Metabolic fate of isoleucine in a rat model of hepatic encephalopathy and in cultured neural cells exposed to ammonia
- First Online:
- Cite this article as:
- Bak, L.K., Iversen, P., Sørensen, M. et al. Metab Brain Dis (2009) 24: 135. doi:10.1007/s11011-008-9123-4
- 146 Views
Hepatic encephalopathy is a severe neuropathological condition arising secondary to liver failure. The pathogenesis is not well understood; however, hyperammonemia is considered to be one causative factor. Hyperammonemia has been suggested to inhibit tricarboxylic acid (TCA) cycle activity, thus affecting energy metabolism. Furthermore, it has been suggested that catabolism of the branched-chain amino acid isoleucine may help curb the effect of hyperammonemia by by-passing the TCA cycle block as well as providing the carbon skeleton for glutamate and glutamine synthesis thus fixating ammonia. Here we present novel results describing [U-13C]isoleucine metabolism in muscle and brain analyzed by mass spectrometry in bile duct ligated rats, a model of chronic hepatic encephalopathy, and discuss them in relation to previously published results from neural cell cultures. The metabolism of [U-13C]isoleucine in muscle tissue was about five times higher than that in the brain which, in turn, was lower than in corresponding cell cultures. However, synthesis of glutamate and glutamine was supported by catabolism of isoleucine. In rat brain, differential labeling patterns in glutamate and glutamine suggest that isoleucine may primarily be metabolized in the astrocytic compartment which is in accordance with previous findings in neural cell cultures. Lastly, in rat brain the labeling patterns of glutamate, aspartate and GABA do not suggest any significant inhibition by ammonia of TCA cycle activity which corresponds well to findings in neural cell cultures. Branched-chain amino acids including isoleucine are used for treating hepatic encephalopathy and the present findings shed light on the possible mechanism involved. The low turn-over of isoleucine in rat brain suggests that this amino acid does not serve the role of providing metabolites pertinent to TCA cycle function and hence energy formation as well as the necessary carbon skeleton for subsequent ammonia fixation in hyperammonemia. The higher metabolism of isoleucine in muscle could, however, contribute to ammonia fixation and thus likely be of value in the treatment of hepatic encephalopathy.
Severe liver disease such as liver cirrhosis may result in hyperammonemia and hepatic encephalopathy (HE), comprizing motor disturbances, personality changes, and lowered consciousness. The pathogenesis of HE is, however, not well understood but it is likely that increased brain ammonia levels play a significant role (Albrecht and Dolinska 2001; Shawcross et al. 2005). In this context, the two enzymes glutamine synthetase (GS) and glutamate dehydrogenase (GDH) are of particular interest in catalyzing the only reactions in the brain able to perform ammonia fixation (Schousboe and Waagepetersen 2007). As GS is exclusively expressed in astrocytes (Norenberg and Martinez-Hernandez 1979) and GDH as judged by immunohistochemistry and measurement of enzyme activity has a higher astrocytic than neuronal expression (Rothe et al. 1994; Zaganas et al. 2001), the astrocytes have a key role in ammonia metabolism (Hawkins et al. 2002). Actually, formation of glutamine is likely to be a key factor in astrocytic swelling seen in relation to HE (Albrecht and Dolinska 2001).
It has been repeatedly suggested that an inhibitory action of ammonia on brain energy metabolism mediated by an effect on the key oxido-reductases pyruvate dehydrogenase and α-ketoglutarate dehydrogenase (McKhann and Tower 1961; Lai and Cooper 1986; Zwingmann et al. 2003) may be of importance for the detrimental symptoms of HE. Additionally, it has been suggested that the effects on glutamate homeostasis may disturb the malate-aspartate shuttle (Hindfelt et al. 1977) that also has a key role in brain energy metabolism (McKenna et al. 2006).
Considering the fact that metabolism of the branched chain amino acid isoleucine leads to formation of a tricarboxylic acid (TCA) cycle intermediary (succinyl-CoA) and acetyl CoA at the same time, it was suggested by Ott et al. (2005) that this amino acid might serve the role of providing metabolites pertinent to TCA cycle function and hence energy formation as well as the necessary carbon skeleton for subsequent ammonia fixation. A recent study in cultured neurons and astrocytes provided evidence that this function of isoleucine may at least to some extent be fulfilled (Johansen et al. 2007). However, no information is available to elucidate to what extent isoleucine may serve such a role in the brain in vivo during hyperammonemia. The present study was accordingly conducted comparing isoleucine metabolism in a rat model of cirrhosis with chronic HE, i.e. bile duct ligated rats with that in cultured neurons and astrocytes exposed to high levels of ammonia. Isoleucine metabolism was monitored using [U-13C]isoleucine and LC-MS technology to follow the metabolic fate of the carbon atoms through the TCA cycle (Johansen et al. 2007). As metabolism of isoleucine in muscle may also be of interest this was also analyzed in the rat model of HE.
