hoaf-caim · 2018-01-16 · graviola, soursop, guanabana, guyabano . soursop leaves come from...

HOAF-CAIM MEDICINA COMPLEMENTARIA, ALTERNATIVA E INTEGRATIVA

ANNONA MURICATA Graviola, Soursop, Guanabana, Guyabano


References and treatment methods support used are from (1) Dr. Carlos Alvarez (2) Dr. Otto Henrich Warburg (3) Dr. Max Gerson (4) Dr. Tulio Simoncini (5) Dr. albert Marti Bosch (6) Dr. John Myers and the HOAF-CAIM database which includes all the

TCM and OM www.healthofamericans.org

ANNONA MURICATA Graviola, Soursop, Guanabana, Guyabano

Soursop leaves come from naturally grown soursop trees. All leaves are carefully handpicked and selected to the right size and maturity level. They are naturally dried, processed and vacuum packed to preserve its most potent active ingredients and natural properties. Leaves have been known to produce many health benefits and can be as nutritious as consuming the fresh fruit.

Chemical free leaves; naturally grown leaves are dark green and healthy; dried under slow heat to preserve their natural properties; whole dried leaves no foreign contaminants. Bioactive products such as soursop leaves are consumed worldwide with multiple health benefits for consumers. Please talk to your physician or Natural health professional for additional recommendations.

Directions to prepare a Soursop Tea Boil 1 cup of water. Introduce 4 or 5 leaves in the hot water Steep between 3 to 5 minutes. Enjoy a beneficent herbal tea! The herb Health Benefits Fights Cancer Cell, Lowers Blood Pressure, Increases Immune System, Makes You Feel Stronger, and Improves Health and Energy.

The Herb description and properties Guyabano belongs to the family of Annonaceae, (A. muricata L.). The flesh of the fruit consist of a white edible pulp that is high in carbohydrates and considerable amounts of Vitamin C, Vitamin B1, Vitamin B2, Potassium and dietary fiber. Guyabano is low in cholesterol, saturated fat and sodium. Not only is guyabano a good health food, it also taste delicious. The tree and fruit is known in various names: Guyabano in Filipino, Soursop in English, Graviola in Brazil, and Guanabana in Spanish.

http://www.healthofamericans.org/

http://www.graviolausa.com/2013/06/graviola-soursop-guanabana-guyabano.html

http://www.graviolausa.com/2013/06/graviola-soursop-guanabana-guyabano.html




Research Studies on Soursop for Cancer Therapy

Studies are being done by leading medical institutes, universities and pharmaceutical companies of the healing properties of soursop against cancers. Initial findings show that certain compounds and chemicals extracted from soursop leaves, seeds, fruit and bark appear to kill cancer cells while leaving normal cells remain unaffected. A Purdue University study showed that soursop leaves killed cancer cells among six human cells lines and were especially effective against prostate, pancreatic and lung cancers. A study conducted at Catholic University of South Korea, and published in the Journal of Natural Products, stated that soursop has one chemical found selectively kill colon cancer cells at "10,000 times potency of Adriamycin (The commonly used chemotherapy drug).", but left healthy cells untouched. Chemotherapy, on the other hand, targets all actively reproducing cells (such as stomach and hair cells) causing nausea and hair loss.

Some Extracts from Research Studies for Apoptosis





Disclaimer: The information on this page is for informational purposes and is not intended to replace the advice of your physician. Do not change or discontinue any medication or treatment without consulting your physician. The statements here are not intended to diagnose, treat, cure or prevent any disease.

AS A SUPPORTING AND SCIENCE BASE RESEARCH STUDIES ABOUT THE ABOVE INFORMATION, PLEASE REVIEW THE FOLLOWING

SEVERAL RESEARCH STUDIES RELAYED WITH ANNONA MURICATA


Journal of Cancer Therapy, 2013, 4, 1244-1250 http://dx.doi.org/10.4236/jct.2013.47146 Published Online September 2013 (http://www.scirp.org/journal/jct)

Annonaa muricata Linn Leaf Induce Apoptosis in Cancer Cause Virus

Okid Parama Astirin1, Anief Nur Artanti1, Meutia Srikandi Fitria1, Eva Agustina Perwitasari1, Adi Prayitno2*

1Department of Biology, Faculty of Mathematics and Natural Science, University of Sebelas Maret, Surakarta, Indonesia; 2Depart- ment of Dental and Oral Disease, Faculty of Medicine, University of Sebelas Maret, Surakarta, Indonesia. Email: *[email protected] Received July 18th, 2013; revised August 20th, 2013; accepted August 28th, 2013 Copyright © 2013 Okid Parama Astirin et al. This is an open access article distributed under the Creative Commons Attribution Li-cense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

ABSTRACT

Introduction: Now many studies conducted on the drug substance from nature that can serve as an anticancer agent as a potential chemoprevention agent, such as Annona muricata Linn leaf escort chemotherapy, which was flaring. The cancer cell in humans was included the loss of p53 protein function due to mutations in the protein gene. Other causes are that the p53 proteins are not functioning due to an increase in protein misfolding event chaperones and degradation events ubiquitous as binding by viral protein. Method: Cytotoxicity assay performed on 24 well plate micro-cultures. HeLa cells are as 2 × 104 cells in 100 mL in RPMI media. Created control is RPMI and solvent DMSO 0.25%. Cyto- toxic Test performed by the method of calculation tryphan blue dye exclusion. Being fasted for 24 hours in the culture medium, then the cells are grown in micro-plate with media plus samples with a non-lethal concentration (LC50) of partition and fractionation Annona muricata Linn leaf. Sampling is performed at 24 hours. Each of these wells is calculated the number of living cells and made the curve of cell number and incubation time. Result: The results showed that HeLa cells are being LC50 partition of leaves Annona muricata Linn in ethyl acetate his cell death rate was higher (2000 µg/ml have 131.89%; 15.625 µg/ml have 11.37%) and in ethanol-distillate water his cell death rate was lower (2000 µg/ml have 35.80%; 15.625 µg/ml have 3.97%). Another results showed that HeLa cells are being LC50 fractionation of leaves Annona muricata Linn in chloroform his cell death rate was higher (2000 µg/ml have 91.86%; 15.625 µg/ml have 2.68%) and in ethyl acetate, his cell death rate was lower (2000 µg/ml have 23.79%; 15.625 µg/ml have 4.69%). Figure regression LC50 of HeLa cell culture treatment with partition or fractionation looks of regression test is the positive regression coefficient. Conclusion: Annona muricata Linn leaf in chloroform is a good candidate for chemoprevention escort chemotherapy for cancer causing viruses. Keywords: Chemoprevention; Annona muricata Linn Leaf; HeLa Cell Culture; Apoptosis

1. Introduction

Many studies conducted on the drug substance from nature that can serve as an anticancer agent as a potential chemoprevention agent’s escort chemotheraphy [1-4]. Many studies have shown the possibility of anti-cancer compounds that have toxicity selectively kill cancer cells without damaging normal cells, the compound derived from acid molecule from nature like fruit, e.g. with dock- ing methods on the component of red fruit and tested for cancer therapy through modeling followed by molecular- dynamics simulations [5,6]. The study on cytotoxic tests at the cellular level has been conducted, but very limited

testing on the molecular level, especially in cancers caused by viral infections. Now it is revealed that Annona muricata Linn members expressed empirically to fight cancer [7-9]. Are after being proven cellular and molecular leaf isolate tailings will produce bioproduct that can be used by the public as a potential chemoprevention agent’s escort agency chemotheraphy?

Cancer is a disease that ranks second leading cause of death in the world. The usual approach to cancer prevention, among others by: prevention of interaction with cancer-causing agents, increasing defense mechanisms against cancer and lifestyle modifications. The main focus of this research is that cancer is caused by a viral infection i.e. cervical cancer (CC) is the number of events, it is *Corresponding author.

Copyright © 2013 SciRes. JCT

Annonaa muricata Linn Leaf Induce Apoptosis in Cancer Cause Virus 1245

quite high in Indonesia is quite high. Cervical Cancer is due to a viral infection known as the Human papilloma virus (HPV). World prevalence rate HPV infection is 99.7% of the CC [10-12]. Cancer begins with a scene gene expression imbalance specific role in apoptosis and cell proliferation, and DNA repair [13]. Understanding these processes provides the basis for chemotherapy apoptosis through induction of cancer cell death [7,14, 15]. The process of apoptosis can occur through multiple pathways. One point that has close links with cancer through the induction of apoptosis is much played by p53 protein [16-18]. Failure of apoptosis regulation is the key principle to the success of carcinogenesis, i.e. inhibition of apoptosis events [19,20]. When the high incidence of the CC in developing countries, including Indonesia, is not addressed, it will have an impact on the rate of increase in morbidity and higher mortality, and at last will reduce the quantity and quality of human resources.

The flow of thought between chemoprevention An- nona muricata Linn, and the incidence of apoptosis in virus CC can be explained by the flow of the discussion pathobiology [21,22]. Expected to chemoprevention bioproduct Annona muricata Linn leaf as Annonaceae members can be deadly virus, which means lower stressor dynamics and increase of the expression of p53, so that events can be enhanced by apoptosis [7,14,15,23,24].

Many stressors, such as hotness, spiciness, wounds and infections, cause or trigger the expression of specific gene. Cells infected with the virus will experience distress increases, and the expression of p53 protein, among others aims to enhance apoptosis. The expression of p53 protein is a form of cell response to a stressor [25-27]. However, these efforts do not always succeed, even progress to cancer. Decreased expression of the p53 protein can cause a decrease in cell apoptosis mechanism. Pro- tein p53, Baxxl, caspase-3 is known as an inhibitor of cancer and is a protein that plays an important role in the regulation of apoptosis. Decrease in apoptotic cells that are not able to offset the increase in cell proliferation would result in the occurrence of cancer cells [28]. Cause of p53 malfunction is clear that the control system as chaperone as plays an important role if the folding process fails or an error occurs, causing folding abnormalities and targeted functionality to accumulate. The accumula- tion of faulty protein folding would harm cells and can result in apoptosis. Many data have shown, how chaperones facilitate transformation towards cancer at molecular level, and support the concept that “there are events of protein function changes in carcinogenesis, which needs serious attention in the development of human cancers” [20,29]. Another cause of p53 malfunction is mutation so the p53 proteins don’t act optimally. Approximately 50% of human cancer cells lose p53 function due to mutations in the protein-coding gene [30,31]. Records of other

cause are that E6 protein of HPV-16 and 18 will result in the inactivation of the p53 gene product through binding mechanism called the ubiquitin-dependent proteolytic pathway (E6AP), resulting in decreased levels of p53 protein (wild-type) [32]. Returning a mutated p53 function could potentially trigger a mass apoptosis, which can kill cancer cells effectively and prove that Actinomycin D (chemotherapy drug) affects implantation failure of Rattus norvegicus [33]. All were expected to explain the role of Annona muricata Linn leaves in decreasing infected virus and increasing expression of p53 protein in the event CC patients infected with the HPV through increased apoptosis.

2. Material and Method

HeLa cells were grown in culture bowls containing RPMI 1640 medium and FBS added 0.5%, Penstrep 2% and fungizon 0.5% and incubated in a 5% CO2 incubator for 24 hours. The cells density is 2 × 104 cells/ml.

Extraction using two methods, namely an extraction method using 96% ethanol and percolation method using the solvent n-hexane, chloroform, ethyl acetate, and ethanol 96%.Results of maceration partitioned using solvents n-hexane, chloroform, ethyl acetate, and ethanol-distillate water. Active extract of the results of the percolation fractionated using vacuum liquid chromatography (VLC) with a mobile phase n-hexane and ethyl acetate, as well as the stationary phase silica gel 60 PF 254.

The stage to develop the test cells HeLa is by Freshney and Gadek method with some modification [34,35]. Cy- totoxicity assay performed on 24 well plate micro-cultures. HeLa cells as 2 × 104 cells in 100 mL RPMI media. Further isolates were given 100 mL at a concentration series ranking triplet. Created are RPMI for medium control and solvent DMSO 0.25%. Cytotoxic Test performed by the method of calculation tryphan blue dye exclusion (MTT). Cells were grown for 24 hours in the culture medium. The cells were grown in micro-plate with media plus samples with a non-lethal concentration (LC50 below) of partition or fractionation Annona muricata Linn leaf. Sampling performed at 24 hours.

Each of these wells is calculated the number of living cells. The percentage of cell death was calculated using the modified formula Abbot [36]. Thus made curve for the number of cells death again incubation time [37].

