Cell-death programs executed in the pancreas under pathological conditions remain largely undetermined, although the severity of experimental pancreatitis has been found to depend on the ratio of apoptosis to necrosis. We have defined mechanisms by which apoptosis is induced in pancreatic acinar cells by the oxidant stressor menadione. Real-time monitoring of initiator caspase activity showed that caspase-9 (66% of cells) and caspase-8 (15% of cells) were activated within 30 min of menadione administration, but no activation of caspase-2, -10, or -12 was detected. Interestingly, when caspase-9 activation was inhibited, activation of caspase-8 was increased. Half-maximum activation (t0.5) of caspase-9 occurred within ∼2 min and was identified at or in close proximity to mitochondria, whereas t0.5 for caspase-8 occurred within ∼26 min of menadione application and was distributed homogeneously throughout cells. Caspase-9 but not caspase-8 activation was blocked completely by the calcium chelator BAPTA or bongkrekic acid, an inhibitor of the mitochondrial permeability transition pore. In contrast, caspase-8 but not caspase-9 activation was blocked by the destruction of lysosomes (preincubation with Gly-Phe β-naphthylamide, a cathepsin C substrate), loss of lysosomal acidity (bafilomycin A1), or inhibition of cathepsin L or D. Using pepstatin A-BODIPY FL conjugate, we confirmed translocation of cathepsin D out of lysosomes in response to menadione. We conclude that the oxidative stressor menadione induces two independent apoptotic pathways within pancreatic acinar cells: the classical mitochondrial calcium-dependent pathway that is initiated rapidly in the majority of cells, and a slower, caspase-8-mediated pathway that depends on the lysosomal activities of cathepsins and is used when the caspase-9 pathway is disabled.
- acinar cells
oxidative stress is known to activate cell death by using different execution pathways (45, 48, 49, 51) and is thought to play an important role in the pathogenesis of acute pancreatitis (6). Increased generation of reactive oxygen species is a well-known initiator of apoptosis in many cell types (33), including mouse pancreatic acinar cells (11, 15, 18, 22, 23). The cellular mechanisms of oxidant-induced apoptosis are not fully understood. However, an increase in the cytosolic free calcium concentration (10, 47) coordinated with opening of the mitochondrial permeability transition pore (mPTP) (21, 44) has been shown to play a crucial role in the induction of apoptosis in pancreatic acinar cells by the oxidant menadione (22).
Caspase activation is a crucial early event in the commitment of a cell to undergo programmed cell death (63). Caspases are aspartate-specific cysteine proteases that when activated cleave numerous cellular proteins, leading to disassembly of the cell. The specificity of caspase activity comes from the unique amino acid sequence each caspase recognizes and cleaves (43). Depending on the particular apoptotic pathway triggered, specific caspases can be activated. There are two main caspase-activation cascades described for apoptosis: the intrinsic and extrinsic apoptotic pathways (65). The intrinsic apoptotic pathway results from alterations in mitochondrial structure and function, including mitochondrial membrane depolarization, release of cytochrome c from the intermembrane space of the mitochondria into the cytosol, and activation of the initiator caspase, caspase-9 (19). Caspase-9 then activates the downstream effector caspases-3, -6, and -7 (36). Activated effector caspases are then responsible for cleaving cellular proteins, leading to the characteristic phenotype of apoptosis (chromatin degradation and condensation, plasma membrane blebbing, formation of apoptotic bodies, etc.) (31). The second apoptotic pathway, the extrinsic pathway, results from activation of death-domain receptors (such as TNF-R, IL-R, Fas, and TNF-related apoptosis inducing ligand-R), the formation of a death-inducing signaling complex, and activation of the initiator caspase, caspase-8 (7, 41). Caspase-8 can then activate downstream effector caspases independent of mitochondria, or it can cleave Bid, a proapoptotic protein, which translocates to the mitochondria and causes release of cytochrome c, thus causing activation of the intrinsic apoptotic pathway (35).
Other caspases, including caspase-2, caspase-10, and caspase-12, can also participate in the apoptotic pathways as initiator caspases. Caspase-2 has been shown to trigger the release of cytochrome c and the intrinsic apoptotic pathway after microinjection into cells or in response to various cytotoxic stimuli (32). Similar to caspase-8, caspase-10 can be activated by death-domain receptors and death-inducing signaling complexes (59) and can activate downstream effector caspases (40). Caspase-12 can be activated in response to endoplasmic reticulum (ER)-stress (such as brefeldin A or thapsigargin) and can induce apoptosis (42).
