In osmoregulating teleost fish, urea is a minor nitrogen excretory product, whereas in osmoconforming marine elasmobranchs it serves as the major tissue organic solute and is retained at relatively high concentrations (∼400 mmol/l). We tested the hypothesis that urea transport across liver mitochondria is carrier mediated in both teleost and elasmobranch fishes. Intact liver mitochondria in rainbow trout (Oncorhynchus mykiss) demonstrated two components of urea uptake, a linear component at high concentrations and a phloretin-sensitive saturable component [Michaelis constant (Km) = 0.58 mmol/l; maximal velocity (Vmax) = 0.12 μmol·h−1·mg protein−1] at lower urea concentrations (<5 mmol/l). Similarly, analysis of urea uptake in mitochondria from the little skate (Raja erinacea) revealed a phloretin-sensitive saturable transport (Km = 0.34 mmol/l; Vmax = 0.054 μmol·h−1·mg protein−1) at low urea concentrations (<5 mmol/l). Surprisingly, urea transport in skate, but not trout, was sensitive to a variety of classic ionophores and respiration inhibitors, suggesting cation sensitivity. Hence, urea transport was measured in the reverse direction using submitochondrial particles in skate. Transport kinetics, inhibitor response, and pH sensitivity were very similar in skate submitochondrial particle submitochondrial particles (Km = 0.65 mmol/l, Vmax = 0.058 μmol·h−1·mg protein−1) relative to intact mitochondria. We conclude that urea influx and efflux in skate mitochondria is dependent, in part, on a bidirectional proton-sensitive mechanism similar to bacterial urea transporters and reminiscent of their ancestral origins. Rapid equilibration of urea across the mitochondrial membrane may be vital for cell osmoregulation (elasmobranch) or nitrogen waste excretion (teleost).
early classical studies by Homer Smith demonstrated that marine elasmobranchs retain urea in their tissues and body fluids at relatively high concentrations (∼400 mmol/l) to maintain their body fluids at slightly higher osmolalities than that of the surrounding seawater (65). This is achieved in part by low branchial urea permeability (14, 28, 33, 55) and reabsorptive mechanisms present in the kidney (13, 41, 66, 81). Regardless, some urea is lost to the environment; therefore, elasmobranchs continuously synthesize urea at high rates via the ornithine-urea cycle (O-UC) primarily in the liver, but also in skeletal muscle tissue (40, 67). Although the O-UC has been retained in teleost fishes, the majority of adult teleosts are ammoniotelic and synthesize urea at low rates via uric acid or arginine catabolism (22, 69, 84) thus maintaining tissue urea at relatively low levels (<10 mmol/l; see Ref. 71).
In piscine species, arginase is the key O-UC enzyme responsible for the production of urea from arginine. Although it is a ubiquitous enzyme, the mitochondria of the liver and kidney has the highest level of activity (4, 18, 23, 27, 43, 51, 58, 87). In contrast, in mammals, the urea precursor, arginine, is shuttled out of the mitochondria to interact with a cytosolic form of arginase (reviewed in Ref. 3). The fundamental differences in subcellular arginase localization suggest that each vertebrate class may require different mechanisms to regulate the movement of urea across the mitochondrial membrane. In fish, urea-produced arginase would be shuttled out of the mitochondria, whereas, in mammals, urea is produced in the cytosol. Although the outer mitochondrial membrane is freely permeable to molecules up to 1.5 kDa because of the presence of large pores formed by the voltage-dependent anion channel (VDAC; 21), it is the inner membrane that may pose a challenge in fish, since this boundary acts as a major barrier to solutes and water moving between the cytoplasm and mitochondrial matrix. In elasmobranchs, the rate of urea synthesis and resulting urea gradient may be sufficiently high for rapid equilibration of urea across the mitochondrial membrane (9, 10, 77). This may be critical for osmotic balance and the avoidance of perturbations on enzyme activity and mitochondrial function. However, in teleosts, the rate of hepatic urea synthesis from argininolysis is considerably lower than elasmobranchs, and rapid urea equilibration may not be of great importance.
Urea permeability of cell membranes is partly dependent on carrier-mediated processes, either specific urea transport proteins or other proteins that transport small-molecular-weight solutes or water [e.g., aquaporins (AQPs)]. Over the past two decades, bidirectional facilitated diffusion urea transporters (UT) have been described in mammalian tissues (15, 26, 46, 63) and fish (47, 57, 75, 77). With respect to elasmobranchs, a UT has been cloned in the kidney of the dogfish shark (Squalus acanthias) that shares considerable sequence similarity to the rat kidney UT-A2 isoform (65). Similar transporters have been isolated from the kidney of the little skate (Raja erinacea; see Refs. 52 and 53), the dogfish Triakis scyllia, (35) and the Atlantic stingray (Dasyatis sabina; see Ref. 38). Evidence for UTs has also been reported in ammoniotelic teleosts such as the plainfin midshipman (Porichthys notatus; see Ref. 48), the Japanese eel (Anguilla japonica, eUT; see Ref. 49), and in the rainbow trout (Oncorhynchus mykiss; see Refs. 47 and 57). Thus there is considerable evidence for the presence of UTs in cells of both ammoniotelic and ureotelic fishes; however, there is no information available on urea transport across the mitochondrial membrane of any animal to our knowledge.
