Background
There are two major apoptotic signaling pathways: the intrinsic mitochondria-mediated pathway and the extrinsic death receptor-induced pathway, and these pathways are linked by the truncated proapoptotic protein Bid [
1]. The extrinsic apoptotic pathway is negatively regulated primarily by the cellular FLICE-inhibitory protein (c-FLIP), including both long (FLIP
L) and short (FLIP
S) forms, through inhibition of caspase-8 activation, whereas the intrinsic apoptotic pathway is negatively regulated by multiple proteins including survivin [
2]. The tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) binds to its receptors: death receptor 4 (DR4, also named TRAIL-R1) and death receptor 5 (DR5, also named TRAIL-R2) to activate the extrinsic apoptotic pathway [
3]. Recently TRAIL has received much attention because it preferentially induces apoptosis in transformed or malignant cells, demonstrating potential as a tumor-selective apoptosis-inducing cytokine for cancer therapy [
3]. Currently TRAIL is being tested in phase I clinical trials. However, some cancer cells are resistant to TRAIL [
4,
5]. Thus agents that can sensitize cancer cells to TRAIL are useful to enhance the efficacy of TRAIL-based cancer therapy or to overcome TRAIL-resistance [
6].
Protein geranylgeranyltransferase type I (GGTase I) is responsible for the posttranslational modification of CAAX motif-containing proteins such as K-Ras, N-Ras, RhoA, RhoC, Rac1, RalA, and Cdc42, which are often involved in cell transformation, tumor development and metastasis [
7,
8]. Thus, GGTase I has been considered a good cancer therapeutic target and thus great efforts have been made to develop GGTase I inhibitors as anticancer drugs [
7,
9]. Earlier studies using the GGTase I inhibitor, GGTI-287, showed K-Ras, a substrate for both GGTase I and farnesyltransferase (FTase), was particularly sensitive to GGTase I inhibition [
10]. This finding was further supported by a genetic approach which has shown that GGTase I deficiency reduces tumor formation and improves survival in mice with K-Ras-induced lung cancer [
11], thus further supporting the importance of GGTase I inhibition as a useful strategy to treat cancer, particularly K-Ras-induced cancer.
It has been documented that GGTase I inhibitors induce apoptosis in human cancer cells [
12‐
16]. However, the underlying mechanisms are largely unknown although it appears to be associated with inhibition of Akt and downregulation of survivin and Mcl-1 [
13,
14]. The present study focused on examining the effects of the GGTase I inhibitor GGTI-298, as well as GGTI-DU40, on induction of apoptosis, particularly induced by TRAIL, in human non-small cell lung cancer (NSCLC) cells and on understanding underlying mechanisms.
Discussion
The present study demonstrates that GGTI-298 inhibits the growth of human NSCLC cells accompanied with induction of apoptosis in addition to G1 arrest, indicating that induction of apoptosis in part accounts for GGTI-298's growth inhibitory effects in human NSCLC cells. GGTI-298 as a single agent moderately induces apoptosis. However, GGTI-298 when combined with TRAIL exerted enhanced apoptosis-inducing activity, indicating that GGTI-298 cooperates with TRAIL to augment apoptosis. In agreement, another highly selective GGTase I inhibitor, GGTI-DU40, but not its inactive analog, SN-DU40, exerted similar effects in augmenting TRAIL-induced apoptosis. To the best of our knowledge, this is the first report demonstrating that inhibition of GGTase I cooperates with TRAIL to augment apoptosis in human cancer cells. Given the recent study that GGTase I deficiency reduces lung tumor formation [
11], our data suggest that GGTase I inhibitors may be used in combination with TRAIL for treatment of human lung cancer through facilitating induction of apoptosis.
