α- and β-Santalols Directly Interact with Tubulin and Cause Mitotic

May 20, 2015 - Southwest Research Institute, 6220 Culebra Road, San Antonio, Texas 78238, United States. § Santalis Pharmaceuticals, 18618 Tuscany ...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/jnp

α- and β‑Santalols Directly Interact with Tubulin and Cause Mitotic Arrest and Cytotoxicity in Oral Cancer Cells Brigette Lee,† Jonathan Bohmann,‡ Tony Reeves,‡ Corey Levenson,§ and April L. Risinger*,† †

Department of Pharmacology, University of Texas Health Science Center, San Antonio, Texas 78229, United States Southwest Research Institute, 6220 Culebra Road, San Antonio, Texas 78238, United States § Santalis Pharmaceuticals, 18618 Tuscany Stone, Suite 100, San Antonio, Texas 78258, United States ‡

S Supporting Information *

ABSTRACT: Head and neck squamous cell carcinoma (HNSCC) is the sixth most common cancer worldwide, with no major advancements in treatment over the past 40 years. The current study explores the biological effects of East Indian sandalwood oil (EISO) and its two major constituents, α- and β-santalol, against a variety of HNSCC lines. All three agents exhibited cytotoxic effects and caused accumulation of cells in the G2/M phases of the cell cycle. Additionally, treatment with these agents caused formation of multipolar mitotic spindles similar to those observed upon treatment of cells with compounds that affect microtubule polymerization. Indeed, the santalols, as well as EISO, inhibited the polymerization of purified tubulin, indicating for the first time that these compounds have the ability to directly bind to tubulin and affect microtubule formation. Modeling studies suggest that the santalols can weakly bind to the colchicine site on tubulin, and topical administration of EISO to a HNSCC xenograft inhibited tumor growth with no observed toxicities. Therefore, santalols can directly interact with tubulin to inhibit the polymerization of microtubules, similarly to established classes of chemotherapeutic agents, albeit with greatly reduced potency that is not associated with the classic toxicity associated with most other compounds that interact directly with tubulin.

S

andalwood oil is the essential oil derived from the Santalum album L. tree, more commonly known as East Indian sandalwood. Sandalwood has been in use for over 4000 years; mentions of the tree have been discovered in Sanskrit writings, and the ancient Egyptians used the tree for embalming, ritual burning, and medicinal purposes. The essential oil derived from the tree is considered safe and is listed by the FDA as a natural flavoring agent for human consumption.1 It is also used heavily in the perfume industry due to its pleasant fragrance.2 Additionally, East Indian sandalwood oil (EISO) has a host of medicinal properties. EISO is used to treat the common cold, bronchitis, fever, and urinary tract infections,1 and it has also been shown to have anti-inflammatory properties3 and cause autophagy in proliferating keratinocytes.4 Moy et al. demonstrated in clinical trials that a mixture of 0.5% salicylic acid and 2.0% EISO reduces acne in patients.5 The major components of EISO are α-santalol and β-santalol (Figure 1), which contribute to the oil’s pleasant fragrance. Together, these two components comprise up to 80% of the essential oil obtained from the East Indian sandalwood tree.2 The purified santalols have also demonstrated medicinal properties. β-Santalol has antiviral effects,6 while α-santalol has been studied for its cytotoxic qualities. In particular, αsantalol reduces angiogenesis and growth of human prostate tumors7 and has also been shown to induce apoptosis by activating caspases and increasing p53 expression.8,9 Additionally, it causes G2/M cell cycle arrest in breast and skin cancer © XXXX American Chemical Society and American Society of Pharmacognosy

Figure 1. Structures of cis-α-santalol (A) and cis-β-santalol (B).

cell lines10,11 and has chemopreventive activity in several models of skin cancer.12−15 Head and neck squamous cell carcinoma (HNSCC) is the sixth most common cancer worldwide, with approximately 600 000 new cases occurring every year. It carries a low fiveyear survival rate of 40−50%.16 Risk factors include tobacco Received: March 6, 2015

A

DOI: 10.1021/acs.jnatprod.5b00207 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

Table 1. GI50, TGI, and LC50 Values for α-Santalol, β-Santalol, and EISO in HNSCC Cell Lines α-santalol (μM) GI50a SCC-4 CAL27 HSC-3 SCC-9 SCC-25 HN5 HSC-2

