The apoptotic pathway triggered by the Fhit protein in lung cancer cell lines is not affected by Bcl-2 or Bcl-x(L) overexpression

Abstract

The expression of the tumour suppressor protein fragile histidine triad (Fhit) is often impaired in many human cancers and its restoration in Fhit-negative cancer cell lines suppresses tumorigenicity and induces apoptosis. Although the proapoptotic function of Fhit is well documented, little is known about its precise mechanism of action and further studies are needed in order to elucidate the putative therapeutic properties of this protein. To this end, we have engineered the lung cancer cell line NCI-H460 in order to express different molecules involved in the control of apoptotic pathways. Infection of these cells with an adenoviral vector transducing the Fhit gene (Ad-Fhit) revealed that complete protection from apoptosis was conferred by the inhibitor of caspases Cytokine response modifier A (CrmA) and by a dominant-negative form of the adapter protein Fas-associated death domain (FADD) and partial protection by a dominant-negative form of caspase-8, while cells over expressing mitochondrial mediators of the apoptotic response such as Bcl-2 or Bcl-x(L) that are resistant to treatment with cisplatin, remained highly susceptible to cell death triggered by Fhit gene transfer. In line to what was observed in H460 cells, Ad-Fhit efficacy was not affected by Bcl-2 overexpression also in two other lung cancer cell lines (A549 and Calu-1). Analysis of cytochrome c release also confirmed that in Bcl-2- or Bcl-x(L)-expressing cells apoptosis could be detected by terminal deoxynucleotidyl-transferase mediated dUTP nick-end labelling (TUNEL) assay before any evidence of mitochondrial membrane perturbation. In conclusion, our analysis indicates that the Fhit protein exerts its oncosuppressor activity through induction of an apoptotic mechanism that seems to be FADD dependent, caspase-8 mediated and independent from mitochondrial amplification.

Introduction

Inactivation of the fragile histidine triad (Fhit) gene is frequently observed in different types of human cancer. Lack or reduction of protein expression has been reported, among others, in a vast percentage of lung (Sozzi et al., 1998), oesophageal (Mori et al., 2000), bladder (Baffa et al., 2000), cervical (Connolly et al., 2000) and head and neck carcinomas (van Heerden et al., 1999).

Transfer of the Fhit gene in Fhit-negative cancer cell lines generally results in the induction of apoptosis and suppression of the tumorigenic phenotype (Ishii et al., 2001a). It has also been shown that the high susceptibility to chemical carcinogenesis in Fhit knockout mice can be prevented by treatment with Fhit-expressing adenoviral or adeno-associated vectors (Dumon et al., 2001a). Taken together, these observations strengthen the hypothesis of a role for the Fhit protein in tumour suppression, but its precise biological function remains largely unknown.

Activation of caspase-8, -9 and -3 has been described in Fhit-induced apoptosis (Dumon et al., 2001b; Ishii et al., 2001b; Roz et al., 2002) and Fhit re-expressing cells have been shown to be more susceptible to external apoptotic stimuli such as serum starvation or Fas treatment compared to the parental counterpart (Roz et al., 2002; Dumon et al., 2001b), but the hierarchy of the events leading to cell death has not been elucidated. Knowledge of the apoptotic pathway induced by different anticancer agents can potentially provide valuable information concerning resistance mechanisms and the possibility of synergies or antagonisms among different treatments. Previous studies on the non-small-cell lung cancer (NSCLC) NCI-H460 (H460) cell line concluded that most DNA-damaging agents currently used in lung cancer chemotherapy induce cell death by a caspase-8-mediated and mitochondria-controlled mechanism (Ferreira et al., 2000). Exceptions to these findings seem to be represented by microtubule-stabilizing agents such as paclitaxel (Huisman et al., 2002) or the novel cytotoxic agents discodermolide and epothilone B (Broker et al., 2002).

In order to investigate the mechanism of Fhit-mediated apoptosis, we have used different clones of H460 cells with modifications in genes involved in the apoptotic pathways and studied their sensitivity to treatment with an adenoviral vector expressing Fhit. Here, we show that the apoptotic mechanism induced by Fhit in susceptible cells triggers early activation of caspase-8, is strictly FADD dependent and does not seem to be controlled at the mitochondrial level, suggesting that this protein could be involved in the regulation of the cytoplasmic response to apoptotic stimuli.

Results

Fhit-induced apoptosis in H460 cells is FADD dependent and is not inhibited by Bcl-2 or Bcl-x(L) overexpression

Parental NSCLC H460 cells are Fhit negative and we and others have previously shown that reintroduction of Fhit expression in these cells induces apoptosis and suppresses their tumorigenic potential (Ji et al., 1999; Sard et al., 1999; Roz et al., 2002).

To investigate the mediators of Fhit-induced cell death, H460 cells were modified using plasmid constructs in order to express different molecules involved in the control of apoptosis and then infected with an adenoviral vector carrying the Fhit transgene (Ad5-Fhit). The establishment of the stable transfectants and their sensitivity to chemotherapy have been described previously (Ferreira et al., 2000): in the present study, we analysed H460 cells expressing inhibitors of the cytoplasmic apoptotic pathway (dominant-negative variants of caspase-8 and FADD) as well as cells expressing proteins involved in the mitochondrial control of apoptosis (Bcl-2 and Bcl-x(L)) and the viral inhibitor of caspase-1 and -8 Cytokine response modifier A (CrmA). Expression of the different transgenes was confirmed by Western blotting (Figure 1a) and transduction with control adenoviral vectors carrying reporter genes (GFP or lacZ) revealed that all clones were equally susceptible to adenoviral infection (data not shown). Efficient expression of exogenous Fhit in all clones was also confirmed by Western analysis (Figure 1b).

