Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Programmed cell death, or apoptosis, can be initiated by the activation of the CD95 (APO-1/Fas) receptor, as well as the closely related tumor necrosis factor receptor I (Galle and Krammer, 1998; Nagata, 1999). These receptors are located in relatively high concentrations in the liver (Watanabe-Fukunaga et al., 1992). Normally, apoptosis is thought to control the number of hepatocytes within the liver (Benedetti et al., 1988). However, under pathological conditions, uncontrolled cell death leads to tissue injury or even organ destruction. Abnormally high amounts of apoptosis occur clinically in a number of liver diseases, including alcoholic hepatitis, hepatitis B or C viral infection, and Wilson's disease, as well as in livers of patients undergoing orthotopic liver transplantation (Kerr et al., 1979; Kawahara et al., 1994; Strand et al., 1998; Tannapfel et al., 1999). Moreover, the CD95 pathway has been implicated in all these conditions (Galle et al., 1995; Strand et al., 1998; Taieb et al., 1998; Tannapfel et al., 1999).

Experimentally, the apoptotic pathway in hepatocytes can be initiated by activation of the CD95 pathway by administration of an anti-Fas antibody (-Fas), which is agonistic for the Fas receptor (Ogasawara et al., 1993). Administration of an -Fas antibody results in marked hepatocellular apoptosis beginning as early as 2 h after administration, with the onset of fulminant hepatitis and lethality occurring near 5 h. Therefore, this model is a valuable system to study Fas-mediated hepatocellular apoptosis. Apoptosis triggered by the Fas pathway occurs, at least in part, by the activation of intracellular enzymes called caspases. Caspases are a family of cysteine proteases that are normally found in an inactive pro-form (Cohen, 1997). Upon activation of the apoptotic pathway, initiator caspases (i.e., caspases 8 and 9) are converted to their active forms, which in turn activate downstream caspases (i.e., caspases 3, 6, and 7). Cellular targets of the effector caspases include endonucleases and cytoskeletal proteins (for review, see Earnshaw et al., 1999). Additionally, Bid, a protein responsible for the release of mitochondrial cytochrome C and amplification of the apoptotic cascade, is also activated by caspases (Luo et al., 1998). Ultimately, this pathway ensures the self-destruction of the cell with the characteristic morphological features of apoptosis, including nuclear fragmentation, cellular shrinkage, and acidophilic staining of the cytoplasm.

Activation of caspases has been demonstrated in the -Fas model, and previous studies have shown that caspase inhibitors [carbobenzyloxy-valinyl-alanyl-aspartyl-fluoromethylketone (Z-VAD-FMK), tyrosinyl-valinyl-alanyl-aspartyl-chloromethylketone (YVAD-CMK), and aspartyl-glutamyl-valinyl-aspartyl-aldehyde (DEVD-CHO)] are effective in eliminating Fas-mediated liver injury (Rodriguez et al., 1996; Rouquet et al., 1996; Kunstle et al., 1997; Suzuki, 1998). However, no reports have thoroughly characterized a caspase inhibitor for efficacy and its effects on caspase activation. The aim of this paper is to more fully characterize the caspase inhibitor IDN-1965 (N-[(1,3-dimethylindole-2-carbonyl)valinyl]-3-amino-4-oxo-5-fluoropentanoic acid) in this model of apoptosis-mediated liver injury, with particular emphasis of its effect on caspase activation. Results clearly indicate that this drug is both potent and effective postinsult, as well as by multiple routes of administration; but, despite its irreversible nature, it has only short-lived effects. The therapeutic potential of this and other caspase inhibitors is briefly discussed for indications where abnormally high amounts of apoptosis occur.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials

-Fas was purchased from PharMingen (Jo-2; PharMingen, San Diego, CA). Pentobarbital (Nembutol) was purchased from Western Medical Supply (Arcadia, CA). Alzet osmotic pumps were purchased from Alza Corp. (Palo Alto, CA). IDN-1965 was synthesized as described (U.S. patent no. 5,869,519). All other materials were purchased from Sigma Chemical Co. (St. Louis, MO) unless otherwise indicated.