4.5 5.5 3.9 3.4 3.5 5.3 4.6

± ± ± ± ± ± ±

0.4 0.6 0.5 0.8 0.3 0.6 0.8

TGIb 7.7 7.6 5.4 5.3 6.0 8.2 9.2

± ± ± ± ± ± ±

0.5 0.5 0.5 0.7 0.4 0.9 1.0

β-santalol (μM) LC50c 14.0 11.0 7.7 9.0 11.6 12.2 10.8

± ± ± ± ± ± ±

0.8 0.7 0.4 0.6 1.7 0.8 0.9

GI50a 5.8 5.7 4.4 4.2 4.2 6.3 7.3

± ± ± ± ± ± ±

TGIb

0.3 0.6 0.5 1.0 0.4 0.5 1.4

10.0 8.4 6.1 6.8 6.3 9.6 9.4

± ± ± ± ± ± ±

0.6 0.7 0.5 0.8 0.5 0.5 2.4

sandalwood oil (μg/mL) LC50c 14.3 12.6 9.1 14.5 11.1 15.9 11.9

± ± ± ± ± ± ±

0.5 0.8 0.4 1.8 0.5 1.5 2.7

GI50a 1.2 1.3 1.0 0.8 1.0 1.5 1.6

± ± ± ± ± ± ±

0.1 0.2 0.1 0.1 0.1 0.2 0.2

TGIb 2.2 1.7 1.4 1.4 1.4 2.1 2.1

± ± ± ± ± ± ±

0.2 0.2 0.1 0.1 0.1 0.1 0.3

LC50c 3.0 2.4 1.9 2.8 2.4 3.0 2.6

± ± ± ± ± ± ±

0.2 0.1 0.1 0.3 0.1 0.3 0.3

a

Concentration that inhibited 50% cell growth. bConcentration that caused total growth inhibition. cConcentration that caused 50% cytotoxicity. Data are from at least three independent experiments, each run in triplicate ± SEM.

Figure 2. α-Santalol, β-santalol, and EISO cause G2/M accumulation. The cell cycle distribution of SCC-4 cells treated with DMSO vehicle, 15 nM paclitaxel, or 1−4× the LC50 concentration of α-santalol, β-santalol, or EISO for 18 h was determined by performing flow cytometry of propidium iodide-stained cells.

and alcohol use17 as well as infection with human papilloma virus (HPV).18 Interestingly, HPV-positive HNSCC patients have a better survival rate.18 Historically, the early stage form of this disease was treated with surgery, sometimes along with radiotherapy. However, this treatment can result in increased difficulty speaking and swallowing. As the disease becomes more advanced, treatments for oral cancers include a combination of surgery, radiotherapy, and chemotherapy. Many drugs are used as chemotherapeutic agents, including cisplatin and cetuximab, an EGFR inhibitor, which is the only targeted therapy currently approved for oral cancer treatment.19 Despite advancements in disease prognosis and treatment, there has been little to no improvement in the treatment of HPV-negative HNSCC tumors in the past 30 years.18 Therefore, there is a need for the development of better treatments for HNSCC. In the current study, each of these compounds was found to be cytotoxic against HNSCC lines,

causing G2/M accumulation and the formation of aberrant mitotic spindles. Most importantly, we demonstrate for the first time that the santalols can directly bind to tubulin to inhibit microtubule polymerization, a well-established mechanism of anticancer activity, and that EISO can slow the growth of tumors in a HNSCC xenograft model.



RESULTS AND DISCUSSION Effects on Proliferation and Cell Viability. The antiproliferative and cytotoxic effects of α-santalol, β-santalol, and EISO were determined from concentration−response curves generated in SCC-4, CAL 27, HSC-3, SCC-9, SCC-25, HN5, and HSC-2 HNSCC lines (Supplemental Figures 1−3). From these data, the GI50 (concentration that caused 50% inhibition of cell growth), the TGI (concentration that caused total growth inhibition), and LC50 (concentration that caused B

DOI: 10.1021/acs.jnatprod.5b00207 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

