Gene expression of urea cycle enzymes following two-thirds partial hepatectomy in the rat

Journal of Hepatology 1995; 22: 349-355 Printed in Denmark . All rights reserved Munksgaard Copenhagen

Copyrighr 0 Journal

of Heparoiogy 1995

Journal of Hepatology ISSN 0168-8278

Gene expression of urea cycle enzymes following twdi.rds partial hepatectomy in the rat Niels Tygstrup’,

Sot-en Bak’, Bjarg Krog’, Antonello

Pietrangelo2

and David A. Shafritz2

‘Department of Medicine A, Rigshospitalet, Copenhagen, Denmark and ‘The Marion Bessin Liver Research Center, Albert Einstein College of Medicine, New York, USA

The effect of reduction of functional liver mass on the expression of enzyme systems for hepatic urea synthesis was assessed in rats following two-thirds partial hepatectomy. Results were related to normal, fed rats and to sham-operated rats, with identical timing for surgery and feeding. Among the five urea cycle enzymes the mRNA steadystate level was higher in hepatectomized than in shamoperated rats for carbamoyl phosphate synthetase and arginino-succinate lyase. The level for albumin mRNA remained close to that of the controls. Relative transcription rates were found to be increased for carbamoyl phosphate synthetase, arginino-succinate syntbase and arginase. For albumin the transcription rate was drastically reduced initially, but recovered gradually during the experimental period. The data indicate that the expression of urea cycle

enzymes, in particular that of carbamoyl phosphate synthetase which is the rate-limiting step is up-regulated by partial hepatectomy. This helps to maintain urea synthesis rate at a normal or near normal level during the period of reduced liver mass, confirming metabolic studies. In contrast, the transcription for albumin was reduced. The immediate increase in urea cycle enzyme expression during the period of acute hepatocyte loss is consistent with the view that it is vitally important that urea synthesis, in contrast to e.g. albumin synthesis, remains intact when the metabolic capacity of the liver is reduced.

T

creased (4,5). This dissociation of certain capacities indicates that liver functions are modulated when the functional reserve is reduced, probably in a way which gives higher priority to functions essential for survival. This modulation may have evolved by natural selection and may be triggered by metabolic signals from the host, in both cases by promotion or suppression of the gene expression for the enzyme systems in question in the liver cells. On this assumption, the partial hepatectomy model may be used to distinguish liver functions which have immediate significance for survival from those which can be deferred temporarily. This knowledge may contribute to understanding of the pathogenesis of acute liver failure, to monitoring of patients with this syndrome, and possibly to development of rational hepatic assist therapy. The present work evaluates the effect of partial hepatectomy on gene expression of urea cycle enzymes and albumin by changes in transcription rate (run-on

hepatectomized rats survive only for a few hours, showing that absence of liver function, even for short periods, is incompatible with life. After two-thirds partial hepatectomy, there is no mortality ascribable to reduced liver function, and the liver mass is restored to normal after 3 weeks. Therefore, a reduced number of liver cells, which at the same time prepare for replication, still maintain the functions required for metabolic homeostasis, including production of acute phase reactants in response to surgical stress. Metabolic studies show that after partial hepatectomy the relative capacity, i.e. per remaining liver mass, is increased for galactose elimination (1) and urea synthesis (2,3), whereas antipyrine clearance is deOTALLY

Received 8 March 1993

Correspondence:

Niels Tygstrup, M.D., Medical Depart-

ment A, Rigshospitalet Denmark.

2152,

DK

2100 Copenhagen,

Key words: Gene expression; Liver function: Partial hepatectomy; Urea cycle. 0 Journal of Hepatology.

349

N. Tygstrup et al.

analysis) and messenger RNA steady-state levels during the first 24 h, i.e. before new liver cells have been formed.

