Amino Acid Uptake and Metabolism by Embryos of the Blenny Zoarces Viviparus

Previous studies have shown that embryos, larvae and juveniles from different teleost species have the ability to take up low molecular weight compounds such as glucose and amino acids from their ambient medium (Siebers and Rosenthal, 1977; Lin and Arnold, 1982; Fauconneau, 1985). By autoradiographic methods, it was recently shown that turbot (Scophthalmus maximus) larvae are able to ingest amino acids from sea water during their yolk-sac stage (Korsgaard, 1991). The significance of these different observations in relation to larval nutritional physiology is at present very uncertain. It has been suggested that for fish larvae, which encounter a metabolic deficit during the transition period up to first feeding, the ability to assimilate low molecular weight compounds may have nutritional importance (Heming and Buddington, 1988). Considering the high mortality of marine fish larvae during early post-yolk-sac life, nutritional requirements and metabolic capacity during early larval development are of particular interest in the field of fish physiology.

Maternal-foetal trophic relationships in viviparous teleosts offer a unique opportunity for the study of nutritional physiology of developing fish embryos (Wourms et al. 1988; Korsgaard and Weber, 1989). Heming and Buddington (1988) suggested that, by rearing embryos from viviparous teleosts, before and after hatching on defined media, it may be possible to obtain valuable information on the nutritional requirements of early life stages.

In the blenny Zoarces viviparus, the embryos are retained in the ovarian cavity for 4–5 months. During this period they undergo extensive growth and development, from a fertilized egg of approximately 20 mg through yolk-sac and post-yolk-sac stages into a young fish of approximately 400 mg at parturition (Korsgaard and Andersen, 1985; Korsgaard, 1986). During most of this time they lie freely in the ovarian lumen surrounded by ovarian fluid, which contains dissolved organic substances such as glucose at a concentration of approximately 1000μmoll⁻¹ (Korsgaard, 1983) and amino acids at approximately 400μmoll⁻¹ (Kristoffersson et al. 1973). These substances are believed to serve as nutritional material for the growing embryos (Kristoffersson et al. 1973; Korsgaard, 1983) In the present work, the capacity of the developing embryos after hatching to assimilate and metabolize L-[¹⁴C]alanine from the ambient medium in vitro or in vivo was investigated. Since uptake, evaluated by a tracer technique only, may not indicate the occurrence of net uptake, medium depletion of labelled L-alanine was compared with net changes in ambient alanine concentrations as assessed by high performance liquid chromatography (HPLC) measurements. In addition, autoradiograms were prepared and studied by dark-field illumination to detect the presence of label in the embryonic tissue. Metabolic conversion of tracer L-alanine was evaluated by the production of ^l4CO₂ and by distribution of labelled carbon in the trichloroacetic acid (TCA)-soluble and TCA-insoluble (protein) fractions of embryonic tissue.

Pregnant female blennies Zoarces viviparus were caught in the Little Belt of Denmark. The fish weighed between 120 and 155 g. In the laboratory they were kept in recirculated, aerated sea water (20‰) at 11°C and allowed an acclimation period of 6 days before experiment.

In the in vitro experiments the embryos were dissected out of a quickly decapitated female, weighed and placed in Ringer’s solution for 1 h at 11°C. Standard incubations consisted of 1.5 g of embryos in 10 ml of Ringer’s solution with 60 or 600 μmol 1⁻¹ unlabelled carrier L-alanine to which 10 (50 μCi) of tracer was added. Labelled L-[U-¹⁴C]alanine (New England Nuclear) with a specific activity of 172 mCi mmol⁻¹ was used in the experiments. The low concentration of unlabelled alanine was chosen because it is similar to the natural concentration found in the ovarian fluid, which is within the range of 51.6±20.3 μmol 1⁻¹(N=5) measured in November (Kristoffersson et al. 1973) to 64.2±7.2_μmoll⁻¹(N=6) as measured in October by HPLC in the present work. The high concentration (600μmoll⁻¹) was chosen to investigate the effect of a different carrier concentration on the uptake rate and turnover time of the labelled L-alanine. The Ringer’s solution consisted of iso-osmotic saline, 170 mmol l⁻¹ NaCl, 4 mmol 1⁻¹ KC1, 4mmoll⁻¹ CaCl₂, lOmmoll⁻¹ NaHCO₃, pH7.5, with a final osmotic concentration of 340mosmol kg⁻¹, in accordance with values obtained by Kors-gaard (1983) in ovarian fluid. All glassware used in the experiments was carefully acid-cleaned, and all solutions were freshly made with double-distilled water. No antibiotics were used in the experimental samples. However, identical samples (two in each experiment) containing penicillin and streptomycin (5000 i.u. and 50 mg l⁻¹) were used as controls for microbial activity in the experiments. These samples were not included in the results since uptake and turnover rates and production of labelled CO₂ in the antibiotic-treated control samples were found to be similar to those of the experimental samples to which no antibiotics were added. Samples were taken at intervals to follow the time course of depletion of labelled and total L-alanine from the medium. 2 ml of 0.02 mol 1⁻¹ HC1 was added to the 0.1ml sample and the mixture was shaken for 20 min to remove ¹⁴CO₂ produced by respiration of the labelled substrate. Scintillation fluor (2 ml of Hydrocount) was added and the radioactivity was counted in a Searle Mark III liquid scintillation counter with automatic quench correction. The quench correction was based on a quench curve constructed on the basis of a quench series and enclosed in the memory of the scintillation counter. The samples were counted after 5h to avoid chemiluminescence in the CO₂ extracts. The concentration of L-alanine in the samples was measured by high performance liquid chromatography (HPLC) according to the method of Godel et al. (1984) using o-phthaldehyde (OPA) as the derivatizing agent and a Waters Novapack C18 (3.9 mmx 15 cm) stainless-steel column. HPLC-graded, filtered NaH₂PO₄ (12.5 mmol 1⁻¹) buffer and acetonitrile/NaH₂PO₄ (1:1) were used as the solvent systems at pH7.5. The L-alanine peak in the samples was identified on the basis of retention time established with an L-alanine standard (Sigma), using internal standards as controls. The concentration of L-alanine was calculated from the peak area of the samples on the basis of the standard peak area.

