Activation of Apoptosis by 1-Hydroxy-5,7-Dimethoxy-2-Naphthalene-Carboxaldehyde, a Novel Compound from Aegle marmelos

Apoptosis is a genetically regulated form of cell death, in which abnormal cells are eliminated from an organism, ensuring normal development of multicellular organisms and maintenance of tissue homeostasis (1). Therefore, an understanding of the molecular mechanisms of apoptosis is not only of scientific importance but may also have therapeutic value. A battery of signaling cascades is involved in the ability of the cell to undergo apoptosis. One such is a family of cysteine proteases, caspases, whose function in apoptotic cell death has been studied in detail (2) and are important mediators of apoptotic processes caused by various inducers (3). They amplify the apoptotic signal and proteolytically process numerous cellular target molecules with different functions (4).

There are two major apoptotic pathways in mammalian cells (5, 6). The extrinsic pathway is initiated by the binding of transmembrane death receptors [e.g., Fas, tumor necrosis factor receptor 1 (TNFR1), and TNF-related apoptosis-inducing ligand (TRAIL) receptors] with cognate extracellular ligands (7). Ligand receptors recruit adaptor proteins [e.g., TNFR-associated death domain (TRADD) and FADD], which interact with and trigger the activation of caspase-8. Activated caspase-8 then cleaves and activates downstream effector caspases, such as caspase-3. In contrast, the intrinsic pathway is characterized by disruption of mitochondrial membrane integrity when cells are exposed to various stresses (e.g., DNA-damaging agents; ref. 8). Mitochondrial membrane integrity triggers apoptosis via both caspase-dependent (e.g., the cytochrome c–caspase-9 pathway) and caspase-independent (e.g., the apoptosis-inducing factor pathway) mechanisms. Cross-talk exists between the extrinsic and intrinsic pathways, as activated caspase-8 can cleave Bid to produce truncated Bid (t-Bid), which then binds to mitochondria and promotes mitochondrial membrane integrity (9, 10). The subsequent release of cytochrome c from mitochondria further facilitates the apoptotic process (11–13).

The importance of cell signaling cascades in normal cell regulation and during disease condition has been realized (14–16), and many well-known targets at the signaling levels have been identified that are critical for the ability of cancer cells to proliferate rapidly. Furthermore, the availability of high-throughput assays based on the above-said molecular targets has further enhanced the process of drug discovery (17). Most anticancer drugs are natural products or their semisynthetic analogues; well known examples include Taxol, epothilone, camptothecin, daunomycin, and Vinca alkaloids (18, 19). Some of the plant-derived anticancer compounds are beginning to enter clinical trials (19).

Here, we report the identification of a novel compound, 1-hydroxy-5,7-dimethoxy-2-naphthalene-carboxaldehyde (HDNC), from Aegle marmelos, a medicinal plant from India used for treatment of cancer-related symptoms in the Indian system of medicine (20). As a common term, we have named it marmelin. HDNC inhibits the growth of HCT-116 colon cancer cell tumor xenografts and angiogenesis in nude mice. In addition, HDNC suppresses cancer cell proliferation while inducing apoptosis through the induction of activation of TNF-α, TNFR1, TRADD, caspase-8, and t-Bid, resulting in cytochrome c release and activation of caspase-3.

Cells and reagents. HCT-116, HT-29 colon, AGS gastric, and HEp-2 alveolar cancer cells obtained from the American Type Culture Collection were grown in DMEM and RPMI 1640 containing 10% heat-inactivated fetal bovine serum (Sigma Chemical Co.) and 1% antibiotic-antimycotic solution (Mediatech, Inc.) at 37°C in a humidified atmosphere of 5% CO₂. The BD RiboQuant multiprobe template set for hCK-3 and hAPO-3c was obtained from BD PharMingen. Caspase-3 Inhibitor IV (Ac-DMQD-CHO) was obtained from EMD Biosciences. Antibodies were obtained from Cell Signaling. The medicinal plant A. marmelos was obtained from south India and the species was authenticated (21, 22).

HCT-116 cell xenograft tumors in mice. Five-week-old male athymic nude mice (The Jackson Laboratory) were maintained with water and standard mouse chow ad libidum and used in protocols approved by the Animal Studies Committee of the University. Animals were injected with 1 × 10⁶ HCT-116 cells in the left and right flank with 100 μL Matrigel and allowed to form xenograft. Marmelin (HDNC, 200 μg/kg body weight) in 5% Na₂HCO₃ buffer alone was administered i.p. daily for 23 d. Tumors were measured weekly with a Vernier caliper and tumor volumes were calculated according to the formula length × width × depth × 0.5236. At the end of treatment, the animals were sacrificed, and the tumors were removed and weighed for use in histology (H&E and CD31) and gene expression studies.

