Role of the glutamate dehydrogenase reaction in furnishing aspartate nitrogen for urea synthesis: studies in perfused rat liver with 15 N

Transcription

1 Biochem. J. (2003) 376, (Printed in Great Britain) 179 Role of the glutamate dehydrogenase reaction in furnishing aspartate nitrogen for urea synthesis: studies in perfused rat liver with N Itzhak NISSIM 1,Oksana HORYN, Bohdan LUHOVYY, Adam LAZAROW, Yevgeny DAIKHIN, Ilana NISSIM and Marc YUDKOFF Division of Child Development and Rehabilitation Medicine, Department of Pediatrics, Children s Hospital of Philadelphia, University of Pennsylvania School of Medicine, Philadelphia, PA 19104, U.S.A. The present study was designed to determine: (i) the role of the reductive amination of α-ketoglutarate via the glutamate dehydrogenase reaction in furnishing mitochondrial glutamate and its transamination into aspartate; (ii) the relative incorporation of perfusate NH 4 Cl, [2- N]glutamine or [5- N]glutamine into carbamoyl phosphate and aspartate-n and, thereby, [ N]urea isotopomers; and (iii) the extent to which perfusate [ N]aspartate is taken up by the liver and incorporated into [ N]urea. We used a liver-perfusion system containing a physiological mixture of amino acids and ammonia similar to concentrations in vivo, with Nlabel only in glutamine, ammonia or aspartate. The results demonstrate that in perfusions with a physiological mixture of amino acids, approx. 45 and 30 % of total urea-n output was derived from perfusate ammonia and glutamine-n respectively. Approximately two-thirds of the ammonia utilized for carbamoyl phosphate synthesis was derived from perfusate ammonia and one-third from glutamine. Perfusate [2- N]glutamine, [5- N]glutamine or [ N]aspartate provided 24, 10 and 10 % respectively of the hepatic aspartate-n pool, whereas perfusate NH 4 Cl provided approx. 37 % of aspartate-n utilized for urea synthesis, secondary to the net formation of [ N]glutamate via the glutamate dehydrogenase reaction. The results suggest that the mitochondrial glutamate formed via the reductive amination of α- ketoglutarate may have a key role in ammonia detoxification by the following processes: (i) furnishing aspartate-n for ureagenesis; (ii) serving as a scavenger for excess ammonia; and (iii) improving the availability of the mitochondrial [glutamate] for synthesis of N-acetylglutamate. In addition, the current findings suggest that the formation of aspartate via the mitochondrial aspartate aminotransferase reaction may play an important role in the synthesis of cytosolic argininosuccinate. Key words: argininosuccinate, carbamoyl phosphate, glutamate dehydrogenase, liver perfusion, mitochondrial aspartate aminotransferase, urea synthesis. INTRODUCTION The process of urea synthesis involves equimolar consumption of NH 4+ and aspartate nitrogen (N) [1,2], but the relative contributions of portal blood ammonia and amino-n for urea-n are still unclear. Therefore the central aim of the present study was to address the following question: what are the primary source(s) of hepatic ammonia and aspartate-n utilized for the synthesis of CP (carbamoyl phosphate) and ASA (argininosuccinate) respectively? Numerous studies have indicated that ammonia and glutamine nitrogen taken up by the liver are the chief sources for the mitochondrial CP synthesis in the periportal hepatocytes [2 7]. However, the source(s) of aspartate-n is obscure, since the incorporation of aspartate into ASA introduced to the liver via the portal vein is negligible when compared with the rate of ureagenesis, owing to hepatic zonation [8]. An example of such zonation is the location of urea-cycle enzymes in periportal hepatocytes, whereas the uptake of aspartate and/or glutamate primarily takes place in the perivenous cells, the site of hepatic glutamine synthetase [3,9]. Therefore a fundamental but yet unresolved question is related to the sources of aspartate-n utilized for urea synthesis [10]. A further question is whether aspartate required for the synthesis of ASA is formed via the Mit-Asp-AT (mitochondrial aspartate aminotransferase) reaction or via the cytosolic aspartate aminotransferase reaction? Recent investigations of the AGTrs (mitochondrial aspartate/glutamate transporters) have indicated that AGTrs play an important role in the regulation of ureagenesis by transporting mitochondrial aspartate into the cytosol for the synthesis of ASA [11,12]. However, an in vivo study indicated that aspartate, derived from glutamate via the Mit-Asp-AT reaction, does not equilibrate with cytosolic aspartate, and only the aspartate formed in the cytosol is used for ASA synthesis [2]. The latter suggestion contradicts previous findings indicating that aspartate required for urea synthesis must be generated in the mitochondria [13]. In addition, our studies, using N-Ps ( N-labelled precursors), have indicated that the mitochondrial metabolism of glutamine may provide between 25 and 30 % of aspartate utilized for synthesis of urea [6,7]. Glutamine is taken up by periportal hepatocytes and metabolized via PDG (phosphate-dependent glutaminase) to glutamate and ammonia [3,6,7]. Subsequently, glutamate is transaminated to aspartate to provide the second nitrogen for urea synthesis [6,7]. In addition, glutamate may be oxidized via the GDH (glutamate dehydrogenase) reaction, thus consuming glutamate and possibly limiting transamination to aspartate. Increased net flux towards oxidative Abbreviations used: AAM, amino acids mixture; AGTs, mitochondrial aspartate/glutamate transporters; AOA, amino-oxyacetate; ASA, argininosuccinate; CP, carbamoyl phosphate; GC MS, gas chromatography mass spectrometry; GDH, glutamate dehydrogenase; HI/HA, hyperinsulinism and hyperammonaemia; α-kg, α-ketoglutarate; Mit-Asp-AT, mitochondrial aspartate aminotransferase; MPE, mol % excess; NAG, N-acetylglutamate; N-P, N- labelled precursor; PDG, phosphate-dependent glutaminase. 1 To whom correspondence should be addressed ( ssitz@mail.med.upenn.edu).