Material and Methods Female Wistar rats, weighing 200 g, were kept on a 12h light/dark cycle and fed Altromin@ pellets ad lib. with free access to water. Groups of three rats were anesthetized with ether and either 2/3 partially hepatectomized by the method of Higgins & Anderson (6) or the abdomen was opened and the liver manipulated for a few minutes (sham-operated). The animals were killed by cervical dislocation on the following morning, 3, 6, 9, 12, 15, 18, 21, and 24 h after the operation. Five control animals were killed simultaneously Food intake after the operation was recorded. In both groups the rats started to eat after about 9 h. After 18 h the intake of sham-operated rats was equal to controls; in hepatectomized rats there was a 10% decrease in food intake. For all animals, about 200 mg of liver tissue was taken from the same site and stored in liquid nitrogen for extraction of RNA. Another 200 mg was kept for chemical determination of RNA and DNA. In a separate experiment, the remaining liver tissue from each group of partially hepatectomized rats and a similar amount of the same lobes from sham-operated and control rats was used for immediate isolation of nuclei. Probes The following cDNA probes were used: mitochondrial carbamoyl phosphate synthetase, ornithine transcarbamylase, argininosuccinate synthase, argininosuccinate lyase, arginase, rat serum albumin, and rat ribosomal 18s RNA. Probes (approximately 25 ng insert) were labelled with 32PdCTP (specific activity 3000 Ci/mmol), using a multiprime labelling kit (Amersham RPN 16012) according to the manufacturer’s instructions, to obtain a specific activity of 46X lo* cpm/pg DNA. Incorporation was measured by scintillation counting. The specificity of the probes was ascertained by autoradiography of Northern blots, showing signals from hybridized mRNA of the expected molecular size. Analysis Total DNA and RNA was measured according to ref. 7. Messenger RNA abundance: The samples were homogenized on ice (Ultraturrax, max. speed for 2 min) in solution D (Promega kit Z 5110). RNA was extracted by adding Na-acetate and phenol chloroforrn/isoamylalcohol (Promega kit Z 51 lo), vigorous mixing, kept on ice for 15 min, and centrifuged at 8000 XG at 4°C for 30 min. The supernatant was collected

350

and re-extracted, and RNA precipitated by equal volume of ice-cold isopropanol, followed by storage at -20°C for 1 h, and centrifugation at 8000 XG for 20 min. The pellet was dissolved in solution D, re-precipitated as above, washed with 75% ethanol, and re-dissolved in 400 ~1 water. RNA was denatured by adding CH3-HgOH 0.02 M, and left at room temperature for 30 min. Northern blots: Each sample, together with loading buffer (borate buffer 2X, glycerol 40%, bromphenol blue 0.04%) was placed in a 1% agarose gel and subjected to 20 V overnight in the electrophoresis buffer (borate buffer IX, formaldehyde 3.7%). The gels were incubated in NH4 acetate 0.1 M with ethidium bromide 0.5 &ml for 30 min, and then in bidistilled water for 30 min. The gel was inspected in UV light to check for loading and degradation of RNA. Gels were blotted to nylon membranes (Hybond N, Amersham) by the wick method (8), the transfer solution consisting of NaC13 M, NaOH 8 mM, pH 11.4011.45, and sarcosylO.2 M. The filters were then soaked in sodium phosphate buffer, pH 6.7, for 5 min and treated with UV light for 5 min. Prehybridization was at 65°C for 2 h with Pipes buffer 50 mM, Na2P04 50 mM, EDTA 1 mM, NaCl 100 mM, and SDS 5%. The labelled probe was immersed in boiling water for 15 min and added to fresh solution for hybridization at 65°C overnight. Filters were washed in 1X SSC and 5% SDS at room temperature for 10 min and with the same solution at 65°C for 30 min, and autoradiograms were prepared. Slot blots: Nitrocellulose filters (BA-S 85) were incubated in bidistilled water for 5 min, then in 10X SSC for 30 min and placed in the S&S Minifold, 10X SSC was added to each slot and emptied by vacuum. The samples (5 pg RNA) were loaded into 3 slots each, controls to 5 slots. After emptying the slots by vacuum 400 ~1 of 10X SSC was added and vacuum applied. Filters were prehybridized and hybridized as above, except that the amount of cDNA probe was doubled. The filters were baked at 80°C for 1 h. Autoradiograms were prepared and read by densitometry (Cream@). Results presented are based on slot blot analysis. No discrepancy between Northern blot and dot blot, as has been described for some probes (9) was observed. Densitometry readings of the slot blots were expressed as per cent of the average for the control rats. Autoradiograms of filters hybridized with rat ribosomal 18s RNA were used to check for loading of the slots. Deviations from the mean value exceeding 15% (the coefficient of variation of double determinations) were taken as evidence of unequal loading and used to adjust the reading of the corresponding sample.

