Glycogenesis from Glucose and Ureagenesis in Isolated Perfused Rat Livers INFLUENCE OF AMMONIUM ION, NORVALINE, AND ETHOXYZOLAMIDE*

The probable involvement of hepatic carbamyl-P in the reciprocal relationship between hepatic ureagenesis and glycogenesis from glucose was explored. Isolated perfused liver preparations from 48-h fasted rats were employed. Moderate (9.2 mm) and relatively high levels of glucose (34 mm) were perfused. Hepatic glycogenesis, glucose-6-P, carbamyl-P, and citrulline levels, hepatic urea formation, and ureagenesis based upon perfusate urea levels were measured. Experimental probes se- lected to modify hepatic ureagenesis and carbamyl-P production and utilization included (a) NH4C1, main- tained at 5 mm by continuous infusion (NH; is a substrate for carbamyl-P synthase I and glutamate dehy- drogenase); (b) norvaline, an inhibitor of ornithine transcarbamylase which catalyzes the first committed step in the urea cycle; and (c) ethoxyzolamide, an inhibi- tor of carbonic anhydrase which produces HCO;, an essential substrate for carbamyl-P synthase I. NH; in- creased ureagenesis and decreased glycogenesis. The inclusion of norvaline value when present in active form. Fructose-1-P binds to the regulatory protein and converts it to the active, inhibitory form (39) while Pi seems to be the major metabolite, at physiologic concentration, capable of converting the regula- tory protein to its non-inhibitory form (41). Thus, it follows that, under appropriate cellular circumstances, the active, inhibitory form of this regulatory protein might bind to glucokinase and, while inhibiting activity, it would produce an increase in the value for the apparent K,n, because of its competitive inhibitory nature.

The probable involvement of hepatic carbamyl-P in the reciprocal relationship between hepatic ureagenesis and glycogenesis from glucose was explored. Isolated perfused liver preparations from 48-h fasted rats were employed. Moderate (9.2 mm) and relatively high levels of glucose (34 mm) were perfused. Hepatic glycogenesis, glucose-6-P, carbamyl-P, and citrulline levels, hepatic urea formation, and ureagenesis based upon perfusate urea levels were measured. Experimental probes selected to modify hepatic ureagenesis and carbamyl-P production and utilization included (a) NH4C1, maintained at 5 mm by continuous infusion (NH; is a substrate for carbamyl-P synthase I and glutamate dehydrogenase); ( b ) norvaline, an inhibitor of ornithine transcarbamylase which catalyzes the first committed step in the urea cycle; and ( c ) ethoxyzolamide, an inhibitor of carbonic anhydrase which produces HCO;, an essential substrate for carbamyl-P synthase I. NH; increased ureagenesis and decreased glycogenesis. The inclusion of norvaline with NH; decreased ureagenesis and increased glycogenesis. Ethoxyzolamide with or without NH: inhibited both ureagenesis and glycogenesis, and decreased the hepatic glucose-6-P level. Glycogenesis was greater at 34 mm than 9.2 mm glucose, increased in norvaline-containing preparations correlative with increased availability of carbamyl-P, and decreased when carbamyl-P formation was inhibited by ethoxyzolamide. Kinetic analysis indicated a K m ,~l C of 31 m~ for glucose phosphorylation preliminary to glycogenesis. Glycogen formation via the "indirect pathway" (Le. involving extrahepatic glycolysis, transport of lactate to the liver, and glyconeogenesis therefrom) was quantitatively insufficient to account for the observed glycogenesis. Glucokinase is contraindicated by the inverse relationship between hepatic glycogenesis and ATP availability in the ethoxyzolamide-treated preparations. In contrast, carbamyl-P:glucose phosphotransferase activity of the glucose-6-phosphatase system has the characteristics to bridge hepatic ureagenesis and glycogenesis. ureagenesis and glycogenesis from glucose, and suggest that the reciprocity of these responses derives from the relative availability of carbamyl-P for the two processes (1). For example, the addition of glutamine to isolated, perfused liver preparations, with glucose and 3-mercaptopicolinate present, stimulates ureagenesis but not glycogenesis. In contrast, added proline stimulates glycogenesis from glucose, but not ureagenesis.
The role of carbamyl-P was studied here by determining ureagenesis and glycogenesis from glucose in response to: ( a ) added NH4C1, a substrate for hepatic carbamyl-P synthase I (2, 3) and glutamate dehydrogenase (4); ( b ) norvaline, an inhibitor of ornithine transcarbamylase (5, 6) which catalyzes the first committed step in the urea cycle; and (c) ethoxyzolamide, an inhibitor of carbonic anhydrase (21, and consequently an inhibitor of carbamyl-P synthesis by carbamyl-P synthase I (6, 7) which requires HCO, as substrate. Our focus (1) upon carbamyl-P as common substrate for both the urea cycle a n d glucose phosphorylation preliminary to glycogenesis from glucose by carbamyl-P:glucose phosphotransferase activity of the glucose-6-phosphatase system (8,9) was prompted by earlier work of Tremblay and colleagues (10,11).
