The extent of liver steatosis in chronic hepatitis C virus infection is mirrored by caspase activity in serum

Authors


  • Potential conflict of interest: Nothing to report.

Abstract

Hepatic steatosis is a frequent histological alteration in chronic hepatitis C virus (HCV) infection that sensitizes the liver to cell injury, inflammation, and fibrosis via unclear mechanisms. Although apoptosis has been implicated in various liver diseases, its importance in HCV-associated steatosis is largely unknown. In this study, we investigated the role of caspases, the key regulators of apoptosis, and employed two novel caspase assays, an immunological and a luminometric enzyme test, to detect hepatic caspase activation in sera from HCV patients with different grades of steatosis. Our data show that increased caspase activation can be found not only in liver biopsies, but also in sera from HCV patients with liver steatosis. Patients with steatosis exhibited significantly higher serum levels of caspase activity compared with normal healthy individuals. Moreover, the extent of steatosis closely correlated with serum caspase activity, whereas in particular in cases of low or moderate steatosis, no correlation was found with aminotransferase levels. In conclusion, apoptotic caspase activation is considerably elevated in HCV-associated steatosis. More importantly, our data imply that measurement of caspase activation might be a sensitive serum biomarker to detect liver steatosis in patients with chronic HCV infection and other liver diseases. (HEPATOLOGY 2005.)

Steatosis is associated with the deposition of triglycerides in the liver and can either be a noninflammatory condition or can be associated with steatohepatitis, which might progress to end-stage liver disease.1 Hepatic steatosis is a common feature in patients with chronic hepatitis C virus (HCV) infection, and its presence is linked to increased liver injury and fibrotic progression.2–6 Higher grades of steatosis are associated with more advanced stages of fibrosis, and worsening of steatosis is a risk factor associated with progression of fibrosis. The pathogenesis of steatosis is complex and includes viral as well as host factors such as alcohol consumption, obesity, insulin resistance, and others.1

The mechanisms that cause increased hepatocellular injury in steatosis patients are largely unknown. It has been postulated that excess fat accumulation in the liver predisposes cells to hepatocellular injury, which is presumably caused by the toxicity of free fatty acids, oxidative stress, and lipid peroxidation. According to the “two-hit model,” increased lipid accumulation can trigger the release of reactive oxygen species and proinflammatory cytokines, DNA damage, adenosine triphosphate depletion, and alterations in mitochondrial membranes.7 There is also increasing evidence that apoptosis is elevated in liver steatosis.8 However, it remains unknown whether apoptosis influences disease progression or is solely a bystander effect.

Apoptosis is essentially controlled by caspases, intracellular cysteine proteases that cleave a selected number of protein substrates, thereby inducing the demise of the cell.9 Among the different substrates are proteins involved in DNA replication, cell cycle control, signal transduction, and structural proteins, including members of the cytokeratin type I family.10 Caspases are synthesized as inactive proenzymes and are proteolytically processed to form an active mature enzyme. Based on their order in cell death pathways, caspases can be divided into two major groups. Initiator caspases, such as caspase-8 and -9, exert apical regulatory roles. Upon binding to signaling molecules, they activate downstream effector caspases such as caspase-3, -6, and -7, which through substrate cleavage produce the typical apoptotic alterations. The activation of caspases is achieved via two principal signaling pathways.11 The extrinsic death pathway involves the ligation of death receptors that leads to the recruitment of caspase-8 into a death-inducing signaling complex. The intrinsic death pathway is initiated at the mitochondrion through the release of cytochrome c, a process that is controlled by proteins of the Bcl-2 family. When released, cytochrome c binds together with apoptosis protease-activating factor 1 to caspase-9 to form the apoptosome. Upon formation of the death-inducing signaling complex or the apoptosome, caspase-8 and -9 are autoproteolytically processed, resulting in the activation of effector caspases.