3. Result

The results showed (Figure 1 and Table 1) that HeLa cells are being LC50 partition of leaves Annona muricata Linn in ethyl acetate his cell death (apoptotic) rate was higher (2000 µg/ml have 131.89%; 15.625 µg/ml have 11.37%) and in ethanol-distillate water his cell death (apoptotic) rate was lower (2000 µg/ml have 35.80%; 15.625 µg/ml have 3.97%)



Figure1. (a) Hela cells culture treatment with DMSO as a control (cell death/apoptotic rate: 3.80%); (b) Hela cells culture after treatment with partition of leaves Annona muricata Linn in chloroform (cell death/apoptotic rate: 65.20%). Table1. The average percentage of death cells associated with many concentration and many solvent from partition the leaves Annona muricata Linn. We can look that partition of leaves Annona muricata Linn in solvent ethyl acetate make cell death/apoptotic rate is higher than other (average percentage is 131.89% cells death in 2000.00 µg/ml and 11.37% death cells in 15.62 µg/ml concentration) but in chloroform cell death/apoptotic rate is number 3 (average percentage 65.20% death cells in 2000.00 µg/ml and 18.42% death cells in 15.62 µg/ml concentration).

No. Solvent Concentration

(µg/ml) Average percentage of death cells (%)

a) 2000.00 131.89 1.

Ethyl acetate b) 15.62 11.37

a) 2000.00 106.53 2. n-heksan

b) 15.62 21.41

a) 2000.00 65.20 3. Chloroform

b) 15.62 18.42

a) 2000.00 35.80 4.

Ethanol distillate water b) 15.62 3.97

Figure regression LC50 of HeLa cell culture treatment

with the partition shown below in Figures 2(a)-(d). From the looks of regression test is the positive regression coefficient. This value is LC50 of four-kind partition An- nona muricata Linn leaves of HeLa cells incubated for 24 hours.

Another results showed (Figure 3 and Table 2) that HeLa cells are being LC50 fractionation in chloroform his cell death (apoptotic) rate was higher (2000 µg/ml have 91.86%; 15.625 µg/ml have 2.68%) and in ethyl acetate his cell death (apoptotic) rate was lower (2000 µg/ml have 23.79%; 15.625 µg/ml have 4.69%).

Figure regression LC50 of HeLa cell culture treatment with the fractionation shown below in Figures 4(a)-(d)). From the looks of regression test is the positive regression coefficient. This value is LC50 of four-kind fractionation Annona muricata Linn leaves of HeLa cells incubated for 24 hours.

(a) (b)

(c) (d)

Figure 2. The curve of (a) Partition Annona muricata Linn leaves in chloroform; (b) Partition Annona muricata Linn leaves in ethanol-distillate water; (c) Partition Annona muricata Linn leaves in n-heksan; (d) Partition Annona muricata Linn leaves in ethyl acetate.

Figure 3. (a) HeLa cell culture treatment with DMSO as a control (cell death/apoptotic rate: 3.80%); (b) HeLa cell culture after treatment with fractionation of leaves Annona muricata Linn in chloroform (cell death/apoptotic rate: 91.86%).

4. Discussion

The incidence of cancer associated with: First, there is increased expression or mutation of gene trigger cancer. Second, there is a decrease in the expression or mutation of gene cancer suppressor. Cancer suppressor gene is normal gene that has an important function in cell homeostasis. If that gene don’t work, then implicated for occur of cancer. Third, the gene associated with cancer is the presence of DNA-repair enzymes. Changes in the function of these enzymes will lead to the occurrence of cancer. Fourth, the process of apoptosis is not normal,



Table 2. The average percentage of death cells associated with many concentration and many solvent from fractionation the leaves Annona muricata Linn. We can look that fractionation of leaves Annona muricata Linn in solvent chloroform cell death/apoptotic rate is higher than other (average percentage is 91.86% death cells in 2000.00 µg/ml and 2.68% death cells in 15.62 µg/ml concentration) but in solvent ethyl acetate cell death/apoptotic rate is lower that other (average percentage is 23.79% death cells in 2000.00 µg/ml and 4.69% in 15.62 µg/ml concentration).

No. Solvent Concentration

(µg/ml) Average percentage of death cells (%)

a) 2000.00 91.86 1. Chloroform

b) 15.62 2.68

a) 2000.00 75.36 2. n-heksan

b) 15.62 8.34

a) 2000.00 34.77 3.

Ethanol distillate water b) 15.62 3.44

a) 2000.00 23.79 4. Ethyl acetate

b) 15.62 4.69

(a) (b)

(c) (d)

Figure 4. The curve of (a) Fractionation Annona muricata Linn leaves in chloroform; (b) Fractionation Annona muricata Linn leaves in ethanol-distillate water; (c) Fractiona- tion Annona muricata Linn leaves in n-heksan; (d) Fraction- ation Annona muricata Linn leaves in ethyl acetate. that happen inhibition of apoptosis [13,16-18,38]. Further developments on the definition of cancer has been suggested that incidence of cancer begins in the disorder at the level of epigenetic (methylation and/or histone modi-

fication) and continues to change at the level of genetic (mutation) [39].

Cell distress by stressor HPV show an increase in the expression of protein chaperones (Hsp70 and Hsp40). Imbalance between increased levels of Hsp70 and Hsp40 with decreased ATP production by the mitochondria so made possible for error events of protein folding of protein denatured or protein newly translated, so that the protein can’t function normally. Abnormal function of these proteins results in the inhibition of apoptosis and increased cell proliferation that triggers carcinogenesis [29,40]. In the state of distress caused by HPV, the cells will express Hsp for the purpose of homeostasis, an imbalance will develop into cancer [41,42].

Various compounds present in Annonaceae familia, including potential anti-cancer compounds. Three compounds in Annona muricata Linn a potential anti-cancer namely are monotetrahydrofuran acetogenins, muricin H, muricin I, and cis-annomontacin [43]. Acetogenin compounds (squamocin A, B, C, and D) and annotemoyin-1 and -2 the Annonaceae have cytotoxic effects [44], platelet agregation inhibitor [45], inhibitor of HIV replica- tion [46], antidiabetic agents (antihiperglikemik) and anti- oxidants [47,48], pesticide [49], and can be used in the treatment of Neisseria gonorrhea [50]. Squamocin serves as insecticides, while the ascimicin have an antileukemia effect [51]. Caryophyllene oxide on the bark has analgesic and anti-inflammatory activity [52], and cyclosqua- mosin D on seed proved showed inhibition of proinflam- matory cytokines in macrophages [45]. Further treatment of cancer with natural ingredients is not always the specific target. Anti-cancer herbal remedies white turmeric (Curcuma zedoaria) affects organs such as the always do mitosis ovarian follicles in mice [53]. Acetogenin the An- nonaceae are composed of fatty acids C32 or C34 long chain fatty acids. Bioactivity acetogenin was diverse as anticancer, immunosuppressive, pesticide, antiprotozoal, and antimicrobial. Acetogenin wall membrane inhibited mitochondrial ATP production, resulting in the production of energy in cancer cells and stops cancer cells even- tually die [54]. Acetogenin was very selective, only at- tacks cancer cells that have excess ATP. These compounds do not attack other cells are normal in the body, disrupting the circulation of cancer cells by reducing the amount of ATP [55]. Soursop leaf cell-killing colon cancer cells to 10,000 times stronger than Adriamycin and other chemotherapy. Some derivatives in different types of structures and some isomers showed significant selectivity cancer cell line, for example, the fight against prostate cancer (PC-3). The main mode of action is acetogenin inhibitor of NADH: oxidoreductase uniquinone, the enzyme complex is important in oxidative phosphorylation in the mitochondria and inhibits NADH oxidase uniquinone the plasma membrane of cancer cells [56].



5. Conclusion

Annona muricata Linn leaf in chloroform is a good ca- didate for chemoprevention escort chemotherapy for cancer cause virus.

6. Acknowledgements

We thank to acknowledge Higher Education Competitive Research Project Ministry of Education and Culture Re- public of Indonesia for Grand Featured Research Univer- sities 2013, LPPT of Gajah Mada University and special thanks to acknowledge, Prof Dr Rafik Karsidi, MSc as a rector of Sebelas Maret University Surakarta Indonesia, Prof Ir Ari Handoko Ramelan, MSc (Hons), PhD as a dean of Faculty of Mathematics and Natural Science of Sebelas Maret University Surakarta Indonesia and thanks to acknowledge Prof Dr Suhartono Taat Putra MS for much inspirations to wrote this article.

REFERENCES [1] G. W. Chang, M. S. H. Wahyuningsih and A. A. A. Yus-

tina, “Effect of Benalu (Dendrophtoe sp.) Leaves Extract on the Male Rat (Rattus Norvegicus) Benzidine Induced Hepatotoxicity,” Journal Kedokteran Yarsi, Vol. 7, No. 1. 1999, pp. 121-132.

[2] D. J. Greenblatt and L. L. von Moltke, “Interaction of Warfarin with Drugs, Natural Substances, and Foods,” Journal of Clinical Pharmacology, Vol. 45, No. 2, 2005, pp. 127-132. doi:10.1177/0091270004271404

[3] P. Imming, “Molecular Targets of Natural Drug Sub- stances: Idiosyncrasies and Preferences,” Planta Medica, Vol. 76, No. 16, 2010, pp. 1794-1801. doi:10.1055/s-0030-1250236

[4] T. Shan, Q. Ma, K. Guo, J. Liu, W. Li, F. Wang and E. Wu, “Xanthones from Mangosteen Extracts as Natural Chemopreventive Agents: Potential Anticancer Drugs,” Current Molecular Medicine, Vol. 11, No. 8, 2011, pp. 666-677.

[5] K. H. Lee, “Anticancer Drug Design Based on Plant- Derived Natural Products,” Journal of Biomedical Sci- ence, Vol. 6, No. 4, 1999, pp. 236-250.

[6] H. Harada, U. Yamashita, H. Kurihara, E. Fukushi, J. Kawabata and Y. Mamei, “Antitumor Activity of Palmitic Acid Found as a Selective Cytotoxic Substance in a Ma- rine Red Alga,” Anticancer Research, Vol. 22, No. 5, 2002, pp. 2587-2590.

[7] B. V. Pardhasaradhi, M. Reddy, A. M. Ali, A. L. Kumari and A. Khar, “Differential Cytotoxic Effects of Annona Squamosa Seed Extracts on Human Tumour Cell Lines: Role of Reactive Oxygen Species and Glutathione,” Journal Biosciences, Vol. 30, No. 2, 2005, pp. 237-244. doi:10.1007/BF02703704

[8] M. Trubus, “Daun Sirsak vs Chemotherapy (Kanker),” Laporan Utama Majalah Trubus, 2011

[9] A. S. N. Formagio, C. A. L. Kassuya, F. F. Neto, C. R. F. Volobuff, E. K. K. Iriguchi, M. do C. Vieira and M. A.

Foglio, “The Flavonoid Content and Antiproliferative, Hypoglycaemic, Anti-Inflammatory and Free Radical Sca- venging Activities of Annona dioica St. Hill,” BMC Complementary and Alternative Medicine, Vol. 13, 2013, p. 14. doi:10.1186/1472-6882-13-14

[10] L. Teppo, E. Pukkala and E. Saxen, “Multiple Cancer— An Epidemiologic Exercise in Finland,” Journal of Na- tional Cancer Institute, Vol. 75, No. 2, 1985, pp. 207- 217.

[11] A. Prayitno, “Cervical Cancer with Human Papilloma Virus and Epstein Barr Virus Positif,” BMC Journal of Carcinogenesis, Vol. 5, 2006, p. 13. doi:10.1186/1477-3163-5-13

[12] D. Subramanya and P. D. Grivas, “HPV and Cervical Cancer: Updates on an Established Relationship,” Post- graduate Medicine, Vol. 120, No. 4, 2008, pp. 7-13. doi:10.3810/pgm.2008.11.1928

[13] J. Samer, Sharkaphaviova, S. Miluska and G. Diana, “Analysis of P53 Status in Human Cell Line Using a Functional Assay in Yeast: Detection of New No Sense p53 Mutattion in Codon 124,” Oncology Report, Vol. 14. 2005, pp. 901-907.

[14] H. F. Chiu, T. T. Chih, Y. M. Hsian, C. H. Tseng, M. J. Wu and Y. C. Wu, “Bullatacin, a Potent Antitumor An- nonaceous Acetogenin, Induces Apoptosis through a Re- duction of Intracellular cAMP and cGMP Levels in Hu- man Hepatoma 2.2.15 Cells,” Biochemical Pharmacology, Vol. 65, No. 3, 2003, pp. 319-327. doi:10.1016/S0006-2952(02)01554-X

[15] G. Jiebin, O. Hirokazu, O. Shigeki, N. Shinobu and N. Shinji, “Arsenic Trioride Induces Apoptosis in Leukemia and Lymphoma Cell Lines via the CD95/CD95L Sys- tem,” Oncology Reports, Vol. 10, 2005, pp. 705-709.