They showed that carbamyl-P generated by carbamyl-P synthase I in mitochondria can become available for pyrimidine biosynthesis in the cytosol. Regulation of pyrimidine synthesis could still be controlled through feedback inhibition by CTP of aspartate transcarbamylase and CTP synthase (12).
Results of the present studies support the previously stated hypothesis (1) that carbamyl-P of mitochondrial origin is important not only as a substrate for hepatic urea synthesis, but also (under some circumstances) as a substrate for glucose phosphorylation via phosphotransferase activity of the glucose-6-phosphatase system as an essential event preliminary to the formation of glycogen from glucose in the perfused liver of the 48-h fasted rat.
EXPERIMENTAL PROCEDURES Materials-Fine chemicals of analytical reagent grade were from sources cited earlier (13). Ethoxyzolamide (6-ethoxyzolamide) was from Sigma. Solutions were prepared in doubly glass-distilled water. Reagent solutions used in perfusates were adjusted to pH 7.4 with concentrated NaOH. Male albino rats of Sprague-Dawley strain (obtained from Harlan Sprague Dawley, Madison, WI), 55-65 days old and weighing between 180 and 250 g, served as liver donors. Red blood cell donors (13) were retired male breeders from the same source. Animals were housed in temperature-constant quarters under a 12-h lighUl2-h dark cycle. They were maintained on tap water and Purina Laboratory Chow, ad libitum. Liver donors were deprived of food for 48 h prior to removal of livers to deplete endogenous hepatic glycogen (14). Livers were removed between 7 and 8 a.m. and immediately perfused.
Liver Perfusions-Isolated livers were perfused at 37 2 0.5 "C by the recycling system ofAlvares and Nordlie (13). The perfusate consisted of rat erythrocytes suspended in Krebs-Ringer bicarbonate buffer (15) (pH  I Effects of ammonium ion, norvaline, and ethoxyzolamide on glycogenesis, ureagenesis, a n d related parameters in isolated livers of 48-h fasted rats perfused with 9.2 mM D-glucose Livers were perfused as described in the text. The perfusion medium for all four groups contained approximately 9.2 m glucose (see Table for  exact values). As indicated, NH4Cl was maintained a t 5 m throughout by initial addition and continuous infusion. Norvaline (10 m) or ethoxyzolamide (3 m) was included as indicated. All metabolite concentrations or rates of formation are expressed on a per g liver basis. Mean 2 S.E. values are presented. N = 14 livers in Groups 1-3, and 10 livers in Group 4. Other details are given in the text. Hepatic glycogen formation was determined in the 44-min period following addition of glucose plus norvaline, or ethoxyzolamide, as indicated.
at the conclusion of perfusion. p < 0.05 for NH4C1-infused group compared with unsupplemented group. p < 0.05 for indicated group compared with NH4C1-infused group.
7.4; hematocrit 19-21%), which contained bovine serum albumin (3%, w/v) and heparin (6000 units/100 ml perfusate). This basic perfusate solution was warmed to 37 "C, the liver was attached, and the system was allowed to equilibrate for 10 min. A small (-0.5 g) liver biopsy sample was taken at this point for analysis of basal glycogen content.
Glucose (to 9.2 or 34 m, along with either norvaline (to 10 m), ethoxyzolamide (to 3 m), or the vehicle (isotonic saline), was added and perfusion was continued for 10 min. At this point, a sample of perfusate was taken for basal urea analysis and immediately NH,C1 (as a concentrated solution) was added to the perfusate (to 5 m), and continuous infusion of NH4Cl solution into the perfusate was begun to maintain the perfusate NH; concentration at 5 mM. Perfusion was continued for another 34 min, a t which time a sample of perfusate for urea analysis was taken, the perfusion terminated, and the liver immediately removed for metabolite (glycogen, glucose-6-P, urea, citrulline, and carbamyl-P) analysis. The flow rate of perfusion was maintained at 3.5 ml/min.g liver (13); adjustment was made when the first biopsy sample was removed. All liver biopsy samples, including those at the termination of perfusion, were frozen in liquid nitrogen (13) immediately after removal.
Measurement of Metabolites-Perfusate urea was measured according to Kerscher and Ziegenhorn (16). Perfusate samples were added to equal volumes of 10% (w/v) trichloroacetic acid and centrifuged at 500 x g for 25 min. The resulting supernatant was removed, pH adjusted to 7.0 with concentrated KHC03, and urea determined (16).
