Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Hyperammonemia resulting from portosystemic shunting and hepatocellular insufficiency is thought to be a major factor in the pathogenesis of hepatic encephalopathy (Lookwood et al., 1979; Jessy et al., 1990; Cooper et al., 1989). CCl₄-induced cirrhosis of rats represents a valuable model with which to study the cause(s) and possible therapeutic prevention of hyperammonemia (Snodgrass, 1989; Gebhardt and Reichen, 1994). In general, changes in both, urea and glutamine synthesis, as well as hemodynamic events can contribute to an insufficient detoxification of ammonia (Gebhardt and Mecke, 1984; Häussinger et al., 1990; Meijer et al., 1990). Although there are some reports on the persistent decrease of the capacity for urea synthesis in cirrhotic livers (Kekomaki et al., 1970; Fischer-Nielsen et al., 1991), recent data from Snodgrass (1989) and our laboratory (Gebhardt and Reichen, 1994) indicated that the activities of the urea cycle enzmyes are nearly normal and cannot fully account for the decrease in converting ammonia to urea commonly observed under these conditions (Kaiser et al., 1988; Krähenbühl and Reichen, 1993). In contrast, it was found that changes in the distribution and activity of glutamine synthetase in CCl₄-induced cirrhosis play a major role in the development of hyperammonemia (Gebhardt and Reichen, 1994). Most interestingly, serum ammonia levels in these cirrhotic animals showed a negative correlation with the specific activity of GS, whereas no such correlation was observed in normal rats (Gebhardt and Reichen, 1994).

Therapy of hepatic encephalopathy is usually based on restricting dietary protein intake and administering lactulose and/or nonabsorbed antibiotics (Conn, 1994; Orlandi et al., 1994; Uribe, 1994). This therapy has a number of limitations and is not considered as ideal. Another strategy mainly aiming at enhancing flux through the urea cycle has suggested the use of OA (Kircheis et al., 1994) which provides critical substrates for the efficient operation of the cycle and, perhaps, for glutamine synthesis (Häussinger et al., 1990). Indeed, the efficacy of OA in reducing elevated ammonia levels has been demonstrated in some animal studies (Hermann, 1972; Zieve et al., 1986) and in controlled clinical trials with hyperammonemic and/or cirrhotic patients (Henglein-Ottermann, 1976; Müting and Relkowski, 1980; Staedt et al., 1993).

In our study, we have used rats with CCl₄-induced liver cirrhosis, to further investigate the influence and the potential mechanisms of action of OA. Apart from the serum levels of urea and ammonia, the influence of OA on blood amino acids and on key enzymes of ammonia detoxification in the liver was determined. To obtain more information on mechanistic aspects, additional in vitro experiments were performed using cultured hepatocytes isolated from the livers of the cirrhotic rats.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Animals. Animal experiments were approved by Swiss supervisory boards and were performed according to strict federal guidelines regulating animal experimentation. Animals were fed standard rat food and tap water ad libitum and maintained on a 12-hr dark-and-light cycle in the facilities of Dr. J. Reichen, Bern, Switzerland. Cirrhosis was induced by exposure to CCl₄-vapors and phenobarbital according to McLean et al. (1969) with modifications as described previously (Reichen et al., 1987, 1988). Treatment was carried out for 10 consecutive wk and was terminated 2 wk before the exposure to OA, a time sufficient for the effects of phenobarbital to disappear (Reichen et al., 1987).

Treatment of the animals with OA. Two weeks after termination of CCl₄ exposure the animals were split into two groups that did not differ in the parameters used to characterize cirrhosis. All animals were housed separately in individual cages, to record individual intake of food and drinking water. Both groups of rats received normal rat food and tap water ad libitum. The drinking water of the experimental group, however, was adjusted to a concentration of OA (Merz + Co. GHmbH & Co., Frankfurt, Germany) sufficient to deliver an amount of 2 g/kg body weight to each rat during a 24-hr drinking period, calculated on the amount of water consumed during the day before. Given the slight variations in the amount of drinking water per day, the daily variations in the intake of OA were less than 6%. Body weight and the amount of food and water consumed were determined each day between 4 and 5 P.M.