[16] S. Haupt, M. Berger, Z. Goldberg and Y. Haupt, “Apop- tosis—The p53 Network,” Journal Cell Science, Vol. 116, No. 20, 2003, pp. 4077-4085. doi:10.1242/jcs.00739

[17] A. Vazquez, E. E. Bond, A. J. Levine and G. L. Bond, “The Genetics of the p53 Pathway, Apoptosis and Cancer Therapy,” Nature Review. Drug Discovery, Vol. 7, No. 12, 2008, pp. 979-987. doi:10.1038/nrd2656

[18] J. D. Amaral, J. M. Xavier, C. J. Steer and C. M. Rodri- gues, “The Role of p53 in Apoptosis,” Discovery Medi- cine, Vol. 9, No. 45, 2010, pp. 145-152.

[19] R. S. Fife, B. T. Rougraff, C. Proctor and G. W. Sledge Jr., “Inhibition of Proliferation and Induction of Apop- tosis by Doxycycline in Cultured Human Osteosarcoma Cells,” Journal of Laboratory and Clinical Medicine, Vol. 130, No. 5, 1997, pp. 530-534. doi:10.1016/S0022-2143(97)90130-X

[20] M. P. Mayer and B. Bukau, “Hsp70 Chaperones: Cellular Functions and Molecular Mechanism,” Cellular and Mo- lecular Life Sciences, Vol. 62, No. 6, 2005, pp. 670-684. doi:10.1016/S0022-2143(97)90130-X

[21] C. R. Young and C. J. Welsh, “Stress, Health, and Dis- ease,” Cell Science, Vol. 2, No. 2, 2005, pp. 132-159.

[22] E. H. Vogel, M. E. Castro, P. A. Solar and F. A. Soto, “Enhancement of Pavlovian Conditioned Immunosupres- sion in Rats,” Acta Neurobiologiae Experimentalis, Vol.


http://dx.doi.org/10.1177/0091270004271404

http://dx.doi.org/10.1055/s-0030-1250236

http://dx.doi.org/10.1007/BF02703704

http://dx.doi.org/10.1186/1472-6882-13-14

http://dx.doi.org/10.1186/1477-3163-5-13

http://dx.doi.org/10.3810/pgm.2008.11.1928

http://dx.doi.org/10.1016/S0006-2952(02)01554-X

http://dx.doi.org/10.1242/jcs.00739

http://dx.doi.org/10.1038/nrd2656

http://dx.doi.org/10.1016/S0022-2143(97)90130-X

http://dx.doi.org/10.1016/S0022-2143(97)90130-X


67, No. 1, 2007, pp. 71-81.

[23] Y. P. Zhang and Y. Xiong, “A p53 Aminoterminal Nu- clear Export Signal Inhibited by DNA Damage-Induced Phosphorylation,” Science, Vol. 292, No. 5523, 2001, pp. 1910-1915. doi:10.1126/science.1058637

[24] Y. H. Zhang, H. Y. Peng, G. H. Xia, M. Y. Wang and Y. Han, “Anticancer Effect of Two Diterpenoid Compounds Isolated from Annona glabra Linn,” Acta Pharmacology Singapore, Vol. 25, No. 7, 2004, pp. 937-942.

[25] W. B. Derry, A. P. Putzke and J. H. Rothman, “Caenor- habditis Elegans p53: Role in Apoptosis, Meiosis, and Stress Resistance,” Science, Vol. 294, No. 5542, 2001, pp. 591-595. doi:10.1126/science.1065486

[26] L. Qu, S. Huang, D. Baltzis, A. M. Rivas-Estilla, O. Plu- quet, M. Hatzoglou, C. Koumenis, Y. Taya, A. Yoshi- mura and A. E. Koromilas, “Endoplasmic Reticulum Stress Induces p53 Cytoplasmic Localization and Pre- vents p53-Dependent Apoptosis by a Pathway Involving Glycogen Synthase Kinase-3beta,” Genes and Develop- ment, Vol. 18, No. 3, 2004, pp. 261-277. doi:10.1101/gad.1165804

[27] J. L. Alberto, M. D. Macario and E. C. de Macario, “Sick Chaperones, Cellular Stress, and Disease,” New England Journal of Medicine, Vol. 353, No. 14, 2005, pp. 1489- 1501. doi:10.1056/NEJMra050111

[28] I. M. Ghobrial, T. E. Witzig and A. A. Adjei, “Targeting Apoptosis Pathways in Cancer Therapy,” CA: A Cancer Journal for Clinicians, Vol. 55, No. 3, 2005, pp. 178-194.

[29] A. Prayitno, E. Asnar, O. P. Astirin, D. Rosmala and S. T. Putra, “Heat Shock Protein 40 (Hsp40) and Hsp70 Protein Expression in Oral Squamous Cell Carcinoma (OSCC),” Journal of Cancer Therapy, Vol. 4, No. 3, 2013, pp. 734-741. doi:10.4236/jct.2013.43090

[30] M. Hollstein, D. Sidransky, B. Vogelstein and C. C. Har- ris, “p53 Mutations in Human Cancers,” Science, Vol. 253, No. 5015, 1991, pp. 49-53.

[31] T. Soussi and G. Lozano, “p53 Mutation Heterogeneity in Cancer,” Biochemical Biophysical Research Communica- tions, Vol. 331, No. 3, 2005, pp. 834-842. doi:10.1016/j.bbrc.2005.03.190

[32] A. Prayitno, R. Darmawan, I. Yuliadi and A. Mudigdo, “The Expression of p53, Rb, and c-myc Protein in Cer- vical Cancer by Immunohistochemistry Stain,” Biodiver- sitas, Vol. 6, No. 3, 2005, pp. 157-159.

[33] O. P. Astirin, “Implantasi Pada Rattus Norvegicus Setelah Perlakuan Dengan Actinomycin D,” Thesis, Faculty of Pasca Sarjana, University of Gajah Mada, Yogyakarta, 1991.

[34] F. R. Ian, “Culture of Animal Cells: A Manual of Basic Techniques,” 4th Edition, Wiley-Lies, Canada, 2000.

[35] Gadek, J. Wesierska, V. Kotala, S. Kuchtickova and M. D. Horky, “Induction of Apoptosis in Human Cervic Carci- noma Cells During Therapy by Cisplatin,” International Symposium on Predictive Oncology and Intervention Strategies, Paris, 9-12 February 2002.

[36] B. N. Meyer, N. R. Ferriqni, J. E. Putnam, L. B. Jacobsen, D. E. Nichols and J. L. McLaughlin, “Brine Shrimp: A Convenient General Bioassay for Active Plant Consti-

tuents,” Planta Medicine, Vol. 45, No. 5, 1982, pp. 31-34. doi:10.1055/s-2007-971236

[37] Mursyidi and Achmad, “Statistika Farmasi Dan Biologi,” Ghalia, 1985.

[38] N. E. Raab-Traub, “Epstein-Barr Virus in the Pathogene- sis of NPC,” Seminars in Cancer Biology, Vol. 12, No. 6, 2002, pp. 431-441. doi:10.1016/S1044579X0200086X

[39] A. P. Feinberg, R. Ohlsson and S. Henikoff, “The Epigenetic Progenitor Origin of Human Cancer,” Nature Reviews. Genetics, Vol. 7, No. 1, 2006, pp. 21-33. doi:10.1038/nrg1748

[40] J. L. Alberto, M. D. Macario and E. C. de Macario, “Sick Chaperones, Cellular Stress, and Disease,” The England Journal of Medicine, Vol. 353, No. 14, 2005, pp. 1489- 1501.

[41] C. Jolly and R. I. Morimoto, “Role of the Heat Shock Response and Molecular Chaperones in Oncogenesis and Cell Death,” Journal of the National Cancer Institute, Vol. 92, No. 19, 2000, pp. 1564-1572. doi:10.1093/jnci/92.19.1564

[42] J. Villar, “Heat Shock Protein Gene Expression and Sur- vival in Critical Illness,” Critical Care, Vol. 4, 2000, pp. 2-5. doi:10.1186/cc643

[43] C. C. Liaw, F.-R. Chang, C.-Y. Lin, C.-J. Chou, H.-F. Chiu, M.-J. Wu and Y.-C. Wu, “New Cytotoxic Mono- tetrahydrofuran Annonaceous Acetogenins from Annona muricata,” Journal of Natural Products, Vol. 65, No. 4, 2002, pp. 470-475. doi:10.1021/np0105578

[44] H. J. Yang, X. Li, N. Zhang, J. W. Chen and M. Y. Wang, “Two New Cytotoxic Acetogenins from Annona squamosa,” Journal of Asian Natural Products Research, Vol. 11, No. 3, 2009, pp. 250-256. doi:10.1080/10286020802682916

[45] Y. L. Yang, K. F. Hua, P. H. Chuang, S. H. Wu, K. Y. Wu, F. R. Chang and Y. C. Wu, “New Cyclic Peptides from the Seeds of Annona squamosa L. and Their Anti- Inflammatory Activities,” Journal of Agricultural and Food Chemistry, Vol. 56, No. 2, 2008, pp. 386-392. doi:10.1021/jf072594w

[46] Y. C. Wu, Y. C. Hung, F. R. Chang, M. Cosentino, H. K. Wang and K. H. Lee, “Identification of ent-16 beta, 17-Dihydroxykauran-19-oic Acid as an Anti-HIV Prin- ciple and Isolation of the New Diterpenoids Annosqua- mosins A and B from Annona squamosa,” Journal of Natural Products, Vol. 59, No. 6, 1996, pp. 635-671. doi:10.1021/np960416j

[47] M. Kaleem, M. Asif, Q. U. Ahmed and B. Bano, “Anti- diabetic and Antioxidant Activity of Annona squamosa Extract in Streptozotocin-Induced Diabetic Rats,” Singa- pore Medicine Journal, Vol. 47, No. 8, 2006, pp. 670- 675.

[48] S. Panda and A. Kar, “Antidiabetic and Antioxidative Ef- fects of Annona squamosa Leaves Are Possibly Mediated through Quercetin-3-O-Glucoside,” BioFactors, Vol. 31, No. 3-4, 2007, pp. 201-210.

[49] A. Jaswanth, P. Ramanathan and K. Ruckmani, “Eva- luation of Mosquitocidal Activity of Annona squamosa Leaves against Filarial Vector Mosquito, Culex quinque-


http://dx.doi.org/10.1126/science.1058637

http://dx.doi.org/10.1126/science.1065486

http://dx.doi.org/10.1101/gad.1165804

http://dx.doi.org/10.1056/NEJMra050111

http://dx.doi.org/10.4236/jct.2013.43090

http://dx.doi.org/10.1016/j.bbrc.2005.03.190

http://dx.doi.org/10.1055/s-2007-971236

http://dx.doi.org/10.1016/S1044579X0200086X

http://dx.doi.org/10.1038/nrg1748

http://dx.doi.org/10.1093/jnci/92.19.1564

http://dx.doi.org/10.1186/cc643

http://dx.doi.org/10.1021/np0105578

http://dx.doi.org/10.1080/10286020802682916

http://dx.doi.org/10.1021/jf072594w

http://dx.doi.org/10.1021/np960416j

Annonaa muricata Linn Leaf Induce Apoptosis in Cancer Cause Virus


1250

fasciatus Say,” Indian Journal of Experimental Biology, Vol. 40, No. 3, 2002, pp. 363-365.

[50] P. Shokeen, K. Ray, M. Bala and V. Tandon, “Prelimi- nary Studies on Activity of Ocimum sanctum, Drynaria quercifolia, and Annona squamosa against Neisseria go- norrhoeae,” Sexually Transmitted Diseases, Vol. 32, No. 2, 2005, pp. 106-111. doi:10.1097/01.olq.0000152821.23777.90

[51] P. J. Schupp, C. Kohlert-Schupp, S. Whitefield, A. Engemann, S. Rohde, T. Hemscheidt, J. M. Pezzuto, T. P. Kondratyuk, E.-J. Park, L. Marler, B. Rostama and A. D. Wright, “Cancer Chemopreventive and Anticancer Evalu- ation of Extracts and Fractions from Marine Macro- and Micro-Organisms Collected from Twilight Zone Waters Around Guam,” Natural Product Communications, Vol. 4, No. 12, 2009, pp. 1717-1728.