Portions of frozen liver biopsy samples were homogenized either in 10% (w/v) trichloroacetic acid (1 g of livedm1 trichloroacetic acid solution) for urea analysis, or in 0.6 M perchloric acid solution (0.2 g of liver/l.5 ml perchloric acid solution) for measurement of glycogen and glucose-6-P (14). For urea analysis, the trichloroacetic acid homogenates were centrifuged for 15 min at 12,900 x g, the pH of the supernatant was adjusted to 7.0 with concentrated KHC03, and assays were performed as described in Ref. 16. Glycogen (17) and glucose-6-P (18) were measured in perchloric acid supernatants as described in the indicated references.
Citrulline and carbamyl-P were measured, beginning with frozen liver samples, exactly as described by Cohen et al. (19). Thawed liver samples were homogenized with 300 m mannitol in 2 m HEPES buffer, pH 7.4 (19). just prior to assay. Standard recovery curves were run with each assay.
Enzyme Assays-Glucokinase plus hexokinase was assayed in a high-speed supernatant fraction as described by Alvares and Nordlie (13) and activities of the glucose-6-phosphatase system were assayed in microsomal preparations as described by Foster et al. (20). All assays were for 10 min a t 30 2 0.1 "C at pH 7.4 in 40 m HEPES buffer. Assay mixtures for the former contained 5 m sodium ATP, 7.5 m MgCl,, 30 mM D-glucose, and 0.60 mg of cytosolic protein in 1.5-ml total volume. Carbamyl-P:glucose phosphotransferase assays contained 1 m lithium carbamyl-P, 30 mM o-glucose, and 0.20-0.35 mg of microsomal protein in 1.2-ml total volume. Glucose-6-P phosphohydrolase assays contained 1 mM sodium glucose-6-P and 25 pg of microsomal protein in 0.3-ml total volume. Ethoxyzolamide was 3 mM and NH,Cl was 5 mM when present. NaCl served as control. Activities of the glucose-6-phosphatase system were adjusted for intactness as described by Arion et al. (21).

RESULTS
The results of perfusions with 9.2 and 34.0 mM glucose are presented in Tables I and 11, respectively. Hepatic glycogen formed in the 44 min following addition of glucose; perfusate urea formation was in the 34 min following addition of NH4Cl to the system; and levels of hepatic metabolites glucose-6-P, citrulline, carbamyl-P, and urea at the termination of perfusion are presented (Tables I and 11). All data were adjusted to a "per g liver" basis. Statistical analysis was as before (22).
Effects of Ammonium Zon-The maintenance of NH; concentration a t 5 mM by continuous infusion into the perfusate (Tables I and 11, Group 2) produced a decrease in glycogenesis from glucose, and an increase in ureagenesis (as indicated by increases in perfusate urea accumulation and in hepatic urea and citrulline, an intermediate of the urea cycle, all compared with control livers (Group 1)). The effects were observed at both concentrations of glucose. Hepatic carbamyl-P was decreased with NH; addition (9.2 M glucose present). Hepatic glucose-6-P was unchanged. In supplemental studies (results not shown), 5 mM NH4Cl was shown to have no significant effect on glucokinase plus hexokinase or on carbamyl-P:glucose phosphotransferase or glucose-6-P phosphohydrolase activity of glucose-6phosphatase.
