It has been shown that the renin-angiotensin system (RAS) plays key roles in the development of fibrosis in numerous organs, including the liver. Other studies have suggested that the RAS also may play roles in diseases of chronic inflammation. However, whether the RAS also can mediate acute inflammation in liver is unclear. The purpose of this study therefore was to determine the effect of the RAS inhibitors captopril and losartan on acute liver damage and inflammation caused by hepatic ischemia and subsequent reperfusion. Accordingly, male rats were subjected to 1 hour of hepatic ischemia (70%) followed by reperfusion; animals were killed 3, 8, or 24 hours after reperfusion. The effect of captopril or losartan (100 or 5 mg/kg intragastrically, respectively) was compared with that of vehicle (saline). The expression of angiotensinogen in liver increased fivefold 3 hours after reperfusion. Indices of liver damage and inflammation (e.g., alanine aminotransferase levels, pathological features, tumor necrosis factor-α levels, and intercellular adhesion molecule-1 expression) all were significantly elevated in vehicle-treated animals after hepatic ischemia and subsequent reperfusion. Ischemia and reperfusion also caused an increase in the accumulation of protein adducts of 4-hydroxynonenal, an index of oxidative stress. Captopril or losartan treatment showed profound protective effects under these conditions, significantly blunting the increase in all these parameters caused by ischemia and reperfusion. In conclusion, RAS inhibitors prevent acute liver injury in a model of inflammation caused by ischemia and reperfusion. These data further suggest that the RAS may play a key role in mediating such responses in the liver and suggest a novel role for this system. (HEPATOLOGY 2004;40:583–589.)
Angiotensin II (AngII), generated by endothelial and circulating angiotensin converting enzyme, is a classic endocrine hormone that plays a central role in the regulation of blood pressure and sodium homeostasis.1 AngII also is known to have a number of blood pressure-independent actions, including mitogenic and trophic effects on cell growth.2 Recent in vivo studies have demonstrated that components of the renin-angiotensin system (RAS) may be involved in organ fibrosis. For example, fibrosis in liver, kidney, lung, and heart have been linked to AngII activity in experimental animal models.3–5 In humans, it has been shown that losartan, an angiotensin II type 1 (AT1) receptor antagonist, not only decreases portal pressure in cirrhosis,6 but also may have antifibrotic effects in liver.7 Therefore, the RAS seems to play key roles in tissue remodeling and scarring, especially after injury.
In addition to the above-mentioned effects of the RAS in development of fibrosis, results of recent work have indicated that this system also may play a proinflammatory role in tissue (see8 for review). For example, AngII has been shown to stimulate expression of proinflammatory chemokines and cytokines,9 adhesion molecules (e.g., intercellular adhesion molecule-1 [ICAM-1]),10 and to activate proinflammatory transcription factors (e.g., activator protein-1 and nuclear factor κB).11 Furthermore, AngII can lead directly to increased superoxide production via nicotinamide adenine dinucleotide phosphate oxidase in macrophages.12 In liver, stellate cells have been shown to respond to AngII to produce proinflammatory and profibrotic cytokines and chemokines and reactive oxygen species.13, 14
Although the role other hepatic cells play in response to the AngII is less understood, Kupffer cells also have been shown to express the AT1 receptor under basal conditions.15 Importantly, chronic systemic infusion of subpressor levels of AngII has been shown to cause a proinflammatory and profibrotic response in rat liver.13 However, mechanisms by which the RAS may play a role in mediating hepatic inflammation remain unclear. The purpose of the current study therefore was to determine the potential role of the RAS in inflammation and damage in acute ischemia and reperfusion injury in rat liver and to test the hypothesis that captopril and losartan exert protective effects under these conditions.
AngII, angiotensin II; RAS, renin-angiotensin system; ICAM-1, intercellular adhesion molecule-1; TNFα, tumor necrosis factor-α; AT1, angiotensin II type 1; MGB, Minor Groove Binder; FAM, 6-carboxyfluorescein; CT, threshold cycles.
Materials and Methods
Animal Husbandry and Treatment.
Male Sprague-Dawley rats (200–250 g) fed laboratory chow were used in all experiments and were housed in pathogen-free barrier facilities accredited by the Association for Assessment and Accreditation of Laboratory Animal Care. The procedures used were approved by the local Institutional Animal Care and Use Committee. Animals were fasted for 24 hours before experiments and received captopril (100 mg/kg), losartan (5 mg/kg), or vehicle (saline) by gavage 24 and 1.5 hours before the surgical procedure. For induction of hepatic ischemia, anesthesia was initiated by isoflurane and body temperatures were maintained by a heating pad. All vessels leading to the left and median hepatic lobes were clamped for 60 minutes using a vascular clamp. Sham surgeries were identical to above in the absence of clamping of the vessels. After 1 hour, the clip was removed, allowing blood to reenter the liver. Animals subsequently were anesthetized with sodium pentobarbital (150 mg/kg intraperitoneally) 3, 6, and 24 h after reperfusion for sample harvesting. Blood was collected after anesthesia just before sacrifice by exsanguination and serum stored at −80°C in a microtube until later analysis. Livers were harvested, with portions snap frozen, and other sections formalin fixed, embedded in paraffin, and mounted on microscope slides for histological assessment.
