The regulatory mechanisms of nontransformed intestinal epithelial cell apoptosis have not been thoroughly investigated. We determined the susceptibility and mechanism of Fas-mediated apoptosis in nontransformed human intestinal epithelial cells (HIPEC) in the presence and absence of inflammatory cytokines. Despite ample expression of Fas, HIPEC were relatively insensitive to Fas-mediated apoptosis in that agonist anti-Fas antibody (CH11) induced a <25% increase in HIPEC apoptosis. Pretreatment of HIPEC with interferon (IFN)-γ, but not tumor necrosis factor-α or granulocyte-macrophage colony-stimulating factor, significantly increased CH11-induced apoptosis of these cells without increasing Fas expression. Increased apoptosis correlated with increased caspase 3 activation but not expression of procaspase 3. Also, there was a significant delay in the onset of Fas-mediated apoptosis in HIPEC, which correlated with the generation of an activated caspase 3 p22/20 subunit. HIPEC required both initiator caspases 8 and 9 activity but expressed significantly less of the zymogen form of these caspases than did control cells. IFN-γ-mediated sensitization of HIPEC occurred upstream of caspase 9 activation and correlated with a small increase in procaspase 8 expression (<1-fold increase) and a significant increase in expression of an intermediate form (p35) of caspase 4 (3.3-fold increase).
- intestinal epithelium
- interferon-γ regulation of apoptosis
the epithelial cellslining the intestine (IEC) are the first line of defense against contamination by bacteria, viruses, and food antigens present in the lumen. Consequently, a regulated process of epithelial cell turnover is critical in maintaining mucosal integrity and avoiding inflammation of the intestinal epithelium. Normally, proliferating progenitor crypt cells differentiate, migrate, and replace the outgoing surface epithelium every 3–5 days. This balanced process of cell renewal in the intestine is believed to be maintained through apoptosis (programmed cell death). In fact, apoptotic IEC have been detected on the tips of the villi of the small intestine, the luminal surface of the colon, and within the crypts at the level of the progenitor cells (17, 40, 48). Altered regulation of either IEC proliferation or apoptosis can lead to crypt hyperplasia, villous atrophy, and disruption of IEC barrier function. In some immune-mediated intestinal disorders, such as ulcerative colitis and graft-versus-host disease, the number of apoptotic IEC may be greatly increased (16, 22, 25, 26), suggesting that apoptosis may contribute to the disruption of intestinal function and the pathophysiology of these diseases. However, the effects of inflammatory cytokines or intestinally derived growth factors on IEC apoptosis, and regulatory mechanisms involved in IEC apoptosis, are not clearly understood.
It has recently been proposed that the death receptor Fas (CD95, APO-I) may play a role in IEC apoptosis. Fas is constitutively expressed on the basolateral membranes of IEC in normal intestinal mucosa (27), and we have previously reported that freshly isolated IEC express Fas on their surface (36). IEC in organ cultures or isolated intestinal crypt epithelial cells have been shown to be sensitive to Fas-mediated apoptosis (42,49). The potential involvement of Fas in programmed epithelial cell death in the intestine is further suggested by the finding that apoptotic, Fas-expressing IEC colocalize with Fas ligand (Fas-L)-expressing mononuclear cells in the intestinal tissue of patients with ulcerative colitis (49, 55).
The pathways for Fas-mediated apoptosis have been extensively studied. Stimulation through Fas initiates the sequential activation of a series of cysteine proteases, called caspases, the effectors of apoptotic cell death (24, 31, 52). Cross-linking of Fas by Fas-L or agonist antibody leads to the formation of a death-inducing signaling complex (DISC) that contains aggregated Fas, the Fas adaptor protein Fas-associated death domain (FADD), and the proenzyme form of caspase 8 (5, 19, 57). Recruitment of procaspase 8 to the DISC results in its autoproteolysis and activation of this initial caspase in the apoptotic cascade. Events downstream of caspase 8 activation include activation of effector caspases 3, 6, and 7 as well as mitochondrial release of cytochrome c(20, 29, 47). The release of cytochrome cfrom mitochondria can induce the activation of an alternative branch of the caspase cascade through the activation of caspase 9. Cytochrome c binds and activates the adaptor protein apoptotic protease activating factor 1 (Apaf-1) such that it can recruit and activate procaspase 9 (21, 61). Caspase 9 then cleaves and activates downstream effector caspases, amplifying the cascade (46). Although both pathways of Fas-mediated caspase activation are functional in most cell types, in some, the mitochondrial events are not required for efficient Fas-mediated apoptosis, but in others, apoptosis is dependent on the release of cytochrome c (44).