Effects of Norvaline-The effect of norvaline, a non-metabolizable analog of valine (6), was determined in perfused livers with NH; infusion into perfusate (Tables I and 11, Group 3). Relative to perfused livers with NH; infusion (Group 21, norvaline increased glycogenesis and concomitantly decreased ureagenesis as indicated by lowered hepatic urea, hepatic citrulline, and perfusate urea values. Carbamyl-P increased with 34 mM glucose. Glucose-6-P levels were unchanged. Effects of Glucose Concentration'-There was an increase in glycogenesis from glucose and in hepatic glucose-6-P levels hepatic glucose phosphorylation is rate-limiting for glycogenesis from An analysis of this sort is predicated upon the presumption that glucose. Studies of Nordlie et al. (14) validate this assumption with livers from 48-h fasted rats perfused in our system. They demonstrated a hyperbolic relationship between glycogenic rates and glucose concentrations ranging from 5.4 to 67 m, and showed a progressive increase in hepatic glucose-6-P concentration with each increment in concentration of glucose perfused, ranging from 5.4 m glucose through 31 m glucose. We also observed these same glucose concentration-dependent increases in hepatic glucose-6-P here, with both 9.2 and 34 m glucose (Tables I and 11, respectively). These observations indicate that with our mainly through a metabolic "push" (Le. exerted at the glucose-phospho-perfusion model, the "glucose effect" upon glycogenesis from glucose is rylative step) rather than by a "pull" (i.e. exerted through glucosedependent activation of glycogen synthase (14)). With the pull effect Effects of ammonium ion, norvaline, and ethoxyzolamide on glycogenesis, ureagenesis, and related parameters in isolated livers of 48-h fasted rats perfused with 34 mM D-glucose Livers were perfused as described in the text. The perfusion medium for all four groups contained approximately 34 m glucose (see Table for exact values). As indicated, NH&l was maintained at 5 m throughout by initial addition and continuous infusion. Norvaline (10 m) or ethoxyzolamide (3 m) was included as indicated. All metabolite concentrations or rates of formation are expressed on a "per g liver" basis. Mean k S.E. values are presented. n = 8 livers in Group 1, 15 in Group 2, 11 in Group 3, and 8 in Group 4. Other details are as in Table I with 34 mM (Table 11) compared to 9.2 mM glucose (Table I) in perfusates with the unsupplemented, NH;-supplemented, and norvaline-treated preparations. Ureagenesis with NH;infusion was higher with 9.2 than 34 mM glucose present. Hepatic carbamyl-P levels were consistently higher with 9.2 than with 34 mM glucose (unsupplemented, NH;-supplemented, and norvaline containing preparations).
The ratios of glycogenesis from glucose with 34 mM glucose t o those with 9.2 mM glucose were calculated from data in Tables  I and 11; i.e. for each of Groups 1-3, the A hepatic glycogen values determined with 34 mM glucose (Table 11) were divided by A hepatic glycogen values determined with 9.2 mM glucose (Table I). Ratios determined in this way are 3.1, 2.2, and 2.2 with unsupplemented, NH;-supplemented, and norvalinesupplemented NH;-infused livers, respectively. Assuming a "traditional" K,,,, G~~ value of 6.0 mM (25) for glucokinase, we calculate an activity ratio, i.e. glucokinase activity with 34 mM glucose/glucokinase activity with 9 mM glucose, of 1.4. The activity ratios calculated from data in Tables I and 11, above, are greater than this value, indicating the involvement of an enzyme or enzymes with apparent Km,Glc values considerably greater than 6 m~ in hepatic glucose phosphorylation preliminary to glycogenesis.
The K,,, value for glucose as precursor for glycogenesis was calculated for the condition where carbamyl-P was most available for non-urea cycle functions, i.e. with NH4Cl plus norvaline present. This was done by substituting the two glucose concentration values and corresponding values for glycogenesis in Tables I and I1 into the simple Michaelis-Menten equation and solving simultaneously the two equations thus generated.
A K,,,, value of 31 mM was obtained.
Comparison of Relative Ureagenic and Glycogenic Values Obserued with 9 a n d 34 mM Glucose-The relative tendency for ureagenesis and glycogenesis with 34 compared with 9 mM glucose, respectively, is made clear in Table 111. Glycogenic values are from Tables I and I1 and ureagenic values are based on A perfusate urea values from Tables I and 11, taking into account that the perfusate volume was 100 ml and adjusting for 44 min rather than 34 min to permit direct comparison with glycogenic rates. Summations of ureagenesis plus glycogenesis are also presented in Table 111, where ureagenic and glycogenic values, as percent of these summation values, are also included. This treatment presupposes that the major, competing uses of carbamyl-P are for ureagenesis and for glucose phospredominant, liver glucose-6-P levels should decrease in a glucose concentration-dependent fashion (14, 23, 24). phorylation preliminary to glycogenesis. Furthermore, it is based on the presumed use of one carbamyl-P molecule per turn of the urea cycle and/or one carbamyl-P molecule per glucosyl unit incorporated into glycogen under the influence of carbamyl-P, and the assumption that ureagenesis remains linear for at least 44 min in our system.
Ureagenesis, calculated as the percent of summation values, was consistently lower with 34 mM glucose than with 9.2 mM glucose. With both concentrations of glucose, the percent of summation values for ureagenesis increased with NH; infusion and decreased with norvaline, while the inverse responses were seen for glycogenesis. Only with the higher (34 mM) concentration of glucose, with NH; limiting (Group l), did glycogenesis exceed ureagenesis.
Effects of Ethoxyzolamide-Ethoxyzolamide, an inhibitor of carbonic anhydrase (2, 71, was tested in perfused livers with NH; infusion (Tables I and 11, Group 4). It inhibited ureagenesis as indicated by decreased hepatic urea and citrulline and a decreased perfusate urea accumulation. This is consistent with its role in blocking formation of HCO;, a substrate for carbamyl-P synthesis (2). Net glycogenesis was blocked' and the hepatic glucose-6-P level was markedly lowered at both glucose concentrations.