2 180 I. Nissim and others deamination of glutamate may occur after the alteration of acid base homoeostasis [14], or the congenital GDH gain-offunction mutation, which causes HI/HA (hyperinsulinism/hyperammonaemia) in infants [,16]. Thus a further aim of the present study was to determine the role of the mitochondrial GDH reaction in furnishing glutamate by allowing reductive amination of α-kg (α-ketoglutarate). Glutamate so formed was then converted into aspartate, which was used for ureagenesis. It has been well established that the GDH reaction was close to equilibrium [17]. However, it was not clear whether the net flux of the hepatic GDH reaction was towards glutamate production via the reductive amination of α-kg, or glutamate consumption via the oxidative deamination of glutamate. This uncertainty was especially true in experiments with a physiological ammonia concentration and other amino acids similar to those that exist in vivo.results reported regarding the net flux of the GDH reaction were inconsistent and, at times, contradictory [,16]. The reason for this inconsistency has been a lack of the precise determination of the true NH 4+ flux towards reductive amination of α-kg or generation of NH 4+ via oxidative deamination of glutamate. However, by using N-Ps and physiological ammonia concentrations similar to those taken up by the liver in vivo, wecandeterminethe net formation of [ N]glutamate from NH 4+ via the reductive amination of α-kg, or the net formation of NH 4+ from [2- N]glutamine via the oxidative deamination of glutamate [4 7]. In the present study, we took advantage of NandGC MS (gas chromatography mass spectrometry) methodology to determine the source(s) of N used for the synthesis of CP or ASA in a liverperfusion system containing a physiological mixture of amino acids similar to those that exist in rat blood, with only glutamine or ammonia labelled with N. Using this methodology, we have shown previously that glutamine is the major source for both the urea-ns [5 7]. Results obtained from isolated hepatocytes [5] may not precisely reflect the structure function relationship and the different metabolic zonations of the liver. Similarly, liver perfusions with a single amino-n [4,7,10] may not reflect the amino acid milieu to which the liver is exposed in vivo. Thisis an important consideration, since a single nitrogen precursor (i.e. glutamine and/or ammonia), in the absence of cognate amino-n, might well affect the metabolism of the test amino acid. For example, it has been shown that leucine influences the GDH reaction and the net production of glutamate from NH 4+ and α-kg [,18,19] and, thereby, may affect hepatic nitrogen metabolism as indicated [19]. Therefore elimination of leucine from the perfusate may change the relative contribution of the amino-n of glutamine and/or glutamate for ammonia and/or aspartate nitrogen. An additional series of experiments was designed to determine the extent to which external (perfusate) aspartate is taken up by the liver and incorporated into urea. As yet, there is no information regarding hepatic uptake of perfusate [ N]aspartate and its relative incorporation into urea or other metabolites, such as glutamate or glutamine. In the present study, perfusions were performed with L-[ N]aspartate, unlabelled ornithine and ammonia, in the absence or presence of either glucagon or insulin. These hormones have a significant role in the regulation of amino acid uptake and their hepatic metabolism [1,6,7,20,21]. We sought to determine whether these hormones modulate hepatic aspartate uptake and its incorporation into urea synthesis. The results demonstrate that approximately two-thirds of the ammonia utilized in CP synthesis was derived from perfusate ammonia and one-third from perfusate glutamine. In addition, perfusate ammonia supplied approx. 37 % of aspartate-n utilized for the synthesis of ASA. Perfusate [2- N]glutamine, [5- N]glutamine and [ N]aspartate provided approx. 24, 10 and 10 % respectively of aspartate-n. MATERIALS AND METHODS Materials and animals Male Sprague Dawley rats (Charles River, Wilmington, MA, U.S.A.) were fed ad lib. onastandard rat chow diet. Chemicals were of analytical grade and obtained from Sigma Aldrich. Enzymes and cofactors for the analysis of urea, lactate, pyruvate, glucose and ammonia were obtained from Sigma. N-labelled NH 4 Cl, aspartate and glutamine [99 MPE (mol% excess)], were from Isotech (Miamisburg, OH, U.S.A.). Liver perfusions and experimental design Livers (9 12 g) from fed male Sprague Dawley rats were perfused in the non-recirculating mode as described in [22]. We employed the single-pass perfusion with antegrade flow direction (3 3.5 ml/g). The basic perfusion medium was a Krebs buffer continuously gassed with O 2 /CO 2 (19:1), containing 2.1 mm lactate and 0.3 mm pyruvate as metabolic fuels. Perfusion flow rate, ph, pco 2 and po 2 (in influent and effluent media) were monitored throughout, and oxygen consumption was calculated. After min of pre-perfusion, we changed to a medium that contained, in addition to the lactate and pyruvate, a mixture of unlabelled amino acids (in mm) including: Ala (0.3), Arg (0.2), Asp (0.05), citrulline (0.05), Cys (0.1), Glu (0.1), Gly (0.2), His (0.1), Ile (0.1), Leu (0.2), Orn (0.05), Ser (0.1), Thr (0.1), Trp (0.1), Tyr (0.1) and Val (0.25). In a separate series of perfusions, the AAM (amino acids mixture) was supplemented with either 0.3 mm NH 4 Cl and 1 mm unlabelled glutamine, or 0.3 mm unlabelled NH 4 Cl and 1 mm [2- N]- or [5- N]-glutamine (99 MPE). To determine whether the absence of AAM would alter the fraction of amino-n of glutamine utilized for aspartate-n, perfusions were performed with 1 mm [2- N]glutamine (99 MPE) plus 0.3 mm NH 4 Cl in the presence or absence of 1 mm AOA (aminooxyacetate), an inhibitor of the aminotransferase reactions. Next, we examined the degree to which perfusate [ N]aspartate is taken up by the liver as well as its relative incorporation into urea (periportal hepatocytes) or glutamate and/or glutamine (perivenous hepatocytes). To this end, perfusions with antegrade flow direction were performed with 0.1 mm [ N]aspartate, 0.05 mm ornithine and 0.3 mm ammonia. An additional series of perfusions was performed with either glucagon or insulin (10 7 M) to determine the action of these hormones on the hepatic uptake of [ N]aspartate, and its incorporation into urea or amino acids. In each of the experiments outlined above, the perfusion was continued for 60 to 70 min. Samples were taken from the influent and effluent media for chemical and GC MS analyses. At the end of the perfusion, the liver was freeze-clamped with aluminium tongs precooled in liquid nitrogen. The frozen liver was ground into a fine powder, extracted into HClO 4,andused for amino acid determination by HPLC, utilizing precolumn derivatization with o-phthalaldehyde [23]. Ammonia [24] and urea levels [22] in the effluent were also assayed. NAG (N-acetylglutamate) level in each liver extract was determined using GC MS and an isotope dilution approach, as described in [22]. GC MS methodology and determination of N-labelled metabolites GC MS measurements of Nisotopic enrichment were performed on a Hewlett Packard 5970 MSD and/or 5971 MSD, coupled with a 5890 HP GC, as described previously [4 7,22].

