|Home | About | Journals | Submit | Contact Us | Français|
A potential application of gliotoxin therapy for liver fibrosis was suggested by its apoptotic effect on fibrogenic activated stellate cells. We investigated if gliotoxin exerts similar effects on hepatic macrophages Kupffer cells.
Effects of gliotoxin on Kupffer cells isolated from the normal liver and in vivo following its administration to CCl4–induced cirrhotic rats were studied.
Gliotoxin caused apoptosis of cultured Kupffer cells, the effect being apparent at 0.3 μM concentration within 1 hour; longer incubation caused necrosis. This effect was associated with mitochondroial cytochrome c release, caspase-3 activation and ATP depletion. Interestingly, inhibition of caspase-3 and serine proteases accelerated and augmented gliotoxin-induced cell death via necrosis. Gliotoxin stimulated nuclear translocation of NFκB, and phosphorylation of p38, ERK1/2 and JNK MAP kinases, but these signaling molecules were not involved in gliotoxin-induced death of Kupffer cells. In vivo administration of gliotoxin to cirrhotic rats caused apoptosis of Kupffer cells, stellate cells and hepatocytes. In control rats, the effect was minimal on the nonparenchymal cells and not apparent on hepatocytes.
In the fibrotic liver, gliotoxin nonspecifically causes death of hepatic cell types. Modification of gliotoxin molecule may be necessary for selective targeting and elimination of activated stellate cells.
Gliotoxin, a fungal metabolite, belongs to the epipolythiodioxopiperazine class of compounds. Gliotoxin (0.3–7.5 μM) was reported to cause apoptosis of activated rat and human hepatic stellate cells (HSCs) by inducing mitochondrial permeability, cytochrome c release and caspase-3 activation (1,2); higher concentrations (>7.5 μM) caused necrosis of activated human HSCs in association with increased oxidative stress (2). These were important observations as activated HSCs play a central role in liver fibrosis and sinusoidal component of portal hypertension (3). HSCs produce several growth factors, cytokines and chemokines (4,5) responsible for initiation and progression of liver pathology. Therefore, strategies to remove selectively activated HSCs from the fibrotic liver are intensely investigated.
Gliotoxin possesses immunomodulating activity: it stimulates lymphocyte proliferation, and induces T cell cytotoxicity and lymphokine release by T cells (6–9). Gliotoxin (10–100 nM) was found to inhibit phagocytic activity of peritoneal and circulating macrophages (6,7,9) at concentrations lower than those required to cause apoptosis of HSCs. High concentration of gliotoxin (>1 μM) induces reactive oxygen species (ROS)-mediated apoptosis of peritoneal macrophages, independent of the inhibition of phagocytosis (8). The role of hepatic macrophages Kupffer cells extends from defense against invading microbes and clearance of toxic substances to liver growth and immune regulation (10–13). These functions of Kupffer cells are even more critical in disease states as their removal can significantly augment pathological development (14). Since gliotoxin-induced elimination of activated HSCs could be therapeutic strategy for liver fibrosis, it is important to rule out its adverse effects on other hepatic cell types. Therefore, we investigated the effects of gliotoxin on rat Kupffer cells in vitro, and in control and CCl4-induced cirrhotic rats in vivo.
Gliotoxin, TPCK, MG132 (Sigma, St. Louis, MO); PD98059, SB203580, SP600125 and PDTC (Calbiochem, La Jolla, CA); Z-DEVD-fmk (R&D, Minneapolis, MN); Mn-TBAP (Calbiochem, La Jolla, CA); anti-rabbit NFκB-p65, ERK1/2 and caspase-3, mouse anti-rat desmin antibodies and CY-3-conjugated goat anti-rabbit secondary antibody (Santa Cruz Biotechnology, Santa Cruz, CA); anti-rabbit monoclonalP-ERK1/2, P-p38, p38, P-JNK and JNK antibodies (Cell Signaling, Danvers, MA); anti-rat ED2 antibody (Serotech, Raleigh, NC); peroxidase-linked anti-rabbit IgG (Amersham-Pharmacia, Piscataway, NJ) were purchased.