Proliferation and apoptosis assays. Cells were grown in 96-well plates and treated with increasing doses of extracts or purified HDNC, and proliferation was assessed by hexoseaminidase assay (23). The plates were read at 405 nm in a Synergy HT microtiter plate reader (Bio-Tek Instruments, Inc.). For apoptosis, caspase-3/7 activity was measured using the Apo-ONE Homogeneous Caspase-3/7 Assay kit (Promega Corp.).

Cell cycle analysis. Cells were plated at a density of 5 × 10⁵ per well on six-well plates. After treatment for 24 h, cells were collected into flow cytometry tubes and stained with propidium iodide at 4°C for 30 min in the dark. Cell cycle analysis was performed with a Becton Dickinson FACScan using an FL2 detector with a bandpass filter at specifications of 585 ± 21 nm. In each analysis, 10,000 events were recorded.

RNA preparation and RNase protection assay. RNA was isolated from the cells using Trizol method (Invitrogen); 20 μg/lane were loaded for RNase protection assay (RPA) using radiolabeled RiboQuant multiprobe template set for human cytokines (hCK-3) or human apoptosis-related genes (hAPO-3c), according to the manufacturer's protocol.

Immunoblot analysis. Total lysates from cells or from tumor xenografts and nuclear extracts were prepared and separated on 10% SDS-polyacrylamide gel and blotted onto Immobilon polyvinylidene difluoride membranes (Millipore). Membranes were incubated overnight with the indicated primary antibody followed by secondary antibody (1:5,000) for 1 h. Specific proteins were detected by the enhanced chemiluminescence system (Amersham Pharmacia Biotech).

ELISA assay. TNF-α levels in the medium were determined by sandwich ELISA assays, using ELISA kits, according to the manufacturer's protocol (Pierce Biotechnology). The TNF-α ELISA kit has a detection range of 15.6 to 1,000 pg/mL and has a sensitivity of <2 pg/mL. Briefly, the samples are incubated with the precoated wells followed by biotinylated second antibody. The plates were subsequently incubated with streptavidin-horseradish peroxidase reagent and 3,3′,5,5′-tetramethylbenzidine substrate solution. The absorbance was read at 450 and 550 nm using the Synergy HT plate reader.

Electrophoretic mobility shift assay analysis. HCT-116 cells were pretreated with the HDNC for 2 h before coincubation with TNF-α (10 ng/mL) for 1 h. Nuclear extracts were prepared using the NE-PER kit (Pierce Biotechnology). Protein concentrations were determined using bicinchoninic acid reagent (Pierce Biotechnology). Nuclear extracts (6 μg) were incubated with ³²P-labeled double-stranded oligonucleotide containing a consensus NF-κB site (5′-AGTTGAGGGGACTTTCCCAGGC-3′; binding site in bold and underlined) at room temperature for 30 min. The protein-DNA complexes were size separated in a 4% native PAGE in 0.5% Tris-borate EDTA buffer and subjected to autoradiography.

Luciferase assay. HCT-116 cells were seeded in six-well dishes and incubated until 80% confluent. Then, the cells were transfected with 3 μg pNFκB-Luc (Clontech) reporter plasmid using Fugene 6 reagent (Roche Diagnostic). After 4 h of incubation, the cells were treated with extracts or purified HDNC in the presence or absence of TNF-α for 24 h. Cells were lysed and luciferase levels were determined using Luciferase Assay System (Promega).

Statistical analysis. All values are expressed as the mean ± SE. Data were analyzed using a paired two-tailed t test. A P value of <0.05 was considered statistically significant.

