Farnesol-Induced Apoptosis in Human Lung Carcinoma Cells Is Coupled to the Endoplasmic Reticulum Stress Response

Farnesol (FOH) and other isoprenoid alcohols induce apoptosis in various carcinoma cells and inhibit tumorigenesis in several in vivo models. However, the mechanisms by which they mediate their effects are not yet fully understood. In this study, we show that FOH is an effective inducer of apoptosis in several lung carcinoma cells, including H460. This induction is associated with activation of several caspases and cleavage of poly(ADP-ribose) polymerase (PARP). To obtain insight into the mechanism involved in FOH-induced apoptosis, we compared the gene expression profiles of FOH-treated and control H460 cells by microarray analysis. This analysis revealed that many genes implicated in endoplasmic reticulum (ER) stress signaling, including ATF3, DDIT3, HERPUD1, HSPA5, XBP1, PDIA4, and PHLDA1, were highly up-regulated within 4 h of FOH treatment, suggesting that FOH-induced apoptosis involves an ER stress response. This was supported by observations showing that treatment with FOH induces splicing of XBP1 mRNA and phosphorylation of eIF2α. FOH induces activation of several mitogen-activated protein kinase (MAPK) pathways, including p38, MAPK/extracellular signal–regulated kinase (ERK) kinase (MEK)-ERK, and c-jun NH₂-terminal kinase (JNK). Inhibition of MEK1/2 by U0126 inhibited the induction of ER stress response genes. In addition, knockdown of the MEK1/2 and JNK1/2 expression by short interfering RNA (siRNA) effectively inhibited the cleavage of caspase-3 and PARP and apoptosis induced by FOH. However, only MEK1/2 siRNAs inhibited the induction of ER stress–related genes, XBP1 mRNA splicing, and eIF2α phosphorylation. Our results show that FOH-induced apoptosis is coupled to ER stress and that activation of MEK1/2 is an early upstream event in the FOH-induced ER stress signaling cascade. [Cancer Res 2007;67(16):7929–36]

Isoprenoids are important in the regulation of cell proliferation, apoptosis, and differentiation (1–6). The nonsterol isoprenoid farnesol (FOH) is produced by dephosphorylation of farnesyl pyrophosphate, a catabolite of the cholesterol biosynthetic pathway (7, 8).

FOH and the related isoprenoids perillyl alcohol (POH), geranylgeraniol (GGOH), and geraniol (GOH) are found in a wide range of fruits and vegetables (9, 10). Each isoprenoid has been shown to inhibit proliferation and induce apoptosis in a number of neoplastic cell lines from different origins (4, 11–14). In addition, these isoprenoids have been reported to be effective in chemoprevention and chemotherapy in various in vivo cancer models (10, 12, 15, 16). FOH has been reported to exhibit chemopreventive effects in colon and pancreas carcinogenesis in rats (9, 17) whereas phase I and II clinical trials have indicated therapeutic potential for POH (16, 18). The mechanisms by which these isoprenoids induce these effects are not yet fully understood. Isoprenoids have been reported to inhibit posttranslational protein prenylation (19) whereas other studies have shown that FOH is a potent inhibitor of the CDP-choline pathway (5, 20, 21). Other effects include inhibition of phospholipase D (22), inhibition of 3-hydroxy-3-methylglutaryl CoA reductase (HMG CoA reductase) activity (6), generation of reactive oxygen species (ROS; refs. 23, 24), and disorganization of the cytoskeleton (4). FOH has also been reported to act as a weak activator of the farnesoid X activated receptor (25).

To obtain greater insight into the mechanism by which FOH induces apoptosis, we did microarray analysis and compared the gene expression profiles between vehicle-treated and FOH-treated human lung adenocarcinoma H460 cells. This analysis showed that a large number of genes associated with the endoplasmic reticulum (ER) stress response are rapidly induced by FOH treatment, suggesting that FOH-induced apoptosis is coupled to the ER stress response. The ER is an organelle responsible for the synthesis, posttranslational modification, and proper folding of membrane and secretory proteins. Disturbance of ER homeostasis results in the activation of the unfolded protein response (26–29). During this response, several prosurvival and proapoptotic signals are activated and, depending on the extent of the ER stress, cells survive or undergo apoptosis. Several (pathologic) conditions, including nutrient deprivation, oxidative stress, changes in calcium homeostasis, failure in posttranslational modifications or transport of proteins, and treatment with a variety of agents, can induce ER stress and trigger the unfolded protein response. ER stress has been implicated in many disease processes, including cancer, diabetes, cardiovascular and neurodegenerative disease, ischemia, and inflammation (30–32).

