JEZ 867 THE JOURNAL OF EXPERIMENTAL ZOOLOGY 279:521–529 (1997) Mitochondrial Citrulline Synthesis in the Ureagenic Toadfish, Opsanus beta, Is Dependent on Carbonic Anhydrase Activity and Glutamine Transport RAYMOND P. HENRY1* AND PATRICK J. WALSH2 Department of Zoology and Wildlife Science, Auburn University, Auburn, Alabama 36849-5414 2 Division of Marine Biology and Fisheries, Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami, Florida 33149-1098 1 ABSTRACT Mitochondria isolated from the liver of the Gulf toadfish Opsanus beta produce scitrulline in the presence of 5 mM glutamine. Citrulline production was inversely related to succinate concentration between 0 and 10 mM, with a maximal rate being achieved at 0.1 mM. When toadfish are induced to become ureagenic by crowding-associated stress, mitochondrial citrulline production increases by approximately 10-fold. Citrulline synthesis is dependent on intramitochondrial carbonic anhydrase (CA) activity, being inhibited by about 75% by both acetazolamide and methazolamide. The addition of exogenous CA did not increase mitochondrial citrulline production. Anesthetizing the fish with MS 222 prior to mitochondrial isolation resulted in the near elimination of the capacity for citrulline synthesis. Mitochondria were also shown to possess an inducible glutamine transport system. The Vmax for glutamine uptake increased threefold and the Km increased four-fold in ureagenic vs. ammoniogenic toadfish. The transport system is the second labile component of the overall ornithine-urea cycle to be identified, and it provides a link between the production of glutamine via cytoplasmic glutamine synthetase and its consumption via mitochondrial carbamoyl phosphate synthetase III. J. Exp. Zool. 279:521 529, 1997. © 1997 Wiley-Liss, Inc. The Gulf toadfish, Opsanus beta, is one of the few teleost fish that has the capacity to become ureotelic on a facultative basis. Under conditions of minimal stress (i.e., individual fish isolated in large holding tanks) toadfish are primarily ammonotelic, excreting 61–80% of their total nitrogen as ammonia, depending on their nutritional status (starved vs. fed, Walsh and Milligan, ’95). However, when animals are either crowded together or confined individually in small, restricted chambers, their pattern of nitrogen excretion is altered such that urea becomes the predominant excretory product, accounting for 90% of the total nitrogen excreted (Walsh et al., ’94; Walsh and Milligan, ’95). Unlike ammonia, which is excreted in a continuous, uniform fashion, urea excretion in toadfish occurs in a periodic manner; most of the excreted urea is released in a single pulse of less than 3 h duration aproximately once every 24 h (Walsh et al., ’90; Wood et al., ’95). The pulse is released from the head region and is believed to originate from the gills (Wood et al., ’95). While the properties of urea excretion have been © 1997 WILEY-LISS, INC. extensively characterized in this species, much less is known about the mechanism, control, and potential lability of urea synthesis. Toadfish are known to possess a functional and complete hepatic ornithine-urea cycle with a glutamine-dependent carbamoyl phosphate synthesis (CPSase III; Mommsen and Walsh, ’89; Anderson and Walsh, ’95). The transition from ammonotelism to ureotelism involves the induction of the enzyme glutamine synthetase (GSase) in the liver which is complete by 24 h after crowding or confinement (Walsh et al., ’94; Hopkins et al., ’95). As a result, ammonia excretion drops to near zero as ammonia-nitrogen is channeled into the ornithine-urea cycle as glutamine (Walsh et al., ’90; Hopkins et al., ’95; Walsh and Milligan, ’95; Anderson and Walsh, ’95). Once the transition to ureotelism is complete, urea synthesis appears to occur at a constant rate with only urea excretion being pulsa- *Correspondence to: Raymond P. Henry, Dept. of Zoology and Wildlife Science, 101 Cary Hall, Auburn University, Auburn, AL 368495414. E-mail: email@example.com 522 R.P. HENRY AND P.J. WALSH tile (Wood et al., ’97). However, it is not known if other components of the urea pathway are altered as well during that transition in order to achieve higher rates of synthesis. One potential point of control could occur between GSase and CPSase III. There are both cytoplasmic and mitochondrial isozymes of GSase (Walsh, ’97a), and it is the cytoplasmic isozyme which is induced (Milligan and Walsh, ’95). Because CPSase III is mitochondrial, it has been speculated that one point of control could occur via the regulation of glutamine transport across