Isolation and cultivation of hepatocytes. Hepatocytes were isolated by the two-step collagenase technique described in detail elsewhere (Gebhardt et al., 1990). After cannulation and washing the livers free of blood, two small lobes were removed and either shock frozen with liquid nitrogen or fixed with 3.5% paraformaldehyde for enzymatic and immunohistochemical studies, respectively. Resonable yield of hepatocytes from the remaining part of the cirrhotic livers could be obtained by rising the collagenase (type BRA, Knoll AG, Ludwigshafen, Germany) concentration to 1.38 mg/ml and extending the digestion period to 25 min. The yield and viability of the hepatocyte suspensions were 272 ± 189 million cells and 73 ± 11 (n = 9), respectively; it did not differ for control and OA-treated animals. Cultivation of the hepatocytes was carried out as recently described (Gebhardt et al., 1994). After 2 hr in serum-containing Williams medium E hepatocytes were incubated in serum-free Williams medium E or with different substrates, OA and ammoniumacetate, in serum-free Hanks' buffered salt solution containing bicarbonate (25 mM) and glucose (5 mM) for evaluating their capacity for urea synthesis for additional 2 hr.

Biochemical measurements. Samples of venous blood were taken from the vena cava inferior before the cannulation of the portal vein for liver perfusion. Blood was allowed to clot and plasma was obtained by centrifugation in an Eppendorf centrifuge. Plasma urea and ammonium levels were determined enzymatically (Gebhardt and Reichen, 1994). Plasma amino acid levels were determined by high-performance liquid chromatography after mixing with orthophthaladehyde as described (Gebhardt et al., 1996). Unfortunately, in some few samples the peaks for Leu and Orn could not be separated by the standard elution protocol; thus, the sum of both amino acids is given in the respective table. However, from the samples with sufficient separation it was obvious that there was no change in the ratio of Leu and Orn under the treatment with OA. The quantitation of urea in the supernatant of the hepatocyte cultures was performed using the same method as above.

Enzymatic measurements. The activities of the urea cycle enzymes and of glutaminase were determined as described (Gebhardt et al., 1988; Gebhardt and Reichen, 1994). Glutamate dehydrogenase activity was determined according to Schmidt (1974). Glutamine synthetase activity was measured as described (Gebhardt and Williams, 1986). Protein content was determined according to Lowry et al., (1951) using bovine serum albumin as standard.

Statistical evaluation. Data were evaluated by Student's t test or analysis of variance

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

All animals showed clear signs of cirrhosis at the beginning of the experimental phase that was confirmed on death by the pronounced nodular appearance of the liver. Initial values of ABT-k, ALT and alkaline phosphatase were comparable to those measured previously in cirrhotic rats (Gebhardt and Reichen, 1994) and did not differ among randomly chosen groups for the evaluation of OA administration (table 1).

During the experimental phase the individual body weight of the animals either remained nearly constant or increased moderately but steadily. If group means are considered, a tendency for a stronger net increase in body weight for the OA group (49.5 vs. 41.3 g, control group) was noted which, however, was not quite significant (P < .06) at the termination of the experiment. It should be noted that one animal in the OA group showed some fluctuations in body weight without any other signs of illness or misbehavior. Ascites was found in two animals of the control group and one animal in the OA group. With respect to other aspects examined in this study these animals did not differ significantly from the other members in the respective groups.

Activities of the urea cycle enzymes in the control group determined at the same time as for the experimental group, i.e., after 14 days, were quite similar to values reported previously (Gebhardt and Reichen, 1994) indicating a comparable status of cirrhosis. On treatment with OA, the activities of two enzymes, e.g., carbamoylphosphate synthetase I and arginase, increased significantly (table 2), thus approaching (carbamoylphosphate synthetase I) or even exceeding (arginase) the values measured in normal control rats (Gebhardt and Reichen, 1994). All three other urea cycle enzymes remained essentially unchanged (table 2) as did glutamate dehydrogenase and glutaminase (fig. 1). Glutamine synthetase, the activity of which had been found to vary considerably in cirrhotic rats in our previous study (Gebhardt and Reichen, 1994), showed an even stronger variability (c.f., fig. 2). In some animals of both groups unusual high activities were detected, although in others extremely low levels were found. This is not surprising and seems to be due to the exceptional fate of GS⁺ hepatocytes recorded during development of cirrhosis (Gebhardt and Reichen 1994). As obvious from figure 1, treatment with OA did not significantly influence GS activities; even quite low values could still be measured (c.f., fig. 2).