3 Sources of urea nitrogen 181 Table 1 Nitrogen balance across the liver Values are means + S.D. determined between 30 and 60 min (steady state) of the perfusion. N/A, not applicable. Perfusions with glutamine (nmol min 1 g 1 )* Perfusions with aspartate (nmol min 1 g 1 ) Control (+) AAM (+) AOA Control (+) Insulin (+) Glucagon Nitrogen uptake Glutamine-N N/A N/A N/A Ammonia-N Aspartate-N N/A N/A Alanine-N N/A N/A N/A N/A N/A Glycine-N N/A N/A N/A N/A N/A Total nitrogen output Urea-N Glutamate and alanine N/A Glutamine-N N/A N/A N/A Nitrogen balance ( ) 340 ( ) 4 ( ) 561 * Control: perfusions with only 1 mm glutamine plus 0.3 mm NH 4 Cl without AOA; (+)AOA:with 1 mm; (+) AAM: perfusions with 1 mm glutamine, 0.3 mm NH 4 Cl and physiologic mixture of amino acids as indicated under the Materials and methods section. Control: perfusions with 0.1 mm aspartate, 0.05 mm ornithine and 0.3 mm NH 4 Cl, without hormone; (+)Insulin: in the presence of insulin; (+)Glucagon: in the presence of glucagon (10 7 M). Values are (nmol min 1 g 1 )ofglutamine. P < 0.05 compared with the respective control, in either perfusions with glutamine plus ammonia or aspartate plus ammonia as nitrogen source. Values are (nmol min 1 g 1 )ofureatimes 2. N-balance is the difference between uptake and output as estimated from the mean values. For measurement of the Nenrichment in urea and amino acids, samples were prepared as described previously [4 7]. Briefly, a 500 µl aliquot of effluent or liver extract was purified via an AG-50 (H + ; mesh; 0.5 cm 2.5 cm) column, and then converted into the t-butyldimethylsilyl derivatives. Ions having the m/z of the urea t-butyldimethylsilyl derivative were monitored as described in [25] for singly (U m+1 )and doubly (U m+2 ) N-labelled urea isotopomers [4,6,7]. Isotopic enrichment in glutamate, aspartate and alanine was monitored using ratios of ions at m/z 433:432, 419:418 and 261:260 respectively. Enrichment in [2- N]glutamine was determined by monitoring m/z 259:258 [26], and [5- N]glutamine by determining the difference between m/z 432:431 and m/z 259:258 ratios. Additionally, doubly labelled glutamine was measured using the m/z 433:431 ratio. Formation of NH 4+ was determined as described in [10]. Calculations and statistical analyses The rate of precursor-n uptake or the output of nitrogen was determined by the measurement of metabolite concentration in the influent and effluent (nmol/ml), normalized to the flow rate (ml/ min) and liver wet weight as described previously [22]. Nitrogen balance across the liver was calculated from the differences between the rate of nitrogen uptake (from ammonia and amino-n) or output (as urea-n and amino-n). Percentage contribution of N-P for N-labelled glutamate, aspartate or ammonia ( N-product) was calculated as [ N- product (MPE)/ N-P (MPE)] 100. The Nenrichment in glutamate or aspartate was determined in the freeze-clamped liver at the end of each perfusion. N-labelled ammonia was determined in the freeze-clamped liver or in the effluent at the end of the perfusion. The output of N-labelled urea or amino-n was the product of Nenrichment (MPE/100) times concentration [nmol min 1 (g wet wt) 1 ]andisexpressed as nmol of Nmetabolite min 1 (g wet wt) 1.The distribution of urea mass isotopomers (e.g. U m, unlabelled urea-n; U m+1,ureacontaining one N; or U m+2,urea containing two N) was also calculated and compared with the observed value, using the mathematical model we have described previously [4]. The percentage of a given N-P transferred into U m+1 or U m+2 at the steady state (between 40 and 70 min of perfusion) was calculated as follows: % N-P transferred [ ] Um+1 output (nmol min 1 g 1 ) = 100 for U N-P uptake (nmol min 1 g 1 m+1 ) % N-P transferred [ ] Um+2 output (nmol min 1 g 1 ) = 100 for U N-P uptake (nmol min 1 g 1 m+2 ) Statistical analyses were performed using In-STAT 1.14 software. We performed 3 or 4 separate liver perfusions for each experimental group outlined above. The Student s t test or ANOVA test was employed to compare two groups or differences among groups as needed. P < 0.05 was taken as indicating a statistically significant difference. RESULTS Nitrogen balance across the liver Atotal of 35 perfusions were performed. Oxygen consumption wasconstant ( µmol min 1 g 1 ) during perfusions, indicating the stability of the preparations. In perfusions containing AAM, all livers were presented with the same concentrations of substrates. Hence, when hepatic metabolite concentrations were assessed, it was possible to combine those with NH 4 Cl and with [2- N]- or [5- N]-glutamine. Table 1 shows the rates of nitrogen uptake and output across the liver during the course of perfusions. In perfusions with AAM, there was net uptake of glutamine-n (313 nmol of N min 1 g 1 ), alanine (1 nmol of N min 1 g 1 ), glycine (118 nmol of N min 1 g 1 )andaspartate

4 182 I. Nissim and others Table 2 Hepatic content of N -acetylglutamate and the primary donors of amino-n for ureagenesis Values are means + S.D. (n = 10 for AAM; n = 3for others), obtained from measurements of liver extract at the end of perfusions. Perfusions with glutamine (nmol min 1 g 1 ) Perfusions with aspartate (nmol min 1 g 1 ) Metabolite Control (+) AAM (+) AOA Control (+)Insulin (+) Glucagon N -Acetylglutamate * * Aspartate * Glutamate Alanine Glutamine * P < 0.05 compared with the respective control (in perfusions with either glutamine or aspartate). P < 0.05 compared with control group (perfusions with glutamine plus ammonia). (54 nmol of N min 1 g 1 ). The uptake of other amino acids was negligible and, therefore, results were not shown. An observation of special importance is that in perfusions with AAM, urea-n output was increased by approx. 55% (P < 0.05), when compared with perfusions with only glutamine and ammonia (Table 1). However, in perfusions with 1 mm AOA, both glutamine and ammonia uptakes were decreased and, subsequently, the outputs of urea-n, glutamate and alanine were significantly decreased (Table 1). The release in the effluent of urea-n, glutamine-n, alanine and glutamate represents the major nitrogenous output (the release of other amino acids was very small). Therefore we calculated the extent to which these compounds account for nitrogen balance across the liver. In perfusions with glutamine plus ammonia (control), AAM or with AOA, there was almost a complete recovery of nitrogen uptake by the release of urea-n, alanine and glutamate (Table 1). Metabolite levels in freeze-clamped livers Table 2 shows metabolite levels in freeze-clamped livers at the end of each perfusion. We present only those metabolites that are directly related to urea synthesis. Although the levels of amino acids (aspartate, glutamate, alanine or glutamine) tended to increase with AAM in the perfusate, these changes were not significant (Table 2). The aspartate level, however, was decreased by approx. 