Eugenol causes melanoma growth suppression through inhibition of E2F1 transcriptional activity.

Metastatic malignant melanoma is an extremely aggressive cancer, with no currently viable therapy. 4-Allyl-2-methoxyphenol (eugenol) was tested for its ability to inhibit proliferation of melanoma cells. Eugenol but not its isomer, isoeugenol (2-methoxy-4-propenylphenol), was found to be a potent inhibitor of melanoma cell proliferation. In a B16 xenograft study, eugenol treatment produced a significant tumor growth delay (p = 0.0057), an almost 40% decrease in tumor size, and a 19% increase in the median time to end point. More significantly, 50% of the animals in the control group died from metastatic growth, whereas none in the treatment group showed any signs of invasion or metastasis. Eugenol was well tolerated as determined by measurement of bodyweights. Examination of the mechanism of the antiproliferative action of eugenol in the human malignant melanoma cell line, WM1205Lu, showed that it arrests cells in the S phase of the cell cycle. Flow cytometry coupled with biochemical analyses demonstrated that eugenol induced apoptosis. cDNA array analysis showed that eugenol caused deregulation of the E2F family of transcription factors. Transient transfection assays and electrophoretic mobility shift assays showed that eugenol inhibits the transcriptional activity of E2F1. Overexpression of E2F1 restored about 75% of proliferation ability in cultures. These results indicate that deregulation of E2F1 may be a key factor in eugenol-mediated melanoma growth inhibition both in vitro and in vivo. Since the E2F transcription factors provide growth impetus for the continuous proliferation of melanoma cells, these results suggest that eugenol could be developed as an E2F-targeted agent for melanoma treatment.

Melanoma is one of the fastest growing cancers in the developing world with the incidence having tripled in the last three decades (1). Chemotherapy, immunotherapy, and vaccines have all produced limited benefits especially since the responses are typically short-lived, with no significant effect on overall survival. As a first step toward developing new com-pounds for effective melanoma management, we screened a panel of naturally occurring compounds for antiproliferative activity toward melanoma cells. From this screen, we have identified 4-allyl-2-methoxyphenol (eugenol) 1 as a potent inhibitor of both anchorage-dependent and anchorage-independent growth of melanoma cells representing the different stages of melanoma progression. The structures of eugenol and the isomeric isoeugenol are shown in Fig. 1.
Eugenol is found in reasonable quantities in the essential oils of different spices such as Syzgium aromaticum (clove), Pimenta racemosa (bay leaves), and Cinnamomum verum (cinnamon leaf). Eugenol has been used as an antiseptic, antibacterial, analgesic agent in traditional medical practices in Asia as well as in dentistry in cavity-filling procedures. Eugenol has been demonstrated to inhibit prostaglandin biosynthesis (2) and to block COX-2 activity with an IC 50 value of 129 mol (3). In long term carcinogenicity experiments by various groups in CD-1 mice and F344 rats, eugenol was not associated with tumor formation (4). Based on numerous long term carcinogenicity studies, the National Toxicology Program concluded that eugenol was not carcinogenic to rats and that there was no evidence that unequivocally proved the carcinogenic nature of eugenol in mice (National Toxicology Program). More recently, in a skin carcinogenesis study using the initiating agent 7,12-dimethylbenz(a)anthracene followed by three times weekly cutaneous applications of eugenol for 63 weeks in a group of female ICR/HA Swiss mice, no carcinomas were found (5). In a skin painting study by Van Duuren and Goldschmidt (6), eugenol was reported as being partially effective in inhibiting benzo(a)pyrene-induced skin carcinomas. Eugenol was shown to inhibit DMBA-croton oil-induced papillomas by about 84% (7). Eugenol is not mutagenic, although the incidence of sister chromatid exchange was found to increase in Chinese hamster ovary cells (8). Eugenol has neither been previously reported to be effective against melanoma nor been systematically tested in other common cancers.
The E2F proteins are a family of transcription factors with an important role in regulating cell cycle progression (9). It has been shown that deregulated transcriptional activity of the E2F family in autonomously growing melanoma cells provides the impetus for continuous proliferation of melanoma cells. Specifically, E2F2 and E2F4 are predominant in actively proliferating melanocytes, melanoma cells, and freshly isolated melanoma tumors. The up-regulated E2F activity in melanoma cells is dependent on persistent cyclin-dependent kinase (CDK) activity and inactivation of the pocket proteins (10). It has also been shown that the members of the E2F family known to cause growth arrest and apoptosis are either absent or expressed at low levels in melanoma cells, therefore providing a growth advantage to melanoma cells (10).