the mitochondrial membrane (Walsh, ’97b). Very little is known about the intramitochondrial aspects of urea synthesis in toadfish as well, although it is clear that they are capable of citrulline synthesis under some circumstances (Anderson and Walsh, ’95). Studies on isolated hepatocytes from toadfish have documented significant levels of urea synthesis, showing that CPSase III is dependent on GSase for nitrogen and on carbonic anhydrase (CA) for HCO3– (Walsh et al., ’89). Inhibition of total hepatocyte CA activity by acetazolamide reduced urea synthesis by a maximum of about 70%, but since both the cytoplasm and mitochondria contain CA activity, it could not be determined if CA in one or both subcellular compartments was necessary for HCO3– supply to CPSase III. Some evidence from mammalian systems exists suggesting that CO2 produced endogenously within the mitochondria may not be sufficient to meet the metabolic carbon demand for urea synthesis. The rate of urea synthesis was reduced by about 60% in isolated guinea pig hepatocytes when external CO2 was reduced from 25 mM to zero (Dodgson and Forster, ’86). Furthermore, Tomera et al. (’82) reported that the rate of 14C fixation from NaH14CO3 into urea in perfused rat liver was much higher than the net rate of CO2 production. This suggests that cytoplasmic CO2 may be diffusing into the mitochondrial matrix to support urea synthesis. If so, cytoplasmic CA may be important in maintaining the transport of CO2 across the mitochondrial membrane (reviewed by Gros and Dodgson, ’88; Henry, ’96). This investigation reports on the mitochondrial component of urea synthesis in the toadfish liver. Citrulline synthesis was used as an indicator of mitochondrial nitrogen metabolism in order to elucidate the factors influencing the induction and regulation of urea synthesis. MATERIALS AND METHODS Collection and maintenance of animals Adult Gulf toadfish Opsanus beta (Goode and Bean) were collected with a roller trawl in Bis- cayne Bay, Florida, by commercial shrimpers during April and May, 1996, for the metabolic experiments, and between July and October, 1996, for the transport experiments. Fish were housed in 80 L aquaria with flowing aerated seawater (detailed procedures given by Walsh and Milligan, ’95). Individuals were fed shrimp but were starved for a period of 72 h before being used in an experiment. Fish were initially held under conditions of low population density (uncrowded). High population density (crowding) was used to initiate ureogenesis: fish were transferred to 6 L plastic tubs (30 cm long x 25 cm wide × 10 cm high) with flowing seawater such that there were four to six individuals per tub (Walsh et al., ’94). Fish were kept under these conditions for a minimum of 4 days before being used. Mitochondria preparation Fish were killed by a blow to the head either with or without having first been lightly anesthetized in MS 222 (tricaine methanesulfonate buffered with NaHCO3). The liver was dissected out, and blood vessels and connective tissue were cleaned away. Fish were used individually when large enough to yield 1.5 gm or more of liver; otherwise, livers for two to three fish were pooled. Mitochondria were prepared according to the method outlined by Anderson and Walsh (’95). Briefly, the liver was suspended in 5.5 volumes of homogenization buffer (250 mM sucrose, 0.5 mM EGTA, 5 mM K2HPO4, 30 mM HEPES, pH = 7.4 at 22°C), minced with scissors, and homogenized using a motor-driven teflon-glass homogenizer (four to six passes). The homogenate was centrifuged for 10 min at 750 × g (SS-34 rotor, Sorvall RC5-B superspeed) at 4°C. The supernatant was saved and centrifuged at 14,500 g for 10 min to produce the initial mitochondrial pellet. This was washed once in nine volumes of homogenization buffer and recentrifuged as above. The final pellet was resuspended in a volume of homogenization buffer plus bovine serum albumin (BSA, 5 mg ml–1), made fresh daily, equivalent to 0.3 ml g–1 of starting weight of liver. Carbonic anhydrase activity A 50 µl sample from each preparation was saved, frozen at –80°C, and shipped to Auburn on dry ice for analysis of CA activity. Samples were thawed on ice, sonicated, and measured using the electrometric delta pH assay (Henry, MITOCHONDRIAL CITRULLINE SYNTHESIS IN TOADFISH ’91). CA activity was reported on the basis of fresh weight of tissue. Citrulline assay The reaction medium for the analysis of mitochondrial citrulline synthesis was slightly modified from that used by Anderson and Walsh (’95; see below). This was made fresh daily and consisted of the