[52] M. J. Chavan, P. S. Wakte and D. B. Shinde, “Analgesic and Anti-Inflammatory Activity of Caryophyllene oxide from Annona squamosa L. Bark,” Phytomedicine, Vol. 17, No. 2, 2010, pp. 149-151.

doi:10.1016/j.phymed.2009.05.016

[53] R. Lobo, K. S. Prabhu, A. Shirwaikar and A. Shirwaikar, “Curcuma zedoaria Rosc. (White Turmeric): A Review of Its Chemical, Pharmacological and Ethnomedicinal Properties,” Journal of Pharmacy and Pharmacology, Vol. 61, No. 1, 2009, pp. 13-21.

[54] Z. Jiang, R.-Y. Chen, Y. Chen and D.-Q. Yu, “Donnaie- nin, a New Acetogenin Bearing a Hydroxylated Tetrahy- drofuran Ring,” Journal of Natural Products, Vol. 61, No. 1, 1998, pp. 86-88. doi:10.1021/np9703196

[55] W. Anne-Isabelle, R. Hocquemiller, A. Laurens and A. Cavé, “Glaucafilin, an Acetogenin from Annona glauca,” Journal of Phytochemistry, Vol. 44, No. 8, 1997, pp. 1537-1540.

[56] M. J. Rieser, Z. M. Gu, X. P. Fang, L. Zeng, K. V. Wood and J. L. McLaughlin, “Five Novel Mono-Tetrahydro- furan Ring Acetogenins from the Seeds of Annona muricata,” Journal of Natural Products, Vol. 59, No. 2, 1996, pp. 100-108. doi:10.1021/np960037q

http://dx.doi.org/10.1097/01.olq.0000152821.23777.90

http://dx.doi.org/10.1016/j.phymed.2009.05.016

http://dx.doi.org/10.1021/np9703196

http://dx.doi.org/10.1021/np960037q

Ar

Y

a

b

c

d

e

a

ARRAA

KATEBC

1

gwr

etd2

(

0d

Journal of Ethnopharmacology 137 (2011) 1283– 1290

Contents lists available at ScienceDirect

Journal of Ethnopharmacology

journa l h o me page: www.elsev ier .com/ locate / je thpharm

nnonacin induces cell cycle-dependent growth arrest and apoptosis in estrogeneceptor-�-related pathways in MCF-7 cells

u-Min Koa, Tung-Ying Wub, Yang-Chang Wub, Fang-Rong Changb, Jinn-Yuh Guhc,d,∗, Lea-Yea Chuange,∗

Graduate Institute of Medicine, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan, ROCGraduate Institute of Natural Products, College of Pharmacy, Kaohsiung Medical University, Kaohsiung, Taiwan, ROCDepartment of Internal Medicine, Faculty of Medicine, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan, ROCDepartment of Internal Medicine, Kaohsiung Medical University Hospital, Kaohsiung Medical University, 100 Zihyou 1st Road, Kaohsiung, Taiwan, ROCDepartment of Biochemistry, College of Medicine, Kaohsiung Medical University, 100 Zihyou 1st Road, Kaohsiung, Taiwan, ROC

r t i c l e i n f o

rticle history:eceived 16 February 2011eceived in revised form 11 July 2011ccepted 28 July 2011vailable online 4 August 2011

eywords:nnonacinamoxifenstrogen receptor-�reast cancerell cycle

a b s t r a c t

Ethnopharmacological relevance: Tamoxifen resistance is common in estrogen receptor-� (ER�)-positivebreast cancers. Pawpaw and soursop are anticancer annonaceous plants in complementary medicine.Thus, we studied the effects of annonacin, an annonaceous acetogenin, in breast cancer cells.Materials and Methods: Cell growth and ER�-related pathways were studied. The effects of annonacinwere tested in MCF-7 xenografts in nude mice.Results: In ER�-positive MCF-7 cells, annonacin (half-effective dose ED50 = 0.31 �M) and 4-hydroxytamoxifen (ED50 = 1.13 �M) decreased cell survival whereas annonacin (0.5-1 �M) increased celldeath at 48 h. Annonacin and 4-hydroxytamoxifen were additive in inhibiting cell survival. Annonacin(0.1 �M) induced G0/G1 growth arrest while increasing p21WAF1 and p27kip1 and decreasing cyclinD1 protein expression. Annonacin (0.1 �M) decreased cyclin D1 protein expression more than 4-hydroxytamoxifen (1 �M). Annonacin (0.1 �M) increased apoptosis while decreasing Bcl-2 proteinexpression. The combination of annonacin (0.1 �M) and 4-hydroxytamoxifen (1 �M) decreased Bcl-2protein expression and ER� transcriptional activity more than annonacin (0.1 �M) did alone. Annonacin,but not 4-hydroxytamoxifen, decreased ER� protein expression. Moreover, annonacin decreased phos-phorylation of ERK1/2, JNK and STAT3. In nude mice, annonacin decreased MCF-7 xenograft tumor size at
7–22 days. Moreover, annonacin decreased ER�, cyclin D1 and Bcl-2 protein expression in the xenograftat 22 days.Conclusions: Annonacin induced growth arrest and apoptosis in ER�-related pathways in MCF-7 cells.Annonacin and 4-hydroxytamoxifen were additive in inhibiting cell survival and ER� transcriptionalactivity. Moreover, annonacin attenuated MCF-7 xenograft tumor growth while inhibiting ER�, cyclin D1and Bcl-2 protein expressions in nude mice.
. Introduction

The pathogenesis of breast cancer includes estrogen and estro-
en receptor-� (ER�)-related pathways (Osborne and Schiff, 2011)hereby nuclear ER� activates target genes via the estrogen-
esponse elements (ERE) (Osborne and Schiff, 2011). Additionally,

Abbreviations: ER�, estrogen receptor-�; ED50, 50% effective dose; ERK,xtracellular signal-regulated kinase; JNK, c-Jun N-terminal kinase; STAT, signalransducers and activators of transcription; GAPDH, Glyceraldehyde 3-phosphateehydrogenase; cdk, cyclin-dependent kinase; MTT, 3-(4,5-dimethylthiazol-2-yl)-,5-diphenyl-tetrazolium bromide; PBS, phosphate-buffered saline.∗ Corresponding authors. Tel.: +886 7 3121101x7353; fax: +886 7 3218309.

E-mail addresses: [email protected] (J.-Y. Guh), [email protected]. Chuang).

378-8741/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.jep.2011.07.056

© 2011 Elsevier Ireland Ltd. All rights reserved.

nonnuclear ER� rapidly activates growth factor downstream sig-nals such as extracellular signal-regulated kinases (ERK1/2), c-JunN-terminal kinase (JNK) and signal transducers and activators oftranscription 3 (STAT3) (Musgrove and Sutherland, 2009; Osborneand Schiff, 2011). Interestingly, STAT integrates cytoplasmic andnuclear estrogen actions (Fox et al., 2009).

Selective estrogen receptor modulators (SERM, such as tamox-ifen) or selective estrogen receptor down-regulators (SERD, such asfulvestrant) are effective treatments for ER�-positive breast can-cers commonly limited by resistance (Osborne and Schiff, 2011).Endocrine resistance may be induced by the loss of ER�, increased
nonnuclear ER� or growth factor receptor signaling, deranged sig-nal transducers (such as ERK1/ERK1, JNK and STAT3), cell cycleregulators (such as cyclin D1 and the cyclin-dependent kinaseinhibitors p21WAF1 and p27kip1) or apoptosis regulators (such as
dx.doi.org/10.1016/j.jep.2011.07.056

http://www.sciencedirect.com/science/journal/03788741

http://www.elsevier.com/locate/jethpharm

mailto:[email protected]


dx.doi.org/10.1016/j.jep.2011.07.056

1 pharm

ta

c(bC2rarcn

apMm

2

2

wtaCwGac4

pnSbSpnc(TtmSdBesw

2

pwtScpim

284 Y.-M. Ko et al. / Journal of Ethno

he anti-apoptotic Bcl-2) (Musgrove and Sutherland, 2009; Osbornend Schiff, 2011).

Pawpaw (Asimina triloba) and soursop (graviola, Annona muri-ata) are annonaceous plants used as anticancer folk therapies inNorth, Central and South) America and Southeast Asia and haveeen studied in a few observational clinical studies (Cassileth, 2008;oothankandaswamy et al., 2010; Liaw et al., 2010; McLaughlin,008). Annonaceous acetogenins are cytotoxic to multidrug-esistant MCF-7 cells (Oberlies et al., 1997). Annonacin (C35H64O7),n annonaceous acetogenin containing a mono-tetrahydrofuraning with two flanking hydroxyls, also inhibits growth in MCF-7ells (Yuan et al., 2003). However, the molecular mechanisms areot understood.

Thus, we studied the growth-inhibitory mechanisms ofnnonacin in terms of ER�-related pathways (p-ERK1/2, p-JNK,-STAT3, cyclin D1, Bcl-2, p21WAF1 and p27kip1) in MCF-7 cells.oreover, the effects of annonacin on MCF-7 xenografts in nudeice were also investigated.

. Materials and methods

.1. Cell culture and reagents

ER�-positive MCF-7 cells and ER�-negative MDA-MB-231 cellsere purchased on February 18, 2009 from Bioresource Collec-

ion and Research Center (Hsinchu, Taiwan), where cells wereuthenticated by DNA fingerprints of short tandem repeat profiling.ells were cultured in DMEM/F-12 (1:1) medium supplementedith 1% penicillin/streptomycin and 10% fetal bovine serum (FBS,ibco, Grand Island, NY, USA) in a humidified 5% CO2 incubatort 37 ◦C. Cells were fasted for 24 h before adding fresh mediumontaining 10% FBS and various concentrations of annonacin or-hydroxytamoxifen.

Cyclin D1, cyclin-dependent kinase 4 (cdk4), cyclin E,21WAF1, p27kip1, Bcl-2, Glyceraldehyde 3-phosphate dehydroge-ase (GAPDH), ER�, STAT3 and JNK antibodies were purchased fromanta Cruz Biotechnology Inc. (Santa Cruz Co., CA). �-tubulin anti-ody was purchased from Lab Vision Corporation (Fremont, CA).erine 118 phosphorylated ER� (pSer118ER�), p-STAT3, ERK1/2,-ERK1/2, JNK and p-JNK antibodies were purchased from Cell Sig-aling Technology (Danvers, MA). We had isolated, purified andharacterized annonacin from the leaves of Formosan A. muricataLiaw et al., 2002) (supplementary Methods and supplementaryable 1), which has a different molecular structure from that ofamoxifen (supplementary Fig. 1). 4-hydroxytamoxifen, an active

etabolite of tamoxifen and other chemicals were purchased fromigma–Aldrich Chemical Company (St. Louis, MO). Annonacin wasissolved in dimethyl sulfoxide (DMSO, 0.1% final concentration).ecause annonacin has not been used in humans clinically, theffective concentration was chosen based on our previous in vitrotudy (Yuan et al., 2003). 17�-estradiol and 4-hydroxytamoxifenere dissolved in ethanol (0.1% final concentration).

.2. Measurement of cell survival and cell death

Cell survival was performed in quadruplicate similar to ourrevious study (Chou et al., 2008). Briefly, viable cell numbersere measured by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-

etrazolium bromide (MTT) assay. MTT was purchased fromigma–Aldrich Chemical Company (St. Louis, MO). Briefly, MCF-7
ells (6500 cells/well) or MDA-MB-231 cells (3500 cells/well) werelated and incubated in 96-well plates. MTT (1 mg/ml) was added
nto each well. Production of insoluble formazan by viable cells waseasured at 540 nm.

acology 137 (2011) 1283– 1290

Dose–response curves for various combinations of 4-hydroxytamoxifen and annonacin and the half-effective dose(ED50) for cell survival at 48 h were analyzed by the four-parameter log-logistic model (Sørensen et al., 2007). The doseranges used were: 0% 4-hydroxytamoxifen (i.e. annonacin, 0.01,0.05, 0.1, 0.5 and 1 �M), 50% 4-hydroxytamoxifen (0.1, 0.5, 1, 5and 10 �M), 91% 4-hydroxytamoxifen (0.1, 0.5, 1, 5 and 10 �M)and 100% 4-hydroxytamoxifen (0.1, 0.5, 1, 5 and 10 �M), respec-tively. The concentration addition model and the isobologrammethod (Sørensen et al., 2007) was used to assess the synergy,additivity or antagonism of the combinations of annonacin and4-hydroxytamoxifen.