A clinical agent that can target the continuous cycling of melanoma cells would be an attractive tool for effective inhibition of melanoma cell growth. Our results show that eugenol is a potent inhibitor of both anchorage-dependent and anchorage-independent growth of melanoma cells in culture, causes significant tumor growth delay (p ϭ 0.0057), decreases size of tumors, and inhibits melanoma invasion and metastasis in B16 xenograft animals. Eugenol also arrests cells in the S phase of cell cycle, induces apoptosis, and inhibits E2F1 transcriptional activity. The growth inhibitory effect of eugenol is partially abrogated by overexpression of E2F1. These results suggest a potential role for E2F1 in eugenol-mediated melanoma management.

EXPERIMENTAL PROCEDURES
Materials-Eugenol was purchased from Sigma; all the polyclonal antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). CellTiter96 Aqueous One solution, DeadEnd colorimetric apoptosis detection system, and Dual-Luciferase assay system were from Promega Corp. (Madison, WI). All other reagents were molecular biology grade from Sigma. The E2F1 plasmids were a gift from Dr. David G. Johnson at University of Texas M. D. Anderson Cancer Center, Smithville, TX.
Cell Lines-The human melanoma cells that represent disease progression (Sbcl2-primary melanoma; WM3211-radial growth phase; primary RGP, WM98-1-radial and vertical growth phase; primary RGP and VGP, WM1205Lu-metastatic melanoma) were a gift from Dr. Meenhard Herlyn at the Wistar Institute in Philadelphia. Cells were cultured in Dulbecco's modified Eagle's medium containing 10% fetal calf serum and 0.5% insulin. All melanoma cell lines were maintained in a humidified incubator with 5% CO 2 at 37°C.
Animal Study-Female B6D2F1 mice bearing established B16 melanomas (ϳ50 mm 3 ) were randomized into a control and a treatment group of eight animals each at Piedmont Research Center in Morrisville, NC. The control group received corn oil (vehicle), and the treatment group received 125 mg of eugenol/kg of body weight twice a week intraperitoneally for the duration of the study. The dose was administered in a volume of 0.2 ml/20 g of bodyweight and was adjusted for the body weight of the animal. The choice of 125 mg/kg of bodyweight of eugenol was based on an earlier maximum tolerated dose study using three different doses. Animals were euthanized when their tumors reached the end point volume of 2,000 mm 3 , and the time to end point was calculated for each mouse. The percentage of tumor growth delay was used as treatment outcome (defined as the percentage of increase in median time to end point of treated versus control mice). Significance of efficacy was calculated using log rank analysis. All animal procedures were conducted in strict adherence to recommendations of the Guide for Care and Use of Laboratory Animals.
Proliferation Assay-Actively growing human melanoma cells were plated in 96-well plates at a density of 4 ϫ 10 3 cells/well in triplicates. After 24 h at 37°C with 5% CO 2 , cells were treated with different concentrations of eugenol (0.5-2.5 M), isoeugenol (0.5-5.0 M), or the solvent (ethanol). Cell proliferation following treatment was carried out with the CellTiter96 Aqueous One solution assay (Promega Corp.) as described elsewhere (11). Briefly, plated cells were incubated with the dye solution containing tetrazolium at 37°C for 4 h. The reaction was terminated with a stop solution that solubilizes the formazan product formed. Absorbance at 570 nm was recorded using a SpectraMaxPlus plate reader (Molecular Devices). Proliferation assays were performed five times in triplicate wells. The trypan blue exclusion assay was initially used to measure cell viability.
Colony Formation Assay-A colony-forming assay as described by Kumar et al. (12) was used. Logarithmically growing melanoma cells were trypsinized and plated at a density of 8,000 cells/ml in 0.5% agarose plates in duplicate. After incubating for 14 days, colonies were stained with 0.02% p-iodonitrotetrazolium. After 6 h, colonies containing more than 50 cells and stained dark pink were counted in eight different fields. The experiment was repeated twice in duplicate.