There is growing evidence that cathepsins, which are lysosomal proteases, also play a role in apoptotic signaling pathways (3, 8, 17, 46, 52, 53, 66). There are three types of cathepsins: serine proteases (cathepsin A and G), aspartic proteases (cathepsin D and E), and cysteine proteases (cathepsin B, C, H, L, and S). Of these lysosomal proteases, cathepsin B (8), D (3, 52), G (53), and L (26, 66) have been shown to play a role in apoptosis. It is not clear how cathepsins participate in apoptotic signaling pathways. However, several studies have shown disruption of lysosomal membranes by measuring the change in distribution of acridine orange, a lysomotropic weak base that accumulates in acidic vacuole compartments, in response to such apoptotic stimuli as H2O2 (1, 20) and blue-light irradiation (9). Immunostaining (3, 28) and fluorescence microscopy (14) studies have also shown that cathepsin D can translocate from lysosomes to the cytosol in response to apoptotic stimuli. Although the cytosolic substrates for cathepsins are unknown, a few studies have shown evidence that cytosolic caspases (26, 55) and proapoptotic proteins (60) may be the targets of lysosomal proteases released into the cytosol.
We have shown previously that oxidant stress induced by menadione can induce mPTP opening and can activate caspase-9 in pancreatic acinar cells. However, in some cells apoptosis appears to occur by a mechanism not dependent on mPTP (22). We hypothesized that other caspases apart from caspase-9 could be activated via a mechanism that is not dependent on induction of the mPTP. We show here that in contrast to the caspase-9 activation mechanism, caspase-8 is activated by a Ca2+-independent and mPTP-independent mechanism that uses lysosomal cathepsins.
Male CD1 mice were killed by cervical dislocation [in accordance with the Animal (Scientific Procedure) Act, 1986], and pancreas was excised. Single cells or small clusters were isolated as previously described (62). Briefly, pancreas was injected with 200 U/ml collagenase solution (Collagenase CLSPA; Worthington Biochemical, Lakewood, NJ) and was incubated at 37°C for 15 min. Pancreas was then agitated by pipette to obtain single cells and small cell clusters. The isolated cells were washed by centrifugation in a standard buffer solution (in mM: 140 NaCl, 1.13 MgCl2, 1 CaCl2, 4.7 KCl, 10 glucose, and 10 HEPES, pH 7.2). All experiments were performed at room temperature (23–25°C), and cells were used within 3–4 h of isolation.
Isolated pancreatic acinar cells were washed and suspended in calcium-free buffer solution (140 mM NaCl, 1.13 mM MgCl2, 4.7 mM KCl, 10 mM glucose, 0.1 M EDTA, and 10 mM HEPES, pH 7.2). Cells were then loaded with fluorescent indicator-linked substrates for activated caspase-2 (10 μM Z-VDAD-R110; Molecular Probes, Eugene, OR), caspase-8 (10 μM Z-IETD-R110; Molecular Probes), caspase-9 (10 μM Z-LEHD-R110; Molecular Probes), or general caspases (10 μM R110-aspartic acid amide; Molecular Probes) at room temperature for 20 min or for caspase-10 (50 μM AEVD-AFC; BioVision, Mountain View, CA) or caspase-12 (50 μM ATAD-AFC; BioVision) at 37°C for 1 h. Caspase substrates (except for the general substrate) used in this study were specific for the relevant initiator caspases, as reported. To avoid activation of substrates by executioner caspases, all experiments were strictly limited to the first 30 min after induction of apoptosis by menadione. After loading, cells were washed and resuspended in calcium-free buffer solution. The isolated cells were placed on a Leica SP2 confocal microscope stage, and fluorescence was imaged over time (excitation 488 nm, emission 505–543 nm for caspase-2, -8, -9, or general caspase substrates; excitation 405 nm, emission 475–600 nm for caspase-10 and -12 substrates). Cells were then treated with 30 μM menadione. To examine colocalization of caspase activation and the position of mitochondria, cells were also loaded with MitoTracker Deep Red 633 (50 nM, excitation 633 nm, emission <650 nm; Molecular Probes) or tetramethyl rhodamine methyl ester (100 nM, excitation 543 nm, emission >600 nm) at 37°C for 15 or 20 min, respectively. Cells were then washed and resuspended in standard buffer solution.
