3-indole induced mRNA expression profile by cDNA microarray.
Experimental A549 RNA was isolated using Trizol reagent (GIBCO BRL, Life Technology, USA). Fromeach sample, total RNA (Control universal human reference RNA (Stratagene, La Jolla, CA) and experimental A549 RNA were used to generate cDNA. Microarray slides were scanned using GenePix 4000B Biochip Analyzer (Axon Instruments, Union City, USA). Changes in gene expression were
presented as logarithmic ratios of fluorescence intensities. The logarithmic ratios of each indicated times were then normalized for each gene to that of Control RNA to obtain the expression pattern (the log-intensity log2R of the red dye versus the log-intensity log2G of the green dye, as well as the log intensity ratio M= log2R/G, experimental Cy5 and Control Cy3). The genes that showed substantial differences after drug treatment were selected based on at least a 2-fold change in expression.
XIII. Subcutaneous Implantation of Cancer Cells in Animals and Monitoring of in Vivo Anti-tumoral Activity after Drug Treatments. Athymic nu/nu female mice (ICR-Foxn1), 4–5 weeks of age, were obtained from the National Laboratory Animal Center (Taiwan, Republic of China). The animals were implanted subcutaneously (s.c.) with 5 × 106 A549 or H1435 lung cancer cells in 0.1 ml Hanks’ balanced salt solution (HBSS) in one flank per mouse.
The size of the tumor mass was measured and the tumor volume was calculated as 1/2 × length × width2 in mm3. In human lung cancer xenograft studies, when tumors attained a mass of ~ 50 mm3, animals were treated intraperitoneally (i.p.) with 3-indole at 0.2 mg/day on days 0, 2, 4, 6, and 8 (final dose, 50 mg/kg) or 0.1 mg/day on days 0, 2, 4, 6, and 8 (final dose, 25 mg/kg), or a solvent (control). A solvent mixture contains DMSO/Cremophor EL/saline (2:1:7). The assay solvent was basically performed according to the method described by Kuo et al.
(Kuo et al., 2004). Tumor size was measured after drug treatment. Prior to being sacrificed, the animals were anesthetized and blood samples
were collected by intracardiac puncture for the mice organ function test.
Before organ dissection, the animals were sacrificed by cervical dislocation. Tumor samples and mice organ tissues (including the lungs and kidneys) were resected, fixed with formalin and embedded in paraffin for histologic examination, stained with hematoxylin and eosin for microscopic evaluation, and examined by a pathologist.
RESULTS
I. 3-indole Apparently Inhibited Growth at Low Concentration and Promoted Cell Death at High Concentration in Various Human Lung Cancer Cells. 1,1,3-tri(3-indolyl)cyclohexane (3-indole) is a novel, 2-step synthetic indole compound of high purity and yield. Its structure is shown in Fig. 1. To test the cytotoxicity effect and future clinical usage of 3-indole, normal human lung fibroblast cells IMR-90 and various human non-small cell lung carcinoma (NSCLC) cells with different p53 status including A549 (p53-wild), H1299 (p53-null), H1435 (p53-mutation), CL1-1 (p53-mutation), and H1437 (p53-mutation) cells were tested. Cells were treated with 0, 1, 2, 5, or 10 μM of 3-indole for 24 h and assessed cell viability by the MTT assay.
Fig. 2 shows that 3-indole caused a concentration-dependent reduction in cell viability. 3-indole could achieve an inhibitory concentration (IC) 50 value at ~ 10 μM in various human NSCLC cells (A549, H1299, H1435, CL1-1, and H1437 cells), whereas did not show apparent cycotoxicity to the IMR-90 cells at this concentration. 3-indole efficacy was similar against all human NSCLC cells tested regardless of the status of p53, which is the most common genetic alteration in human cancers. These results show that 3-indole were efficacious against human NSCLC cells with various status of p53. The anti-cancer efficacy of 3-indole was also noted in various human esophageal squamous cell carcinoma cell lines, including KYSE170, KYSE50, KYSE510, and KYSE70 (appendix Figure 1). Furthermore, to examine whether the cytotoxicity observed for 3-indole was due to cell growth inhibition or
cell death, various human lung cancer cells including A549, H1299, H1435, CL1-1, and H1437 were treated with 0, 2, 10, or 30 μM of 3-indole for the indicated times and cell proliferation was assessed by the MTT assay. Fig. 3 shows that 3-indole caused a concentration-dependent reduction in cell proliferation with apparent inhibition of growth at low concentration (10 μM) and promotion of cell death at high concentration (30 μM) in various human lung cancer cells.