IACUC, University of Pittsburgh approved the protocols in accordance with the NIH guidelines. Kupffer cells, prepared essentially as described previously (15), were suspended in William’s medium E containing 10% fetal calf serum and plated at a density of 0.5 × 106/cm2. Medium was renewed after 3h and following day; cells were used on the third day. Purity as determined by immunostaining for ED2 (Kupffer cells), desmin (stellate cells) and SE-1 (endothelial cells) was greater than 95%. Rat liver-derived HSCs were activated in culture and used in passage 3–5 as describe previously (16).
Cell viability was determined by MTT assay as described (16). To determine apoptosis, the cells were washed with PBS, fixed in 4% parafarmaldehyde and incubated with 5 μg/ml propidium iodide at room temperature for 5 min. Morphological evaluation of propidium iodide-stained nuclei was performed by excitation at 380 nm and emission at 520 nm (17). For DNA fragmentation analysis, attached, loosely attached and detached cells were pooled. Total DNA was extracted and following electrophoretic separation (16), ethidium bromide-stained DNA bands were visualized under UV light.
LDH release and cellular ATP were measured using LDH Liqui-UV Kit [Stanbio Labs (2940-430)] and Cell Viability Assay Kit-ATP (Dojindo Molecular Technologies, Gaithersburg, MD) respectively. For additional determination of membrane damage, cells from the culture supernatant were pelleted by centrifugation, and the attached cells were trypsinized and washed. The two fractions were pooled, suspended in culture medium and mixed with trypan blue. Trypan blue-positive and negative cells were counted.
Quantikine murine immunoassay kit (R&D, Minneapolis, MN) and caspase fluorescent assay kit (BD Biosciences-Clonetech, San Jose, CA) respectively were used to measure cytosolic cytochrome c and Caspase-3-like activity.
Western analysis for caspase-3 and MAPKs was performed as described previously (18).
The cells were washed and loaded with 10 μm hydroethidine (O2.−) or dihydro-2′7′-dichlorofluorescein diacetate (H2O2) for 30 min at 37°C (20,21). After washing, the cells were treated with gliotoxin, collected by trypisization, and washed with PBS. Fluorescence was measured by flow cytometry within 30 minutes in a Coulter Epics XL Flow Cytometer.
Cirrhosis was induced in male Sprague-Dawley rats by 8 weeks of CCl4/phenobarbital treatment as described (22). Gliotoxin (3 mg/kg) was administered intraperitoneally 24h after the last CCl4 injection. The liver was excised 6h later, rinsed with cold PBS, cut into approximately 1 cm3 pieces, fixed in 4% paraformaldehyde for 2h, placed in 30% sucrose for 16–24h and frozen in 2-methylbutane cooled with liquid nitrogen. Cryostat sections (5 μm) were cut, washed with PBS, incubated in TUNEL reagents (Roche) (30 min; 37°C), followed by streptavidin CY3 (30 min; 22°C). After blocking in 2% BSA solution, sections were incubated with primary antibodies (1:200 in BSA solution) followed by Cy3-conjugated secondary antibody (Jackson Laboratories), then treated with Hoechst dye (1 mg/100 ml), coverslipped with Gelvatol, and analyzed under Olympus BX51 fluorescence microscope.
Statistical significance between the groups was determined by one-way ANOVA followed by a test for linear trend. A P value of <0.05 was considered statistically significant.
Viability of control cells did not change over 24h (Figure 1A). Gliotoxin (0.03 μM) caused significant reduction in viability only at 24h. At 0.3 μM and 3.0 μM, gliotoxin caused time-dependent loss of viability that was significant at 3h and 1h respectively (Figure 1A); incubation beyond 3h at both concentrations caused detachment of the majority of cells. Gliotoxin caused equivalent loss of viability of Kupffer cells even in the presence of serum (Figure 1B). Pancreatic elastase (apoptotic agent for Kupffer cells) (23) reduced the viability moderately, while necrosis-inducing chlorpromazine (1), caused profound loss of viability (Figure 1B).
We determined if apoptosis is a mechanism of gliotoxin-induced reduced viability of Kupffer cells. Apoptotic cells increased progressively upon treatment with 0.03 to 3 μM gliotoxin (Figure 2A and 2B). DNA fragmentation analysis confirmed substantial genomic DNA damage at 3h of incubation with gliotoxin (Figure 2C). Incubation for 4h with elastase caused very mild DNA fragmentation; gliotoxin treatment of activated HSCs showed classical DNA fragmentation (Figure 2D).
