Incorporation of 15N-ammonia into five nitrogenous components (Ammonia-N, Amide-N, Amino-N, Urea-N and Protein-N) in tissue was compared among the blood, brain, liver, gill, skin and muscle of the mudskipper, Periophthalmus modestus, following immersion in 15 mM 15N-ammonium sulfate (99.7 atom%) dissolved in diluted sea water for 24 to 168 hr. Total net 15N-uptake (μmol-15N/g wet mass) into tissue was greater in the order of the brain, liver, gill, muscle, skin and blood. Among the components in each tissue, amount of 15N in the form of Amino-N was the highest during the first 24 hr. Subsequently, that of Protein-N greatly increased, particularly in the liver, gill and brain. Amount of 15N in the form of Urea-N was negligible throughout the experimental period. In addition to having the highest ammonia content, the muscle showed the highest 15N-Concentrations (36–38 atom% excess) in Ammonia-N and Amide-N during the course of immersion. Amount of 15N in the form of Ammonia-N in the blood remained at the lowest levels among the tissues examined until 96 hr at which the nitrogen in the muscle reached equilibrium with 15N. It may be concluded that under ammonia loading conditions, the muscle plays the most important role not only in trapping a large amount of ammonia, but also in producing glutamine and other amino acids.
INTRODUCTION
Ammonia is a predominant nitrogen end product in most teleosts and can be readily excreted to the aquatic environment. In particular circumstances such as high ambient ammonia or aerial exposure, however, fish can hardly excrete ammonia, and toxic ammonia is accumulated in fish tissues. Most teleosts, except a very few fishes such as the toadifish [32, 36] and the Lake Magadi tilapia [32], have virtually no functional urea productive ability [8, 20, 26], so that it is very interesting to know how ammonotelic teleosts manage ammonia when they are faced with ammonia loading situations.
The mudskipper, Periophthalmus modestus, is a typical amphibious teleost and able to stay for a long period out of water. During aerial exposure, a large amount of ammonia as well as amino acids, but not urea is accumulated in its body, particularly in the muscle [22, 23, 29]. Furthermore, this fish shows a high tolerance to high ambient ammonia [23].
In mammals, the metabolic fates of exogenous ammonia in different tissues are well studied: in addition to hepatic route [13, 14], a large proportion of exogenous ammonia loaded is known to be taken up by muscle and utilized for the formation of glutamine which is the most important intertissue vehicle for nitrogen transport [17, 24, 33]. In contrast, ammonia is the main intertissue vehicle for nitrogen transport in teleosts [27, 39]. Therefore, it would be expected that ammonia administrated in a fish is widely distributed among tissues and incorporated into various nitrogen components.
In the present study, after the mudskipper was kept in 15 mM 15N-ammonium sulfate for 24 to 168 hr, we measured 15N-enrichment of five nitrogenous components (ammonia, amide nitrogen, free amino acids, urea and protein) in tissues using a mass spectrometer, and compared the changes in those distribution among the blood, brain, liver, gill, muscle, skin and whole body with the time course of immersion. In particular, we attempted to shed light on the role of fish muscle in managing ammonia under ammonia loading conditions.
Sensitivity of 15N detection is low as compared with a radioactive tracer, so that a large amount of 15N was necessary for a reproducible 15N measurement. As the mudskipper can tolerate a high concentration of ammonia, this fish, in spite of its small size, may be a good experimental animal for studying the metabolic fate of 15N-ammonia.
MATERIALS AND METHODS
Experimental animal
Mudskippers, Periophthalmus modestus (formerly P. cantonesis, see Murdy [30]) weighing 2.5–3.5 g were collected at the estuary of Shirahama in Wakayama Prefecture, Japan, and they were kept in 20% (7‰ salinity) sea water (SW) for a week before experimentation. During this period, the mixture of minced frozen krill and TetraMin (Tetra Werke, Germany) (2:1 w/w) was supplied daily. No attempt was made to separate the sexes.