Although enormous progress has been made in identifying effectors and regulators of Fas-mediated apoptosis in various cell systems, our understanding of how these processes are regulated in maintaining epithelial integrity (under normal physiological conditions) or in causing epithelial injury (in diseased states) is still incomplete. Also, the caspases required for Fas-mediated apoptosis of IEC have so far not been completely defined. Studies using colon carcinoma and adenocarcinoma cell lines have allowed some insight into the role of Fas in IEC renewal (1, 2, 34). However, these studies do not always represent the physiological scenario since immortalized cell lines frequently display abnormalities in apoptosis. Also, the findings for one malignant epithelial cell line are frequently not reproducible for another malignant/transformed cell line (35). These abnormalities in apoptosis may contribute to the immortality of transformed cell lines in vitro as well as contribute to their resistance to tumor-specific immune responses, including apoptosis through the Fas death receptor, in vivo. In fact, it has been shown that most colon carcinomas are resistant to Fas-mediated apoptosis, and the mechanisms by which these lines have acquired resistance are numerous (reviewed in Ref. 32). It has also been shown that this resistance can be overcome by exposing cultures to inflammatory cytokines such as interferon (IFN)-γ (1, 2, 33, 34). Multiple mechanisms of cytokine-mediated Fas sensitization have been identified using transformed cell lines (32). These mechanisms of cytokine-mediated sensitization include regulation of expression of Fas (1, 2, 10, 54, 58), Fas-L (6, 28), procaspases (7, 34), and Bcl-2 family proteins (33,34). Cytokines might also regulate the expression of proteins involved in DISC formation, as well as inhibitors of caspases such as the inhibitor of FADD-like interleukin-1β-converting enzyme (I-FLICE) (33, 34). Recently, Ruemmele et al. (42) have demonstrated that some of these mechanisms of cytokine-mediated sensitization may take place in nontransformed IEC cultures as well.
Alternative in vitro models for the study of Fas-mediated apoptosis of IEC include primary cultures of freshly isolated IEC, intact colonic crypts, or organ culture. Isolation of IEC requires the disruption of tissue integrity, which can result in stress-induced forms of apoptosis (13, 14), making it difficult to investigate the mechanism of apoptosis induced by alternative stimuli such as through death receptor ligation. Although it has been shown that IEC in organ cultures and intestinal crypts are sensitive to Fas-mediated apoptosis (42, 49), the IEC-specific regulatory effects of cytokines are difficult to determine using multicellular specimens. Also, isolated IEC do not survive in culture long enough to determine the effects of cytokine on these cells.
We have recently developed a protocol for the isolation and maintenance of nontransformed human intestinal primary epithelial cells (HIPEC) in long-term culture and have established a number of HIPEC lines from small and large intestine (37). Using this cellular model, we determined the susceptibility of HIPEC to anti-Fas agonist antibody (CH11), identified the caspases required for Fas-mediated apoptosis of HIPEC, and systematically determined whether the mechanisms of cytokine-mediated sensitization for Fas-mediated apoptosis previously reported for transformed cell lines were involved in regulation of HIPEC susceptibility.
MATERIALS AND METHODS
Human IEC were isolated from surgical specimens, and HIPEC were derived and maintained as previously described (37). Briefly, isolated crypt cells were cultured in mucosal tissue-derived growth factor containing F-12 medium supplemented with epidermal growth factor, insulin, transferrin, retinoic acid, and hydrocortisone (HIPEC medium). Cells were grown in serum-free HIPEC medium for at least four passages and then in HIPEC medium supplemented with 2% dialyzed fetal calf serum (Summit Biotech, Atlanta, GA). Jurkat cells, LS180, and HT-29 were maintained in RPMI (GIBCO BRL, Grand Island, NY) with 10% fetal calf serum (Sigma, St. Louis, MO), supplemented with glutamine (2 mM), penicillin (50 U/ml), and streptomycin (50 μg/ml). Adherent cells were harvested using 0.25% trypsin containing 1 mM EDTA.
Antibodies and reagents.
Monoclonal anti-Fas (DX2, IgG1), isotype control murine IgG1, polyclonal anticaspase 3 (procaspase and large subunits), monoclonal anticaspase 8 (procaspase), polyclonal (procaspase and large subunit) anti-caspase 9, monoclonal anticaspase 1 (proenzyme and intermediate fragment), monoclonal anti-FADD, and polyclonal anti-I-FLICE were purchased from BD Pharmingen (San Diego, CA). Polyclonal anti-caspase 8 (procaspase and large subunit), antibodies to the inhibitor of apoptosis proteins c-IAP1 and c-IAP2, Bcl-2, Bclx, Bak, Bax, caspase 4 (proenzyme and p20 fragment), and horseradish peroxidase-labeled anti-goat antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA) and anti-XIAP (human IAP-like protein) was purchased from BD Transduction Laboratories (San Diego, CA). Agonist anti-Fas antibody CH11 was purchased from Upstate Biotechnology (Lake Placid, NY). Horseradish peroxidase-labeled anti-mouse and anti-rabbit antibodies were from Amersham (Arlington Heights, IL). FITC-labeled anti-mouse IgG was purchased from Jackson Immunoresearch Laboratories (West Grove, PA). Anti-human α-actin antibody and murine ascites IgM were purchased from Sigma. Additional reagents, unless otherwise stated, were from Sigma.
Cultures were harvested, washed in PBS containing 0.1% bovine serum albumin and 0.1% sodium azide (PBSA), and then incubated with either DX2 anti-Fas monoclonal antibody or isotype control murine IgG1 for 30 min on ice. Cells were washed twice in PBSA and incubated with FITC-labeled secondary antibody for another 30 min on ice. Cells were washed three times in PBSA and then resuspended in PBS or PBS with 2% paraformaldehyde and analyzed on a FacScan (Becton Dickinson, San Jose, CA).
Induction and detection of apoptosis.