In the present study, we tested the hypothesis that urea movement across the hepatic mitochondria is carrier mediated in the little skate (R. erinacea) and rainbow trout (O. mykiss). Hepatic mitochondria were isolated from rainbow trout and little skates using differential centrifugation. Characterization of urea transport kinetics was performed by exposing the mitochondria to a variety of external urea concentrations in the presence of [14C]urea, and influx was measured by rapid filtration techniques as described by Perry and Flik (56). Urea transport was also measured in response to a number of competitive and noncompetitive urea transport inhibitors, mitochondrial ionophores, metabolic inhibitors, and pH gradients. Furthermore, urea efflux was studied in skate mitochondria using a submitochondrial particle (SMP) preparation to generate inverted mitochondrial vesicles; [14C]urea uptake was then measured using the same treatments described above.
METHODS AND MATERIALS
Rainbow trout (O. mykiss Walbaum 1792) were purchased from Alma Aquaculture Research Station (Alma, Ontario) and were maintained in constantly flowing aerated fresh water at the Hagen Aqualab on the campus of the University of Guelph (Guelph, Ontario, Canada). Water temperatures were kept at 10 ± 1°C (pH 8.1) during the period of experiments. Trout were fed commercial trout pellets one time daily for the period of the experiment.
Little skates (R. erinacea Mitchell 1825) were collected by otter trawl from Passamaquoddy Bay, New Brunswick, Canada during the summer months of July and August in 2001 and 2003. Skates were kept at the Huntsman Marine Science Centre (St. Andrew's, New Brunswick, Canada) in 1,000-liter outdoor tanks exposed to the natural photoperiod. Tanks were supplied with unfiltered and aerated seawater (12–14°C, pH 8.0). For some experiments, a number of skates were transported to and held at the Hagen Aqualab at the University of Guelph. These fish were held under similar conditions to the natural photoperiod in recirculating artificial seawater (32 ppt, pH 8.2, 10°C; Crystal Sea, Baltimore, MD). Skates were fed filleted, chopped herring three times a week. Treatment of animals followed approved protocols and guidelines of the University of Guelph Animal Care Committee.
Isolation of mitochondria.
Mitochondria were isolated following a modified differential centrifugation protocol from Ballantyne (8). Slightly different protocols were required for each species. Briefly, trout (250–400 g) and skate (250–400 g) were stunned by a blow to the head and killed by severance of the spinal cord. All subsequent steps were carried out at 0–4°C. Whole livers were immediately excised from the animals and submerged in isolation media. The trout isolation media (TIM) was composed of 120 mmol/l KCl, 10 mmol/l HEPES, 0.1% albumin, and 5 mmol/l EGTA and was adjusted to a pH of 7.4. The skate isolation media (SIM) contained 180 mmol/l sucrose, 135 mmol/l KCl, 30 mmol/l HEPES, 1% albumin, 10 mmol/l KH2PO4, and 345 mmol/l d-mannitol, and the pH was adjusted to 7.2. Livers were diced into ∼1 mm2 using a razor blade and placed in a homogenization tube filled with 15 ml of isolation media. Homogenization consisted of three 10-s passes with a Potter-Elvehjem tissue homogenizer and a tight-fitting pestle attached to a high-torque hand drill with a constant speed of 100 revolutions/min. The homogenate was centrifuged at 570 g for 10 min to separate cellular debris such as erythrocytes, connective tissue, and fat. The supernatant was filtered through a cheesecloth, decanted in a clean centrifuge tube, and centrifuged at 5,700 and 11,000 g for trout and skate, respectively. The resulting supernatant was discarded, and the final pellet of isolated mitochondria was resuspended in 1 ml of isolation media for subsequent measurement of respiratory control ratios (RCRs), protein assays, and rapid filtration procedures.
Validation of isolated viable mitochondria.
RCRs were measured to ensure that intact, viable mitochondria were isolated. This ratio represents a comparison between the rate of respiration in the presence of a substrate and ADP (state 3) to the rate following the phosphorylation of ADP (state 4). Typically, ratios less than two indicate isolation of poor-quality mitochondria (49). RCRs were measured following the method of Moyes et al. (54) using Clark-type electrodes. Isolated mitochondria were adjusted to a protein concentration of 1–2 mg/ml using a respiratory media (100 mmol/l KCl, 10 mmol/l Tris base, 25 mmol/l KH2PO4, and 1% albumin, pH 7.4). Following the addition of a 5-μl aliquot of 50 mmol/l ADP to stimulate state 3 respiration, 10 μl of 1 mol/l succinate followed by 2 μl of 10 mmol/l malic acid were added to fuel the tricarboxylic acid cycle (state 4 respiration). RCRs were calculated by dividing the slope of the state 3 respiration by the slope of the state 4 respiration rate.