Experimental protocol
After the rearing period, fish were divided into five groups. Fish in the first group (6 fish) were starved for 8 days in 20%SW (control). Fish in the second to fifth group were immersed in 15 mM (15NH4)2SO4 (99.7 atom%, Shoko Tsusho Co Japan) dissolved in 20% SW (pH adjusted to 7.5 with NaOH) for 24 (8 fish), 48 (5 fish), 96 (5 fish) and 168 hr (5 fish), respectively. Prior to immersion, fish in each group were starved for 7, 6, 4 and 1 day in 20%SW, to adjust the starvation period. 15N-ammonium sulfate solution was renewed every 24 hr, and temperature was held at 25 ± 1°C throughout the experiment. In disagreement with the previous report [23], two fish died during the course of immersion at 48 and 168 hr.
After completing the immersion, fish were rinsed with distilled water and blotted with wet guaze, and weighted. The whole blood samples were taken into heparinized hematocrit capillary tubes by causdal peduncle severance, weighed and homogenized in cold 80% ethanol using a sonicator (Ultrasonics, England). The brain, liver, gill, skin, dorsolateral muscle and carcass (all of the remaining tissue) were removed, weighed, and immediately frozen on an aluminum block cooled at −40°C. Frozen tissues were stored at −70°C until extraction was made.
Frozen tissues were homogenized in cold 80% ethanol with mortar and pestle (used for soft tissues) or Ulta-Turrax homogenizer (Janke and Kunkel Gmbh & Co. Germany), and centrifuged at 4,000 rpm. Pellets (residue) were washed twice with cold 80% ethanol. Original supernatant and the wash were pooled, and then stored at − 20°C. Because of small sample sizes, the whole blood, brain, liver, gills and skin from 4 to 8 individuals were pooled. To remove lipid, the residue from each tissue was washed with 100% ethanol, the mixture of ethanol and ether (1:1), and ether in the order and then dried. After weighing, 4–8 mg (dry weight) of each residue was hydrolyzed with 6 N HCl at 105°C for 24 hr in a sealed glass tube, and then dried by bubbling nitrogen at 50°C for later determination of 15N-enrichment in protein.
The 80% ethanol supernatant was evaporated to dryness in vacuo under acidic conditions (50 μl 1 M HCl was added to each 10 ml supernatant). The dried sample was dissolved with a small amount of cold water (1 to 5 ml) using the sonicator, and then centrifuged at 12,000 g at 2°C for 30 min. The resultant supernatant (non-protein fraction) was stored at −20°C for later determination of 15N-enrichment.
Determination of ammonia, amino acids and urea contents
The total ammonia content (NH3 + NH4+) in the non-protein fraction was determined by Ion-chromatography (HIC-6A, Shimadzu Co. Japan) with a Shim-pack IC-Cl column. Asparagine, glutamine and glutamic acid contents were measured using Shimadzu Amino acid Analysis system (ISC-07/S1504 Li type column). Urea content was spectrophtometrically measured by the method of Ceriotti and Spandrio [9].
Isolation of nitrogenous components
Free ammonia was isolated by a modification of the micro-diffusion method described by Campbell et al. [7]: the ammonia diffusion vial used in the present study was composed of a glass vial (25 ml) and a rubber stopper from which a slip of glass fiber filter (Whatman CF/C) wetted with 1% H2SO4 was suspended. The glass fiber filter used for absorbing ammonia was previously baked at 450°C for 24 hr. The nitrogen in the amide group of glutamine plus asparagine (Amide-N) in the non-protein fraction was hydrolyzed by a modified method of Buttery et al. [5]. No attempt was made to separate the amino acid and protein fractions into individual amino acid.