To measure changes in ATP levels as described previously (34), we loaded isolated pancreatic acinar cells with 4 μM Mg Green (Molecular Probes) at room temperature for 30 min. Cells were washed and resuspended in buffer solution. Cells were treated with 50 μM bongkrekic acid for 45 min before cell suspensions were placed onto the confocal microscope stage for imaging (excitation 488 nm, emission >505 nm) before and after treatment with 10 μM acetylcholine.
Redistribution of cathepsin D.
To measure movement of cathepsin D, cells were loaded with a cathepsin D inhibitor linked to a fluorescent probe [pepstatin A-BODIPY FL conjugate, 1 μM; Invitrogen, Paisley, UK; binds cathepsin D at acidic pH (14) at 37°C for 30 min]. Cells were washed and resuspended in standard buffer solution and were incubated at 37°C for another 30 min. After incubation, cells were washed and resuspended in calcium-free buffer solution, and fluorescence was imaged on a confocal microscope (excitation 476 nm, emission 500–600 nm).
The chemicals used in this study were menadione (30 μM, Sigma), BAPTA-AM (25 μM, Molecular Probes), bongkrekic acid (50 μM, Calbiochem, La Jolla, CA), Gly-Phe β-naphthylamide (GPN, 50 μM, Sigma), bafilomycin A1 (100 nM, Sigma), cathepsin B inhibitor (CA074-Me, 50 μM, Sigma), cathepsin G inhibitor 1 (10 μM, Calbiochem), cathepsin L inhibitor (10 μM Z-FF-FMK, Calbiochem, and NapSul-Ile-Trp-CHO, Biomol), and cathepsin D inhibitor (10 μM, pepstatin A, Sigma). Chemicals, where necessary, were dissolved in ethanol or DMSO (concentration <0.1%).
Data are presented as means ± SE of whole cell fluorescence and percentages of cell populations positive for caspase activation. Cells were considered positive for caspase activity when fluorescence of a treated cell was higher than the average fluorescence of control cells plus three standard deviations. One-way ANOVA was used for statistical comparison between control and treatment groups. P < 0.05 was considered significant.
Caspase activation in response to oxidative stress.
To determine which initiator caspases are activated in response to oxidative stress in the pancreatic acinar cell, we loaded cells with fluorescent substrates for caspase-2, -8, -9, -10, and -12 (Fig. 1). After 30 min of treatment with menadione, fluorescence did not change in cells loaded with substrates for caspase-10 (Fig. 1D) and caspase-12 (Fig. 1E), and only 1 out of 92 cells was positive for caspase-2 activity (P > 0.35; Fig. 1A). In contrast, a high proportion of cells (66 ± 12%) was positive for activation of caspase-9 (significant difference with control P < 0.006; Fig. 1C), in agreement with data reported previously (22). The level of activation of caspase-9 was similar to the measurements obtained with the general caspase substrate (49 ± 5%; Fig. 1F). These results suggest that caspase-2, -10, and -12 are not activated in the acute response to menadione-induced oxidative stress in these cells, as opposed to a high activation of caspase-9, the classical intrinsic apoptosis pathway-initiator caspase.
Marked caspase-8 activation was demonstrated in a minority of isolated pancreatic acinar cells by using a fluorescent probe-linked caspase-8 substrate (Z-IETD-R110). Menadione (30 μM) induced a significant increase (P < 0.00018) in the fluorescence of 15 ± 2% of acinar cells (Fig. 1B). Apoptosis-positive cells displayed a high increase in fluorescence (∼10.9 ± 0.8-fold) compared with control cells (Fig. 2A). Fluorescence of the caspase-8 substrate increased within 30 min after application of menadione (Fig. 2B). These data show that caspase-8 was activated in response to the menadione-induced oxidative stress. Confocal-microscopy images (Fig. 2C) show homogeneous distribution of fluorescence throughout the cell, suggesting predominantly cytosolic localization of activated caspase-8.