Apoptosis has been recently identified as an important feature of liver injury in chronic HCV infection.12–16 Although hepatocyte apoptosis can occur through a variety of mechanisms, death receptor–mediated apoptosis is a particularly prominent process in the liver.17 In hepatocytes, these mechanisms seem to require a mitochondrial amplification pathway, and HCV proteins can modulate this process, either to sensitize cells to apoptosis or to exert inhibitory effects. Several death receptor systems such as the CD95 ligand, tumor necrosis factor (TNF), and TNF-related apoptosis-inducing ligand pathways have been implicated in HCV-mediated liver injury. It has been shown that both CD95 and its ligand CD95L are upregulated in liver biopsies of chronic HCV infection, correlating with the severity of liver inflammation.18–20 An elevated number of apoptotic cells as assessed by terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling as well as an increased expression of CD95 and TNF have been found in alcoholic and nonalcoholic steatohepatitis.21, 22 Furthermore, in nonalcoholic steatohepatitis a positive correlation between hepatocyte apoptosis and hepatic fibrosis as well as inflammatory activity was observed. We have recently reported that caspase activation is considerably elevated in hepatocytes from chronic HCV patients compared with healthy individuals and correlated with the degree of liver injury.23, 24 These studies suggest that hepatocyte apoptosis is an important mechanism of liver damage that might contribute to liver fibrogenesis and cirrhosis.

The evaluation of liver biopsies is currently the standard procedure for assessing apoptosis and HCV-related liver disease. However, because this intervention is associated with a risk of clinical complications, there is a need for the development of novel noninvasive techniques to detect early liver damage. In this context, we have recently used the monoclonal antibody M30 that selectively recognizes a caspase-generated neoepitope of cytokeratin-18 (CK-18) which is abundantly expressed in hepatocytes.24–26 We detected increased caspase activation not only in liver biopsies, but also in serum samples from patients with chronic HCV infection.24

Because steatosis has been recognized as an important cofactor influencing disease progression in chronic HCV infection, we investigated the correlation between caspase activation in liver biopsies and in sera with different degrees of liver steatosis. By employing immunohistochemistry, an ELISA for caspase-generated CK-18 fragments, and a sensitive enzyme assay for caspase activity, we now show that caspase activation is an early and prominent feature of steatosis in HCV infection. The strict correlation found between serum caspase activity and the degree of steatosis suggests that caspase activation might be a sensitive serum marker to detect early liver injury, particularly in patients with a risk of progressing fibrosis.

Abbreviations

HCV, hepatitis C virus; TNF, tumor necrosis factor; CK-18, cytokeratin 18; RLU, relative light unit; AST, aspartate aminotransferase; ALT, alanine aminotransferase; ULN, upper limit of normal.

Patients and Methods

Patients.

We investigated sera from 66 patients (24 women, 42 men aged 15–73 years [mean age 45.21 ± 14.2]) with chronic HCV infection. The study was performed according to the guidelines of the Ethics Committee of the Hannover Medical School. The diagnosis was based on histological examination by two pathologists as well as the presence of anti-HCV antibodies and HCV RNA in serum for at least 6 months. The patients had no other causes of liver disease and did not have elevated levels of glucose, cholesterol, or triglycerides at the time of investigation. The mean value of the body mass index of all patients was 25.3 ± 0.52 (range, 18.2–38.1). An elevated body mass index (>25) was observed in 26 of 66 patients. None of the patients showed regulatory or excessive consumption of alcohol. The investigated patients did not receive HCV-specific therapy or drugs with liver toxicity at the time of investigation. The presence of anti-HCV antibodies and HCV RNA was quantified via ELISA (Abbott Diagnostics, Wiesbaden, Germany) and the Amplicor HCV Monitor test (Roche Diagnostics, Mannheim, Germany), respectively. Genotyping of HCV was performed via reverse hybridization assay (Inno LiPA HCV II; Innogenetics, Ghent, Belgium). The histological disease activity was assessed according to the ISHAK score27 using hematoxylin-eosin and Masson trichrome stains of formalin-fixed, paraffin-embedded liver. We selectively included patients with minimal necroinflammatory disease activity and low stages of fibrosis. Steatosis was assessed in steps of 5% by an independent pathologist and was graded as follows: no or minimal steatosis, 0% to 10% of hepatocytes affected; moderate steatosis, 10% to 40% of hepatocytes affected; severe steatosis, more than 40% of hepatocytes affected. The main clinical, biochemical, virological, and histological features in the patients at the time of liver biopsy are summarized in Table 1. Sera from 10 healthy volunteers served as control. Sera from patients and control individuals were stored at −20°C.