Mitochondrial matrix volume measurements.
Mitochondrial matrix volumes were measured as an additional indicator of mitochondrial viability and stability. A 2-mg protein sample of mitochondria was incubated in 1 ml of either SIM or TIM containing 10 μl (1 μCi) of 3H20 and 10 μl (0.1 μCi) of [14C]sucrose for 2 min in a 2-ml aliquot at 4°C. The mitochondrial suspension was sedimented by centrifugation at 12,000 g for 2 min. A 500-μl aliquot of supernatant was transferred to a 5-ml scintillation vial, and 3.5 ml of scintillation fluid (Fisher Scintisafe Econo F) were added. The remaining supernatant was decanted, and the walls of the tubes were cleaned to remove any residual supernatant. A 40-μl aliquot of Triton X-100 was added to the pellet and was resuspended by vortexing. The mixture was transferred to a 5-ml scintillation vial, and 3.5 ml of scintillant were added. The radioactivity (3H and 14C) of the supernatant and pellet suspension was measured using a two-channel liquid scintillation counter (6500 Beckman Liquid Scintillation Counter). Mitochondrial matrix volumes were calculated as described by Brand (16) and expressed as microliter per milligram protein.
Urea transport in intact mitochondria.
Transport of [14C]urea was measured by rapid filtration techniques carried out at 10°C as described by Perry and Flik (56). Freshly isolated mitochondria were resuspended in 10 ml of a modified isolation media composed of either TIM containing 290 mmol/l d-mannitol or SIM with 370 mmol/l d-mannitol for trout and skate, respectively, and allowed to equilibrate on ice for 45 min. Mitochondria were collected by centrifugation at 5,700 g (trout) or 11,000 g (skate) for 10 min at 4°C. The trout mitochondria were resuspended in 290 mM d-mannitol TIM and adjusted to a final protein concentration of ∼10 mg/protein ml. Initial experiments demonstrated that high sucrose levels in SIM clogged the filters during the filtration phase of the procedure. As such, skate mitochondria were resuspended in a modified SIM that contained no sucrose (135 mmol/l KCl, 30 mmol/l HEPES, 1% albumin, 10 mmol/l KH2PO4, and 370 mmol/l d-mannitol adjusted to a pH of 7.2) and adjusted to a concentration of 10 mg protein/ml. Transport was initiated by rapidly mixing the 10 μl of mitochondria with 40 μl of isolation media (TIM or SIM) containing 1 mmol/l [14C]urea (activity of 50 μCi/ml) at 10°C. Preliminary studies were completed to establish the relationship between urea uptake and time across the mitochondrial membrane. In both trout and skate mitochondria, urea uptake at 5 s was within the linear range of the relationship before uptake reached a plateau of maximum filling capacity. Therefore, in subsequent experiments, flux was terminated at 5 s. Termination of transport was accomplished by adding 1 ml of ice-cold stop solution. The stop solution for the rainbow trout mitochondria was composed of TIM containing 290 mmol/l urea. Similarly, the stop solution for the skate was made from the modified SIM with 370 mmol/l cold urea in place of 370 mmol/l d-mannitol. The high urea content of the stop solutions reduced the nonspecific binding of [14C]urea to the filters. Immediately following the termination of transport, the diluted mitochondrial solution was filtered through prewetted filters (type 0.4 μm HTTP; Millipore). The filters were rinsed two times using 3 ml of the ice-cold stop solution to remove any residual [14C]urea. Filters were placed in scintillation vials and filled with 15 ml of scintillation fluid (Fisher Scintisafe Econo F). Counts per minute were determined by using a liquid scintillation counter (Rackbeta LKB Wallac 1211 LS or 6500 Beckman Liquid Scintillation Counter). Uptake of [14C]urea (μmol·h−1·mg protein−1) was determined by the equation: where cpmf represents the counts per minute present on the filters, Amedia is the activity of the assay media in counts per minute per micromole, P corresponds to the amount of protein in milligrams present in the mitochondrial assay suspension, and t is the time of incubation (hours). Amedia was calculated for each stock urea concentration by measuring the counts per minute of the solution and dividing it by the micromole amount of urea, a sum of both nonradiolabeled and radiolabeled (sp act 55 mCi/mmol) urea components present in the assay media.