With the help of fluorescent probes for caspase activation, we were able to examine the time course and localization of caspases activated in response to oxidative stress. The results for caspase-9 activity showed that 30 μM menadione induced a significant increase (∼4.1 ± 0.1-fold) in fluorescence (Fig. 3A) in apoptosis-positive cells. Figure 3B shows that fluorescence of the caspase-9 substrate develops very quickly, within 2 min of the application of menadione. Confocal-microscopy images (Fig. 3C) show fluorescence from caspase-9 activation localized at or near the mitochondria. To determine this relationship more precisely, we loaded pancreatic acinar cells with a mitochondrial indicator (MitoTracker Deep Red 633) simultaneously with the caspase-9 fluorescent substrate. After treatment with 30 μM menadione, caspase-9 substrate fluorescence was very closely localized to the fluorescence of the MitoTracker Deep Red (Fig. 3C). These data suggest that caspase-9 is activated very close to or at the surface of mitochondria. Interestingly, near-mitochondrial distribution of caspase-9 was detected shortly after induction of apoptosis, at 2–15 min after application of menadione, and then gradually became more diffused, resembling distribution of caspase-8 (Fig. 2C).
To further determine the timing and the order of caspase activation, we loaded pancreatic acinar cells with a fluorescent probe-linked general caspase substrate to determine whether or not another initiator caspase is activated before caspases-9 and -8. Fluorescence of the general caspase substrate increased significantly (∼2.3 ± 0.1-fold) in apoptosis-positive acinar cells (Fig. 4A), an increase that occurred quickly, within 2 min (Fig. 4B) of administration of 30 μM menadione. General caspase substrate fluorescence (Fig. 4C) was close to mitochondria at first, as with caspase-9 (Fig. 3C), but at a later time it was also seen throughout cells, with a distribution similar to that of caspase-8 activation (Fig. 2C).
To compare the time-to-activation of the different caspases, we have calculated the length of time to reach half-maximum activation (t0.5) for general caspase, caspase-8, and caspase-9. Figure 4D shows that caspase-9 (t0.5 = 129 ± 43 s; n = 12) is activated with a similar time course to that of the first active caspase detected by the general caspase substrate measurements (t0.5 = 104 ± 48 s; n = 19). There was no significant difference between t0.5 for caspase-9 substrate fluorescence and the general caspase substrate fluorescence (P > 0.15), which is consistent with caspase-9 being the first caspase activated in response to menadione-induced oxidative stress. Caspase-8 is activated at a significantly later time (t0.5 = 1,560 ± 56 s; n = 8; P < 0.000001), suggesting that either caspase-8 activation occurs as a result of caspase-9 activation or is activated independently of caspase-9 at a later time point.
Calcium dependence of caspase-8 and -9.
Previously published data (22) have shown that activation of caspase-9 is completely inhibited by the calcium chelator BAPTA. To compare the calcium dependence of menadione-induced caspase-8 and -9 activation, we preloaded cells with an intracellular calcium chelator, BAPTA in AM form (25 μM). Figure 5A shows that the absence of any alteration in the cytosolic free calcium concentration did not change the profile of menadione-induced caspase-8 activation (9 ± 3 and 8.7 ± 5% of caspase-8-positive cells in the absence and presence of BAPTA-AM, respectively; P > 0.96). These data show that caspase-8 activation in response to oxidative stress from menadione is independent of changes in cytosolic free calcium concentration. However, chelating cytosolic calcium with BAPTA did prevent any increase in fluorescence of the caspase-9 substrate within all cells tested in response to 30 μM menadione compared with 47 ± 10% caspase-9-positive cells in the absence of BAPTA (Fig. 5B), indicating that caspase-9 activation by menadione is calcium dependent (22). Control experiments with 10 μM of heavy metal chelator N,N,N,N-tetrakis(2-pyridylmethyl)ethylenediamine (TPEN; Fig. 5C) show no significant difference in caspase-9 activation (55 ± 9% positive cells against 46 ± 3%; n = 26–44/group; P > 0.46).
Caspase-8 is activated independently of caspase-9.
