Incubation with ¹⁴CD₃

Figures 1 –4 show the presence of several metabolites more polar than the substrate added before the incubation procedure. All species analysed metabolized vitamin D₃ into more polar products. At least four metabolites were produced and three of these coeluted with the standards 25OHD₃, 24,25(OH)₂D₃ and 1,25(OH)₂D₃, with retention times of 15, 18 and 25 min, respectively. An unidentified metabolite eluting at 10 min (A in the figures) was recorded. The identification of the metabolites 24,25(OH)₂D₃, 1,25(OH)₂D₃ and 25OHD₃ was confirmed by coeluting with the respective standards at 7.0, 8.6, and 15.5 min on reversed-phase HPLC and binding to serum vitamin D-binding protein and chick duodenal receptor. The relative amounts of the various metabolites produced within organs from one species showed a similar pattern among the species: 24,25(OH)₂D₃ was significantly the major metabolite in salmon and mackerel ( Figs 1 and 2) and, in most tissues. The most abundant metabolite in most organs from cod and halibut was 25OHD₃ ( Figs 3 and 4), although the significance was obvious only in cod. Furthermore the amount of 24,25(OH)₂D₃ in halibut was not significantly different from the amount of 25OHD₃.

Incubation with ³H25OHD₃

During incubation with ³H25OHD₃ in salmon, halibut, and cod all organs produced four products of greater polarity than the substrate; two of these coeluted with the standards, 24,25(OH)₂D₃ and 1,25(OH)₂D₃. The other two had retention times of 14 and 20 min, B and C, respectively. C eluted between 24,25(OH)₂D₃ and 1,25(OH)₂D₃. In the mackerel samples only 24,25(OH)₂D₃, C, and 1,25(OH)₂D₃ were found. Production of metabolites is shown in Figs 5 –8. The hydroxylation of ³H25OHD₃ demonstrated a pattern more closely related to species than to organs. The production of metabolites among organs within a species showed similarities but compared with the ¹⁴CD₃ incubation, the pattern was somewhat more diffuse, especially in halibut. The greatest amounts of 24,25(OH)₂D₃ were produced in salmon and mackerel, with significant differences in kidney, head kidney, gills and spleen of salmon, and in head kidney, gills and spleen of mackerel ( Figs 5 and 6). Fewer significant differences were found in the samples from halibut and cod for the amounts of 24,25(OH)₂D₃. However, the tendency of 24,25(OH)₂D₃ as dominating metabolite was still obvious ( Figs 7 and 8).

There were, however, some problems in the quantitative determinations owing to a negative correlation between the concentration of homogenates and the production of metabolites (data not shown). Homogenates with high concentrations of tissue gave a lower measurable activity of hydroxylases. This correlation was confirmed by the inverse relationship of 1,25(OH)₂D₃ to the concentration of tissue in the additional incubation of the kidney from cod, r = 0.99 (Fig. 9).

Discussion

In the present study, hydroxylated products of the substrates were detected in each of the tissues analysed from all four species. Qualitatively, the method demonstrated clearly the activity pattern of the hydroxylases. The identities of the metabolites were confirmed by their coelution with the standards, as well as by being specifically bound to the binding proteins. However, in discussing these results it must be taken into account that the degree of hydroxylation was negatively correlated with the concentration of tissue homogenate in the reaction mixture. The explanation for this phenomenon was probably that the homogenates contained amounts of nonradioactive substrate competing with the radioactive substrate for hydroxylation. This circumstance limited the possibility of comparing the absolute production of metabolites per gram tissue within the different organs. To our knowledge, this aspect has been paid little attention in the literature, even though several concentrations of homogenates have been used (bbr rid="b4">Bailly du Bois et al. 1988) and, sometimes, the concentrations are not given ( Blondin et al. 1967 ; Henry & Norman 1975; Kenny et al. 1977 ; Kobayashi et al. 1991 ). Livers from marine fish contain large amounts of vitamin D₃ (Bills, 1927) and, in every species analysed, liver was the only organ where difficulties in registration of the hydroxylated products occurred. These samples still produced products that were more polar than the substrate and all other organs clearly demonstrated metabolism of vitamin D₃ and 25OHD₃. Such activity has not been recorded in organs other than liver and kidney in previous studies. It has been suggested, however, that the liver is the major hydroxylation site. This may be the reason for the low-affinity receptors most fish possess in their vitamin D₃ transport system ( Sundell et al. 1992 ), in addition to the low level of 25OHD₃ in fish serum ( Takeuchi et al. 1991a ). Robertson (1993) supported these conclusions and suggested that the fish kidney is less essential for the hydroxylation of vitamin D₃ than the kidneys in birds and mammals. The results in the present study demonstrated activity of the hydroxylases in all organs analysed and activity was also found in species and organs in which previously it has been questionable or such sites of hydroxylation has been unknown. Takeuchi et al. (1991b ) confirmed the hydroxylations to be enzymatic by heating some of the samples to 100°C for 15 min: preheating eliminated the peaks corresponding to hydroxylation products such as 25OHD₃ and 1,25(OH)₂D₃.