In this study, we further show that FOH treatment of H460 cells results in the activation of several mitogen-activated protein kinases (MAPK), including p38, extracellular signal–regulated kinase (ERK)-1/2, and c-jun NH₂-terminal kinase (JNK)-1/2. Treatment with chemical inhibitors and short interfering RNA (siRNA) knockdown experiments showed that activation of MAPK/ERK kinase (MEK)-1/2 is an early event that is upstream of the activation of the ER stress signaling cascade by FOH.

Materials. Trans,trans-FOH, GOH, GGOH, and (S)-(−)-POH were purchased from Sigma-Aldrich. U0126 was obtained from Promega and SB203580 was purchased from Calbiochem.

Cell lines and cell culture. The human lung carcinoma cell lines H460 (adenocarcinoma), H1355 (adenocarcinoma), H82 (small cell carcinoma), and Calu6 (alveolar anaplastic carcinoma) and the immortalized human bronchial epithelial cell line BEAS-2B were obtained from American Type Culture Collection. The carcinoma cells were grown in RPMI 1640 (Life Technologies, Inc.) supplemented with 10% heat-inactivated fetal bovine serum (Atlanta Biologicals) and 100 units/mL each of penicillin and streptomycin (Sigma-Aldrich). BEAS-2B cells were grown in KGM (Cambrex).

Proliferation and cell death assays. Cell proliferation and viability were evaluated using the Cell Proliferation Kit II (XTT) following the manufacturer's protocol (Roche). Absorbance was measured at 450 nm using a microplate reader (Molecular Devices Corp.). Apoptosis was measured with a Cell Death Detection ELISA kit (Roche).

Flow cytometry. H460 cells treated with FOH for different time intervals were harvested, resuspended in PBS, and then fixed in 70% ethanol. Fixed cells were washed twice in PBS and then resuspended in 0.5-mL propidium iodide solution consisting of 50 μg/mL propidium iodide, 0.1 mg/mL RNase A, and 0.05% Triton X-100 in PBS for 30 min. Cell cycle analysis was done with a FACSort flow cytometer (Becton Dickinson). Data were analyzed using CellQuest software (Becton Dickinson).

Microarray analysis. Total RNA was isolated using TriReagent (Sigma) following the manufacturer's protocol. Gene expression analyses were conducted by the NIEHS Microarray Group (NMG) on Agilent Whole Human Genome microarrays (Agilent Technologies). Total RNA was prepared from H460 cells treated for 4 h with 250 μmol/L FOH or vehicle (DMSO). RNA from two independent experiments was used and each microarray analysis was done in duplicate. Further experimental details of the microarray analysis are provided in Supplementary Materials and Methods.

Western blot analysis. Cells were harvested and lysed in lysis buffer containing 50 mmol/L Tris-HCl (pH 7.4), 150 mmol/L NaCl, 1% NP40, and 0.1% SDS, supplemented with protease and phosphatase inhibitor cocktails I and II (Sigma). After centrifugation, proteins were examined by Western blot analysis with the antibodies indicated (Supplementary Materials and Methods). The blots were developed with a peroxidase-conjugated secondary antibody and enhanced chemiluminescence detection reagent (GE Healthcare Life Sciences).

Northern blot analysis and reverse transcription-PCR. Total RNA was isolated using TriReagent following the manufacturer's protocol. RNA was separated on a 1.2% agarose gel containing 0.5% formaldehyde in 1× MOPS buffer and then transferred onto a nylon membrane (Sigma). After UV cross-linking, the membrane was hybridized to ³²P-labeled probes. The membrane was then washed and exposed to Hyperfilm (Amersham Bioscience) at −70°C. Probes used in Northern blot analysis were generated by PCR as described in Supplementary Materials and Methods. The nonconventional splicing of XBP1 mRNA was examined by reverse transcription-PCR (RT-PCR) using 5′-CCTTGTAGTTGAGAACCAGG and 5′-GGGGCTTGGTATATATGTGG as primers. This will amplify both unspliced (XBP1u) and spliced (XBP1s) XBP1 mRNAs.