Animals and Treatment

Male BALB/c mice (Harlan Sprague-Dawley, Indianapolis, IN) weighing approximately 20 to 25 g were provided food and water ad libitum before the experiment. Animals were acclimated for at least 7 days before experimentation.

-Fas Model. In all studies, -Fas was injected i.v. at a dose of 200 µg/kg via the tail vein. IDN-1965 was administered in either saline, DMSO:saline (1:1 ratio), or DMSO:water (1:1) at times indicated in the figures. Controls were administered similar volumes of appropriate vehicle (4 ml/kg). Preliminary experiments determined that different vehicles alone do not affect -Fas-mediated liver injury in this model (data not shown). When comparing efficacy by multiple routes of administration (Table 1), IDN-1965 was administered i.p., p.o., or i.v. in sterile saline immediately after administration of -Fas at concentrations indicated in Table 1. In osmotic pump studies, isoflurane-anesthetized mice received Alzet osmotic pumps implanted into the peritoneum and were allowed to recover for 24 h. Pumps contained either 50% water:DMSO or IDN-1965 (30 mg/ml) in 50% water:DMSO. In most experiments, animals were euthanized with pentobarbital (50 mg/kg, i.p.) 1.5 to 24 h after the administration of antibody. In lethality studies, surviving mice were allowed to live for 7 days after -Fas administration before euthanization with pentobarbital. Plasma was harvested for measurement of alanine aminotransferase (ALT) activities. Portions of liver were fixed in formalin (10% in neutral-buffered saline) for immunohistochemical and morphological analyses, as well as frozen in liquid nitrogen for caspase activity determination.

View this table: [in this window] [in a new window]   TABLE 1 Potency of IDN-1965 by multiple routes of administration in the -Fas model of liver injury -Fas was administered i.v. (200 µg/kg) and immediately following injection, mice were treated with either saline vehicle or IDN-1965 (0.003-3 mg/kg). Mice were killed 6 h after -Fas administration, and ALT activities were measured in plasma.

Assays

Caspase Activity. Livers were frozen in liquid nitrogen immediately upon collection and stored until processed. Tissues were homogenized in ice-cold hypotonic buffer (10 mM HEPES, pH 7.4; 42 mM KCl; 50 mM MgCl₂·6H₂O) containing 1 mM dithiothreitol (DTT), 0.5% (w/v) CHAPS (Sigma Ultrapure), and a cocktail of protease inhibitors [100 µM EGTA, 100 µM EDTA, 1 µg/ml pepstatin A, 10 µg/ml aprotinin, 1 µg/ml leupeptin, and 1 mM phenylmethylsulfonyl fluoride (PMSF)] and centrifuged at 12,000g for 15 min at 4°C. Protein concentrations of the resulting supernatant suspension were determined routinely using a bicinchoninic acid method (Pierce, Inc., Rockford, IL) with bovine serum albumin as the standard. The tissue sample extracts were assayed for caspase enzyme activity by monitoring the fluorescent product of the cleavage of DEVD-amc. Briefly, 200 µg of protein/well was added to 96-well microtiter plates and diluted to 50 µl by addition of ICE buffer [consisting of 20 mM HEPES (pH 7.5), 1 mM EDTA, 10% sucrose, and 0.1% CHAPS] to which DTT was added fresh before the start of every assay at a final concentration of 10 mM. The enzyme assay was initiated by addition of the DEVD-amc substrate to a final concentration of 12.7 µM (100 µl of a 20 µM stock). The substrate cleavage activity was measured at time 0 and every 30 min thereafter for 2 h in a fluorescent plate reader (excitation wavelength: 360/40 nm; emission wavelength: 460/40 nm).

Immunohistochemical Analysis of Activated Caspase 3. CM-1 antibody, which recognizes the p18 subunit of the cleaved caspase 3 (activated caspase 3), was generated in rabbits, purified, and characterized as previously described (Srinivasan et al., 1998b). Formalin-fixed, paraffin-embedded livers were cut into 5-µm sections and rehydrated by routine methods. After elimination of endogenous peroxidase activity with 3% H₂O₂, tissues were subjected to an antigen retrieval step (boiling in 0.1 M sodium citrate for 15 min), followed by incubation in blocking buffer for 1 h. Tissue sections were then exposed to 0.4 µg/ml antibody (CM-1) for 1 h in blocking buffer. After extensive washing, amplification of the signal was performed using a biotinylated antibody kit (Vector Laboratories, Burlingame, CA) with diaminobenzidine as the indicator. Tissues were counterstained with Mayer's hematoxylin. Negative controls consisted both of tissues incubated without primary antibody or with nonspecific IgG antibody.