Figure 1
figure1

Expression of apoptosis modulators and efficacy of adenoviral-mediated Fhit gene transfer in H460 transfectants. (a) Western blot analysis of NCI-H460 cells transfected with different plasmids demonstrating overexpression of the indicated transgenes in selected clonal cell lines (transfectants, T) compared to empty vector transfected cells (controls, C). (b) Western blots showing high level of Fhit expression in different H460 clones 3 days after infection with an adenoviral vector carrying the Fhit cDNA (Ad5-Fhit) compared to uninfected parental H460 cells (H460-wt) that have undetectable level of Fhit protein (first lane)

Cells were transduced at a multiplicity of infection (MOI) of 10 and the characteristic morphological changes associated with apoptosis were observed under microscopic examination and quantified by cytofluorimetric analysis using the TUNEL assay, 3 and 5 days after infection. An adenoviral vector expressing lacZ (Ad-lacZ) was also used in all experiments to confirm that a similar number of cells were infected for all clones and as a control for possible cell death induced by viral infection. Under these experimental conditions, the infection rate is close to 90% of the cells and the apoptotic rate in parental (wt) H460 cells is usually around 40–50% after 5 days under serum starvation conditions. In this study, similar results were obtained with all the control clones (cells transfected with empty vectors), while complete protection from Fhit-induced apoptosis was observed in cells expressing the dominant-negative form of FADD (FADD-DN) and a partial protection was conferred by the dominant-negative form of caspase-8 (Figure 2). In the representative experiment shown, 43% of parental H460 cells, 51 and 55% of empty vectors control cells stained positive at the terminal deoxynucleotidyltransferase-mediated dUTP nick-end labeling (TUNEL) analysis while only 15% of caspases-8-DN- and 2% of FADD-DN-expressing cells displayed signs of apoptotic cell death. On the other hand, cells over expressing mitochondrial mediators of the apoptotic response, such as Bcl-2 or Bcl-x(L), remained highly sensitive to Ad5-Fhit treatment with 35 and 40% of apoptotic cells, respectively. As expected, cells expressing the strong inhibitor of caspases CrmA were resistant to apoptosis (1% TUNEL-positive cells) while a loss-of-function mutant form of the same gene did not confer any protection (72% apoptotic cells).

Figure 2
figure2

Fhit-induced apoptosis is FADD dependent and mitochondria independent in H460 cells. A total of 2 × 106 cells of the different H460 transfectants (indicated on the left) or of the parental H460 cells (H460-wt) were infected at an MOI of 10 with Ad5-Fhit (Fhit), Ad5-lacZ (lacZ) or mock infected (PBS) and left under serum starvation conditions for 5 days. The presence of apoptotic cells was analysed using the TUNEL assay and the percentage of apoptotic cells (indicated above the bar) was estimated by cytofluorimetric analysis. The same fluorescence threshold was used for PBS-, lacZ- or Fhit-treated cells. The open graph superimposed to the Fhit histogram represents the negative control reaction where no enzyme is included (TdT−). The results shown are representative of experiments performed at least three times in which the standard deviation was less than 15%

The different cell populations were also analysed for their sensitivity to the anticancer drug cisplatin. Confirming the functionality of the transgenes and the findings of our previous study (Ferreira et al., 2000), cisplatin-induced apoptosis was strongly inhibited in Bcl-2- and Bcl-x(L) expressing clones, while FADD-DN cells displayed a sensitivity similar to the parental cell line (data not shown). The different apoptotic stimuli employed in this study (Fhit gene transfer and cisplatin treatment) resulted therefore in an almost complementary pattern of toxicity, indicating different modalities of engagement of the apoptotic machinery. In summary, in H460 cells, cisplatin triggers apoptosis in a mitochondria-controlled manner while Fhit-induced apoptosis seems to be dependent only on cytoplasmic mediators of apoptotic response and independent from mitochondria.

In vivo studies of the tumorigenic potential of Fhit-transduced cells

H460 cells are highly tumorigenic when injected subcutaneously in the nude mouse but this potential is strongly reduced after Ad5-Fhit transduction (Ji et al., 1999). In a previous study, we have shown that in vitro transduced cells completely lost the potential to form tumours, with 20 out of 20 animals remaining tumour free 3 months after inoculation, while control animals (that received either untransduced cells or Ad-lacZ-transduced cells) all developed large tumours within 4 weeks (Roz et al., 2002). To test whether interference with the cytoplasmic apoptotic signalling could also interfere with the in vivo tumour suppressor activity of Fhit, H460 cells expressing FADD-DN and CrmA (that are protected from Fhit-induced apoptosis) were transduced in vitro with Ad5-Fhit and Ad5-lacZ and then injected subcutaneously into the right flank of nu/nu mice (Figure 3). For FADD-DN-expressing cells, the mean tumour volume 4 weeks after inoculation was 515±39 and 534±63 mm3 (P=0.77) for Ad5-Fhit- and Ad5-lacZ-infected cells, respectively, showing that inhibition of FADD signalling antagonizes the oncosuppressive properties of Fhit. The same was also true for CrmA-expressing cells with tumour volumes reaching 267±16 and 365±51 mm3 (P=0.55). As expected, suppression of the tumorigenic potential was instead observed when parental H460 cells (untransfected) were treated with Ad5-Fhit. Antagonizing caspase-8 activity either through its interaction with the adapter molecule FADD or by direct inhibition with CrmA therefore seems to abrogate the in vivo oncosuppressive properties of Fhit, suggesting a central role for caspase-8 signalling in the biochemical pathway stimulated by Fhit reintroduction.