The effects of ethoxyzolamide on net glycogenesis, ureagenesis, and hepatic glucose-6-P level also were tested with control livers without NH; infusion (Table IV, Group 5). For comparison, analogous values also are included in Table IV for unsupplemented control (Group 1) and ethoxyzolamide-supplemented NH;-infused livers (Group 4). Ethoxyzolamide by itself markedly lowered net glycogenesis, ureagenesis, and hepatic glucose-6-P (compared Table IV, Group 5 with Group 1). As is made clear from data in Table IV, the inclusion of NH; infusion with ethoxyzolamide further decreased net glycogenesis. Glucose-6-P level was not further altered. Ureagenesis, which was decreased by ethoxyzolamide by itself (Group 5 uersus Group 1, Table IV), was unchanged from the unsupplemented control value when NH; infusion was combined with ethoxyzolamide addition (compare Table IV, Groups 4 and 1). Glucose was initially 34 mM in studies in Table Iv. Comparable results were obtained with 9.2 mM glucose (data not shown).
Note that A glycogen is a net value, the resultant difference between total hepatic glycogen synthesis and glycogenolysis. This value may be negative when the breakdown of existing glycogen exceeds its synthesis. Although rats were fasted 48 h to remove glycogen, small amounts often remained, and were supplemented by a small glycogenesis during initial phases of the perfusion process.  Tables I and 11. Net urea formation values have been calculated from "A expressed as micromole of urea formed per g.44 min or micromole of glucosyl units incorporated in the glycogen per g. 44 min. The numbers in parentheses pertain to individual values immediately above, calculated as percent of the comparable summation value. These percent of summation values serve as a n index of the relative tendency for ureagenesis or glycogenesis to occur under the defined condition. Mean * values are presented. ' p < 0.05 for indicated glycogenic value compared with ureagenic value for the identical group (i.e. Group 1, Group 2, Group 3). Statistical analysis from other relevant comparisons between ureagenesis and glycogenesis among Groups 1,2, and 3, and between these values with 9.2 mM compared with 34 m~ glucose, have been made in Table I and 11, and are not repeated here. p < 0.05 for indicated value compared with comparable value immediately above.

TABLE IV
Effects of ethoxyzolamide, and ofethoxyzolamide plus ammonium ion, on glycogenesis, ureagenesis, and hepatic glucose-6-P in isolated liuers from 48-h fasted rats perfused with 34 mM D-glucose Livers were perfused as described in the text. In Group 5, ethoxyzolamide (3 mM) was included. Data for Group 1 (unsupplemented) and Group 4 (ethoxyzolamide-supplemented, NH4C1-infused) are from Table I1 Table I. * p < 0.05 for indicated compared with control value. ' p < 0.05 for indicated compared with control + ethoxyzolamide value.
Effects of Ethoxyzolamide on Glucokinase plus Hexokinase, Carbamyl-P:Glucose Phosphotransferase, and Glucose-6-P Phosphohydrolase-The effects of 3 m M ethoxyzolamide upon the enzymes of ATP-dependent glucose phosphorylation (glucokinase plus hexokinase), carbamyl-P-dependent glucose phosphorylation (carbamyl-P:glucose phosphotransferase), and glucose-6-P phosphohydrolase, were studied, in vitro ( Table Vj. No significant effect was noted with the glucokinase plus hexokinase, or with glucose-6-P phosphohydrolase activity of the glucose-6-phosphatase system. Carbamyl-P:glucose phosphotransferase activity of the latter system was inhibited less than 10%. Potential Contributions to Glycogenesis by the "Indirect Pathway"-Current concepts of hepatic glycogenesis (25, 26) include both the "direct" formation of glycogen from glucose through a series of enzyme-catalyzed reactions in the liver, and formation via the indirect pathway involving glucose phosphorylation in some peripheral tissue, glycolysis to lactate there, return of lactate to the liver, and ultimate formation of glycogen via hepatic glyconeogenesis from this lactate. The compound 3-mercaptopicolinic acid inhibits this process by inhibiting the enzyme phosphoenolpyruvate carboxykinase which is integral to the glyconeogenic process (1, 25, 26). Here, the possible contribution to glycogenesis from glucose via the indirect pathway must be considered, because 3-mercaptopicolinic acid was not included (1, 25), and a glycolytic tissue, erythrocytes, was present. This was done using Sukalski and Nordlie's (25) estimate of the maximum rate of glycolysis by red blood cells of 0.76 pmol of glucose/min.lOO ml of perfusate with the perfusion system unattached to livers, independent of glucose concentration. Making the extreme assumption that the entire amount of glucose fluxing through the red blood cells is used for glycogen synthesis via glyconeogenesis and the indirect pathway, it follows that 0.76 pmol of glucose/min x 44 min (the period of perfusion used here), i.e. a maximum of 33.4 pmol of glucosyl units/44 min.perfusate system, could be due to the indirect pathway.