SiRNA knockdown. Knockdown of MEK1/2 and JNK1/2 expression in H460 cells was achieved by transfection of siRNAs. The siRNAs of human MEK-1, MEK-2, JNK-1, and JNK-2 were purchased from Santa Cruz Biotechnology. The silencer-negative control siRNA was purchased from Ambion. Transfection of siRNA was done with DharmaFECT 4 transfection reagent. H460 cells were plated in six-well dishes at a density 3.3 × 10⁵ per well. The next day, cells were treated with the siRNA transfection mixtures following the DharmaFECT General Transfection Protocol. After 48-h incubation, cells were treated with or without FOH as indicated and harvested for Western and Northern blot analyses.

Statistics. Values are presented as means ± SD. Statistical analysis was done with the Student t test.

FOH induces apoptosis in lung carcinoma cells. XTT analysis showed that treatment of human lung carcinoma H460 and Calu6 cells and of the immortalized human bronchial epithelial BEAS-2B cells with FOH greatly reduced cell proliferation and viability, extending previous observations (refs. 2, 5, 6, 11, 14, 20, 33; Fig. 1A). FOH also reduced the growth and viability of lung adenocarcinoma H1355 and small cell lung carcinoma H82 cells (not shown). The concentrations at which FOH inhibited cell proliferation ranged from 100 to 300 μmol/L. FOH inhibited cell proliferation of H460 cells in a time-dependent manner (Fig. 1B). At 225 μmol/L, FOH reduced cell viability by ∼50% after 24 h of treatment, whereas at 250 μmol/L, <10% of the cells were viable. Comparison of the growth inhibitory effect of FOH with those of GOH, GGOH, and POH showed that FOH was the most effective in inhibiting the proliferation of H460 cells followed by GGOH whereas treatment with GOH and POH had little effect (Fig. 1C).

To determine whether the FOH-induced cell death in H460 cells involved apoptosis, we analyzed the cell cycle distribution and the induction of nucleosome fragmentation. Flow cytometric analysis showed that treatment of H460 cells with 250 μmol/L FOH induced a change in cell cycle distribution (Fig. 1D). The percentage of sub-G₀-G₁ cells in FOH-treated H460 cultures steadily increased over a period of 24 h whereas the number of cells in G₁, S, and G₂-M greatly diminished. At 12 and 24 h, respectively, 49% and 71% of the cells were in sub-G₀-G₁. The cell cycle profile did not significantly change over a 24 h period in cells treated with vehicle only (not shown). Nucleosome fragmentation analysis determined by Cell Death Detection ELISA confirmed that H460 cells treated for 6 h with 250 μmol/L FOH undergo apoptosis (Supplementary Fig. S1). Pretreatment with the pan-caspase inhibitor z-VAD-fmk effectively inhibited nucleosome fragmentation, indicating that FOH-induced cell death is caused by a caspase-dependent apoptotic pathway.

FOH induces ER stress and unfolded protein response. To obtain insight into the early events involved in FOH-induced apoptosis, we did microarray analysis to determine the gene expression profiles of H460 cells treated for 4 h with vehicle (DMSO) or 250 μmol/L FOH. Comparison of the gene expression profiles identified a large number of genes that were induced or repressed by FOH in H460 cells.³

Of the 1,480 changes in gene expression, 537 genes were found to be induced ≥2-fold and 320 genes were down-regulated ≥2-fold in FOH-treated H460 cells. A selective list of genes induced or repressed by FOH is shown in Table 1. Closer analysis of the gene expression profiles revealed that among the genes highly up-regulated by FOH, many were typical for ER stress response signaling. These included the transcription factors DNA damage–inducible transcript 3 [DDIT3; also named GADD153], activating transcription factor 3 (ATF3), and X-box binding protein 1 (XBP1); the chaperones HSPA5 (also named GRP78 or BiP) and HSP90B1 (also named GRP94), protein disulfide isomerase A4 (PDIA4), and homocysteine-inducible, ER stress–inducible, ubiquitin-like domain member 1 (HERPUD1; also named HERP); and pleckstrin homologue-like domain family A1 (PHLDA1, also named TDAG51). The expression of several members of the Fos/Jun and a number of early response genes, including the nuclear receptors Nurr1 (NR4A2) and Nur77 (NR4A1), and members of the Egr family of transcription factors were also greatly up-regulated. Many of these genes have been reported to act downstream in the ER stress signaling cascade (26, 28, 29, 32).