Figure 3
figure3

In vivo tumorigenicity of H460 transfectants treated with Ad5-Fhit. H460 parental cells (H460-wt) and its transfectants expressing CrmA or a dominant-negative form of FADD (FADD-DN) were transduced in vitro with Ad5-Fhit, Ad5-lacZ or mock infected (PBS) and after 2 days inoculated subcutaneously into nude mice (at least four animals per group). The tumorigenic potential of parental H460 cells is abrogated when Fhit expression is restored, while CrmA- and FADD-DN-expressing cells retain their full tumorigenicity

Activation of caspase-8 in response to Ad5-Fhit treatment

During apoptosis induced by exogenous Fhit expression in Fhit-negative cells, caspase-8 activation has been consistently detected in lung (Roz et al., 2002), pancreatic (Dumon et al., 2001b) and oesophageal (Ishii et al., 2001b) cancer cell lines. Also in this study, Western blot analysis of Ad5-Fhit-infected cells revealed a correlation with caspase-8 activity and cell death, with the appearance of activated forms of caspase-8 in the cell lysates from susceptible clones. To gain insights into the ordering of the molecular events leading to the execution of the Fhit-induced apoptotic programme, caspase-8 activation was also analysed 3 days after adenoviral infection, a time point when there are only a small percentage of apoptotic cells (10–15%). As expected, caspase-8 could not be efficiently activated in the FADD-DN-expressing cells, while activation was detected in Bcl-2- and Bcl-x(L)-transfected cells after Ad5-Fhit infection (Figure 4a). Interestingly, in drug induced apoptosis, efficient activation of caspase-8 in H460 cells has been reported to be dependent on mitochondrial amplification of the apoptotic signal (Ferreira et al., 2000), while our results show that, following Fhit treatment, caspase-8 might act as the initiator signal in a pathway that is independent from mitochondria. Fhit reintroduction seems therefore to facilitate the execution of a cytoplasmic apoptotic programme.

Figure 4
figure4

Activation of the cytoplasmic apoptotic pathway in Ad5-Fhit-infected cells. (a) Caspase-8 activation. H460 transfectants were infected with adenoviral vectors expressing Fhit or the control protein lacZ and analysed for the activation of caspase-8 in the early phases of apoptosis induction (3 days after infection). In Fhit re-expressing cells there is evidence of cleavage of the inactive procaspase-8 (p55) into the active forms (p43, p41 and the p18 subunit). (b) Bid cleavage. Following adenoviral infection, H460 clones were analysed for the status of the Bid protein. In accordance with the observed caspase-8 activation, reduction in the quantity of the full-length protein indicates cleavage of Bid

Bid cleavage in Ad5-Fhit-infected cells

The cytoplasmic and mitochondrial apoptotic pathways are not completely separated at the cellular level. One of the best understood connections between the two is represented by the cleavage of the proapoptotic Bid protein by activated caspase-8 (Luo et al., 1998; Li et al., 1998). Truncated Bid translocates into the mitochondria and mediates the release of cytochrome c, thereby triggering the activation of caspase-9. In some cell types (type II cells), this mitochondrial amplification of the death signal is required for irreversible activation of the apoptotic cascade following a cytoplasmic stimulus (Scaffidi et al., 1999) and in these cells overexpression of Bcl-2 can antagonize death-receptor-induced apoptosis because its mitochondrial protecting activity creates a block of the amplification loop. Of interest is the fact that, in H460 cells, the toxic effects of cisplatin were shown to be dependent on caspase-8 activation, but in a mitochondria-controlled manner (Ferreira et al., 2000).

To study the crosstalk between the cytoplasmic and mitochondrial pathways in our model, we investigated Bid cleavage in cells undergoing Fhit-induced apoptosis and found that following caspase-8 activation Bid was cleaved efficiently in all apoptosis-susceptible clones (Figure 4b). As expected, overexpression of the dominant-negative forms of caspase-8 and FADD interfered with this process (data not shown). These findings indicate that, also in this model, in H460 cells the interaction between the different apoptotic pathways is normally used to amplify the apoptotic response.

Cytochrome c release is not required in the early phases of Fhit-induced apoptosis