Table 1. Demographic and Clinical Features of Patients
 TotalNo or Minimal SteatosisModerate SteatosisSevere Steatosis
No. of patients66212025
Mean age ± SEM45.2 ± 1.837.6 ± 2.952.4 ± 2.845.8 ± 2.6
Sex (% male)64526076
Mean body mass index ± SEM25.3 ± 0.5223.8 ± 0.7825.1 ± 0.6726.5 ± 0.99
Stage of fibrosis ISHAK F (mean ± SEM)1.2 ± 0.10.4 ± 0.21.2 ± 0.31.9 ± 0.2
Necroinflammatory score ISHAK A-D (mean ± SEM)5.2 ± 0.24.6 ± 0.34.9 ± 0.45.9 ± 0.4
Mean AST (× upper limit of normal ± SEM)2.0 ± 0.21.2 ± 0.11.5 ± 0.12.8 ± 0.3
Mean ALT (× upper limit of normal ± SEM)2.9 ± 0.22.0 ± 0.32.1 ± 0.24.1 ± 0.4
Mean viral load (× 106/mL ± SEM)1.39 ± 0.21.33 ± 0.31.55 ± 0.31.32 ± 0.3
No. of patients with HCV genotype 144191312
No. of patients with HCV genotype 316169
No. of patients with unknown HCV genotype6114

Immunohistochemistry.

Frozen sections (5 μm) of liver biopsy specimens from patients with different grades of steatosis were examined. After blocking of endogenous peroxidase with 0.3% hydrogen peroxidase in methanol and washing in dimineralized water, nonspecific binding was blocked with 1% bovine serum albumin for 1 hour. Next, the M30 monoclonal antibody (0.5 μg/mL in 1% bovine serum albumin; Roche Diagnostics), recognizing the neoepitope formed by caspase cleavage of CK18, was incubated overnight at 4°C. After repeated washings, bound antibody was detected with peroxidase-labeled dextran polymers coupled to goat anti-mouse immunoglobulin G (EnVision; DAKO, Hamburg, Germany). After 30 minutes, the sections were washed in PBS and stained in freshly prepared substrate solution (4 mg aminoethylcarbazole in 10 mL sodium acetate buffer [pH 4.9], 500 μL dimethylformamide, 0.03% hydrogen peroxide) for 10 minutes. The reaction was stopped by extensive rinsing in demineralized water. Sections were subsequently counterstained with hematoxylin and mounted in glycerol/gelatin.

Serological Detection of Caspase Activity.

Caspase activity in serum samples was determined via two independent methods. For the quantitative measurement of the apoptosis-associated neoepitope in the C-terminal domain of CK-18 (amino acids 387–396), we used the M30-Apoptosense ELISA kit (Peviva; Alexis, Grünwald, Germany)24, 28 according to the manufacturer's instructions. In addition, we established a novel luminescent substrate assay for active caspase-3 and -7 (Caspase-Glo assay; Promega, Mannheim, Germany). The assay provides a luminescent substrate with the caspase cleavage sequence DEVD in a reagent optimized for caspase activity and luciferase activity. Following cleavage of the substrate at the DEVD peptide by caspase-3 and caspase-7, aminoluciferin is released, resulting in light production in a luciferase reaction, which can be measured in relative light units (RLU) by a luminometer. First, sera from 66 patients with chronic HCV infection and 10 healthy individuals were diluted 1:1 in buffer containing 50 mmol/L Tris-HCl [pH 7.4], 10 mmol/L KCl, and 5% glycerol. Then, 10 μL of the diluted serum were incubated with 10 μL caspase substrate for 3 hours at room temperature. Finally, the luminescence of the samples was measured in a luminometer. A statistical analysis comparing the concentration of the M30 antigen (U/L), caspase activity (RLU) or aminotransferase levels (U/L) with the degree of steatosis was performed using the Mann-Whitney U test. A P value less than .05 was considered significant. All assays were performed in duplicate. Intra-assay variations were less than 4%, while interassay variations were less than 10%.