5-fold in perfusions with AOA. An observation of special interest is that the NAG level was significantly (P < 0.05) higher after perfusions with AAM, when compared with perfusions with only glutamine and ammonia (control, Table 2). In perfusions with [ N]aspartate and NH 4 Cl as the only nitrogen substrates, the levels of glutamate, alanine and glutamine were significantly lower compared with perfusions with AAM or glutamine plus ammonia (Table 2). The addition of glucagon or insulin had no effect on the level of glutamate, alanine or glutamine. However, the level of aspartate was significantly increased in perfusions with glucagon when compared with perfusions without the hormone. Similarly, in perfusions with [ N]aspartate plus glucagon, there was approx. a 2-fold increase (P < 0.05) in the NAG level compared with perfusions without glucagon (control) or perfusions with insulin (Table 2). Fate of the N-labelled glutamine and ammonia Figure 1 illustrates the Nenrichment of the primary metabolites (products) in freeze-clamped livers. In perfusions with AAM and NH 4 Cl, the enrichment of [ N]glutamate was approx. 26 MPE, similar to that in perfusions with [2- N]glutamine (Fig- Figure 1 Enrichment (MPE) of N-labelled amino acids in freeze-clamped livers at the end of perfusion (A) Perfusions with NH 4 Cl, amino acid mixture and unlabelled glutamine; (B) asin(a), but with [2- N]glutamine and unlabelled ammonia; (C) asin(b), but with [5- N]glutamine; (D) perfusions with [2- N]glutamine, ammonia and 1 mm AOA; (E) perfusion with [ N]aspartate, ammonia and ornithine; (F) asin(e) plusinsulin; and (G) asin(e) plus glucagon. Nenrichment in aspartate is presented in Table 3. Bars are means + S.D. from three livers. ures 1A and 1B). Glutamate was more heavily labelled by [2- N]glutamine than by [5- N]glutamine (Figures 1B and 1C). This is to be expected, as the PDG reaction produces [ N]glutamate. In perfusions with AAM, the percentage contributions of perfusate NH 4 Cl, [2- N]glutamine and [5- N]glutamine to the total hepatic glutamate pool were approx. 40, 34 and 9% respectively. [ N]Glutamate is derived from NH 4+ following reductive amination of α-kg via the GDH reaction [4,5]. The current observation indicates a net production of [ N]glutamate via the reductive amination of α-kg. The addition of AOA did not affect the relative formation of [ N]glutamate from [2- N]glutamine (Figure 1D). The ratio of Nenrichment in glutamate/ aspartate or glutamate/alanine always exceeds 1 (Figure 1 and

5 Sources of urea nitrogen 183 Table 3 Contribution of perfusate N-labelled ammonia, glutamine or aspartate to hepatic (Hep) aspartate or ammonia pool Values are means + S.D. from 3 or 4 perfusions; n.d., not detected. N-P [ N]Aspartate Contribution to [ N]Ammonia Contribution to Perfusate/ N-precursor* (MPE) (MPE) Hep aspartate (%) (MPE) Hep ammonia (%) (I) NH 4 Cl + AAM N/A 65 (II) [5- N]Glutamine + AAM (III) [2- N]Glutamine + AAM (IV) [5- N]Glutamine + NH 4 Cl (V) [2- N]Glutamine + NH 4 Cl Control (+)AOA (VI) [ N]Aspartate + NH 4 Cl Control N/A 9 n.d. n.d. (+)Insulin N/A 9 n.d. n.d. (+)Glucagon N/A 12 n.d. n.d. *(I)Perfusions with 0.3 mm N-labelled ammonia, a physiological mixture of amino acids (AAM) and unlabelled glutamine or 1 mm N-labelled glutamine (II or III) plus 0.3 mm NH 4 Cl and AAM as indicated under the Materials and methods section; (IV) perfusions with 1 mm [5- N]glutamine plus 0.3 mm NH 4 Cl [22]; (V) perfusions with 1 mm [2- N]glutamine plus 0.3 mm NH 4 Cl with or without AOA; and (VI) perfusions with 0.1 mm [ N]aspartate, 0.05 mm ornithine and 0.3 mm NH 4 Cl without hormone (control) or in the presence (+)ofinsulin or glucagon (10 7 M). N-P, N-labelled precursor 1 in freeze-clamped livers at the end of each perfusion. Calculation of percentage contribution as indicated under the Materials and methods section, using the mean value (MPE) in the precursor and products. In most cases, Nenrichment was similar in freeze-clamped liver or effluent ammonia at the steady state, i.e min of the perfusion. N/A, similar to 5 N-P value in the first column. Experiments with [5- N]glutamine have appeared previously in [22], and results are included here to facilitate comparison. Table 4 Contribution of perfusate N-labelled ammonia, glutamine or aspartate to [ N]urea isotopomers Values are means + S.D. from 3 or 4 perfusions; n.d., not detected. Transfer of P N into urea (%) Perfusate/ N-precursor* U m+1 U m+2 U m+1 U m+2 U m+1 U m+2 (I) NH 4 Cl + AAM (II) [5- N]Glutamine + AAM (III) [2- N]Glutamine + AAM (IV) [5- N]Glutamine +NH 4 Cl (V) [2- N]Glutamine + NH 4 Cl Control (+)AOA n.d n.d. 2 0 (VI) [ N]Aspartate + NH 4 Cl Control n.d n.d. 5 0 (+)Insulin n.d n.d. 8 0 (+)Glucagon n.d n.d *Experimental conditions are as described in Table 3. N-enrichment in singly labelled urea (U m+1 )ordoubly labelled urea (U m+2 ). Values are from effluent urea released at the steady state, i.e. between 40 and 70 min of the perfusion. U m+1 and U m+2 are expressed in terms of MPE. The product of total urea-n released at the steady state (Table 1), times MPE/100. U m+1 and U m+2 are expressed in terms of nmol of N min 1 g 1. Percentage of precursor-n uptake (P N )transferred into urea was calculated as indicated in the Materials and methods section. Experiments with [5- N]glutamine have previously appeared in [22], and results are included here to facilitate comparison. Table 3), indicating a probable precursor product relationship between glutamate and these amino acids, as well as a rapid equilibrium between glutamate and these amino acids via the corresponding aminotransferase reaction. In perfusions with AAM, the contribution of perfusate NH 4 Cl, [2- N]glutamine or [5- N]glutamine to the total hepatic aspartate pool was approx. 37, 24 and 10% respectively (Table 3). In the absence of AAM, the relative contribution of [2- N]glutamine to the hepatic aspartate pool was increased to 37%. However, the addition of AOA decreased the formation of [ N]aspartate from [2- N]glutamine by approx. -fold. Therefore the results demonstrate that glutamine-n (amino-n and amido-n) and perfusate ammonia contributed approx. 71% of aspartate nitrogen. Similar calculations for the sources of ammonia indicate that in experiments with AAM, [5- N]glutamine and [2- N]glutamine contributed approx. 35and4% respectively to the total ammonia pool (Table 3). The remaining 60 65% of the ammonia was derived from perfusate NH 4 Cl. Therefore in perfusions with AAM, approximately one-third of the ammonia was derived from glutamine and another two-thirds from perfusate ammonia. The incorporation of N-P into urea and the formation of [ N]urea isotopomers is presented in Table 4. The output