50% cytotoxicity as compared to the time of treatment) were calculated for each compound/cell line pair (Table 1). The antiproliferative GI50 values were in the range 3.4−5.5 μM for α-santalol and were slightly higher for β-santalol at 4.2− 7.3 μM. EISO had GI50 values in the range 0.8−1.6 μg/mL in these cell lines. Assuming that the oil has an average molecular weight equivalent to the santalols (given that over 69% of the oil is made up of α- and β-santalol), this translates into an approximate antiproliferative potency of 3.6−7.3 μM for EISO. Therefore, not surprisingly, the potency of the oil is actually quite similar to the individual santalols. TGI values were in the range 5.3−9.2 μM for α-santalol, 6.1−10.0 μM for β-santalol, and 1.4−2.2 μg/mL for EISO, while LC50 values were in the range 7.7−14.0 μM for α-santalol, 9.1−15.9 μM for β-santalol, and 1.9−3.0 μg/mL for EISO. Although similar, β-santalol was consistently less potent than α-santalol against each cell line tested. All treatments caused cytotoxicity, as observed by the concentration−response curves dipping below the dashed line in Supplemental Figures 1−3, which indicates the density of cells at the time of treatment. All 21 of the concentration−response curves in this study were characterized by steep slopes where the concentrations that initiated cytotoxicity were less than 2-fold greater than the antiproliferative concentrations (Table 1). The concentration− response curves obtained for each substance were similar for each cell line tested, suggesting that α-santalol, β-santalol, and EISO affect a universal feature of all seven molecularly diverse HNSCC cell lines used in this study. Due to the similar potency and efficacy of the compounds in all seven HNSCC cell lines tested, additional mechanistic studies were primarily performed with the commonly used SCC-4 line, which is also amenable to growth in xenograft models. Effects on Cell Cycle. The effects of EISO and the purified santalols on the cell cycle profile of SCC-4 cells were determined 18 h after treatment. At the LC50, EISO did not have a major effect on the cell cycle distribution of SCC-4 cells as compared to vehicle-treated controls (Figure 2). However, some G2/M accumulation was evident at 2× the LC50, which became even more pronounced at 3−4× the LC50, which was similar to the degree observed in the presence of paclitaxel, a microtubule stabilizer that efficiently arrests cells in mitosis (Figure 2). A concentration-dependent G2/M accumulation was also observed for both α- and β-santalol; however αsantalol was more potent in this regard, as it caused a similar level of G2/M arrest at the 2× LC50 concentration to what was observed for EISO or β-santalol at 3× their respective LC50 values. The higher concentrations required to see G2/M arrest as compared to cytotoxicity may be due to the shorter duration of treatment used to observe effects on the cell cycle prior to the cytotoxicity observed 48 h after treatment at this concentration. This concentration-dependent G2/M arrest of SCC-4 cells after treatment with the santalols or EISO was observed in three independent experiments, and similar results were observed in the HN5 HNSCC cell line as well (Supplemental Figure 4), supporting previous reports that the santalols are able to cause G2/M accumulation in multiple cell types.11 To determine whether cells were accumulating in the G2 and/or M phases of the cell cycle, SCC-4 cells were visualized 18 h after treatment with 4× the LC50 of either EISO, α-santalol, or β-santalol, concentrations that caused maximal G2/M arrest in Figure 2. All three substances caused the accumulation of mitotic cells, as evidenced by the increased number of cells with a rounded morphology (Figure 3).

Figure 3. α-Santalol, β-santalol, and EISO alter mitotic spindles. The microtubules in SCC-4 cells were visualized through indirect immunofluorescence for β-tubulin 18 h after treatment with vehicle (A), 55.6 μM α-santalol (B), 61.2 μM β-santalol (C), or 12.8 μg/mL EISO.

Additionally, microtubules were visualized by immunofluorescence. In contrast to vehicle-treated control cells that contained a normal interphase microtubule network with bipolar spindles observed in mitotic cells (Figure 3A), the mitotic cells resulting from treatment with any of the three agents contained multiple aberrant microtubule spindles (Figure 3B−D). The spindles formed after treatment with αsantalol were more punctate and numerous (Figure 3B), while cells treated with β-santalol contained larger multipolar spindles or a single large spindle (Figure 3C). Interestingly, cells treated with EISO contained a mix of cells containing either numerous punctate spindles similar to α-santalol (arrows) or larger, more diffuse spindles similar to β-santalol (arrowheads). Both types of spindles have been associated with the treatment of cells with other classes of microtubule-destabilizing agents.20 Direct Interaction of Santalols with Tubulin. Following these data, as well as previous reports indicating that the santalols cause G2/M arrest and disrupt cellular microtubules,11 we tested the effects of EISO and the purified santalols directly on tubulin polymerization. All treatments inhibited the polymerization of purified tubulin at a concentration of 50 μM, with α-santalol causing the greatest inhibition of tubulin polymerization followed by EISO and, last, β-santalol (Figure 4A). The concentration-dependent inhibition of tubulin polymerization by α-santalol was further demonstrated (Figure 4B). Interestingly, the concentrations of these compounds that directly inhibit tubulin polymerization correlate with those that cause G2/M arrest and aberrant microtubule spindles in cells. These data indicate, for the first time, that α-santalol, β-santalol, and EISO directly inhibit tubulin polymerization, which is a well-established anticancer mechanism that may underlie some of the biological effects ascribed to these agents. Further studies were undertaken to determine whether the santalols bind to either of the well-established microtubuledestabilizing sites on tubulin, the colchicine or vinblastine sites. Surprisingly, the santalols were unable to efficiently displace either drug from purified tubulin, even when present at concentrations up to 7000-fold greater than vinblastine or colchicine. Therefore, the santalols may bind to a novel site on tubulin to cause their destabilizing effects or, more likely, bind C