following: 88 mM sucrose, 0.175 mM EGTA, 6 mM K2HPO4, 90 mM KCl, 5 mM NaHCO3, 2 mM MgCl2, 0.15 mM ATP, 10 mM ornithine, 0.1 mM succinate, and 36 mM HEPES, pH = 7.40 at 22°C. Seventy five microliters of the mitochondrial suspension were added to 500 µl of the reaction medium in a 1.5 ml microcentrifuge tube; this was allowed to equilibrate to room temperature (22°C) for 20 min. The reaction was started by adding 50 µl of a 50 mM glutamine (GLN) stock solution and gently mixing. After 30 min the reaction was stopped by the addition of 10 µl of 70% perchloric acid (PCA). The tubes were centrifuged for 30 sec at 10,000 g, and the supernatant was assayed for citrulline concentration. The citrulline assay reagent consisted of a 1:1 (v:v) mixture of antipyrene (2,3-dimethyl-1-phenyl-3pyrazolin-5-One, 200 mg in 94 ml H2O, 6 ml concentrated H2SO4, stored in a dark bottle at room temperature) and diacetyl monoxime (900 mg 2,3 butanedione monoxime in 95 ml H2O, 5 ml glacial acetic acid, stored refrigerated in a dark bottle) made fresh before use. One ml of this reagent was added to 50 µl of sample in a glass tube, mixed, loosely capped with a marble, and boiled in the dark for 30 min. The tubes were then cooled in an ice bath in the dark. Samples were transferred to semi-micro cuvettes, and absorbance was read at 464 nm (Perkin-Elmer Lambda 2). The increase in citrulline concentration was determined by comparison against a standard curve of 0 to 50 µM citrulline. All chemicals were purchased from Sigma (St. Louis, MO) and were reagent grade. Experimental protocol Citrulline synthesis was measured for mitochondria from toadfish that had been held either under uncrowded conditions for at least 7 days or crowded conditions for a minimum of 4 days. Because the initial measurements of the rates of citrulline synthesis were very low in both groups, we reinvestigated the assay conditions used previously (Anderson and Walsh, ’95). Both the linearity of citrulline accumulation over time and the rate of citrulline synthesis were tested against 523 varying concentrations of GLN and succinate. Furthermore, even after the optimal assay conditions were established, citrulline synthesis within a treatment group of fish was at times highly variable, with some individual preparations showing no apparent citrulline synthesis. In order to determine if this was a result of the anesthetic (MS 222) depressing one or more components of the citrulline pathway, synthesis was measured on mitochondria prepared from fish that were either killed after being anesthetized using MS 222 or killed immediately without anesthesia. For the remainder of the experiments, toadfish were killed without prior exposure to MS 222. Mitochondria from these preparations were used to determine if citrulline synthesis was dependent on mitochondrial and/or cytoplasmic CA activity. The CA inhibitors acetazolamide (Az) and methazolamide (Mz) were used to inhibit total CA activity (within the mitochondrial matrix and any residual CA activity in the general reaction medium). Stock solutions (10 mM) were made up in reaction medium; these were added to the final mitochondrial suspensions (also in reaction medium) in a 1:10 dilution to give a working concentration of 1 mM. The mitochondria were allowed to incubate at room temperature for 30 min, and the reaction was started by the addition of GLN. Another CA inhibitor, quaternary ammonium sulfanilamide (QAS), was used to selectively inhibit CA activity in the reaction medium, leaving CA within the matrix functional (Henry, ’87). A stock solutin of QAS in reaction medium (100 mM) was diluted 1:10 for a final working concentration of 10 mM; mitochondria were incubated for 30 min before GLN addition. In order to determine the potential importance of extra-mitochondrial (i.e., cytoplasmic) CO2 in citrulline synthesis, exogenous CA activity was added to the reaction medium. Purified bovine red cell CA (BCA) was added to the reaction medium to a final concentration of 0.1 mg ml–1. In a separate set of experiments, a series of HCO3– concentrations in the reaction medium (0, 5, and 25 mM) was used in order to determine the sensitivity of citrulline production to external CO2 concentrations. These measurements were carried out in the presence of 0.1 mM succinate with and without BCA, and in the absence of succinate with and without BCA. For each experiment, a blank consisting of the final mitochondrial suspension without the addition of GLN was carried out to control for preexisting levels of citrulline. 