Female Wistar rats (mean body weight 200 g (range 192–214 g), age between 12 and 13 weeks were obtained from Møllegård Breeding Centre, University of Aarhus (Denmark) and 7-day-old NMRI mice were obtained from Taconic M&B (Ry, Denmark). Tissue culture utensils were purchased from NUNC A/S (Roskilde, Denmark), fetal calf serum from Seralab., Ltd. (Succex, UK) and culture medium as well as poly-D-lysine from Sigma Chem. Co. (St. Louis, MO, USA). [U-13C]isoleucine (>98% enriched) was obtained from Isotec/Sigma Chem. Co. (St. Louis, MO, USA). The Phenomenex EZ-faast LC-MS kit was used for amino acid analysis. All other chemicals used were of the purest grade available from regular commercial sources.
Experimental cirrhosis in rats
The rats were housed for a week before surgery in the animal facilities of Aarhus University at 22±2°C, 55 ± 10% relative humidity, air change 8–10 times/h, 12:12 h light–dark cycle (06:30–18:30 h light), fed standard food (Altromin diet no. 1324; Chr. Petersen, Slagelse, Denmark), and with three animals per cage. The rats were subjected to either bile duct ligation (BLD) or sham-operation (SHAM). Since approx. 10% of the BDL-rats died in the postoperative period, the number of animals allocated to the two procedures was adjusted accordingly. Surgical procedures were performed under inhalation anaesthesia: Induction was performed with the rat in a custom-made glass cylinder through which a mixture of oxygen, 0.5 l /minute, and atmospheric air, 1.5 l/minute, containing 5% Isoflurane (Forene; Abbott Laboratories, UK) was blown. During surgery, anaesthesia was maintained with 2% Isoflurane in oxygen and atmospheric air as above, given on a mask covering the nose of the animal. With the animal placed in a supine position on an operating table and using sterile technique, a midline abdominal incision was made and the hepatic ligament was exposed. In the BLD-rats the common bile duct was ligated at two places with 4–0 non-absorbent surgical sutures (Seraslex Silk 4.0) and cut. SHAM-rats had the hepatic ligament manipulated, but the bile duct was not ligated. The abdomen was closed with 4–0 absorbent suture in two layers single knots (Ethicon Vicryl 4.0). After the operation the animal received a subcutaneous injection of a long lasting NSAID, 1 mg Carprofen (Rimadyl; Pfizer Animal Health, Exton, USA), 30 mg Ampicillin (Pentrexyl; Bristol-Meyers Squibb, Lyngby, Denmark) and 1.5 ml isotonic saline. After the operation, the animals were kept in the animal facilities for 6 weeks for cirrhosis to develop in the BLD-rats. The study was undertaken in accordance with local and national guidelines for animal welfare.
Tracer injection studies
The rats were fasted for 3 h prior to the tracer injection studies, weighed and given intraperitoneal injections of 0.2 mmol [U-13C]isoleucine four times during 24 min. Two min after the last injection, the animals were subjected to inhalation anaesthesia as mentioned above and the abdomen was opened. 1 ml blood was taken from the aorta for measurements of arterial blood ammonia (van Anken and Schiphorst 1974) and enrichment of [U-13C]isoleucine.
The rats were decapitated and the brains removed and frozen in liquid nitrogen within 40–50 s after the blood sampling. Within 1 week the brains were dissected in forebrain and cerebellum. The tissue samples (brain and muscle, see below) were extracted in 70% v/v ethanol (ice–cold) and the extracts centrifuged (20.000 g, 20 min). The pellets were dissolved in 1 M KOH and used for determination of protein (Lowry et al. 1951) using bovine serum albumin as the standard. The supernatants were used for LC-MS analysis (Johansen et al. 2007) using a Shimadzu LC-MS-2010 mass spectrometer coupled to a Shimadzu 10A VP HPLC system. The Phenomenex EZ-fast amino acid analysis kit for LC-MS was used for analysis of percentage 13C labeling in relevant amino acids.
Simultaneously with the removal of the brain, muscle tissue samples were frozen in liquid nitrogen. Liver tissue was sampled for histological examination using Haematoxylin & Eosin and Masson’s Trichrome staining. Three primary changes for cirrhosis were assessed: disturbed architecture, bile duct proliferation with displacement of hepatocytes, and formation of septa between portal areas containing bile duct cells and connective tissue. According to the degree of these changes in liver architecture, cirrhosis was evaluated as light or severe.
Cell culture studies
Results from a previous study of [U-13C]isoleucine metabolism in cultured neurons and astrocytes are used for comparison in the present study. Experimental details can be found in Johansen et al. (2007).