Urea transport was measured over a range of physiological urea concentrations for both trout and skate to establish a concentration dependence relationship. Trout mitochondria were exposed to concentrations ranging from 1 to 20 mmol/l (0, 0.01, 0.025, 0.05, 0.075, 0.1, 0.25, 0.5, 1, 2, 5, 10, 15, 20 mmol/l), and urea uptake was measured in skate mitochondria at concentrations of 0, 0.05, 0.1, 0.25, 0.75, 1, 2, 5, 10, 25, 100, 250, 370 mmol/l. To further characterize the properties of urea transport across the mitochondrial membrane, noncompetitive and competitive inhibitors were tested. A stock solution of the noncompetitive inhibitor phloretin (10 mmol/l) was prepared by first dissolving phloretin in ethanol and then by dilution in isolation media to concentrations that ranged from 0 to 0.25 mmol/l. Final ethanol concentrations did not exceed 0.25% vol/vol. Immediately before the addition of the radioactive mixture, 10 μl of phloretin solution was added to the mitochondria and vortexed. The radioactive mix containing urea (50 μCi/ml [14C]urea) was added, and transport was measured as described above. Phloretin-sensitive transport was measured at urea concentrations of 0.35 mmol/l in trout and 1 mmol/l in skate.
Mitochondria were incubated in 0.35 and 1 mmol/l urea for trout and skate, respectively, with twofold stronger concentrations of individual urea analogs such as acetamide, thiourea, and N-methylurea, as used in previous studies (57, 82, 89). Nitrophenylthiourea (NPTU) is a potent urea analog; as such, mitochondria were exposed to a significantly lower concentration (0.08 mmol/l) and incubated in the abovementioned urea concentrations. Preceding the addition of the radioactive urea solution, 10 μl of the analog (trout 0.70 mmol/l; skate 2 mmol/l) was added to the mitochondria and vortexed. Following the addition of the competitive inhibitors, the radioactive urea solution (50 μCi/ml [14C]urea) was added, and transport was measured as previously described.
The linkage between urea transport and mitochondrial function was investigated using a variety of compounds that influence the ionic permeability of lipid membranes. The chemical agent trifluoromethoxyphenylhydrazone (FCCP) was used to discharge both the mitochondrial pH gradient and membrane potential (45). However, the electroneutral K+/H+ exchanger nigericin was used to target and disrupt the pH gradient and valinomycin, a K+ selective transport compound, was employed to discharge the membrane potential (reviewed in Ref. 59). Mitochondria were exposed to 1 mmol/l cyanide to inhibit complex IV [cytochrome c oxidase (CCO)], the proton extrusion pump that is actively involved in establishing the proton gradient that drives ATP synthesis. FCCP (0.002 mmol/l with a final ethanol concentration of <0.14% v/v), nigericin (0.006 mmol/l with a final ethanol concentration of <0.17% vol/vol), valinomycin (0.001 mmol/l with a final ethanol concentration of <0.0085% vol/vol), and cyanide (1 mmol/l) were added individually (10 μl) to the mitochondria and mixed before the addition of 50 μCi/ml [14C]urea. Mitochondria were incubated in 0.65 and 1 mmol/l urea in trout and skate, respectively.
To further characterize the involvement of protons on urea transport, experiments varying the external pH and comparing the rate of urea uptake across the mitochondrial membrane were performed. Rainbow trout mitochondria were incubated in 0.65 mmol/l urea for 5 s; this solution was adjusted to a variety of pH values that ranged from 6.8 to 8.0. Hepatic mitochondria of the skate were incubated in 1 mmol/l urea solutions of varying pH (pH 6.5–7.8).
Urea uptake in SMP in skates.
Components of [14C]urea influx in skate mitochondria exhibited a higher degree of complexity than that of the trout. As such, a further aim of this study was to examine urea efflux in the hepatic mitochondria of the skate using a SMP preparation that generated inverted mitochondrial vesicles. SMPs were isolated following a modified method from D'Souza and Wilson (25). In brief, frozen mitochondria from R. erinacea, stored at −80°C, were thawed and suspended in 10 ml of SIM. All subsequent steps were carried out at 0–4°C. The mitochondrial suspension was sonicated with a Vibracell VC 50 sonicator (setting of 70 W; Sonics and Materials, Danbury, CT) for four 10-s passes on ice. Following sonication, the suspension was centrifuged at 47,813 g at 4°C for 60 min. The resulting supernatant was decanted, and the pellet was resuspended in a modified SIM that was adjusted to a higher pH (180 mmol/l sucrose, 135 mmol/l KCl, 30 mmol/l HEPES, 1% albumin, 10 mmol/l KH2PO4, and 345 mmol/l d-mannitol, pH 8.2), which was equivalent to the pH of the matrix in the intact mitochondria. The suspension was passed through a 22-gauge needle (4 times) to aid SMP resealing and formation.