ATP content did not decrease significantly with 0.03 μM gliotoxin (Figure 3A). However, 0.3 μM gliotoxin caused rapid and progressive decrease in ATP that was profound between 3 and 6h. Robust ATP depletion occurred in cells treated with 3 μM gliotoxin even at 30 min. ATP decreased by about 20% compared to control upon 4h elastase treatment, and could not be detected in chlorpromazine-treated cells at 3h (not shown).
Profound apoptosis and ATP depletion lead to secondary necrosis characterized by membrane damage, LDH release and trypan blue uptake. The time-course and concentration-dependence experiment showed no LDH release at low concentration of gliotoxin (0.03 μM) (Figure 3B). At 0.3 μM gliotoxin, no LDH release occurred at 3h (Figure 3B) although decreased viability and DNA fragmentation was noted (Figures 1 and and2);2); LDH release occurred at 4.5h, and was maximal at 6h of incubation. At 3 μM gliotoxin, LDH release occurred earlier (at 3h) and was maximal at 6–12h. Chlorpromazine caused LDH release equivalent of that in cells treated with 3 μM gliotoxin and elastase did not exert any effect (not shown).
No difference in the number of cells taking up trypan blue was observed between control and 0.03 μM gliotoxin-treated cells (Figure 3C). Progressive increase in trypan blue uptake was caused by 0.3 μM gliotoxin after 3h. These results suggest that gliotoxin-induced apoptosis (before 3h) precedes cell membrane damage (LDH release and trypan blue uptake after 3h). Treatment with elastase and chlorpromazine for 3h caused trypan blue uptake by 8 ± 3 and 93 ± 7% Kupffer cells. Significant increase in trypan blue uptake was observed in cells treated with 3 μM gliotoxin only for 1h. Thus at high concentration, necrosis seems to be the predominant mode of gliotoxin-induced death of Kupffer cells.
Gliotoxin-induced apoptosis of activated HSCs is associated with increased mitochondrial permeability suggesting the release of cytochrome c (24), a caspase-3 activator (25). In gliotoxin-treated Kupffer cells, cytochrome c release occurred at 15 min and progressively thereafter (Figure 4A). Caspase-3 activity, as determined enzymatically, increased time-dependently from 15 min (Figure 4B). However, cleavage of caspase-3 into the active form was observed at 2 and 3h by Western analysis (Figure 4C).
To determine the role of caspases in gliotoxin-induced apoptosis, the cells were preincubated with caspase inhibitor Z-DEVD.fmk. Surprisingly, Z-DEVD.fmk augmented gliotoxin-induced death of the Kupffer cells (Figure 5); a serine protease inhibitor TPCK also exacerbated gliotoxin-induced death of Kupffer cells. Because such effects of these inhibitors were unexpected, we determined if the inhibition diverts the death pathway to necrosis. Indeed, Z-DEVD.fmk exacerbated gliotoxin-induced LDH release and trypan blue uptake by Kupffer cells (Figure 6). Interestingly, TPCK pretreatment caused profound LDH release and trypan blue uptake by gliotoxin-treated cells even at 1h. At 6h, gliotoxin itself caused robust LDH release and trypan blue uptake by Kupffer cells.
Several mechanisms of eukaryotic cell regulation involve signal transduction via mitogen-activated protein kinases (MAPK) such as extracellular signal-regulated protein kinases 1/2 (ERK1/2), p38, and the c-Jun NH2-terminal kinase/stress-activated protein kinases (JNK/SAPK) (26–29). To explore the possibility of the role of MAPK in gliotoxin-induced death of Kupffer cells, activation of p-38, JNK and ERK1/2 was determined. Gliotoxin caused activation of all 3 MAPKs with earliest increase in phosphorylation of p38 occurring at 15 min and of JNK and ERK at 30 min (Figure 7A). However, pretreatment with specific MAPK inhibitors prior to the addition of 0.3 μM gliotoxin was found not to affect gliotoxin-induced death of Kupffer cells (Figure 7B).
Nuclear translocation of NFκB is important in cellular signaling associated with inflammatory stimuli and cell survival. We hypothesized that constitutive activation of NFκB may be inhibited by gliotoxin as in HSCs (1). However, gliotoxin actually stimulated nuclear translocation of NFκB as determined by immunohistochemistry (Figure 8A) as well as by gel shift analysis (Figure 8B). Even at 3h, when the cells were shrunk indicative of apoptosis, strong immunolabeling for nuclear NFκB was seen. In gel shift analysis the specificity of gliotoxin effect was demonstrated by its inhibition with MG132.