Flow Cytometric Analysis-Actively growing cells were plated at a density of 10 6 cells in 100-mm dishes. Cells at ϳ70% confluency were treated with either 0.5 M eugenol in ethanol or solvent (ethanol) for 20 and 36 h. Cells were harvested and resuspended in 1 ml of Krishan stain containing 1.1 mg/ml sodium citrate, 46 g/ml propidium iodide, 0.01% Nonidet P-40, and 10 g/ml RNase (13). These cells were subject to flow cytometric analysis on a Beckman Coulter XL flow cytometer (Beckman Coulter) at the University of Colorado Comprehensive Cancer Center Flow Cytometry core facility. Data analysis was carried out with the Modfit LT software (Verity Software House, Topsham, ME). Three independent flow cytometry analyses were carried out.
Apoptosis Detection-Melanoma cells treated with ethanol or 0.5, 1, and 2.5 M eugenol in ethanol for 18 h and observed by phase contrast microscopy. From this initial experiment, we chose to demonstrate the induction of apoptosis at the biochemical level using the malignant melanoma cell line WM1205Lu. The DeadEnd colorimetric apoptosis detection system (Promega Corp.) was used to detect apoptosis as described previously (11). This assay uses a modified terminal deoxynucleotidyltransferase-mediated dUTP nick end-labeling (TUNEL) method to detect apoptotic cells in situ. Cells were grown and treated as described above. Cells were washed and fixed on poly-L-lysine-coated slides in 2% formaldehyde solution. After permeabilizing cells with Triton X-100, double-stranded breaks were labeled with the biotinylated nucleotide mix and the terminal deoxynucleotidyltransferase enzyme. After the reactions were terminated, apoptosis was detected as the dark brown color of the horseradish peroxidase-labeled streptavidin bound to the biotinylated nucleotides. Two independent TUNEL assays were carried out with the DNase I-treated cells as positive control and a negative control without the terminal deoxynucleotidyltransferase enzyme. We also performed DNA fragmentation analysis to assess apoptosis induction by eugenol.
Whole Cell Extracts and Western Blotting-Cells were either treated with 0.5 M eugenol or left untreated for 18 h. Cells were harvested and lysed in a buffer containing 50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 0.5% Nonidet P-40, 50 mM NaF, 1 mM NaVO 4 , 1 mM phenylmethylsulfonyl fluoride, 25 g/ml leupeptin, 25 g/ml aprotinin, 25 g/ml pepstatin, and 1 mM dithiothreitol. Lysed cells were passed through a 25-gauge needle, and the released material was centrifuged at 12,000 rpm for 30 min. Protein content in the supernatant was determined by the method of Bradford (14). Western blotting was carried out as described elsewhere (11). Briefly, equal amounts of whole cell extracts were fractionated on a 10% SDS-polyacrylamide gel and transferred to a nitrocellulose membrane. Antibodies used were as follows: polyclonal anti-E2F1, -E2F2, -E2F3, -E2F4, -E2F5, and -E2F6. We used the Western Lightning chemiluminescence reagent Plus for detection according to the manufacturer's instructions (PerkinElmer Life Sciences). All the blots were probed with ␤-actin to normalize for loading differences. Western blotting was repeated twice with different batches of extracts.
Transfection and Transient Expression Assay-Transient transfection assays were performed in WM1205Lu cells using a Lipofectamine reagent (Invitrogen) as published previously (15). Wild type or mutant E2F1-luciferase reporter plasmids (1 g/well) and pRL-TK plasmid (50 ng/well) were incubated with Lipofectamine for 30 min and then added to the cells. 24 h after transfection, the cells were treated with 1 M eugenol in ethanol for 4 h. Following treatment, luciferase reporter assay was performed with the Dual-Luciferase reporter assay system from Promega Corp. As per the manufacturer's recommendation, cells were harvested and then lysed in passive lysis buffer. Cell lysate was cleared from debris by centrifugation at 10,000 rpm for 5 min at 4°C. Luciferase activity was assessed in duplicate samples containing equal amounts of protein. The assay contained 20 l of cell lysate and 50 l of firefly luciferase buffer (luciferase assay reagent II). Firefly luciferase was measured on the Victor plate reader. 50 l of Renilla luciferase (stop and glow) buffer was added, and Renilla luciferase activity was measured. Renilla luciferase activity was used to normalize transfection efficiency. Results are expressed as the ratio of firefly luciferase/ Renilla luciferase at equal amounts of protein.