Clinical Analyses and Histological Examination.
Serum alanine transaminase levels were determined using standard kits (Thermotrace, Melbourne, Australia). For histological analysis, sections (6 μm) were stained with hematoxylin and eosin. Infiltrating and apoptotic (condensed and fragmented nuclei) cells were determined morphologically by counting from 1,000 cells in hematoxylin and eosin-stained slides.16 Adducts of 4-hydroxynonenal (lipid peroxidation) were detected by immunohistochemistry as described previously.17 Neutrophil accumulation in the livers was assessed by staining tissue sections for chloracetate esterase, a specific marker for neutrophils, using the naphthol AS-D chloracetate esterase kit (Sigma, St. Louis, MO).18
Real-Time Polymerase Chain Reaction Analysis of messenger RNA Expression.
Total RNA was isolated from liver tissue using RNA STAT 60 (Tel-Test, Inc., Friendswood, TX) at sacrifice according to manufacturer's instruction. Polymerase chain reaction (PCR) primers (see Table 1) for angiotensinogen, tumor necrosis factor-α (TNFα), and ICAM-1 were designed using Primer Express (version 1.5; Applied Biosystems, Foster City, CA). Primers were designed to cross introns to ensure only cDNA and not DNA was amplified. The fluorogenic Minor Groove Binder (MGB) probe was labeled with the reporter dye 6-carboxyfluorescein (FAM). The TaqMan Universal PCR Master Mix (Applied Biosystems) was used to prepare the PCR mix. The 2 × mixture is optimized for TaqMan reactions and contains AmpliTaq gold DNA polymerase, AmpErase, dNTPs with UTP, and a passive reference. Primers and probe were added to a final concentration of 300 nM and 100 nM, respectively. cDNA was made from 1 μg of total RNA using the Advantage for RT for PCR kit (BD Biosciences, Palo Alto, CA) following the manufacturer's instructions. Controls (no DNA template) were run to ensure that there was no amplification of contaminating DNA. The amplification reactions were carried out in the ABI Prism 7700 sequence detection system (Applied Biosystems, Foster City, CA) with initial hold steps (50°C for 2 min, followed by 95°C for 10 min) and 40 cycles of a two-step PCR (92°C for 15 sec, 60°C for 1 min). The fluorescence intensity of each sample was measured at each temperature change to monitor amplification of the target gene. The comparative CT method was used to determine fold differences between samples. The comparative CT method determines the amount of target, normalized to an endogenous reference (β-actin) and relative to a calibrator (2−ΔΔCt). The purity and specificity of PCR products were verified by gel filtration and sequencing.
Table 1. Primer and Probe Sequences for TNFα, ICAM-1, Angiotensinogen, (ANG) and β-Actin
Forward Primer (5′-3′)
Reverse Primer (5′-3′)
Abbreviation: ANG, angiotensinogen.
Results are reported as means ± SEM (n = 4–6). Two-way ANOVA with Bonferroni's post hoc test was used for the determination of statistical significance between treatment groups, as appropriate. For comparison of pathological scores, the Mann-Whitney rank sum test was used. A P value less than .05 was selected before the study as the level of significance.
Effects of Ischemia and Reperfusion and of Captopril on Release of Transaminases and Liver Pathological Features.
Serum alanine aminotransferase levels before ischemia reperfusion were in the normal range for rats (approximately 25 U/L). Sham surgery caused a moderate but significant increase in these values, peaking 6 hours after the procedure, with averages of approximately 100 U/L (Fig. 1); treatment of animals with captopril had no significant effect on these parameters. As expected, 1 hour of ischemia, followed by reperfusion, led to a progressive increase in serum alanine aminotransferase levels, peaking at 6 hours of reperfusion, at which time values were almost 20-fold higher than in sham-treated animals (Fig. 1, top panel). Although the effect was less pronounced 24 hours after reperfusion, serum alanine aminotransferase levels still were significantly elevated compared with sham-treated animals (Fig. 1, bottom panel). Treatment with captopril or losartan caused a significant blunting of this effect (Fig. 1).
