renal prostaglandin (PG) excretion in experimental cir- rhosis. In an additional group of animals, including nine rats chronically exposed to CC1, (CC1, rats) and ...
Clinical Science (1988) 75,263-269
263
Longitudinal study of renal prostaglandin excretion in cirrhotic rats: relationship with the renin-aldosterone system JAVIER SOLA, JORDI CAMPS, VICENTE ARROYO, FRANCISCO GUARNER, JOAN GAYA, FRANCISCA RIVERA AND JOAN ROOES Liver Unit and Hormonal Laboratory, Hospital Clinic i Provincial, University of Barcelona, Barcelona, Spain
(Received 24 September 1987/20 January 1988; accepted 16 February 1988) SUMMARY 1. A cross-sectional study (protocol A ) was performed in 19 rats with cirrhosis, induced by carbon tetrachloride (CCl,), and ascites and in 10 control animals to assess renal prostaglandin (PG) excretion in experimental cirrhosis. In an additional group of animals, including nine rats chronically exposed to CC1, (CC1, rats) and six control rats, a longitudinal study (protocol B) was performed to investigate the temporal relationship between changes in renal PG excretion, the renin-aldosterone system and renal function. 2. Urinary PG excretion was assessed by specific radioimmunoassay of PGE,, PGF,,, 6-keto-PGF1,and thromboxane (TX) B, after extraction with octadecyl silica cartridges and h.p.1.c. purification. Recoveries for each prostanoid (61f 8% for PGE,, 64 k 12% for PGF,,, 65 k 11% for 6-keto-PGFIa and 66 k 17% for TXB,) were determined in every sample by adding tritiated standards, and the final values were corrected according to the individual recoveries. 3. Cirrhotic rats with ascites in protocol A showed a significantly higher plasma renin and aldosterone concentrations and urinary excretion of 6-keto-PGF1, and TXB, than did control animals. Urinary excretion of PGE, and PGF,,, however, was significantly reduced in cirrhotic animals as compared with controls. 4. In CCI, rats included in protocol B, there was a close chronological relationship between the activation of the renin-aldosterone system, as estimated by urinary aldosterone excretion, the onset of sodium retention and the increase in urinary excretion of 6-keto-PGF,, and TXB,. The urinary excretion of PGE, and PGF,, in CCl, rats was reduced throughout the study. 5. These results suggest that in rats with CC1,-induced cirrhosis, the increased renal production of prostacyclin Correspondence: Dr V. Arroyo, Unidad de Hepatologia, Hospital Clinic i Provincial, Villarroel, 170, 08036-Barcelona, Spain.
and TXA, is a consequence of the stimulation of endogenous vasoconstrictor systems. The impaired urinary excretion of PGE, and PGF,, in CCI, rats may be secondary to an inhibitory effect of CCl, on PG synthesis in the renal tubules. Key words: cirrhosis, prostacyclin, prostaglandin E,, prostaglandin F,, sodium retention, thromboxane A,. Abbreviations: NAG, N-acetyl-B-D-glucosaminidase; PG, prostaglandin; PRC, plasma renin concentration; TX, thromboxane; U,,, V, urinary excretion of aldosterone 18-glucuronide; U,, V,urinary excretion of sodium. INTRODUCTION Boyer et al. [l]and Zipser et al. [2] first suggested that renal prostaglandins (PGs) play an important role in the maintenance of renal haemodynamics in cirrhosis. These authors reported that the administration of indomethacin, a prostaglandin synthetase inhibitor, to patients with cirrhosis and ascites produced an acute reversible reduction of renal blood flow and glomerular filtration rate. Subsequent studies confirmed their findings [3] and demonstrated that urinary excretion of 6-keto-PGF1, and PGE, in cirrhosis with ascites correlates with renal haemodynamics [4].Additionally, other investigators have shown that PGs play a critical role in renal water metabolism and in the renal response to diuretics in these patients [5-71. It has been suggested that PGs participate in the control of renal function in cirrhosis by antagonizing the renal vascular and tubular effects of angiotensin 11, sympathetic nervous activity and antidiuretic hormone [3, 8-10]. The development of animal models of cirrhosis allows the investigation of several aspects of the pathogenesis of renal dysfunction in chronic liver disease that cannot be studied in man. In the current study we report the results of a longitudinal study in one of these models (rats chroni-