Ammonia fraction (Ammonia-N): to 0.5 ml of the filtrate of the non-protein fraction in the glass vial, 0.1 ml of 0.1 M borate buffer, pH 10.4 (0.1 M Na2B4O7 adjusted pH with NaOH) was added. Immediately after the addition of borate buffer, the vial was tightly capped with the rubber stopper as mentioned above, and left for 20–24 hr at room temperature (20–25°C). Liberated ammonia was absorbed on the slip of glass fiber filter, and then the glass fiber filter was dried in a desiccator at room temperature.
Amide-N fraction (Amide-N): after removing Ammonia-N from the filtrate, 0.1 ml of 10% H2SO4 was added into the vial, and the vial covered with aluminum foil was heated at 100°C in a boiling water bath for 15 min to hydrolyze the amide nitrogen. After cooling, 0.2 ml of 2 M Na2CO3 was added into the vial, and liberated ammonia was absorbed on the glass fiber filter as described above.
Free amino acid fraction (Amino-N): after removal of Ammonia-N and Amide-N as described above, the filtrate was adjusted to pH 2–4 with HCl. It was placed on an ion-exchange column (10 × 120 mm; Dowex 50W-X8, H+ type, 100–200 mesh) and washed with deionized water. Amino acids retained on the column was eluted with 2 M NH4OH. After addition of 0.1 ml 1 M NaOH, the effluent from the column was evaporated to dryness with a vacuum evaporator at 50°C, and dissolved in a small amount of deionized water, and evaporated again, in order to remove trace ammonia. The final dried effluent was dissolved in a small amount of deionized water and absorbed on the pre-baked glass fiber filter, and then dried in a desiccator.
Urea fraction (Urea-N): after removal of Ammonia-N and Amide-N (in this case, 2 M Na2CO3 was used for liberating free ammonia), the filtrate was neutralized with HCl, and 0.2 ml 0.2 M phosphate buffer (pH 7.2) and 7.5 units/ml of urease (Sigma, Type IX) were added. Then, they were incubated at 37°C for 1 hr, and the liberated ammonia was trapped on the glass fiber filter as described above. Because of limitations of amount of filtrates, the filtrates from carcass after 24 hr of immersion were used for isolation of Urea-N.
Protein fraction (Protein-N): to exclude any contamination of free 15N-ammonia into the protein fraction, 0.5 ml of the hydrolysate dissolved in deionized water was put on the ion-exchange column, and eluted with 2 M NH4OH, and then absorbed on the glass fiber filter as in Amino-N mentioned above.
Detection of isotope ratio in each fraction
Measurements of 15N-Concentration in each fraction were made by a quadrupole type mass spectrometer equipped with a CN analyzer (Anerva, TE-360B, Japan). The total nitrogen contents in Amino-N and Protein-N were also determined using this apparatus.
Recovery test
Using 0.5 ml of known concentrations (10–14 mM) of ammonium sulfate, glutamine and glutamic acid, the recovery efficiency for Ammonia-N and Amide-N of the present micro-diffusion method was examined. Ammonia liberated and absorbed on the glass fiber filter was dissolved again in a definite volume of deionized water, and then ammonia content was determined as described above. The resultant recovery percent of Ammonia-N was virtually 100%: 98.9% ± 0.2 (Mean ± SE, n = 6). However, about 3% of the amide nitrogen in glutamine was hydrolyzed during the isolation procedure of Ammonia-N, so that the recovery percent of Amide-N was 96.7% ± 0.5 (n = 6). On the other hand, no hydrolysis occurred in α-amino nitrogen in glutamic acid during either the isolation procedure.
Statistical analysis
The data for muscle and whole body (Σ tissue values × weight in each tissue sample) are presented as mean ± standard error (SE). Significance of difference between means was assessed by Duncan's multiple range test [15], when appropriate. P < 0.05 was considered to represent a significant difference.