The sensitivity to apoptosis of Bcl-2 and Bcl-x(L) overexpressing H460 cells could be explained by means of different mechanisms: the Fhit protein could participate in the stimulation of an apoptotic cascade acting principally at the cytoplasmic level and therefore unaffected by the mitochondrial protection conferred by the antiapoptotic Bcl-2 family members or, alternatively, Fhit could exert a strong effect on the connection between cytoplasmic and mitochondrial pathways ultimately leading to mitochondrial dysfunction. Cytochrome c release in the cytoplasm is one of the earliest events involved in mitochondrial-controlled apoptosis and results in the activation of caspase-9 after aggregation of the Apaf-1-containing apoptosome complex (Zou et al., 1999). Although there is still debate on the precise mechanism of action, it is widely believed that the antiapoptotic potential of Bcl-2 and Bcl-x(L) is linked to their ability to prevent cytochrome c release after mitochondrial stimulation (reviewed in Kaufmann and Hengartner, 2001). We have analysed the kinetics of the appearance of cytochrome c in the cytoplasm in different phases of Ad5-Fhit-induced apoptosis to understand how the system is regulated in this model. In control H460 cells (parental or empty vectors clones) infected with Ad5-Fhit, release of cytochrome c could be easily detected as soon as the apoptotic mechanism was active, but in Bcl-2- and Bcl-x(L)-expressing cells, apoptosis could be detected in the absence of cytochrome c release (Figure 5a and b). Interestingly, in these cells Bid cleavage could be demonstrated (Figure 4b), indicating that caspase-8 activation would normally signal also through the mitochondrial amplification loop, but that this is not required for Fhit-induced apoptosis. When the same cells were treated with cisplatin, functional activity of Bcl-2 and Bcl-x(L) was also confirmed inasmuch as their expression was able to antagonize the action of the drug at the mitochondrial level as indicated by the lack of cytochrome c release in the cytoplasm after drug treatment (Figure 5c). These results suggest therefore that in H460 cells the apoptotic mechanism triggered by Fhit is usually amplified at the mitochondrial level through Bid cleavage by activated caspase-8, but that this is not a necessary step and that the early phases of this cell death process might be independent of caspase-9 activation.

Figure 5
figure5

Cytochrome c release and apoptosis analysis. (a) Cytochrome c release after adenoviral infection. The indicated H460 clones or the parental H460 line (H460-wt) were infected with Ad5-Fhit (Fhit) or Ad5-lacZ (lacZ) and analysed for cytochrome c release in the cytosol in the early phases of apoptosis induction (3 days postinfection). Clones susceptible to Fhit-induced apoptosis generally showed evidence of cytochrome c release (upper panel), while there was no release in the resistant FADD-DN clone or in the susceptible Bcl-2 and Bcl-x(L) clones (lower panel). This indicates functional protection of the mitochondria in the Bcl-2 and Bcl-x(L) clones. Membranes were also hybridized with an anti-actin antibody to confirm equivalent loading and with an anti-COX IV antibody to verify the purity of the cytosolic extracts. TL=total lysate. (b) TUNEL analysis of apoptosis. The same cells as in (a) were analysed for the presence of apoptotic cells. The reported values are the result of the subtraction of the percentage of apoptotic cells measured in Ad5-Fhit-treated cells and the percentage measured in the control cells (Ad5-lacZ treated). Interestingly, apoptosis is detectable in Bcl-2 and Bcl-x(L) clones even in the absence of cytochrome c release (panel a). (c) Cytochrome c release after cisplatin treatment. To confirm susceptibility or resistance to different apoptotic stimuli, the indicated H460 clones were cultured for 48 h in the absence (−) of presence (+) of 10 μ M cisplatin and then the cytosolic extracts were analysed for the presence of cytochrome c. As expected, an increase in the quantity of cytochrome c in the cytoplasm is present in the FADD-DN clone, but not in the cisplatin resistant clones overexpressing Bcl-2 or Bcl-x(L), confirming that the two latter clones are protected from a mitochondrial apoptotic stimulus

Bcl-2 overexpression does not protect NSCLC cell lines from Fhit-induced apoptosis

To generalize the data obtained with Ad-Fhit in H460 cells (large cell carcinoma) in the key aspect of apoptosis induction even in the presence of high levels of Bcl-2, we tested the effect of Bcl-2 expression in cell lines derived from other histological subtypes of NSCLC: Calu-1 (squamous cell carcinoma) and A549 (adenocarcinomas). Both lines are Fhit negative and we have previously demonstrated that they are susceptible to Fhit-induced apoptosis. Independent Bcl-2 overexpressing clones were generated by stable transfection of both cell lines with the same plasmid construct used for H460 cells. All the clones were tested for their sensitivity to cisplatin treatment to confirm functionality of the Bcl-2 transgene. Ad-Fhit infection was performed using an MOI of 10 (that results in transduction of more than 90% of the cells) and apoptosis assessed 5 days after infection. These experimental conditions have similar toxicity for both treatments inducing apoptosis in 30–40% of the cells in the parental lines.

For A549 cells four clones were studied (Clones 2, 6, 7, 9) one of which (Clone 6) lost expression of the Bcl-2 transgene during culture and therefore represented an additional negative control. As observed for H460 cells, the three A549 clones with high Bcl-2 levels were protected from cisplatin toxicity (compared to the parental cell line or to the nonexpressing clone), while they remained sensitive to Fhit-induced apoptosis (Figure 6a). The mean sensitivity of the Bcl-2-positive clones to cisplatin treatment was 23% that of the parental cell line (range 12–33%), confirming the involvement of the mitochondrial pathway in mediating chemotherapy effects in NSCLC. On the other hand, sensitivity to Ad-Fhit remained 75% that of controls (range 70–81%), suggesting that the apoptotic response triggered by Fhit re-expression is not inhibitable by mitochondrial protection.