6 184 I. Nissim and others approx. 21 MPE of U m+1 and approx. 5 MPE of U m+2.similarly, there were no differences in the labelling of U m+1 and U m+2 from N-labelled glutamine in the presence or absence of AAM (Table 4). In perfusions with AAM, the percentage transfer of [5- N] or [2- N] into urea-n indicates that approx. 100% of N- labelled glutamine uptake was recovered in N-labelled urea isotopomers, whereas in perfusions without AAM only 50% of glutamine-n uptake was accounted for by the production of [ N]urea (sum of U m+1 and U m+2 )(Table4). In perfusions with [2- N]glutamine plus AOA, the output of [ N]urea (nmol of N min 1 g 1 )wassignificantly decreased when compared with perfusions without AOA. Only U m+1,with approx. 2 MPE, was detected in the presence of AOA (Figure 2B and Table 4). The current results demonstrate that the inhibition of the aminotransferase reaction decreased the conversion of [ N]glutamate into [ N]aspartate, thereby decreasing [ N]urea synthesis from [2- N]glutamine. Since the formation of [ N]glutamate from [2- N]glutamine or NH 4+ is strictly mitochondrial, the results suggest that the mitochondrial [ N]glutamate is an important source for the synthesis of cytosolic ASA. Figure 2 [ N]Urea production (sum of U m+1 and U m+2 ) during the course of liver perfusion (A) Perfusions with NH 4 Cl, AAM and unlabelled glutamine (, n = 3), with [2- N]glutamine and unlabelled ammonia (, n = 3) or with [5- N]glutamine and unlabelled ammonia (, n = 3). (B) Perfusions with only [2- N]glutamine and ammonia without (, n = 4) or with 1 mm AOA (, n = 3). (C) Perfusions with [ N]aspartate without hormone (, n = 3), with insulin (, n = 3) or with glucagon (, n = 3). The results are means + S.D. of [ N]urea (nmol of N min 1 g 1 ) during the course of the perfusion is presented in Figure 2. In the presence of AAM and NH 4 Cl, there was an immediate and massive production of [ N]urea over the range of nmol of N min 1 g 1 (Figure 2A). The [ N]urea output comprised approx. 43 MPE of U m+1 and approx. 23 MPE of U m+2.thesumofu m+1 and U m+2 (nmol of N min 1 g 1 )indicates that approx. 45% of the total urea-n was derived from perfusate NH 4 Cl (Tables 1 and 4). With N-labelled glutamine and AAM, the output of U m+1 + U m+2 was approx. 300 nmol of N min 1 g 1 at the steady state (between 30 and 70 min of perfusion), regardless of whether [5- N]- or [2- N] glutamine was used as the Nprecursor (Figure 2A and Table 4), indicating that approx. 30% of total urea-n output was derived from glutamine-n (sum of amino-n and amido-n). There were no differences in the labelling (MPE) of U m+1 and U m+2 with [5- N]- or [2- N]-glutamine as precursors, which comprised Fate of perfusate [ N]aspartate Perfusions with [ N]aspartate plus ammonia demonstrated that [ N]aspartate was taken up by the liver at the rate of nmol min 1 g 1 (Table 1). The uptake of [ N]aspartate was increased by approx. 2-fold in perfusions with insulin or glucagon. Perfusate [ N]aspartate was 0.1 mm, and in the effluent, 0.09, 0.07 and mm in control, perfusion with glucagon and perfusion with insulin respectively. This observation indicates that the liver took up approx. 10, 30 and 25% of perfusate aspartate in control, perfusions with glucagon, and perfusion with insulin respectively. The isotopic enrichment of hepatic [ N]aspartate at the end of the perfusions was approx. 9 MPE in control and in the presence of insulin, and 12 MPE after the addition of glucagon (P = 0.35), indicating that approx. 9 12% of the hepatic aspartate pool was derived from perfusate aspartate, regardless of the hormonal addition (Table 3). Figure 3 illustrates the output of the main N-labelled metabolites in the effluent during the course of the perfusion. With glucagon or insulin there was a higher (P < 0.05) isotopic enrichment in [ N]glutamate between 30 and 60 min (Figure 3, top panel). The increased [ N]glutamate enrichment was accompanied by an increased enrichment in U m+1,especially in perfusions with glucagon (Figure 3, bottom panel). The isotopic enrichment in U m+1 was 1, 2.1 and 1.7 MPE in control perfusions or perfusions with glucagon or insulin respectively. In control, the output of [ N]urea amounted to 3 5 nmol of N min 1 g 1.Thisvalue was increased to 8 9 nmol of N min 1 g 1 with insulin and to 20 nmol of N min 1 g 1 with glucagon, between 30 and 60 min of the perfusion (Figure 2C). In control perfusions, approx. 5% of [ N]aspartate uptake was transferred into U m+1.thisvalue was increased to 8 and 13% in perfusions with insulin or glucagon respectively (Table 3). Other metabolites of [ N]aspartate are [2- N]glutamine, [ N]alanine and [ N]glutamate, which accounted for 60% of [ N]aspartate uptake (calculated from data in Table 1 and Figure 3), despite hormonal treatment. The unaccounted portion of [ N]aspartate uptake may be utilized for the synthesis of purines and pyrimidines, which are in the range of µmol/g of tissue [27]. [2- N]Glutamine was the major product of perfusate [ N]aspartate. The isotopic enrichment (MPE) in [2- N]glutamine was approx. 2-fold higher than that in [ N]glutamate. The ratio [2- N]glutamine (effluent)/[ N]aspartate (extract) was 2:1 (Figure 3 and Table 3). Similarly, the ratio effluent [2- N]glutamine/