RESULTS
Incorporation of 15N-ammonia into Ammonia-N, Amide-N and Amino-N
Blood 15N-Concentration (atom% excess) in Ammonia-N in the whole blood showed the lowest values among the tissues examined during the first 48 hr after immersion (Fig. 1A), whereas ammonia content showed a peak at 24 hr (Fig. 1B). At 96 hr, 15N-Concentration in Ammonia-N reached at upper plateau level. Similar trends were also found in 15N-Concentrations in Amide-N and Amino-N (Fig. 1A), although the values for Amino-N fluctuated. Both glutamine and glutamic acid contents increasd with time, while asparagine content (0.14 μmol/g wet mass) hardly changed during the course of immersion (Fig. 1B).
Fig. 1
(A): 15N-Concentrations in Ammonia-N (open circles), Amide-N (open tringles) and Amino-N (open squares) fractions and (B): Ammonia (solid circles), glutamine (solid triangles) and glutamic acid (solid squares) contents in the blood (pooled sample) following immersion 15N-Ammonium sulfate.
Brain As in the blood, ammonia content in the brain showed a peak at 24 hr, and then decreased to a lower steady level (Fig. 2B). The reverse was true for changes in glutamic acid content. Glutamine content increased up to 18-fold that of control at 96 hr, and then decreased at 168 hr (Fig. 2B). Asparagine content was low (0.4 μmol/g wet mass) throughout the experimental period. 15N-Concentration in Ammonia-N remained at a lower stable level after showing a peak at 24 hr, while 15N-Concentrations in Amide-N and Amino-N increased with time and reached upper plateau levels at 96 hr (Fig. 2A). It should be noted that of the tissues examined, only the brain had higher 15N-Concentrations in Amide-N and Amino-N than in Ammonia-N.
Fig. 2
(A): 15N-Concentrations in Ammonia-N, Amide-N and Amino-N fractions and (B): Ammonia, glutamine and glutamic acid contents in the brain (pooled sample) following immersion in 15N-ammonium sulfate. Symbols are the same as in Fig. 1.
Gill Ammonia content in the gill showed a peak at 24 hr as in the blood and brain (Fig. 3B). 15N-Concentration in Ammonia-N greatly increased following immersion and reached an upper plateau level at 96 hr (Fig. 3A), as in the blood. Similarly, 15N-Concentrations in Amide-N and Amino-N increased with increases in glutamine and glutamic acid levels which rose by nearly 10-fold those of control at 96 hr. Liver Ammonia content in the liver did not change greatly during the first 48 hr (Fig. 4B), while 15N-Concentration in Ammonia-N greatly increased during this period (Fig. 4A). On the other hand, 15N-Concentrations in Amide-N and Amino-N increased with increases in glutamine and glutamic acid contents. Following the brain, the liver had the highest 15N-Concentration in Amino-N (Fig. 4A). Asparagine content was quite low (0.1 μmol/g wet mass) throughout the expeimental period.
Fig. 3
(A): 15N-Concentrations in Ammonia-N, Amide-N and Amino-N fractions and (B): Ammonia, glutamine and glutamic acid contents in the gill (pooled sample) following immersion in 15N-ammonium sulfate. Symbols are the same as in Fig. 1.
Fig. 4
(A): 15N-Concentrations in Ammonia-N, Amide-N and Amino-N fractions and (B): Ammonia, glutamine and glutamic acid contents in the liver (pooled sample) following immersion in 15N-ammonium sulfate. Symbols are the same as in Fig. 1.