524 R.P. HENRY AND P.J. WALSH Mitochondrial glutamine transport Mitochondria were isolated as above with slight modification. Livers from four to five fish were pooled to give 3.5 to 5 g starting weight. The homogenization medium consisted of 250 mM sucrose, 0.5 mM EGTA and 30 mM HEPES, pH = 7.40 at 22°C. The washed mitochondrial pellet was suspended to 0.15 ml g–1 original starting weight using homogenization buffer plus 5 mg ml–1 BSA and kept on ice until use. The rapid mixing/rapid filtration method of Goldstein and Boylan (’78) was used to measure the rate of GLN uptake. In general, 50 µl of the mitochondrial suspension was drawn into a 100 µl Hamilton syringe followed by 50 µl of mitochondrial transport medium (99 mM sucrose, 0.175 mM EGTA, 90 mM KCl, 2 mM MgCl2, 36 mM HEPES, pH = 7.40 at 22°C plus the metabolic inhibitors 2 µg ml–1 rotenone and 1.5 µg ml–1 antimycin A). The contents were ejected into a 1.5 ml microcentrifuge tube and quickly withdrawn again to facilitate mixing of the mitochondria with the transport medium. The mixture was allowed to stand for 1 min prior to injection into the uptake vessel in order to facilitate the action of the metabolic inhibitors. A second syringe contained 100 µl of the mitochondrial transport medium plus 0.2 µCi of L-[14C(U)]-GLN (200 mCi mmol–1, NEC-451, DuPont NEN) and twice the appropriate final concentration of unlabeled GLN. The first experimental series involved a fixed concentration of 0.2 mM GLN and variable uptake times between 2.5 and 20 sec to determine the appropriate time to use for kinetic measurements. The second experimental series used variable final GLN concentrations between 0.2 and 5 mM and a fixed uptake time of 4 sec. For sucrose space determinations, 0.2 µCi of [14C(U)]-sucrose (4.5 mCi mmol–1, NEC-100, DuPont NEN) was substituted for GLN in the transport medium. Briefly, the procedure involved the very rapid and simultaneous injection of both syringes into a small mixing chamber, the bottom of which is fitted with a Millipore filter supported on a wire mesh (see Goldstein and Boylan, ’78 for details of construction of the apparatus). Immediately after mixing, vacuum was applied to the bottom of the chamber, trapping the mitochondria on the filter; the filter was then washed with 10 ml of isolation medium containing no substrate. The entire procedure was completed within 2.5 sec in our hands. The filter was then rapidly removed and placed in 10 ml Ecolume (ICN) and counted in a Tracor Analytic LSC. Uptake rates of GLN were calculated according to the equations of Goldstein and Boylan (’78) as follows: matrix GLN dpms = [GLNf] – [SUCROSEf/SUCROSEm] × [GLNm] (1) where GLNf = the total GLN dpms of the filter, SUCROSEf = the dpms on the filter in the sucrose incubations, SUCROSEm = the total sucrose dpms in the medium, and GLNm = the GLN dpms in the medium. And: nmol GLN matrix = matrix GLN dpms/GLN specific activity (2) where specific activity is expressed as dpms nmol–1. Uptakes were then divided by 4 sec and normalized on a per mg mitochondrial protein basis. Protein was determined on an aliquot of mitochondrial suspension by sonicating 1:1 in 50 mM HEPES with 0.1% Triton X-100 followed by assay according to the bicinchoninic acid method (Pierce Chemical Co., Rockfield, IL) of Smith et al. (’85) using BSA as a standard. All measurements were made in triplicate and averaged. RESULTS Optimal citrulline assay conditions Mitochondrial citrulline synthesis was inversely related to the concentration of succinate used in the reaction medium between values of 0.1 and 10 mM (Fig. 1). The peak rate of synthesis occurred at 0.1 mM, but the rates for 5 and 10 mM were only slightly above those measured in the complete absence of succinate. With 0.1 mM succinate, citrulline synthesis reached a maximum rate using 5 mM GLN (Fig. 2.). Under these assay conditions, mitochondria from uncrowded toadfish produced citrulline in a linear fashion but very slowly and only to a slight degree (Fig. 3A). There was no apparent difference in either the rate of accumulation or the final amount of citrulline regardless of the amount of succinate used (0, 0.1, or 10 mM). The actual rate of synthesis declined over time (Fig. 3B). For crowded toadfish, mitochondrial citrulline synthesis was virtually undetectable when either 0 or 10 mM succinate was used (Fig. 3C,D). For 0.1 mM succinate, citrulline production was linear over 25 min with a disproportionate increase occurring over the last 5 min of the assay (Fig. 3C). The linear portion of citrulline production occurred MITOCHONDRIAL CITRULLINE SYNTHESIS IN TOADFISH 525 g–1 min–1, n = 5) vs. crowded (29.4 ± 6.7 µmol g–1 min–1, n = 8) fish. These values were slightly lower but in the general range of values previously reported for toadfish liver mitochondria CA activity using very similar isolation and assay procedures (Walsh et al., ’89). Mitrochondrial citrulline synthesis Fig. 1. Citrulline synthesis from mitochondria isolated from crowded toadfish vs succinate concentration in the reaction medium. Points are the average of triplicate measurements. at a constant rate of synthesis of about 0.08 µmol gm–1 min–1 between 5 and 25 min (Fig. 3D); the rate actually increased by nearly 2 fold over the final 5 min of the assay. Mitochondrial CA activity Mitochondria from all preparations contained significant CA activity. Activity was not different in mitochondria from uncrowded (31.8 ± 7.8 µmol Mitochondria from crowded fish synthesized citrulline at a rate of about 0.12 µmol g–1 min–1; this was nearly 10-fold greater than the rate for mitochondria from uncrowded fish (Fig. 4). Both CA inhibitors, Az and Mz, reduced citrulline synthesis by approximately 75%; treatment with QAS resulted in a slight but statistically insignificant reduction in citrulline production (P > 0.05, student’s t-test; Fig. 4). The addition of BCA to the reaction medium had no effect on citrulline synthesis (Fig. 4). The use of 5 mM NaHCO3 appeared to produce maximal citrulline synthesis regardless of the presence or absence of exogenous CA. For mitochondria from crowded toadfish, citrulline synthesis was 0.04, 0.12, and 0.11 µmol g–1 min–1 for 0, 5, and 25 mM HCO3–, respectively, using 0.1 mM succinate and with no CA added to the reaction medium (n = 2–3). There was no apparent difference in citrulline synthesis for the same three HCO3– concentrations in the presence of exogenous CA (0.04, 0.13, and 0.08 µmol g–1 min–1, respectively, n = 2–3). Preliminary data indicate that, in the absence of succinate, HCO3– concentration can have an effect on citrulline synthesis: the rates were appoximately doubled at 5 mM NaHCO3 in the presence or absence of exogenous CA (0.03, 0.05, and 0.03 µmol g–1 min–1 for 0, 5, and 25 mM NaHCO3, respectively, n = 2). Interestingly, mitochondria from crowded fish exposed to MS 222 had significantly depressed rates of citrulline synthesis. MS 222 treatment resulted in an 80% inhibition of citrulline production (Fig. 4). Mitochondrial glutamine uptake Fig. 2. Citrulline synthesis from mitochondria isolated from crowded toadfish vs glutamine concentration in the reaction medium. All measurements made in the presence of 0.1 mM succinate. Points are the average of triplicate measurements. Mitochondrial GLN-derived radioactivity increased in a near linear fashion over time from 2.5 to 10 sec and then levelled off through 20 sec (Fig. 5), confirming 4 sec as a reproducible and short enough time to be on the most linear portion of the curve for the subsequent experiments. However, given the slight displacement of the 2.5 to 10 sec values from a direct line from the origin, there is possibly an even faster uptake component which was not detected by the current methodology. Uptake rates as a function of GLN concentration followed Michaelis-Menton kinetics, 526 R.P. HENRY AND P.J. WALSH Fig. 3. Citrulline production over time in mitochondria isolated from uncrowded (A) vs. crowded (C) toadfish, and the rate of citrulline synthesis in mitochondria from uncrowded (B) vs. crowded (D) toadfish. Open circles = 0 mM succinate; triangles = 0.1 mM succinate; squares = 10 mM succinate. All points are the average of triplicate measurements. so Lineweaver-Burke transformations were used to estimate Km and Vmax. Toadfish liver mitochondria Vmax values (Table 1) were at the low end of the range of those observed for rat kidney mitochondria (Goldstein and Boylan, ’78), and Km values were in the range reported for toadfish liver GLN content in vivo (Walsh and Milligan, ’95). Km increased three-fold and Vmax increased fourfold in crowded toadfish, indicating that circumstances that activate ureogenesis and ureotely also enhance mitochondrial transport capacity for the substrate glutamine. parent when the colorimetric assay was run under conditions (notably high succinate concentrations) used previously for a radioisotopic citrulline assay. Anderson and Walsh (’95), using the radioisotopic assay which measures incorporation of 14C-HCO3– into an acid-stable fraction, reported production of 0.35 µmol of citrulline over a 25 min incubation period for 0.75 mg mitochondrial protein, using 10 mM succinate. That amount was similar to values reported here using 10 mM succinate