Cell death was assessed by lactate dehydogenase releasethrough using the CytoTox 96 nonradioactive cytotoxicity assaykit according to the manufacturer’s instructions (Promega Corp.,Madison, WI).

2.3. Cell cycle analysis

Cell cycle analysis was performed similar to our previous study(Chuang et al., 2006). Briefly, cells were trypsinized and suspendedin cold phosphate-buffered saline (PBS). Suspended cells werewashed twice with cold PBS and fixed with 70% ice-cold ethanoland placed at −20 ◦C overnight. Cells were then centrifuged andresuspended with nuclear staining buffer (0.1% Triton X-100 in PBS,200 �g/ml RNase, and 100 �g/ml propidium iodide) for 30 min at37 ◦C. For each sample, at least 1 × 104 events were recorded. Cellcycle profiles were obtained with a FACScan flow cytometer (BectonDickinson Co., San Jose, CA) and data were analyzed with WinCyclesoftware (Phoenix Flow Systems Inc., San Diego, CA).

2.4. Immunoblotting

Immunoblotting was performed similar to our previous study(Guh et al., 2003). Briefly, a 30 �g sample of cell lysates waselectrophoresed on 10% sodium dodecyl sulfate-polyacrylamidegels, transferred to polyvinylidene difluoride membranes. Afterblocking, blots were incubated with antibody in blocking solutionovernight (phospho-antibodies) or for 2 h (other antibodies) fol-lowed by 5 min wash twice in PBS containing 0.1% Tween 20 andthen incubated with horseradish peroxidase-conjugated secondaryantibodies (Santa Cruz Biotechnologies Inc., Santa Cruz, CA) for 1 h.Enhanced chemiluminescence reagents were employed to depictprotein bands on the membrane.

2.5. Measurement of apoptosis

Apoptosis was measured by flow cytometry by using theVybrantTM Apoptosis Assay Kit #2 (Alexa Fluor® 488 annexinV/Propidium Iodide kit #2) supplied by Molecular Probes Inc.(Eugene, OR) according to the manufacturer’s instructions. This kitdetects the externalization of phosphatidylserine in apoptotic cellsusing the green-fluorescent Alexa Fluor® 488 annexin and the red-fluorescent propidium iodide nucleic acid stain. Propidium iodidestains necrotic cells with red fluorescence. After treatment withboth probes, apoptotic cells show green fluorescence, dead cellsshow red and green fluorescence, and live cells show little or nofluorescence.

2.6. Transient transfection and luciferase assay

Transient transfection was performed similar to our previousstudy (Chou et al., 2008). Briefly, MCF-7 cells were plated into 6-well plates at density of 1.2 × 105 cells/well in DMEM/F12 mediumand grown overnight. Cells were transfected with 0.2 �g of the

Y.-M. Ko et al. / Journal of Ethnopharmacology 137 (2011) 1283– 1290 1285

Fig. 1. Time-dependent effects of annonacin on cell cycle distribution and cell death in MCF-7 cells. MCF-7 cells (2.5 × 105 cells/dish) were plated and incubated in 6 cmdishes. Cells were fasted for 24 h before adding fresh medium (10% FBS) containing 0.1% DMSO or 0.1 �M annonacin. Cell cycle was measured by flow cytometry. Apoptosiswas measured by flow cytometry by using the VybrantTM Apoptosis Assay Kit #2 (Alexa Fluor® 488 annexin V/Propidium Iodide kit #2). (A) Time-dependent (1–48 h) effectsof annonacin (0.1 �M) on cell cycle distribution. (B) Time-dependent (6–48 h) effects of annonacin (0.1 �M) on apoptosis. (C) Dose-dependent effects of annonacin on celld Cells w0 ath wm

ET(at

2

pT2ititotTamTh

eath. MCF-7 cells (6500 cells/well) were plated and incubated in 96-well plates.

.1% DMSO (open bars) or various concentrations of annonacin (gray bars). Cell deean ± SEM of three independent experiments. *P < 0.05 versus DMSO.

R� transcriptional activity reporter plasmid containing 3X ERE-ATA luc (Addgene Inc., Cambridge, MA) with LipofectAMINE 2000Life Technology, Gaithersburg, MD) and luciferase activities weressayed by integrating the total light emission over 10 s by usinghe Dynatech ML1000 luminometer.

.7. In vivo tumor xenograft study

Female nude mice (4 weeks old; BALB/cA-nu [nu/nu]) wereurchased from the National Laboratory Animal Center (Taipei,aiwan) and were housed in specific pathogen-free conditions for

weeks. Afterwards, MCF-7 cells (5 × 106 cells in 200 �L PBS) werenjected subcutaneously into the flanks, and tumors were allowedo develop for 30 days. Then eleven mice were randomly dividednto two groups. Annonacin-treated mice (N = 6) were intraperi-oneally injected daily with annonacin (50 mg/kg/day) in 200 �Lf 25% polyethylene glycol. Control mice (N = 5) were intraperi-oneally injected daily with 200 �L of 25% polyethylene glycol.umor volume was measured using calipers at 0, 3, 7, 11, 15, 19
nd 22 days. Tumor volume was estimated by the following for-ula: tumor volume (mm3) = L × W × W/2 (L: length; W: width).
he xenografts of the annonacin-treated and the control mice werearvested and fixed in 4% formaldehyde, embedded in paraffin

ere fasted for 24 h before adding fresh medium (10% FBS, open bars) containingas measured by lactate dehydrogenase release. The results were expressed as the

for immunohistochemistry at 22 days. All animal procedures wereapproved and done in accordance with the national guidelines andthe guidelines by the Kaohsiung Medical University Animal Exper-iment Committee.

2.8. Immunohistochemistry

Paraffin-embedded tumor tissues were cut to 4 �m sections forimmunohistochemistry. The sections were treated with microwaveat 100 ◦C for 30 min, and blocked nonspecific response. The sectionswere incubated at 4 ◦C overnight with primary antibodies (ER�,cyclin D1 and Bcl-2). After washing twice 10 min with PBS contain-ing 0.2% Tween 20, the sections were incubated with biotinylatedsecondary antibodies for 1 h. After washing twice with PBS con-tain 0.2% Tween 20 for 10 min, sections were stained by UniversalDAB + kit/HRP (Dako Corp., Carpinteria, CA) and counterstainedwith hematoxylin.

2.9. Statistical analysis

The results were expressed as the mean ± standard errors ofthe mean. Unpaired Student’s t-tests were used for the compar-ison between two groups. P < 0.05 was considered as statistically

1286 Y.-M. Ko et al. / Journal of Ethnopharm

Fig. 2. Effects of annonacin on cell cycle regulatory proteins in MCF-7 cells. MCF-7 cells (2.5 × 105 cells/dish) were plated and incubated in 6 cm dishes. Cells werefasted for 24 h before adding fresh medium (10% FBS) containing 0.1% DMSO or0.1 �M annonacin. Cell cycle regulatory proteins (cyclin D1, cdk4, cyclin E, p21WAF1

and p27kip1) were measured by immunoblotting. (A) Time-dependent effects ofaefi

sm((

3

3i

ip(d(d26

3p

iad((d(

3a

bdt

To confirm the growth-inhibitory effects of annonacin in vivo,

nnonacin (0.1 �M) on p21WAF1 and p27kip1 protein expression. (B) Time-dependentffects of annonacin (0.1 �M) on cyclin D1, cdk4 and cyclin E protein expression. Thisgure is representative of three independent experiments.

ignificant. Concentration addition model and the isobologramethod (Sørensen et al., 2007) were assessed by the drc

dose–response curves) package in the R statistical programKnezevic et al., 2009).

. Results

.1. Effects of annonacin on cell cycle distribution and cell deathn MCF-7 cells

Annonacin (0.1 �M) time-dependently (6–48 h) arrested cellsn the G0/G1 phase of the cell cycle (Fig. 1A) while increasing apo-tosis at 48 h (Fig. 1B). Additionally, annonacin dose-dependently0.5–1 �M) increased cell death at 48 h (Fig. 1C). Annonacin alsoose-dependently (0.05–1 �M) decreased cell survival at 48 hsupplementary Fig. 2A). Moreover, annonacin (0.1 �M) time-ependently (24–48 h) decreased cell survival (supplementary Fig.B). However, annonacin (0.1 �M) did not affect cell survival at–48 h in MDA-MB-231 cells (supplementary Fig. 2C).

.2. Time-dependent effects of annonacin on cell cycle regulatoryroteins in MCF-7 cells

Annonacin (0.1 �M) time-dependently (30 min to 12 h)ncreased p21WAF1 but decreased p21WAF1 protein expressiont 24–48 h (Fig. 2A). In contrast, annonacin (0.1 �M) time-ependently increased p27kip1 protein expression at 1–6 hFig. 2A). Additionally, annonacin (0.1 �M) time-dependently1–48 h) decreased cyclin D1 and time-dependently (12–48 h)ecreased cdk4 protein expression, but not that of cyclin EFig. 2B).

.3. Effects of annonacin and 4-hydroxytamoxifen on cell survivalnd cyclin D1 or Bcl-2 protein expression in MCF-7 cells

Annonacin (0.1 �M), 4-hydroxytamoxifen (1 �M) and a com-ination of annonacin (0.1 �M) plus 4-hydroxytamoxifen (1 �M)ecreased cell survival to a similar degree at 24 h (Fig. 3A). However,he combination of annonacin (0.1 �M) plus 4-hydroxytamoxifen

acology 137 (2011) 1283– 1290

(1 �M) decreased cell survival more than annonacin (0.1 �M) didalone at 48 h (Fig. 3A).

The estimated half-effective doses from dose–responsecurves at 48 h were: 0% 4-hydroxytamoxifen (i.e. annonacin,ED50 = 0.31 �M), 50% 4-hydroxytamoxifen (ED50 = 0.45 �M), 91%4-hydroxytamoxifen (ED50 = 1.01 �M), 100% 4-hydroxytamoxifen(ED50 = 1.13 �M), respectively. Isobologram analysis showedthat the ED50 of the various combinations (50% and 91% 4-hydroxytamoxifen) coincided with the estimated concentrationaddition isobole (Fig. 3B). In other words, the combination effectsof annonacin and 4-hydroxytamoxifen were additive instead ofsynergistic (ED50 of the various combinations shifted to left) orantagonistic (ED50 of the various combinations shifted to right).

As shown in supplementary Fig. 3A, annonacin (0.01–2 �M)evenly decreased cyclin D1 and Bcl-2 protein expression at48 h. Additionally, annonacin (0.1 �M) time-dependently (1–48 h)decreased Bcl-2 protein expression (supplementary Fig. 3B).Annonacin (0.1 �M) decreased cyclin D1 protein expressionmore than 4-hydroxytamoxifen (1 �M) did at 48 h (Fig. 3C). Incontrast, annonacin (0.1 �M) and 4-hydroxytamoxifen (1 �M)decreased Bcl-2 protein expression to a similar degree at 48 h(Fig. 3C). Moreover, the combination of annonacin (0.1 �M) plus4-hydroxytamoxifen (1 �M) decreased Bcl-2 protein expressionmore than annonacin (0.1 �M) did alone at 48 h (Fig. 3C).

3.4. Effects of annonacin and 4-hydroxytamoxifen on ER˛transcriptional activity, ER˛ protein expression andphosphorylation in MCF-7 cells

As shown in Fig. 4A, annonacin (0.1 �M) and 4-hydroxytamoxifen (1 �M) decreased ER� transcriptional activityto the same degree at 24–48 h. Moreover, the combination ofannonacin (0.1 �M) plus 4-hydroxytamoxifen (1 �M) decreasedER� transcriptional activity more than annonacin (0.1 �M) didalone at 24–48 h (Fig. 4A).

As shown in supplementary Fig. 4A, 17�-estradiol increased ER�transcriptional activity whereas annonacin (0.1 �M) attenuatedER� transcriptional activity at 12–48 h. Additionally, annonacin(0.1 �M) decreased ER� protein expression at 24 h and decreasedER� protein serine 118 phosphorylation at 12 h (supplementaryFig. 4B). Annonacin dose-dependently (0.01–2 �M) decreasedER� protein expression at 48 h (Fig. 4B). In contrast, 4-hydroxytamoxifen (1 �M) did not affect ER� protein expression at48 h (Fig. 4C). Finally, the combination of annonacin (0.1 �M) plus4-hydroxytamoxifen (1 �M) decreased ER� protein expression toa similar degree as annonacin (0.1 �M) did alone at 48 h (Fig. 4C).