Jurkat cells (0.5 × 106/ml), and HIPEC, LS180, and HT-29 monolayers were grown in 12-well tissue culture plates and incubated with medium alone or medium containing either recombinant human IFN-γ (200 U/ml), human tumor necrosis factor (TNF)-α (10 ng/ml), or human granulocyte-macrophage colony-stimulating factor (GM-CSF; 20 ng/ml) (R&D Systems, Minneapolis, MN) for 18–20 h before the addition of CH11 agonist anti-Fas antibody. CH11 was used at a concentration of 100 ng/ml, approximately three times the concentration required for maximum apoptosis of Jurkat after 20 h incubation. Cytokine-primed and unprimed HIPEC were incubated with CH11 for 18–24 h before both adherent and nonadherent cells were harvested for assays to measure apoptosis or for preparation of lysates. Apoptosis was measured by staining with FITC-labeled annexin V (Pharmingen) following the manufacturer's directions. Alternatively, cells were stained with FITC-labeled dUTP in terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) assays, following the manufacturer's instructions for the Death Detection kit from Boehringer Mannheim (Mannheim, Germany). Labeled cells were quantitated by flow cytometry.
Western blot analysis.
Cell pellets were lysed in 10 mM NaH2PO4, 150 mM NaCl buffer containing 1% SDS, 1% Triton X-100, 1% sodium deoxycholate, 2 mM EDTA, and protease inhibitors (2 mM phenylmethylsulfonyl fluoride, 10 mM benzamidine, and 10 mg/ml trypsin inhibitor, leupeptin, antipain, and aprotinin). Total protein was determined by the BCL method (Pierce Chemical, Rockford, IL). Protein (20–50 μg) was run on 12% SDS-PAGE gels (Bio-Rad, Hercules, CA) and then transferred to Trans-Blot nitrocellulose membranes (Bio-Rad). Membranes were blocked with 5% milk in Tris-buffered saline (TBS) before incubation with antibodies overnight at 4°C. After being washed (TBS with 0.5% Tween), blots were incubated with horseradish peroxidase-labeled secondary antibody, washed, and developed by the enhanced chemiluminescence method (Amersham). Autoradiographs were analyzed by densitometry. Blots were then stripped and reprobed with murine anti-human α-actin, against which all immunoblot data were normalized.
Peptide inhibition of caspase activity.
Caspase activity was irreversibly inhibited with fluoromethylketone (fmk)-derived tetrapeptides containing benzyloxycarbonyl (z) groups to enhance cell permeability (R&D Systems). Untreated and IFN-γ-treated HIPEC or Jurkat cells (control) were incubated with peptide inhibitors for 15 min before the addition of CH11 antibody. After 20-h incubation, cells were harvested and stained with FITC-labeled annexin V. Inhibitory peptides were dissolved in DMSO (final dilution of DMSO was >1:400). Therefore, cultures treated with vehicle alone were run as controls in each experiment. Background apoptosis was defined as the percentage of annexin V-positive cells in cultures that did not contain CH11 antibody or inhibitors. Percent inhibition of apoptosis was calculated using the formula
Caspase activity assay.
Colon HIPEC were grown in 100-mm tissue culture dishes, and confluent monolayers were stimulated with IFN-γ, CH11, or IFN-γ followed by CH11 as described in Induction and detection of apoptosis. Cultures were lysed in 50 mM HEPES, 100 mM NaCl, 0.1% CHAPS, 1 mM EDTA, 10% and glycerol containing 1.0% NP-40. An aliquot was reserved for protein quantitation (BCL assay), and then dithiothreitol was added to a final concentration of 10 mM. Caspase activity was assayed using caspase 3 (Ac-DEVD-pNA) colorimetric substrate (BIOMOL Research Laboratories, Plymouth Meeting, PA) following the manufacturer's recommendations. Recombinant active caspase 3 was used as a control. Optical density readings were performed at 0, 1, 2, and 3 h. The data are reported as the change in optical density for lysates from treated cells compared with untreated cells (defined as 1.00) normalized to total protein.
HIPEC lines are relatively insensitive to Fas-mediated apoptosis.
We examined whether cross-linking of Fas antigen could induce HIPEC apoptosis. As seen in Fig. 1, treatment of HIPEC lines with agonist anti-Fas antibody CH11 for 18–20 h induced a <25% increase in the number of annexin V-positive cells above background. This increase in cell death was significant in jejunum HIPEC (P < 0.001) but not in colon HIPEC (P > 0.05). Incubation with control IgM antibody had no effect on HIPEC lines (data not shown). Treatment of HIPEC lines with up to 300 ng/ml CH11 did not induce any further increase in the number of apoptotic cells, indicating that the difference in HIPEC and Jurkat sensitivity to CH11 was not due to differences in antibody dose requirements (data not shown). These results suggest that additional stimuli may be required for sensitization of HIPEC for Fas-mediated apoptosis. Therefore, we also assessed the effects of proinflammatory cytokines on HIPEC sensitivity to Fas-mediated apoptosis. HIPEC lines were incubated with IFN-γ or TNF-α for 18–20 h before the addition of CH11 (Fig. 1). Cytokines alone did not induce apoptosis of either colon or jejunum HIPEC. Addition of CH11 to IFN-γ-treated colon and jejunum HIPEC induced significant numbers of annexin V-positive cells compared with either medium or IFN-γ alone. Also, the level of apoptosis in jejunum HIPEC treated with IFN-γ and CH11 was significantly increased compared with jejunum HIPEC treated with CH11 alone. Similar results were obtained using another colon HIPEC line (data not shown). In contrast to IFN-γ, treatment with TNF-α did not alter HIPEC susceptiblity to CH11-induced cell damage (Fig. 1). HIPEC lines were not primed for Fas-mediated apoptosis if treated with IFN-γ for only 6 h before addition of CH11 or if treated with IFN-γ and CH11 concurrently for 18–20 h (data not shown). In addition, we determined the effects of GM-CSF and combinations of TNF-α + IFN-γ on HIPEC susceptibility (data not shown). Preincubation with GM-CSF did not enhance CH11-induced apoptosis of HIPEC lines. Incubation of HIPEC with a combination of IFN-γ + TNF-α induced a significant level of cell death by 48 h. However, preincubation with IFN-γ + TNF-α (20 h) followed by CH11 induced a level of cell death that was less than the added values for IFN-γ + TNF-α and IFN-γ followed by CH11. This indicated that IFN-γ + TNF-α was no better at priming HIPEC for Fas-mediated apoptosis than was IFN-γ alone and that IFN-γ-treatment primed HIPEC for TNF-α-mediated cell damage as well as Fas-mediated cell apoptosis.