Figure 6
figure6

Sensitivity of A549 and Calu-1 Bcl-2 overexpressing clones to cisplatin and Ad-Fhit treatment. The presence of apoptotic cells was analysed using the TUNEL assay and the percentage of apoptotic cells was estimated by cytofluorimetric analysis. Results are expressed as the percentage of apoptotic cells relative to the parental cell line and represent the mean of four independent analyses. Western blot analysis at the bottom of the graphs demonstrates the presence of high levels of Bcl-2 transgene in the selected clonal cell line (A549 clone 6 lost Bcl-2 expression during culture and maintains sensitivity to cisplatin). (a) Sensitivity of A549 Bcl-2 overexpressing clones to cisplatin treatment (upper panel) or to Ad-Fhit infection (lower panel). The clones were treated with cisplatin 32 μ M and incubated for 48 h in normal culture conditions or transduced at an MOI of 10 with Ad-Fhit and left under serum starvation conditions for 5 days. The mean sensitivity of Bcl-2-expressing clones to cisplatin: 23%; Ad-Fhit: 75%. (b) Results for Calu-1 Bcl-2 over expressing clones treated with cisplatin 64 μ M for 48 h (upper panel) or infected at an MOI of 10 with Ad-Fhit and kept under serum starvation conditions for 5 days (lower panel). The mean sensitivity of Bcl-2 expressing clones to cisplatin: 35%; Ad-Fhit: 116%

Similar results were also obtained with the analysis of five different Bcl-2 positive clones from Calu-1 cells (Figure 6b): cisplatin treatment efficacy was consistently reduced (35% that of controls, range 19–46%) and conversely Ad-Fhit-induced apoptosis was not affected (mean apoptosis of the five clones relative to controls was 116%, range 97–139%).

The effects of Bcl-2 overexpression were also evaluated in transient transfections and on early passage bulk populations of stable transfectants of both cell lines with similar results (data not shown).

In conclusion, for the three NSCLC cell lines tested (H460, A549 and Calu-1), Bcl-2 overexpression did not seem to confer protection from the proapoptotic effect of Ad-Fhit treatment, although it was sufficient to induce considerable resistance to cisplatin toxicity.

Discussion

The molecular basis underlying cellular responses to anticancer treatment is currently investigated with growing interest in order to ameliorate the understanding of the rationale of some therapeutic protocols, to devise better strategies of intervention and, in certain cases, to provide ‘tailored’ therapies (Johnstone et al., 2002). Much attention has been devoted to the study of programmed cell death (or apoptosis) because it has been shown that the apoptotic machinery is often stimulated by drugs used in chemotherapy and that defects in its intracellular signalling could provide the basis for tumour resistance to therapy (Makin and Dive, 2001). Apoptosis is a tightly regulated physiological programme characterized by specific biochemical and morphological changes and executed by cystein aspartyl-specific proteases (caspases) that cleave specific substrates following activation by an apoptotic stimulus (Hengartner, 2000). Two major apoptotic pathways have so far been delineated: the cytoplasmic or death receptor pathway (headed by caspase-8) and the mitochondrial pathway (headed by caspase-9). Alterations in either the cytoplasmic (‘extrinsic’) or mitochondrial (‘intrinsic’) apoptotic pathways have been found in different forms of human cancers, indicating a common mechanism for tumour development and a possible hurdle for therapies aiming at exploiting these systems to kill cancer cells (Johnstone et al., 2002).

Lung cancer is often resistant to standard chemotherapy and survival rates for late-stage disease are very low: only 10–15% of patients with stage II–IV lung cancer survive over 5 years and therefore new systems for early diagnosis and novel therapies are needed to reduce mortality from this disease (Peto et al., 1996). We have previously reported that an adenoviral vector carrying a Fhit transgene can effectively kill lung cancer cells in vitro and suggested its possible use as a novel therapeutic tool (Roz et al., 2002). In order to better understand its mechanism of action and its links with cellular mechanisms that might be altered in cancer cells, we have studied its efficacy in the presence of alterations of the cellular apoptotic network. Cells undergoing Fhit-mediated apoptosis show activation of caspase-8, especially under low serum conditions, and restoration of Fhit expression in Fhit-negative lung cancer cells induces strong susceptibility to Fas-mediated apoptosis, suggesting a role for this protein in facilitating cytoplasmic responses to external apoptotic stimuli (Roz et al., 2002). In the present study, we have observed that in H460 lung cancer cells, overexpression of the viral inhibitor of caspases-1 and -8 CrmA or of a dominant-negative form of FADD protects from Fhit treatment, pointing to a mechanism connected with caspase activation originating at the cytoplasmic level. Consistent with this finding, caspase-8 inhibition with a dominant-negative construct also protected cells from apoptosis, although protection was not complete, indicating a possible role for other caspases in the initiation of this process. On the other hand, cells overexpressing Bcl-2 or Bcl-x(L) that are resistant to cisplatin treatment remain highly sensitive to Fhit-induced apoptosis. Analysis of mediators of Fhit activity revealed early activation of caspase-8 with subsequent cleavage of Bid, suggesting that in H460 cells the cytoplasmic and mitochondrial pathways are usually connected to amplify the apoptotic response headed by caspase-8. When the execution of the mitochondrial apoptotic programme was blocked by Bcl-2 or Bcl-x(L) overexpression, however, apoptosis could still be detected, indicating that this amplification step is not therefore a necessary requisite for Fhit activity. It is thus possible to hypothesize a treatment with Fhit-expressing viral vectors also in Bcl-2 overexpressing tumours. Consistent with this hypothesis also other cell lines from NSCLC (A549 and Calu-1), modified to express high levels of Bcl-2 that could protect them from cisplatin treatment, remained susceptible to Ad-Fhit-induced apoptosis.