The orientation of the SMPs was assessed using the mitochondrial marker enzyme CCO. In brief, the percentage of inside-out SMPs (IO-SMPs) was determined by measuring CCO activity of a 50-μl aliquot of SMPs in the presence or absence of 10 μl of 2% vol/vol Triton X. The following calculation was used: where CCOTriton X is the CCO activity of SMPs treated with Triton X. CCO represents the enzyme activity of untreated SMPs, and CCO activity was measured following the method of Blier and Guderley (12).
Urea uptake rates were measured in SMPs under identical treatments to those used in intact mitochondria (concentration dependence, analogs, chemical inhibitors and uncouplers, and in the presence of a pH gradient).
Liver samples were collected and stored at −80°C for up to 6 mo before analysis of urea content. Extracts were prepared by grinding the tissue to a fine powder with a mortar and pestle under liquid nitrogen (88). Samples were deproteinized using ice-cold 8% perchloric acid and centrifuged at 14 000 g for 10 min at 4°C. Urea concentrations were analyzed as described previously by Rahmatullah and Boyde (61). Tissue water content was determined from the liver tissue wet and dry weights. Urea concentrations were corrected for tissue water content and are expressed as micromoles per liter of tissue water.
Urea content of mitochondria was measured before the transport assays to confirm that the levels of urea within the mitochondria were negligible. Different compounds in the TIM and SIM adversely affected the color development of the urea assays. Therefore, urea levels were determined by conversion to ammonia using a urease protocol described by Wood et al. (83). Ammonia content was subsequently analyzed according the method of Verdouw et al. (73).
[14C]urea was purchased from Amersham Life Science (Baie d'Urfé, Quebec, Canada). Phloretin, FCCP, nigericin, valinomycin, [14C]sucrose, and 3H2O were all obtained from Sigma through Sigma-Aldrich. N-methylurea and NPTU were acquired from Aldrich through Sigma-Aldrich. All other chemicals were of reagent grade and procured from either Fisher Scientific or Sigma Chemical (St. Louis, MO).
Values are expressed as means ± SE. Nonlinear or linear regressions were completed to determine the line of best fit. Regression analysis was performed for the concentration dependence experiments using the method of least squares, and the significance of the correlation coefficient value, r2, was assessed. Total transport over the entire range of urea concentrations was defined as a sum of two regressions, a linear element (y = mx + b) and a saturation element [y = ax/(c + x)]. Michaelis constant (Km) and maximal velocity (Vmax) values were calculated using an Eadie-Hofstee plot, since higher rates of error are associated with the more conventional Lineweaver-Burke method (24, 60). For all trials focusing on the effects of transport and metabolic inhibitors, an estimated diffusion component at the experimental urea concentration was subtracted from the individual values of urea uptake measured at each concentration, thus giving a rate of uptake due to urea transport proteins alone.
A one-way ANOVA was performed to compare the differences among urea uptake of mitochondria exposed to various competitive urea analogs. A Dunnett's test was used as a comparison test for analogs against the control values. A Student's paired t-test was used to compare the differences between the control rates and the rates of the mitochondria exposed to the chemical uncouplers. Results were considered statistically significant if the P value was <0.05.
On average, freshly isolated mitochondria from trout had average RCR values ranged from 3.0 to 7.7 with an average value of 5.11 ± 0.87 (n = 4), whereas mitochondria from the little skate had a range of 2.8–4.3, with a mean of 3.11 ± 0.58 (n = 4). Further comparisons with values obtained after a 45-min incubation period in d-mannitol yielded values of 4.21 ± 0.47 (range 2.8–6.3; n = 4) for trout and 3.03 ± 0.29 (range 2.5–4.4; n = 4) for the skate, which were not significantly different from values from freshly isolated mitochondria (P > 0.05). The mitochondrial matrix volumes were 1.03 ± 0.11 and 1.15 ± 0.18 μl/mg for trout and skate, respectively.
Urea content of trout liver was 0.60 ± 0.16 mmol/l tissue water (n = 12), whereas the value in skate liver was 278.5 ± 23.76 mmol/l tissue water (n = 12). Mitochondrial urea content following incubation in d-mannitol was essentially zero in trout and skate (n = 4).
Mitochondrial urea influx was measured over a range of external urea concentrations, revealing two components of uptake in trout. Over the entire range of concentrations from 0 to 20 mmol/l, urea uptake in trout mitochondria was linearly dependent on the external urea concentration (r2 = 0.91; Fig. 1A). However, at lower urea concentrations within the physiological range (<5 mmol/l), urea uptake exhibited saturation kinetics (r2 = 0.97; Fig. 1B). Transformation of the data using an Eadie-Hofstee plot (Fig. 1C) revealed a Km of 0.59 mmol/l and a Vmax of 0.12 μmol·h−1·mg protein−1 (Table 1).