Urea cycle expression after hepatectorny

ALB pslrxSham

sh 12-h 1

?@h 21h 2447 Fig. I. Autoradiograms of a representative RNA slot-blot experiment showing mRNA levels for urea cycle enzymes and albumin in liver extracts from control rats (Cont) and rats at different periods after two-thirds partial hepatectomy (pHx) or sham operation (Sham). The observations are included in Fig. 2 as per cent of controls (adjusted for unequal loading as required, see Material and Methods for details). CPS=carbamoyl phosphate synthetase; OTC=ornithine transcarbamylase; ASS=argininosuccinate synthase; ASL=argininosuccinate lyase; ARG=arginase; ALB=albumin.

Transcription rate: Tissue was cut into pieces of about 1 mm3, rinsed twice in sucrose 0.25 M, MgC12 1 mM, heparin 10 U/ml, homogenized at 4°C (Potter Elvehjelm) in sucrose 0.3 M, Tris HCl 10 mM, MgC12 5 mM, Triton O.l%, dithiothreitol 5 mM, Triton and dithiothreitol added immediately before homogenization. The homogenate was filtered through four layers of sterile gauze and centrifuged at 4°C for 5 min at 800 G. The pellets were resuspended in the same buffer, but lacking Triton, and sucrose was added to a final concentration of 1.65 M. The mixture was centrifuged at 25 000 G for 60 min at 4°C through 3 ml of sucrose 2 M, Tris HCl pH 7.9 10 mM, and MgC12 2 mM. Pellets were resuspended in glycerol 50%, MgC12 5 mM, EDTA 0.1 mM, and Hepes 50 mM. Nuclei were counted in a counting chamber. Samples were diluted to contain the same concentration of nuclei as the sample with the lowest value. For labelling of mRNA transcripts produced by the isolated nuclei, reactions were performed in tubes containing 2X lo7 nuclei, MgC12 5 mM, Hepes 100 mM; dithiothreitol 5 mM; KC1 100 mM, guanosine triphos-

phate 0.5 mM, cytosine triphosphate 0.5 mM, adenosine triphosphate 0.5 mM; RNAsin@ 300 U/ml, 32Puridine triphosphate 50 ,&i (Amersham 3000 ,&i/ mmol) incubated at room temperature for 40 min on a shaker. Reactions were terminated by treatment with DNase 240 U for 5 min, and labelled RNA isolated by sodium dodecyl sulfate-proteinase K digestion at 42°C for 40 min, followed by phenolchloroform extraction and ethanol precipitation. Hybond filter N, 0.45 micron and Whatman 2 MM paper were soaked in distilled water for 20 min, then in sodium chloride sodium citrate solution 6~ for 5 min, and placed in a slot-blotter (S&S Minifold II, SRC072/0). Labelled cDNA (5 pugfor each sample, including the pBR322 plasmid) was dissolved in TE buffer and NaOH 2N 9 : 1, kept at room temperature for 5-10 min, then heated in boiling water for 15 min and diluted with ice-cold sodium chloride sodium citrate solution 6~ 10 : 1. Each slot was loaded with 320 ~1 and drained by suction. DNA was fixed by UV light for 7 min. Filters were prehybridized with formamide 50%, 3.51

N. Tygstrup

et al.

mRNA

levels

CPS

OTC Sham

0

0

0

I

I

I

I

I

I

I

I

3

6

9

12

15

18

21

24

01

0

3

6

9

12

15

18

21

24 Hours

Hours

ASS

ASL 0 0

0’

0

I

I

I

I

I

I

1

I

3

6

9

12

15

18

21

24

J

01 0

PHX

Sham

I

I

I

I

I

I

I

I

3

6

9

12

15

18

21

24

Hours

Hours

ARG

ALB

2oo1 2ool

0’

0

I

I

1

I

1

I

I

3

6

9

12

15

18

21

I

24 Hours

0

0

I

I

I

t

I

I

I

1

3

6

9

12

15

18

21

24 Hours

Fig. 2. Densitometry of autoradiograms slot-blot experiments, showing mRNA levels for urea cycle enzymes and albumin in liver extracts from control rats (zero time) and rats at dlyferent periods after two-thir& partial hepatectomy (a) and shamoperated rats (0). Each point is the mean of observations from three animals (including one shown in Fig. 1) as per cent of five controls, with SEA4 (vertical bars). Abbreviations as in Fig. 1.

