Cirrhosis induced in rats either by carbon tetrachloride/phenobarbital administration or bile duct ligation is the standard experimental model for decompensated cirrhosis.1–4 Although the rat model has solved numerous problems, genetic manipulation of rats has many limitations. Within the last 8 years, there have been more than 100 reports of fibrosis, inflammation, hepatic necrosis, apoptosis, nitric oxide synthase, peptide kinetics, and altered Kupffer cell function in bile duct–ligated wild-type and transgenic mice. There is, thus, every reason to congratulate Ackermann and co-workers on their success in creating a cholestatic mouse model of decompensated cirrhosis in male CD 1 mice.5 The use of a mouse model has several advantages over other mammalian animal models: A large number of genetically inbred mouse strains are available, and it is possible to generate stable gene-specific targeting by homologous recombination in murine embryonic stem cells.6 Despite the obvious differences between human and mouse, the 2 genomes are remarkably similar, and in most cases a gene can be substituted without any detectable difference in gene function or regulation, which further strengthens the use of transgenic mouse models for human diseases. Moreover, mice are easy to breed, have a short generation time, and produce large litters. Because of their small size, mice can be housed in larger numbers, which cuts down the costs of animal housing, and they are generally easy to handle. A mouse model of experimental cirrhosis may, therefore, have a hitherto unseen potential with respect to the investigation of basal and advanced pathophysiology, including molecular and regulatory aspects of renal sodium retention.
Various animal models of cirrhosis and sinusoidal-portal hypertension illustrate different potential pathophysiological mechanisms. Thus, the venoconstricted dog or pig model may chiefly illustrate underfilling, secondary to ascites formation.7, 8 In the portal vein–ligated rat model, vasodilatation and a hyperdynamic systemic circulation has been shown to precede renal sodium retention,9 and effective hypovolemia has been measured directly in this model.10 Animals with toxic experimental cirrhosis (carbon tetrachloride/phenobarbital in the rat, dimethylnitrosamine in the dog) exhibit vasodilatation and various degrees of hyperdynamic systemic circulation, but most often have normal renal perfusion and glomerular filtration rate,1, 11 the latter being in contradistinction to human cirrhosis. The toxic models of experimental cirrhosis may predominantly illustrate overflow mechanisms of sodium retention, but may have some other problems inherent in the model. For example, carbon tetrachloride is known to be nephrotoxic and cardiotoxic. Moreover, sodium retention may exist in the dimethylnitrosamine dog model of experimental cirrhosis even with completely normal portal pressure,11 a feature that has not been described in patients with cirrhosis with normal portal pressure. In the bile duct–ligated rat model of experimental cirrhosis, the thick ascending limb of Henle has recently been identified as a main site of sodium reabsorption, and down-regulation of the enzyme 11beta-hydroxysteroid dehydrogenase, with inappropriate activation of tubular sodium reabsorption by endogenous glucocorticoids, has been described as a potential pathophysiological mechanism.12
The bile duct–ligated mouse model of cirrhosis described by Ackermann and co-workers contains definite elements of primary renal sodium retention. However, the presence of systemic hemodynamic alterations such as vasodilatation/vasoconstriction, sinusoidal-portal hypertension, renal perfusion, glomerular filtration, etc. have not been reported. Moreover, some assertions in the article regarding underfill, overflow, and arterial vasodilatation theories are not completely balanced. Also ignored in the discussion is the substantial evidence supporting effective hypovolemia, and the role of aldosterone in sodium reabsorption in cirrhosis.13, 14 This phenomenon is especially important because the widely recognized clinical effects of spironolactone and amiloride in patients with cirrhosis with sodium retention and ascites14 cannot be explained by the new mouse model.
The sham-operated mice in the study by Ackermann and co-workers showed a positive sodium balance (although less so than the bile duct–ligated mice). This is surprising, unless positive sodium balance follows growth and gain in body weight. It could also be assumed that the experimental set-up may favor sodium retention or reflect problems inherent in balance studies. The bile duct–ligated rat and the toxic rat models of experimental cirrhosis demonstrate a positive sodium balance before ascites develops.1, 12 This contrasts with the current mouse model, where no significant sodium retention was found in the bile duct–ligated mice without ascites, as compared to the sham-operated mice. This may suggest differences between the different animal models of experimental cirrhosis, or in the time course between the development of sodium retention and the formation of ascites.
Using a sophisticated methodology, Ackermann and co-workers localized the cortical collecting ducts as the main site of sodium reabsorption, independent of increased aldosterone, corticosteroid, and activation of the mineralocorticoid receptor. Urinary excretion of aldosterone was significantly increased in the nonascitic mice and even more so in the animals with ascites. This is consistent with several studies in patients with early cirrhosis, with normal or moderately elevated circulating renin and aldosterone,13–15 and may actually indicate the presence of effective systemic hypovolemia also in the bile duct–ligated mouse, a mechanism that may work alongside a primary sodium retention in the cortical collecting duct. The present mineralocorticosteroid clamp (adrenalectomy plus steroid infusion) shows very convincingly, in concert with the applied molecular techniques, that sodium retention can take place independently of the presence of aldosterone and without stimulation of the mineralocorticoid receptor. However, in vivo, the excretion (and thereby production) of aldosterone was increased in the bile duct–ligated mice, which would contribute to sodium and water retention, rather than the opposite. If the increased aldosterone production in the mouse model were a result of renal insensitivity to aldosterone, this would be a finding contrary to most previous studies in humans. Further experiments with suppression and stimulation of the aldosterone system may determine whether there is renal insensitivity to aldosterone in this mouse model.
Proximal tubular sodium excretion may be reduced in the patient with preascitic phase and ascitic phase of cirrhosis,15–17 a finding recognized by researchers who favor overflow mechanisms.18 Sodium/potassium–adenosine triphosphatase (Na+/K+-ATPase) activity is an important indicator of sodium transport along the nephron. An important finding in the present study is reduced Na+/K+-ATPase activity in the proximal tubules of the bile duct–ligated mouse model. This could well be due to a reduced load of glomerular filtrated sodium in addition to disturbances of nitric oxide–mediated mechanisms.15, 19 Reduced glomerular filtration has been observed in the bile duct–ligated rat.20 Therefore, the sodium retention demonstrated in the mouse model of experimental cirrhosis could be caused by a combination of reduced filtrated sodium (secondary to reduced renal perfusion and glomerular filtration, both potential mechanisms of effective hypovolemia), a fixed enhanced retention of sodium in the cortical collecting ducts, and, in fact, some action of aldosterone in vivo. These possibilities, which evidently exist from the data presented, should be considered in a balanced view of this important model of early and advanced experimental cirrhosis. Manipulation of the portal and systemic circulation is important in order to solve afferent, efferent, and other regulatory mechanisms in the integral renal sodium retention, because alteration in one part of the nephron may be counterbalanced by changes in other parts, making the picture complex. Such studies are now awaited with interest.
The multicatheterized, exercising, water-immersed rat model was introduced more than 25 years ago,21 and experimental cirrhosis in the rat has now solved many problems relating to decompensated liver disease in humans. We now look forward to an advanced mouse model of experimental cirrhosis, where genetic and pathophysiological manipulations can be performed to further elucidate local, systemic, and regulatory aspects of the complicated process of sodium-water retention on the integral, cellular, and molecular level in liver disease in order to improve our understanding of the complications of cirrhosis in humans.