No pathological changes were observed in liver tissue 24 hours after sham operation, either in the presence or absence of captopril (Fig. 2A). In contrast, progressive tissue damage was observed in liver specimens from vehicle-treated animals peaking 24 hours after reperfusion; pathological changes consisted of focal necrosis and infiltration of lymphocytes (Fig. 2B). Captopril or losartan treatment blunted the pathological changes at all time points caused by ischemia and reperfusion; indeed, the number and size of necroinflammatory foci were decreased dramatically 24 hours after reperfusion (Fig. 2C, D). The number of condensed nuclei (Fig. 3, top panel) and infiltrating lymphocytes (Fig. 3, bottom panel) were quantitated in these samples 24 hours after reperfusion as indices of apoptosis or oncosis and inflammation, respectively. The number of condensed nuclei and infiltrating lymphocytes was low in sham-treated animals, as expected. Ischemia and reperfusion led to a dramatic increase in both of these parameters in vehicle-treated animals. The effect of ischemia and reperfusion was almost completely blocked by captopril or losartan treatment (Fig. 3).
Increases in Gene Expression Caused by Ischemia and Reperfusion.
The relative hepatic messenger RNA levels of angiotensinogen, ICAM-1, and TNFα were quantitated after reperfusion (or sham surgery) by real-time PCR and were calibrated to expression of β-actin as a housekeeping gene. Changes in the expression of all of these parameters were elevated 3 hours after reperfusion (Fig. 4), but returned to levels comparable with that of sham groups 6 and 24 hours after reperfusion (not shown). Ischemia and reperfusion caused an approximately fourfold increase in angiotensinogen expression 3 hours after reperfusion (Fig. 4, “ANG,” top panel). Captopril or losartan treatment did not affect the increase in angiotensinogen caused by ischemia and reperfusion. As has been observed by others,19, 20 expression of ICAM-1 and TNFα in rat liver was increased 3 hours after reperfusion. Captopril or losartan treatment completely blocked the increase in expression of ICAM-1 and TNFα caused by ischemia and reperfusion (Fig. 4).
Neutrophil Accumulation Caused by Ischemia and Reperfusion Is Blunted by Captopril.
Figure 5 comprises representative photomicrographs depicting neutrophil accumulation in the liver 24 hours after ischemia and reperfusion, as determined by naphthol AS-D chloracetate esterase staining (see Materials and Methods). This technique stains neutrophils a dark pink color against a counterstain of hematoxylin (blue). There were few neutrophils observed in any region of livers from sham-operated animals (200×; Fig. 5A). As expected, there was a robust accumulation of neutrophils in necroinflammatory foci that peaked 24 hours after ischemia and reperfusion in vehicle-treated animals (Fig. 5B). The relative amount of neutrophils in these foci was approximately 10% of total cells present. Conversely, neutrophil accumulation in captopril- or losartan-treated animals after ischemia and reperfusion was far less (Fig. 5C, D) in both the amount and relative percent of necroinflammatory foci. Furthermore, red blood cells, which stain yellow with this technique, also were highly abundant after ischemia and reperfusion in necroinflammatory foci in livers from vehicle-treated rats (Fig. 5B), indicative of edema. The number of these cells also was fewer in these necroinflammatory regions in captopril- or losartan-treated animals (Fig. 5C, D).
Accumulation of Lipid Peroxidation (4-Hydroxynonenal) Protein Adducts in Liver after Ischemia and Reperfusion.
The accumulation of 4-hydroxynonenal protein adducts was determined as an index of lipid peroxidation and oxidative stress 24 hours after reperfusion; this time point was determined in preliminary studies to be this signal peak after reperfusion in vehicle-treated animals. The number of adducts in vehicle-treated sham animals was low, comprising approximately 1% of total cellular area (Fig. 6A). Levels of staining were similar in sham-treated animals receiving captopril or losartan (not shown). Ischemia and reperfusion caused a dramatic increase in the amount and extent of 4-hydroxynonenal adduct staining in vehicle-treated animals (Fig. 6B). Staining was localized not only in the necroinflammatory foci, but also in other regions of the liver that appeared to be undamaged by pathological assessment and comprised approximately 30% of total cellular area. Compared with vehicle-treated animals, the amount and extent of adduct staining was substantially less in captopril- and losartan-treated animals (Fig. 6C, D). Specifically, although staining still occurred in the necroinflammatory foci in this treatment group, there was almost no increase in staining over livers from sham-treated animals in undamaged areas of the hepatic lobule. Because of this factor, coupled with the fewer and smaller foci in this group compared with vehicle-treated animals after ischemia and reperfusion, 4-hydroxynonenal comprised less than 5% of total cellular area under these conditions in the presence of captopril or losartan.