DOI: 10.1021/acs.jnatprod.5b00207 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

Figure 4. α-Santalol, β-santalol, and EISO inhibit tubulin polymerization. The polymerization of purified tubulin (18 μM) was monitored turbidimetrically in the presence of 50 μM α-santalol, 50 μM β-santalol, or 11 μg/mL EISO (A) or with 6−50 μM α-santalol (B).

Figure 5. Modeling of α- and β-santalol into the colchicine site on tubulin. (A) The affinity (log Ki) for combretastatin A4 (1), CHEMBL143850 (2), colchicine (3), and CHEMBL143680 (4) was plotted against their predicted binding site locations in the colchicine site, to make a standard curve relating the two values. The predicted log Ki values for α- and β-santalol shown in the plot were independently determined from this curve using only their corresponding Rhodium binding locations. (B) Rhodium-predicted binding pose of α-santalol (magenta) and colchicine (blue) in the colchicine binding pocket.

with binding site location (r = 0.9), when the coordinate (center of mass of the ligand) was expressed as distance from the protein’s center of mass along the principal axis (z-axis) of the crystal structure. The model also predicts that the binding site of vinblastine is quite distinct from these and the other colchicine site binding agents. Intriguingly, the maximum cavity-filling principal axis coordinate for α- and β-santalol in the colchicine site predicts that they would have Ki’s of 5500 and 6600 nM, respectively. This finding is consistent with α-santalol being slightly more potent than β-santalol for tubulin-dependent effects (Figure 4) and with the inability to detect colchicine displacement from tubulin by these agents. The predicted binding pose of αsantalol in the colchicine binding pocket of β-tubulin as selected from Rhodium docking is depicted in Figure 5B. Xenograft Studies. The ability of EISO to inhibit the growth of HNSCC tumors in vivo was determined using the fast growing SCC-4 xenograft model (Figure 6). Daily topical administration of 5 μL of a 50% EISO solution in DMSO caused little to no inhibition of tumor growth on days 0−2 and was therefore increased to 10 μL on days 3 and 4. This higher EISO dose resulted in a lag in tumor growth as compared to tumors treated with DMSO alone and was further increased to 20 μL for the duration of the trial. While the EISO-treated animals appeared to have slower tumor growth, the tumors did increase in size even with almost daily treatment, suggesting that EISO can slow the growth of, but not completely inhibit, tumor growth in this model. Importantly, antitumor effects were not observed when EISO administration was initiated

with such a weak affinity that we could not detect displacement in our assays. Molecular Modeling. Molecular modeling studies were undertaken to determine whether there was a reasonable expectation that the santalols could weakly bind within either the colchicine or vinblastine sites on tubulin. The studies were performed with the Rhodium protein docking simulation program, using a fully automated search over the entire tubulin crystal structure (PDB ID 4O28) to predict the binding site. The predicted binding site locations used in the analysis correspond to globally maximized cavity filling of the binding molecule. This method is distinct from other docking programs, which globally minimize specific enthalpic components in the free energy, such as hydrogen bonds and hydrophobic interactions. By way of distinction, Rhodium’s maximum cavity-filling poses roughly correspond to the maximum potential to liberate surface water upon ligand binding (apolar desolvation), an important entropic factor that might outweigh specific enthalpic interactions in determining the most likely pose.21 The docking simulations predict that the santalol binding location is in the colchicine site and not the vinblastine site on tubulin. Separately, using the same methods, we found a linear correlation between the simulated binding site and ligand affinity of other compounds that bind within the colchicine site on tubulin. Using a training set of four distinct colchicine site binding agents that have Ki values ranging from 180 to 3690 nM,22 a model-based standard curve was constructed using the Rhodium docking simulations (Figure 5A). The affinity of the training set ligands could be correlated D