A previous report suggested that inhibition of Akt and subsequent downregulation of survivin are important for GGTI-298-induced apoptosis in human ovarian cancer cells [
14]. In our study, we observed that GGTI-298 inhibited Akt phosphorylation and downregulated survivin levels primarily in A549 cells, but not in other tested NSCLC cell lines (e.g., Calu-1, H157 and H226), indicating a cell line-specific modulation of Akt and survivin. In Calu-1 cells, GGTI-298 even increased p-Akt levels. Since Calu-1 and H157 cells were also sensitive to induction of apoptosis by either GGTI-298 alone or GGTI-298 plus TRAIL, inhibition of Akt and survivin is unlikely to be a common mechanism for GGTI-298- or GGTI-298 plus TRAIL-induced apoptosis although they may play roles in individual cell lines (e.g., A549). Currently, we do not know why GGTI-298 decreases Akt phosphorylation in some cells lines (e.g., A549), but increases Akt phosphorylation in others (e.g., Calu-1). Similar phenomenon was also documented from studies on FTase inhibitors. Whereas FTase inhibitors decrease Akt phoshoporylation in some cells [
33,
34], we previously reported that the FTase inhibitor SCH66336 increases Akt phosphorylation in some human lung cancer cell lines [
35].
Importantly, we found that GGTI-298 apparently downregulated the levels of c-FLIP, particularly FLIP
S and induced the expression of DR4 and DR5, primarily in GGTI-sensitive cell lines (e.g., A549, Calu-1 and H157), suggesting an association between modulation of these proteins and cell sensitivity to GGTI-298 or GGTI-298 plus TRAIL. We noted that downregulation of c-FLIP preceded induction of DR4 and DR5, implying that c-FLIP downregulation may play a critical role in mediating GGTI-298-induced apoptosis and sensitization of TRAIL-induced apoptosis. Indeed, we found that enforced expression of ectopic FLIP
S reduced cell sensitivity to GGTI-298-induced apoptosis, whereas siRNA-mediated silencing of c-FLIP further increased cell sensitivity to GGTI-298 (Fig.
6). Similarly, enforced expression of ectopic FLIP
S also protected cells from GGTI-298 plus TRAIL-induced apoptosis (Fig.
5). Thus, we conclude that c-FLIP, particularly FLIP
S, downregulation plays a crucial role in mediating apoptosis induced by either GGTI-298 or GGTI-298 combined with TRAIL.
Both DR4 and DR5 are receptors for TRAIL and can transduce apoptotic signaling upon binding to TRAIL. Interestingly, we found that silencing of DR5, but not DR4, abrogated cooperative induction of apoptosis by the combination of GGTI-298 and TRAIL (Fig.
7). Surprisingly, silencing of DR4 even sensitized cells to apoptosis induced by GGTI-298 plus TRAIL as we demonstrated using three DR4 siRNAs targeting different regions of the DR4 gene; however, this sensitization did not occur in cells where both DR4 and DR5 expression were silenced (Fig.
7). Taken together, these results indicate that DR5 induction is also critical for cooperative induction of apoptosis by the combination of GGTI-298 and TRAIL.
In this study, it appears that DR4 plays an antagonistic role in GGTI-298 plus TRAIL-induced apoptosis. Given that both DR4 and DR5 can mediate the TRAIL-induced NF-κB survival signaling pathway while initiating death signaling [
30,
31], it is possible that DR4 may be the major receptor that mediated TRAIL-induced NF-κB activation during TRAIL-induced apoptosis. Hence, silencing of DR4 would prevent NF-κB activation and enhance GGTI-298 plus TRAIL-induced apoptosis. Indeed, the GGTI-298 and TRAIL combination augmented IκBα phosphorylation, IκBα degradation and subsequent NF-κB activity (Fig.
8). Unexpectedly, knockdown of DR5, but not DR4, abrogated GGTI-298 plus TRAIL-induced IκBα degradation. Thus, the GGTI-298 and TRAIL combination induces a DR5-dependent IκBα degradation or NF-κB activation. Accordingly, we concluded that DR4 is unlikely to antagonize GGTI-298 plus TRAIL-induced apoptosis through activation of NF-κB.