for both uncrowded and crowded toadfish (Fig. 3A,C). The use of 0.1 mM succinate (vs. 0 or 10 mM) did not alter either the accumulation of citrulline in uncrowded toadfish or its rate of synthesis. Also, it was not possible to discern a difference in citrulline synthesis between conditions in which succinate was present or absent from DISCUSSION A 10-fold increase in citrulline synthesis was observed in mitochondria from crowded vs uncrowded toadfish, but this difference was not ap- MITOCHONDRIAL CITRULLINE SYNTHESIS IN TOADFISH 527 TABLE 1. Kinetic constants for glutamine uptake by mitochondria isolated from livers of uncrowded or crowded toadfish1 Condition Uncrowded Crowded Km (mM) 7.21 ± 1.73 22.24 ± 5.84* Vmax (nmoles mg protein–1 sec–1) 0.200 ± 0.023 0.887 ± 0.271* 1 Mean ± SEM, n = 5, where each pooled preparation is one sample. *Significantly different from uncrowded (P < 0.05, student’s t-test). Fig. 4. Citrulline synthesis in isolated toadfish mitochondria. Bars represent the Mean ± SEM for the following conditions: UC = uncrowded control (n = 4); CC = crowded control (n = 10); CAz = crowded plus 1 mM acetazolamide (n = 5); CMz = crowded plus 1 mM methazolamide (n = 5); CCA = crowded plus 0.1 mg ml–1 bovine CA (n = 7); CQAS = crowded plus 10 mM QAS (n = 4); CMS222 = crowded fish treated with MS 222 before mitochondrial isolation (n = 2). the assay, indicating that mitochondrial citrulline production in uncrowded toadfish is minimal, even in the presence of high levels of GLN. Toadfish hepatocyte urea synthesis has been linked to an induction of cytoplasmic glutamine synthase (GSase; Walsh and Milligan, ’95; Hopkins et al., ’95); however, it appears that there is also another component of the ornithine-urea cycle that is induced during the transition to ureagenesis. This is indicated by the results with citrulline production in mitochondria from crowded toadfish. In the presence of 0.1 mM succinate, citrulline production was about 10-fold higher than Fig. 5. Plot of glutamine accumulation vs time in mitochondria isolated from liver of crowded toadfish. Values are mean ± SEM (n = 3). in uncrowded fish (Fig. 4). However, citrulline synthesis appeared to be inhibited by higher succinate concentrations, the rate for 10 mM succinate being no different from that in the complete absence of succinate (Fig. 3D). The inhibitory effect of succinate could be occurring either by a competitive inhibition of GLN transport into the mitochondria or by inhibiting one of the reactions in the citrulline pathway. Although the mechanism was not tested directly, the delay in the inhibitory effect (maximal inhibition takes about 15–20 min to occur, Fig. 3B) suggests that succinate is working on the reaction pathway via a slow accumulation within the intramitochondrial space. Furthermore, even though high rates of citrulline synthesis were observed using 0.1 mM succinate, the true optimal concentration for maximal citrulline synthesis in vivo may be even lower. Both the accumulation of citrulline and its rate of synthesis doubled over the last 5 min of the measurement (between 25 and 30 min; Fig. 3C,D), suggesting that citrulline synthesis increased once a certain amount of the available succinate was consumed and its in vivo concentration decreased below a critical point. This possibility deserves more systematic examination. This information suggests that the rate of citrulline production reported by Anderson and Walsh (’95) is very conservative. As such, actual in vivo rates of citrulline synthesis are probably well above those necessary to support both urea production in isolated hepatocytes and whole organism rates of urea excretion (Barber and Walsh, ’93). The radioisotopic assay may be sensitive enough to detect small differences in citrulline synthesis in the presence of 10 mM succinate, but the colorimetric assay is not. As proposed by Walsh (’97b), the stress-activation of urea production in toadfish liver appears to not only involve an increase in cytoplasmic glutamine production by GSase, but also a coordinated increase in the ability of the mitochondria to take up GLN so produced (Table 1). A transient increase in plasma levels of cortisol, dur- 528 R.P. HENRY AND P.J. WALSH ing the initial 24 h of crowding, has been shown to be necessary for the increase in GSase (Hopkins et al., ’95). Whether the induction of GLN transport is also under the influence of cortisol, and whether the induction of the enzyme and the transport protein are coordinated or independent events is not known at this time. It is informative to attempt to place the