Glycogenic values from Tables I and I1 were multiplied by NH,' , Norvaline, Ethoxyzolamide, and Hepatic Glycogenesis 7883 mean weights of livers, in grams, in each perfusion group to express the data on a "pmol of glucosyl groupsi44 min.perfusate system." The maximum contribution, as percentage of total net glycogen formation which may be explained based on the indirect pathway, was calculated by dividing 33.4 pmol of glucosyl units/44 min.perfusate system by the individual values for net glycogen formation expressed "per 44 min.perfusate system" and multiplying by 100 to convert to percent.
From the analysis (shown in Table VI), it is apparent that the maximum percent contribution of the indirect pathway to total net glycogen formation is considerably lower with 34 m M glucose than with 9.2 m M glucose. Also, at either glucose concentration, the addition of NH; increases the percent maximum contribution by the indirect pathway to glycogen formation, and the addition of norvaline to NH;-infused preparations lowers the percentage of possible contribution of the indirect pathway to total net glycogen formation.

DISCUSSION
Here, ureagenesis and glycogenesis respond inversely to NH;, without and with norvaline, in the isolated perfused liver. Similar inverse responses were seen previously with glutamine and proline (1). We have suggested that the inverse character of these responses is due to the competitive requirements for carbamyl-P in the biosynthesis of urea and glycogen (1). A further analysis of this, and other possible mechanisms underlying this reciprocity of response of ureagenesis and glycogenesis, is given here. A metabolic scheme depicting the various pathways and effects of the varied parameters considered is presented in Fig. 1. Effects of Ammonium Zon and Norvaline-NH; may stimulate ureagenesis in at least two ways: (i) directly by functioning as substrate for carbamyl-P synthase I (Reaction 1, Fig. 1); and (ii) by serving as a substrate for mitochondrial glutamate dehydrogenase (Reaction 2; Fig. l), glutamate then, through transamination by glutamate-oxalacetate transaminase (Reaction 3; Fig. l), being converted to L-aspartate which functions as a urea cycle intermediate (Reaction 4; Fig. 1) provides the second amino group for ureagenesis. With added NH; and glucose, accelerated ureagenesis through mechanisms i and ii is adequate to explain the metabolic responses observed (Tables I, 11, and IV). Accordingly, the addition of NH; to the perfusates would not only stimulate carbamyl-P synthesis, but would also accelerate its utilization via enhanced urea cycle activity resulting from increased availability of the essential intermediate L-aspartate. Thus, both the observed increased ureagenesis and diminished hepatic carbamyl-P level (with 9.2 mM glucose) may be rationalized.

Effects of ethoxyzolamide on glucokinase plus hexokinase, carbamyl-Pcglucose, phosphotransferase, and glucose-6-P phosphohydrolase actiuities, in vitro
Discrimination between the two NHI-utilizing activities cannot be explained by the differences in K,, NH; (2 m M for carbamyl phosphate synthase I (28) and 3.2 m M for glutamate dehydrogenase (33)). However, large differences exist in reported v, , , values for carbamyl-P synthase I (340 pm0Vh.g wet liver; Ref. 28) and glutamate dehydrogenase (3000 pmoV h.g wet liver; Ref. 34). Furthermore, carbamyl-P synthase I is inhibited by ADP (28) and glutamate dehydrogenase is activated by ADP (35). Both of these effects would favor a more extensive stimulation by NH; of flux through the glutamate dehydrogenase reaction than through the carbamyl-P synthase I reaction. In addition, the aspartate utilizing enzyme, argininosuccinate synthase, appears rate-limiting for ureagenesis (28) and is dependent upon aspartate concentration over a broad range of concentrations (36) spanning the physiologic (28). Following this reasoning, we conclude that the greatest impact of the added NH; is to elevate aspartate formation and thus stimulate use of available carbamyl phosphate in the urea cycle, to the disadvantage of glucose phosphorylation for glycogenesis, as is evident from the observed decreased glycogenesis and increased ureagenesis observed in the presence of added ammonium ion (Tables I and 11, Group 2 uersus 1).