As mentioned, previous work has shown that a small population of pancreatic acinar cells could still undergo menadione-induced apoptosis in the presence of bongkrekic acid, considered to be an inhibitor of the opening of the mPTP (22, 39). Since caspase-8 was activated in a small proportion of pancreatic acinar cells in response to menadione, we have tested here whether or not caspase-8 activation could be blocked by bongkrekic acid. Figure 5D shows that after treatment with menadione, cells pretreated with 50 μM bongkrekic acid showed an activation profile (13 ± 9% caspase-8-positive) similar to cells not treated with bongkrekic acid (11 ± 4% caspase-8-positive; P > 0.8), suggesting that caspase-8 is activated independently of mPTP opening, which is, however, required for caspase-9 activation. Thus caspase-9 activation was virtually abolished by bongkrekic acid (2 ± 2% of caspase-9-positive cells compared with 52 ± 2% of caspase-9-positive cells without inhibitor; P < 0.004; Fig. 5E). Using a Mg Green protocol described by Leyssens et al. (34), we found that pancreatic acinar cells maintained their ATP levels for at least 45 min following exposure to bongkrekic acid (i.e., more than the length of all caspase experiments; data not shown). These controls confirm the validity of our experiments with bongkrekic acid, the only known effective mPTP inhibitor in pancreatic acinar cells (22).
Inhibition of caspase-8 and -9.
We conducted control experiments with inhibitors of caspase-8 and -9. As expected, an inhibitor of caspase-8 completely abolished activation of caspase-8 by menadione (Fig. 6A) and activation of caspase-9 was abolished by an inhibitor of caspase-9 (Fig. 6B). We have also tested the dependence of caspase-8 activation on the activity of caspase-9 and caspase-3 (37) by pretreating cells with inhibitors of caspase-9 and caspase-3. Neither a caspase-9 inhibitor (Fig. 6C) nor a caspase-3 inhibitor (P > 0.8; data not shown) reduced caspase-8 activation in pancreatic acinar cells. Thus caspase-8 activation by menadione is independent of the activity of caspase-9 and caspase-3, as well as of the mPTP opening. Interestingly, application of the caspase-9 inhibitor potentiated activation of caspase-8 (Fig. 6C), suggesting that caspase-8 may serve as a backup apoptosis-induction pathway when the classical intrinsic mechanism is blocked. In reverse, however, inhibition of caspase-8 did not change the activity of caspase-9 (not shown). Application of both inhibitors virtually abolished activation of caspases (as measured with the general caspase substrate; Fig. 6D).
Lysosomal role in caspase activation.
Since caspase-8 activation was found to be independent of cytosolic calcium changes, mPTP opening, or caspase-9 activation, we investigated alternative mechanisms of caspase-8 activation, notably a possible role for lysosomes. We disrupted lysosomes by pretreating cells with GPN, a substrate for cathepsin C that accumulates inside lysosomes and causes their collapse (27). Caspase-8 activation was significantly inhibited in response to menadione when cells were pretreated with GPN (1 ± 1% of caspase-8 positive cells compared with 11 ± 2% in the absence of GPN; P < 0.007; Fig. 7A). We also tested the effect of GPN on caspase activation and found that caspase-9 activity in response to menadione did not change when cells were pretreated with GPN (57 ± 10 and 50 ± 8% positive for caspase-9 activity in absence and presence of GPN, respectively; P > 0.43; Fig. 7B). To confirm that inhibition of caspase-8 activation was due to lysosomal dysfunction, we pretreated cells with bafilomycin A1, an inhibitor of the vacuolar H+-ATPase, which disrupts lysosomal acidity and thus lysosomal function. Figure 7, C and D shows that, similar to GPN, bafilomycin A1 blocks caspase-8 activation (only 1 ± 1% positive compared with 15 ± 5% of caspase-8-positive cells in the absence of bafilomycin A1; P < 0.03). In contrast, caspase-9 activity did not change significantly (38 ± 13% of caspase-9-positive cells compared with 50 ± 14% in the absence of bafilomycin A1; P > 0.4; Fig. 7D). Bafilomycin A1 alone in control experiments did not induce activation of either caspase-8 or -9 (P < 0.02 and P < 0.007, respectively, compared with menadione-treated cells) after 30 min of incubation. These results indicate that caspase-8 activation in response to menadione-induced oxidative stress depends on functional lysosomes. Caspase-9 activation, however, is independent of lysosomal function.
Cathepsin dependence of caspase-8 activation.