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Comparison of the specific activities of glutamine synthetase (GS), glutamate dehydrogenase (GlDH) and glutaminase (GLUnase) in livers of cirrhotic controls (C, open bars) and cirrhotic rats treated with L-Orn-L-Asp (OA, closed bars). Means ± S.E.M. from six rats in each group are depicted.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Relationship between the specific activity of glutamine synthetase and the serum ammonia concentration in cirrhotic controls (upper panel) and cirrhotic rats treated with L-Orn-L-Asp (lower panel). Linear regression lines and the correlation coefficients are shown together with individual values. Note the absence of a significant correlation in the cirrhotic rats treated with L-Orn-L-Asp (lower panel).

With respect to the serum concentrations of compounds related to nitrogen metabolism, urea levels were significantly increased by approximately 34% after treatment with OA (fig. 3). However, individual values in the OA group showed a larger variation than those in controls (c.f., fig. 2). Ammonia levels in cirrhotic controls were also quite variable (figs. 2 and 3), but showed a similar negative correlation with the GS activity (fig. 2) as found previously (Gebhardt and Reichen, 1994). On treatment with OA a tendency to decrease (about 35%) was noted (P < .05) (fig. 3), although the variability of individual values remained high. However, a striking change brought about by OA was that the negative correlation with the activities of GS was not apparent any more for this group (fig. 2).

View larger version (11K): [in this window] [in a new window]   Fig. 3.   Comparison of the concentrations of urea and ammonia in the serum of cirrhotic controls (C, open bars) and cirrhotic rats treated with L-Orn-L-Asp (OA, closed bars). Means ± S.E.M. from six rats in each group are depicted. *Different from respective control, P < .03; **Different from respective control, P < .004.

Plasma amino acid concentrations determined 20 hr after the last supplementation of the drinking water with OA, in general, were slightly lower in rats of the OA group, but were significantly reduced only for glutamate, glutamine and arginine (table 3). There was no correlation between the concentration of arginine and arginase activity or urea levels. Because blood samples were taken during morning hours when the consumption of OA from drinking water was lower, it is not surprising that aspartate and ornithine did not show higher blood levels in the OA group.

The isolation of hepatocytes from the cirrhotic livers was successful with relatively high yield and viability when the concentration of collagenase was doubled. These cells attached with high plating efficiency and formed monolayers morphologically indistinguishable from cultures of normal hepatocytes. When these cultures were assayed for urea secreted into the culture medium, no difference in the basal urea production was found for hepatocytes from control and OA-treated animals (table 4). Addition of 250 µM ammoniumacetate, 1 mM OA or the combination thereof to the culture medium resulted in a comparable stimulation of urea production leading to about twice the basal rate in case of the combination (table 4). In the presence of Williams medium E, urea production determined in cultures from OA-pretreated rats was significantly higher than that in cultures from control animals. Under these conditions, production rates were 4.5- and 3.3-fold basal rates for cells from the OA group and control group, respectively.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

In our study, CCl₄-induced cirrhotic rats were used to assess the antihyperammonemic efficiency of OA. Based on several criteria, all animals used were cirrhotic and seemed to be in a comparable metabolic state as in previous experiments (Gebhardt and Reichen, 1994). As expected, body weight was increasing after the withdrawal of CCl₄ and administration of OA seemed to further support the recovery of the rats. This effect of OA does not seem to be due to the additional intake of amino acids (the amount of which is too low), but may rather be a consequence of the lowering of blood ammonia levels and, thus, less restricted cellular metabolism.

The major finding of this study is that treatment of cirrhotic rats with OA for 2 wk enhanced urea synthesis and lowered serum ammonia concentrations, although the latter effect was somewhat less strong than the former which resulted in serum ammonia levels even exceeding those of normal control rats (Gebhardt and Reichen, 1994). With respect to the mechanism of action, two effects of OA were remarkable:

First, the activities of two urea cycle enzymes were enhanced in the hepatocytes in an apparently selective manner. This result is interesting, since during administration for 2 days only Snodgrass and Lin (1981) did not find a significant influence of these amino acids on urea cycle enzymes. Furthermore, although some enzyme-specific induction phenomena were reported for some other amino acids (Snodgrass and Lin, 1981), it is commonly believed that dietary effects on urea cycle enzymes are mediated by endocrine events and affect all five enzymes in a coordinated manner (Gebhardt and Mecke, 1979; Morris; 1992). Obviously, the induction by OA observed herein should increase the overall capacity of the urea cycle.