HIPEC lines express surface Fas, which is not altered by treatment with proinflammatory cytokines.
To determine if HIPEC resistance to Fas-mediated apoptosis was related to a decrease in Fas expression in culture, colon and jejunum HIPEC lines were stained with anti-Fas DX2 monoclonal antibody or an isotype control and analyzed by flow cytometry. As seen in Fig.2, colon and jejunum HIPEC lines consistently expressed high levels of Fas antigen, comparable with that of freshly isolated colon IEC. Fas expression on HIPEC was also comparable with that of Jurkat cells, the prototypical Fas-sensitive cell line. Previous studies have demonstrated that proinflammatory cytokines may induce or enhance the expression of Fas by IEC lines (1, 2, 34, 42). Therefore, we evaluated whether a similar regulatory mechanism of Fas expression existed in our HIPEC system. In contrast to these previous reports, treatment of HIPEC with IFN-γ, TNF-α, or GM-CSF had no effect on Fas antigen expression (Fig. 2). Addition of lipopolysaccharide or the combination of IFN-γ + lipopolysaccharide to HIPEC cultures did not alter Fas expression either (data not shown). Lack of response to TNF-α and GM-CSF was not due to a lack of receptor expression because, in a series of independent experiments, TNF-α and GM-CSF were both found to bind HIPEC in a specific manner (data not shown). Also, IFN-γ treatment induced upregulation of HIPEC major histocompatability complex class II expression, providing functional evidence for expression of IFN-γ receptor on HIPEC (data not shown).
Fas-mediated death of HIPEC is apoptotic.
To confirm that the HIPEC death caused by IFN-γ + CH11 stimulation was truly apoptotic and not necrotic, we assayed for the presence of caspase 3 activation fragments or subunits. We chose to look at caspase 3 for two reasons. First, caspase 3 is an effector caspase commonly activated by many apoptotic stimuli. Secondly, we wanted to determine whether IFN-γ-mediated increase in apoptosis correlated with an increase in effector caspase activation. Jurkat was chosen as a control line because it is known to utilize caspase 3 activation for Fas-mediated apoptosis (5, 57). As seen in Fig.3 A, Jurkat cells (control) treated with CH11 for 4 h contained caspase 3 proenzyme (36 kDa) and three large subunits, p25, p22, and p20. These fragments correspond to the p24 intermediate form and activated p20 and p17 subunits, previously reported (9, 23). HIPEC contained ample amounts of caspase 3 proenzyme, comparable to that found in the Fas-sensitive Jurkat cell line (Fig. 3 A) as well as HT-29 (data not shown). Untreated or IFN-γ-treated HIPEC contained no large subunits of caspase 3, and IFN-γ treatment did not induce increased expression of procaspase 3. HIPEC treated with CH11 alone contained proenzyme and the intermediate p25 fragment, indicating that caspase 3 cleavage had taken place and implying caspase 3 activation. However, only low levels of p22, and no p20 fragment, were detected in lysates from CH11-treated HIPEC. In contrast, HIPEC primed with IFN-γ followed by CH11 contained more completely activated caspase 3, as indicated by the detection of increased levels of the p25 intermediate fragment as well as the detection of p22 and p20 activation fragments. These results confirm that Fas-mediated death of HIPEC was apoptotic and indicate that the level of caspase activation correlated with the level of apoptosis in HIPEC lines.
Sensitivity of HIPEC to CH11-induced apoptosis was also assessed by TUNEL assay for the detection of DNA fragmentation (Fig.3 B). The IEC adenocarcinoma line HT-29 was chosen as a control because it is of epithelial cell lineage and because it has been shown to be sensitized for Fas-mediated apoptosis by preincubation with IFN-γ (1, 2, 34). Treatment with medium or IFN-γ alone induced no detectable apoptosis of HIPEC or HT-29, whereas CH11 treatment resulted in the generation of low amounts of fragmented DNA in HIPEC but not HT-29. Priming with IFN-γ followed by the addition of CH11 induced significant apoptotic death of both HIPEC and HT-29. These results further confirm that Fas-mediated death of HIPEC was apoptotic.
To further determine if caspase 3 activation correlated with apoptosis, lysates from treated and untreated colon HIPEC cultures were assessed for caspase 3 activity using Ac-DEVD-pNA as substrate. Caspase 3 activity was 1.733 ± 0.480-fold (mean ± SE) higher in colonic HIPEC treated with CH11 alone and 5.017 ± 0.329-fold higher in cultures treated with IFN-γ and CH11 than in untreated cultures (n = 3). By comparison, caspase 3 activity was 10.845 ± 2.185-fold higher in CH11 treated Jurkat cultures (n = 2). These results further demonstrate that HIPEC were relatively resistant to Fas-mediated apoptosis but that caspase 3 is active in HIPEC treated with CH11 alone. It also confirms that IFN-γ-mediated sensitization of HIPEC for Fas-mediated apoptosis correlated with an increase in caspase 3 activity.