Inhibition of E2F1 by Eugenol
E2F1, WM1205Lu cells in triplicates were transfected with 5 and 10 g of pCDNA3E2F1-overexpressing plasmid or pCDNA3 (control) plasmid. 24 h after transfection, the cells were treated with eugenol for 4 h. The number of live cells was counted in duplicates using the trypan blue exclusion assay. All transfection experiments were carried out twice each with two different plasmid preparations.
Gene Expression Analysis-WM1205Lu cells at 80% confluency were treated with 0.5 M eugenol in ethanol for 18 h. Total RNA was isolated with the TRIzol reagent (Ambion, Austin, TX) according to vendor instructions. To determine changes in gene expression following eugenol treatment, we used the cell cycle pathway-specific gene expression profiling system from SuperArray Bioscience Corp. (Frederick, MD). We made biotinylated cDNA from total RNA by RT-PCR according to instructions provided by SuperArray Bioscience Corp. The denatured probe was hybridized to the membranes containing 96 cDNA fragments involved in cell cycle regulation. Data were extracted from the raw image and analyzed with the GEArray analyzer software (Super Array Bioscience Corp. (Frederick, MD)). All raw signal intensities were corrected for background by subtracting the signal intensity of pUC18 (negative control) and normalized to Homo sapiens peptidylprolyl isomerase A; PPIA (housekeeping gene). The corrected and normalized signal was used to estimate the relative abundance of transcripts. The array results were verified by quantitative RT-PCR.
Preparation of Nuclear Extracts-Nuclear extracts were prepared as described elsewhere (16). Eugenol-treated (0.5 M) and control cells were harvested and homogenized in a buffer containing 10 mM Hepes, pH 7.9, 1.5 mM MgCl 2 , 10 mM KCl, 1.0 mM dithiothreitol, 1.0 mM phenylmethylsulfonyl fluoride, and protease inhibitors including 1 mM phenylmethylsulfonyl fluoride, 25 g/ml leupeptin, 25 g/ml aprotinin, 25 g/ml pepstatin. Nuclei were separated from cell debris and lysed in a buffer containing 20 mM Hepes, pH 7.9, 25% (v/v) glycerol, 0.42 M NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, 1.0 mM dithiothreitol, and the protease inhibitors. Nuclear protein was collected by centrifugation at 15,000 ϫ g for 45 min. Protein content in the supernatant was determined by the method of Bradford (14).
Immunocytochemistry and Immunohistochemistry-Untreated or 1 M eugenol-treated cells were fixed in paraformaldehyde on poly-Llysine-coated slides. For immunocytochemical detection of E2F1, we used 1:200 dilution of the primary antibody and followed the manufacturer's recommendations for the Histostain Plus kit (Zymed Laboratories Inc., San Francisco, CA). For immunohistochemical detection of E2F1 in melanoma tumor tissues, we deparaffinized tissue sections with xylene and rehydrated in a graded series of ethanol. After the appropriate washes in phosphate-buffered saline, immunochemical detection was carried out in the same way as cytochemical detection of E2F1. We used an E2F1 blocking antibody as a negative control for immunocytochemical as well as immunohistochemical detection of E2F1. The immunocytochemical assay was performed twice with cells from different passages. A Coolpix digital camera attached to a Nikon T1-SM or Zeiss microscope was used to obtain photographs of the cells and tissues.

Eugenol Inhibits Anchorage-dependent and -independent Growth of Melanoma Cells in Culture-
The CellTiter96 proliferation assay was used to measure proliferation of the cells for up to 72 h following the addition of increasing concentrations of eugenol. As shown in Fig. 2, A-D, eugenol addition inhibited the growth of all the human melanoma cell lines tested. The Sbcl2 and WM3211 cells showed 50% growth inhibition in 0.5 M eugenol after 24 h, whereas the WM98-1 and WM1205Lu cells needed twice as much time for 50% growth inhibition at this concentration of eugenol. At 72 h, however, there was no difference in response between the different cell lines using 0.5 M eugenol. Fig. 2E shows that isoeugenol, an isomer of eugenol, did not inhibit the growth of any of the human melanoma cell lines at concentrations up to 5 M.