Results

Detection of Caspase-Mediated CK-18 Cleavage in Liver Biopsies From Patients With HCV-Associated Steatosis.

Apoptosis has been recently recognized as an important feature of liver injury in HCV infection.12–15 To investigate the role of apoptotic caspase activation in HCV-associated steatosis, 66 patients with different grades of steatosis were recruited. We selectively included patients with minimal necroinflammatory disease activity (ISHAK A-D 5.2 ± 0.2) and low stages of fibrosis (ISHAK F 1.2 ± 0.1). In the first experiments, we evaluated caspase activation by immunohistochemistry of liver biopsies of representative patients, and therefore used the monoclonal antibody M30 that selectively detects a caspase-generated neoepitope of the caspase substrate CK-18. Almost no immunoreactivity was evident in liver biopsies with no or minimal steatosis (Fig. 1A). In contrast, caspase activity was considerably elevated in liver biopsies with advanced steatosis. Areas with moderate steatosis (Fig. 1B) showed lower caspase activity compared with those with higher steatosis (Fig. 1C). The majority of hepatocytes showed a granular staining for the caspase-generated CK-18 cleavage fragments, which could be seen in ballooned hepatocytes, but also in cells that were not fat-engorged. The localization of the CK-18 fragments is in line with previous studies in cultured tumor cells, demonstrating that caspase-mediated cleavage results in the disruption of the keratin network and the formation of spheroidal inclusion bodies.29 Thus, the data provide morphological evidence that apoptotic caspase activation occurs in hepatocytes and is elevated in steatotic patients.

Figure 1.

Immunohistochemical detection of caspase-generated CK-18 cleavage fragments in liver biopsies from patients with chronic HCV infection and different grades of steatosis. Liver tissue sections were immunostained with a monoclonal antibody recognizing a specific caspase-generated neoepitope of CK-18 (original magnification ×400). (A) Weak immunoreactivity was detected in areas of liver sections from patients with minimal or no steatosis. In contrast, a prominent granular staining pattern of caspase-mediated CK-18 cleavage fragments was seen in representative liver sections of an area with (B) moderate steatosis that was even more pronounced in areas with (C) severe steatosis. Note that many hepatocytes, albeit positive for caspase activation, do not exhibit an overt apoptotic morphology, indicating that CK-18 cleavage marks a very early event in the apoptotic process. The use of a control antibody revealed no staining in either case (data not shown). Representative examples of the different groups of patients are shown. Arrowheads indicate immunoreactive hepatocytes with a fat-transformed phenotype.

Correlation of Serum Caspase Activity and the Grade of HCV-Associated Liver Steatosis.

We next investigated whether the extent of caspase activation in the serum of patients with chronic HCV infection correlated with the extent of hepatocyte steatosis. Serological caspase activity was measured with a novel ELISA using the same antibody, which we used for immunohistochemistry to detect the caspase-generated neoepitope of the CK-18 proteolytic fragment. Patients with no or minimal steatosis (0%-10%) (n = 21) were compared with patients with moderate steatosis (10%–40%) (n = 20) and patients with severe steatosis (>40%) (n = 25). As shown in Fig. 2A, healthy individuals (n = 10) who served as a control showed only low serum levels of CK-18 cleavage fragments (mean, 74.8 ± 5.4 U/L; range, 50–100 U/L). In contrast, caspase activity was elevated in the total number of HCV patients compared with normal healthy control individuals, as indicated by increased concentrations of truncated CK-18 in serum (260.0 ± 26.2 U/L; P < .005). Moreover, significantly higher caspase activity was detected in patients with severe steatosis (mean, 387.3 ± 49.0 U/L; range, 104–1,000 U/L) compared with individuals with moderate steatosis (mean, 229.3 ± 41.1 U/L; range, 79–875 U/L) or minimal or no steatosis (mean, 137.6 ± 16.7 U/L; range, 64–318 U/L).

Figure 2.