All labeling data were corrected for natural abundance of 13C by subtracting the enrichment of 13C measured in a standard sample of the relevant amino acid. Isotopic enrichment was calculated according to Biemann (1962). Analysis of 13C labeling data was performed by employing Microsoft Excel 2007 and GraphPad Prism v5 software. All data are presented as averages ± standard error of the mean (SEM). Statistical differences were analyzed employing one-way ANOVA followed by Tukey-Kramer post hoc test. A P-value of 0.05 or less was considered significant.
Bile duct ligated rats (BDL) had yellowish color of the tail and the fur, and one third had ascites. Liver histology showed cirrhosis in all BDL rats, four with light and two with severe changes in cytoarchitecture. They all had macroscopic characteristics, such as fatty liver, and bile duct dilation. Liver weight was significantly higher in the BDL rats (on average 17 g, range 14–20 g) than in the control rats (on average 8 g, range 5–10 g) (P < 0.001). BDL rats had a similar gain in weight, from about 200 g, to on average 240 g (range 218–259 g) as the control rats, weighing on average 226 g (range 211–241 g).
Plasma levels of ammonia and enrichment of [U-13C]isoleucine
Values for labeling (%) of [U-13C]isoleucine and concentration (µM) of ammonia in arterial plasma from SHAM-operated and BDL-rats
69.4 ± 5.4
29.0 ± 3.3
69.1 ± 1.0
81.0 ± 17.9*
Labeling (%) of 13C isotopomers of glutamate in muscle from SHAM- and BDL-rats
5.3 ± 1.0
9.8 ± 2.2
3.7 ± 0.9
4.4 ± 1.1
7.3 ± 1.8
2.7 ± 0.7
Labeling (%) of 13C isotopomers of amino acids in forebrain extracts from rats (SHAM and BDL) after repeated i.p. injections of [U-13C]isoleucine
1.17 ± 0.47
0.84 ± 0.10
0.17 ± 0.04
1.02 ± 0.39
0.85 ± 0.11
0.26 ± 0.11
1.50 ± 0.32
0.74 ± 0.07
1.58 ± 0.23*
0.63 ± 0.10
1.20 ± 0.19
0.48 ± 0.14
1.08 ± 0.19
0.48 ± 0.14
1.60 ± 0.17
0.27 ± 0.07
0.26 ± 0.04
1.46 ± 0.16
0.31 ± 0.08
0.28 ± 0.05
Labeling (%) of 13C isotopomers of amino acids in cerebellar extracts from rats (SHAM and BDL) after repeated i.p. injections of [U-13C]isoleucine
1.30 ± 0.20
1.18 ± 0.25
0.18 ± 0.05
1.37 ± 0.09
1.22 ± 0.17
0.23 ± 0.03
2.32 ± 0.41#
0.99 ± 0.10
2.29 ± 0.39#
0.93 ± 0.11
1.98 ± 0.25*
0.68 ± 0.21
1.72 ± 0.39
0.65 ± 0.24
1.96 ± 0.26
0.61 ± 0.06
0.30 ± 0.04
2.64 ± 0.42*
0.66 ± 0.16
0.30 ± 0.05
Metabolism of [U-13C]isoleucine in muscle
Metabolism of [U-13C]isoleucine in brain
[U-13C]Isoleucine will give rise to a combination of triple labeled succinyl CoA and double (uniformly) labeled acetyl CoA. The latter will be diluted to a large extent by unlabeled acetyl CoA originating from glucose metabolism and is therefore considered less important for labeling of the amino acids glutamate, glutamine, aspartate and GABA, the carbon skeletons of which are derived from α-ketoglutarate or oxaloacetate (Fig. 1). The prevailing labeling pattern will be mono, double or triple labeled glutamate/glutamine, and aspartate while GABA will be essentially only mono and double labeled. As seen from Tables 3 and 4 this was indeed the labeling patterns obtained analyzing extracts from forebrain and cerebellum from both SHAM-operated and bile duct ligated rats. It is seen that no effect of the bile duct ligation was observed in any of the two brain regions. However, mono-labeling of aspartate was statistically significantly (P < 0.001) higher in cerebellum compared to forebrain in BDL-rats. In addition, mono-labeling of GABA was significantly higher in cerebellum as compared to forebrain in SHAM-operated animals (P < 0.05).
It should be noted that in both brain regions, mono-labeled glutamine could not be detected in spite of significant mono-label in glutamate, its precursor (Tables 3 and 4). Additionally, double-labeling was statistically significantly (P < 0.01) more pronounced in glutamine than in corresponding glutamate in both brain areas in BDL-rats and for cerebellum also in sham-operated rats (Tables 3 and 4).