kinetic parameters for mitochondrial GLN uptake and citrulline synthesis in an in vivo context. First, the hepatic GLN content in toadfish rises from about 2.7 to about 5 mmol Kg–1 upon confinement (Walsh and Milligan, ’95). Interestingly, although the Km for mitochondrial uptake of GLN in uncrowded toadfish is above these in vivo concentrations, there is still an adaptive increase in both Km and Vmax of GLN transport, such that the transport system continues to have a great deal of reserve capacity for GLN, and would likely not be saturated in vivo. Indeed, the increase in transport capacity reported here is in close agreement with the three-to-four fold increase in cytoplasmic GSase (Walsh and Milligan, ’95). Notably when transport rates for GLN are converted to the units of µmol g–1 min–1 used for citrulline production (by multiplication of the values in Table 1 by a factor of 720), it is clear that transport capacity exceeds the rate of mitochondrial citrulline production by approximately five-fold. Furthermore, since the rate of mitochondrial citrulline production saturates at 5 mM GLN, the maximum concentration observed in vivo, these results would appear to rule out GLN transport as being rate-limiting to urea production. Anderson and Walsh (’95) concluded that urea production was limited by nitrogen availability, and the results presented here support their original proposal that the rates of urea synthesis are regulated by a balance between GLN production by cytoplasmic GSase and consumption by mitochondrial CPSase III. Our results also indicate that CPSase III is dependent on a second enzyme for substrate supply, namely intramitochondrial CA which supplies HCO3–. As was reported for urea synthesis in intact hepatocytes (Walsh et al., ’89), carbonic anhydrase (CA) is essential for normal rates of citrulline synthesis in isolated mitochondria. Both acetazolamide (Az) and methazolamide (Mz), CA inhibitors which permeate into the mitochondrial matrix, inhibit citrulline production by about 70– 80%. This is very similar to the degree of inhibition for citrulline synthesis in isolated guinea pig liver mitochondria (Dodgson et al., ’83), and for urea synthesis in isolated perfused rat and human liver (Haussinger and Gerok, ’85; Haussinger et al., ’86), and in isolated hepatocytes from the liver of guinea pigs (Dodgson and Forster, ’86) and toadfish (Walsh et al., ’89). QAS, a CA inhibitor that is excluded from the mitochondrial matrix, had no effect on citrulline synthesis. Furthermore, the addition of bovine CA to the reaction medium did not increase citrulline production. These results support the role of mitochondrial matrix CA in citrulline synthesis exclusive of any supporting function of cytoplasmic CA activity. It does not appear that mitochondrial synthesis is HCO3– limited even under conditions of peak urea production, and there is no evidence indicating that CO2 transport from the cytoplasm into the mitochondria is necessary to maintain peak citrulline synthesis. Citrulline synthesis appears to be dependent on intramitochondrial CO2 production and on the activity of matrix CA to maintain chemical equilibrium between CO2 and HCO3– in order to keep CPSase III supplied with substrate. Mitochondrial CA activity is not induced in crowded toadfish, indicating that the enzyme is there in excess, a condition that is considered almost universal for CA activity, regardless of its localization or function. Treatment of crowded toadfish with MS 222 prior to use in an experiment resulted in severely depressed mitochondrial citrulline synthesis. This could be a result of the anesthetic working at two possible levels of inhibition. First, MS 222 is known to directly inhibit CA activity (Christensen and Tucker, ’76); the compound has been shown to cause an approximate 20% inhibition of red cell CA activity in catfish. Given the very high total concentration of CA in catfish blood (Henry et al., ’88), 20% is a significant amount of inhibition. Since mitochondrial CA activity is much lower than that in red cells, it is quite possible that enough MS 222 could have permeated into the mitochondria to cause a great enough degree of enzyme inhibition to result in depressed citrulline synthesis. A second possibility could be that MS 222 could be interfering with mitrochondrial GLN uptake. MS 222, like other anesthetics, probably acts as a general membrane stabilizing agent, and this could have impacted on a number of transport events, including the uptake of GLN. Regardless of its mechanism of action, however, the effects are clear, and it