Since anchorage-independent growth is a hallmark of cancer cells, we tested to see whether eugenol could inhibit colony formation on soft agar. Data presented in Fig. 2F show that eugenol inhibited colony formation in all the human melanoma cells. The metastatic melanoma cell line WM1205Lu showed the lowest inhibition in colony formation as compared with the other melanoma cells at 0.5 M eugenol. Taken together, these results show that eugenol inhibits both anchorage-dependent and anchorage-independent growth of melanoma cells.

Eugenol Causes Significant Tumor Growth Delay, Decreases Tumor Size, and Prevents Tumor Metastasis in B16F10 Xe
nograft Mice-We tested the effect of eugenol in vivo in the B16 melanoma xenograft model system as shown in Table I. As shown in Table II tumors in the vehicle group grew rapidly to the end point volume, and four mice died due to metastasis between days 8 and 14. The median time for the tumor to grow to end point for the control group was 12.6 days. There was a highly significant 19% tumor growth delay (p ϭ 0.0057) in the group treated with eugenol. The size of tumors in the treatment group was about 62% of that of the tumors of control animals on day 15. Very significantly, 50% of the animals in the control group developed non-treatment-related metastases, whereas none of the animals in the treatment group showed any signs of invasion or metastasis. Daily examination and body weight measurements of the animals showed no difference in body weight between control and treatment group animals, indicating that eugenol was well tolerated.
Tumor tissue (Fig. 3) from control group animals stained with hematoxylin and eosin (Fig. 3A) demonstrates numerous confluent collections of atypical cells with large pleomorphic nuclei, clumped chromatin, and atypical mitotic figures. Interspersed between the melanoma cells are numerous melanincontaining melanophages. Following treatment with eugenol (panel ii), diffuse areas of tumor necrosis intermixed with occasional small aggregates of damaged melanocytes that demonstrate nuclear pyknosis and condensation of the cytoplasm are seen. Diffuse areas of tumor necrosis with no viable-appearing melanoma cells are also seen. We used the modified TUNEL assay to determine whether the tumors in the treatment group were undergoing apoptosis. In the modified TUNEL assay, streptavidin-labeled dUTP is incorporated into the 3Ј-OH ends of apoptotic DNA by the enzyme terminal deoxynucleotidyltransferase to produce brown staining in cells undergoing apoptosis. As shown in Fig. 3B (panel i), tumor sections of the control animals showed negligible brown staining. On the other hand, in Fig. 3B, panel ii, intense brown staining is visible, indicating that eugenol treatment induces apoptosis in melanoma tumors.
Eugenol Blocks Cell Cycle Progression and Induces Apoptosis-Logarithmically growing WM1205Lu cells were treated with eugenol and subjected to flow cytometry analysis as described under "Experimental Procedures." Results in Fig. 4A show that there was a 40% increase in cells in the S phase accompanied by a decrease in the G 1 phase cells with no significant change in the G 2 /M phase cells following eugenol treatment. Further these data also show that the cells remain blocked in the S phase up to 36 h. Other human melanoma cells (Sbcl2, WM3211, and WM98-1) also showed a similar pattern of S phase block upon eugenol treatment (data not shown). Further we found the consistent appearance of a sub-G 1 peak during flow cytometry analysis that is indicative of apoptotic cells (data not shown).
We treated melanoma cells with 0.5, 1, and 2.5 M eugenol for 18 h and then observed the cells under a phase contrast microscope. Fig. 4B shows the morphological changes that occur in a representative cell line WM1205Lu following eugenol treatment. As shown in Fig. 4B, panels ii-iv, all the cells treated with eugenol showed blebbing of membranes, shrinkage of cytoplasm, and condensation of nuclear material, as well as gradually lifting off the dishes. These characteristic features of apoptosis occurred in a dose-dependent manner and was absent in untreated cells (Fig. 4B, panel i). To confirm the induction of apoptosis, we performed the modified TUNEL assay to detect apoptotic cells in situ using the metastatic melanoma cell line, WM1205Lu. As shown in Fig. 4C, panels ii-iv, increasing the concentration of eugenol from 0.5 to 2.5 M produced an increase in the number of brown stained cells, indicating that these cells were undergoing apoptosis. In Fig.   FIG. 2. Eugenol inhibits growth of melanoma cells. In A-D, the anchorage-dependent proliferation of cells representing the different phases of progression Sbcl2 (primary melanoma), WM3211 (primary radial growth phase), WM98-1 (primary vertical growth phase), and WM1205Lu (metastatic melanoma) are shown. Cells were treated with either ethanol or 0.5, 1, or 2.5 M eugenol as described under "Experimental Procedures." CellTiter96 Aqueous One solution assay was used to determine the conversion of tetrazolium salt into a formazan product in proliferating cells. Absorbance of the formazan product was measured at 570 nm every 24 h. E shows the effect of isoeugenol on human melanoma cell proliferation. Cells were treated with increasing concentrations of isoeugenol (from 0. 4C, panel i, the untreated control showed no staining. Cells treated with DNase I were used as a positive control (data not shown). A characteristic feature of apoptotic cells is the activation of endonucleases that attack internucleosomal DNA resulting in DNA fragments that are 180 -200 bp. We found that DNA isolated from eugenol-treated cells, but not from control cells, showed the laddering effect (data not shown). Taken together, the data presented in Fig. 4 clearly show that eugenol blocks cells in the S phase of the cell cycle and induces apoptosis in the human melanoma cells, WM1205Lu. It is not known, however, whether the S phase block is essential for induction of apoptosis.