3.52

Urea cycle expression after hepatectomy

SET buffer 5X, sodium dodecyl sulfate 0.5%, Denhardt’s solution (Sigma, 50 mg each of bovine serum albumin, ficoll and polyvinylpyrrolidone) 2 X, and yeast tRNA 250 &ml at 42°C for 4 h, and hybridized at 42°C for 36 h in fresh hybridizing solution containing the labelled RNA. Filters were washed twice with sodium chloride sodium citrate solution 2X at room temperature for 15 min, twice with sodium chloride sodium citrate solution 2X and sodium dodecyl sulfate 0.1% at 65°C for 30 min, once with sodium chloride sodium citrate solution 1X, twice with sodium chloride sodium citrate solution 0.1 X, all with sodium dodecyl sulfate 0.1% at 65°C for 30 min, once with sodium chloride sodium citrate solution 0.1 X at room temperature for 30 min, and finally treated with 100 ,~l RNase solution (5 &ml) in sodium chloride sodium citrate solution 2~ 200 ml. Autoradiograms were prepared and read by densitometry (Cream@).

Results Liver weight was reduced to 30% after hepatectomy, remaining constant in both groups. The concentration

of total hepatic RNA and DNA was not significantly different between the groups, nor did it change during the experiment. Levels of mRNA for urea cycle enzymes and albumin in liver tissue during the first 24 h after partial hepatectomy and sham operation, respectively, was examined with three separate RNA slot-blot experiments. Fig. 1 shows an autoradiogram of a representative experiment and Fig. 2 the densitometer readings. The steady-state level of mRNA for carbamoyl phosphate synthetase was increased 2-3 times after hepatectomy, half as much after sham operation. Both groups showed an early peak for argininosuccinate lyase, and the fold induction was higher for argininosuccinate lyase after hepatectomy. For ornithine transcarbamylase and argininosuccinate synthetase there was no consistent change during the experiment, nor a difference between the groups. In sham-operated rats arginase showed an early peak as with argininosuccinate lyase, but the difference between the groups was small. The average albumin mRNA in the hepatectomized rats fell slightly during the experimental period.

2Fh

2

Fig. 3. Autoradiograms of a representative run-on transcription experiment showing relative transcription rate for urea cycle enzymes and albumin in isolated liver cell nuclei from control rats (Cont) and rats at different periods after two-thirds partial hepatectomy (pHx) or sham operation (Sham). (See Material and Methods for details). Abbreviations as in Fig. 1.

353

N. Tygstrup et al.

Representative autoradiograms of run-on experiments are shown in Fig. 3. Signals for ornithine transcarbamylase were very weak, but for carbamoyl phosphate synthetase, argininosuccinate synthetase, argininosuccinate lyase, and arginase the signal at most periods of observation in separate experiments was consistently stronger in hepatectomized than in sham-operated rats. Since the signals for argininosuccinate synthetase and arginase in the sham-operated group were very low, it was not possible to perform quantitative comparison by densitometry. The relative transcription rate for albumin was suppressed initially in both groups, more strongly so in nuclei from the partially hepatectomized rats, and gradually returned to control values during the experimental period.