7 Sources of urea nitrogen 185 the liver and physiological flow direction [4,7,10,22]. The model also avoids the problem of substrate recycling from perivenous to periportal hepatocytes [9]. The chief findings of the present study are summarized in Scheme 1 and discussed below. Figure 3 Formation of N-labelled amino acids and singly labelled urea (U m+1 ), during the course of perfusion with [ N]aspartate Nenrichment (MPE) was obtained in experiments without hormone (, n = 4), with glucagon (, n = 3) or with insulin (, n = 3). Bars are means + S.D. from three livers. [ N]glutamate was approx. 2:1. These calculations suggest that N-labelled glutamate and glutamine were formed in different compartments in the liver. The effluent [2- N]glutamine might have been formed in the perivenous hepatocytes, whereas the effluent [ N]glutamate may represent a mixture of unlabelled glutamate released from periportal hepatocytes and [ N]glutamate (formed from [ N]aspartate) in perivenous hepatocytes. The unlabelled glutamate was probably formed from perfusate ammonia via the reductive amination of α-kg mainly in the periportal region, whereas the uptake of [ N]aspartate and its transamination into glutamate was predominantly in the perivenous cells, the site of hepatic glutamine synthetase. This observation is in accordance with the notion of liver zonation [3,9,28]. DISCUSSION The present study focuses on some specific questions. (i) What is the role of mitochondrial glutamate metabolism, via aspartate aminotransferase or the GDH reaction, in furnishing aspartate for ASA synthesis? (ii) What is the relative incorporation of perfusate NH 4 Cl, [2- N]glutamine or [5- N]glutamine into CP or aspartate-n, and thereby, into [ N]urea isotopomers? (iii) What is the source of the aspartate that is required for the synthesis of ASA? Does this aspartate form via the Mit-Asp-AT or the cytosolic aspartate aminotransferase reaction? (iv) What is the extent of perfusate [ N]aspartate uptake and its incorporation into urea? To address these questions, we took advantage of GC MS methodology and analysis of the N-P/ N-product relationship as in our previous studies [5 7]. We used a liver-perfusion system containing a physiological mixture of amino acids and ammonia that resembles the milieu to which the liver is exposed in vivo, with only one precursor labelled with N. This approach offers an important opportunity to determine the source(s) of nitrogen used for the synthesis of CP or ASA under physiological conditions. The single-pass perfusion system with an antegrade flow was used, since this model preserves the normal lobular microcirculation of Metabolism of perfusate NH 4 Cl Results obtained from experiments with AAM and NH 4 Cl indicated that the net flux through the GDH reaction is predominantly towards glutamate production. There was a net production of [ N]glutamate from NH 4 Cl via the reductive amination of α-kg (Figure 1). Current observations suggest that the mitochondrial GDH reaction may have a key role in the regulation of urea synthesis by furnishing glutamate and, thereby, aspartate-n for the synthesis of ASA. This conclusion is derived on the basis of the following findings. With a physiological mixture of amino acids, approx. 65% of the hepatic [ N]ammonia and 37% of the hepatic pool of aspartate-n was derived from perfusate NH 4 Cl (Table 3). Subsequently, the output in the effluent of [ N]urea comprises approx. 42 MPE of U m+1 and approx. 23 MPE of U m+2 (Scheme 1 and Table 4). The results indicate that the isotopic enrichment in U m+1 represents the direct incorporation of perfusate NH 4 Cl into CP. In addition, results in Figure 1 and Table 3 demonstrate a 1:1:1 ratio among the isotopic enrichments in [ N]glutamate, [ N]aspartate and U m+2. Therefore U m+2 was probably formed following a sequence of metabolic reactions: (i) incorporation of NH 4 Cl into α-kg to form [ N]glutamate via the GDH reaction; (ii) transamination of [ N]glutamate to form [ N]aspartate; and (iii) translocation of mitochondrial [ N]- aspartate to cytosol and its incorporation into ASA (Scheme 1). Thus the current findings suggest that aspartate required for ASA synthesis is formed in the mitochondria, and is unidirectionally transported out into the cytosol and incorporated into urea. This conclusion is in agreement with the study of Meijer et al. [13], as well as recent studies indicating that AGTrs may play an important role in the regulation of ureagenesis by transporting the mitochondrial aspartate into the cytosol for the synthesis of ASA [11,12]. The net flux of the GDH reaction towards glutamate production may have multiple roles in the regulation of systemic ammonia detoxification. The formation of glutamate from ammonia may: (i) serve as a scavenger for excess ammonia; (ii) improve the availability of mitochondrial [glutamate] for synthesis of NAG; and (iii) lead to the formation of aspartate via the Mit-Asp-AT reaction. Increased synthesis of NAG and availability of aspartate would be expected to stimulate ammonia detoxification via urea synthesis. In addition, the current findings may shed new light on the mechanism responsible for HA in cases such as the congenital GDH gain-of-function mutation. Children suffering from this defect experience a 2 10-fold increase in blood ammonia concentration [,29]. It has been speculated that stimulated oxidative deamination of glutamate via the hepatic GDH reaction, and thereby decreased NAG synthesis, is the primary cause for the impaired ureagenesis and HA [,29]. However, as yet, no data have been reported to support this hypothesis. The current findings suggest that, in addition to a possible decrease in NAG levels, the gain-offunction mutation of GDH may lead to decreased net flux towards glutamate synthesis, thereby diminishing production of aspartate required for ASA synthesis, since approx. 37% of the hepatic aspartate pool was derived from perfusate ammonia (Table 3), secondary to glutamate formation via the GDH reaction. Metabolism of perfusate N-labelled glutamine With the addition of AAM and 0.3 mm NH 4 Cl, the contribution of [2- N]glutamine and [5- N]glutamine to the total hepatic