In Kupffer cells, gliotoxin did not stimulate SO or H2O2 production (Figure 9A, 9B), which are shown to cause DNA damage and death of several cell types (30–32). Moreover, antioxidants PDTC, Mn-TBAP, vitamin E and N-acetylcysteine did not affect gliotoxin-induced decrease in Kupffer cell viability (Figure 10).
Administration of gliotoxin to rats causes apoptosis of activated HSCs in the fibrotic liver, but not of quiescent HSCs in control liver (1). Considering the strong apoptosis/necrosis of gliotoxin-treated Kupffer cells in vitro, we investigated effect of gliotoxin on Kupffer cells in vivo. As shown in Figure 11A, glotoxin did not affect viability of hepatocytes in the control liver. Although few apoptotic hepatocytes were observed in DMSO-treated cirrhotic liver (not shown), morphological analysis revealed greater number of apoptotic hepatocytes in gliotoxin-treated cirrhotic liver (Figure 11D). TUNEL staining also showed several apoptotic hepatocytes in gliotoxin-treated fibrotic (Figure 11E, F), but not in the normal liver (Figures 11B, C). Consistent with a previous report (1), apoptotic HSCs were found in gliotoxin-treated fibrotic liver (Figure 11E). Several apoptotic Kupffer cells were also found to undergo apoptosis in the fibrotic liver after gliotoxin administration (Figure 11F). In control rats, gliotoxin caused minimal apoptosis of HSCs and Kupffer cells (<0.01%) (Figure 11B, C). The numbers of apoptotic hepatocytes, HSCs and Kupffer cells in gliotoxin-treated cirrhotic liver were 8 ± 2%, 45 ± 9% and 53 ± 12% respectively.
The above results demonstrated that gliotoxin causes death of Kupffer cells activated in vivo. Although activation of cultured Kupffer cells upon short-term treatment with endotoxin (33) did not sensitize them to gliotoxin up to 0.03 μM concentration, further reduction in their viability occurred at 0.3 and 3.0 μM concentration (Figure 12).
Here we report that gliotoxin causes death of cultured Kupffer cells via apoptosis and secondary necrosis. Previously, activated rat and human HSCs were found to undergo apoptosis and exclude trypan blue at 0.3–7.5 μM gliotoxin (1,2) but necrosis was observed at concentration above 7.5 μM (2). Hepatocytes are resistant to lower gliotoxin concentration but undergo necrosis at high concentration (50 μM). Apoptosis of Kupffer cells by lower concentration of gliotoxin (0.03–0.3 μM) indicates their greater sensitivity than HSCs and hepatocytes, and also peritoneal macrophages, which undergo apoptosis at higher (1 μM) concentration (9).
Gliotoxin-induced apoptosis of Kupffer cells involved classical apoptotic pathway of mitochondrial cytochrome c release, caspase-3 activation and ATP depletion. ATP is required for apoptosis but its depletion below a threshold concentration leads to secondary necrosis (34,35). Thus profound loss of ATP might be a mechanism of secondary necrosis of Kupffer cells at later times during gliotoxin treatment. Curiously, although gliotoxin stimulated caspace-3 activation, its inhibition exacerbated gliotoxoin-induced death of Kupffer cells. This finding contrasts the blockade of gliotoxin-induced nucleosomal ladder formation (but not morphological changes) in activated HSCs by caspase inhibitor Z-VAD.fmk (1). In addition, serine protease inhibition by TPCK also exacerbated, even more strongly, the effect of gliotoxin. These unusual and unexpected findings do not have definitive explanation. It is likely that inhibition of caspase-3 activation or serine proteases may promote another pathway (necrotic) of gliotoxin-induced death. Indeed, preincubation of the cells with Z-DEVD.fmk or TPCK caused strong early increase in LDH release and uptake of trypan blue upon gliotoxin challenge.