Cathepsins, important lysosomal proteases, have been shown to contribute to apoptosis (3, 52, 53, 66). Because we had found that lysosomal disruption prevents caspase-8 activation, we investigated the role of cathepsins in menadione-induced caspase-8 activation. Since cathepsins B, D, L, and G have previously been shown to participate in apoptosis, we pretreated cells with inhibitors of these cathepsins and examined caspase-8 activation in response to menadione. Caspase-8 was still activated in the presence of a cathepsin B inhibitor (8 ± 1 and 6 ± 2% positive in the absence and presence of a cathepsin B inhibitor, respectively; P > 0.28; Fig. 8A) and a cathepsin G inhibitor (21 ± 11 and 25 ± 15% positive cells in the absence and presence of the cathepsin G inhibitor, respectively; P > 0.8; Fig. 8C). Inhibition of cathepsin D completely abolished activation of caspase-8 (0% compared with 8 ± 3% in the absence of inhibitor; P < 0.02; Fig. 8B). Cathepsin L inhibition also virtually abolished menadione-induced caspase-8 activation (1 ± 1% positive cells for cathepsin L inhibitor Z-FF-FMK and 0% positive cells for NapSul-Ile-Trp-CHO; P < 0.03 and P < 0.014, respectively; Fig. 8, D and E).
Monitoring cathepsin D in the lysosomes.
Because caspase-8 activation is cytosolic and cathepsin D and L are lysosomal, we investigated whether cathepsin localization changed in cells treated with menadione. We loaded cells with a fluorescent probe specifically targeting cathepsin D in the lysosomes (pepstatin A-BODIPY FL conjugate) and measured fluorescence of the probe within the lysosomes before and after treatment with menadione. Note that this cathepsin D substrate only shows fluorescence when bound to cathepsin D in the lysosomes at acidic pH and hence does not report cathepsin D activity within the cytosol (14). Figure 9A shows a significant decrease in fluorescence (28 ± 1% compared with 4 ± 1% in control; P < 0.0001) after administration of menadione, suggesting substantial loss of cathepsin D from the lysosomes in response to menadione. Although the fluorescence significantly decreased within lysosomes, loss was only partial (Fig. 9B), suggesting that lysosomes retain some cathepsin D.
Here we have performed a comprehensive study of the activation of initiator caspases that induce apoptosis in pancreatic acinar cells exposed to oxidative stress and demonstrated a significant role for the extrinsic pathway. In addition to caspase-9, the classical intrinsic apoptotic pathway initiator activated in the majority of cells, only caspase-8, the classical initiator of the extrinsic pathway, was activated in ∼15% of cells in response to oxidative stress. Other caspases that can initiate either intrinsic or extrinsic pathways, namely caspase-2, -10, and -12, were not activated by menadione. Caspase-8 activation was shown to occur throughout the cytosol within ∼26 min. Caspase-8 activation has not been observed (25) or examined so far in studies with menadione but has been seen in various cell types in response to other forms of oxidant stress (11, 29, 38, 64), including a rat model of cerulein-induced pancreatitis (37).
Using fluorescent probe-linked substrates, we were able to examine caspase activation in real time and to monitor the distribution of activated caspases. Our data, using the fluorescent substrate for caspase-8, show a predominantly cytoplasmic localization of substrate, in contrast to caspase-9, which was found at or close to mitochondria. Others have previously reported a cytosolic distribution of activated caspase-8, using green fluorescent protein-tagged caspases (57) and antibodies in cell-fractionation studies (58). However, Chandra et al. (12) observed active caspase-8 associated with the mitochondrial membrane during apoptosis induced by etoposide in breast cancer cells.
We found half-maximal activation of caspase-9 within 2 min of treatment with menadione. Early activation of the general caspase substrate showed the same timing and distribution, suggesting that caspase-9 is likely to be the first caspase activated by oxidant stress induced by menadione. In the pancreatic acinar cell, oxidant stress-induced caspase-9 activation was localized to the mitochondria. Susin et al. (61) have found procaspase-9 localized within mitochondria of T cells, whereas active caspase-9 was then released from mitochondria on opening of the mPTP by atractyloside. Chandra and Tang (13) also showed localization of active caspase-9 in the mitochondria. Others (57, 58), however, have been unable to confirm any mitochondrial localization of caspase-9 and found a cytosolic distribution of this caspase. We have approached these questions comprehensively, comparing the time course and distribution of fluorescence from caspase-9, caspase-8, and general caspase substrates, only detecting activated caspases, while using the different substrates as controls for each other. We have shown that activated caspase-9 is localized very close to mitochondria, whereas activated caspase-8 is predominately cytoplasmic. The distribution of general caspase substrate fluorescence was time dependent; at first it was similar to the early (∼2 min) caspase-9 distribution, but at a later time it became more homogeneous, likely reflecting activation of caspase-8, the half-maximum of which occurred at ∼26 min.