DOI: 10.1021/acs.jnatprod.5b00207 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

cells were provided by Dr. Cara Gonzales (UTHSCSA, San Antonio, TX, USA). HSC-2 cells were provided by Dr. Shivani Ruperel (UTHSCSA). HN5 cells were provided by Dr. Michael Story (UT Southwestern, Dallas, TX, USA). Analysis of Antiproliferative and Cytotoxic Effects. The sulforhodamine B assay25 was used to determine the antiproliferative and cytotoxic effects of α-santalol, β-santalol, and EISO. Cells were plated in 96-well plates and incubated with a range of concentrations of α-santalol, β-santalol, or EISO for 48 h. The GI50 (the concentration that inhibited 50% cell growth), TGI (the concentration for total growth inhibition), and LC50 (the concentration that caused 50% cytotoxicity as compared to the time of treatment) were calculated from log concentration−response curves of three to nine independent experiments, each performed in triplicate with at least six concentrations per experiment using GraphPad Prism 6 software (GraphPad Software, La Jolla, CA, USA). Cell Cycle Analysis. Flow cytometry was used to determine the effects of α-santalol, β-santalol, and EISO on the cell cycle of SCC-4 and HN5 cells. Cells were treated with vehicle (DMSO), 15 nM paclitaxel (as a positive control), and the LC50, 2× LC50, 3× LC50, and 4× LC50 values for α-santalol, β-santalol, and EISO for 18 h. Krishan’s reagent was then used to label the DNA of the cell with propidium iodide.26 DNA content was measured using a FACSCalibur flow cytometer (BD Biosciences) and analyzed using ModFit LT 3.0 software (Verity Software, Topsham, ME, USA) by plotting propidium iodide intensity versus the number of cellular events. Indirect Immunofluorescence. Indirect immunofluorescence was used to observe the effects of α-santalol, β-santalol, and EISO on cellular microtubules. SCC-4 cells were grown on glass coverslips and treated for 18 h with vehicle (DMSO), α-santalol, β-santalol, or EISO. After incubation, cells were fixed with methanol and microtubules stained using a monoclonal β-tubulin antibody (T4026; Sigma, St. Louis, MO, USA) and observed using a Nikon Eclipse 80i fluorescence microscope. Images were acquired using NIS Elements AR 3.0 software. Tubulin Polymerization Assay. The effects of α-santalol, βsantalol, and EISO on the polymerization of purified porcine brain tubulin (Cytoskeleton, Denver, CO, USA) were assessed as previously described.20 The ability of the santalols to directly inhibit tubulin polymerization was replicated in three independent experiments. Molecular Docking. Docking simulation at the colchicine site of bovine brain tubulin was performed using the tubulin crystal structure 4O2B. Hydrogen atoms were added with Pymol. Assay data and molecular structures of four tubulin binding molecules, used in building the docking affinity model, were retrieved from BindingDB,27,28 assay ID 50035270.22 The coordinates were optimized using the MMFF94 force field with the OpenBabel oboptimize program. Multiple conformers of each structure were generated with the confab method in OpenBabel, using an energy cutoff of 4 kcal/mol. Rhodium docking simulations were performed on all conformers using an automatic search for maximal cavity-filling binding site location. The location was measured using coordinates along the principal axis of the tubulin crystal structure. In Vivo Antitumor Assay. Female athymic nude mice 6−8 weeks of age were bilaterally injected sc with 1 × 106 SCC-4 cells into each flank. When tumors reached an average of 108 mm3, they were divided into two treatment groups (n = 7 tumors). The tumors in the test group were treated with topical application of 5 μL (days 0−2), 10 μL (days 3−4), or 20 μL (days 7−11 and 14−18) of a 50:50 solution of EISO/DMSO. Tumors in the vehicle group were treated topically with DMSO vehicle alone. Mice were 8−10 weeks of age and 22−26 g in weight when dosing was initiated. Over the course of dosing, the weight of animals was monitored and tumor mass was calculated in mm3 from caliper measurements of tumor length (mm) × width (mm) × height (mm). Mice were maintained in an AAALAC approved facility under approval of the IACUC at UTHSCSA, provided food and water ad libitum, and euthanized before the tumor volume exceeded 2500 mm3.