Following subtraction of the diffusion component of transport, saturable urea influx in trout mitochondria demonstrated concentration sensitivity to the urea transport inhibitor phloretin (Fig. 2). The IC50 of phloretin in the trout was <0.01 mmol/l. There was no change to the control rate of urea transport with the addition of ethanol vehicle alone (P > 0.05). Furthermore, exposure to 0.35 mmol/l urea and various urea analogs (acetamide, thiourea, N-methylurea, and NPTU) caused a 1.6- to 1.7-fold significant decrease in the rate of uptake relative to rates of urea alone (P < 0.05, Fig. 3A). Investigations focusing on the metabolically active components of mitochondrial function demonstrated that saturable urea uptake in trout mitochondria was not influenced by a number of ionophores and inhibitors such as FCCP, nigericin, valinomycin, and cyanide (P > 0.05; data not shown). In addition, varying the external pH from 6.8 to 8.0 had no influence on urea influx (P > 0.05; data not shown).
Total urea transport in skate mitochondria between 0 and 370 mmol/l demonstrated a dual pattern, as observed in trout. At urea concentrations >5 mmol/l, there was a strong linear relationship between urea concentration and urea uptake (r2 = 0.99; Fig. 4A). Analysis at low urea concentrations (<5 mmol/l) revealed a second saturable component to transport (r2 = 0.92; Fig. 4B). With the use of the Eadie-Hofstee transformation (Fig. 4C), the calculated Km value was 0.34 mmol/l with a Vmax of 0.054 μmol·h−1·mg protein−1 (Table 1).
Exposure to phloretin resulted in a 50% decrease in saturable urea influx rates at a concentration of 0.16 mmol/l (Fig. 2). The use of competitive urea analogs caused a 1.9- to 3.7-fold decrease in urea uptake (P < 0.05; Fig. 3B). Alteration of membrane potential (FCCP, valinomycin), the pH gradient (nigericin), or respiration (cyanide) significantly inhibited urea influx by 5-, 2-, 2-, and 1.3-fold, respectively, relative to control rates (Fig. 5A). Further investigations that varied external pH revealed a proton-dependent component to saturable urea transport in intact skate mitochondria. Rates of total urea uptake were inversely correlated to external pH in intact mitochondria (r2 = 0.99), with rates of influx inversely proportional to external pH (Fig. 6A).
CCO activity revealed that following isolation, ∼92.23 ± 2.23% of SMPs were in the correct inside-out formation. Urea uptake by skate SMPs corroborated results observed in intact mitochondria. Two components of total urea uptake were observed [a linear diffusion constituent (Fig. 7A) and a saturable constituent (Fig. 7B)]. The rate of urea uptake was linear over a wide range of urea concentrations (5–370 mmol/l) but was significantly 2- to 5-fold higher in SMPs at various urea concentrations compared with intact mitochondria (P < 0.05). Eadie-Hofstee analysis (Fig. 7C) of low urea concentrations (<5 mmol/l) revealed a calculated Km value of 0.65 mmol/l and a Vmax of 0.058 μmol·h−1·mg protein−1 (Table 1). Phloretin was effective at significantly decreasing rates of urea uptake (P < 0.05); however, the IC50 was lower with a value of 0.075 mmol/l (Fig. 2). Transport rates of urea in SMPs were sensitive to analogs such as acetamide, thiourea, N-methylurea and NPTU, causing a 2.2- to 2.5-fold decrease relative to control rates (P < 0.05; Fig. 3C). As well, urea movement was significantly inhibited by FCCP (1.9-fold; P < 0.05) and nigericin (1.8-fold; P < 0.05) but not valinomycin (P > 0.05; Fig. 5B). External pH influenced the rate of urea transport, specifically resulting in an inversely proportional relationship (r2 = 0.99) with rates of urea uptake in SMPs decreasing with increases in external pH (Fig. 6B).
Validation of preparation.
The outer mitochondrial membrane is freely permeable to relatively large compound pores formed by the VDACs, which limit the trafficking of molecules >1 kDa (21). However, it is the inner membrane that poses a significant barrier for many smaller organic and inorganic solutes moving between the cytoplasmic and mitochondrial matrix compartments. For this reason, we hypothesized that a UT would be present in this membrane to facilitate the movement of urea out of the mitochondria. By using isolated mitochondria, it was possible to characterize the transport kinetics of urea across the inner membrane. In the present study, the isolation procedure consistently yielded intact and metabolically coupled trout and skate mitochondria as judged by RCR values with succinate as a substrate. Measurements of RCR values following incubation in d-mannitol demonstrated that, before the transport assays, the mitochondria were still energetically viable, and values were not significantly different from freshly isolated mitochondria. Typically, succinate yields lower RCR values than other substrates such as malate or glutamine (19); however, the values obtained in this study, with succinate as a substrate, were well within the normal range for both teleost and elasmobranch species (e.g., see Refs. 11, 19, 68, 70). Measurement of matrix volumes confirmed that the treatments used in this study did not cause mitochondrial swelling, since measured values in this study were consistent with those reported in the piscine literature (10). Furthermore, analysis of urea content in the matrix of isolated mitochondria following incubation in d-mannitol revealed that internal concentrations were negligible, therefore confirming that a urea gradient was present during the transport assays.