Discussion The aim of this study was to investigate whether the liver modulates gene expression for urea synthesis when the functional liver mass is reduced. The first 24 h after partial hepatectomy, i.e. before new liver cells contribute to function, was selected for study. During this period, most of the hepatocytes will enter the cell cycle under the influence of growth factors. At the same time, the liver responds to the stress of anaesthesia and surgery by producing acute phase reactants, mediated by glucagon, corticoid hormones and cytokines. Unless the liver has the capacity to keep all homeostatic functions at a normal live1 (by trebling the work of the remaining liver cells), it might be expected that, by natural selection, gene regulation would favour vital functions at the expense of less vital functions. Sham-operated rats served as controls in order to distinguish the effect of a stress response on changes in gene expression from that of responses due tp reduced functional liver mass. However, stress induced by sham operation and partial hepatectomy may not be the same, presumably being greater after the latter. Furthermore, partial hepatectomy may in itself modify the acute phase response with depressed expression of several ‘positive’ acute phase reactants, such as thiostatin, az-macroglobulin, and p-fibrinogen (10). Stress hormones, notably glucagon and glucocorticoids, known to be increased after hepatectomy, are also physiological regulators for expression of metabolic functions like ureagenesis and gluconeogenesis. Therefore, response to stress and to reduced liver function cannot be completely separated. Within 24 h after hepatectomy, the mRNA levels for argininosuccinate synthetase, argininosuccinate lyase, and particularly carbamoyl phosphate synthetase were higher than in the control rats. The expression of all five urea cycle enzymes is known to be stimulated by

354

glucagon and (except for ornithine transcarbamylase) by corticosteroids (1 l), both of which are increased during stress. In sham-operated rats, only carbamoyl phosphate synthetase mRNA was above the control level during the whole period. Fasting also stimulates expression of urea cycle genes, in 1 day it increases argininosuccinate lyase mRNA three-fold (11). The rise in mRNA for argininosuccinate lyase at 6 h (cf. Fig. 2), i.e. before the rats start to eat, may be an early signal of fasting. In the partially hepatectomized rats, the transcription rate of urea cycle enzymes, except ornithine transcarbamylase, is increased in relation to control rats as well as to sham-operated rats (Fig. 3). The difference in transcription rate is most pronounced for carbamoyl phosphate synthetase and argininosuccinate lyase, and enhanced transcription of the genes may explain the difference in mRNA steady-state level for these enzymes. The sharp decrease in the transcription rate for carbamoyl phosphate synthetase 9 h after sham operation coincides with intake of food, but is absent after partial hepatectomy. The increase in carbamoyl phosphate synthetase mRNA is consistent with the finding of increased enzyme activity in hepatectomized rats (3). Carbamoyl phosphate synthetase is rate limiting for the urea cycle, which normally operates at 20-50% of capacity (12). Reduction of the functional liver mass to one third may therefore reduce urea formation unless more enzyme is formed. After 70% partial hepatectomy, the capacity for urea formation per g liver was doubled (2). The Vmax of ornithine transcarbamylase is 40 times greater than that of carbamoyl phosphate synthetase (13), and the unchanged mRNA level for this enzyme after hepatectomy, or even the reduction, as found by some authors (10,14) is not contradictory. Argininosuccinate synthetase, argininosuccinate lyase and arginase have a lower Vmax and tend to have higher mRNA levels in hepatectomized than in sham-operated rats. The slow turnover of the enzymes (3-9 days, (15)) may enhance the effect of moderate changes in mRNA. Ornithine is increased in the regenerating liver (16) and metabolized to polyamine by ornithine decarboxylase, which is also increased (17). It is questionable whether this process interacts with ureagenesis (13). The transcription rate for albumin is suppressed after partial hepatectomy, as also found by Milland et al. (lo), and slightly less so after sham operation. According to Rosa et al. (18) transcription was equal to controls 6 h after partial hepatectomy and only slightly reduced after 30 h. It seems reasonable to conclude that, under the conditions studied, the reduced liver mass upgrades urea

Urea cycle expression after hepatectomy

production through a selective and early activation of the transcription rate of key enzymes, such as carbamoyl phosphate synthetase, and/or posttranscriptionally in the case of the other enzymes in the urea cycle. This occurs at the expense of the other functions, possibly including production of albumin and other serum enzymes. Whether it also is at the expense of the growth response can only be shown in a model where the stimuli for growth and for maintenance of urea synthesis can be separated.

Acknowledgements The study was supported by the Benzon Foundation, the Danish Health Research Foundation Grant 120853 and NIH grants DK 17609 and Core Center Grant DK 41296. We wish to thank S.M.J. Morris for providing plasmids containing CPS, OTC, ASS, ASL, and ARG cDNA sequences. Preliminary data presented at Cold Spring Harbor Laboratory Symposium, May 1993, and European Association for the Study of the Liver, Paris, August 1993.

