In vitro hydroxylation of vitamin D3 and 25-hydroxy vitamin D3 in tissues of Atlantic salmon Salmo salar, Atlantic mackerel Scomber scombrus, Atlantic halibut Hippoglossus hippoglossus and Atlantic cod Gadus morhua
GRAFF, LIE & AKSNES
1 Institute of Nutrition, Directorate of Fisheries, Bergen, Norway 2 Department of Pediatrics, Haukeland Sykehus, Norway
The in vitro metabolism of 14CD3 and 3H25OHD3 was investigated in different tissues from Atlantic salmon Salmo salar, Atlantic mackerel Scomber scombrus, Atlantic halibut Hippoglossus hippoglossus and Atlantic cod Gadus morhua. The tissues analysed were liver, kidney, head kidney, gills, spleen and intestine. The metabolites were extracted in methanol–chloroform and separated by normal-phase high-pressure liquid chromatography (HPLC) followed by scintillation counting. Identification of the metabolites was by comigration with standards on normal and reversed-phase HPLC systems and by protein-binding assays. All tissues from all species analysed produced hydroxylated derivatives identified as 25OHD3, 24,25(OH)2D3 and 1,25(OH)2D3. In addition, some unidentified derivatives were recorded, one probably being 25,26(OH)2D3. Organs producing great amounts of one metabolite also produced considerable amounts of the other possible derivatives, suggesting a lower degree of specificity in fish organs than in human organs. The predominating metabolite was 24,25(OH)2D3 in all organs from salmon and mackerel during incubation with 14CD3 and within most organs from all species during 3H25OHD3 incubation. The latter observation probably results from the need for decreasing rather than increasing the calcium absorption in these species, which live at least some periods of life in a marine environment.
The metabolism of vitamin D3 in mammals is well understood. The vitamin normally undergoes two hydroxylations, the first in the liver, transforming vitamin D3 to 25OHD3, and the second in the kidneys, which hydroxylate 25OHD3 to 1,25(OH)2D3 ( DeLuca & Schnoes 1983).
The 1,25(OH)2D3 compound is the most active metabolite of vitamin D3 and regulates the calcium and phosphate balance in mammals ( Fraser 1980; DeLuca & Schnoes 1983; Fraser 1995). 25OHD3 is also hydroxylated to 24,25(OH)2D3, but the function of this metabolite remains unclear, although there are suggestions that it has an antihypercalcaemic role in hypercalcaemic patients ( Sherrard et al. 1985 ). Normal bone growth is probably also dependent on 24,25(OH)2D3, at least in birds ( Norman & Hurwitz 1993).
(OH)2D3 was identified as a circulating metabolite in teleost fish ( Avioli et al. 1981 ). Several sets of data suggest that this compound is at least one of the active vitamin D3 metabolites in fish. In vivo administration of 1,25(OH)2D3 resulted in increased serum phosphate concentration in European eel Anguilla anguilla ( MacIntyre et al. 1976 ), intestinal calcium absorption in American eel Anguilla rostrata ( Fenwick et al. 1984 ) and had positive effect on weight gain and reduced the incidence of tetany in rainbow trout Oncorhyncus mykiss ( Barnett et al. 1982a , b). The studies of the physiological role of vitamin D3 in fish have focused mainly on freshwater species, although the highest concentrations of the vitamin are found in the liver of the marine species ( Bills 1927). However, Sundell et al. (1992 , 1993) have identified 1,25(OH)2D3 receptors in calcium-regulatory tissues such as gills and intestine in Atlantic cod and observed increased calcium absorption after 1,25(OH)2D3 administration in vivo. In addition, Sundell & Björnsson (1990) and Larsson et al. (1995 ) suggested a hypocalcaemic role for 24,25(OH)2D3in vitro in Atlantic cod. They presume that the effect of 24,25(OH)2D3 results from the needs of fish in a hypercalcaemic environment to limit calcium absorption.