Muscle, Skin and Whole body As shown in Figure 5B, ammonia content in the muscle increased sharply during the first 48 hr and reached a plateau level (no significant difference among the values at 48, 96 and 168 hr), nearly external ammonia content. Glutamine content increased by 8-fold that of control at 96 hr and then decreased slightly at 168 hr (significant difference between 96 and 168 hr; P < 0.01), whereas asparagine levels (ca. 1.1 μmol/g wet mass) hardly changed after immersion (Fig. 5B). In the whole body, ammonia and glutamine contents showed similar changes to those found in the muscle, although the levels were 70 to 80% of those of the muscle (Fig. 7B). It should be noted that except for 96 hr, the ratios of glutamine to ammonia contents in both the muscle and whole body were nearly constant values (0.44 and 0.59 for the muscle and whole body, respectively) throughout the experimental period, suggesting that mass action is important for glutamine synthesis. In both the muscle (Fig. 5B) and whole body (Fig. 7B), glutamic acid contents except for the first 48 hr increased 3- to 5-fold those of control (significant difference; P < 0.05). The total nitrogen contents in Amino-N in each tissue, however, did not change significantly with time. In Table 1, the mean values of the total nitrogen contents in Amino-N throughout the experiment are listed.
Fig. 5
(A): 15N-Concentrations in Ammonia-N, Amide-N and Amino-N fractions and (B): Ammonia, glutamine and glutamic acid contents in the muscle following immersion in 15N-ammonium sulfate. Symbols are the same as in Fig. 1. #indicates a significant difference in 15N-Concentrations between Ammonia-N and Amide-N. *indicates a significant difference from the control value. Numbers and bars in the figures indicate the sample size and standard error, respectively.
Table 1
Total nitrogen contents in Amino-N and Protein-N fractions. Each value represents the mean contents through-out the experimental period
15N-Concentration in Ammonia-N in the muscle greatly increased and reached an upper plateau level during the first 48 hr (no significant difference among the values at 48, 96 and 168 hr) (Fig. 5A). On the other hand, 15N-Concentration in Amide-N in the muscle reached an upper plateau level at 96 hr (Fig. 5A): 15N-Concentration in Amide-N at 96 hr was significantly different from that at 48 hr P < 0.01), but not at 168 hr; only the 15N-Concentration in Amide-N at 48 hr was significantly lower than that in Ammonia-N (P < 0.01). It should be noted that 15N-Concentrations in both Ammonia-N and Amide-N in the muscle after 24 hr showed the highest values among the tissues examined. 15N-Concentration in Amino-N in the muscle was low as compared with Amide-N, but steadily increased until 96 hr, and then slightly decreased (significant difference between 96 and 168 hr; P < 0.01) (Fig. 5A).
In the skin, the changes in 15N-Concentrations in Ammonia-N and Amino-N and their contents were similar to those found in the muscle (Fig. 6A, B). However, 15N-Concentration in Amide-N and glutamine content in the skin were much lower than those in the muscle.
Fig. 6
(A): 15N-Concentrations in Ammonia-N, Amide-N and Amino-N fractions and (B): Ammonia, glutamine and glutamic acid contents in the skin (pooled sample) following immersion in 15N-ammonium sulfate. Symbols are the same as in Fig. 1.
Fig. 7
(A): 15N-Concentrations in Ammonia-N, Amide-N and Amino-N fractions and (B): Ammonia, glutamine and glutamic acid contents in the whole body following immersion in 15N-ammonium sulfate. Symbols are the same as in Fig. 5.
In the whole body, the changes in 15N-Concentrations in Ammonia-N and Amide-N were essentially similar to those in the muscle (Fig. 7A), however, in addition to 48 hr, 15N-Concentration in Amide-N at 24 hr was significantly lower than that of Ammonia-N (P < 0.01). 15N-Concentration in Amino-N in the whole body increased with time and reached the maximum level at 96 hr (no significant difference between 96 and 168 hr).
Incorporation of 15N-ammonia into Urea-N
15N-Concentration in Urea-N was very low throughout the experimental period, although the value at 96 hr increased by 2.6-fold that at 48 hr (significant difference between 96 and 48 hr; P < 0.05, but not 168 hr; Table 2). Furthermore, urea content hardly changed with the time course of immersion (no significant difference with time).