To study the apoptotic pathway triggered by Fhit, we used adenoviral-mediated gene transfer that results in high levels of expression. Using an inducible expression system, we recently demonstrated that Fhit expression had a precise dose-dependent effect in reducing the growth of Calu-1 lung cancer cells and that a threshold had to be reached before effects on proliferation could be seen (Cavazzoni et al., 2004). This level of expression was already higher than that observed in normal bronchial epithelial cells (NHBE). Although the levels of expression needed to obtain an apoptotic effect are even higher, it has to be considered that established cancer cell lines often harbour many additional genetic changes that may impair the efficacy of the reintroduction of a single gene, especially in highly deregulated pathways such as cell growth and apoptosis. The phenotypic effects caused by Fhit reintroduction in Fhit-negative cell lines seem therefore dose dependent with growth inhibitory effects through action on the cell cycle observed at lower levels of expression, while prolonged and sustained expression is required to restore sensitivity to apoptosis.

Previous data on apoptosis triggered by the Fhit protein showed involvement of both cytoplasmic and mitochondrial pathways (Dumon et al., 2001b; Ishii et al., 2001b). These findings are consistent with the fact that in those studies the analysis was performed on cells in the late stages of the apoptotic process, when the connection of the different apoptotic pathways leads to evidence of activation of multiple caspases (both initiators such as caspase-8 or -9, and effectors such as caspase-3) and cleavage of several target substrates (e.g. PARP, Bid).

In this study, we have shown that, even in a cell type that usually relies on mitochondrial amplification of death signals (type II cells), Fhit-induced apoptosis can be effective in the presence of a mitochondrial blockade, with early signs of cell death preceding the release of cytochrome c into the cytoplasm.

In conclusion, the apoptotic mechanism induced by Fhit in susceptible cells seems to be mediated by caspase-8 activation, can be inhibited by expression of negative regulators of the cytoplasmic apoptotic pathway, such as CrmA or a dominant-negative form of FADD, and is not dependent on mitochondrial mediators of apoptosis. The elucidation of the mechanism of action of a protein investigated as a novel therapeutic tool can provide valuable information for the accurate planning of future clinical applications.

Materials and methods

Expression vectors

The cDNAs encoding CrmA wild type, loss of function mutant CrmA (T291R), Bcl-2 and Bcl-x(L), subcloned into the expression vector pEF FLAGpGKpuro, have been described previously (Ekert et al., 1999; Huang et al., 1997). pcDNA3 vectors encoding active site mutant caspase-8 (Flice-DN or caspase-8-DN) and FADD-DN have also been described (Chinnaiyan et al., 1995, 1996).

Cell lines and transfections

The human NSCLC cell lines NCI-H460 (H460), A549 and Calu-1 used for the study were purchased from ATCC. Cells were cultured in RPMI 1640 supplemented with 10% heat-inactivated FCS (Bio-Whittaker Europe, Belgium). Cells from exponentially growing cultures were used in all of the experiments. For the generation of stable transfectants, H460 cells were transfected with 10 μg of plasmid DNA, using Superfect reagent (Qiagen) according to the manufacturer's protocol. After 24 h, cells from each transfection were split into five separate culture dishes to ensure that independent lines were established. Selection was made using increasing concentrations of Puromycin (Sigma, Saint Louis, MO, USA) ranging from 1 to 2 μg/ml or 200 to 400 μg/ml Geneticin (Invitrogen, Carlsbad, CA, USA), depending on the transfected plasmid used. Transfections of A549 and Calu-1 cells were performed using Lipofectamine™ with Plus™ Reagent (Invitrogen) using standard techniques and clones were selected in 1 μg/ml of Puromycin. Independent clones were collected and allowed to grow in six-well plates. Clones were tested for the expression of the constructs by Western blotting and were selected for use in the desired experiments. The sensitivity to chemotherapy of the different cells has been reported previously (Ferreira et al., 2000) and independent clones expressing the same construct gave comparable results: representative clones already well characterized were therefore selected for the present study. Cells were maintained either in 1–2 μg/ml puromicyn (Bcl-2, Bcl-X(L), CrmA and mut-CrmA) or 200–400 μg/ml geneticin (caspase-8-DN, FADD-DN) and the appropriate empty vector controls (‘puro’ or ‘neo’, respectively) were used in all the experiments.

Electrophoresis and Western blotting

Western blot analysis was performed using standard techniques. From each sample, 25 μg of proteins was separated on 8–15% SDS–PAGE and electroblotted onto polyvinylidene difluoride membranes (Amersham Biosciences, UK). Subsequently, membranes were incubated for 1 h at room temperature in a solution of PBS supplemented with 5% nonfat dry milk. For immunodetection, the following antibodies were used: anti-Fhit polyclonal antibody 71-9000 (Zymed, San Francisco, CA, USA; 1 : 1000); anti-FADD monoclonal antibody (mAb) clone 1 (Transduction Laboratories, Lexington, KY, USA; 1 : 250); anti-caspase-8 mAb antibody C-15 (gift from Professor P Krammer, Deutsches Krebsforschungszentrum, Heidelberg, Germany; 1 : 5); anti-Flag M2 mAb and anti-actin A2066 (Sigma, Saint Louis, MO, USA; 1 : 1000); anti-Bcl-x(L) polyclonal antibody (1 : 1000) and anti-cytochrome c mAb clone 7H8.2C12 (Pharmingen, San Diego, CA, USA; 1 : 100); anti-Bcl-2 mAb clone 120 (Dako, Santa Barbara, CA, USA; 1 : 250); anti-Bid polyclonal antibody (Cell Signaling, Beverly, MA, USA; 1 : 500) and anti-COX IV mAb clone 20E8 (Molecular Probes, Eugene, OR, USA; 1 : 500). After overnight incubation at 4°C with the primary antibody, membranes were washed in TBST (10 mM Tris-HCl (pH 8.0), 0.15 M NaCl and 0.05% Tween-20), followed by horse-radish peroxidase-conjugated goat-antimouse or goat-anti-rabbit antibody. ECL (Amersham Biosciences, UK) was used for detection. Protein loading equivalence was assessed by the expression of β-actin.