After 24h of labelling in the tracer medium, the embryos were rinsed, blotted on absorbent tissue paper and placed in tracer-free Ringer’s solution (5 ml) in closed flasks. The effect of the rinsing procedure was tested by taking a sample of the Ringer’s solution immediately after transfer of the embryos and counting it in the scintillation counter. 0.2 ml of 10% KOH was added to paper wicks as the ¹⁴CO₂-trapping system in the flasks. The release experiment was stopped after 2 h by removing the embryos and the ¹⁴CO₂ and dissolved organic carbon (DOC) were measured as described by Korsgaard and Andersen (1985). Embryonic tissue was disrupted, extracted overnight in 10% TCA, washed twice in 5% TCA made up in 50% ethanol and digested in Lumasolve. Radioactivity in the resulting TCA-insoluble and TCA-soluble fractions of the embryonic tissue was determined by scintillation counting.

L-alanine as measured by HPLC.

In the in vivo experiments, 70 μl of the radioactive compound was injected into the ovary of the pregnant fish simultaneously with 1 ml of 60 or 600 μl mol 1⁻¹ carrier alanine solution. Samples of the ovarian fluid were collected by syringe at intervals. The disappearance of labelled L-alanine was measured by scintillation counting. After 24 h of intraovarian incubation, the embryos were quickly dissected out, rinsed and blotted. They were then immediately placed in unlabelled Ringer’s solution in closed flasks as described above. The production of ¹⁴CO₂ and DOC was measured and calculated as a percentage of the radioactivity in the embryos at the beginning of the release experiment. Radioactivity in the TCA-soluble and TCA-insoluble fractions of the embryos was determined as described above.

For autoradiography, embryos were taken after 24 h of exposure to labelled L-alanine and prepared according to the methods described by Korsgaard (1991). Briefly, the embryos were dehydrated then embedded in paraffin and serially sectioned in 5-μm sections. Sections were coated under darkroom conditions with Kodak fine-grain stripping film, placed in black boxes with silica gel desiccant and stored for 3 weeks. After development of the autoradiograms, embryonic sections were photographed using dark-field illumination. Control sections from embryos which had not been exposed to radiolabelled alanine were prepared to evaluate positive and negative artefacts. Results were subjected to statistical analysis by Student’s t-test.

Fig. 1 shows that after 24 h L-[¹⁴C]alanine is distributed throughout the tissues of the embryo. In Fig. 2B,D the autoradiograms represent sections of intestinal and muscular tissue from embryos in late development exposed for 24 h to labelled L-alanine. Fig. 2A,C shows corresponding sections from embryos not exposed to radiolabelled alanine, but otherwise prepared by the same autoradiographic method. A careful analysis of the spread of grain densities over background areas was performed on several sections to evaluate the difference between label and background. The autoradiograms in the figures, viewed by dark-field illumination, confirm that label is incorporated into the embryonic tissues well above back-ground levels. Thus, the experiment clearly demonstrates that the embryos are able to assimilate alanine from ambient micromolar solutions.