becomes important to avoid the use of anesthetics such as MS 222 in studies involving the measurement of metabolic processes. MITOCHONDRIAL CITRULLINE SYNTHESIS IN TOADFISH ACKNOWLEDGMENTS This work was supported by NSF IBN 93-04844 to R.P.H. and to NSF IBN 95-07239 to P.J.W. We thank Jimbo Luznar for collecting the toadfish and John Paupe for technical assistance above and beyond the call of duty. LITERATURE CITED Anderson, P.M., and P.J. Walsh (1995) Subcellular localization and biochemical properties of the enzymes carbamoyl phosphate and urea synthesis in the batrachoidid fishes Opsanus beta, Opsanus tau, and Porichthys notatus. J. Exp. Biol., 198:755–766. Christensen, G.M., and J.H. Tucker (1976) Effects of selected water toxicants on the in vitro activity of fish carbonic anhydrase. Chem.-Biol. Interactions, 13:181–192. Dodgson, S.J., and R.E. Forster (1986) Carbonic anhydrase inhibition results in decreased urea production by hepatocytes. J. Appl. Physiol., 60:646–652. Dodgson, S.J., R.E. Forster II, D.A. Schwed, and B.T. Storey (1983) Contribution of matrix carbonic anhydrase to citrulline synthesis in isolated guinea pig liver mitochondria. J. Biol. Chem., 258:7696–7701. Goldstein, L., and J.M. Boylan (1978) Renal mitochondrial glutamine transport and metabolism: studies with a rapidmixing, rapid-filtration technique. Am. J. Physiol., 234: F514–F521. Gros, G., and S.J. Dodgson (1988) Velocity of CO2 exchange in muscle and liver. Ann. Rev. Physiol., 50:669–694. Haussinger, D., and W. Gerok (1985) Hepatic urea synthesis and pH regulation: role of CO2, HCO3–, pH and the activity of carbonic anhydrase. Eur. J. Biochem., 152:381–386. Haussinger, D., S. Kaiser, T. Stehle, and W. Gerok (1986) Liver carbonic anhydrase and urea synthesis: the effects of diuretics. Biochem. Pharmacol., 35:3317–3322. Henry, R.P. (1987) Quaternary ammonium sulfaniliamide: a membrane-impermeant carbonic anhydrase inhibitor. Am. J. Physiol., 252:R959–R965. Henry, R.P. (1991) Techniques for measuring carbonic anhydrase activity in vitro: The electrometric delta pH and pH stat methods. In: The Carbonic Anhydrases: Cellular Physiology and Molecular Genetics, S.J. Dodgson, R.E. Tashian, G. Gros and N.D. Carter, eds. Plenum, New York, pp. 119–125. Henry, R.P. (1996) Multiple roles of carbonic anhydrase in 529 cellular transport and metabolism. Ann. Rev. Physiol., 58:523–538. Henry, R.P., N.J. Smatresk, and J.N. Cameron (1988) The distribution of branchial carbonic anhydrase and the effects of gill and erythrocyte carbonic anhydrase inhibition in the channel catfish Ictalurus p[unctatus. J. Exp. Biol., 134:201–218. Hopkins, T.E., C.M. Wood, and P.J. Walsh (1995) Interactions of cortisol and nitrogen metabolism in the ureogenic Gulf toadfish Opsanus beta. J. Exp. Biol., 198:2229–2235. Mommsen, T.P., and P.J. Walsh (1989) Evolution of urea synthesis in vertebrates: the piscine connection. Science, 243:72–75. Smith, P.K., R.I. Krohn, G.T. Hermason, A.K. Mallia, and F.H. Gartner (1985) Measurement of protein using bicinchoninic acid. Anal. Biochem., 150:76–85. Tomera, J.F., P.G. Goetz, W.M. Rand, and H. Brunegraber (1982) Underestimation of metabolic rates owing to reincorporation of 14CO2 in the perfused rat liver. Biochem. J., 208:231–237. Walsh, P.J. (1997a) Purification and properties of hepatic glutamine synthase from the ureotelic Gulf toadfish, Opsanus beta. Comp. Biochem. Physiol., 115B:523–532. Walsh, P.J. (1997b) Evolution and regulation of urea synthesis and ureotely in (Batrachoidid) fishes. Ann. Rev. Physiol., 59:299–323. Walsh, P.J., and C.L. Milligan (1995) Effects of feeding and confinement on nitrogen metabolism and excretion in the Gulf toadfish Opsanus beta. J. Exp. Biol., 198:1559–1566. Walsh, P.J., J.J. Parent, and R.P. Henry (1989) Carbonic anhydrase supplies bicarbonate for urea synthesis in toadfish (Opsanus beta) hepatocytes. Physiol. Zool., 62:1257–1272. Walsh, P.J., E. Danulat, and T.P. Mommsen (1990) Variation in urea excretion in the gulf toadfish Opsanus beta. Mar. Biol., 106:323–328. Walsh, P.J., B.C. Tucker, and T.E. Hopkins (1994) Effects of confinement/crowding on uroegenesis in the Gulf toadfish Opsanus beta. J. Exp. Biol., 191:195–206. Wood, C.M., T.E. Hopkins, C. Hogstrand, and P.J. Walsh (1995) Pulsatile urea excretion in the ureagenic toadfish Opsanus beta: an analysis of rates and routes. J. Exp. Biol., 198:1729–1741. Wood, C.M., T.E. Hopkins, and P.J. Walsh (1997) Pulsatile urea excretion in the toadfish (Opsanus beta) is due to a pulsatile excretion mechanism, not a pulsatile production mechanism. J. Exp. Biol., 200:1039–1046.