The fact that teleost fish, especially marine species, possess large hepatic stores of vitamin D3 has been adequately demonstrated ( Urist 1976). The specific metabolism of vitamin D3 has not been clearly elucidated and the results presented to date are contradictory. Oizumi & Monder (1972) did not observe metabolism of vitamin D3 in goldfish Carassius auratus and Holick et al. (1982 ) did not find significant levels of the metabolites in serum from rainbow trout. The first demonstration of metabolism of vitamin D in both fresh and saltwater fish tissues came in an extensive in vitro study where 25OHD-1-hydroxylase was detected ( Henry & Norman 1975). Furthermore, Yanda & Ghazarian (1981) observed metabolism of vitamin D in vivo and in vitro in rainbow trout, even though they could not completely identify the metabolites. Later studies identified the metabolites as being 25OHD3, 1,25(OH)2D3, 24,25(OH)2D3 and 25,26(OH)2D3 in plasma from European eel ( Marcocci et al. 1982 ), rainbow trout ( Hayes et al. 1985 ) and channel catfish Ictalurus punctatus ( Brown & Robinson 1992). The more recent studies ( Hayes et al. 1986 ; Bailly du Bois et al. 1988 ; Takeuchi et al. 1991b ) have all demonstrated the production of more polar products of vitamin D3in vivo and in vitro in at least liver and kidney in various fish species. Some of these derivatives have not been found in mammals. The aim of the present study was to investigate the metabolism of vitamin D3 and 25OHD3in vitro in four species of anadrome and saltwater fish and to clarify if any other tissue than liver and kidney are involved in this metabolism.
Materials and methods
Materials
All reagents and chemicals were of analytical grade and were purchased from general laboratory suppliers. Non-radioactive standards 25OHD3, 1,25(OH)2D3 and 24,25(OH)2D3 were obtained from F. Hoffman-LaRoche and Co., Basle, Switzerland. Incubation medium, RPMI 1640, was obtained from Bio Whittaker, Inc, Walcesville, MD, USA, and [4–14C]vitamin D3, 57 mCi mmol–1 and 25-hydroxy-[26(27)-methyl-3H]cholecalciferol, 15 Ci mmol–1 were purchased from Amersham, England, and stored on ethanol (950 g L–1 at –20°C. Owing to limited availability, we were forced to use two different radioactive marks on the vitamin D3 compounds, but this was not considered a problem.
Fish
Atlantic salmon Salmo salar weighing 188 ± 23 g were obtained from Matre Aquaculture Research Centre (Institute of Marine Research, Bergen, Norway). They were anaesthetized in a solution of metomidate (7 g L–1; Wildlife Laboratories, Fort Collins, USA) and dissected. Mackerel Scomber scombrus, halibut Hippoglossus hippoglossus, and cod Gadus morhua weighing 226 ± 33 g, 566 ± 191 g and 390 ± 116 g, respectively, were obtained from Austevoll Aquaculture Research Centre (Institute of Marine Research, Bergen, Norway). They were killed by cranial percussion and transported (1.5 h) on ice to Bergen where they were dissected.
Incubation procedure
Liver, kidney, head kidney, gills, spleen and intestine were removed immediately, weighed and homogenized in RPMI-medium by use of a Potter–Elvehjem homogenizer (Kebo Lab, Bergen, Norway) on ice. For each species, organs from five fish were used. The homogenates had a concentration of tissue between 0.03 and 0.65 g mL–1. From each homogenate 1 mL was transferred to two 10 mL Pyrex tubes containing 1.1 × 10–6 mmol (25 ?L) 14CD3 or 2.1 × 10–8 mmol (25 ?L) 3H25OHD3, respectively. Two samples containing only RPMI medium and isotopes were used as controls. The tubes were incubated in a water bath, 30°C, for 60 min and the incubation was terminated by adding 2 mL of methanol (modified from Bailly du Bois et al. 1988 ; Takeuchi et al. 1991b ).