Western Immunoblotting. Frozen liver tissues from control and -Fas-treated mice were thawed in lysis buffer [50 mM KCl, 50 mM Pipes (pH 7.4), 10 mM EGTA, 2 mM MgCl₂, 1 mM DTT, 0.1 µM PMSF, 1 mM cytochalasin B, 2 µg/ml leupeptin, 1 mg/ml pepstatin A, and 10 mg/ml aprotinin] and homogenized on ice using a Polytron tissue homogenizer. Ground tissue was then further homogenized using a Dounce homogenizer (100 strokes), transferred to 1.5-ml microfuge tubes, and subjected to five rounds of dry ice/37°C freeze-thaw cycles. Samples were centrifuged at 10,000g at 4°C for 20 min to remove solid material. Supernatants were centrifuged for 1 h at 100,000g at 4°C. Protein concentrations of the S100 supernatants were measured using the Protein Assay Kit II (Bio-Rad, Hercules, CA) with bovine serum albumin as the calibration standard. Western blotting of the S100 supernatants was done as previously described (Krebs et al. 1999). Briefly, liver extracts (60 µg of protein per lane) were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis on 12% or 16% gels (Novex, La Jolla, CA) and transferred to Immobilon polyvinylidene difluoride membranes (Millipore, Bedford, MA). Membranes were blocked in phosphate-buffered saline/0.1% Tween 20 (PBST) + 0.4% casein (I-block, Tropix, Bedford, MA). Blots were incubated in 1 µg/ml primary antibody diluted in PBST/casein for 1 h. Following three washes in PBST, blots were incubated for 1 h in 1:15,000 dilutions of horseradish peroxidase-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch Laboratories, West Grove, PA) in PBST/casein. Following the secondary antibody incubations, the blots were developed using the ECL-plus kit (Amersham Pharmacia Biotech, Piscataway, NJ). The blots were visualized using the STORM fluorescence imager (Molecular Dynamics, Sunnyvale, CA). The antibodies used were against caspase 8 (Srinivasan et al., 1998a), caspase 3 (Srinivasan et al., 1998b), caspase 3 large subunit (Srinivasan et al., 1998a), human Bid (Ref., kind gift of Dr. X. Wang, University of Texas Southwestern Health Science Center, Dallas, TX), and caspase 6 large subunit (C6LS-1, Idun Pharmaceuticals). The rabbit polyclonal antibody to the large subunit of caspase 6 was generated using a protocol identical to that used to generate CM-1 (Srinivasan et al., 1998b), except in the peptide sequence of the immunogen. The sequence of the peptide used to generate the rabbit sera was NH₂-CRGNQHDVPVIPLDVVD-COOH conjugated to keyhole limpet hemocyanin. The sequence corresponds to amino acids 163 through 179 of human caspase 6. All antibodies used were affinity-purified rabbit polyclonals, except the antibody to human Bid, which was used as an unpurified rabbit serum.

Morphological Analysis. Liver sections (5 µm) were stained with H&E for morphological analysis. Representative photomicrographs are shown in the figures.

ALT Activities. ALT activities were determined using a standard diagnostic kit (Sigma).