The activation of caspase-8 was found to be calcium independent, as contrasted with the completely calcium-dependent activation of caspase-9 by menadione. These results, together with the observation that caspase-8 can still be activated when the mitochondrial intrinsic apoptotic pathway is blocked (with bongkrekic acid or a caspase-9 inhibitor), indicates that caspase-8 is activated independently of the intrinsic apoptotic pathway via caspase-9. A recent study by Sharma et al. (56) indicated that caspase-9, but not caspase-8, is activated in a photoreceptor-derived cell line in response to increased intracellular calcium concentrations induced by A-23187, suggesting that the differences in calcium dependence of these caspases (and independent activation of each) are not limited to the pancreatic acinar cell.
The requirement for functional lysosomes for the activation of caspase-8 but not caspase-9 further separates the activation of the intrinsic and extrinsic apoptotic pathways in pancreatic acinar cells in response to oxidant stress. The role that lysosomes play in apoptotic signaling is not fully understood. However, several studies have shown that reactive oxygen species can lead to an increase in the permeability of lysosomal membranes, as seen by redistribution of acridine orange from lysosomes only to lysosomes and cytosol (1, 20). In our study, we show that cathepsins exit lysosomes to activate caspases on treatment with menadione, but cathepsin B, a lysosomal enzyme that contributes to the pathogenesis of pancreatitis (23), does not contribute to caspase activation in response to menadione, consistent with findings made following hyperstimulation of cathepsin B knockout mice (4, 24). Other studies have also suggested a contribution from lysosomal cathepsins to apoptosis (3, 14, 30). An elegant study by Roberg et al. (52) showed that caspase activation is induced by microinjection of active cathepsin D into the cytosol of human foreskin fibroblasts in a pH-independent manner. Another interesting recent study by Beaujouin et al. (3) showed that overexpression of active or mutated inactive cathepsin D both enhanced etoposide-induced apoptosis in a cancer cell line. These studies suggest that either there is a factor in the cytosol that may allow activity of cathepsins within a more neutral pH or that another property of cathepsins other than their catalytic activity may play a role in apoptotic signaling pathways. An alternative hypothesis is that procaspases or other potential substrates of cathepsins may reside very close to lysosomes, where microdomains of lower pH may be located during leakage of lysosomal content. Leaked cathepsins in the vicinity of lysosomes are more likely to cleave cytosol-located procaspase-8 but are unlikely to affect procaspase-9 within mitochondria. Microdomains of cathepsin activity might also explain a lower probability of activation of caspase-8 and the subsequently low percentage (∼15%) of apoptosis induced by activation of lysosomal intrusion into the second step of the extrinsic apoptotic pathway. Potentiation of caspase-8 and subsequently the extrinsic apoptotic pathway in situations where the intrinsic mechanism was inhibited is another interesting component of our findings. Such a backup system could be very useful in the pancreas, where necrotic cell death is extremely dangerous (5, 48). Activation of both caspases-8 and -9 and their cross-talk has been reported previously (2, 37, 50), possibly through inhibitor of apoptosis proteins; however, the precise mechanism remains unclear. Our results are in agreement with the notion that oxidative damage can occur in lysosomes and microsomes, as well as mitochondria (16, 54).
In summary, as shown in Fig. 10, both of the principal apoptosis pathways are activated independently by oxidative stress in the pancreatic acinar cell. The caspase-9-mediated classical intrinsic pathway is fast and calcium (as well as mitochondria) -dependent. In contrast, the caspase-8-mediated pathway observed in a smaller proportion of cells is slower, is calcium- and mitochondria-independent, and appears to depend entirely on the lysosomal activity of cathepsins.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society