Figure 6. Antitumor effects of EISO. A 50:50 solution of EISO/ DMSO was applied topically to tumors in mice bearing bilateral SCC4 xenografts. Tumors in the EISO treatment group received 5 μL (days 0−2), 10 μL (days 3−4), or 20 μL (days 7−11 and 14−18) of the 50% EISO solution in DMSO. Corresponding volumes of DMSO were applied topically to the vehicle tumors. Tumor volume was monitored using caliper measurements.

after tumors had reached sizes greater than 200 mm3. No weight loss or any other signs of toxicity were observed in either the vehicle or EISO treatment groups over the course of the trial. In line with previous studies in skin cancer models,12−15 these results support the potential of EISO in chemoprevention of early stage tumors as opposed to the advanced stages, where other, more potent microtubuletargeting agents are currently indicated.



CONCLUSION Microtubule-targeted agents are some of the most effective chemotherapeutic agents used today, although they are also associated with dose-limiting toxicities ranging from neutropenia to peripheral neuropathy. Interestingly, most microtubule interacting compounds were originally isolated as minor, highly potent constituents from plants, including the vincas, paclitaxel, and colchicine. In contrast, we show that the santalols are much less potent microtubule-destabilizing compounds that are the major constituents present in EISO. It is intriguing to speculate that the presence of low-abundance, highly potent microtubuletargeted agents, such as paclitaxel, endows the Pacific Yew tree the same effective dose as the high-abundance, low-potency santalols in the East Indian sandalwood tree. Similarly, cruciferous vegetables have been found to contain the lowpotency microtubule destabilizers erucin and sulforaphane, which have also been implicated in cancer prevention.23,24 Our finding that low-potency microtubule-destabilizing compounds are the major component of EISO, which has been deemed safe for human dietary, cosmetic, and medicinal use, and may slow the growth of oral tumors with topical administration suggests that future studies exploring its utility in the treatment or prevention of cancer may be warranted.



EXPERIMENTAL SECTION

Materials. α-Santalol, β-santalol, and EISO were obtained from Santalis Pharmaceuticals (San Antonio, TX, USA). Purity was determined through GC/FID analysis (98.2% for α-santalol, 93.2% for β-santalol). The α-santalol content of EISO was 48.32%, and the βsantalol content was 20.73% as determined by GC/FID analysis. Cell Culture. SCC-4, CAL 27, HSC-3, SCC-9, SCC-25, HN5, and HSC-2 cells were grown in Dulbecco’s modified Eagle medium (Life Technologies, Carlsbad, CA, USA) supplemented with 10% FBS and 50 μg/mL gentamicin. SCC-4, CAL 27, and SCC-25 cells were obtained from ATCC (Manassas, VA, USA). HSC-3 cells were provided by Dr. Brian Schmidt (NYU, New York, NY, USA). SCC-9 E

DOI: 10.1021/acs.jnatprod.5b00207 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products



Article

(23) Azarenko, O.; Jordan, M. A.; Wilson, L. PLoS One 2014, 9, e100599. (24) Azarenko, O.; Okouneva, T.; Singletary, K. W.; Jordan, M. A.; Wilson, L. Carcinogenesis 2008, 29, 2360−2368. (25) Skehan, P.; Storeng, R.; Scudiero, D.; Monks, A.; McMahon, J.; Vistica, D.; Warren, J. T.; Bokesch, H.; Kenney, S.; Boyd, M. R. J. Natl. Cancer Inst. 1990, 82, 1107−1112. (26) Krishan, A. J. Cell Biol. 1975, 66, 188−193. (27) Chen, X.; Lin, Y.; Liu, M.; Gilson, M. K. Bioinformatics 2002, 18, 130−139. (28) Liu, T.; Lin, Y.; Wen, X.; Jorissen, R. N.; Gilson, M. K. Nucleic Acids Res. 2007, 35, D198−201.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jnatprod.5b00207.



AUTHOR INFORMATION

Corresponding Author

*Phone: 210-567-6267. Fax: 210-567-4300. E-mail: risingera@ uthscsa.edu (A. L. Risinger). Notes

The authors declare the following competing financial interest(s): Dr. Levenson is an employee of Santalis Pharmaceuticals, which provided funding for this study.