264
J. Sola et al.
cally exposed to carbon tetrachloride, CCl,) aimed at investigating the temporal relationship between renal PGs, the activity of endogenous vasoactive systems, as estimated by the degree of stimulation of the renin-aldosterone system, and renal function. The goal of the investigation was to ascertain if the increased renal PG synthesis in cirrhosis correlated chronologically with the activation of endogenous vasoactive systems and with the onset of sodium retention. MATERIALS AND METHODS Animals
Cirrhosis was induced in male Sprague-Dawley rats weighing 160-180 g by the method of Lopez-Novoa et al. [ll]. CCl, was used as hepatotoxin, and phenobarbital shortened the time required to induce cirrhosis. Phenobarbital (0.3 g/l) was added to the drinking water during induction. After 1 week on phenobarbital, inhalation of CCl, was started. Subsets of five rats were placed in a glass chamber (70 cm x 25 cm X 30 cm). Compressed air was passed via a flowmeter (1 litre/min) into the glass chamber after bubbling through a flask containing CCI, at room temperature. Animals were exposed to CCl, twice weekly (Monday and Saturday) during the study, starting with 0.5 min of bubbled air and 0.5 min in a gas atmosphere. In the fourth inhalation session, there was an increase to 1 min of bubbled air and 1 min in a gas atmosphere. Afterwards, the dosage was increased by 1 min every three sessions until 5 min of bubbled air and 5 min in a gas atmosphere was reached. Male control rats of matched weight and age received only phenobarbital added to the drinking water throughout the study. Protocol A
This protocol was designed to assess renal PG excretion, plasma renin concentration (PRC), plasma aldosterone concentration and renal function in cirrhotic rats with sodium retention and in control animals. The aim of the protocol was to investigate whether renal PG excretion is altered in this animal model. The protocol was performed in 19 cirrhotic rats and 10 controls. On the twelfth week after starting treatment with CCI, and/or phenobarbital, animals were placed in individual metabolic cages and studies were performed after 1 week to allow adaptation. The presence of ascites in animals submitted to the cirrhosis induction programme (CCI, rats) was determined by abdomjnal paracentesis. A previous investigation in our laboratory [ 121 has shown that CC1, rats develop sodium retention and ascites within 8-10 weeks after starting the cirrhosis induction programme. A 24 h urine volume was collected in glass testtubes under dry ice in order to ensure immediate freezing of the urine after voiding. A portion of each 24 h urine sample was frozen at -30°C until assayed for 6-ketoPGF,. (a stable metabolite of prostacyclin), PGE,, PGF,, and thromboxane (TX)B, (a stable metabolite of TXA,), sodium, creatinine and N-acetyl-B-D-glucosaminidase
(NAG). Sodium balance was estimated by subtracting the urinary sodium excretion from the sodium intake. After collection of urine, the animals were killed by decapitation. The time which elapsed from the removal of the animal from the cage until decapitation was always less than 10 s [ 131. During exsanguination blood samples were collected under ice, in ethylenediaminetetra-acetate (potassium salt) tubes, to measure plasma creatinine, plasma aldosterone and PRC. These samples were centrifuged at 4°C and the plasma was frozen at - 30°C until assayed. Liver and kidney biopsies were performed immediately after killing the rats. The tissue was fixed in formalin and sectioned conventionally. Liver tissue was stained with haematoxylin and eosin, Masson’s trichrome stain and reticulin stain for histological examination. Kidney tissue was stained with haematoxylin and eosin. Protocol B