Caspase activation in sera from patients with chronic HCV infection and different grades of liver steatosis. Caspase activation was assessed via ELISA detecting caspase-generated CK-18 fragments (A) as well as by a novel luminescent substrate assay for active caspase-3/-7 (B). Patients (n = 21) with no or minimal steatosis (0%-10%) were compared with patients with moderate steatosis (10%–40%) (n = 20) and patients with severe steatosis (>40%) (n = 25). Normal healthy individuals (n = 10) served as control. Caspase activity is significantly elevated in patients with liver steatosis compared with healthy control individuals (P < .005) as indicated by both the ELISA for truncated CK-18 and the luminescent assay for active caspase-3/-7. Caspase activity was most strongly evident in patients with severe hepatocyte steatosis compared with those with moderate or minimal steatosis. Data represent the mean values ± SEM. *P < .05. **P < .005. RLU, relative light units.

To verify the data obtained via ELISA, we additionally established a novel luminometric enzyme assay for the measurement of caspase activity. To this end, serum samples were incubated with the caspase substrate DEVD-aminoluciferin, which upon cleavage by caspase-3 and -7 generates a luminescent signal that can be measured in a luciferase reaction. As shown in Fig. 2B, healthy individuals only showed low caspase activities (mean, 2,598 ± 272 RLU; range, 1,554–4,094 RLU), whereas in all HCV patients caspase activity was strongly elevated (mean, 12,359 ± 1,190 RLU; P < .005). Significantly higher caspase activity (P < .005) was measured in patients with severe steatosis (mean, 19,693 ± 2,368 RLU; range, 6,267–52,811 RLU) compared with patients with moderate steatosis (mean, 9,645 ± 913 RLU; range, 3,189–19,991 RLU) or no or minimal steatosis (mean, 6,214 ± 516 RLU; range, 3,409–13,033 RLU). Regression analysis of the 66 patients with different grades of hepatocyte steatosis (Fig. 3) revealed that the percentage of hepatocyte steatosis correlated positively with the serum concentration of truncated CK-18 (r = 0.55; significant correlation at 0.01 level/two-tailed) as well as with serum activity of caspases as detected by the luminometric assay (r = 0.61; significant correlation at 0.01 level/two-tailed).

Figure 3.

Regression analysis correlating (A) caspase-mediated CK-18 cleavage fragments or (B) caspase-3/-7 activity with the percentage of liver steatosis. A significant correlation (at 0.01 level/two-tailed) was observed between caspase-mediated CK-18 fragment cleavage (r = 0.55) as well as between the direct detection of serum caspase proteolytic activity (r = 0.61) and the percentage of hepatocyte steatosis. RLU, relative light units.

Aminotransferase Levels Do Not Correlate With the Grade of Liver Steatosis.

The previous data clearly show that higher grades of steatosis are associated with strongly elevated caspase activity. We therefore asked whether different grades of steatosis are also associated with significant changes in aminotransferase levels. As shown in Fig. 4, in contrast to serological caspase activity, no significant differences (P > .1) in aspartate aminotransferase (AST) or alanine aminotransferase (ALT) levels were observed, when patients with minimal or no steatosis (AST: mean, 1.2 ± 0.1, upper limit of normal [ULN]; range, 0.7–2.2 ULN; ALT: mean, 2.0 ± 0.3 ULN; range, 0.5–5.1 ULN) were compared with patients with moderate steatosis (AST: mean, 1.5 ± 0.1 ULN; range, 0.6–2.8 ULN; ALT: mean, 2.1 ± 0.2 ULN; range, 1.0–4.4 ULN). Aminotransferase levels were significantly elevated (P < .05) only in the serum from patients with severe steatosis (AST: mean, 2.8 ± 0.3; range, 0.7–7.2 ULN; ALT: mean, 4.1 ± 0.4 ULN; range, 0.9–11.2 ULN) compared with those with moderate or minimal or no liver steatosis (Fig. 4).

Figure 4.

Detection of (A) AST or (B) ALT levels in patients with different grades of liver steatosis. No significant differences (P > .1) were observed between patients with no or minimal steatosis (0%–10%) and moderate steatosis (10%-40%). A significant elevation of aminotransferases (P < .05) was observed only when patients with severe (>40%) steatosis were compared with those with no or minimal steatosis or moderate steatosis. **P < .005. AST, aspartate aminotransferase; ALT, alanine aminotransferase; n.s., nonsignificant.