HDNC inhibits cancer cell proliferation. Recent studies have shown that administration of hydroalcoholic extract of A. marmelos to Ehrlich ascites carcinoma-bearing Swiss albino mice resulted in a dose-dependent remission of the tumor (21). Furthermore, a recent study showed no toxicity in rats administered extracts of leaves from this plant (22). These data, taken together, suggest that the plant encodes antitumor activity with a low toxicity profile. Accordingly, we aimed to identify an active compound in A. marmelos that is a potent inhibitor of solid tumors. Powdered A. marmelos plant material was extracted with ethanol and fractionated into hexane, dichloromethane, ethyl acetate, and n-butanol soluble fractions (data not shown). The effect of the solvent fractions on the proliferation of cancer cells was evaluated for up to 72 h. Dose-response analysis revealed 10 μg/mL of ethyl acetate fraction (Fig. 1A). The active ethyl acetate fraction was subsequently fractionated by silica column chromatography using a combination of hexane and ethyl acetate solvents. According to their TLC profiles, they were pooled into 12 fractions (data not shown). Fraction 5 showed maximum activity in 24 h. This compound was characterized as HDNC with a molecular mass of 232.23 (Fig. 1B) by nuclear magnetic resonance (NMR) and mass spectrometry (MS) analysis. We have also given the compound a common name marmelin because it was isolated from A. marmelos. Dose-response analysis revealed significant reduction in cell proliferation of multiple cancer cells at a dose of 50 μmol/L HDNC (Fig. 1A; Supplementary Fig. S1). In contrast, HDNC did not inhibit proliferation of normal mouse embryo fibroblasts for doses up to 500 μmol/L (data not shown). Together, these data show that the purified HDNC is a potent antiproliferative molecule with activity against epithelial cancer cells of different origin.

Caspase-3 activation is essential for HDNC-mediated cell apoptosis. Caspase-3 is a key effector in the apoptosis pathway, amplifying the signal from initiator caspases, such as caspase-8 (24, 25). To determine whether cells undergo apoptosis, we first performed a caspase-3/7 assay after the cells were incubated with purified HDNC. There was a dose-dependent increase in caspase-3/7 activity, suggesting that the compound induces cells to undergo apoptosis (Fig. 1C,, top). To further determine whether the apoptotic activity is due to caspase-3, Western blot analysis was performed with extracts from HEp-2 cells. Significant induction in caspase-3 activation was observed in cells treated with either the ethyl acetate fraction or purified HDNC (Fig. 1C,, bottom). To confirm that caspase-3 activation is required for HDNC-mediated apoptosis, cell cycle was performed. A significant increase in cells in sub-G₁ phase was observed when the cells were treated with either the ethyl acetate fraction (10 μg/mL) or purified HDNC (50 μmol/L; Fig. 1D). Furthermore, coincubation with a specific inhibitor of caspase-3, Ac-DMQD-CHO, resulted in a significant reduction in cells in sub-G₁ phase. Similar results were obtained with HCT-116 and HT-29 cells (data not shown). These data suggest that HDNC-mediated induction of apoptosis is mediated by caspase-3 activation.

HDNC inhibits tumor growth. To evaluate the role of HDNC in tumor proliferation in vivo, we examined the ability of the compound to suppress the growth of human cancer cell xenografts in nude mice. Colon cancer cell–induced xenograft tumors were allowed to develop and grow to a size of 500 mm³, following which HDNC was administered i.p. for 3 weeks daily. HDNC significantly inhibits the growth of the tumor xenografts (Fig. 2A). Whereas the control tumors grew during the treatment period, reaching a size of 2,000 mm³, tumors in the HDNC-treated animals were smaller throughout. The excised tumors from the HDNC-treated animals averaged ∼1.3 g, whereas those from the control group weighed between 2.0 and 2.5 g (Fig. 2C). In addition, tumor volumes were significantly decreased (Fig. 2B and C). There was no apparent change in liver weight, spleen weight, or body weight in the wild-type animals treated with HDNC (data not shown). These data imply that HDNC is a potential therapeutic for treatment of colon cancers and that it is relatively nontoxic to mice.

HDNC inhibits angiogenesis. A key feature of tumors is the development of microcapillaries. Given the reduction in the size of the tumors following HDNC treatment, we determined the effect of the compound on tumor vascularization by staining for the endothelial-specific antigen CD31. HDNC treatment leads to lower levels of CD31 staining and to the obliteration of the normal vasculature that is associated with tumor angiogenesis (Fig. 2D). Furthermore, HDNC treatment resulted in a significant reduction in microvessel density (Supplementary Fig. S2). Two key factors that regulate capillary growth in the tumors are vascular endothelial growth factor (VEGF) and interleukin-8 (IL-8), and prostaglandins and the other tumors promoting mediators are known to induce their expression in epithelial cells. Prostaglandin production is regulated by cyclooxygenase-2 (COX-2), a protein that is overexpressed at multiple cancers (26–28). Therefore, we next determined the effects of HDNC treatment on COX-2, VEGF, and IL-8 expression. HDNC treatment reduced COX-2, VEGF, and IL-8 mRNA levels in the tumor xenografts (Supplementary Fig. S3).