From Fig. 3A it appears that the embryos during their yolk-sac stage already have the ability to take up alanine from a 60μmoll⁻¹ ambient solution, which is approximately the concentration normally found in the ovarian fluid (an average of 64μmol 1⁻¹ as measured by HPLC in the present experiment). From Table 1, however, it appears that the disappearance rate of the tracer (0.52 μmolg⁻¹ h⁻¹) is significantly higher than the net disappearance rate of L-alanine (0.33 μmol g^−lh⁻¹) as measured by HPLC. The difference may be due to a concomitant release of cold alanine from the embryos. That the embryos do not use exogenous alanine as efficiently during their yolk-sac stage is also reflected by a smaller production of labelled CO₂ and a larger production of dissolved organic carbon (DOC) (Table 1), when compared with later stages of development (Table 2)

Fig. 3B shows substrate depletion in the ambient medium in vitro by embryos in late development, approximately 2 months later, when the yolk-sac has been absorbed. It appears that net flux measured by HPLC closely follows the uptake of the tracer. The calculated rates of uptake (Table 2) measured by the tracer method (0.60μmol g⁻¹ h⁻¹) and the HPLC method (0.54μmol g⁻¹h⁻¹) at a carrier concentration of 60μmol 1⁻¹ show a close correspondence. The production of labelled CO₂ at 60μmol 1⁻¹ alanine appears to be larger (2.36%) than that observed during the yolk-sac stages (1.46%). The results indicate that there is a higher metabolic conversion of alanine during late development of the embryos and that their capacity for uptake of amino acids increases with increasing metabolic demands during development.

The in vivo experiments were carried out to investigate whether the embryos take up and metabolize L-[^l4C]alanine when they are enclosed in the ovarian sac. The tracer alanine was completely cleared during 24 h in both carrier groups. However, different time courses were found for label in the ovarian fluid at the two carrier concentrations (Fig. 4). The labelled alanine was taken up faster at the low carrier concentration as indicated by the initial part of the curve (at 0.25 and 0.5 h). After 24 h of intraovarian labelling, production of labelled CO₂ and DOC by the embryos was measured for 2h (Table 3). The production of ¹⁴CO₂, calculated as a percentage of the radioactivity in the embryos immediately after dissection from the ovary, lies within the same low range as was observed in vitro.

Anabolic metabolism of L-alanine was indicated by the large incorporation of radioactivity into molecules insoluble in TCA (protein) (Table 4). Results from the in vitro experiments were included in Table 4 for comparative reasons. The fraction of the tracer L-alanine that is precipitated as protein in post-yolk-sac embryos appears to be of the same magnitude at 60μmol l^−l alanine both in vitro and in vivo. The TCA-insoluble fraction of the embryonic tissue amounts to approximately 35% of the total radioactivity in the two 60 μmol l⁻¹ alanine carrier groups and 24% in the 600μmoll⁻¹ alanine groups and in embryos in early development.

In the present work the term ‘embryo’ has been used for the entire posthatch development in the ovary, as it normally is in viviparous fish (Wourms et al. 1988). The embryos of Zoarces viviparus are not provided with any specific external structures for absorption as are embryos from many other viviparous fish. The alimentary canal is therefore believed to be the principal pathway for uptake of nutrients from the ovarian fluid, at least during later stages of development, when the skin of the embryos is relatively thick and non-vascularized (Kristoffersson et al. 1973). These authors describe the greatly expanded and richly folded hindgut as the only exceptional structure for nutrient uptake in the developing embryos.

Drinking activity by embryos of Z. viviparus may thus be one of the prerequisites for uptake of nutritional substances such as amino acids, which can be detected in the ovarian fluid. This has also been suggested by Kristoffersson et al. (1973), who observed that red blood cells, which may be found in the ovarian fluid as a result of minor bleeding, could also be found in the alimentary tract of the embryos during later development. Their observations are confirmed by the present autoradiographic studies, which show that label can be observed in the intestinal tissue of the embryos (Fig. 2) exposed to [^l4C]alanine for 24 h. Similarly, autoradiographic studies performed on the teleosts Clinus superciliosus and C. dorsalis during intrafollicular gestation showed that embryos absorbed nutrients from the embryotrophe secreted by the follicular wall and that most of the nutrient absorption took place through the alimentary canal (Veith, 1980; Cornish, 1985). Drinking activity has also been shown to occur in yolk-sac and first-feeding larval stages of various oviparous teleosts and is normally related to osmoregulation (Mangor-Jensen and Adoff, 1987; Tytler and Blaxter, 1988). Drinking may, however, also represent a way of supplying nutrients or other substances via the gut, as suggested by Tytler et al. (1990) and Korsgaard (1991). However, during early development of Z. viviparus embryos, the skin and gills may also be of importance as absorptive surfaces.