II. 3-indole Induced Cell Cycle Arrest and Apoptosis in Various Human NSCLC Cells. Microtubules are highly dynamic polymers composed of α–tubulin and β-tubulin heterodimers that constantly assembly (polymerization) or shortening (depolymerization). Indole-like compounds are known to arrest cells in G1 or G2-M, and substantially induce apoptosis (Brandi et al., 2003; Kuo et al., 2004). To determine whether the anti-cancer effect of 3-indole was associated with cell cycle deregulation, the cell cycle distribution was analyzed by flow cytometry, we investigated whether cell cycle arrest and/or apoptosis could be induced in various human lung cancer cells (A549, H1299, H1435, CL1-1, and H1437 cells) treated with 3-indole at 0, 10, and 30 μM for 24 or 48 h. Flow cytometry indicated that 10 μM of 3-indole caused most cancer cell lines, except for H1437, to accumulate in G1 phase and a substantial increase in the sub-G1 region (an apoptosis indicator) resulted from treatment with 30 μM of 3-indole at 24 h (Fig. 4). Furthermore, the partially G2-M arrest efficacy of treatment 3-indole with 30 μM at 48 h was also noted in H1437 cells (Fig. 5).
III. Activation of the p53/p21 Pathway Is Required for the Induction of G1 Cell Cycle Arrest in 3-Indole. The function of p53 as a tumor suppressor has been demonstrated by experiments showing that p53 correlates with G1 or G2 cell cycle regulation after DNA damage (Kastan et al., 1991; Liebermann et al., 1995; Park et al., 2001; Liu et al., 2003). Some report has been demonstrated that p21, a tumor suppressor, is response to upregulation by p53 or by p53-independent remains cell cycle in G1 or G2-M allowing time for DNA repair.
Therefore, we performed Western blot to confirm whether the p53/p21 pathway was activated after 3-indole treatment. The preliminary data in Fig. 6A shows that treatment A549 cells (p53-wild) with 10 μM 3-indole increased the expression of p53 protein after 2 h and subsequently p21 increase expression in 8 h. The similar observations were also noted in H1437 cells (p53-mutation) (Fig. 6B). Furthermore, treatment with 10 μM 3-indole can induced cyclin B1 (an G2-M indicator) expression partial increases at 48 h in H1435 cells (p53-mutation) (Fig. 6C). Interestingly, 3-indole treatment for 12 h initially decreased cyclin B1 (Fig. 6C).
IV. 3-indole Induced Apoptosis Through the Activation of the Intrinsic Mitochondrial Pathway. To confirm that the sub-G1 region was caused by apoptosis, we performed a DNA ladder analysis, and found that ladders appeared in various human NSCLC cells (A549, H1299, H1435, and CL1-1 cells) at 24 h, and in H1437 cells at 48 h after 3-indole treatment (Fig. 7). These results suggest that 3-indole maybe induce cell death via G1 or G2-M arrests in various human
NSCLC cells. Furthermore, using Western blot analysis to investigate the mechanism of 3-indole induced apoptosis, we found that treatment of A549 cells with 30 μM of 3-indole resulted in a time-dependent reduction in the levels of the anti-apoptotic protein, Bcl-2. At the same time, the level of the pro-apoptotic protein, Bax and Bad, was concomitantly increased compared with the cells that were not treated with 3-indole (Fig. 8A upper panel). The expression decrease of Bcl-2 was also noted in H1437 cells (Fig. 8B lower panel). To further dissect the apoptosis pathway induced by 3-indole, we performed Western blot analysis for cytochrome c release and caspase protein expression, and used different fluorogenic tetrapeptide substrates (Ac-DEVD-pNA, Ac-LETD-pNA, and Ac-LEHD-pNA) to measure the activity of caspases -3, -8, and -9, respectively. 3-indole increased the release of cytochrome c from mitochondria to cytosol in 8 h and stimulated caspases -3, -9 (an indicator of the intrinsic mitochondria pathway) and to a lesser extent caspase -8 (an indicator of the extrinsic membrane receptor pathway) activities in A549 cells (Figs. 8 and 9). Together, these results showed that 3-indole induced the execution of apoptosis through the activation of the mitochondrial pathway.