HDNC induces TNF-α, which exerts its activity through recruitment of TRADD. TNF-α is pleiotropic cytokine involved in inflammatory and apoptotic responses. To determine the effect of HDNC on TNF-α expression, we performed a RPA analysis. TNF-α levels were significantly up-regulated in HEp-2 cells following treatment with HDNC (Fig. 3A). Similarly, incubation of HCT-116 and AGS cells resulted in increased expression of TNF-α at the RNA and protein levels (Fig. 3B and C). Binding of TNF-α to its cognate receptor TNFR1 results in receptor trimerization and subsequent binding to an adapter molecule, TRADD. TRADD in turn recruits TNF-associated factor 2, and subsequent activation of the nuclear factor-κB (NF-κB) and c-Jun NH₂-terminal kinase (JNK) pathways. In addition, activated TRADD has been shown to recruit a second adaptor (FADD), which in turn activates caspase-8. To determine whether these proteins are affected by HDNC, a RPA analysis with TNF signaling templates was performed. RPA analysis revealed the activation of TNFR1 (p55) receptor, TRADD, and Fas. Further, we also observed the activation of caspase-8 (Fig. 4A). Immunoblot analysis also showed activation of TRADD on treatment with both ethyl acetate fraction and HDNC compared with the untreated controls (Fig. 4B). Similarly, activation of caspase-8 was confirmed (Fig. 4B).

HDNC induces Bid cleavage leading to cytochrome c release. Activated caspase-8 can cleave Bid, a Bcl2-interacting protein, at the COOH terminus. The activated t-Bid subsequently translocates to the mitochondria where it triggers cytochrome c release (29). Because caspase-8 was induced following HDNC treatment, we next analyzed whether Bid is cleaved. Western blot analysis showed significant levels of t-Bid (Fig. 4B). Furthermore, activation of cytosolic cytochrome c was observed following HDNC treatment (Fig. 4B). This observation further strengthens the hypothesis that a cross-talk exists between intrinsic and extrinsic pathways in case of HDNC-induced apoptosis.

Suppression of AKT and extracellular signal-regulated kinase activation in colon cancer cells. AKT is a serine/threonine kinase involved in cell proliferation, survival/apoptosis inhibition, and angiogenesis (30, 31). Similarly, the p42/p44 extracellular signal-regulated kinase (ERK) mitogen-activated protein kinase (MAPK) signaling pathway controls cell growth, differentiation, and survival (32). Because TNF-α modulates cellular responses through MAPK to induce NF-κB, we next determined whether HDNC-mediated induction of TNF-α resulted in increased AKT and/or ERK activity. Western blot analyses showed TNF-α–mediated increase in AKT and ERK phosphorylation (Fig. 5A and B). However, pretreatment with HDNC or ethyl acetate fraction inhibited TNF-α–mediated induction of AKT and ERK phosphorylation (Fig. 5A and B; Supplementary Fig. S4). Further confirmation that HDNC inhibits AKT and ERK phosphorylation was obtained when extracts of tumor xenografts were subjected to Western blot analyses (Fig. 5C and D; Supplementary Fig. S4). Taken together, these data suggest that HDNC is a potent suppressor of TNF-α–mediated signaling.

Suppression of TNF-α–induced NF-κB activation in colon cancer cells. A key downstream target of TNF-α–mediated NF-κB activation is IL-8 (33–35). To determine the effect of HDNC on TNF-α–mediated NF-κB activation, we first determined the expression of IL-8 in HCT-116 cells. Although HDNC did not affect baseline IL-8 expression, it inhibited TNF-α–mediated induction of IL-8 expression in the cells (Fig. 6A). To determine if this is due to loss of NF-κB activation, we performed an electrophoretic mobility shift assay (EMSA) analysis with nuclear extracts of HCT-116 cells. HDNC suppressed TNF-α–induced NF-κB activation (Fig. 6B). Further confirmation of this inhibition was shown by Western blot analysis for RelA, the p65 component of NF-κB. RelA levels were significantly reduced in the nucleus of cells treated with HDNC even in the presence of TNF-α (Fig. 6C,, left). This was even more evident when the expression was quantitated and compared with controls (Fig. 6C,, right). Finally, to show that HDNC inhibits NF-κB activation, we transiently transfected HCT-116 and HT-29 cells with a luciferase reporter gene plasmid under the control of a minimal TATA promoter fused to NF-κB response element. NF-κB transcriptional activity was attenuated in both cell lines treated with either the extract or HDNC when compared with the buffer-treated controls. Furthermore, they also inhibited TNF-α–induced NF-κB transcriptional activity (Fig. 6D). Because TNF-α–mediated inflammatory and pro-proliferative responses involve activation of NF-κB, these data suggest that HDNC enhances TNF-α–mediated apoptotic response in cancer cells by inhibiting the NF-κB activation pathway.