We investigated the involvement of NFκB, MAPKs and ROS in gliotoxin-induced death of Kupffer cells. Gliotoxin inhibits TNF-α-inducible nuclear binding of NFκB but not constitutively activated NFκB in activated HSCs and promotes apoptosis (1), probably by inhibiting TNF-α-induced activation of antiapoptotic genes (36–39). In contrast, PDTC, an inhibitor of NFκB activation (40,41) and an antioxidant (42,43), blocked gliotoxin-induced death of activated HSCs (1). Since NFκB promotes cell survival, its nuclear translocation by gliotoxin suggested adaptive stimulation of survival pathway, leading to the hypothesis that its inhibition should augment the toxic effect of gliotoxin. However, inability of PDTC to affect gliotoxin-induced death of Kupffer cells indicates that NFκB activation may not be related to this phenomenon or that the overwhelming activation of pro-death mechanisms overrides them. Interestingly, NFκB activation was shown to mediate elastase-induced apoptosis of Kupffer cells (23).
Most evidence indicates that p38 MAPK inhibits cell proliferation (44,45) and promotes apoptosis (46–48); its activation was also shown to be associated with GdCl3-induced apoptosis of Kupffer cells (49). Activation of JNK was found to be associated with apoptosis, presumably by phosphorylation of antiapoptotic Bcl2 members (29,48). There is also evidence for the role of ERK activation in apoptosis (50). However, inhibition of these MAPK pathways did not prevent or inhibit gliotoxin-induced Kupffer cell death suggesting that their activation may be related to other responses.
Gliotoxin-induced apoptosis of peritoneal macrophages was reported to involve ROS (8). Although 1.5–7.5 μM gliotoxin was found to cause apoptosis of culture-activated and immortalized HSCs respectively in the absence or at low level of oxidative stress (2, 24), higher concentrations induce oxidative stress and necrotic death of the immortalized HSCs (2). Necrosis of hepatocytes induced by high concentrations of gliotoxin (>10 μM) was prevented by antioxidants catalase and superoxide dismutase (24). However, gliotoxin did not stimulate SO and H2O2 generation in Kupffer cells, and several antioxidants did not prevent gliotoxin-induced death of these cells. Together, these observations indicate that the mechanism of gliotoxin-induced death of Kupffer cells(apoptotic and necrotic) is different from that for peritoneal macrophages and HSCs with respect to NFκB or ROS involvement, and is independent of MAPK activation.
Gliotoxin caused apoptosis of Kupffer cells as well as HSCs in the fibrotic rat liver; the effect was minimal in the normal liver. This may be because activated Kupffer cells and HSCs are more sensitive to gliotoxin, and the numbers of both cell types increase in cirrhosis (51). The profound effect of gliotoxin observed on Kupffer cells isolated from normal liver as compared to the minimal effect in vivo is interesting. A plausible explanation is that Kupffer cells may be activated in culture similar to those in vivo during hepatic injury (52,53). However, the activation does not seem to be complete as preincubation of the cells with endotoxin further sensitized Kupffer cells to gliotoxin-induced death. These observations contrast a recent study reporting comparable gliotoxin-induced apoptosis of Kupffer cells and HSCs in normal and cirrhotic livers (54). The reason for this discrepancy may be that liver slices were incubated for 8h with gliotoxin in vitro (54), while we and others (1,24) investigated the effect of gliotoxin in vivo. We also found significant gliotoxin-induced apoptosis of hepatocytes in the fibrotic liver but not the normal liver. Since hepatocytes undergo necrosis, but not apoptosis, upon treatment with high (>10 μM) concentration of gliotoxin in vitro (1,24), these results suggest that hepatocytes are either modified during fibrosis to become sensitive to pro-apoptotic effect of gliotoxin or that gliotoxin may release proapoptotic stimuli from nonparenchymal cells that then cause apoptosis of hepatocytes.
In summary, our results show that gliotoxin is a nonspecific pro-death agent for HSCs, Kupffer cells and hepatocytes in the fibrotic liver. The results also demonstrate a novel death-augmenting effect (via necrosis) of caspase-3 and serine protease inhibition on gliotoxin-induced apoptosis of Kupffer cells. Considering the importance of macrophages in resolution of fibrosis (55), removal of Kupffer cells might delay this process or even adversely affect the liver. Further research to investigate whether specific derivatives of gliotoxin cause apoptosis of activated HSCs selectively, excluding other liver cell types, is necessary for clinical application.
This work was supported by two grants from National Institutes of Health: RO1-DK54401 to CRG and CA76541 to DBS. We thank Dr. C. Thirunavukkarasu for assistance with apoptotic assays and Mr. Mark Ross for immunostaining of liver sections.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.