Net uptake rates for L-alanine by embryos in vitro are considerably lower during their yolk-sac stages than rates found after absorption of the yolk-sac. This observation indicates that the yolk-sac embryos do not utilize ambient substrate as effectively as they do later in development. This difference may simply be due to an increased surface area for absorption in the gut or an increased drinking rate in the post-yolk-sac embryos. In contrast, Fauconneau (1985) found the highest rate of amino acid incorporation into body protein in yolk-sac larvae of the coregonid fish (Coregonus scinzii pallea) when he compared different developmental stages of the fish after immersion in solutions containing [^L4C]arginine. In Zoarces viviparus, the rate of embryonic growth during early development is probably a function of mixed feeding, defined as an uptake of yolk nutrients from the yolk mass and a simultaneous uptake of low molecular weight compounds from the ambient ovarian fluid. Thus, mixed feeding, in viviparous as well as in oviparous teleost larvae, may prevent any potential metabolic deficit prior to complete yolk absorption and may enhance growth and survival during the terminal phase of the yolk-sac stages (Eldridge et al. 1981,1982; Heming et al. 1982; Wiggins et al. 1985; Yin and Blaxter, 1987).

During late embryonic development, the rate of disappearance of [¹⁴C]alanine from the ambient medium seems to be an accurate measurement of net flux, when compared with the medium depletion of alanine measured by HPLC. The considerable increase in net uptake rates from 0.33 μmol g⁻¹ h⁻¹ during yolk-sac stages to 0.54μmolg⁻¹ h⁻¹ during late development and the concomitant increase in the production of ¹⁴CO₂ from 1.46% to 2.36% per 2h indicate that the importance of amino acids for nitrogen anabolism and catabolism has increased in relation to the increasing weight of the embryos. Similarly, the amount of nitrogen per embryo was shown to increase linearly with time after hatching of the Z. viviparus embryos, indicating that nitrogenous energy sources are involved in the maternal nutrition of the embryos in the ovary (Korsgaard, 1986). According to observations by Terner (1968), the capacity of fertilized eggs from rainbow trout (Oncorhynchus my kiss) to oxidize ¹⁴C-labelled acetate, pyruvate and glucose also increased continuously during development.

The release of labelled CO₂ and dissolved organic carbon in the present work was in general very low, indicating that amino acids are used mostly for synthetic activities. This agrees well with previous studies on embryos from Zoarces viviparus, which showed that glucose contributed approximately 14% of the average total aerobic respiration, whereas the amino acid glycine contributed only insignificantly to respiration (Korsgaard and Andersen, 1985). The oxidation rate of L-alanine in the present work is in accordance with the oxidation rate observed in turbot larvae during their yolk-sac stage (Korsgaard, 1991). Fyhn and Serigstad (1987) found free amino acids in developing eggs and early yolk-sac stages of the larvae in cod (Gadus morhua) to be the main substrate for aerobic energy production during the first 19 days after hatching. The mass of body protein did not change significantly during this period, whereas the pool of free amino acids was reduced by 80% over the period of investigation.

Complete clearance of tracer L-alanine from the ovarian fluid was observed during 24 h in vivo. These experiments were performed to compare anabolic and catabolic metabolism of embryos labelled under natural conditions in the ovarian fluid with metabolism of embryos labelled under in vitro conditions. Calculated on the basis of total label in the embryos after 24 h, catabolism evaluated by production of ¹⁴CO₂ was higher in the embryos labelled in vivo, whereas incorporation of radioactivity into TCA-soluble and TCA-insoluble fractions was surprisingly similar. It is difficult to obtain turnover times for alanine in the ovarian fluid because the volume of fluid and number of embryos are highly variable in different mother fish. Also, the ovarian fluid is not a static pool, but undergoes a rapid circulation and exchange of metabolites with the maternal organs (Korsgaard, 1983). Thus, under natural conditions, turnover of alanine in the ovary may be much faster, since the number of embryos per unit volume of ovarian fluid is higher in the ovary than in the in vitro incubation systems. The experiments performed in the present work with simultaneously injected tracer and carrier solutions indicate, however, that uptake mechanisms can operate efficiently at the natural concentration (óOmoll⁻¹) as well as at a higher concentration of carrier alanine.

In conclusion, the present data indicate that embryos from the blenny Z. viviparus have the ability in their early and late development to assimilate ambient L-alanine in vitro and in vivo. The capacity for uptake of L-alanine may increase during embryonic development in the ovary. The results show that label is distributed in the tissues of the embryos and that the labelled amino acid contributes to the general metabolism of the embryos in the ovarian cavity.

The author wishes to thank laboratory technician Jette Porsgaard for excellent technical assistance and the Danish Natural Science Research council for grants for the purchase of the HPLC system.