Akt activation is often associated with apoptosis resistance [
36]. In this study, we also found that the combination of GGTI-298 and TRAIL inhibited Akt phosphorylation. Interestingly, respective knockdown of DR4 and DR5 generated opposite results on GGTI-298 plus TRAIL-induced p-Akt reduction; i.e., knockdown of DR5 inhibited p-Akt reduction in cells exposed to the combination of GGTI-298 and TRAIL, whereas knockdown of DR4 enhanced the ability of the GGTI-298 and TRAIL combination to decrease p-Akt levels (Fig.
8C). Thus, the GGTI-298 and TRAIL combination induces a DR5-dependent Akt inhibition. Here it seems that DR4 may play an opposite role to DR5 in positively regulating Akt activity. Hence, it is possible that DR4 may mediate Akt activation, at least in part leading to cell resistance to GGTI-298 plus TRAIL-induced apoptosis. To the best of our knowledge, this is the first study to show the regulation of Akt activity by the TRAIL death receptors.
The finding on the opposite roles of DR4 and DR5 in regulation of GGTI-198/TRAIL-induced apoptosis is intriguing. However, the underlying molecular mechanism(s) have not been elucidated. It is possible that DR5 and DR4 can form a heterodimer, which is less active than a DR5 homodimer and even functions as an antagonist of DR5 homodimers. Knockdown of DR4 will favor DR5 homodimer formation while reducing the DR5/DR4 heterodimer formation, leading to sensitization of cells to GGTI-298/TRAIL-induced apoptosis. Nonetheless, our findings warrant further study in this direction.
Both GGTase I and FTase are cytosolic heterodimers that share a common α subunit but use different β subunits and are all involved in prenylation of CAAX proteins [
7]. Our previous study showed that the FTase inhibitors SCH66336 and R115777 upregulate DR5 expression and augment TRAIL-induced apoptosis [
25,
26]. The current study further shows that the GGTase I inhibitor GGTI-298 exerts similar effects. Therefore, future studies should address whether there is a relationship between inhibition of protein prenylation and modulation of DR5 and c-FLIP. In this study, we have shown that GGTI-DU40, but not SN-DU40, exhibits identical results as GGTI-298 in modulation of c-FLIP levels, DR5 expression and TRAIL-induced apoptosis. Both GGTI-298 and GGTI-DU40 have similar GGTase I-inhibitory activity; however they possess distinct chemical structures [
17,
37]. All of these suggest a possible link between inhibition of protein geranylgeranylation and modulation of DR5 and c-FLIP.
Given that FTase and GGTase I inhibitors affect different set of proteins [
8], the question is why both types of inhibitors exert similar effects on DR5 upregulation. One possible is that both FTase and GGTase I inhibitors affect a common protein or pathway, leading to DR5 upregulation. In this study, we investigated the possible role of RhoB in GGTI-298-induced DR5 upregulation since RhoB is modified and regulated by both farnesylation and geranylgeranylation [
28]. We found that GGTI-298 increases RhoB expression (Fig.
3A), which was paralleled with DR5 induction (Fig.
2C). Moreover, siRNA-mediated blockade of RhoB upregulation partially inhibited DR5 induction by GGTI-298 (Fig.
3B). These results suggest that RhoB in part mediates DR5 induction by GGTI-298. Because RhoB siRNA transfection completely blocked RhoB upregulation in our study, but failed to abolish GGTI-298's ability to increase DR5 expression (Fig.
3B), other mechanisms may also account for GGTI-298-induced DR5 upregulation and need to be studied further.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
SC, LF, SMR and PY designed and conducted experiments as well data analysis. FRK participated in discussion of the data and draft of the manuscript. SYS participated in experimental design, coordination, data analysis and draft of the manuscript. All authors read and approved the final manuscript.