Eugenol Modulates Expression of E2F Family Members-
Since we found that eugenol blocks cells in the replication phase of the cell cycle, we examined the cell cycle regulatory genes involved in the eugenol response using the pathwayspecific gene expression system as described under "Experimental Procedures." The raw data shown in Fig. 5A were corrected for background by subtracting the signal intensity of pUC18 (negative control) and normalizing to peptidylprolyl isomerase A (PPIA; housekeeping gene), and the relative abundance of transcripts between control and eugenol-treated samples was determined (Fig. 5B). To determine genes that are experimentally and biologically relevant, we filtered the data for those genes whose expression level increased or decreased at least 2-fold. Based on this analysis, we determined that members of the E2F family of transcription factors are modulated by eugenol treatment. Quantitative-RT-PCR with glyceraldehyde-3-phosphate dehydrogenase as an internal control was used to validate the array results. Results presented in Fig. 5C show the amplification products of the E2F family (E2F1-E2F6) along with glyceraldehyde-3-phosphate dehydrogenase (bottom band). Transcript abundance was calculated as  Photographs are at ϫ20 magnification taken through a Zeiss microscope using a CCD camera. Photographs shown are a representative of three independent experiments. the ratio of the target transcript to that of its internal control. A graphical representation of the relative abundance of transcripts in eugenol-treated and untreated cells is shown in Fig.  5D. We found that E2F1, E2F2, and E2F3 were all downregulated 2-fold or more following treatment with 1 M eugenol. E2F6 was the only E2F family member that was upregulated in response to eugenol treatment. These results were also validated by Western blotting using whole cell extracts from eugenol-treated and control cells. As shown in Fig. 5E, protein levels of E2F1, E2F2, and E2F3 were down-regulated by eugenol treatment, confirming the array data. The downregulation of E2F4 and E2F5 as well as the up-regulation of E2F6 were not discernable at the protein level.
We used immunocytochemical analyses to determine the level of E2F1 protein expression following eugenol treatment. Results shown in Fig. 6A show that more than 90% of the cells express E2F1 in untreated WM1205Lu cells (panel i), whereas the number of cells expressing E2F1 following 1 M eugenol treatment decreased to less than 1% (panel ii). Nuclear extracts made from control and eugenol-treated cells showed a 90% decrease in the E2F1 protein following eugenol treatment (Fig.  6B). This was confirmed in melanoma tumor samples by immunohistochemistry. E2F1 expression was found to be greater in the control animals (Fig. 6C, panel i) as compared with the treatment group (Fig. 6C, panel ii). A comparison of the data presented in Fig. 6, A and C, shows that E2F1 immunoreactive staining was more diffuse in tumor tissues as compared with tumor cells in culture, indicating the heterogeneous nature of tumor growth.
