyroid hormone (PTH), hypophosphatemia, and hypocalcemia. Thus, enzyme activity is markedly increased in vitamin D-deficient states, but whether this stimulation is a direct response to the vitamin D deficiency or only occurs following the associated changes in plasma calcium, phosphate, or PTH is unclear. We tested whether vitamin D deficiency per se influences 1-hydroxylase activity in renal cortical slices using a normocalcemic rat model of vitamin D deficiency. Weanling male rats were fed one of the following three diets: (A) 0.8% Ca, 0.5% P, 2.2 IU vitamin D3/g; or vitamin D-deficient diets containing, (B) 0.8% Ca, 0.5% P; and (C) 2.0% Ca, 1.25% P, 20% lactose. Vitamin D-deficient rats fed diet B were hypocalcemic with elevated PTH at both test periods, and 1-hydroxylase activity was increased more than 100-fold compared with rats fed diet A. Plasma calcium, phosphate, and PTH levels were the same in groups A and C, but 1-hydroxylase activity was also substantially elevated in group C versus group A rats (104- and 17-fold increases after 10 and 19 diet weeks, respectively). These data lead to the important conclusion that severe deficiency of vitamin D metabolites per se provides a strong and independent stimulus to renal 1-hydroxylase activity in rats, perhaps due to the absence of 1,25(OH)2D3-mediated enzyme inhibition.     

1,25-Dihydroxyvitamin D3 enhances the effect of refeeding on steady state preproinsulin messenger ribonucleic acid levels in rats

To elucidate the regulatory mechanism of vitamin D action on insulin biosynthesis and secretion, we examined preproinsulin (ppI) mRNA levels in the pancreas of normal rats, vitamin D-deficient rats, and rats supplemented with 1,25-dihydroxyvitamin D3 [1,25-(OH)2D3] or calcium (Ca) for 3 days. The ppI mRNA levels determined by dot blot analysis in vitamin D-deficient, 1,25-(OH)2D3-replete, and Ca-replete rats were 39.1%, 68.7%, and 66.7%, respectively, of values in normal rats. These results concur with previously reported levels of insulin secretion in the perfused rat pancreas. The reduced level of ppI mRNA should lead to a decrease in insulin biosynthesis and, thus, impair insulin secretion in vitamin D-deficient rats. The observed partial recovery of ppI mRNA levels through supplementation of 1,25-(OH)2D3 or Ca may be one mechanism by which insulin secretion is restored in rats after 1,25-(OH)2D3 or Ca repletion. We examined further the time course of ppI mRNA accumulation in rats after a single administration of 1,25-(OH)2D3. When fasting was continued for an additional 24-h period after an overnight fast, ppI mRNA levels were not changed significantly in either vitamin D-deficient or replete rats. However, in the rats that were pair-fed after overnight fasting, ppI mRNA levels in 1,25-(OH)2D3-replete rats increased at 8 and 24 h, whereas ppI mRNA in vitamin D-deficient rats increased only at 24 h. Moreover, the increment at 24 h was significantly larger in 1,25-(OH)2D3-replete rats than in vitamin D-deficient rats. We conclude that 1,25-(OH)2D3 enhances steady state levels of ppI mRNA only under conditions of refeeding and during feeding.