The aim of this protocol was to investigate the temporal relationship between the activation of the renin-aldosterone system, as estimated by the urinary excretion of aldosterone- 18-glucuronide ( u A , d V ) , renal PG excretion, sodium retention and ascites formation in rats submitted to the cirrhosis induction programme. Twelve rats exposed to CCl, and phenobarbital and six control rats were initially included in the protocol. However, three CCl, rats died before the onset of hyperaldosteronism. Therefore, protocol B included nine CCl, rats and six control animals. One week after starting the gas inhalation in CCl, rats, these animals and control rats were placed in individual metabolic cages and fed ad libitum with a normal sodium chow (125 mmol of sodium/kg of food) and distilled water as drinking fluid. After allowing 1 week for adaptation, measurements of the 24 h sodium intake and urinary sodium excretion (U,,V) were made on three consecutive days each week (Tuesday, Wednesday and Thursday) throughout the study. Since faecal sodium excretion was not measured, sodium balance was estimated by subtracting the daily U,,V from the daily sodium intake. The urine was collected in glass test-tubes under dry ice and the volume was measured gravimetrically. A portion of each 24 h urine sample was frozen at -30°C until assayed for aldosterone, 6-keto-PGF,,, PGE,, PGF,, and TXB,. In CCl, rats the appearance of ascites was explored once weekly (Monday) by abdominal paracentesis. Thirteen weeks after starting the protocol, animals were killed. Paraffin-embedded tissue samples of liver were stained with haematoxylin and eosin, Masson’s trichrome stain and reticulin stain for histological examination. Measurementsand statistical analysis U A l d V and plasma aldosterone were measured by radioimmunoassay (CIS Sorin Biomedica, Saluggia, Italy). Urine samples (0.5 ml) were adjusted to pH 1.0 with 1 ml of 0.2 mol/l HC1 and were kept for 20 h at 30°C. Using this procedure, most aldosterone-18-glucuronide is trans-
Prostaglandinsand the renin system in cirrhosis formed into aldosterone [14].The coefficient of variation was 10% for intra-assay and 14% for interassay determinations. PGs were extracted from urine and then analysed by h.p.1.c. and radioimmunoassay. The method involved the addition of [3H]PGE,, [3HlpcF,,, 6-ket0-[~HlPGF,, and [3HlTXB, (1000 c.p.m. each) to a 2.5 ml portion of each urine sample to estimate recoveries after extraction and chromatography. Urine samples were then acidified to pH 3.0 and extracted on octadecyl silica cartridges (Sep-Pak, Waters Associates, Milford, MA, U.S.A.). PGs were isolated by h.p.1.c. on a reverse-phase octadecyl silica column (Nova-Pak, Waters Associates, Milford, MA, U.S.A.), using an isocratic solvent system composed of triethylamine/formic buffer (pH 3.0) and acetonitrile (7525, v/v) at a flow rate of 2 ml/min. The percentage of recovery for PGE,, PGF,,, 6-keto-PGFl, and TXB, after the combined extraction and chromatography averaged 6 1 f8%, 64 f 12%, 65 f 11% and 66 f 17% (mean fSD), respectively. The PGE, and PGF,, fractions were analysed by an equilibrium radioimmunoassay using highly specific antibodies (Institute Pasteur, Paris, France), [3H]PGE,and [3H]PGF2,as tracers, and charcoal-dextran as separation system. 6-Keto-PGF,, and TXB, fractions were determined by equilibrium radioimmunoassay (New England Nuclear, Boston, MA, U.S.A.), using iodinated PGs as labelled ligands and 16% (w/v) polyethylene glycol as the separation system. The accuracy measured by the percentage of recovery of authentic PG standards added was 90 f lo%, 91 f 13%, 100 f7% and 102 f 13% (meansfSD), respectively, the inter-assay coefficient of variation 19%, 16%, 17% and 16%, respectively, and the sensitivity 1.3,1.8,2.0 and 1.5 pg/tube, respectively. PRC was determined by radioimmunoassay as previously described [ 131. Sodium concentration was measured by flame photometry (Instrumentation Laboratory model 943) and creatinine concentration by the Jaff6 reaction (Boehringer Mannheim Diagnostica, Mannheim, West Germany). NAG was determined by a colorimetric method [ 151.