Eugenol Inhibits E2F1 Transcriptional Activity-Since E2F1 is a key regulator of genes involved in cell cycle progression, we determined whether eugenol affected the transcriptional regulation of E2F1. We used transient transfection assays to determine the effect of eugenol on E2F1 transcription activity as described under "Experimental Procedures." 24 h after transfection with wild type or mutant E2F1 reporter plasmid, WM1205Lu cells were treated with 1 M eugenol for 6 h. Normalized Renilla Luciferase activity presented in Fig. 7A shows that eugenol treatment decreased the E2F1 transcriptional activity by more than 50% as compared with that of untreated cells. Cells transfected either with the pGL3 control vector or with the E2F1 mutant reporter showed no transcriptional activity and no response to eugenol treatment. This result clearly demonstrates that eugenol inhibits transcriptional activity of E2F1. Overexpression of E2F1 Restores Melanoma Cell Proliferation-The observation that eugenol-induced inhibition of cell proliferation decreased E2F1 expression and its transcriptional activity suggested that E2F1 is involved in eugenol-mediated inhibition of melanoma cell proliferation. Therefore we determined whether overexpression of E2F1 would restore cell proliferation. Cells were transfected with increasing amounts of the E2F1 expression plasmid (pCDNA3E2F1) or the vector control. As shown in Fig. 8, vector control cells responded to eugenol treatment with about 90% decrease in cell viability. In cells that were transfected with increasing concentrations of E2F1 cDNA, the percentage of viable cells increased to about 75% of vector transfected cells treated with eugenol ( Fig. 8 and data not shown). This indicates that overexpression of E2F1 is able to reverse the growth inhibitory effect of eugenol in a dose-response manner.

DISCUSSION
Melanoma incidence rate has been increasing at the rate of 2.8%/year at a time when most other cancer incidence rates are either holding steady or declining. The target population for melanoma is fairly young, with people in their 40s and 50s being affected. Therefore it is estimated that the average productive life lost from melanoma is about 18.8 years. Treatment choices for malignant melanoma are very limited and hindered by the short response rates that do not affect overall diseasefree survival. In an effort to develop molecular mechanismbased treatment strategies, we have identified eugenol (a volatile oil isolated from culinary spices) as a compound with the ability to inhibit the growth of melanoma cells in culture. Our results show that eugenol inhibits both anchorage-dependent and anchorage-independent growth of human and mouse melanoma cells. Surprisingly, however, isoeugenol, an isomer of eugenol with similar structure and functionality, did not exhibit this antiproliferative activity. This difference in activity of the two related molecules may therefore be due to the differences in metabolism of the two isomers. In studies of skin sensitization to eugenol and isoeugenol in mice, it was concluded that eugenol was probably being demethylated and then oxidized to a reactive o-quinone, whereas isoeugenol was probably being directly oxidized to a p-quinone methide (17). The IC 50 for eugenol at 72 h varied from 0.3 to 0.5 M in human melanoma cells to 75 nM in the B16 mouse melanoma cells. Treatment of female BDF1 mice bearing B16 melanoma allografts with 125 mg of eugenol/kg of body weight resulted in a small but highly significant (p ϭ 0.0057) 2.4-day tumor growth delay. Furthermore the fact that the treated animals had no deaths that were attributed to metastasis or tumor invasion is FIG. 6. A, immunocytochemical detection of E2F1. Control and eugenol-treated WM1205Lu cells were fixed in paraformaldehyde, and immunocytochemistry was carried out as described under "Experimental Procedures." Brown staining in the left panel represents E2F1 expression. In the eugenol-treated cells in the right panel, there is very little staining, indicating that eugenol treatment decreases E2F1 protein expression. The experiment was repeated twice. Photographs shown are at ϫ40 magnification through a Nikon T1-SM microscope fitted with a Coolpix Nikon digital camera. B, detection of E2F1 protein in nuclear extracts. WM1205Lu cells were treated or not with eugenol as described for immunocytochemistry. Nuclear extracts were made and resolved by gel electrophoresis as described under "Experimental Procedures." Blots were probed with the E2F1 antibody, and the experiment was repeated twice with extracts from two different passage cells. C is control, and T is eugenol-treated. C, immunohistochemical detection of E2F1. Melanoma tumor sections from two different animals in the control and treatment groups were subject to immunohistochemistry as described. E2F1 staining in the tissue of the representative control animal is shown in the left panel. Tissue from a representative eugenol-treated animal is shown in the right panel. Photographs were taken with a Coolpix Nikon digital camera attached to a Zeiss microscope at ϫ20 magnification.

FIG. 7. Eugenol inhibits E2F1 transcriptional activity.