8 186 I. Nissim and others Scheme 1 Schematic representation of the primary amino-n and ammonia-n flow into various metabolites in either the cytosolic (left) or mitochondrial (right) compartment of the periportal hepatocytes The scheme depicts the perfusion with NH 4 Cl, unlabelled glutamine and AAM, as indicated under the Materials and methods section. Glutamine is metabolized in the mitochondria via the PDG reaction (1) to form [ 14 N]glutamate and 14 NH 4 +.Simultaneously, the perfusate NH 4 Cl mixes with the unlabelled 14 NH 4 +,resulting in a mitochondrial ammonia pool with Nenrichment of approx. 64 MPE. This NH 4 + pool is incorporated into CP via the CP synthetase-i reaction (4). [ N]CP is then incorporated into ornithine to form [ N]citrulline via the ornithine carbamoyltransferase reaction (5). Subsequently, [ N]citrulline is incorporated into the urea cycle and forms U m+1, which comprises approx. 42 MPE. Concurrently, the mitochondrial pool of NH 4 + is utilized to form [ N]glutamate via the reductive amination of α-kg through the GDH reaction (2), thus creating a mitochondrial glutamate pool with Nenrichment of 26 MPE. This glutamate pool comprises [ 14 N]glutamate formed from unlabelled glutamine and [ N]glutamate formed from NH 4 +. The mitochondrial [ N]glutamate is utilized to form [ N]aspartate via the Mit-Asp-AT reaction (3). The mitochondrial [ N]aspartate is transported via the AGTr to the cytosol and is then incorporated into ASA and form U m+2, which comprises approx. 23 MPE. Some of the aspartate may be formed in the cytosol via the Cit-Asp-AT reaction (7). However, [ 14 N]aspartate formed via the Cit-Asp-AT reaction does not incorporate into urea, as indicated by the ratio between U m+2 MPE (product) and [ N]aspartate (precursor) MPE. Note that the metabolic reactions and nitrogen flow would be as indicated in this scheme, irrespective of the fact whether N-labelled ammonia or glutamine was used as N-P. However, the degree of the isotopic enrichment in glutamate, ammonia, aspartate, and thereby U m+1 and U m+2,will be different with different N-Ps, as illustrated in Tables 3 and 4. aspartate pool was approx. 24 and 10% respectively. Simultaneously, the contribution to the ammonia pool was approx. 35 and 4% from [5- N]glutamine and [2- N]glutamine respectively (Table 3). Subsequently, the output in the effluent of [ N]urea isotopomers comprised approx. 21 MPE of U m+1 and approx. 5 MPE of U m+2 (Table 4). [5- N]Glutamine primarily entered into urea via NH 4+ incorporation into CP, whereas [2- N]glutamine was predominantly incorporated via [ N]aspartate [7]. This is evident from the relative enrichments of N-labelled aspartate and ammonia generated from either [2- N]- or [5- N]-glutamine (Table 3). Calculation of the individual urea isotopomers as described previously [4] by using the Nenrichment in aspartate in freeze-clamped livers and effluent NH 4+ after 70 min of perfusion with either [2- N]glutamine, [5- N]glutamine or NH 4 Cl (Table 3), indicated an excellent agreement between predicted and observed U m,u m+1 and U m+2 (results not shown). Therefore our theoretical model for the incorporation of labelled nitrogen into urea [4] is further substantiated in perfusion systems that replicate the mixture of amino acids taken up by the liver in vivo. Data in Tables 1 and 4 and Figure 2 indicate that approx. 30% of total urea nitrogen output was derived from glutamine-n (sum of the amino-n and amido-n) in perfusions with AAM and approx. 36% without AAM. Similarly, the relative (MPE) contribution of perfusate N-labelled glutamine to [ N]urea isotopomer production shows little difference in the presence or absence of AAM (Table 4). These results further substantiate our previous findings that glutamine is the primary amino acid utilized for urea synthesis [5,6], regardless of the fact as to whether glutamine is supplemented as the sole amino acid, or in the presence of a physiological mixture of amino acids. Metabolism of perfusate [ N]aspartate Aspartate is required for the cytosolic synthesis of ASA in periportal hepatocytes [1]. However, the sources of cytosolic aspartate are vague, especially since the periportal uptake of aspartate is negligible relative to the rate of urea synthesis [8]. In the present study, we assessed the extent to which [ N]aspartate is taken up by the liver and subsequently incorporated into [ N]urea. In addition, we investigated whether glucagon or insulin modulates hepatic aspartate uptake and metabolism. The present observations indicate that the uptake of perfusate aspartate is quite limited relative to the rate of urea synthesis (Figure 2 and Table 1), and, therefore, cannot provide sufficient aspartate-n to maintain the synthesis of urea in periportal hepatocytes. However, our investigation demonstrates that approx. 10% of the hepatic aspartate pool was derived from perfusate [ N]aspartate (Table 3), and that perfusate aspartate contributed approx. 5, 8 and 13% of U m+1 output

9 Sources of urea nitrogen 187 in control, perfusion with insulin and perfusion with glucagon respectively (Table 4). The portion of perfusate aspartate transferred to [ N]urea synthesis may be considered insignificant in the normal physiological metabolic state. However, this amount could be crucial for ammonia detoxification in pathological situations such as the congenital HI/HA [,29]. In this case, the mitochondrial glutamate conversion into aspartate-n could be limited owing to the increased net flux through oxidative deamination of glutamate via the GDH reaction []. Insight into the regulation of urea synthesis and nitrogen balance across the liver We noted greater synthesis and release of urea in perfusions using AAM compared with perfusions using only glutamine plus ammonia (Table 1 and Figure 2). Similarly, in perfusions using [ N]aspartate plus glucagon, the output of urea was significantly higher compared with perfusions without glucagon (Table 1). Although the augmented urea synthesis in perfusions with AAM may be mediated by the provision of additional nitrogenous substrates, in perfusions with only aspartate plus glucagon or with AAM, the increase in urea synthesis was associated with increased NAG levels in freeze-clamped livers (Tables 1 and 2). A regression analysis between NAG levels and urea synthesis, under all experimental conditions (data of Tables 1 and 2), demonstrates a linear correlation (r = 0.77), suggesting that the increased urea synthesis in perfusions with either AAM or aspartate plus glucagon was mediated via increased synthesis and availability of NAG. We have demonstrated previously that glucagon stimulates NAG synthesis from perfusate glutamine [7]. Our previous results suggest that the increased NAG synthesis was secondary to the stimulation of glutaminase by glucagon and increased availability of glutamate [7]. However, results in Table 2 demonstrate that although [glutamate] was approx. 