The data support our hypothesis that urea transport is carrier mediated in the little skate and trout liver mitochondria. This is the first report of UT in mitochondria, or in any subcellular membrane to our knowledge. Mitochondrial urea uptake by skate hepatocytes is saturable at low urea concentrations (<5 mmol/l), indicating the presence of a carrier-mediated transport system. Kinetic analysis of urea transport for intact skate mitochondria and SMP preparations revealed Km values (0.34 and 0.65 mmol/l, respectively) that were significantly lower than physiological concentrations observed in hepatocytes. Similar trends have been reported for gill tissue of the spiny dogfish S. acanthias (28) and kidney of the little skate (52). These high-affinity UT appear to substantially reduce the efflux of urea from excretory surfaces by actively “scavenging” intracellular urea and returning it to the plasma, overall, facilitating urea retention in these species. In the case of mitochondria, the putative urea transport system may facilitate the rapid equilibration of urea across the inner mitochondrial membrane, potentially reducing pronounced fluctuations in mitochondrial volume due to the high endogenous rates of urea synthesis in elasmobranchs. This is supported by evidence from Ballantyne and Moon (10) who have shown that, in R. erinacea, urea is quickly and uniformly distributed across the mitochondrial and cytosolic compartments, whereas many solutes, such as K+, Cl−, sucrose, and trimethylamine oxide, are unable to penetrate the inner mitochondria or do so very slowly (10, 29, 34, 44, 50).
Urea transport kinetics in trout mitochondria showed similar trends to skate mitochondria and were also consistent with other teleost cell membrane studies (47, 57). Trout mitochondria showed saturation kinetics at low urea concentrations (<5 mmol/l) with a Km value of 0.59 mmol/l, similar to physiological concentrations in the liver. Because urea is a relatively minor component of nitrogen metabolism in adult teleosts (reviewed in Refs. 42, 80, and 86), equilibration of urea across the mitochondrial compartment would probably be less of a concern compared with elasmobranch species. However, it is possible that UT play a more prominent role during early life stages, when the production and excretion of urea are believed to be essential mechanisms for ammonia detoxification (69, 88).
Submillimolar concentrations of phloretin, as low as 0.0125 mmol/l in trout and 0.05 mmol/l in skate, caused a significant inhibition of urea uptake in both trout and skate mitochondria. Interestingly, urea uptake was not inhibited >50% in each preparation, despite the increasing concentration of phloretin up to 0.25 mmol/l. Although phloretin inhibition is a hallmark of facilitated urea transport systems, numerous in vitro studies in fish have reported varying levels of inhibition to phloretin (18–85% relative to control rates) at comparable concentrations (28, 47, 52, 77). These dissimilarities may be caused in part by an interaction between phloretin and lipid regions of the membrane (39), which has been shown to modulate the conformation of urea transport proteins and aqueous pores (71). In the present study, species-specific variations in mitochondrial lipid content (30) may have contributed to the diverse effects of phloretin; however, further experiments are required to confirm this hypothesis.
Urea uptake in both trout and skate mitochondrial preparations was significantly inhibited following exposure to acetamide, thiourea, N-methylurea, and NPTU. In trout, the pattern of analog sensitivity was urea > NPTU ≥ acetamide ≥ thiourea ≥ N-methylurea. In comparison, skate mitochondria, both intact and SMP preparations, demonstrated a pattern of urea > acetamide ≥ thiourea > NPTU > N-methylurea. Sensitivities to NPTU and N-methylurea have been reported in the branchial tissue of S. acanthias (28), whereas urea transport in the dorsal section of the skate kidney was only sensitive to NPTU (52). In teleosts, in vitro preparations have shown similar variations in analog handling (47, 75, 76). Quantitative differences in urea analog reactivity may depend on a variety of factors, including variations in membrane transport proteins, the regulatory state of the same transport protein (e.g., phosphorylation or methylation), and even the differences in lipid composition of the membranes (20, 46).
We used FCCP to eliminate both the mitochondrial pH gradient and membrane potential (45), and nigericin, an electroneutral K+/H+ exchanger, to disrupt the pH gradient (reviewed in Ref. 59) to understand the energetic basis of urea transport in mitochondria. We have shown that urea transport in intact skate mitochondria and SMPs was significantly inhibited by FCCP and nigericin. Although this indicates that urea transport is influenced by the movement of protons, it also suggests that transport may potentially be dependant on the movement of other cations, such as K+, across the inner mitochondrial membrane. In addition to FCCP and nigericin sensitivity, urea uptake in intact skate mitochondria, but not SMPs, was significantly inhibited by valinomycin. Although valinomycin, a K+ selective ionophore, dissipates membrane potential in respiring mitochondria (reviewed in Ref. 59), there is also evidence that this compound indirectly decreases the pH gradient at high K+ concentrations due to an increase in general ionic permeability (1, 7). The concentrations of K+ used in this study were similar to those used by Ahmed and Booth (1); therefore, valinomycin may have dissipated both the membrane potential and the pH gradient. Hence urea movement in skate mitochondria appears to be linked to the transport of protons. This interpretation is consistent with the data collected from both intact skate mitochondria and SMPs where urea transport was enhanced in response to acidification of the external media.