REFERENCES

(1) Burdock, G. A.; Carabin, I. G. Food Chem. Toxicol. 2008, 46, 421−432. (2) Christenson, P. A.; Secord, N.; Willis, B. J. Phytochemistry 1981, 20, 1139−1141. (3) Sharma, M.; Levenson, C.; Bell, R. H.; Anderson, S. A.; Hudson, J. B.; Collins, C. C.; Cox, M. E. Phytother. Res. 2014, 28, 925−932. (4) Dickinson, S. E.; Olson, E. R.; Levenson, C.; Janda, J.; Rusche, J. J.; Alberts, D. S.; Bowden, G. T. Arch. Biochem. Biophys. 2014, 558, 143−152. (5) Moy, R. L.; Levenson, C.; So, J. J.; Rock, J. A. J. Drugs Dermatol. 2012, 11, 1403−1408. (6) Paulpandi, M.; Kannan, S.; Thangam, R.; Kaveri, K.; Gunasekaran, P.; Rejeeth, C. Phytomedicine 2012, 19, 231−235. (7) Saraswati, S.; Kumar, S.; Alhaider, A. A. Mol. Cancer 2013, 12, 147. (8) Arasada, B. L.; Bommareddy, A.; Zhang, X.; Bremmon, K.; Dwivedi, C. Anticancer Res. 2008, 28, 129−132. (9) Kaur, M.; Agarwal, C.; Singh, R. P.; Guan, X.; Dwivedi, C.; Agarwal, R. Carcinogenesis 2005, 26, 369−380. (10) Santha, S.; Bommareddy, A.; Rule, B.; Guillermo, R.; Kaushik, R. S.; Young, A.; Dwivedi, C. PLoS One 2013, 8, e56982. (11) Zhang, X.; Chen, W.; Guillermo, R.; Chandrasekher, G.; Kaushik, R. S.; Young, A.; Fahmy, H.; Dwivedi, C. BMC Res. Notes 2010, 3, 220. (12) Bommareddy, A.; Hora, J.; Cornish, B.; Dwivedi, C. Anticancer Res. 2007, 27, 2185−2188. (13) Dwivedi, C.; Guan, X.; Harmsen, W. L.; Voss, A. L.; GoetzParten, D. E.; Koopman, E. M.; Johnson, K. M.; Valluri, H. B.; Matthees, D. P. Cancer Epidemiol. Biomarkers Prev. 2003, 12, 151−156. (14) Dwivedi, C.; Valluri, H. B.; Guan, X.; Agarwal, R. Carcinogenesis 2006, 27, 1917−1922. (15) Santha, S.; Dwivedi, C. Photochem. Photobiol. 2013, 89, 919− 926. (16) Leemans, C. R.; Braakhuis, B. J.; Brakenhoff, R. H. Nat. Rev. Cancer 2011, 11, 9−22. (17) Li, H.; Wawrose, J. S.; Gooding, W. E.; Garraway, L. A.; Lui, V. W.; Peyser, N. D.; Grandis, J. R. Mol. Cancer Res. 2014, 12, 571−582. (18) Rothenberg, S. M.; Ellisen, L. W. J. Clin Invest. 2012, 122, 1951−1957. (19) Urban, D.; Corry, J.; Rischin, D. Cancer Metastasis Rev. 2014, 120, 1462−1470. (20) Risinger, A. L.; Westbrook, C. D.; Encinas, A.; Mulbaier, M.; Schultes, C. M.; Wawro, S.; Lewis, J. D.; Janssen, B.; Giles, F. J.; Mooberry, S. L. J. Pharmacol. Exp. Ther. 2011, 336, 652−660. (21) Olsson, T. S.; Williams, M. A.; Pitt, W. R.; Ladbury, J. E. J. Mol. Biol. 2008, 384, 1002−1017. (22) Li, Q.; Woods, K. W.; Claiborne, A.; Gwaltney, S. L., 2nd; Barr, K. J.; Liu, G.; Gehrke, L.; Credo, R. B.; Hui, Y. H.; Lee, J.; Warner, R. B.; Kovar, P.; Nukkala, M. A.; Zielinski, N. A.; Tahir, S. K.; Fitzgerald, M.; Kim, K. H.; Marsh, K.; Frost, D.; Ng, S. C.; Rosenberg, S.; Sham, H. L. Bioorg. Med. Chem. Lett. 2002, 12, 465−469. F

DOI: 10.1021/acs.jnatprod.5b00207 J. Nat. Prod. XXXX, XXX, XXX−XXX