Apoptosis of HIPEC was delayed and correlated with the generation of the p22/20 activation fragment of caspase 3.
In initial experiments, we found that when untreated or IFN-γ-treated HIPEC were incubated with CH11 for 4–6 h there was no increase in the number of annexin V-positive cells above background. By comparison, CH11-induced apoptosis of Jurkat was morphologically evident by 2 h and IFN-γ-primed HT-29 showed evidence of apoptosis 4–6 h after addition of CH11, as previously reported (2,34). To further investigate the kinetics of Fas-mediated apoptosis in HIPEC, we performed a time-course study using IFN-γ-treated HIPEC as well as Jurkat cells for comparison. Annexin V-positive Jurkat cells were detected 2 h after the addition of CH11 (Fig. 4 A). By comparison, apoptosis was not detected in IFN-γ-primed HIPEC lines until after 8-h incubation with CH11. Immunoblots for the detection of caspase 3 subunits (Fig. 4 B) show that the p25 subunit was generated by 4 h, but p22/20 subunits were not detected until the onset of apoptosis at or after 8 h of treatment. These data demonstrate that the onset of Fas-mediated apoptosis in HIPEC was significantly delayed and correlated with the generation of the active subunits of caspase 3.
The data thus far indicate that untreated HIPEC are comparatively resistant to Fas-mediated apoptosis and that the onset of apoptosis in unprimed and IFN-γ-primed HIPEC was significantly delayed compared with the prototypical Fas-sensitive cell line Jurkat as well as the colonic adenocarcinoma epithelial cell line HT-29. The medium for the culture of HIPEC contained multiple growth factors, some of which may inhibit apoptosis (4,12). To determine whether HIPEC sensitivity to Fas-mediated apoptosis was effected by the culture medium, colon HIPEC and a colonic adenocarcinoma line, LS180, were grown in HIPEC medium or HIPEC medium without growth factors (F-12). LS180 was used instead of HT-29 because LS180 constitutively expressed Fas and is sensitive to Fas-mediated apoptosis and because HT-29 viability significantly decreased when cultured in HIPEC medium. There was no significant difference in Fas sensitivity between cultures (untreated or IFN-γ-treated) grown in media with and without growth factors (data not shown). Therefore, HIPEC insensitivity to Fas-mediated apoptosis was not due to the presence of exogenous growth factors in the culture medium.
Initiator caspases 8 and 9 as well as effector caspase 3 are required for Fas-mediated apoptosis of HIPEC.
Fas-mediated apoptosis may require activity of both initiator caspases 8 and 9 (44). To identify the specific caspases required for CH11-induced apoptosis in HIPEC, we used synthetic peptide inhibitors, z-IETD-fmk, z-LEHD-fmk, and z-DEVD-fmk, to irreversibly inhibit caspase 8, 9, and 3 activity, respectively (Fig.5). Jurkat cells were run as controls, and our findings were consistent with those of previous reports (5, 19, 57). All three peptides inhibited CH11-induced apoptosis in untreated as well as IFN-γ-treated HIPEC lines, suggesting that HIPEC require both initiator caspases 8 and 9 as well as effector caspase 3 for Fas-mediated apoptosis. All of the peptides tested inhibited apoptosis in a dose-dependent fashion from 1 μM to 50 μM (1, 15, 30, and 50 μM) in primed and unprimed HIPEC (data not shown). It should also be noted that z-IETD-fmk (caspase 8) and z-LEHD-fmk (caspase 9) were more inhibitory in untreated and IFN-γ-treated HIPEC than in Jurkat cells. Although not all of these differences were statistically significant, the findings suggest that HIPEC lines might contain less active caspase 8 and 9 than Jurkat. In addition, z-DEVD-fmk inhibited more apoptosis in untreated HIPEC than in IFN-γ-treated HIPEC or Jurkat. Again, although the differences were not statistically significant, the data suggest that untreated HIPEC contained less activated caspase 3 than did Jurkat cells, and that IFN-γ treatment of HIPEC resulted in increased production of active caspase 3. Indeed, as shown in Fig. 3 A, lysates from Jurkat cells contained more active caspase 3 (p22/20 fragment) than lysates from untreated HIPEC. Furthermore, lysates from IFN-γ-treated HIPEC contained increased amounts of p22 and p20 subunits compared with untreated HIPEC (Fig. 3 A).
HIPEC express comparatively low levels of procaspase 8, and IFN-γ treatment induces a less than onefold increase in procaspase 8 expression but does not alter expression of FADD or I-FLICE.
The peptide inhibition data indicated that HIPEC required caspase 8 activity for Fas-mediated apoptosis and suggested that HIPEC lines might contain less activated caspase 8 than Jurkat cells. To determine if IFN-γ-mediated sensitization of HIPEC for apoptosis was dependent on upregulation of procaspase 8 expression, we performed immunoblot analysis of lysates from untreated and IFN-γ-treated HIPEC as well as HT-29 and Jurkat cells (controls). As seen in Fig. 6 A, HIPEC expressed two isoforms of procaspase 8 (45). HIPEC expressed substantially lower levels of procaspase 8 than either Jurkat or HT-29. Also, IFN-γ treatment of HIPEC resulted in a less than onefold increase in expression of procaspase 8 compared with untreated HIPEC (Fig. 6 B). IFN-γ treatment also induced an increase in procaspase 8 in HT-29 (Fig. 5 A), as previously reported (34).