3.5. Time-dependent effects of annonacin on phosphorylation ofERK1/2, JNK and STAT3 protein in MCF-7 cells

Annonacin (0.1 �M) time-dependently (1–48 h) decreasedERK1/2 protein phosphorylation (Fig. 5A). Annonacin (0.1 �M) alsotime-dependently (1–48 h) decreased JNK protein phosphoryla-tion (Fig. 5B). Additionally, annonacin (0.1 �M) time-dependently(30 min to 48 h) decreased STAT3 protein phosphorylation (Fig. 5C).

3.6. Effects of annonacin on MCF-7 xenograft tumor size andexpression of ER˛, cyclin D1 and Bcl-2 proteins in nude mice

MCF-7 cells were grafted into the flanks of nude mice as xenografts.As shown in Fig. 6A, annonacin decreased tumor size at 7–22 days.As shown in Fig. 6B, annonacin attenuated the expression of ER�,cyclin D1 and Bcl-2 protein in the nude mice at 22 days.


Fig. 3. Combination effects of annonacin and 4-hydroxytamoxifen on cell survival and cyclin D1 and Bcl-2 protein expression in MCF-7 cells. MCF-7 cells (6500 cells/well)were plated and incubated in 96-well plates. Cells were fasted for 24 h before adding fresh medium (10% FBS) containing 0.1% DMSO or 0.1% ethanol. Cell survival wasmeasured by the MTT assay. (A) Effects of annonacin (A, 0.1 �M), 4-hydroxytamoxifen (T, 1 �M) or their combination (A + T) on cell survival at 24 h (lanes 4–6) or 48 h (lanes7–9). *P < 0.05 versus control. #P < 0.05 versus annonacin alone. (B) Isobologram analysis of various combinations (50% 4-hydroxytamoxifen and 91% 4-hydroxytamoxifen)of annonacin and 4-hydroxytamoxifen at 48 h. Dose combinations (dotted lines) achieving ED50 (dots) with standard errors together with the estimated isoboles are shown.The estimated concentration addition isobole is shown as the solid line connecting the 100% and 0% 4-hydroxytamoxifen ED50 dots. MCF-7 cells (2.5 × 105 cells/dish) werep measu( and

e

4

i2dtSat

a(mSgaau(

sap2

lated and incubated in 6 cm dishes. Expression of cyclin D1 and Bcl-2 protein was

A, 0.1 �M), 4-hydroxytamoxifen (T, 1 �M) or their combination (A + T) on cyclin D1xperiments.

. Discussion and conclusions

This study adds mechanistic insights to our previous find-ng that annonacin inhibits growth in MCF-7 cells (Yuan et al.,003).We found that annonacin induced cell death only at highoses (≥0.5 �M). In contrast, low-dose (0.1 �M) annonacin, likeamoxifen and fulvestrant (Cariou et al., 2000; Musgrove andutherland, 2009), induced cell-cycle-dependent (G0/G1) growthrrest concomitantly with the induction of p21WAF1 and p27kip andhe inhibition of cyclin D1 protein expression.

Cell cycle (G0/G1 transition) inhibitors p21WAF1 and p27kip

re implicated in the pathogenesis or prognosis of breast cancerCaldon et al., 2006). For example, the loss of p21WAF1 or p27kip

ay mediate tamoxifen resistance in breast cancer (Musgrove andutherland, 2009). Conversely, transfection of p21WAF1 or p27kip

enes inhibits MCF-7 cell growth (Jiang et al., 2009). Surprisingly,nnonacin inhibited both p21WAF1 and cyclin D1 protein expressiont 24–48 h. However, inhibition of p21WAF1 alone cannot atten-ate cyclin D1 inhibition-induced growth arrest in cancer cellsMasamha and Benbrook, 2009).

Our finding that annonacin inhibited cyclin D1 protein expres-
ion is compatible with the notion that cyclin D1 overexpression isssociated with poor prognosis and tamoxifen resistance in ER�-ositive breast cancer (Butt et al., 2008; Musgrove and Sutherland,009). Interestingly, cyclin D1 overexpression induces whereas
red by immunoblotting and normalized to that of GAPDH. (C) Effects of annonacinBcl-2 protein expression at 48 h. This figure is representative of three independent

cyclin D1 knockout inhibits breast cancer in mice (Butt et al., 2008).Similarly, cyclin D1 overexpression increases whereas cyclin D1inhibition decreases proliferation in MCF-7 cells (Grillo et al., 2006).

Annonacin, like tamoxifen and fulvestrant (Lam et al., 2009;Musgrove and Sutherland, 2009), also induced apoptosis whileinhibiting Bcl-2 protein expression. Note that cell survival, celldeath and cell cycle pathways are interconnected in cancers(Maddika et al., 2007). For example, apoptosis is a cell-cyclecheckpoint during cell injury and Bcl-2 is an anti-apoptotic pro-tein which also inhibits the G1/S checkpoint of the cell cycle(Maddika et al., 2007; Musgrove and Sutherland, 2009). Transfec-tion of the Bcl-2 gene inhibits whereas inhibition of the Bcl-2 geneenhances chemosensitivity of breast cancer cells (Emi et al., 2005).Moreover, Bcl-2 protein expression is associated with tamoxifenresistance in breast cancer patients (Musgrove and Sutherland,2009).

The observation that annonacin decreased cell survival inER�-positive MCF-7 cells, but not ER�-negative MDA-MB-231cells, suggests that ER� is required for the effects of annonacin.Interestingly, annonacin (ED50 = 0.31 �M) was more potent than4-hydroxytamoxfien (ED50 = 1.13 �M), while annonacin and 4-
hydroxytamoxifen were additive, in inhibiting cell survival.Similarly, annonacin was more potent than 4-hydroxytamoxifen,and annonacin and 4-hydroxytamoxifen were additive, in inhibit-ing ER� transcriptional activity.

1288 Y.-M. Ko et al. / Journal of Ethnopharmacology 137 (2011) 1283– 1290

Fig. 4. Effects of annonacin and 4-hydroxytamoxifen on ER� transcriptional activity and ER� protein expression in MCF-7 cells. MCF-7 cells (1.2 × 105 cells/well) were platedand incubated in 6-well plates. Cells were fasted for 24 h before adding fresh medium (M, 10% FBS) containing 0.1% DMSO (D) or 0.1% ethanol. (A) Effects of annonacin(0.1 �M), 4-hydroxytamoxifen (1 �M) or their combination on ER� transcriptional activity at 24 h or 48 h. Cells were transfected with 3X ERE-TATA luc with LipofectAMINE2000 and luciferase activities were assayed by integrating the total light emission over 10 s by using the Dynatech ML1000 luminometer. *P < 0.05 versus control. #P < 0.05versus annonacin alone. MCF-7 cells (2.5 × 105 cells/dish) were plated and incubated in 6 cm dishes. Cells were fasted for 24 h before adding fresh medium containing 0.1%DMSO, 0.1% ethanol, annonacin or 4-hydroxytamoxifen. Expression of ER� protein was measured by immunoblotting. (B) Dose-dependent effects of annonacin (A, 0.01-2 �M) on ER� protein expression at 48 h. Expression of ER� was normalized to that of �-tubulin. (C) Effects of annonacin (A, 0.1 �M), 4-hydroxytamoxifen (T, 1 �M) or theircombination (A + T) on ER� protein expression at 48 h. Expression of ER� was normalized to that of GAPDH. Both figures B and C are representative of three independentexperiments.

Fig. 5. Time-dependent effects of annonacin on phosphorylation of ERK1/2, JNK and STAT3 protein in MCF-7 cells. MCF-7 cells (2.5 × 105 cells/dish) were plated and incubatedin 6 cm dishes. Cells were fasted for 24 h before adding fresh medium (M, 10% FBS) containing 0.1% DMSO (D) or annonacin for 30 min to 48 h. Phosphorylation of ERK1/2,JNK and STAT3 protein was measured by immunoblotting and normalized to that of ERK1/2, JNK and STAT3. (A) Time-dependent effects of annnonacin (0.1 �M) on p-ERK1/2protein phosphorylation. (B) Time-dependent effects of annnonacin (0.1 �M) on p-JNK protein phosphorylation. (C) Time-dependent effects of annnonacin (0.1 �M) onp-STAT3 protein phosphorylation. This figure is representative of three independent experiments.


Fig. 6. Effects of annonacin on MCF-7 xenograft tumor size and expression of ER�, cyclin D1 and Bcl-2 protein in the nude mice. MCF-7 cells (5 × 106 cells in 200 �L PBS) weregrafted into the flanks of the nude mice as xenografts, and tumors were allowed to develop for 30 days. Afterwards, annonacin (50 mg/kg/day) or vehicle (25% polyethyleneg suredd for thm .

aFbHptaiidsttoi

MSto2ttai

iEI(

lycol) was administered intraperitoneally for 21 days. (A) Tumor volume was meaays). (B) Tumor was excised at 22 days and immunohistochemistry was performedean ± SEM of 5 mice in each group. *P < 0.05, **P < 0.01, ***P < 0.001 versus vehicle

The mode of action of annonacin can be studied by ER� bindingssay and ER� transcriptional activation assay (Dang et al., 2011).or example, SERM and SERD usually act by inhibiting ER� ligand-inding. We did not study the ER�-binding ability of annonacin.owever, we found that annonacin, like SERD, also decreases ER�rotein expression. Thus, annonacin, like SERD, may be useful inamoxifen resistance (Osborne and Schiff, 2011). We also found thatnnonacin decreased ERE-dependent ER� transcriptional activ-ty. Note that ERE is indispensable for ER�-induced phenotypesn breast cancer cells (Nott et al., 2009). Moreover, annonacinecreased ER� serine 118 phosphorylation, which is required forome ER�-induced effects (Duplessis et al., 2011). In contrast, bothamoxifen and fulvestrant increase ER� serine 118 phosphoryla-ion (Lipfert et al., 2006; Maggi, 2011). However, there are manyther possible modes of action whereby annonacin can inhibit ER�-nduced effects (Shapiro et al., 2011).

Annonacin, unlike tamoxifen (Ishii et al., 2008; Lam et al., 2009;usgrove and Sutherland, 2009), inactivated ERK1/2, JNK and

TAT3 in this study. Interestingly, inhibition of ERK1/2 attenuatesamoxifen resistance (Ghayad et al., 2010) and inhibition of JNKr STAT3 induces apoptosis in breast cancer cells (Kunigal et al.,009; Mingo-Sion et al., 2004). In summary, annonacin differs fromamoxifen in terms of ER� protein abundance and phosphoryla-ion, p-ERK1/2, p-JNK and p-STAT3. These observations may partlyccount for the finding that annonacin and tamoxifen were additiven inhibiting MCF-7 cell growth.

The in vitro effects of annonacin were corroborated by our find-
ngs that annonacin attenuated tumor size and the expression ofR�, cyclin D1 and Bcl-2 protein in MCF-7 cell-grafted nude mice.n this regard, cyclin D1 is overexpressed in 50% of breast cancersCaldon et al., 2006). Breast epithelial cell-specific overexpression
every 3–4 days using calipers at 0, 3, 7, 11, 15, 19 and 22 days (inset: tumor at 22e expression of ER�, cyclin D1 and Bcl-2 protein. The results were expressed as the

of cyclin D1 induces breast cancer in mice, while cyclin D1-null miceare resistant to oncogene-induced breast cancer (Butt et al., 2008).Moreover, inhibition of Bcl-2 in combination with chemotherapywas effective in some breast cancer patients in a clinical trial (Fatoet al., 2008).

In conclusion, annonacin induced cell-cycle-dependent growtharrest and induced apoptosis in ER�-related pathways (ERK1/2,JNK, STAT3, cyclin D1, Bcl-2, p21WAF1 and p27kip1) in MCF-7 cells.Annonacin and 4-hydroxytamoxifen were additive in inhibitinggrowth and ER� transcriptional activity. Moreover, annonacinattenuated MCF-7 xenograft tumor growth while inhibiting ER�,cyclin D1 and Bcl-2 protein expression in nude mice.

Acknowledgement

This work was supported by the National Science Council ofTaiwan (NSC-94-2321-B-037-006) to Lea-Yea Chuang.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.jep.2011.07.056.

References

Butt, A., et al., 2008. Cell cycle machinery: links with genesis and treatment of breast
cancer. Advances in Experimental Medicine and Biology 630, 189–205.
Caldon, E., et al., 2006. Cell cycle control in breast cancer cells. Journal of CellularBiochemistry 97, 261–274.