WM1205Lu cells were transfected with E2F1 wild type (wt) reporter or E2F1 mutant (mut) reporter plasmid and treated with eugenol for 4 h. Luciferase activity was measured and normalized to Renilla luciferase. The graph shows normalized luciferase activity in pGL3 (control plasmid) treated with no eugenol and with 1 M eugenol, E2F1 wild type untreated and treated with 1 M eugenol, and E2F1 mutant untreated and treated with 1 M eugenol. The data shown are an average of two independent experiments Ϯ S.D. with two different plasmid preparations each time. indicative of the ability of eugenol to suppress melanoma metastasis. This has great significance for melanoma management since it is metastatic malignant melanoma that is fatal, accounting for about 78% of melanoma deaths. Eugenol was also well tolerated in this study based on the fact that there were no treatment-related deaths and changes in body weight were comparable with the control group animals. Tumor size in eugenol-treated animals was about 40% smaller than animals in the control group. Staining of the tumor tissue from the control animals demonstrated extensive areas of viable melanoma cells with only focal areas of necrosis, whereas the tumors from the eugenol-treated animals demonstrated extensive areas of tumor necrosis with only a few small areas of viable tumor cells. Resistance of melanoma cells to apoptosis induction is a common problem associated with successful melanoma therapy. Our results show that eugenol induces apoptosis in melanoma tumors as demonstrated by the TUNEL assay data presented in Fig. 3B. All of these results taken together provide evidence for the potential development of eugenol as an agent for the treatment of melanoma.
Mechanistically, eugenol blocks cells in the replication phase, suggesting that cells stop to repair DNA damage and either re-enter the cell cycle or activate apoptosis in case of massive DNA damage. Results presented here show that melanoma cells treated with eugenol remain blocked in the S phase and undergo apoptosis. We also found that eugenol treatment up-regulated numerous enzymes involved in the base excision repair pathway (data not shown). Using cDNA array analysis, we found that the E2F family of transcription factors is differentially regulated in response to eugenol treatment. Under normal growth conditions, these factors regulate cell cycle transition through the activity of the CDKs by phosphorylating and inactivating the pocket proteins, the retinoblastoma tumor suppressor protein (pRb), p107, and p130. The cyclin-dependent kinase inhibitors negatively regulate the CDK holoenzymes and their partner cyclins by site-specific phosphorylation and dephosphorylation (18,19). Therefore the loss of these inhibitors facilitates the downstream hyperphosphorylation events that keep pRb phosphorylated and promotes uninterrupted cycling of cells through the cell cycle. Since the loss of functional p16 is common in melanoma (20 -24), pRb is either constitutively hyperphosphorylated or expressed at extremely low levels in melanoma cells (25,26). Hyperphosphorylation of the pocket proteins promotes the release of the E2F family of transcription factors. This leads to an increase in E2F transcriptional activity and up-regulation of genes involved in cell cycle progression, DNA synthesis, and transcription factors that participate in the induction of early and late responsive genes (27)(28)(29)(30)(31). E2F6, on the other hand, is not associated with the pocket proteins, and overexpression of E2F6 acts as a transcriptional repressor of E2F-responsive genes and arrests cell cycle transition (32)(33)(34).
It has been shown that deregulated E2F transcriptional activity in autonomously growing melanoma cells provides the impetus for the continuous proliferation of melanoma cells. Further it has been shown that the E2F2 and E2F4 proteins are predominant in actively proliferating melanocytes, melanoma cells, and freshly isolated melanoma tumors, that growth-arrested melanocytes manifested a E2F4-p130/pRb growth inhibitory complex, and that up-regulated E2F activity in melanoma cells was dependent on persistent CDK activity and inactivation of the pocket proteins. In addition, it has been found that members of the E2F family that are known to cause growth arrest and apoptosis, namely E2F1 and E2F6, are deregulated and therefore provide a growth advantage to melanoma cells (26). Therefore the ability of eugenol to disrupt the transcriptional activity of E2F1 demonstrated here underscores the importance of the potential use of eugenol in deregulating cell cycle progression in melanoma cells.
By overexpressing E2F1 in melanoma cells, we were able to restore the proliferation ability of the melanoma cells following eugenol treatment. We have, however, not identified other target(s) that work in concert with E2F1 to allow these melanoma cells to proliferate. Work is in progress to identify other members of the E2F family that may be involved and to identify the downstream targets of the E2F family that are involved in eugenol-induced cell cycle block and apoptosis.