The recruitment and activation of procaspase 8 by Fas ligation is dependent on the binding of the Fas adaptor protein FADD (60). Therefore, it was of interest to determine whether IFN-γ primed HIPEC for Fas-mediated apoptosis by inducing an increase in expression of FADD. As seen in Fig. 6 C, HIPEC expressed only a slightly lower level of FADD than did Jurkat and HT-29, and treatment with IFN-γ did not alter FADD expression in either colon or jejunum HIPEC. Because Fas-mediated apoptosis is dependent on caspase 8 activity (56) and IFN-γ did not significantly alter either procaspase 8 or FADD expression, we determined whether IFN-γ induced any change in expression of the caspase 8 inhibitor I-FLICE (15). As seen in Fig.6 C, HIPEC expressed I-FLICE, but again treatment with IFN-γ did not alter its expression.
IFN-γ-mediated sensitization of HIPEC is independent of procaspase 9, Bcl-2 family proteins, or IAP expression.
Because the peptide inhibition data suggested that Fas-mediated apoptosis of HIPEC was dependent on caspase 9 activity and, by extension, release of mitochondrial cytochrome c, we determined whether IFN-γ treatment modulated expression of procaspase 9 or Bcl-2 family proteins. These data (Fig.5) also suggested that HIPEC expressed decreased levels of caspase 9. As seen in Fig.7 A, HIPEC expressed less procaspase 9 than either Jurkat or HT-29 and IFN-γ-priming did not upregulate expression of this initiator caspase. Also, HIPEC expressed BCLx, BAX, and BAK, but not BCL-2 (Fig.7 B). However, IFN-γ treatment did not induce an increase in expression of any of these apoptosis regulators. IAP functions, in part, by binding to procaspase 9 (8, 9). However, as seen in Fig. 7 C, IFN-γ-treatment did not alter expression of XIAP or c-IAP2 (HIAP-1). In addition, we found that HIPEC did not express c-IAP1 (HIAP-2), and IFN-γ treatment did not induce its expression (data not shown).
The data thus far suggest that IFN-γ treatment sensitized HIPEC for Fas-mediated apoptosis without regulating mitochondrial events or caspase 9 expression. To further determine which part of the caspase activation pathway is affected by IFN-γ, untreated and IFN-γ-treated HIPEC were incubated with CH11, camptothecin, or etoposide and then assayed for apoptosis by annexin V-FITC staining. HT-29 was run as a control, and, as previously reported (33, 34), IFN-γ primed HT-29 for drug-induced apoptosis as well as Fas-mediated apoptosis (Fig.8). Like HT-29, colon HIPEC were resistant to drug-induced apoptosis. However, in contrast to HT-29, colon HIPEC were primed only for Fas-mediated apoptosis and not for apoptosis induced by etoposide or camptothecin. This finding supports the above data, indicating that IFN-γ primed HIPEC for Fas-mediated apoptosis without directly affecting mitochondrial events or caspase 9 activation.
IFN-γ treatment induced an increase in expression of caspase 4 intermediate subunit but not caspase 1.
It has been suggested that caspase 1 or caspase 4 may play a role in Fas-mediated apoptosis (18, 32-34). To assess whether these caspases were involved in IFN-γ-mediated sensitization of HIPEC to Fas-mediated apoptosis, untreated and IFN-γ-treated colon HIPEC were incubated with z-YVAD-fmk peptide inhibitor before addition of anti-Fas antibody and assayed for apoptosis by annexin V-FITC staining. As seen in Fig.9 A, z-YVAD-fmk inhibited apoptosis in a dose-dependent fashion in both untreated and IFN-γ-treated colon HIPEC, suggesting that an interleukin-1β-converting enzyme (ICE) family caspase (caspase 1-like) was involved in Fas-mediated apoptosis of HIPEC. Western blotting for caspase 1 showed that, although caspase 1 was presumably activated during Fas-mediated apoptosis, as indicated by the detection of a p33 intermediate fragment, IFN-γ treatment did not increase the level of procaspase 1 expression nor the level of its activation. In contrast, the proenzyme form (p45) and intermediate-sized fragments (p40 and p35) of caspase 4 were consistently detected in both untreated and CH11-treated HIPEC. IFN-γ treatment did not induce a significant increase in procaspase 4 and a less than onefold increase in the p40 band. However, IFN-γ treatment did induce a significant increase in expression of the p35 intermediate fragment (up to a 3.3-fold increase above untreated HIPEC) (Fig.9 C). Despite the apparent activation of caspase 4, as suggested by the detection of an intermediate fragment, the active fragment of caspase 4 (p20) was not detected in lysates from any HIPEC culture.
We have used an in vitro model of nontransformed HIPEC to study Fas-mediated epithelial apoptosis in the intestine. As previously reported for IEC (27, 36), HIPEC constitutively expressed Fas antigen on their surface. However, Fas expression did not correlate with susceptibility to Fas-mediated apoptosis in that agonist antibody induced only marginal amounts of HIPEC death, as detected by two independent types of assays for apoptosis: detection of phosphatidylserine inversion (annexin V staining; Fig. 1) and fragmentation of DNA (TUNEL assay; Fig. 3 B). These results are supported by the finding that caspase 3 was poorly activated after addition of anti-Fas, as indicated by the low level of p22/20 fragment detected by Western blot as well as by caspase 3 activity assays. Previous work using organ culture of isolated intestinal crypts suggested that IEC are sensitive to Fas-mediated apoptosis (42, 49), and although it is known that many transformed IEC lines are often resistant to Fas-mediated apoptosis, it is somewhat surprising to find that nontransformed primary cells are relatively resistant as well. In addition, the onset of apoptosis in HIPEC was significantly delayed compared with Jurkat as well as HT-29, further demonstrating that HIPEC lines are insensitive to Fas-mediated apoptosis.