V. 3-indole Induced Cell Cycle Arrest and Apoptosis by Reactive Oxygen Species Production and DNA Double-Strand Breaks in A549 or H1299 Cells. A number of studies have shown that loss of the mitochondrial membrane potential (MMP) in cells triggers mitochondrial disruption and the generation of reactive oxygen species (ROSs) (Herrera et al., 2001; Gupta et al., 2003; Gong et al., 2006). ROSs are known to
damage many molecules including proteins, RNA, and DNA (Salmon et al., 2004; Pan et al., 2005). We examined the changes in the MMP and mitochondrial localization using DiOC6, a cationic fluorescent probe. A concentration-dependent change in MMP was observed at 15-30 min in A549 cells (Fig. 10, upper panel). The data in Fig. 10 lower panel shows that Treatment of 10 or 30 μM of 3-indole decreased the MMP in 4 h but only high concentration (30 μM) of 3-indole continuously decreased the MMP. Moreover, using cell fluorescence staining, mitochondrial localization was detected by a MitoTracker. In untreated cells, mitochondria were evenly distributed in the cytoplasm. In 3-indole (30 μM) treated A549 cells, aggregated mitochondria increased after 12 h, and dendrite-like structures disappeared (Fig. 11).
Next, we examined the changes in ROS production and DNA damage in cells treated with various concentration of 3-indole for the indicated times. A significant increase in ROS production was observed in various human NSCLC cells A549 and H1299 at 2 h with 10 μM 3-indole (Fig. 12). In addition, A549 cells were treated with 3-indole and rotenone (0.05 μM, an inhibitor of mitochondrial respiratory chain complex I) or 3-indole and N-acetylcysteine (NAC) (5 mM, a hydroxyl radical scavenger). The results indicated a partial reversal of ROS production by rotenone (Fig. 13) and reduced apoptosis during co-treatment with rotenone or NAC compared to 3-indole treatment alone (Fig. 14). Since there was an increase in ROS production, we decided to assess the degree of DNA strand break damage using pulsed-field gel electrophoresis (PFGE). A549 cells, following 30 μM 3-indole treatment, exhibited a change in DNA damage at 24 h (Fig. 15 left panel). In
addition, we treated A549 cells with both 3-indole and rotenone (0.05 μM). The data indicated that co-treatment with rotenone reduced DNA damage compared to 3-indole treatment alone (Fig. 15 right panel).
VI. cDNA Microarray Analysis to Search For Differential Expressed Genes After 3-indole Treatment. To reveal more potential targets and pathways involved in 3-indole treatment, we performed cDNA microarray analysis on untreated A549 cells at 0 h and A549 cells treated with 30 μM of 3-indole at 4, 8, and 12 h and harvested the RNA. The dose chosen were close to the dose needed for apoptosis induction. The rationale for the indicated times was to capture the expression profiles of genes that involved in the apoptotic processes. We found many differentially expressed genes, which are related to cell cycle, apoptosis, and cell signaling pathways (Fig. 16). For example, we found that 30 μM of 3-indole caused changes in the mRNA levels of several mitogen-activated protein kinase (MAPK) signaling proteins such as p38β and JNK2. Further, that 3-indole affected other signaling pathways, such as PI3K-Akt and Wnt signaling pathways. In addition, 3-indole treatment reduced the expression of histone deacetylase 1 (HDAC1).
VII. Activation of the JNK Signaling Pathways Is Required for the Induction of Apoptosis in 3-indole Treated A549 Cells. cDNA microarray data revealed that expression of several proteins in MAPK pathway changed after 3-indole treatment. In addition, ROS has been shown to induce various biological processes, including activation of the
MAPK pathway (Kamata et al., 2005; Gong et al., 2006). Therefore, we performed Western blot to confirm whether the MAPK signaling pathway was activated after 3-indole treatment and whether ROS was involved in 3-indole induced MAPK activation. Cell lysates were subjected to Western blot analysis using anti-phospho-MAPK antibodies (ERK1/2, JNK, and p38) to detect phosphorylated activated MAPK family proteins.
The data in upper panel of Fig. 17 shows that 3-indole increased the protein level of phosphor-JNK1 in 4 h and phosphor-JNK2 in 8 h. In addition, we found that 3-indole increased the phosphorylation of c-Jun, a major nuclear factor of the JNK signaling pathway in 4-12 h. Furthermore, we co-treated A549 cells with rotenone (0.05 μM), SP600125 (20 μM, an inhibitor of JNK) or U0126 (10 μM, an inhibitor of ERK). The results indicated that co-treatment of 3-indole with rotenone or SP600125 reduced protein level of phosphor-JNK and c-Jun protein expression compared to 3-indole treatment alone (Fig. 17 lower panel). The data in Fig. 18 shows that treatment of A549 cells with a combination of the JNK inhibitor and 3-indole, caused a significant reduction in 3-indole-induced apoptosis when compared to the cells treated with 3-indole alone, whereas no effect of ERK inhibitor on 3-indole-induced apoptosis was seen. In addition, co-treatment with 3-indole and SP600125 reduced DNA damage compared to 3-indole treatment alone (Fig. 15 right panel).