Norvaline is a non-metabolizable amino acid that inhibits ornithine transcarbamylase (5, 6 ) , the first committed step in the urea cycle. When this amino acid is included with NH;, the use of carbamyl-P via the urea cycle should decrease and make carbamyl-P more available for other hepatic metabolic processes, e.g. glucose phosphorylation by carbamyl-P:glucose phosphotransferase activity of the glucose-6-phosphatase system ( 8 ) preliminary to glycogenesis from glucose. This expectation was supported by the reduced ureagenesis and enhanced glycogenesis observed when norvaline was added to the NH;-infused system, where the carbamyl-P level also increased Enzymes and reactions of the urea cycle below the membrane are cytosolic (28); carbamyl-P:glucose phosphotransferase activity of the glucose-6-phosphatase system is associated with the endoplasmic reticulum (9). NH,Cl supplied to the system provides the nitrogen directly for carbamyl-P synthesis via carbamyl-P synthase I, 111.
Although the forgoing considerations implicate the relative availability of carbamyl-P to, and utilization by, the ureagenic and glycogenic processes, they do not by themselves unequivocally establish carbamyl-P:glucose phosphotransferase activity as a major determinant in hepatic glycogenesis from glucose. Because 2 ATP molecules are required for the synthesis of one carbamyl-P molecule via carbamyl-P synthase I (Reaction 1, Fig.  l), a n enhanced synthesis and utilization of carbamyl-P in response to added NH; could drain off available ATP. This lowered ATP availability could result in a lessened glycogenesis from glucose (as observed) if glucokinase were a major determinant even under 48-h fasted conditions where the level of this enzyme was diminished by more than two-thirds (13). The noted effects of norvaline in increasing net glycogenesis from glucose thus also might be explained on this same basis. Norvaline inhibits ornithine transcarbamylase and hence ureagenesis, lowering the utilization of carbamyl-P by the urea cycle. This reduced utilization of carbamyl-P in turn would place a lessened demand on ATP for its synthesis, making more ATP available for glucose phosphorylation via residual glucokinase thus leading to increased glycogenesis from glucose. However, considerations of varied glucose concentrations and studies with ethoxyzolamide serve to obviate this contention (see below).
Effects of Varied Glucose Concentration-Because of possible ambiguities of interpretation of data to this point, responses of the various parameters measured with 34 mM glucose were compared with those same parameters determined with 9.2 mM glucose (Table I1 compared with Table I). Consistent with our hypothesis that carbamy1P:glucose phosphotransferase activity of the glucose-6-phosphatase system contributes importantly to glycogenesis from glucose are the observations that carbamyl-P levels were consistently lower in the livers perfused with 34 mM glucose than with 9.2 mM glucose in the unsupplemented, NH;-infused, and norvaline-treated, NHi-infused livers. These observations correlate with the direct utilization of carbamyl-P as a substrate for glucose phosphorylation by the relatively high K,, Gle (about 50 mM; Ref. 37); phosphotransferase activity of the glucose-6-phosphatase system in hepatic glucose phosphorylation was preliminary to glycogenesis. More carbamyl-P would be used for this purpose when the glucose concentration is high.
The second approach, also consistent with this hypothesis, involves a more formal kinetic analysis. With livers continuously infused with a source of nitrogen, i.e. NH,Cl, in the presence of the urea cycle inhibitor norvaline, an apparent K,, value of 31 mM was determined for glucose as a precursor of glycogen. This value is consistent with the apparent K,, GIc value for carbamyl-P:glucose phosphotransferase activity of the glucose-6-phosphatase system determined at low near-physiologic carbamyl-P concentration (50 mM; Ref. 371, and with the value of 30 m~ recently observed for glucose as precursor of glycogen in isolated perfused livers with 3-mercaptopicolinic acid added to inhibit glucose-6-P rehydrolysis (38). The K,, GIc values determined with intact livers no doubt contain input not only from carbamyl-P:glucose phosphotransferase, but from residual glucokinase and hexokinase also. The K,, Glc values for these last two activities are considerably lower than the carbamyl-P:glucose phosphotransferase.
The recent discovery by van Schaftingen (39) of a regulatory protein affecting glucokinase3 complicates the interpretation of observed K,, GIe values, however. The present authors concur with van Schaftingen (39) that "it is presently impossible to calculate the precise contribution of glucokinase to the formation of glucose-6-P from measurements of this enzyme made in cell-free systems." We must thus conclude that any interpretation from the apparent K,, GIc determined here is equivocal.