Discussion

Although liver cell apoptosis was previously demonstrated in HCV infection, the influence of steatosis has not thoroughly been investigated. We now show that caspase activation is a prominent feature of HCV infection that is closely associated with steatosis. To investigate the role of apoptotic caspases, we employed two novel tools that allowed for the assessment of caspase activity in serum samples. In the first assay, we measured the concentration of a caspase-generated fragment of CK-18 via ELISA. Because CK-18 is restricted to epithelial cells and is abundantly expressed in hepatocytes, this assay does not detect apoptosis of nonepithelial cells such as lymphocytes. We previously showed that the M30 antibody specifically labels early apoptotic but not necrotic cells and, moreover, immunoprecipitates cleaved CK-18 in HCV serum samples.24, 26 The second assay analyzed the proteolytic activity of caspases via a luminometric technique. This is the first study in which caspase activity was directly measured in serum samples. So far, caspase activation is mostly detected via colorimetric or fluorimetric substrate assays in cell lysates. However, due to the quenching of the signal by hemoglobin, these techniques are not suitable for serum samples. The luminometric assay not only provides a high sensitivity, it also circumvents this quenching problem.

Because we detected caspase activity in the serum of steatosis patients, we have to conclude that active caspases are somehow released from apoptotic hepatocytes. The exact mechanism leading to the secretion of caspases and CK-18 fragments into the blood of HCV patients has yet to be determined. During execution of apoptosis, caspase activation results in the disruption of the keratin network and the formation of small spheroidal cytoplasmic inclusions containing cleaved CK-18 as well as activated caspases.29 The sequestration of caspases and CK-18 into inclusion bodies suggests that both molecules are released from apoptotic cells via similar mechanisms. Consistent with this, the degree of steatosis correlated with the levels of caspase-mediated CK-18 fragments and the caspase activity in a similar manner. This not only underlines the reliability of our measurements, but also indicates that a concurrent caspase activation, which might have occurred in nonhepatocytes, was not detected by the luminescent caspase assay.

In contrast to caspase activation, no significant differences were observed between no or minimal steatosis and moderate steatosis regarding aminotransferase levels. Only patients with severe steatosis showed significantly higher aminotransferase levels. There are several possibilities that could explain the lack of a strict correlation between the serum levels of aminotransferases and caspase activity. Whereas aminotransferases are also released during necrosis, generation of CK-18 fragments requires apoptotic caspase activation, and their release might occur during secondary necrosis. In vivo studies demonstrated that hepatocyte apoptosis is associated with increased aminotransferase values, but that the release of aminotransferases is lower in apoptosis than in necrosis.13 Thus, the relative differences in the occurrence of apoptosis and necrosis could account for the lack of a strict correlation between aminotransferases and lower CK-18 fragment levels.

Several studies have provided evidence for the existence of multiple factors underlying steatosis in chronic HCV infection.2–6, 30–35 In patients with genotype 3, there is a “viral” steatosis that is directly linked to HCV replication and that disappears upon successful antiviral therapy.30–34 Steatosis in these patients seems to depend on a HCV-mediated disturbance of lipoprotein secretion and low apolipoprotein B levels. Indeed, in our study the majority of patients with genotype 3 (56%) had a more severe steatosis compared with patients with genotype 1 (27%). Patients with genotype 3 also showed higher serum caspase activities than patients with HCV genotype 1 (data not shown), although these differences were not significant in our study, presumably because of the low incidence of genotype 3 infection. In contrast, in patients with genotype 1, steatosis is generally associated with a metabolic syndrome due to obesity, hypercholesterinemia, alcohol abuse, and insulin resistance.2, 31, 35 The severity of fat accumulation in these cases is often proportional to body mass index and is not responsive to antiviral therapy. However, in our study, patients did not show a metabolic syndrome, and the average body mass index did not significantly differ in patients with low, moderate, or severe steatosis. Furthermore, only a weak correlation (r = 0.16) between caspase activation (as measured via ELISA) and body mass index was observed. Thus, additional factors might account for the progression of liver injury in HCV patients with liver steatosis.