Extraction and chromatography
The samples were extracted with 4 mL chloroform by mixing for 45 min, 1 mL of water was added and the tubes were cooled in a refrigerator (4°C) for 30 min and centrifuged for 10 min at 400 g. The chloroform layer was aspirated and dried under nitrogen. The residue was suspended in 250 ?L of n-hexane–ethanol–isopropanol (96:2:2, v/v) and applied to a normal-phase high-pressure liquid chromatography (HPLC) system (4.6 × 150 mm of Supelcosil, 3 ?m, with a 20 mm Supelguard column Supelco, Inc., Bellefonte, PA, USA) with a UV-detector operating at 265 nm (modified from Aksnes & Aarskog 1980). Samples incubated with 14CD3 were eluted from start to 7.4 min by n-hexane–ethanol–isopropanol (99:0.5:0.5, v/v) 1.5 mL min–1 and then the mixture was changed to 94:3:3 (v/v). From 25.1 min to 30.1 min there was a linear change from 94:3:3 (v/v) to 85:7.5:7.5 (v/v) and at 30.1 min the latter solution was used until 45.1 min, when the column was equilibrated for 15 min with the 99:0.5:0.5 (v/v) mixture before applying the next sample. For samples incubated with 3H25OHD3, elution started with n-hexane–ethanol–isopropanol at 98:1:1 (v/v) and had the same times of change as the previous samples, but changed from 98:1:1 (v/v) to 94:3:3 (v/v) and had a linear change to 85:7.5:7.5 (v/v), which was used until the column was equilibrated with 98:1:1 (v/v) (modified from Aksnes & Aarskog 1980). Thirty-six 1.5 mL fractions were collected and suspended in 3 mL toluene-based scintillant followed by counting for radioactivity in Tricarb 1900 CA (Packard Instrument Company Inc., the Netherlands). The metabolites were also identified by coelution with standard 25OHD3, 24,25(OH)2D3 and 1,25(OH)2D3 by a reversed-phase HPLC system, using a 4.6 × 250 mm Supelcosil RP-18, 5 ?m, column with a 20 mm Supelguard column, eluted with methanol–water (85:15, v/v) at 1,5 mL min–1.
Identification of metabolites by protein binding assay
An additional identification of the metabolites were carried out through a protein-binding assay (PBA). Kidney (39.02 g) from an Atlantic salmon (6.76 kg) from Matre Research Station was homogenized and a tissue homogenate of 0.03 g mL–1 was incubated with 14CD3 and extracted according to the procedure described above. The samples were analysed after separation on both normal-phase HPLC and on reverse-phase HPLC. Fractions from the normal-phase HPLC were collected and an aliquot of each fraction were counted for radioactivity. The other aliquots were handled according to the PBA described by Aksnes (1980). They were evaporated to dryness and solubilized in 100 ?L ethanol (950g L–1) and mixed either with vitamin D binding protein (DBP) from a normal pregnant woman's diluted serum to achieve identification of 25OHD3 and 24,25(OH)2D3, or with chick duodenal cytosol-receptor protein to bind 1,25(OH)2D3 ( Aksnes 1980).
Additional analysis of kidney from cod
Because of a negative correlation between concentration of the homogenates and the production of metabolites, a kidney from an Atlantic cod was analysed to uncover any systematic dependence. The cod was bought at the fish market in Bergen; the weights of the fish and the kidney were 3.55 kg and 5.5 g, respectively. The kidney was homogenized, and incubations were performed with 2.1 × 10–8 mmol (25 ?L) 3H25OHD3 following exactly the same procedures for extraction and chromatography as for the organs mentioned above. The concentrations of kidney tissue in the samples were 5, 10, 43, 90, 200, and 500 mg mL–1 in duplicates.
Statistical analysis
The data were statistically evaluated by means of one-way ANOVA after being tested for homogeneity of variance (Levene's test). Any data found to be heteroscedastic were subjected to logarithmic or square root transformation. Tukey's HSD (honest significant difference) test were used for testing differences in the amounts of metabolites produced within each organ. Spearman rank order correlations were calculated to determine the correlations between concentration of tissue and production of metabolites. The significance level was set to P ? 0,05, and data are presented as means ± SD.
Email
Print
Permalink