Urea is an uncharged molecule; as a result, some prokaryotes have evolved systems where extracellular urea is translocated through the cell membrane proton-dependent and/or active processes. For example, in the prokaryotes Corynebacterium glutamicum and Klebsiella pneumonaie, urea movement is accomplished by a secondary transport system that is linked to the proton motive force (37, 64). Although both skate mitochondrial preparations were significantly inhibited by cyanide, it must be noted that this compound inhibits respiration by irreversibly binding to complex IV in the electron transport chain. Complex IV CCO is a proton extrusion pump that is actively involved in establishing the proton gradient that drives ATP synthesis. In the present study, exposure to cyanide may have caused a disruption of the H+ gradient therefore diminishing proton-dependent urea transport in skate mitochondrial preparations. Proton-gated urea channels have been reported in both Helicobacter pylori and H. hepaticus (78, 79). It is not unreasonable to postulate that similar mechanisms may be present in skate mitochondria, since it has been theorized that the ancestry of mitochondria may have originated from the Archaebacteria (reviewed in Ref. 9). Interestingly, proton-sensitive urea transport has also been reported in one vertebrate species, the green toad, Bufo viridis (62). It appears that active urea transport across the skin of the toad is accomplished by a phloretin-sensitive cotransport system with protons. Although the mitochondrion is characterized by the constant extrusion of protons that is accomplished by multisubunit complexes, protons also reenter the matrix space, the ATP synthase, which is the driving force of ATP synthesis. Based on the findings in the present study, further experiments are required to determine whether urea in skate mitochondria is transported in a symporter, antiporter, or pH-gated fashion.
Previous studies have identified a subgroup of AQPs (AQP3, -7, -9, and -10), termed aquaglyceroporins, that transport urea, glycerol, and other small solutes in addition to water (reviewed in Ref. 74). These channels are highly expressed in the mammalian brain, testis, and kidney (5, 36, 72). Interestingly, a recent study by Amiry-Moghaddam and colleagues (2) has localized the expression of AQP9 to the inner mitochondrial membrane of astrocytes and midbrain neurons in rats. Similar channels may be present on the inner mitochondrial membrane in skate hepatocytes.
Perspectives and Significance
In the past few years, significant knowledge has been gained about the physiological roles and phylogenetic relationship of UT, which have been cloned in bacteria, yeast, amphibians, fish, and mammals. Structural analysis of UT among lower vertebrates reveals that they have high homology (53–71%) with the mammalian UT-A2 transporter, suggesting that they all evolved from a common ancestral form (reviewed in Ref. 6). The movement of urea by the majority of these proteins relies on facilitated transport mechanisms, similar to the putative teleost mitochondrial UT described in this study. The mitochondrial location of piscine arginase predicates that the mitochondria-to-cytosol urea gradient in trout would be favorable, and a simple facilitated transporter would carry out this task. This can be contrasted against the situation in skates where high rates of urea production may result in elevated levels of urea in the mitochondria, and efflux may be further complicated by the maintenance of high cytosolic urea levels that is the result of the ureosmotic strategy used by marine elasmobranchs. Therefore, a link between the transport of urea and tightly controlled H+ movements across the mitochondria may be advantageous, similar to the H+-gated urea channel in Helicobacter sp. that is coupled to the enzyme urease (78, 79). These findings present valuable groundwork and direction for the question of whether mitochondrial urea transport genes represent another isoform of the piscine UTs or whether these transporters are a novel family of mitochondrial UTs that arose from an ancestral form, similar to those characterized in prokaryotes. Interestingly, prokaryotic UT genes have minimal identity (20–25%) with higher organisms, and no homologous counterparts of these genes have been found in vertebrates, to date (6). This may be due in part to the limitations of mRNA or genomic screening approaches used in the numerous studies. These putative and potentially novel mitochondrial UT genes may be one of the 5% of mitochondrial proteins that are not encoded by nuclear DNA; therefore, future investigations should focus on mining both nuclear and mitochondrial piscine genomes.
This work was funded by a Natural Sciences and Engineering Research Council Discovery grant to P. A. Wright and a W. B. Scott Ichthyology scholarship to T. M. Rodela.
We thank F. Purton (Huntsman Marine Science Centre) for logistical support.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society