Table 2
15N-Concentration in Urea-N fraction and urea content in the carcass following immersion in 15N-ammonium sulfate solution
Incorporation of 15N-ammonia into Protein-N
At the first 24 hr after immersion, atom percent in Protein-N in the skin, muscle and carcass were not statistically different from the 15N natural abundance. However, 15N-enrichment in Protein-N in the liver, gill and brain exceeded by far the natural abundance within 24 hr, and then those levels greatly increased with time (Fig. 8). At 48 hr, 15N-enrichment in Protein-N in the muscle, skin and carcass significantly exceeded the natural abundance (P < 0.01), although the value in the muscle was the lowest among the tissues examined. The total nitrogen contents in Protein-N in each tissue did not change significantly with time, so that the mean values were listed in Table 1.
Distribution of 15N-fraction within tissue
Using the data of 15N-Concentration in each fraction and its contents (Figs. 1–8, Tables 1–2), the net 15N-uptake (μmol-15N/g wet mass; 15N-Concentration × Nitrogen content in each fraction) in each tissue in terms of each fraction was calculated (Fig. 9).
Fig. 9
Net 15N-uptake (μmol-15N/g wet mass) in tissues (A: blood, B: brain, C: liver, D: skin, E: muscle and F: whole body) in terms of each fraction (Ammonia-N, Amide-N, Amino-N and Protein-N) with the time course of immersion in 15N-ammonium sulfate. Bars indicate standard error.
Blood The net 15N-uptake in Ammonia-N (0.25 μmol-15N/g wet mass) remained at the lowest levels among the tissues examined during the first 48 hr. Subsequently, those levels showed 3- to 4-fold increases at 96 and 168 hr (Fig. 9A). The net 15N-uptake in Amino-N, as in other tissues, comprised a major part (40–60%) of the total net 15N-uptake (Ammonia-N + Amide-N + Amino-N + Protein-N) in the blood, although the value at 96 hr suddenly decreasd. In contrast, the net 15N-uptake in Amide-N showed the lowest among the tissues examined. The net 15N-uptake in Protein-N occupied rather constant portions (30–40%) of the total net 15N-uptake throughout the experimental period.
Brain A relative amount of 15N in the form of Ammonia-N in the brain after the first 48 hr was extremely low, which showed only 0.6% of the total net 15N-uptake (Fig. 9B). On the contrary, the net 15N-uptake in Amino-N, Protein-N and Amide-N occupied large parts, which accounted for 50–60, 30–40 and 10–15% of the total net 15N-uptake, respectively. The net 15N-uptake in the non-protein fraction (Ammonia-N + Amide-N + Amino-N) reached a plateau level around 96 hr. The brain showed the highest total net 15N-uptake per unit of tissue mass among the tissues examined.
Liver The net 15N-uptake in Ammonia-N and Amide-N accounted for only 3–9% of the total net 15N-uptake in the liver (Fig. 9C). On the contrary, the net 15N-uptake in Amino-N, particularly in Protein-N greatly increased with time and became occupied 21 and 70% of the total net 15N-uptake at 168 hr, respectively. Similar trends were also seen in gill (data not shown).
Muscle, Skin and Whole body In the muscle, the net 15N-uptake in the non-protein fraction accounted for 86–100% of the total net 15N-uptake during 96 hr (Fig. 9E). Of the non-protein fraction, the net 15N-uptake in Ammonia-N and Amide-N occupied relatively constant portions (Ammonia-N: 33%; Amide-N: 16%) of the total net 15N-uptake in the muscle throughout the experimental period. The net 15N-uptake in Protein-N accounted for 14–21% of the total net 15N-uptake after the first 24 hr, which were the lowest among the tissues examined (Fig. 9E). The total net 15N-uptake in the muscle reached the maximum level (equilibrium state) at 96 hr (no significant difference between 96 and 168 hr), while the time at which the net 15N-uptake in Ammonia-N attained to a plateau was earlier than those of the other fractions.