HIPEC are relatively resistant not only to Fas-mediated apoptosis but also to apoptosis induced by TNF-α, camptothecin, and etoposide, and although the level of Fas apoptosis is significantly increased in IFN-γ-sensitized cells, the onset of apoptosis is still delayed. The comparatively low level of caspase 3 activation fragments detected in lysates from CH11-treated HIPEC lines, and the observation that caspase 3 activity in IFN-γ + CH11-treated HIPEC is about half that of CH11-treated Jurkat, suggests that caspase activation was inefficient in HIPEC. Although the level of expression of the proenzyme form of a caspase is not indicative of its activation status, the availability of that procaspase form may be a limiting factor in the efficiency of activation of the caspase cascade. HIPEC lines express a level of procaspase 3 comparable to that of Jurkat and HT-29 cell lines, indicating that this is not a limiting factor in the effector caspase activation. In contrast, our results demonstrate that the levels of procaspases 8 and 9 are decreased in HIPEC compared with both Jurkat and HT-29. In addition, caspase 8 and 9 activation fragments could not be detected in lysates from unprimed CH11-treated HIPEC and these fragments were just barely detectable in lysates from IFN-γ-primed, CH11-treated cells, even when 75–100 μg of total protein were loaded onto gels for Western blotting (data not shown). These observations lend support to the hypothesis that caspase activation is inefficient in HIPEC lines and suggest that decreased initiator procaspase expression may contribute to HIPEC resistance to apoptosis. Also to be considered is the observation that HIPEC express multiple inhibitors of apoptosis, including the caspase 8 inhibitor I-FLICE (15) as well as the caspase 9 and 3 inhibitors XIAP and c-IAP2 (8, 9). Whether HIPEC resistance to apoptosis and poor activation of the caspase cascade is due to low initiator caspase expression or the action of apoptosis inhibitors has yet to be determined.
IEC express receptors for multiple cytokines (38, 41) that regulate IEC proliferation, differentiation, and function (39,43). In transformed cell lines, inflammatory cytokines have been reported to modulate apoptotic pathways by regulating the expression of Fas (1, 2, 10, 34, 42, 54, 58) and Fas-L (6, 28). This IFN-γ-induced increase in Fas expression frequently correlates with increased sensitivity to apoptosis induced by agonist anti-Fas antibody (1, 2, 34). In a recent report, Ruemmele et al. (42) report that both IFN-γ and TNF-α upregulate Fas expression and sensitize nontransformed, primary cultures of human IEC for Fas-mediated apoptosis. Consistent with these reports (1, 2, 34,42), we found that IFN-γ treatment primed HIPEC for anti-Fas-stimulated apoptosis. However, in direct contrast to these reports, IFN-γ-mediated sensitization of HIPEC did not correlate with regulation of Fas expression (1, 2, 34,42). Our finding that proinflammatory cytokines do not induce an increase in HIPEC Fas expression is in accordance with the observation that Fas expression is not upregulated on epithelial cells of the intestinal mucosa in patients with inflammatory bowel disease compared with normals (49, 55).
In addition to regulation of surface death receptor expression, inflammatory cytokines have been shown to sensitize IEC lines for apoptosis by modulating intracellular regulators, or effectors, of apoptosis (32-34). We found that IFN-γ priming of HIPEC resulted in enhanced generation of caspase 3 activation fragment p22/20 and a 2.9-fold increase in caspase 3 activity above that for unprimed HIPEC. Also, the onset of Fas-mediated apoptosis correlated with appearance of the caspase 3 p22 activation subunit, and inhibition of caspase 3 activity blocks Fas-mediated apoptosis of HIPEC. These data suggest that caspase 3 is the principle effector caspase in Fas-mediated apoptosis in HIPEC. However, we have not excluded the possibility that IFN-γ-enhanced apoptosis was due to upregulation of expression or activation of additional caspases, such as caspase 6 or 7. In cell-free systems, caspase 8 and 9 have been shown to directly or indirectly cleave and activate caspases 2, 3, 6, 7, and 10 (31, 52). Also, z-DEVD-fmk is not an exclusive inhibitor of caspase 3 (11, 53). Nonetheless, our data strongly suggest that caspase 3 is a primary effector caspase in Fas-mediated apoptosis of HIPEC.
HIPEC expressed high levels of procaspase 3, comparable to that of Jurkat (Fig. 3) and HT-29 (data not shown), and IFN-γ priming did not induce any further increase in procaspase 3 expression. This indicates that IFN-γ-mediated sensitization occurs proximal to caspase 3 cleavage. The data indicate that Fas-mediated apoptosis of HIPEC is dependent on two initiator caspases, caspase 8 and caspase 9. Therefore, IFN-γ treatment could have upregulated apoptosis by affecting either caspase 8 or caspase 9 activation.