References 1. Yildirim SI, Pot&en HE. Quantitative liver functions after 70% hepatectomy. Em J Clin Invest 1981; 11: 469-72. 2. Hansen BA, Poulsen HE. Changes in the hepatic capacity of urea-N synthesis, galactose elimination and antipyrine clearance following 70% hepatectomy in the rat. Stand J Clin Lab Invest 1986; 46: 233-7. 3. Goodman MW, Zieve L. Carbamoyl-phosphate synthetase I activity and ureagenesis in regenerating liver of the normal rat. Am J Physiol Gastrointest Liver Physiol 1985; 248: G501-6. 4. Poulsen HE, Pilsgaard H. Antipyrine metabolism during hepatic regeneration in the rat. Liver 1985; 5: 1969. 5. Zieve L, Lyftogt C, Draves K, Raphael D. Urea excretion, galactose elimination, and aminopyrine disappearance during normal liver regeneration after partial hepatectomy. J Lab Clin Med 1984; 104: 2434.

6. Higgins GM, Anderson RM. Experimental pathology of the liver. I. Restoration of the liver of the white rat following partial surgical removal. Arch Path01 1931; 12: 186202. 7. Schneider WC. Determination of nucleic acids in tissues by pentose analysis. Meth Enzymol 1957; 3: 6804. 8. Chomczynski l? One-hour downward alkaline capillary transfer for blotting of DNA and RNA. Anal Biochem 1992; 201: 134-9. 9. Tsykin A, Thomas T, Milland J, Aldred AR, Schreiber G. Dot blot hybridization using cytoplasmic extracts is inappropriate for determination of mRNA levels in regenerating liver. Nucl Acids Res 1990; 18: 382. 10. Milland J, Tsykin A, Thomas T, Aldred AR, Cole T, Schreiber G. Gene expression in regenerating and acute-phase rat liver. Am J Physiol 1990: 259: G340-7. 11 Morris SMJ, Moncman CL, Rand KD, Dizikes GJ, Cederbaum SD, O’Brian WE. Regulation of mRNA levels for five urea cycle enzymes in rat liver by diet, cyclic AMP, and glucocorticoids. Arch Biochem Biophys 1987; 256: 343-53. 12. Grisolia S, Wallace R, Mendelson J. Correlation between in vivo and in vitro metabolic measurements. Maximum capacity for urea synthesis. Physiol Chem Physics 1975; 7: 21923. 13. Meijer AJ, Lamers WH, Chamuleau AFM. Nitrogen metabolism and ornithine cycle function. Physiol Rev 1990; 70: 70148. 14. Ito Y, Hayashi H, Taira M, Tatibana M, Tabata Y, Isono K. Depression of liver-specific gene expression in regenerating rat liver: a putative cause for liver dysfunction after hepatectomy. J Surg Res 1991; 51: 143-7. 15. Nebes VL, Morris SMJ. Regulation of messenger ribonucleic acid levels for five urea cycle enzymes in cultured rat hepatocytes. Requirements for cyclic adenosine monophosphate, glucocorticoids, and ongoing protein synthesis. Mol Endocrino1 1988; 2: 444-51. 16. Van Dijk MA, Rodenburg RJT, Holthuizen P, Sussenbach JS. The liver-specific promoter of the human insulin-like growth factor II gene is activated by CCAAT/enhancer binding protein (C/EBP). Nucleic Acids Res 1992; 20: 30999104. __ __.. 17. Tomiya T, Sato Y, Murakami Y, Nagoshi S, Ugata S, Pujiwara K. Ornithine decarboxylase induction in partially hepatectomized rat liver and modes of its stimulation by glucagon and insulin. J Biochem 1991; 110: 173-8. 18. Rosa JL, Tauler A, Lange AJ, Pilkis SJ, Bartrons R. Transcriptional and posttranscriptional regulation of 6-phosphofructo-2-kinase/fructose-2,6_bisphosphatase during liver regeneration. Proc Nat1 Acad Sci 1992; 89: 374650.

355