Cariou, S., et al., 2000. Down-regulation of p21WAF1/CIP1 or p27Kip1 abro-gates antiestrogen-mediated cell cycle arrest in human breast cancer cells.

http://dx.doi.org/10.1016/j.jep.2011.07.056

1 pharm

C

C

C

C

D

D

E

F

F

G

G

G

I

J

K


Proceedings of the National Academy of Sciences of the United States of America97, 9042–9046.

assileth, B., 2008. Complementary therapies, herbs, and other OTC agents. Oncology(Williston Park) 22, 1202.

hou, W.-W., et al., 2008. Arecoline-induced growth arrest and p21WAF1 expressionare dependent on p53 in rat hepatocytes. Toxicology 243, 1–10.

huang, L.-Y., et al., 2006. Role of nitric oxide in high glucose-induced mitogenicresponse in renal fibroblasts. Molecular Endocrinology 20, 2548–2558.

oothankandaswamy, V., et al., 2010. The alternative medicine pawpaw and its ace-togenin constituents suppress tumor angiogenesis via the HIF-1/VEGF pathway.Journal of Natural Products 73, 956–961.

ang, Z., et al., 2011. Comparison of chemical-induced transcriptional activation offish and human estrogen receptors: regulatory implications. Toxicology Letters201, 152–175.

uplessis, T.T., et al., 2011. Phosphorylation of estrogen receptor � at serine 118directs recruitment of promoter complexes and gene-specific transcription.Endocrinology 152, 2517–2526.

mi, M., et al., 2005. Targeted therapy against Bcl-2-related proteins in breast cancercells. Breast Cancer Research 7, R940–R952.

ato, R., et al., 2008. Generation of reactive oxygen species by mitochondrialcomplex I: implications in neurodegeneration. Neurochemical Research 33,2487–2501.

ox, E., et al., 2009. Novel actions of estrogen to promote proliferation: integrationof cytoplasmic and nuclear pathways. Steroids 74, 622–627.

hayad, S., et al., 2010. Endocrine resistance associated with activated ErbB sys-tem in breast cancer cells is reversed by inhibiting MAPK or PI3K/Akt signalingpathways. International Journal of Cancer 126, 545–562.

rillo, M., et al., 2006. Validation of cyclin D1/CDK4 as an anticancer drug targetin MCF-7 breast cancer cells: effect of regulated overexpression of cyclin D1and siRNA-mediated inhibition of endogenous cyclin D1 and CDK4 expression.Breast Cancer Research and Treatment 95, 185–194.

uh, J.-Y., et al., 2003. Beta-hydroxybutyrate–induced growth inhibition and col-lagen production in HK-2 cells are dependent on TGF-beta and Smad3. KidneyInternational 64, 2041–2051.

shii, Y., et al., 2008. Tamoxifen stimulates the growth of cyclin D1-overexpressingbreast cancer cells by promoting the activation of signal transducer and activatorof transcription 3. Cancer Research 68, 852–860.

iang, D.-D., et al., 2009. [Effects of transfected pIRES-p21waf1-p27kip1 gene on theduplication of centrosomes and the proliferation of breast cancer cell line MCF-7]. Zhonghua Yi Xue Za Zhi 89, 3295–3298.

nezevic, S.Z., et al., 2009. Utilizing R software package for dose–response studies:the concept and data analysis. Weed Technology 21, 840–848.

acology 137 (2011) 1283– 1290

Kunigal, S., et al., 2009. Stat3-siRNA induces Fas-mediated apoptosis in vitro andin vivo in breast cancer. International Journal of Oncology 34, 1209–1220.

Lam, L., et al., 2009. Tamoxifen and ICI 182,780 increase Bcl-2 levels and inhibitgrowth of breast carcinoma cells by modulating PI3K/AKT, ERK and IGF-1Rpathways independent of ERalpha. Breast Cancer Research and Treatment 118,605–621.

Liaw, C.-C., et al., 2002. New cytotoxic monotetrahydrofuran annonaceous aceto-genins from Annona muricata. Journal of Natural Products 65, 470–475.

Liaw, C.C., et al., 2010. Historic perspectives on Annonaceous acetogenins from thechemical bench to preclinical trials. Planta Medica 76, 1390–1404.

Lipfert, L., et al., 2006. Antagonist-induced, activation function-2-independent estro-gen receptor alpha phosphorylation. Molecular Endocrinology 20, 516–533.

Maddika, S., et al., 2007. Cell survival, cell death and cell cycle pathways are inter-connected: implications for cancer therapy. Drug Resistance Updates 10, 13–29.

Maggi, A., 2011. Liganded and unliganded activation of estrogen receptor and hor-mone replacement therapies. Biochimica et Biophysica Acta 1812, 1054–1060.

Masamha, C.P., Benbrook, D.M., 2009. Cyclin D1 degradation is sufficient to induceG1 cell cycle arrest despite constitutive expression of cyclin E2 in ovarian cancercells. Cancer Research 69, 6565–6572.

McLaughlin, J.L., 2008. Paw paw and cancer: annonaceous acetogenins from discov-ery to commercial products. Journal of Natural Products 71, 1311–1321.

Mingo-Sion, A., et al., 2004. Inhibition of JNK reduces G2/M transit independentof p53, leading to endoreduplication, decreased proliferation, and apoptosis inbreast cancer cells. Oncogene 23, 596–604.

Musgrove, E., Sutherland, R., 2009. Biological determinants of endocrine resistancein breast cancer. Nature Reviews. Cancer 9, 631–643.

Nott, S.L., et al., 2009. Genomic responses from the estrogen-responsive element-dependent signaling pathway mediated by estrogen receptor alpha are requiredto elicit cellular alterations. Journal of Biological Chemistry 284, 15277–15288.

Oberlies, N.H., et al., 1997. Structure−activity relationships of diverse annonaceousacetogenins against multidrug resistant human mammary adenocarcinoma(MCF-7/Adr) cells. Journal of Medicinal Chemistry 40, 2102–2106.

Osborne, C.K., Schiff, R., 2011. Mechanisms of endocrine resistance in breast cancer.Annual Review of Medicine 62, 233–247.

Sørensen, H., et al., 2007. An isobole-based statistical model and test for syner-gism/antagonism in binary mixture toxicity experiments. Environmental andEcological Statistics 14, 383–397.

Shapiro, D.J, et al., 2011. Small molecule inhibitors as probes for estrogen and andro-gen receptor action. Journal of Biological Chemistry 286, 4043–4048.

Yuan, S.-S., et al., 2003. Annonacin, a mono-tetrahydrofuran acetogenin, arrests can-cer cells at the G1 phase and causes cytotoxicity in a Bax- and caspase-3-relatedpathway. Life Sciences 72, 2853–2861.

Ar

Y

a

b

c

d

e

a

ARRAA

KATEBC

1

gwr

etd2

(

0d

Journal of Ethnopharmacology 137 (2011) 1283– 1290

Contents lists available at ScienceDirect

Journal of Ethnopharmacology

journa l h o me page: www.elsev ier .com/ locate / je thpharm

nnonacin induces cell cycle-dependent growth arrest and apoptosis in estrogeneceptor-�-related pathways in MCF-7 cells

u-Min Koa, Tung-Ying Wub, Yang-Chang Wub, Fang-Rong Changb, Jinn-Yuh Guhc,d,∗, Lea-Yea Chuange,∗

Graduate Institute of Medicine, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan, ROCGraduate Institute of Natural Products, College of Pharmacy, Kaohsiung Medical University, Kaohsiung, Taiwan, ROCDepartment of Internal Medicine, Faculty of Medicine, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan, ROCDepartment of Internal Medicine, Kaohsiung Medical University Hospital, Kaohsiung Medical University, 100 Zihyou 1st Road, Kaohsiung, Taiwan, ROCDepartment of Biochemistry, College of Medicine, Kaohsiung Medical University, 100 Zihyou 1st Road, Kaohsiung, Taiwan, ROC

r t i c l e i n f o

rticle history:eceived 16 February 2011eceived in revised form 11 July 2011ccepted 28 July 2011vailable online 4 August 2011

eywords:nnonacinamoxifenstrogen receptor-�reast cancerell cycle

a b s t r a c t

Ethnopharmacological relevance: Tamoxifen resistance is common in estrogen receptor-� (ER�)-positivebreast cancers. Pawpaw and soursop are anticancer annonaceous plants in complementary medicine.Thus, we studied the effects of annonacin, an annonaceous acetogenin, in breast cancer cells.Materials and Methods: Cell growth and ER�-related pathways were studied. The effects of annonacinwere tested in MCF-7 xenografts in nude mice.Results: In ER�-positive MCF-7 cells, annonacin (half-effective dose ED50 = 0.31 �M) and 4-hydroxytamoxifen (ED50 = 1.13 �M) decreased cell survival whereas annonacin (0.5-1 �M) increased celldeath at 48 h. Annonacin and 4-hydroxytamoxifen were additive in inhibiting cell survival. Annonacin(0.1 �M) induced G0/G1 growth arrest while increasing p21WAF1 and p27kip1 and decreasing cyclinD1 protein expression. Annonacin (0.1 �M) decreased cyclin D1 protein expression more than 4-hydroxytamoxifen (1 �M). Annonacin (0.1 �M) increased apoptosis while decreasing Bcl-2 proteinexpression. The combination of annonacin (0.1 �M) and 4-hydroxytamoxifen (1 �M) decreased Bcl-2protein expression and ER� transcriptional activity more than annonacin (0.1 �M) did alone. Annonacin,but not 4-hydroxytamoxifen, decreased ER� protein expression. Moreover, annonacin decreased phos-phorylation of ERK1/2, JNK and STAT3. In nude mice, annonacin decreased MCF-7 xenograft tumor size at
7–22 days. Moreover, annonacin decreased ER�, cyclin D1 and Bcl-2 protein expression in the xenograftat 22 days.Conclusions: Annonacin induced growth arrest and apoptosis in ER�-related pathways in MCF-7 cells.Annonacin and 4-hydroxytamoxifen were additive in inhibiting cell survival and ER� transcriptionalactivity. Moreover, annonacin attenuated MCF-7 xenograft tumor growth while inhibiting ER�, cyclin D1and Bcl-2 protein expressions in nude mice.
. Introduction

The pathogenesis of breast cancer includes estrogen and estro-
en receptor-� (ER�)-related pathways (Osborne and Schiff, 2011)hereby nuclear ER� activates target genes via the estrogen-
esponse elements (ERE) (Osborne and Schiff, 2011). Additionally,

Abbreviations: ER�, estrogen receptor-�; ED50, 50% effective dose; ERK,xtracellular signal-regulated kinase; JNK, c-Jun N-terminal kinase; STAT, signalransducers and activators of transcription; GAPDH, Glyceraldehyde 3-phosphateehydrogenase; cdk, cyclin-dependent kinase; MTT, 3-(4,5-dimethylthiazol-2-yl)-,5-diphenyl-tetrazolium bromide; PBS, phosphate-buffered saline.∗ Corresponding authors. Tel.: +886 7 3121101x7353; fax: +886 7 3218309.

E-mail addresses: [email protected] (J.-Y. Guh), [email protected]. Chuang).

378-8741/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.jep.2011.07.056

© 2011 Elsevier Ireland Ltd. All rights reserved.

nonnuclear ER� rapidly activates growth factor downstream sig-nals such as extracellular signal-regulated kinases (ERK1/2), c-JunN-terminal kinase (JNK) and signal transducers and activators oftranscription 3 (STAT3) (Musgrove and Sutherland, 2009; Osborneand Schiff, 2011). Interestingly, STAT integrates cytoplasmic andnuclear estrogen actions (Fox et al., 2009).

Selective estrogen receptor modulators (SERM, such as tamox-ifen) or selective estrogen receptor down-regulators (SERD, such asfulvestrant) are effective treatments for ER�-positive breast can-cers commonly limited by resistance (Osborne and Schiff, 2011).Endocrine resistance may be induced by the loss of ER�, increased
nonnuclear ER� or growth factor receptor signaling, deranged sig-nal transducers (such as ERK1/ERK1, JNK and STAT3), cell cycleregulators (such as cyclin D1 and the cyclin-dependent kinaseinhibitors p21WAF1 and p27kip1) or apoptosis regulators (such as
dx.doi.org/10.1016/j.jep.2011.07.056


http://www.elsevier.com/locate/jethpharm



dx.doi.org/10.1016/j.jep.2011.07.056

fcdsadmin

Highlight

1 pharm

ta

c(bC2rarcn

apMm

2

2

wtaCwGac4

pnSbSpnc(TtmSdBesw

2

pwtScpim


he anti-apoptotic Bcl-2) (Musgrove and Sutherland, 2009; Osbornend Schiff, 2011).