Adenoviral vectors

The preparation of recombinant, E1 and E3 deleted, adenoviral vectors expressing the Fhit protein or the control proteins lacZ and GFP was performed in foetal kidney 293 cells (Microbix Biosystems Inc.,. Toronto, Canada) using standard techniques as described (Roz et al., 2002). For Ad5-Fhit a 707 bp fragment of Fhit cDNA was amplified by RT–PCR from human placental cDNA and cloned into an adenoviral shuttle vector purchased from Qbiogene, Montreal, Canada (pQBI-AdCMV5). The absence of replication-competent viruses was confirmed by PCR and plaque assays on nonpermissive cell lines and vectors were expanded by sequential rounds of infection on 293 cells and purified to a titre of approximately 109 PFU/ml.

TUNEL assay of adenoviral vector or cisplatin-treated cells

Analysis of apoptosis was performed by using TUNEL. Cells were plated at 2 × 106 cells/100 mm Petri dish and the following day adenoviral vectors were added at a minimum MOI of 10 in 4 ml of culture medium without serum. After 6 h, new medium was added and the cells were cultured 3 to 5 days before analysis. For cisplatin treatment, cells were plated in 100 mm Petri dishes, allowed to adhere before the addition of cisplatin diluted in culture medium at a final concentration of 10 μ M (H460), 32 μ M (A549) or 64 μ M (Calu-1) and harvested 48 h after treatment. FACScalibur (BD Biosciences, San Jose, CA, USA) analysis of apoptosis was performed as follows: 2 × 106 cells per sample were fixed with 2% paraformaldehyde in PBS (10 min on ice), washed three times with TBS (50 mM Tris-HCl in saline solution, pH 7.5), permeabilized with ice-cold acetone (1 min on ice) and washed twice in TBS and once in distilled water. Staining was performed by incubating cells for 1 h at 37°C in 25 μl (final volume) of TUNEL reaction mixture (In Situ Cell Death Detection Kit, Fluorescein; Roche). Cells with fragmented DNA appeared positive at the analysis performed on FACScalibur using CELLQuest software (BD). Apoptotic cells were defined by establishing fluorescence thresholds based on the negative controls represented by cells mock infected, infected with control adenovirus (lacZ) and by cells treated with TUNEL reaction mixture without the enzyme.

In vivo studies

All animal studies were performed following institutional guidelines. Clones of H460 cells were transduced in vitro at an MOI of 10 with Ad5-Fhit, Ad5-lacZ or mock infected. At 2 or 4 days after transduction, cells were counted and 1 × 106 viable cells were injected into the right flank of 6-week-old female nude mice, four mice per group in two separate experiments that gave similar results. Tumour growth was monitored three times a week with a linear caliper and volume estimated using the equation V=(a × b2)/2, where a is the larger dimension and b the perpendicular diameter.

Preparation of cytosolic extracts and cytochrome c release analysis

To study the release of cytochrome c in the cytosol, 2 × 106 cells were pelleted in a microcentrifuge tube and immediately resuspended in 300 μl of buffer A (50 mM Tris, 1 mM EGTA, 5 mM 2-mercaptoethanol, 0.2% BSA, 10 mM KH2PO4, pH 7.6, 0.4 M sucrose). Cells were kept on ice for 30 min and then disrupted by 30 passages through a 26-gauge needle. The extract was then clarified by centrifugation at 15 000 r.p.m. at 4°C for 30 min and 200 μl of the supernatant was transferred to a new tube taking care not to disturb the pellet. Cytosolic extracts were separated on a 14% SDS–PAGE gel and after Western blotting the membranes were incubated with the anti-cytochrome c antibody. As a control for the purity of the extracts, the same membranes were always hybridized with anti-COX IV antibody to exclude the presence of mitochondrial proteins.

Abbreviations

Fhit:

fragile histidine triad

NSLC:

non-small-cell lung cancer

FADD:

Fas-associated death domain

CrmA:

cytokine response modifier A

MOI:

multiplicity of infection

TUNEL:

terminal deoxynucleotidyltransferase-mediated dUTP nick-end labelling

References

  1. Baffa R, Gomella LG, Vecchione A, Bassi P, Mimori K, Sedor J, Calviello CM, Gardiman M, Minimo C, Strup SE, McCue PA, Kovatich AJ, Pagano F, Huebner K and Croce CM . (2000). Am. J. Pathol., 156, 419–424.

  2. Broker LE, Huisman C, Ferreira CG, Rodriguez JA, Kruyt FA and Giaccone G . (2002). Cancer Res., 62, 4081–4088.

  3. Cavazzoni A, Petronini PG, Galetti M, Roz L, Andriani F, Carbognani P, Rusca M, Fumarola C, Alfieri R and Sozzi G . (2004). Oncogene, Sep 13 (Epub ahead of print).

  4. Chinnaiyan AM, O'Rourke K, Tewari M and Dixit VM . (1995). Cell, 81, 505–512.

  5. Chinnaiyan AM, O'Rourke K, Yu GL, Lyons RH, Garg M, Duan DR, Xing L, Gentz R, Ni J and Dixit VM . (1996). Science, 274, 990–992.