Statistical Analyses. Data are expressed as the mean ± S.E.M. ED₅₀ dose-response curves were calculated by plotting ALT as a percentage of -Fas-treated positive controls-vehicle controls, followed by nonlinear regression analysis using GraphPad Prism Software, Inc. (San Diego, CA). When comparing significant differences between groups, either a one-way or two-way analysis of variance was performed followed by either a Student Newman-Keuls test or Bonferroni's post hoc test. Differences were considered significant if p < 0.05.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Previous studies demonstrated that administration of -Fas antibody results in massive hepatocellular apoptosis and peak ALT activities near 6 h postadministration (Ogasawara et al., 1993). When administered simultaneously with -Fas (200 µg/kg, i.v.), i.p. administration of IDN-1965 dose dependently decreased in ALT activities when measured 6 h after -Fas administration. A representative experiment is shown in Fig. 1, and the average ED₅₀ based on three experiments for IDN-1965 was 0.14 ± 0.08 mg/kg. To determine the potency of IDN-1965 by multiple routes of administration, the ED₅₀ for liver injury after i.p., p.o., and i.v. administration was determined (Table 1). The potency was similar for i.p. and i.v. administration, but p.o. administration was less potent. Histologically, massive hepatocellular apoptosis and hemorrhaging was noted in livers treated with -Fas after 6 h, whereas only a few hepatocytes were apoptotic in livers treated with -Fas and IDN-1965 at doses of 1 mg/kg or higher (Fig. 2, A and B).

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Dose-dependent inhibition of -Fas-induced liver injury by IDN-1965. Mice were administered -Fas (200 µg/kg; i.v.) in saline and cotreated simultaneously with either saline vehicle or IDN-1965 (0.01 to 1.0 mg/kg, i.p.). Mice were killed 6 h later, and plasma was harvested for measurement of ALT activities. Data are expressed as (drug/-Fas-vehicle)/(-Fas-vehicle) × 100%. ED50 is expressed with 95% confidence limits in parentheses. n = 6 to 10/group.

View larger version (125K): [in this window] [in a new window]   Fig. 2.   Morphological analysis of livers from mice treated with -Fas: protection by IDN-1965. Sectioned livers were stained with H&E. A, representative liver from an -Fas-treated mouse (6 h post--Fas administration). Note massive hemorrhaging, numerous apoptotic hepatocytes, and loss of normal architecture. B, representative liver from an -Fas/IDN-1965 (1 mg/kg)-treated mouse (6 h post--Fas administration). Note only a few apoptotic hepatocytes (arrowheads) and normal morphology. Original magnification, 20×.

To investigate the effect of IDN-1965 on caspase activation, caspase 3-like activities were measured in the liver between 1.5 and 6 h after -Fas administration. Caspase activities markedly increased 3 h after -Fas administration and remained elevated over the 6-h period (Fig. 3A). Coadministration of 3 mg/kg IDN-1965 by i.p. injection inhibited caspase activities at 3 h, but caspase activity increased to similar levels as the -Fas-treated livers by 6 h. In contrast, a single i.p. injection of 10 mg/kg IDN-1965 completely inhibited caspase activities over the 6-h period (Fig. 3A). ALT activities at 6 h were inhibited by 93 and 98% in the 3 and 10 mg/kg dose groups, respectively (Fig. 3B).

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Effect of IDN-1965 on ALT and caspase 3-like activities after -Fas administration. Mice were injected with -Fas (200 µg/kg, i.v.) in saline vehicle and cotreated simultaneously with either vehicle or IDN-1965 (3 to 10 mg/kg, i.p.). Mice were killed 1.5 to 6 h later. Plasma and livers were harvested for measurement of ALT and caspase activities, respectively. A, ALT activities. B, caspase activities. Data are expressed as the mean ± S.E.M. n = 3 to 7/group. Control represents plasma or liver harvested 6 h after administration of vehicle.

An immunohistochemical analysis of livers for activated caspase 3, using an antibody (CM-1) that recognizes the processed (active) form of caspase 3 in the liver of -Fas-treated mice also revealed a similar time course of caspase activation. Positive staining occurred in a few hepatocytes 1.5 h after -Fas administration, but was pronounced in a number of hepatocytes by 3.0 h (Fig. 4, A and B). Detection of activated caspase 3 protein was extensive by 4.5 h post--Fas administration (Fig. 4C) and remained high over the 6-h period (not shown). These data coincide with caspase activities measured by DEVD-amc cleavage (Fig. 3B). When -Fas-treated mice were cotreated with 10 mg/kg IDN-1965, few CM-1 positive cells were noted even 6.0 h after -Fas injection (Fig. 4D). Negative controls (using nonspecific IgG instead of CM-1 antibody) had little or no staining, indicating the specificity for the antibody (not shown).