The comparative rates of ureagenesis and glycogenesis, relative to one another, are made clear both by the absolute values and percent of summation values in Table 111. Glycogen is formed at the expense of urea synthesis by increasing the level of glucose from 9 to 34 mM, and by inhibiting (with norvaline) the intramitochondrial usage of carbamyl-P in the ornithine transcarbamylase reaction. Ureagenesis is markedly favored at the expense of glycogenesis when NH; is infused. The summation of ureagenesis plus glycogenesis is very similar with 9 and 34 mM glucose, without additions to the system (Control, Group 1). However, when norvaline is present to inhibit the use of carbamyl-P in the ornithine transcarbamylase reaction, the total glycogenesis plus residual ureagenesis is much higher with 34 than with 9 m~ glucose, consistent with an increase in Because this protein is a competitive inhibitor uersus glucose (401, it would serve to increase the apparent K,,,g,c value when present in active form. Fructose-1-P binds to the regulatory protein and converts it to the active, inhibitory form (39) while Pi seems to be the major metabolite, at physiologic concentration, capable of converting the regulatory protein to its non-inhibitory form (41). Thus, it follows that, under appropriate cellular circumstances, the active, inhibitory form of this regulatory protein might bind to glucokinase and, while inhibiting activity, it would produce an increase in the value for the apparent K,n, because of its competitive inhibitory nature. glucose concentration-dependent use of available carbamyl-P (e.g. for glucose phosphorylation preliminary to glycogenesis) under these conditions. Effects of Ethoxyzolamide-Ethoxyzolamide and related compounds have been shown (2, 7) to inhibit liver carbonic anhydrase V, thereby lowering the production of HCO,, an essential substrate for carbamyl-P synthase I, resulting in lowered rates of ureagenesis (7). Data presented in Tables I, 11, and IV confirm these earlier findings. Furthermore, if carbamyl-P is important for hepatic glucose phosphorylation preliminary to glycogenesis, then addition of ethoxyzolamide to NH;-perfused livers should also diminish glycogenesis. Both hepatic glucose-6-P concentration and glycogenesis were markedly lowered by ethoxyzolamide included either with (Tables I  and 11) or without NH; infusion (Table IV). Because HCO, is required for the essential gluconeogenic enzyme pyruvate carboxylase (2, 421, a part of the inhibitory effect of ethoxyzolamide on hepatic glycogenesis could be through inhibition of the indirect pathway. Note, however, that maximally the indirect pathway may contribute only a portion of observed net glycogen formation ( Table VI). The marked decrease in hepatic glucose-6-P level seen with ethoxyzolamide (Tables I, 11, and IV) also is inconsistent with an effect of ethoxyzolamide manifest exclusively upon the indirect pathway of glycogenesis.
The decrease in hepatic glucose-6-P concentration as well as decreased glycogenesis from glucose in ethoxyzolamide-perfused livers supports carbamyl-P-dependent glucose phosphorylation as an important determinant in hepatic glycogenesis from glucose. Both decrease correlatively with an ethoxyzolamide-induced decrease in carbamyl-P production as evidenced also by reduced ureagenesis (Tables I and 11). Further consistent with this contention are the observations in Table IV which show that: ( a ) ethoxyzolamide added to unsupplemented control livers significantly lowers rates of both net glycogenesis and ureagenesis, as well as reduces levels of hepatic glucose-6-P; and ( b ) the presence of NH;, which enhances ureagenesis and lowers net glucogenesis when included by itself, further lowered net glycogenesis and abolished the ethoxyzolamiderelated decrease in ureagenesis when included with ethoxyzolamide. We interpret these observations to indicate that ethoxyzolamide, through its inhibition of bicarbonate production by carbonic anhydrase, inhibits glycogenesis by carbamyl-P:glucose phosphotransferase-initiated glucose phosphorylation and by the indirect pathway, and that the addition of NH; further lowers net glycogenesis by directing any residual carbamyl-P to the urea cycle at the expense of glucose phosphorylation by carbamyl-P:glucose phosphotransferase activity.
The ambiguity introduced in the earlier interpretation of our data by the possible impact of altered carbamyl-P synthesis upon availability of ATP for glucose phosphorylation by glucokinase seems obviated by our observations with ethoxyzolamide. Here, carbamyl-P synthesis is lowered as demonstrated by markedly reduced ureagenesis. Under this circumstance, using the previous line of reasoning, there would be a much lesser drain on cellular ATP for carbamyl-P synthesis. Yet, while ATP will be more available, glycogenesis from glucose is markedly reduced.
Had an effect of ethoxyzolamide been to inhibit glycogen synthase or modify one of its regulatory enzymes, or to activate glycogen phosphorylase or modify one of its regulatory enzymes (any of which could explain the ethoxyzolamide-related decrease in net glycogenesis), then the hepatic glucose-6-P level should have increased in response to ethoxyzolamide. In fact, the hepatic glucose-6-P level dropped extensively in ethoxyzolamide-treated livers. This observed decrease in liver glucose-6-P thus logically resulted from an ethoxyzolamide-associated