The induction of the expression of several ER stress–related genes by FOH was confirmed by Northern blot analysis. Figure 2A shows a time course of the induction of DDIT3, ATF3, HSPA5, XBP1, HERPUD1, and PHLDA1 mRNA expression. An increase in the expression of these mRNAs was observed within 2 h after the addition of FOH, and the expression of most mRNAs reached a maximum at 4 h of FOH treatment. These observations suggest that the FOH-induced cell death program is coupled to activation of the ER stress signaling pathway.

To compare the ability of several alcohol isoprenoids to induce an ER stress response, H460 cells were treated with FOH, GOH, GGOH, and POH for 4 h at the concentrations indicated and the induction of DDIT3 and ATF3 expression was examined (Fig. 2B). FOH was the most effective inducer of DDIT3 and ATF3 mRNA expression in H460 cells followed by GGOH. Treatment with GOH and POH did not induce these ER stress marker genes. These results correlate with the effects of these isoprenoids on the viability of H460 cells and suggest that both FOH and GGOH induce the unfolded protein response.

ER stress triggers several specific signaling pathways including the unfolded protein response (26–29, 32). The latter involves the activation of several proteins including inositol requiring protein 1 (IRE1 or ERN1) and PKR-like ER kinase (PERK). Activation of the RNase activity of IRE1 initiates splicing of XBP1u into XBP1s mRNA, which is subsequently translated into a potent transcription factor (34, 35). Our results showed that FOH induced the generation of XBP1s transcripts (Fig. 2A,, bottom). The ER stress sensor PERK induces the phosphorylation and inactivation of the translation initiation factor eIF2α, resulting in an attenuation of the rate of general translational initiation (36). As shown in Fig. 2C, FOH induced phosphorylation of eIF2α in H460 cells in a time-dependent manner. These observations suggest that FOH induces the activation of two important unfolded protein response sensors, IRE1 and PERK.

Several caspases have been reported to be activated in the unfolded protein response (26–29, 37). Examination of caspase activation in FOH-treated H460 cells showed that caspase-3 and caspase-9 were significantly activated at 4 h, similar to the time course of poly(ADP-ribose) polymerase (PARP) cleavage (Fig. 2D). In addition, FOH treatment reduced the level of unprocessed caspase-4 by ∼70%, indicating that FOH induces activation of caspase-4. No activation of caspase-8 was detected.

FOH induces activation of p38, ERK, and JNK. Activation of MAPKs are involved in many aspects of the control of cellular proliferation and apoptosis and have been implicated in the regulation of gene expression in the ER stress signaling cascade (26–29, 32, 38–40). We therefore examined the effects of FOH on the activation of several MAPK pathways. As shown in Fig. 3A, treatment of H460 cells with FOH induced the level of phosphorylated ERK, p38, and JNK whereas it did not change the level of total ERK, p38, and JNK protein. An increase in ERK phosphorylation was observed at 5 to 15 min and reached a maximum at 60 min. This activation seemed to precede the phosphorylation of p38 and JNK.

To determine the role of the activation of p38 and ERK1/2 in FOH-induced ER stress response, H460 cells were pretreated with the MEK1/2 inhibitor U0126 or the p38 MAPK inhibitor SB203580 and their effects on the induction of several stress response genes examined. As shown in Fig. 3B, induction of DDIT3, ATF3, HSPA5, HERPUD1, and PHLDA1 mRNA by FOH was effectively inhibited by U0126. These results suggest that activation of MEK1/2 is important for FOH-induced ER stress response and is an upstream event. SB203580 had a small effect on the increase in DDIT3 and HERPUD1 mRNA expression but significantly reduced the induction of ATF3 expression. The latter is in agreement with previous observations showing that p38 MAPK is involved in induction of ATF3 by stress signals (41).