View larger version (158K): [in this window] [in a new window]   Fig. 4.   CM-1 immunohistochemical analysis of livers from mice treated with -Fas: protection by IDN-1965. Representative livers from mice treated with -Fas for 1.5 to 4.5 h (A-C) or -Fas-treated mice receiving IDN-1965 (10 mg/kg, i.p.). Procedure described under Experimental Procedures. Representative photomicrographs of livers from -Fas-treated mice (1.5 to 4.5 h post--Fas administration, A-C, respectively). Note increasing numbers of CM-1 positive hepatocytes. By 4.5 h (C), note extensive immunopositivity and loss of liver architecture. D, liver of -Fas- and IDN-1965-treated mouse (4.5 h post--Fas administration). Note few CM-1 positive hepatocytes. Original magnification, 40×.

Increases in caspase activation 6 h after -Fas administration, despite cotreatment of IDN-1965, raised the question whether liver injury would occur at later time points. Therefore, the effect of a single dose of IDN-1965 (1 mg/kg) on survival and liver injury in -Fas-treated mice was assessed over a 7-day period. This was also compared with a continuous administration of IDN-1965 via an osmotic minipump implanted i.p. over the same period of time. In this experiment, 7 of 10 -Fas-treated mice died over a 7-day period after -Fas administration, with the majority of mice dying between 4 and 12 h post-treatment. ALT activities 6 h after -Fas administration in the surviving mice (n = 3) were 6215 ± 939 IU/l (Fig. 5) and decreased to 1364 ± 542 IU/l by 24 h. In contrast, a single dose of 1 mg/kg IDN-1965 completely eliminated -Fas-induced lethality up to 7 days post-treatment and markedly lowered ALT activities at 6 h (152 ± 140 IU/l). However, ALT activities were not significantly different than -Fas-treated mice at 24 h (969 ± 251 IU/l). By 48 h, ALT activities of the -Fas-treated mice decreased to control levels. Administration of IDN-1965 via an osmotic minipump (0.36 mg/kg/h) 24 h before -Fas injection significantly blocked liver injury at both the 6 and 24 h time points (Fig. 5). -Fas-induced lethality was also completely blocked by constant infusion, compared with mice receiving vehicle from the osmotic pump (n = 6).

View larger version (25K): [in this window] [in a new window]   Fig. 5.   Seven-day time course of -Fas-induced liver injury: comparison of protection of a single administration or constant infusion with IDN-1965. One set of mice was injected with -Fas (200 µg/kg, i.v.) in saline vehicle and cotreated simultaneously with either vehicle (50% DMSO), or IDN-1965 (1 mg/kg, i.p.) in vehicle. Another set of mice was implanted i.p. with osmotic minipumps delivering either vehicle (50% DMSO) or IDN-1965 (30 mg/ml) at a rate of 0.36 mg/kg/h 24 h before administration of -Fas. Surviving mice were killed from 6 h to 7 days after -Fas administration. Plasma was harvested for ALT activities. Data are expressed as the mean ± S.E.M. n = 3 to 10/group. *Indicates significant difference from IDN-1965-treated control. **Indicates significant difference from IDN-1965 bolus-treated and vehicle-treated controls.

Histologically, this time course study confirmed that -Fas-treated livers contained numerous apoptotic hepatocytes at 6 h, which progressed to massive areas of necrosis by 24 h. By 7 days, fibrotic areas were present with inflammatory infiltrates and numerous mitotic cells, indicating hepatocellular regeneration (Fig. 6A). In contrast, a single dose of 1 mg/kg IDN-1965 blocked apoptosis in most hepatocytes at 6 h, but numerous apoptotic, as well as necrotic, areas were evident by 24 h. By 7 days, livers had areas of fibrosis with mild inflammation and mitotic hepatocytes, but the extent was much less than that of -Fas treatment alone at all time points (data not shown). Lastly, livers were essentially normal in -Fas-treated mice coadministered IDN-1965 via the osmotic pump 7 days after administration, with few apoptotic hepatocytes at the 6 h time point (Fig. 6B). Livers from control mice receiving either IDN-1965 or vehicle via osmotic mini-pumps were unremarkable.