Caspase 8 activation requires autoproteolysis. This activation through self-cleavage is dependent on procaspase 8 recruitment to the DISC, which in turn is dependent on the adaptor protein FADD (5, 19,57, 60). Caspase 8 activity is also regulated by I-FLICE, a mammalian homologue of a family of viral proteins called FLIPs (FLICE-inhibitory proteins) that inhibits autoproteolysis by blocking caspase 8 recruitment by FADD (15). We determined that the expression of FADD and I-FLICE is unaltered by incubation with IFN-γ. However, IFN-γ treatment did induce a small increase in expression of procaspase 8 (<1-fold). It is possible that this small increase in caspase 8 concentration was sufficient to promote more efficient autoactivation with a subsequent increase in activation of downstream caspases (caspase 9 and 3).
IFN-γ treatment of HIPEC did not induce an increase in procaspase 9 expression. However, this does not indicate whether caspase 9 activity was increased or not. Regulators of caspase 9 activation or activity include Bcl-2 family proteins (3) and IAP family proteins (8, 9). IFN-γ priming of HIPEC did not alter expression of proapoptotic BAX or BAK or antiapoptotic BCLx, suggesting that IFN-γ treatment did not affect the release of cytochrome c and subsequently the activation of caspase 9. IAP proteins, such as XIAP, have been shown to bind the proenzyme form of caspase 9 as well as the activated form of caspase 3 and inhibit proteolytic activity (8, 9). However, we found that although HIPEC express XIAP and c-IAP2, IFN-γ treatment did not modulate their expression. These data suggesting that IFN-γ did not prime HIPEC through the direct regulation of caspase 9 activation are supported by the finding that IFN-γ did not sensitize HIPEC for apoptosis induced by either camptothecin or etoposide (50). It also indicates that the target(s) of IFN-γ-mediated priming was upstream of caspase 9 activation. These findings differ from those of several studies using IEC tumor lines in which IFN-γ treatment sensitized for both Fas- and drug-mediated apoptosis (32-34).
Caspases 1 and 4, two closely related proteases with similar amino acid sequences and substrate specificity (30, 52), have previously been implicated in Fas-mediated apoptosis (18,32-34). IFN-γ-mediated sensitization of IEC tumor lines has been shown to involve upregulation of caspase 1 or 4 expression (33, 34). Although caspase 1 is activated in CH11-treated HIPEC, we found no evidence of IFN-γ-mediated regulation of caspase 1 expression or activation. In contrast, IFN-γ-mediated sensitization of HIPEC to Fas-mediated apoptosis correlated with the detection of significantly increased levels of an intermediate-sized fragment (p35) of caspase 4. Interestingly, two intermediate-sized fragments, p40 and p35, were detected in untreated as well as CH11-treated colon HIPEC cultures. It has been suggested that, like for caspase 1 (59), these intermediate-sized fragments are indicative of caspase 4 activation (18) or perhaps partial activation. It is probable that, as for caspase 1 (59) and caspase 3 (23), the intermediate fragment of caspase 4 is proteolytically active, but to a lesser extent than the fully activated p20/p10 subunit aggregate. Also, this intermediate fragment may not have the same substrate specificity as the fully activated form but is primarily involved in autoproteolysis (59). However, a caspase 4 p20 activation fragment was not detected in lysates from either treated or untreated HIPEC. The concentration of p20 fragment may have been too low for detection by Western blotting but high enough for proteolytic activity, in which case it could cleave and activate effector caspase 3 (18). However, the fact that IFN-γ treatment alone does not induce apoptosis of HIPEC or sensitize cultures for drug-induced apoptosis suggests that the priming effect of elevated caspase 4 p35 expression requires caspase 8 activity. An alternative possibility is that an increase in expression of caspase 4, independent of its activation status, may contribute to the activation of caspase 8. This has been demonstrated for caspase 1 (51). HeLa cells transfected with only the prodomain of caspase 1 are more sensitive to Fas-mediated apoptosis, and this sensitization was specific for apoptosis triggered through Fas but not by etoposide (51). Tatsuta et al. (51) demonstrated that this caspase 1 prodomain-mediated Fas sensitization correlated with an increase in activity of caspase 8 as well as downstream increases in caspase 3 activity. Therefore, in the case of IFN-γ-sensitized HIPEC, a caspase 4-mediated increase in caspase 3 activity may occur through the enhanced activation of caspase 8 and not require caspase 4 activation (generation of a p20 subunit). How an increase in caspase 4 p35 expression can regulate caspase 8 activation remains to be determined.
Our finding that nontransformed IEC are relatively resistant to Fas-mediated apoptosis has physiological relevance in that it suggests that Fas may not play a significant role in normal IEC turnover. However, in conditions in which inflammatory cytokines are increased in the intestinal mucosa, such as inflammatory bowel disease or graft-versus-host disease, Fas-mediated apoptosis may contribute significantly to the pathophysiology of these diseases. Also, our finding that IFN-γ-mediated sensitization of nontransformed IEC correlates with expression of caspase 4 intermediate subunit contributes to our understanding of the diversity of regulatory mechanisms that may be involved in controlling apoptosis in the intestinal mucosa.
This work was supported by a First Award Grant (to A. Panja) from the Crohn's and Colitis Foundation of America and Winthrop-University Hospital Intramural Funding (to C. A. Martin and A. Panja).
Address for reprint requests and other correspondence: A. Panja, Gastrointestinal Research Laboratory, Winthrop-University Hospital, 222 Station Plaza North, Suite 511, Mineola, NY 11501 (E-mail:).
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 © 2002 the American Physiological Society