  6. Connolly DC, Greenspan DL, Wu R, Ren X, Dunn RL, Shah KV, Jones RW, Bosch FX, Munoz N and Cho KR . (2000). Clin. Cancer Res., 6, 3505–3510.

  7. Dumon KR, Ishii H, Fong LY, Zanesi N, Fidanza V, Mancini R, Vecchione A, Baffa R, Trapasso F, During MJ, Huebner K and Croce CM . (2001a). Proc. Natl. Acad. Sci. USA, 98, 3346–3351.

  8. Dumon KR, Ishii H, Vecchione A, Trapasso F, Baldassarre G, Chakrani F, Druck T, Rosato EF, Williams NN, Baffa R, During MJ, Huebner K and Croce CM . (2001b). Cancer Res., 61, 4827–4836.

  9. Ekert PG, Silke J and Vaux DL . (1999). EMBO J., 18, 330–338.

  10. Ferreira CG, Span SW, Peters GJ, Kruyt FA and Giaccone G . (2000). Cancer Res., 60, 7133–7141.

  11. Hengartner MO . (2000). Nature, 407, 770–776.

  12. Huang DC, Cory S and Strasser A . (1997). Oncogene, 14, 405–414.

  13. Huisman C, Ferreira CG, Broker LE, Rodriguez JA, Smit EF, Postmus PE, Kruyt FA and Giaccone G . (2002). Clin. Cancer Res., 8, 596–606.

  14. Ishii H, Dumon KR, Vecchione A, Fong LY, Baffa R, Huebner K and Croce CM . (2001a). JAMA, 286, 2441–2449.

  15. Ishii H, Dumon KR, Vecchione A, Trapasso F, Mimori K, Alder H, Mori M, Sozzi G, Baffa R, Huebner K and Croce CM . (2001b). Cancer Res., 61, 1578–1584.

  16. Ji L, Fang B, Yen N, Fong K, Minna JD and Roth JA . (1999). Cancer Res., 59, 3333–3339.

  17. Johnstone RW, Ruefli AA and Lowe SW . (2002). Cell, 108, 153–164.

  18. Kaufmann SH and Hengartner MO . (2001). Trends Cell Biol., 11, 526–534.

  19. Li H, Zhu H, Xu CJ and Yuan J . (1998). Cell, 94, 491–501.

  20. Luo X, Budihardjo I, Zou H, Slaughter C and Wang X . (1998). Cell, 94, 481–490.

  21. Makin G and Dive C . (2001). Trends Cell Biol., 11, S22–S26.

  22. Mori M, Mimori K, Shiraishi T, Alder H, Inoue H, Tanaka Y, Sugimachi K, Huebner K and Croce CM . (2000). Cancer Res., 60, 1177–1182.

  23. Peto R, Lopez AD, Boreham J, Thun M, Heath Jr C and Doll R . (1996). Br. Med Bull., 52, 12–21.

  24. Roz L, Gramegna M, Ishii H, Croce CM and Sozzi G . (2002). Proc. Natl. Acad. Sci. USA, 99, 3615–3620.

  25. Sard L, Accornero P, Tornielli S, Delia D, Bunone G, Campiglio M, Colombo MP, Gramegna M, Croce CM, Pierotti MA and Sozzi G . (1999). Proc. Natl. Acad. Sci. USA, 96, 8489–8492.

  26. Scaffidi C, Schmitz I, Zha J, Korsmeyer SJ, Krammer PH and Peter ME . (1999). J. Biol. Chem., 274, 22532–22538.

  27. Sozzi G, Pastorino U, Moiraghi L, Tagliabue E, Pezzella F, Ghirelli C, Tornielli S, Sard L, Huebner K, Pierotti MA, Croce CM and Pilotti S . (1998). Cancer Res., 58, 5032–5037.

  28. van Heerden WF, Swart TJ, van Heerden MB, van Rensburg EJ, Engelbrecht S, Dreyer L and Huebner K . (1999). J. Oral. Pathol. Med., 28, 433–437.

  29. Zou H, Li Y, Liu X and Wang X . (1999). J Biol. Chem., 274, 11549–11556.

Download references

Acknowledgements

We thank Professor P Krammer (Deutsches Krebsforschungszentrum, Heidelberg, Germany) for the anti-caspase-8 antibody, Drs Massimo Zeviani and Valeria Tiranti (Istituto Nazionale Neurologico ‘C Besta’, Milan, Italy) for the anti-COX IV antibody and Drs Delia Mezzanzanica and Paola Perego (Istituto Nazionale Tumori, Milan, Italy) for helpful discussions and for providing the anti-cytochrome c and anti-Bcl-X(L) antibodies, respectively. We are also grateful to Dr Hideshi Ishii for technical advice and Professor Carlo M Croce (both at Kimmel Cancer Center, Jefferson Medical College, Philadelphia, PA, USA) for comments and suggestions. This work was partially supported by grants from Associazione Italiana per la Ricerca sul Cancro (AIRC) and from Ministero Italiano della Sanita' to GS. LR and FA are recipients of fellowships from AIRC and Ministero Italiano della Sanita'.

Author information

Correspondence to Luca Roz.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Roz, L., Andriani, F., Ferreira, C. et al. The apoptotic pathway triggered by the Fhit protein in lung cancer cell lines is not affected by Bcl-2 or Bcl-x(L) overexpression. Oncogene 23, 9102–9110 (2004). https://doi.org/10.1038/sj.onc.1208142

Download citation

Keywords

  • Fhit
  • lung cancer
  • apoptosis control

Further reading