In the skin, the changes in the net 15N-uptake in Ammonia-N, Amino-N and Protein-N were similar to those of the muscle, although the net 15N-uptake in Amide-N was very low (2.5%) (Fig. 9D).
In the whole body, the net 15N-uptake in Protein-N showed a 7% of the total net 15N-uptake at 24 hr and then increased up to 47% at 168 hr (Fig. 9F). The net 15N-uptake in the nitrogen other than Ammonia-N accounted for 75–83% of the total net 15N-uptake during the course of immersion. However, the net 15N-uptake in Urea-N was extremely low, which accounted for only 0.1–0.2% of the total net 15N-uptake. The total net 15N-uptake in the whole body reached an equilibrium state at 96 hr (no significant difference between 96 and 168 hr). As shown in Figure 9F, the total net 15N-uptake into the whole body during 96 hr after immersion increased linearly with time at the rate of 0.17 (μmol-15N/hr/g wet mass).
DISCUSSION
In agreement with the previous reports [22, 23, 29], urea content in the mudskipper (P. modestus) hardly changed under ammonia loading conditions. It is known that the first step of the nitrogen entry into the ornithine-urea cycle in teleostean fishes in mediated by CPS III (carbamoyl phosphate synthetase) requiring glutamine as a nitrogen donor [26]. 15N-Concentration in Amide-N in the liver after immersion reached as much as 21 atom% excess (Fig. 4A). Nevertheless, 15N-Concentration in Urea-N in the carcass (an approximation of whole body) showed only 1.4 atom% excess or less (Table 2), although Urea-N in the liver might have had a higher 15N-Concentration. It may be concluded that the ornithine-urea cycle in the mudskipper (if present) bears little ammonia detexifying function.
It is well known that the formation of glutamine is the most efficient way of ammonia detoxification in vertebrate brains. During 96 hr of immersion, the mudskipper brain responded with 18-fold increase in glutamine content (Fig. 2B). Except for the first 24 hr, the brain was the sole tissue having higher 15N-Concentrations in Amide-N than in Ammonia-N (Fig. 2A). As in other fishes [2, 25], glutamic acid content in the mudskipper brain before immersion was higher than glutamine. However, following immersion for 24 hr, glutamic acid content decreased, which was accompanied with the rise of ammonia and glutamine contents as well as their 15N-Concentrations, indicating that at the early period of immersion, a glutamine synthetic rate was far greater than that for glutamic acid. However, the activities of glutamate dehydrogenase and glutamine synthetase in the mudskipper brain were about 10- and 2-fold those of water breathing gobiid fishes, respectively [23]. A similar marked fall of glutamic acid level was observed for rainbow trout brain exposed to a high ambient ammonia for 24 hr [2] and for mammalian brain [19]. In mammalian brain after a short-term infusion of labeled ammonia, the nitrogen of the amide group in glutamine is known to be more heavily labeled than that of α-amino group and glutamic acid [4, 12]. It is noteworthy that the brain showed the highest weight specific net 15N-uptake, and those values were much higher in the form of Amino-N and Protein-N than in Amide-N (Fig. 9B), in spite of a great increase in glutamine content. These results indicate that exogenous ammonia trapped by a powerful glutamate dehydrogenase and glutamine synthetase system was rapidly transformed into the nitrogen in various amino acids and proteins.
In the mudskipper liver, glutamine contents after immersion were relatively low (Fig. 4B), which were quantitatively similar to those reported by Levi et al. [25] and Arillo et al. [2] in the livers of ammonia treated goldfish and rainbow trout, respectively. Furthermore, the net 15N-uptake in Amide-N was low throughout the experimental period (Fig. 9C). From these findings, it seems unlikely that an ammonotelic teleost liver play an important role in glutamine synthesis. Similarly, in the liver of an ammonotelic toad, Xenopus laevis exposed to a desiccated stress, injected 15N-ammonia was heavily incorporated into aspartic acid, whereas glutamine was not well labeled [35].