The results indicated that inhibition of JNK activation protects against the cytotoxic effects of 3-indole and that ROS may play a role in JNK activation. In addition, we found that 3-indole decreased other growth singling pathway relation protein such as COX-2 and Akt (Fig. 19). The preliminary data indicated that that 3-indole is a multi-target inhibitor
compound for cancer treatment.
VIII. 3-indole Effectively Inhibited the Growth of Human A549 and H1435 Xenografts. To examine whether 3-indole treatment inhibited A549 cell growth in vivo, we followed the tumor growth in 3-indole and vehicle-treated animals (ICR-Foxn1). 3-indole was dissolved in Solvent (a vehicle mixture of DMSO/Cremophor EL/saline, 2:1:7) before further treatment. To further determine the effect of 3-indole over an extended treatment period, tumor size was measured in each animal. In the meantime, 3-indole-treated animals were sacrificed and processed for evaluation of any possible changes in histopathology and serum biochemistry.
Fig. 20 shows the tumor growth in solvent-treated animals (control) compared with 3-indole treatments. Treatment with 3-indole (final dose of 25 or 50 mg/kg i.p.) resulted in tumor growth inhibition, compared to that produced by solvent (control) treated animals bearing A549 cell xenografts (Fig. 20 upper panel). The same observations were also noted in a H1435 cell xenograft model (Fig. 20 lower panel). Evaluation of numerous histologic sections of these tissues from animals bearing human A549 xenografts did not indicate any detectable pathologic abnormalities, as examined by H&E staining (Fig. 21). In addition, 3-indole therapy caused no detectable toxicity on tissues and did not affect organ functions. The organ functions tests included liver function tests, such as glutamic oxalacetic transaminase (GOT), glutamic pyvuvic transaminase (GPT), and albumin levels, and renal function tests, such as blood urea nitrogen (BUN)
and creatinine levels. The organ functions were similar between the 3-indole-treated and the vehicle-treated groups (Fig. 22).
DISCUSSION
We evaluated the biological activities, especially the mechanisms, involved in the anti-cancer growth of 1,1,3-tri(3-indolyl)cyclohexane (3-indole) in cell and animal models. 3-indole caused a concentration-dependent reduction in cell viability. 3-indole could achieve an IC50 value at ~ 10 μM in various human NSCLC cells (A549, H1299, H1435, CL1-1, and H1437 cells), whereas did not show apparent cycotoxicity to the IMR-90 cells. The anticancer efficacy of 3-indole was also noted in various human esophageal squamous cell carcinoma cell lines. Furthermore, 3-indole caused most cancer cell lines, except for H1437, an accumulation in the G1 phase at a low concentration (10 μM), and increased in the sub-G1 region (an apoptosis indicator) at a high concentration (30 μM) at 24 h. Furthermore, the G2-M arrest by 48 h treatment with 30 μM 3-indole was also noted in H1437. The multi-effect of an anticancer drug concentration dependent on G1 or G2-M cell cycle arrest has also been shown for other compounds (Liebermann et al., 1995;
Giannakakou et al., 2001; Blajeski et al., 2002). Cells respond to DNA damage by activating cell cycle checkpoints. p53 is one of the most commonly mutated genes found in human tumors (Friend 1994;
Greenblatt et al., 1994). The function of p53 as a tumor suppressor has been demonstrated by experiments showing that the loss of p53 correlates with the loss of G1-S cell cycle transition regulation after DNA damage (Kastan et al., 1991; Park et al., 2001; Liu et al., 2003). The G2-M checkpoint induced by DNA damage can occur by either p53-dependent
or -independent through inhibition of RB phosphorylation mechanisms (Agarwal et al., 1995; Paules et al., 1995). Furthermore, the p21 protein at several sites in the cell cycle, targeting CDKs (4, 6, and 2), regulate cell cycle checkpoint can induces cell cycle arrest, is response to upregulation by p53 and by p53-independent mechanisms. In contrast to synthetic small-molecule compounds with an indole structure, such as vinorelbine, which induce almost complete G2-M arrest, 3-indole causes mainly G1 arrest.