Second, the predominant control of serum ammonia levels by the low activities of GS that was recently described for CCl₄-induced cirrhosis and is reflected in a negative correlation between serum ammonia levels and GS activities (Gebhardt and Reichen, 1994), is replaced by a more balanced control including both, ureogenesis and glutamine synthesis, characterized by the absence of such a negative correlation similar to what is found in normal rats. Apparently, the rise in the capacity of the urea cycle noted above is one reason for this convergence toward the normal control status. In addition, OA may increase the level of N-acetylglutamate (Lund and Wiggins, 1986) or of other stimulators of the urea cycle. Furthermore, an enhanced flux through the cycle, because of higher substrate levels, may also contribute to this situation.

This latter point is suggested by the in vitro experiments using cultured hepatocytes isolated from the cirrhotic rats. These hepatocytes clearly responded to the addition of ammoniumacetate and OA to the culture medium. Interestingly, neither these compounds nor their combination led to differences in the urea production of hepatocytes from control and OA-treated animals. Only in the presence of Williams medium E, i.e., in the presence of the full spectrum of amino acids and substrates, could such a difference be revealed indicating that additional cellular and regulatory aspects of ammonia detoxification had been affected by the treatment with OA. Current research is focussing on the question whether mitochondrial function in general might be affected or whether other enzymes such as transaminases are also induced by chronic administration of OA. Nevertheless, although urea excretion in vivo was not determined, these in vitro data suggest that the elevated serum urea levels reflect a higher urea production. It should be noted that the extra intake of nitrogen brought about by the administration of OA was less than 15% of total amino acid nitrogen intake calculated from food consumption. Therefore, this extra intake cannot fully account for the higher formation of urea and the higher serum urea levels observed in vivo. In turn, it can be assumed that the improved urea production contributes to the reduction of serum ammonia levels.

Another indicator for an improved nitrogen balance under treatment with OA is the fact that amino acids in general, particularly glutamate, glutamine and arginine concentrations in blood were significantly reduced. Thus, the higher weight gain of the OA-treated cirrhotic animals may reflect an increased utilization of amino acids for protein synthesis. Whether this conclusion is correct must be confirmed by further studies. However, it is interesting to note that in a clinical study on patients with cirrhosis carried out by Staedt et al. (1993) plasma levels of several amino acids also decreased, although glutamate concentration went up and glutamine remained. Because glutamine plays a significant role in the compensatory stimulation of urea formation by compensatory metabolic alkalosis in cirrhosis (Häussinger et al., 1992), the reduced concentration found in our study may limit ammonia detoxification via the urea cycle. Whether the lower level of glutamate, in turn, might limit glutamine synthesis and, thus, directly and indirectly the further reduction of blood ammonia remains to be established. The main reason for the moderate reduction of serum ammonia levels, however, seems to be the fact that neither the portocaval shunting nor the scattered distribution of GS⁺ cells that was found to be characteristic for this type of cirrhosis (Gebhardt and Reichen, 1994) were altered by the administration of OA (R. Gebhardt, unpublished observations). Thus, ammonia not removed by the low-affinity urea cycle pathway (Häussinger, 1990) may still escape the high-affinity glutamine synthetic pathway the capacity of which is also not enhanced by OA in this model.

In summary, treatment of cirrhotic rats with OA resulted in an improved ureogenesis and a concomitant lowering of blood ammonia levels. Because of the severe architectural changes in cirrhotic livers characterized by pronounced portocentral shunting of blood and the scattering of pericentral GS⁺ hepatocytes, it is doubtful whether prolonged administration of OA might cause further improvement, particularly with respect to glutamine synthesis, in this model of hyperammonemia. However, in view of its effects as an inducer of enzyme activities and its role as stimulator of urea cycle function revealed by this study, one might expect OA to considerably improve the nitrogen status in types of hyperammonemia associated with less severe changes in liver architecture.