View larger version (130K): [in this window] [in a new window]   Fig. 6.   Morphological evaluation of livers from -Fas-treated mice. One set of mice was injected with -Fas (200 µg/kg, i.v.) in saline vehicle and cotreated simultaneously with either vehicle or IDN-1965 (1 mg/kg, i.p.). Another set of mice was implanted with osmotic minipumps containing either vehicle or IDN-1965 24 h before administration with -Fas antibody. Surviving mice were killed either 6 h to 7 days after -Fas administration. Livers were harvested, stained with H&E and morphologically evaluated. A, representative liver from an -Fas-treated mouse 7 days after -Fas administration. Note mitotic hepatocyte (larger arrow) and region of inflammatory cell accumulation and fibrosis (arrowheads). B, representative liver from an -Fas-treated mouse 7 days after -Fas administration also receiving IDN-1965 via an osmotic minipump. Note normal histology. Original magnification, 40×.

To determine a window of efficacy of IDN-1965 in this model, the caspase inhibitor was administered both several hours before and after -Fas administration. In predosing experiments, IDN-1965 significantly protected the liver from -Fas when administered up to 1 h before -Fas administration (Fig. 7A). Treatment with 3 mg/kg IDN-1965 3 or 5 h before -Fas resulted in no protection of the liver. To determine how effective IDN-1965 is when administered after the -Fas insult, mice were treated with IDN-1965 0 to 4 h after -Fas. As shown in Fig. 7B, IDN-1965 significantly inhibited -Fas-induced liver injury when administered as late as 3 h post--Fas administration. Administration of IDN-1965 3 h post--Fas still reduced liver injury by 61%, but administration 4 h after was not efficacious.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Effect of pre- and postadministration of IDN-1965 on -Fas-induced liver injury. Mice were administered -Fas (200 µg/kg, i.v.) in saline vehicle at t = 0. A, effect of predose: IDN-1965 (3 mg/kg, i.p. in DMSO) was administered 5, 3, 1, or 0 h before -Fas administration. B, effect of postdose: IDN-1965 was administered 0, 1, 2, 3, and 4 h after -Fas treatment. Mice were killed 6 h after -Fas administration, and plasma was harvested for measurement of ALT activities. Data are expressed as percentage of -Fas control. *Significant difference from -Fas control. n = 6 to 8/group.

To determine whether IDN-1965 alters caspase processing, Western blot analysis was used to examine the effect of IDN-1965 on -Fas-induced processing of caspases 8, 3, and 6, 3 h after -Fas administration. The effect of IDN-1965 cotreatment on Bid cleavage was also determined. -Fas administration resulted in the disappearance of pro-caspase 8, which was inhibited by IDN-1965 (Fig. 8A). Additionally, the processed forms of caspases 3 and 6 present after -Fas administration were not detected when IDN-1965 was coadministered with -Fas (Fig. 8, B and C). The activated form of caspase 3 was detected only in livers of -Fas-treated mice, but not in mice cotreated with IDN-1965 (Fig. 8D). Lastly, Bid, a substrate for various caspases, was cleaved in livers of -Fas-treated mice, but not in livers of mice cotreated with IDN-1965.

View larger version (42K): [in this window] [in a new window]   Fig. 8.   Western blot analysis for caspases 3, 6, and 8, as well as Bid 3 h after -Fas administration: effect of IDN-1965. Livers from mice treated with either saline vehicle, -Fas (200 µg/kg, i.v. in saline), IDN-1965 (3 mg/kg, i.p.), or -Fas and IDN-1965 simultaneously were harvested 3 h after -Fas administration and processed for Western blot analysis as described under Experimental Procedures. A, pro-caspase 8. Note marked disappearance of pro-caspase 8 only in the -Fas-treated group. B, caspase 6. Note disappearance of pro-caspase 6 and appearance of caspase long subunit. C, pro-caspase 3. Note appearance of both the short-chain and long subunit in only the -Fas-treated group. D, activated caspase 3. Note detection of the caspase 3 long subunit only in the -Fas-treated group. E, Bid. Note detection of cleaved Bid in only the -Fas-treated group.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