It is known that both NH3 and NH4+ are permeable across fish cell membranes, and among tissues, muscle has a great capacity to store ammonia [40, 41]. In addition, NH4+ is more permeable in the seawater adapted fish membranes than in fresh water ones [38]. In the previous results [23] based on the non-labeled substances, it was found that ammonia located in the muscle occupied 78% of that in the whole body under ammonia loading conditions. Similarly, the distribution of 15N in the form of Ammonia-N, Amide-N and Amino-N in the muscle at 96 hr accounted for 84, 72 and 65% of those in the whole body, respectively, while those in the blood were only 0.5, 0.2 and 1.5%, respectively. Moreover, the net 15N-uptake as well as 15N-Concentration in Ammonia-N in the blood remained at the lowest levels among the tissues examined until 96 hr at which those levels increased by nearly 3-fold (Figs. 1A and 9A). On the other hand, the total net 15N-uptake in the muscle increased linearly with time, and reached an equilibrium state at 96 hr (Fig. 9E). These findings suggest that the muscle efficiently trapped 15N-Ammonia in the blood until the nitrogen in itself was equilibrated with 15N.
In mammals, particularly in cirrhosis patients, it is well documented that skeletal muscle is highly efficient in trapping ammonia and produces glutamine and other amino acids [16–18, 24, 33]. However, in fish, it is still unclear whether or not muscle is the main site of glutamine production. In the present study, 15N-Concentration in Amide-N in the muscle was by far the highest among the tissues examined. Glutamic acid content in the muscle was low, but tended to increase with time, in the face of a large increase in glutamine content as well as the net 15N-uptake in Amide-N and Amino-N (Figs. 5B and 9E), suggesting that glutamic acid was continuously supplied for glutamine synthesis. These results confirm the earlier conclusions [23] that the muscle is able to store a great amount of ammonia and produce glutamine and other amino acids, which play the major role in lowering ammonia levels in the remaining tissues, particularly in the blood.
In mammals, it is known that the rate of glutamine and other amino acid releases from muscle closely approximates net amino acid synthesis in its tissue, so that there is little change in tissue amino acid levels even under ammonia loading conditions [18, 24, 33]. On the contrary, in aggreement with the previous reports [22, 23], the mudskipper muscle accumulated a large amount of the amide and amino nitrogen. The reason for this discrepancy between mammalian and fish muscles is unclear; fish muscle may have a special capacity to accumulate amino acids and other related compound, since fish muscle, particularly from euryhaline species is known to have large free amino acid pools which serve for cell volume regulation [1, 3, 11, 21].
Mudskipper has a unique skin in relation to amphibious mode of life in that there are numerous blood capillaries distributed in the outermost layer of epidermis and several strata of enormous cells (20–30 μm in length) with large vacuole located in the middle of epidermis [34, 42]. Furthermore, the mudskipper can excrete a considerable amount of ammonia and urea through the skin [28]. Following the muscle, the mudskipper skin stored the largest amount of 15N in the form of Ammonia-N and Amino-N, which accounted for 7 and 13% of those of the whole body at 96 hr, respecively. The accumulation of these substances in the skin seems to play an important role in ameliorating an ammonia load in the other tissues, as it does in muscle.
Several investigations [6, 10, 37] showed that amount of ammonia excretion after a fish was exposed to ammonia loading conditions was far less than would be expected on the calculated ammonia accumulation in fish body. Hitherto, the question of this missing ammonia has been largely unexplored. The present study clearly revealed that exogenous ammonia entered in the blood was transferred into intracellular compartments and converted mainly into the nitrogen in various amino acids and proteins.
Acknowledgments
The authors would like to thank Mss. Kazuko Mizoguchi and Yuko Ikebe and Mr. Hiroyuki Hayashi of Faculty of Education, Wakayama University for their invaluable assistance. We are also grateful to Dr. Tali A. Conine, University of British Columbia for improving the manuscript.