Induction of ER stress response and apoptosis by FOH is dependent on MEK1/2 activation. To analyze the role of MAPKs further, we examined the effect of MEK1/2 and JNK1/2 knockdown by siRNA on the ER stress signaling cascade. As shown in Fig. 4A, the MEK1/2 and JNK1/2 siRNAs reduced significantly the expression of MEK1/2 and JNK1/2, respectively. In addition, these siRNAs very effectively inhibited FOH-induced apoptosis in H460 cells (Fig. 4B) and the cleavage of PARP and several caspases (Fig. 4C). The inhibition of the apoptotic cascade was confirmed by morphologic observations (Supplementary Fig. S2).

We next examined the effects of MEK1/2 and JNK1/2 knockdown on the induction of several ER stress–related genes. As shown in Fig. 5A, siMEK1/2 significantly reduced the induction of DDIT3, ATF3, HSPA5, HERPUD1, and PHLDA1 mRNA expression by FOH, consistent with the observed inhibition by U0126 (Fig. 3B). In contrast, knockdown of JNK1/2 by siJNK1/2 had little effect on the expression of these ER stress–related genes. Transfection of siMEK1/2 also inhibited XBP1 mRNA splicing whereas siJNK1/2 had no effect. These observations suggest that activation of MEK1/2 is upstream of IRE1 activation whereas JNK activation is a downstream event. The latter conclusion is in agreement with previous studies (27–29, 32, 39, 40).

As expected, knockdown of siMEK1/2 greatly reduced total MEK1/2 protein and, as a consequence, the level of phosphorylated MEK1/2. Knockdown of MEK1/2 expression also inhibited the induction of phosphorylated JNK by FOH without affecting the total level of JNK1/2 protein (Fig. 5B). In addition, PERK-induced phosphorylation of eIF2α by FOH was greatly diminished in H460 cells transfected with siMEK1/2. These results are in agreement with the conclusion that activation of MEK1/2 is upstream in the ER stress signaling cascade.

In this study, we show that FOH is able to induce apoptosis in a number of human lung carcinoma cell lines, as indicated by the activation of several caspases, and the induction of PARP cleavage and nucleosomal degradation. The mechanism by which this isoprenoid induces apoptosis in a variety of cell systems is not yet fully understood (4, 5, 11–14). The objective of this study was to obtain greater insight into the mechanisms involved in the induction of apoptosis by FOH. We therefore did microarray analysis and compared the gene expression profiles between human lung carcinoma H460 cells treated for 4 h with FOH or vehicle (DMSO). This analysis showed that FOH induced the expression of a great number of genes typically associated with ER stress response signaling, including ATF3, DDIT3, XBP1, HSPA5, HSP90B1, PDIA4, HERPUD1, and PHLDA1 (26, 28, 29, 32). These observations suggested that FOH induces an ER stress response.

The ER plays a critical role in the regulation of protein synthesis, protein folding, and trafficking. A wide variety of signals have been reported to disrupt ER function and induce ER stress, which is associated with an accumulation of unfolded or malfolded proteins in the ER (26, 28, 29, 32). Activation of the ER stress response leads to attenuation of protein synthesis to prevent the accumulation of more proteins, the translocation of unfolded or malfolded proteins and their degradation by the ubiquitin/proteasome system, induction of chaperone synthesis to increase folding capacity, and induction of apoptosis. The ER stress response is a balance between prosurvival and proapoptotic signaling pathways. When the prosurvivor responses fail, cells undergo apoptosis, as seems to be the case in FOH-treated H460 cells.

ER stress triggers the activation of several sensor proteins, including PERK, IRE1, and ATF6 (27–29, 32). FOH treatment seems to activate both IRE1 and PERK because FOH induces splicing of XBP1 mRNA and phosphorylation of eIF2α. Splicing of XBP1 mRNA has been reported to depend on the activation of the RNase activity of IRE1 (34, 35) and to result in the synthesis of a potent XBP1 transcription factor. Phosphorylation of eIF2α is catalyzed by PERK. Phosphorylated eIF2α causes a decrease in the rate of translation initiation of many proteins but enhances the translation of the transcription factor ATF4 (26–29, 32, 42). ATF4 and XBP1 in combination with ATF6 regulate the transcription of several ER stress–related genes, including ER luminal chaperones HSPA5, HSP90B1, and PDIA4 and the transcription factor DDIT3. These observations support the conclusion that FOH induces an ER stress response.