A number of anti-microtubule chemotherapy compounds characterized by the presence of an indole core nucleus have been obtained (Brancale & Silvestri 2007). Microtubules are crucial in G2-M phase and cell division (Jordan & Wilson 2004; Pellegrini & Budman 2005). The mechanism of action of many anti-microtubule drugs is interference with the normal formation of the mitotic spindle by either increasing microtubule depolymerization or tubulin polymerization leading to cell cycle arrest (Sorger et al., 1997). The different sensitivity of tumor and normal cells to anti-microtubule agents could possibly be due to (a) deficient function of G1 checkpoint (Trielli et al., 1996) and (b) deficiency of p53 tumor suppressor genes (Di Leonardo et al., 1997) in tumor cells. Our preliminary data shows that 3-indole induce the transient activation of p53 in early time and subsequently p21 increase expression in A549 or H1437 cells. Furthermore, the G2-M arrest by 48 h treatment with 30 μM 3-indole was also noted in H1437 cells (p53-mutation).
3-indole can induce G2 checkpoint protein cyclin B1 expression increases in H1435 cells (p53-mutation). Indeed, it has been found that breast
cancer cells with inactivated p53 failed to arrest in G2-M after paclitaxel treatment (Bacus et al., 2001). Therefore, we hypothesized that difference activation of p21 expression pathway is required for the induction of cell cycle arrest in 3-indole treated lung cancer cells with differ p53 status.
Characterization of 3-indole-induced G2-M arrest in more cells with mutant p53 backgrounds with various treatment time of 3-indole is under investigation. In addition, microtubulin binding site of 3-indole will be further verified.
DNA ladders appeared in various human lung cells in a time-dependent manner after 3-indole treatment. Apoptosis is a major control mechanism by which cells die if DNA damage is not repaired.
Apoptosis occurs through two main pathways. The first pathway involves a member of the TNF receptor superfamily (extrinsic) and the second pathway involves the mitochondrial (intrinsic) pathway (Ghobrial et al., 2005). The Bcl-2 family of proteins constitutes a critical mediator in the mitochondrial pathway of apoptosis. Our results showed that treatment of A549 cells with 30 μM of 3-indole resulted in a time-dependent reduction in the levels of the anti-apoptotic Bcl-2 protein. Concomitantly, the level of pro-apoptotic Bax and Bad protein was increased. The decrease expression of anti-apoptotic Bcl-2 was also noted in a H1437 cells.
Furthermore, the progression of apoptosis involves the activation of a cascade of proteases called caspases. Theoretically, the extrinsic pathway is related to the activation of caspase -8 and the intrinsic pathway is associated with activation of caspase -9. Both pathways converge to a common pathway involving the activation of caspase -3. As shown in our
data, 3-indole apparently stimulated caspases -3, caspase -9 and to a lesser extent caspase -8 activities in A549 cells. Together, these results suggested that 3-indole induced the execution of apoptosis through the activation of the intrinsic mitochondrial pathway.
Various physical and chemical environmental stresses can activate apoptosis (Lavrik et al., 2005). One example of environmental stress-induced apoptosis is the loss of the MMP in cells and the subsequent induction of ROS by electron leakage from the mitochondrial electron transport chain. Various cancer cells have low-expression of some antioxidant enzymes (i. e. catalase and superoxide dismutase) (Ahmad et al., 2005), suggesting that induction of ROS in cancer cells may exhibit a potential target effect. Our data indicated that the MMP was decreased within 15-30 min in A549 cells and a significant increase in ROS production was observed by 2 ~ 8 h in various human lung cancer cells. Furthermore, the ROS induced by 3-indole can be partially reduced by an inhibitor of mitochondrial respiratory chain complex I and a hydroxyl radical scavenger. Considerable evidence indicated that ROS, as signaling transduction molecules, induced apoptosis by the mitochondria pathway and DNA damage activation. Therefore, we hypothesized that 3-indole may cause DNA damage. PFGE analyses showed that 3-indole triggers DNA strand breaks in treated-A549 cells in a time-dependent manner and the triggered DNA damage can be partially recovered by incubation of 3-indole treated cells with ROS inhibitor. In addition, 3-indole-induced apoptosis can be rescued by co-treatment with ROS inhibitors. Together, these results suggested that oxidative stress may
potentially trigger 3-indole-induced DNA damage and may lead to apoptosis.
ROS have been shown to induce various biological processes, including activation of the MAPK (Kamata et al., 2005; Gong et al., 2006). JNK-induced apoptosis has been shown to occur through the
ROS have been shown to induce various biological processes, including activation of the MAPK (Kamata et al., 2005; Gong et al., 2006). JNK-induced apoptosis has been shown to occur through the