Although ischemia can damage cells directly, liver cells have defense mechanisms to protect against such insults if the ischemic time is relatively brief.21 However, even if the liver cells survive the ischemic insult, reintroduction of bloodflow (reperfusion) often leads to cellular damage.22, 23 The hypothesis is that this second “hit” to the liver during reperfusion is mediated predominantly by inflammation and generation of reactive oxygen species.24 In support of this hypothesis, pharmacological or genetic (i.e., knockout mice) suppression of adhesion molecule (e.g., ICAM-1) expression has been shown to protect against inflammation and subsequent hepatic damage caused by ischemia and reperfusion in liver.25, 26 Indeed, correlations between indices of inflammation (Figs. 3–5) and those of tissue damage (Figs. 1–3) were observed here.
As mentioned above (see Introduction), it has been proposed that the RAS may contribute to inflammation in target organs, including the liver. However, most work investigating such a role has been in models of chronic tissue damage (e.g., renal and hepatic fibrosis). Although some studies have suggested that the RAS is involved in some models of acute inflammation in liver,27, 28 the potential mechanisms by which this system may contribute to inflammation and damage are unclear. Here, 3 hours after the initiation of reperfusion, hepatic expression of angiotensinogen already was elevated fourfold over sham-treated animals (Fig. 4). Captopril or losartan had no effect on angiotensinogen expression, as would be expected, because RAS inhibitors work “downstream” of the formation of angiotensinogen. The timing of this response is similar to increased expression of known mediators of inflammation under these conditions (TNFα and ICAM-1; Fig. 4). Importantly, captopril or losartan completely blocked this increase in expression of TNFα and ICAM-1 (Fig. 4). RAS inhibition also had profound protective effects on subsequent liver damage (Figs. 1–3) and recruitment of inflammatory cells (Figs. 3 and 5).
The mechanisms by which the RAS mediates inflammation under this effect, however, are unclear. As detailed above, AngII has been shown directly to stimulate expression of proinflammatory mediators.9–11 Conversely, the RAS-dependent inflammation observed here could be the result of an indirect effect of this system on bloodflow. For example, Bailey et al.29 showed that hepatic damage in a pig model of cardiac shock can be prevented either by surgically impairing the RAS axis or by angiotensin converting enzyme inhibition with teprotide. The protection of these two procedures correlated with improved postischemic bloodflow to the organ. Nevertheless, whether the effects observed here are mediated by the local or systemic RAS, or both, these data support the hypothesis that the RAS is involved in mediating acute inflammation and damage to liver after ischemia and reperfusion.
As mentioned above, oxidative stress after reperfusion seems to be a critical component in subsequent tissue damage.24 There most likely are numerous sources of reactive oxygen (and nitrogen) species during reperfusion. For example, reperfusion itself leads to the generation of free radicals via enzymes such as xanthine oxidase.30 Furthermore, intrinsic inflammatory cells (e.g., macrophages) and recruited inflammatory cells (e.g., neutrophils) generate reactive oxygen species in liver. In addition to preventing the inflammation and damage, captopril or losartan also protected against oxidative stress, as determined by accumulation of 4-hydroxynonenal protein adducts in liver (Fig. 6). This protective effect was observed not only in the necroinflammatory foci, but also in hepatocytes that appeared undamaged by histological assessment. Taken together, these data clearly show that captopril or losartan had an antioxidant effect under these conditions.
The mechanism of this protective effect, however, is unclear. Although captopril is a thiol-containing drug that has been shown to scavenge free radicals directly, the relative rate of the this reaction has been proposed to be too slow to outcompete with other biologic molecules even at pharmacological doses.31 Furthermore, similar protective effects were observed with losartan, which does not possess direct free radical scavenging activity like captopril does. Therefore, it is unlikely that direct reactive oxygen species scavenging is the main mechanism by which these drugs prevented 4-hydroxynonenal adduct accumulation. A broader definition of antioxidants includes not only therapies that directly intercept free radicals, but also those that prevent the formation of these species.32 In this context, it is likely that prevention of the infiltration of inflammatory cells by captopril or losartan (Figs. 3 and 5) contributed to the observed antioxidant effect.
In summary, the results obtained here support the hypothesis that the RAS is involved in the acute inflammatory response in liver after ischemia and reperfusion. The observed protection seems to occur early in the response to ischemia and reperfusion, because early expression of mediators of inflammation (e.g., TNFα and ICAM-1) were blunted by RAS inhibition. Taken together, these data suggest an additional rationale for pharmacological blockade of the RAS in hepatic diseases involving inflammation, as well as fibrosis, such as in alcoholic liver disease.