Treatment with FOH, as has been shown for other ER stress–inducing conditions, results in the activation of the MAPK family members p38, ERK, and JNK (29, 32, 39, 40). In response to ER stress, IRE1 becomes activated through autophosphorylation and recruits tumor necrosis factor receptor–associated factor 2 (TRAF2), an adaptor protein involved in the signaling pathways of proinflammatory cytokines. This complex in turn activates apoptosis signal–regulating kinase 1 (ASK1) and subsequently the MAPK kinase 4-JNK pathway (39, 40) and, possibly, the p38 pathway. The IRE1-TRAF2-ASK1-JNK pathway is an important pathway in ER stress–induced apoptosis. Our results show that knockdown of JNK1/2 expression did not inhibit activation of IRE1 but inhibited caspase activation and the induction of apoptosis in H460 cells by FOH. However, JNK1/2 siRNAs had little effect on the induction of most ER-stress–related genes analyzed (Fig. 5A). Our observations are consistent with the concept that JNK activation is dependent on IRE1 activation and suggest that activation of IRE1-TRAF2-ASK1-JNK pathway plays a critical role in FOH-induced apoptosis.

Activation of the MEK-ERK signaling pathway is generally considered as a prosurvival signal. However, activation of ERK1/2 has been reported to play a significant role in the induction of apoptosis in renal, neuronal, and hepatoma cells by a variety of conditions (43, 44). Our study shows the importance of the MEK-ERK signaling pathway in the FOH-induced ER stress response. This is indicated by the inhibition of the induction of ER stress–related genes by the MEK1/2 inhibitor U0126. Moreover, knockdown of MEK1/2 expression in H460 cells by siRNA inhibited the FOH-induced splicing of XBP1 mRNA, eIF2α phosphorylation, activation of JNK1/2 and several caspases, the induction of several ER stress–related genes, and apoptosis. These observations suggest that activation of the MEK1/2 signaling pathway is an early event and critical in triggering FOH-induced ER stress.

This raises the question on what mechanism FOH activates the MEK-ERK pathway. FOH has been reported to generate ROS in Saccharomyces cerevisiae (24, 45) and ROS has been shown to induce activation of ERK1/2 in mammalian cells under several conditions (46, 47). Preliminary results showed that FOH enhances ROS in human lung carcinoma cells.⁴

Therefore, generation of ROS by FOH may result in the activation of the ERK1/2 pathway and subsequently trigger the ER stress response. However, activation of the ERK1/2 pathway might be related to other actions of FOH, including inhibition of the CDP-choline pathway (5, 20, 21, 48) and inhibition of phospholipase D (22) or HMG CoA reductase activity (6). Future studies have to provide insight into the mechanism by which FOH activates the MEK-ERK pathway.

FOH also induces the expression of several members of the Egr family, transcription factors that are important regulators of apoptosis and cell proliferation. Egr1 has previously been shown to be induced during ER stress (49) and has been implicated in the regulation of several genes, including ATF3 (50). Therefore, the induction of ATF3 in FOH-treated H460 cells might be related to the observed increase in Egr1 expression. The induction of ATF3 in FOH-treated H460 cells also depends on the activation of p38 MAPK, in agreement with a previous report showing that ATF3 is a p38 target gene (41). These observations indicate that the transcriptional regulation of genes downstream in the ER stress signaling cascade might be regulated by multiple kinase pathways and transcription factors that affect their transcription either synergistically or antagonistically.

In summary, in this study we show that FOH-induced apoptosis in human lung carcinoma cells is coupled to the activation of an ER stress response that includes activation of the sensors IRE1 and PERK. In addition, we show that this induction is dependent on the activation of the MEK-ERK signaling pathway, suggesting that it is an early event in FOH-induced ER stress and apoptosis.

Grant support: Intramural Research Program of the NIEHS, NIH.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Dr. Carl Bortner for his advice with Flow Cytometry, Drs. Kyung-Soo Chun and Carl Bortner for their valuable comments on the manuscript, and Dr. Gary S. Bird and Johnny F. Obie for their advice.