According to the “two-hit” model, deposition of hepatic steatosis represents the first hit, which requires additional factors as a second hit for disease progression to inflammation, liver damage, and fibrosis. A strong candidate for this second hit is oxidative stress leading to the generation of proinflammatory products of lipid peroxidation as well as to cytokine release.5 Whether apoptosis is a primary mechanism promoting steatohepatitis or is a secondary phenomenon resulting from tissue inflammation cannot fully be elucidated, because in the present study we selectively included patients with minimal necroinflammatory disease activity and low stages of fibrosis. It is well known that there is a significant association between steatosis and inflammation in liver biopsies.1, 5, 36 Liver injury triggers the recruitment and activation of macrophages and neutrophils, which produce a number of injurious factors that may promote hepatocyte injury. The most critical of these may be cytokines such as TNF, TNF-related apoptosis-inducing ligand, and CD95 ligand, all of which are proinflammatory but are also potent inducers of apoptosis. Many of these death ligands and other cytokines are induced by oxidative stress through the activation of a different signaling pathway.37, 38 Our observation that many hepatocytes showing caspase activation were not fat-engorged might argue for a more indirect and immunomediated mechanism.

In addition, there may be several vicious cycles, and apoptosis could provide an extra second hit that exacerbates disease progression. A hallmark of apoptosis is the removal of apoptotic bodies by phagocytes without eliciting an inflammatory response. However, there is increasing evidence that apoptosis may also promote inflammation, and that inflammation can occur in response to hepatocellular apoptosis.39 Several studies have shown that apoptosis triggers neutrophil recruitment into liver parenchyma.40, 41 Thus unorchestrated and continuous apoptosis may contribute to necroinflammation and liver damage. A recent study demonstrated a significant independent association between steatosis and increased apoptosis in HCV infection,20 indicating that lipid acccumlation might also have direct effects on hepatocyte death. Accumulation of lipid in nonadipose tissues has been reported to trigger the generation of specific proapoptotic lipid species or signaling molecules (e.g., ceramides, reactive oxygen species, nitric oxide), leading to mitochondrial dysfunction or cell death.42 In addition, in several rodent models and human cells, long-chain fatty acids have been shown to suppress antiapoptotic factors such as Bcl-2 and to activate proapoptotic mediators, including p53 and Bax.43–46

Steatosis has been linked to both apoptosis and fibrosis, although the relative contribution of steatosis and apoptosis to fibrogenesis has yet to be delineated.47 Steatosis may promote cellular injury and death independently from its effect on fibrosis, and apoptosis, driven by steatosis and/or inflammation, may enhance fibrogenesis. In support of this latter idea, engulfment of apoptotic bodies by hepatic stellate cells is associated with activation of these cells, resulting in increased transforming growth factor β and collagen production.48 Additionally, inhibition of hepatocyte apoptosis has been shown to reduce fibrogenesis in animal models of cholestasis.49 Convincing evidence for a link between apoptosis and fibrogenesis has recently been established by the hepatocyte-specific genetic disruption of Bcl-xL, which led to hepatocyte apoptosis and liver fibrotic responses.50 In line with these findings, we have previously shown that caspase activation is associated with more progressive fibrotic liver alterations, even in patients with normal aminotransferase levels.24

In conclusion, our present results are consistent with recent findings that hepatocyte apoptosis is elevated in HCV-infected liver biopsies with advanced steatosis.20 Moreover, we now show that measurement of caspase activation in sera from patients with HCV infection might serve as an early marker to detect the extent of hepatocyte steatosis and progressing liver disease. Detection of caspase activation in serum may also be useful for monitoring the therapeutic efficacy of current and future antiviral therapies. It will be interesting to see whether inhibition of apoptosis can prevent the fatal consequences of chronic liver disease. In this respect, novel caspase inhibitors that are currently tested in different clinical trials for prevention of HCV-mediated liver injury will be of particular interest.

Acknowledgements

The authors wish to thank H. Kreipe for providing materials, H. Wedemeyer and H. Tillmann for helpful comments, and B. Vaske for statistical analysis.

Ancillary