Human tissues show great organ specificity for the various conversions of vitamin D₃ to more polar metabolites. The 25-hydroxylation mainly takes place in the liver, whereas the 1?-hydroxylation is located in the kidney ( Walters 1992). Takeuchi et al. (1991b ) suggested a content of hydroxylases in carp kidney similar to that in the mammalian kidney, presuming that the 1?-hydroxylase is the predominant hydroxylation enzyme. In the present study, we observed a pattern of metabolites produced in different organs, which suggested a lower degree of specificity in fish organs compared with mammalian organs. A number of metabolites was produced in each fish organ and at least three hydroxylated metabolites were produced from the added substrate in all the organs analysed. This indicates that the presence of one hydroxylated product in a fish organ suggests the existence of other metabolites. Some fish species probably have a more general distribution of the enzymes or the enzymes can hydroxylate in more than one position. It has been observed in previous studies that the carp liver contains both 25-hydroxylase and 1?-hydroxylase ( Takeuchi et al. 1991b ). Furthermore, it has been postulated earlier that at least two hydroxylations of vitamin D₃ are due to a common active site in a single rat protein ( Ohyama et al. 1991 ). This is not in agreement with earlier studies ( Dahlbäck 1988). The metabolite C in the histograms, which eluted between 24,25(OH)₂D₃ and 1,25(OH)₂D₃, is probably 25,26(OH)₂D₃. Previous studies recorded production of this metabolite in rainbow trout when incubated with ³H25OHD₃ ( Hayes et al. 1985 , 1986), and according to Aksnes & Aarskog (1980) the retention time of 25,26(OH)₂D₃ is in the interval between 24,25(OH)₂D₃ and 1,25(OH)₂D₃, independent of the mobile phase. The function of this metabolite is not elucidated, although it was isolated and identified from in vivo sources in 1970 ( DeLuca & Schnoes 1983). We were unable to identify metabolite A and B. A is, however, less polar than 25OHD₃ and B has an intermediate polarity between 25OHD₃ and 24,25(OH)₂D₃. Incubations with ³H25OHD₃ in previous studies resulted in metabolites that were less polar than 25OHD₃, but the authors had no clear suggestions about the functional and metabolic identity ( Yanda & Ghazarian 1981; Bailly du Bois et al. 1988 ).

Among the hydroxylated compounds 24,25(OH)₂D₃ appears to be particularly interesting. Partly because it is quantitatively the most important one in salmon and mackerel incubated with ¹⁴CD₃ and it is also predominant in most tissues incubated with ³H25OHD₃. Bailly du Bois et al. (1988 ) recorded that the compound which eluted close to 24,25(OH)₂D₃ was the major metabolite in the European eel. Even if they did not succeed in identifying the compound as 24,25(OH)₂D₃, one cannot exclude this possibility. According to the results of the present study, Bailly du Bois et al. (1988 ) may have eluted a mixture of 24,25(OH)₂D₃ and 25,26(OH)₂D₃. In humans, 24,25(OH)₂D₃ is produced in the kidneys in amounts inversely related to the amount of 1,25(OH)₂D₃ ( Okuda et al. 1995 ). The role of 24,25(OH)₂D₃ in mammals has not been clearly defined and whether 24,25(OH)₂D₃ is biologically active in humans or whether it is simply a part of a catabolic pathway has been a topic of active debate ( Walters 1992).

Fish species living in a marine environment have a large calcium supply and use a great deal of energy to get rid of this from the body. Freshwater species have the opposite problem, needing a calcium transport from the water to the animal ( Kryvi 1990). If the vitamin D metabolites regulate the calcium balance and 24,25(OH)₂D₃ has a hypocalcaemic role, different amounts of 24,25(OH)₂D₃ and 1,25(OH)₂D₃ within the species from the two separate environments may be expected. The presence of 24,25(OH)₂D₃ has been demonstrated only once in fish serum ( Takeuchi et al. 1991a ), although several studies still indicate a hypocalcaemic function of this metabolite in fish ( Wendelaar Bonga et al. 1983 ). Support for this hypothesis is the lowered intestinal calcium uptake in Atlantic cod shortly after 24,25(OH)₂D₃ administration, independent of pharmacological (Sundell et al. 1990) or physiological concentrations ( Larsson et al. 1995 ). Hayes et al. (1986 ) have also found a reciprocal relationship between the production of 1,25(OH)₂D₃ in rainbow trout and the calcium content in the water. The species analysed in the present study do all spend at least periods of their lives in hypercalcaemic environments and elevated levels of 24,25(OH)₂D₃ during incubation with ¹⁴CD₃ and ³H25OHD₃ supports 24,25(OH)₂D₃ as a hypocalcaemic compound.

Because of the aforementioned problems regarding concentration of tissue homogenates in the samples, it is not possible to derive any conclusions about quantitative hydroxylations of vitamin D₃ in the tissues analysed. Earlier studies used other concentrations of homogenates and it is therefore difficult to compare the results of the present study with previous publications. To minimize the methodological problems, the content of substrate in the tissue prior to incubation has to be controlled. This can be done by analysing the exact content of the nonradioactive substrate and taking this into account during handling of the results. Another and perhaps better alternative is to make the nonradioactive substrate inaccessible by the use of proteins which bind the substrate specifically. Furthermore, an alternative solution might be to remove the lipids in the sample prior to incubation. More research is needed in this field. We have, however, demonstrated metabolism of vitamin D₃ and 25OHD₃ in six tissues in the four fish species analysed and observed hydroxylation in tissues which previously have been questionable or unknown as hydroxylation sites. Among the hydroxylated products, 25OHD₃, 24,25(OH)₂D₃, 1,25(OH)₂D₃ and perhaps 25,26 (OH)₂D₃ were identified. It was found that 24,25(OH)₂D₃ was the quantitatively predominant metabolite and this may be explained by the need of the species in calcium-rich environment to decrease rather than increase their calcium absorption.

Acknowledgements

The authors thank K. Ask for skilful technical assistance during the handling and incubation of the fish tissues.