265
Data are presented as means ~ S E M .Comparison between groups was performed using the Student's t-test or the non-parametric Mann-Whitney test. Comparison within the same group was performed using the paired ttest. Correlations between data were made by twovariable regression analysis. RESULTS Protocol A Table 1 shows the renal function, PRC, plasma aldosterone and urinary excretion of PGs in the two groups of rats studied. All cirrhotic rats had ascites at the time of the study. Serum creatinine, creatinine clearance and urine volume were similar in the two groups of animals. Cirrhotic rats showed marked sodium retention at the time of the study. PRC, plasma aldosterone and urinary excretion of 6-keto-PGF, ,and TXB, were significantly higher in cirrhotic rats than in control rats. In contrast, urinary excretion of PGE, and PGF,, was significantly reduced in cirrhotic rats as compared with controls. The post-mortem examination of the liver of CCl, rats showed cirrhosis in every instance. Kidney histology was normal in all animals. The urinary excretion of a marker of tubular necrosis, NAG, was similar in both groups (19.4 f2.5 i.u./g of creatinine in control rats vs 15.0 f2.8 in cirrhotic rats). Protocol B
Table 2 shows the temporal relationship between sodium balance, U,,,V and urinary excretion of 6-ketoPGF,,, TXB,, PGE, and PGF,,, in CC14 rats and control animals included in protocol B. U,,, V, U,, V and sodium balance in control rats remained steady throughout the study. However, in CCl, rats U,,, V increased between the seventh and tenth week after exposure to CCl, (four rats had increased U,,,V at the seventh week, two at the eighth, two at the ninth and one at the tenth week). In
Table 1. Renal function, activity of the renin-aldosterone system and urinary excretion of PGs in control and cirrhotic rats included in protocol A Results are means fSEM.Abbreviation: NS, not s i m c a n t . Control rats (n=10) Plasma creatinine (mg/dl) Creatinine clearance (ml/min) Urine volume (ml/day) Sodium excretion (mmol/day) Sodium balance (mmol/day) PRC (ngof ANG I h - ' ml-I) Plasma aldosterone (pg/ml) Urinary 6-keto-PGF,, (ng/day) Urinary TXB, (ng/day) Urinary PGE, (ng/day) Urinary PGF,, (ng/day)
0.64 f 0.02 2.12 f 0.07 14.8 f 1.7 1.49 f 0.02 --0.19f 0.10 13.3 f 3.6 89.5 f 19.6 7.3 k 0.6 8.1 f 0.3 63.3 f 9.9 96.3 f 8.2
Cirrhotic rats (n=19)
P
0.60 f 0.02 2.04 f 0.10 12.6 f 0.78 0.91 kO.10 0.43f0.13 44.6 f 6.1 1159.2 f 2 1.7 52.7 f 8.9 26.6 k 2.5 36.4 f 6.9 66.6 f 10.4
NS NS NS
< 0.00 1 < 0.005 < 0.005 < 0.005 < 0.00 1 < 0.001 < 0.050 < 0.050
J. Sola et al.
266
Table 2. Sodium balance, UAld Vand urinary excretion of PGs in control and CCI, rats included in protocol B Results represent the arithmetic mean ( k SEM) of all values obtained weekly in each group of rats ( 18 samples from control rats and 27 from CCI, rats). In order to evaluate the temporal relationship between hyperaldosteronism,sodium retention, ascites formation and PG excretion in CCI, rats, values obtained in this group are shown according to the onset of hyperaldosteronism: week 0 represents the week in which an increase in U A I d V was first detected; weeks - 2 and - 1 are the 2 consecutive weeks before the appearance of hyperaldosteronism. In normal rats, values are presented in relation to the initiation of the protocol. Statistical significance: "P