Excess apoptosis, which appears to be a detrimental process in a number of liver diseases, may be effectively treated by inhibition of caspases. Liver diseases where abnormally large amounts of apoptosis occur include Wilson's disease, allograft rejection, and both viral and alcoholic hepatitis (Kerr et al., 1979; Kawahara et al., 1994; Strand et al., 1998; Tannapfel et al., 1999). Moreover, CD95/Fas-mediated apoptosis has been associated with these diseases. For example, increased CD95 expression was seen in hepatocytes from patients with alcoholic liver disease, as well as in livers of patients with acute liver failure diagnosed with Wilson's disease (Galle et al., 1995; Strand et al., 1998; Taieb et al., 1998; Tannapfel et al., 1999). The beneficial effects of caspase inhibitors in models of apoptotic liver injury are now well established, but reports have not fully characterized a caspase inhibitor in these systems. Therefore, the major objective of this study was to more fully evaluate a caspase inhibitor in the -Fas model of liver injury and lethality where apoptosis via the FAS/APO1/CD95 system predominates. Additionally, the effect of this drug on the activation of a number of caspases and Bid within the liver was also investigated.

IDN-1965 is an irreversible caspase inhibitor with potent activity on a number of caspases. For example, the dissociation constant of enzyme inhibitor complex (K_i) for IDN-1965 is less than 0.08 and 0.03 µM for the apical caspases 8 and 9, respectively (Wu and Fritz, 1999). Based on the nature of the caspase cascade, inhibition of these apical caspases may be critical for the apoptotic cascade to be halted. Caspase 8 associates with FADD/MORT1 adaptor and is the first caspase activated in the Fas pathway (Fernandes-Alnemri et al., 1996). Self-activation of caspase 8 leads to the direct activation of caspase 3 and cleavage of Bid (Li et al., 1998). Cleaved Bid induces cytochrome c release from the mitochondria which in turn activates caspase 9 in the presence of apaf-1 (Li et al., 1997). Caspase 9 then can also activate caspase 3 directly and thus the cascade amplifies for destruction of the cell. Because of the potency and effectiveness of IDN-1965 to inhibit the apical caspases, efficacy was predicted in this model when administered simultaneously with -Fas. This was evidenced by ED₅₀ values of approximately 0.1 to 0.3 mg/kg after i.p. and i.v. routes of administration, respectively. Orally administered IDN-1965 was also effective, but at an order of magnitude lower than i.p. administration, possibly due to the lower bioavailability of the drug.

Since IDN-1965 also has a relatively low K_i for caspase 3 (0.96 µM) and caspase 6 (0.04 µM; Wu and Fritz, 1999), the drug may also have the potential to inhibit apoptosis after activation of the apical caspases has occurred. This was apparent, as IDN-1965 inhibited liver injury/lethality when administered as late as 3 h after -Fas administration, well after downstream caspases have been activated and morphological signs of apoptosis were detected in a number of hepatocytes. With respect to caspase 3, its activation has been suggested to be the "point of no return" for cells, as this caspase has a number of vital cellular targets (Kothakota et al., 1997; Liu et al., 1997). Therefore, the potency of an irreversible caspase inhibitor with strong activity toward caspase 3 may be essential for continued potency. While the K_i for IDN-1965 with respect to caspase 3 is higher than for other caspases, the first-order rate constant for covalent binding to caspase 3 is 0.78 min¹, resulting in a high rate of inactivation (Wu and Fritz, 1999). Therefore, the high activity toward caspase 3 may also explain the effectiveness of IDN-1965 in this model, further supporting the hypothesis that broad-spectrum caspase inhibition may be necessary for effective therapy due to the amplifying nature of the caspase cascade. Additionally, recent data suggest that caspase 3 can activate the upstream caspases 8 and 9, as well as the BH3 domain containing protein Bid, suggesting caspase 3 can further amplify the caspase cascade (Woo et al., 1999). Inhibition of caspase 3 activity may be a crucial element, but not the only one, to block further programmed cell death. Therefore, although caspase 3 inhibition is important, caspase 3 (/) mice were still susceptible to -Fas-mediated liver injury and lethality, although the -Fas-mediated effects were delayed (Woo et al., 1999).