4-fold lower in perfusions using aspartate, there were only minor changes in the levels of NAG, when compared with perfusions using glutamine plus ammonia (control group). Thus, in perfusions with [ N]aspartate, the increased hepatic NAG level by glucagon is not mediated via an increased flux through glutaminase, but probably via direct action of glucagon on NAG synthesis. Alternatively, increased hepatic NAG level and total urea output could have resulted by glucagon stimulation of hepatic proteolysis as indicated previously [21]. Similarly, the increased NAG levels in perfusions with AAM were not necessarily due to an increased availability of glutamate, since hepatic [glutamate] was similar in all perfusions with glutamine (Table 2). However, since 0.05 mm arginine was included in perfusions with AAM, the formation of agmatine via the arginine decarboxylase would be expected to stimulate NAG synthesis, as demonstrated previously [22]. In perfusions with [ N]aspartate and unlabelled NH 4 Cl there was a negative nitrogen balance across the liver, regardless of the hormonal addition (Table 1). Therefore perfusate aspartate failed to provide aspartate-n at a rate necessary to maintain urea synthesis. The calculations indicate that approx % of N output was derived from intra-hepatic sources (i.e. proteolysis) to satisfy the balance between N uptake and N output (Table 1). However, in perfusions with glutamine, with or without the supplementation of AAM, there was almost a complete balance between nitrogen uptake and output (Table 1). In the mitochondria, glutamine is metabolized to ammonia and glutamate, after which the latter is transaminated to form aspartate. The results indicate that approx. 34 and 47% of the hepatic aspartate pool was derived from perfusate glutamine (sum of amino-n and amido-n) in perfusions with or without AAM respectively (Table 3). This amount, together with approx. 37% of aspartate formed from perfusate ammonia (Table 3), indicates that 75 80% of aspartate-n utilized for urea synthesis was derived from perfusate ammonia and glutamine. The remaining portion (i.e %) ofaspartate-n may be derived from other amino acids such as alanine and glycine (Table 1). Therefore, as indicated in our previous study [10], the current findings suggest that the intensity of hepatic proteolysis is determined, at least in part, by the need to furnish aspartate nitrogen for urea synthesis. In conclusion, the present study suggests that mitochondrial glutamate formed from ammonia via the reductive amination of α-kg, as well as from glutamine following the activity of the PDG reaction, may provide between 75 and 80% of cytosolic aspartate-n required for ASA synthesis. In addition, glutamine and ammonia nitrogen provide approx. 75% of total urea nitrogen output. The current findings provide new insight into the mechanism of defective ammonia detoxification and urea synthesis in cases such as the congenital HI/HA. The results suggest that decreased net flux of the GDH reaction towards glutamate formation may diminish mitochondrial [glutamate], thus limiting the synthesis of NAG and/or ASA. This work was supported by The National Institutes of Health grants DK and CA (to Itzhak Nissim). REFERENCES 1 Meijer, A. J., Lamers, W. H. and Chamuleau, R. A. F. M. (1990) Nitrogen metabolism and ornithine cycle function. Physiol. Rev. 70, Yang, D., Hazey, J. W., David, F., Singh, J., Riochum, R., Sreem, J. M., Halperin, M. L. and Brunengraber, H. (2000) Integrative physiology of splanchnic glutamine and ammonium metabolism. Am. J. Physiol. 278, E469 E476 3 Haussinger, D. (1983) Hepatocyte heterogeneity in glutamine and ammonia metabolism and the role of intercellular glutamine cycle during ureogenesis in perfused rat liver. J. Biochem. (Tokyo) 133, Brosnan, J. T., Brosnan, M. E., Charron, R. and Nissim, I. (1996) A mass isotopomers study of urea and glutamine synthesis from N-labeled ammonia in the perfused rat liver. J. Biol. Chem. 271, Nissim, I., Cattano, C., Nissim, I. and Yudkoff, M. (1992) Relative role of the glutaminase, glutamate dehydrogenase, and AMP-deaminase pathway in hepatic ureagenesis: studies with NandGC-MS. Arch. Biochem. Biophys. 292, Nissim, I., Yudkoff, M. and Brosnan, J. T. (1996) Regulation of [ N]urea synthesis from [5- N]glutamine: role of ph, hormones and pyruvate. J. Biol. Chem. 271, Nissim, I., Brosnan, M. E., Yudkoff, M., Nissim, I. and Brosnan, J. T. (1999) Studies of hepatic glutamine metabolism in the perfused rat liver with N-labeled glutamine. J. Biol. Chem. 274, Stoll, B., McNelly, S., Buscher, H. P. and Haussinger, D. (1991) Functional hepatocyte heterogeneity in glutamate, aspartate and α-ketoglutarate uptake: a histoautoradiographical study. Hepatology 13, Haussinger, D. and Gerok, W. (1983) Hepatocyte heterogeneity in glutamate uptake by isolated perfused rat liver. Eur. J. Biochem. 136, Brosnan, J. T., Brosnan, M. E., Yudkoff, M., Nissim, I., Daikhin, Y., Lazarow, A., Horyn, O. and Nissim, I. (2001) Alanine metabolism in the perfused rat liver. J. Biol. Chem. 276, Palmieri, L., Pardo, B., Lasorsa, F. M., del Arco, A., Kobayashi, K., Iijima, M., Runswick, M. J., Walker, J. E., Saheki, T., Satrustegui, J. et al. (2001) Citrin and aralar1 are Ca 2+ -stimulated aspartate/glutamate transporters in mitochondria. EMBO J. 20, Iijima, M., Jalil, A., Begum, L., Yasuda, T., Yamaguchi, N., Xian, Li, M., Kawada, N., Endou, H., Kobayashi, K. and Saheki, T. (2001) Pathogenesis of adult-onset type II citrullinemia caused by deficiency of citrin, a mitochondrial solute carrier protein: tissue and subcellular localization of citrin. Adv. Enzyme Regul. 41, Meijer, A. J., Gimpel, J. A., Deleeuw, G., Tischler, M. E., Tager, J. M. and Williamson, J. R. (1978) Interrelationships between gluconeogenesis and ureagenesis in isolated hepatocytes. J. Biol. Chem. 253, Nissim, I. (1999) Newer aspects of glutamine/glutamate metabolism: the role of acute ph changes. Am. J. Physiol. 277, F493 F497 Kelly, A. and Stanley, C. A. (2001) Disorders of glutamate metabolism. Ment. Retard. Dev. Disabil. Res. Rev. 7,