The ability of cancer cells to avert the apoptotic program has been identified as one of the major mechanisms for the development of cancers (15). In a previous study, extracts from A. marmelos were found to inhibit the proliferation of multiple human tumor cell lines in in vitro assays, including the leukemic K562, erythroleukemic HEL, melanoma Colo38, and breast cancer MCF7 and MDA-MB-231 cell lines (36). In the present study, we have also shown that the ethyl acetate fraction of A. marmelos significantly inhibits proliferation of many cells (Fig. 1A). Subsequently, we have identified a single molecule, which was structurally characterized as HDNC, by detailed NMR and MS analysis (Fig. 1B). This pure compound was also capable of inducing apoptosis to a variety of cell lines. Our in vivo studies have further shown that HDNC inhibits the growth of HCT-116 colon cancer xenografts in nude mice, with a significant reduction in the number of microcapillaries (Fig. 2). Furthermore, we have determined that HDNC acts to inhibit tumor growth by inducing TNF-α–mediated apoptosis program.

Two major apoptotic pathways exist in mammalian cells (5). The extrinsic pathway is initiated by the binding of transmembrane death receptors (e.g., Fas, TNFR1, and TRAIL receptors) with cognate extracellular ligands. Ligand-bound receptors recruit adaptor proteins (e.g., FADD), which interact with and trigger the activation of caspase-8. Activation of TNFR p55 expression by both the crude fraction and pure HDNC and activation of TNF-α imply that this compound was capable of inducing apoptosis. TNF-α is known to trigger the expression of a large battery of genes in diverse cell types (37–39), including its specific role in triggering apoptotic death via TNFR1 receptor (40). Our studies show that the presence of the compound causes significantly higher expression levels of TNF-α in the cells coupled with overexpression of TRADD driving the cell into apoptosis via TNF-α–TRADD cascade. In this regard, it should be noted that TNFR1 has been the only receptor expressed in HEp-2 cells (41). However, the colon cancer cells used in the study do express both TNF receptors, but only TNFR1 was found to be induced, further strengthening our conclusion that HDNC acts through the TNFR1 receptor to induce apoptosis.

The intrinsic pathway is characterized by disruption of mitochondrial membrane integrity when cells are exposed to stresses such as DNA-damaging agents (6). Disruption of mitochondrial membrane integrity can trigger apoptosis via both caspase-9–dependent pathway that includes the truncation of Bid protein, which then binds to mitochondria and promotes mitochondrial membrane integrity disruption. Although it is not known whether HDNC can activate caspase-9, our findings suggest the possibility of a cross-talk between the two pathways, as activated caspase-8 can cleave Bid to produce t-Bid.

NF-κB/Rel transcription factors can block apoptosis induced by TNF-α (37, 42, 43). The antiapoptotic activity of NF-κB is also crucial for immunity, tumorigenesis, and cancer chemoresistance (37). With respect to TNF-α, the NF-κB–mediated suppression of apoptosis involves inhibition of JNK cascade. The control of TNFR-induced killing serves as paradigm for the antiapoptotic activity of NF-κB (44, 45). TNF-α normally has less or no cytotoxic effect unless NF-κB activation or protein synthesis is blocked (45). In line with this, EMSA and reporter gene analyses reveal suppression of TNF-α–induced NF-κB activation in colon cancer cells. We infer that this may aid in TNF-α–induced apoptosis by HDNC. However, further studies are required to determine whether knockdown of the NF-κB activation is essential for HDNC activity.

Caspases are activated in a branched protease cascade and control distinct downstream processes in TNF-α–induced and Fas-induced apoptosis (4, 46). To confirm whether induction of apoptosis is due to caspase-3 activation, inhibitor studies using specific caspase-3 reveal marked reduction in the induction of HDNC-mediated apoptosis. It would, however, also be interesting to determine whether additional pathways exist for HDNC-mediated caspase-3 activation by using dominant-negative mutant forms of caspase-8 and caspase-9. Nevertheless, the ability of the compound HDNC from A. marmelos to inhibit cell proliferation and induce apoptosis by a multiprolonged cascade of events suggests that it might be an ideal therapeutic for many solid tumors.

No potential conflicts of interest were disclosed.

Grant support: Department of Biotechnology, Government of India (A. Balakrishnan) and NIH grants CA109269 (S. Anant) and AT004118 (R.P. Ramanujam).

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.