Despite its irreversible nature, the effect of IDN-1965 on inhibition of -Fas-induced caspase activity in the liver appears to be short-lived. This is supported by the observation that administration of 1 or 3 mg/kg IDN-1965 significantly inhibited liver caspase 3-like activation 3 h after -Fas administration but by 4.5 h, caspase activation was equivalent to those of -Fas-treated mice (despite near complete inhibition of liver injury at 6 h). A number of possibilities can explain this finding, including a relatively rapid clearance of the drug. Preliminary studies have determined plasma half-life in rats after i.v. administration for IDN-1965 is 40 min (data not shown). The predose study further supports this hypothesis since no efficacy was demonstrated if drug was administered 3 h before -Fas. Transient inhibition may also be due to ongoing expression (de novo synthesis) of one or a number of caspases, ultimately resulting in active caspase, but these studies have not been reported. There may be other protective effects of IDN-1965, but activity of IDN-1965 against other serine/cysteine proteases is low. The K_i of IDN-1965 against cathepsin B, calpain I, and calpain II are 1360, 24.3, and 11.6 µM, respectively (Joe Wu, unpublished data). Regardless of the mechanism(s), caspase activity was initially inhibited by IDN-1965 at lower doses, and either multiple or continuous dosing of a caspase inhibitor may be necessary for optimal long-term efficacy.

The increases in caspase activity without a significant rise in ALT activity at the 6 h time point raised the question whether -Fas-induced injury continues to occur after the 6 h time point, despite treatment with IDN-1965. This was addressed by comparing Fas-mediated liver injury after a single dose of IDN-1965 to continuous administration of the drug. Indeed, although there was little evidence of -Fas-induced liver injury at 6 h with coadministration of a single dose of IDN-1965, ALT activities at 24 h were similar to -Fas-treated controls. Additionally, at this time point, both apoptosis and necrosis were evident in both treatment groups. Therefore, these data suggest that the increase in caspase activity initially inhibited by IDN-1965 will eventually lead to liver apoptosis at later time points. Continuous administration of IDN-1965 via an osmotic minipump (a delivery equaling 0.36 mg/kg/h), however, was extremely effective in reducing liver injury, as evidenced by both ALT activities and histological evaluation throughout the 7-day period. It appears therefore that as long as adequate concentrations of inhibitor are achieved, the caspase cascade and ultimately apoptosis in this model will be inhibited.

Another objective of this study was to determine which caspases in the liver become activated after in vivo -Fas administration and how IDN-1965 affected their activation. In particular, the upstream caspase 8, as well as the downstream caspases 3 and 6 were examined. Consistent with previous reports demonstrating the processing of pro-caspase 8 and pro-caspase 3 after -Fas administration (Yin et al., 1999), the cleaved form of caspase 3 using the CM-1 antibody was also clearly detected. In contrast, processing of caspase 6 was also detected, a finding contrary to Woo et al. (1999). The reason for the discrepancy is not known, but could be due to the use of different primary antibodies and methodology. These studies also indicated that IDN-1965 inhibited the processing of all three caspases. Finally, the effect of caspase inhibitors on the processing of Bid was investigated. Bid is activated by caspase 8 and is an essential component of the caspase cascade, responsible for the release of cytochrome c from the mitochondria (Li et al., 1998). Bid (/) mice are significantly protected from -Fas-induced liver injury/lethality (Yin et al., 1999) and are therefore an essential component in the amplification of the apoptotic cascade. In these studies, Bid cleavage was inhibited by IDN-1965, theoretically due to caspase 8 inhibition, but it may also be due to the drug's ability to inhibit caspase 3 (Woo et al., 1999).

Recent data also suggest that IDN-1965 is not only effective in protecting hepatocytes, but also sinusoidal endothelial cells. Natori et al. showed that IDN-1965 inhibited sinusoidal endothelial cell apoptosis in a model of cold ischemia/warm reperfusion (Natori et al., 1999). Sinusoidal cell apoptosis has been shown to be an early and critical event after liver transplantation (Gao et al., 1998). Indeed, it was also demonstrated that IDN-1965 increased the survival time of rats receiving donor livers (Natori et al., 1999), suggesting that inhibition of apoptosis may prove beneficial in the transplantation process. In conclusion, this study supports the hypothesis that caspase inhibition may protect cells from abnormally large amounts of apoptosis seen in a number of disease states. Continual treatment with a caspase inhibitor may preserve the loss of critical numbers of cells responsible for the development of caspase-mediated cell death.