Metabolism of [U-13C]isoleucine in cultured neural cells
It is seen that the extent of labeling was generally much higher in the cultured neural cells than in rat brain (cf. Tables 3 and 4). Moreover, while bile duct ligation had no effect on the metabolism of isoleucine in vivo, neuronal isoleucine metabolism was reduced by ammonia in the cultured neurons possibly caused by increased glycolysis and subsequent metabolism of the resulting unlabeled acetyl CoA in the TCA cycle (see Johansen et al. 2007). In the astrocytes, aspartate labeling was also reduced by ammonia while that of glutamine was increased.
The finding that the BDL-rats had cirrhosis and a significantly increased plasma ammonia level shows that this model can be used to study the effect of chronically increased ammonia levels on metabolic processes. Moreover, the finding that the isotopic enrichment in isoleucine was similar in blood, muscle and brain indicates that it is possible to compare the metabolic events in muscle and brain. It should be noted that the elevated level of ammonia did not influence the basic metabolism of isoleucine in the two tissues.
A mixture of branched-chain amino acids (Generaid®) has been employed in the treatment of patients suffering from chronic HE; however, the exact mechanism remains unclear. The hypothesis (Ott et al. 2005) that isoleucine might be able to partly ameliorate the adverse effects of ammonia during HE due to its ability to supply a TCA cycle intermediate as well as acetyl CoA thereby rescuing energy metabolism and providing a precursor for glutamate and glutamine synthesis was only partly supported by the results of the in vivo study. The level of metabolic activity of [U-13C]isoleucine in muscle tissue of both control (SHAM-operated) and BDL rats was found to be rather low compared to the high enrichment of the amino acid itself. It would also have been anticipated that isoleucine should have been metabolized to a considerable extent in the rat brain both under normal conditions and during hyperammonemia induced by bile duct ligation. This was clearly not the case since only about 2% incorporation of 13C from [U-13C]isoleucine into TCA cycle derived amino acids was observed, a value much lower than that found in the muscle tissue. This level of incorporation in the two brain regions was much lower than that seen in the cultured neurons as well as astrocytes (Johansen et al. 2007). However, the fact that relatively high levels of mono and double labeled glutamate, aspartate and GABA were found suggests that no significant inhibition of the TCA cycle was evident, as these metabolites are formed from second turn-derived intermediaries of the TCA cycle (see Fig. 1). This is in accordance with the findings in neuronal cell cultures, where only a slight inhibition of the astrocytic TCA cycle was found whereas neuronal TCA cycling (and glycolysis) actually increased (Johansen et al. 2007). This finding may, however, be contradictory to the notion that ammonia has an inhibitory action on oxidative metabolism (Zwingmann et al. 2003) and that the cerebral oxygen consumption appears down-regulated during human liver failure associated with elevated ammonia levels (Iversen et al. 2009). One possible explanation may be that the actual level of isoleucine metabolism observed in the present study was low and that this therefore may not reflect a possible general reduction of the oxidative metabolism. It was interesting that double-labeled glutamine was more prevalent than its corresponding precursor, glutamate. This finding indicates that TCA cycle metabolism of isoleucine is more pronounced in the astrocytic than the neuronal compartment of the rat brain. That this may well be the case is supported by the previous in vitro study (Johansen et al. 2007) in which astrocytic metabolism of isoleucine was more pronounced than its neuronal metabolism. It should, however, be noted that while ammonia in vitro was found to have an effect on isoleucine metabolism (Johansen et al. 2007), this was not observed in the rat brain in vivo. One important distinction between the in vitro and the in vivo studies is, that the in vitro experiments probably reflect an acute load of ammonia whereas the BDL-rats represent a chronic form of hyperammonemia/hepatic encephalopathy. This may partly explain the lack of observable effect in the in vivo study; thus, future studies should include BDL-rats subjected to an acute load of ammonia.
The reason why isoleucine was not avidly metabolized in the rat brain regardless of the degree of hyperammonemia is not quite clear. One possibility is that the i.p. injected [U-13C]isoleucine did not easily penetrate the blood brain barrier. This is, however, not likely to be the case as the large essential neutral amino acid transporter carrying the branched chain amino acids is present (Hawkins et al. 2006). In agreement with this, the isotopic enrichment of isoleucine was found to be essentially the same in plasma, muscle and brain. Further studies are needed to investigate this question. In the meantime it is concluded that while it was found in the cultured neural cells that isoleucine facilitated ammonia detoxification by facilitating astrocytic glutamine formation (Johansen et al. 2007) this effect of isoleucine seemed less pronounced in the brain in vivo.
The expert secretarial assistance of Ms Hanne Danø and the skilful technical assistance of Ms Ann Lene Vigh are highly appreciated. The experimental work has been supported by grants from the Danish Medical Research Council (22-04-0314 and 271-07-0267) and the Lundbeck, Hørslev and Carlsberg Foundations. Professor Stephen Hamilton Dutoit, Institute of Pathology, Aarhus University Hospital is cordially thanked for help evaluating the liver biopsies.