Pawpaw (Asimina triloba) and soursop (graviola, Annona muri-ata) are annonaceous plants used as anticancer folk therapies inNorth, Central and South) America and Southeast Asia and haveeen studied in a few observational clinical studies (Cassileth, 2008;oothankandaswamy et al., 2010; Liaw et al., 2010; McLaughlin,008). Annonaceous acetogenins are cytotoxic to multidrug-esistant MCF-7 cells (Oberlies et al., 1997). Annonacin (C35H64O7),n annonaceous acetogenin containing a mono-tetrahydrofuraning with two flanking hydroxyls, also inhibits growth in MCF-7ells (Yuan et al., 2003). However, the molecular mechanisms areot understood.

Thus, we studied the growth-inhibitory mechanisms ofnnonacin in terms of ER�-related pathways (p-ERK1/2, p-JNK,-STAT3, cyclin D1, Bcl-2, p21WAF1 and p27kip1) in MCF-7 cells.oreover, the effects of annonacin on MCF-7 xenografts in nudeice were also investigated.

. Materials and methods

.1. Cell culture and reagents

ER�-positive MCF-7 cells and ER�-negative MDA-MB-231 cellsere purchased on February 18, 2009 from Bioresource Collec-

ion and Research Center (Hsinchu, Taiwan), where cells wereuthenticated by DNA fingerprints of short tandem repeat profiling.ells were cultured in DMEM/F-12 (1:1) medium supplementedith 1% penicillin/streptomycin and 10% fetal bovine serum (FBS,ibco, Grand Island, NY, USA) in a humidified 5% CO2 incubatort 37 ◦C. Cells were fasted for 24 h before adding fresh mediumontaining 10% FBS and various concentrations of annonacin or-hydroxytamoxifen.

Cyclin D1, cyclin-dependent kinase 4 (cdk4), cyclin E,21WAF1, p27kip1, Bcl-2, Glyceraldehyde 3-phosphate dehydroge-ase (GAPDH), ER�, STAT3 and JNK antibodies were purchased fromanta Cruz Biotechnology Inc. (Santa Cruz Co., CA). �-tubulin anti-ody was purchased from Lab Vision Corporation (Fremont, CA).erine 118 phosphorylated ER� (pSer118ER�), p-STAT3, ERK1/2,-ERK1/2, JNK and p-JNK antibodies were purchased from Cell Sig-aling Technology (Danvers, MA). We had isolated, purified andharacterized annonacin from the leaves of Formosan A. muricataLiaw et al., 2002) (supplementary Methods and supplementaryable 1), which has a different molecular structure from that ofamoxifen (supplementary Fig. 1). 4-hydroxytamoxifen, an active

etabolite of tamoxifen and other chemicals were purchased fromigma–Aldrich Chemical Company (St. Louis, MO). Annonacin wasissolved in dimethyl sulfoxide (DMSO, 0.1% final concentration).ecause annonacin has not been used in humans clinically, theffective concentration was chosen based on our previous in vitrotudy (Yuan et al., 2003). 17�-estradiol and 4-hydroxytamoxifenere dissolved in ethanol (0.1% final concentration).

.2. Measurement of cell survival and cell death

Cell survival was performed in quadruplicate similar to ourrevious study (Chou et al., 2008). Briefly, viable cell numbersere measured by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-

etrazolium bromide (MTT) assay. MTT was purchased fromigma–Aldrich Chemical Company (St. Louis, MO). Briefly, MCF-7
ells (6500 cells/well) or MDA-MB-231 cells (3500 cells/well) werelated and incubated in 96-well plates. MTT (1 mg/ml) was added
nto each well. Production of insoluble formazan by viable cells waseasured at 540 nm.

acology 137 (2011) 1283– 1290

Dose–response curves for various combinations of 4-hydroxytamoxifen and annonacin and the half-effective dose(ED50) for cell survival at 48 h were analyzed by the four-parameter log-logistic model (Sørensen et al., 2007). The doseranges used were: 0% 4-hydroxytamoxifen (i.e. annonacin, 0.01,0.05, 0.1, 0.5 and 1 �M), 50% 4-hydroxytamoxifen (0.1, 0.5, 1, 5and 10 �M), 91% 4-hydroxytamoxifen (0.1, 0.5, 1, 5 and 10 �M)and 100% 4-hydroxytamoxifen (0.1, 0.5, 1, 5 and 10 �M), respec-tively. The concentration addition model and the isobologrammethod (Sørensen et al., 2007) was used to assess the synergy,additivity or antagonism of the combinations of annonacin and4-hydroxytamoxifen.

Cell death was assessed by lactate dehydogenase releasethrough using the CytoTox 96 nonradioactive cytotoxicity assaykit according to the manufacturer’s instructions (Promega Corp.,Madison, WI).


Cell cycle analysis was performed similar to our previous study(Chuang et al., 2006). Briefly, cells were trypsinized and suspendedin cold phosphate-buffered saline (PBS). Suspended cells werewashed twice with cold PBS and fixed with 70% ice-cold ethanoland placed at −20 ◦C overnight. Cells were then centrifuged andresuspended with nuclear staining buffer (0.1% Triton X-100 in PBS,200 �g/ml RNase, and 100 �g/ml propidium iodide) for 30 min at37 ◦C. For each sample, at least 1 × 104 events were recorded. Cellcycle profiles were obtained with a FACScan flow cytometer (BectonDickinson Co., San Jose, CA) and data were analyzed with WinCyclesoftware (Phoenix Flow Systems Inc., San Diego, CA).

2.4. Immunoblotting

Immunoblotting was performed similar to our previous study(Guh et al., 2003). Briefly, a 30 �g sample of cell lysates waselectrophoresed on 10% sodium dodecyl sulfate-polyacrylamidegels, transferred to polyvinylidene difluoride membranes. Afterblocking, blots were incubated with antibody in blocking solutionovernight (phospho-antibodies) or for 2 h (other antibodies) fol-lowed by 5 min wash twice in PBS containing 0.1% Tween 20 andthen incubated with horseradish peroxidase-conjugated secondaryantibodies (Santa Cruz Biotechnologies Inc., Santa Cruz, CA) for 1 h.Enhanced chemiluminescence reagents were employed to depictprotein bands on the membrane.

2.5. Measurement of apoptosis

Apoptosis was measured by flow cytometry by using theVybrantTM Apoptosis Assay Kit #2 (Alexa Fluor® 488 annexinV/Propidium Iodide kit #2) supplied by Molecular Probes Inc.(Eugene, OR) according to the manufacturer’s instructions. This kitdetects the externalization of phosphatidylserine in apoptotic cellsusing the green-fluorescent Alexa Fluor® 488 annexin and the red-fluorescent propidium iodide nucleic acid stain. Propidium iodidestains necrotic cells with red fluorescence. After treatment withboth probes, apoptotic cells show green fluorescence, dead cellsshow red and green fluorescence, and live cells show little or nofluorescence.

2.6. Transient transfection and luciferase assay

Transient transfection was performed similar to our previousstudy (Chou et al., 2008). Briefly, MCF-7 cells were plated into 6-well plates at density of 1.2 × 105 cells/well in DMEM/F12 mediumand grown overnight. Cells were transfected with 0.2 �g of the


Fig. 1. Time-dependent effects of annonacin on cell cycle distribution and cell death in MCF-7 cells. MCF-7 cells (2.5 × 105 cells/dish) were plated and incubated in 6 cmdishes. Cells were fasted for 24 h before adding fresh medium (10% FBS) containing 0.1% DMSO or 0.1 �M annonacin. Cell cycle was measured by flow cytometry. Apoptosiswas measured by flow cytometry by using the VybrantTM Apoptosis Assay Kit #2 (Alexa Fluor® 488 annexin V/Propidium Iodide kit #2). (A) Time-dependent (1–48 h) effectsof annonacin (0.1 �M) on cell cycle distribution. (B) Time-dependent (6–48 h) effects of annonacin (0.1 �M) on apoptosis. (C) Dose-dependent effects of annonacin on celld Cells w0 ath wm

ET(at

2

pT2ititotTamTh

eath. MCF-7 cells (6500 cells/well) were plated and incubated in 96-well plates.

.1% DMSO (open bars) or various concentrations of annonacin (gray bars). Cell deean ± SEM of three independent experiments. *P < 0.05 versus DMSO.

R� transcriptional activity reporter plasmid containing 3X ERE-ATA luc (Addgene Inc., Cambridge, MA) with LipofectAMINE 2000Life Technology, Gaithersburg, MD) and luciferase activities weressayed by integrating the total light emission over 10 s by usinghe Dynatech ML1000 luminometer.

.7. In vivo tumor xenograft study

Female nude mice (4 weeks old; BALB/cA-nu [nu/nu]) wereurchased from the National Laboratory Animal Center (Taipei,aiwan) and were housed in specific pathogen-free conditions for

weeks. Afterwards, MCF-7 cells (5 × 106 cells in 200 �L PBS) werenjected subcutaneously into the flanks, and tumors were allowedo develop for 30 days. Then eleven mice were randomly dividednto two groups. Annonacin-treated mice (N = 6) were intraperi-oneally injected daily with annonacin (50 mg/kg/day) in 200 �Lf 25% polyethylene glycol. Control mice (N = 5) were intraperi-oneally injected daily with 200 �L of 25% polyethylene glycol.umor volume was measured using calipers at 0, 3, 7, 11, 15, 19
nd 22 days. Tumor volume was estimated by the following for-ula: tumor volume (mm3) = L × W × W/2 (L: length; W: width).
he xenografts of the annonacin-treated and the control mice werearvested and fixed in 4% formaldehyde, embedded in paraffin

ere fasted for 24 h before adding fresh medium (10% FBS, open bars) containingas measured by lactate dehydrogenase release. The results were expressed as the

for immunohistochemistry at 22 days. All animal procedures wereapproved and done in accordance with the national guidelines andthe guidelines by the Kaohsiung Medical University Animal Exper-iment Committee.

2.8. Immunohistochemistry

Paraffin-embedded tumor tissues were cut to 4 �m sections forimmunohistochemistry. The sections were treated with microwaveat 100 ◦C for 30 min, and blocked nonspecific response. The sectionswere incubated at 4 ◦C overnight with primary antibodies (ER�,cyclin D1 and Bcl-2). After washing twice 10 min with PBS contain-ing 0.2% Tween 20, the sections were incubated with biotinylatedsecondary antibodies for 1 h. After washing twice with PBS con-tain 0.2% Tween 20 for 10 min, sections were stained by UniversalDAB + kit/HRP (Dako Corp., Carpinteria, CA) and counterstainedwith hematoxylin.


The results were expressed as the mean ± standard errors ofthe mean. Unpaired Student’s t-tests were used for the compar-ison between two groups. P < 0.05 was considered as statistically

1286 Y.-M. Ko et al. / Journal of Ethnopharm

Fig. 2. Effects of annonacin on cell cycle regulatory proteins in MCF-7 cells. MCF-7 cells (2.5 × 105 cells/dish) were plated and incubated in 6 cm dishes. Cells werefasted for 24 h before adding fresh medium (10% FBS) containing 0.1% DMSO or0.1 �M annonacin. Cell cycle regulatory proteins (cyclin D1, cdk4, cyclin E, p21WAF1

and p27kip1) were measured by immunoblotting. (A) Time-dependent effects ofaefi

sm((

3

3i

ip(d(d26

3p

iad((d(

3a

bdt

To confirm the growth-inhibitory effects of annonacin in vivo,

nnonacin (0.1 �M) on p21WAF1 and p27kip1 protein expression. (B) Time-dependentffects of annonacin (0.1 �M) on cyclin D1, cdk4 and cyclin E protein expression. Thisgure is representative of three independent experiments.

ignificant. Concentration addition model and the isobologramethod (Sørensen et al., 2007) were assessed by the drc

dose–response curves) package in the R statistical programKnezevic et al., 2009).

. Results

.1. Effects of annonacin on cell cycle distribution and cell deathn MCF-7 cells

Annonacin (0.1 �M) time-dependently (6–48 h) arrested cellsn the G0/G1 phase of the cell cycle (Fig. 1A) while increasing apo-tosis at 48 h (Fig. 1B). Additionally, annonacin dose-dependently0.5–1 �M) increased cell death at 48 h (Fig. 1C). Annonacin alsoose-dependently (0.05–1 �M) decreased cell survival at 48 hsupplementary Fig. 2A). Moreover, annonacin (0.1 �M) time-ependently (24–48 h) decreased cell survival (supplementary Fig.B). However, annonacin (0.1 �M) did not affect cell survival at–48 h in MDA-MB-231 cells (supplementary Fig. 2C).

.2. Time-dependent effects of annonacin on cell cycle regulatoryroteins in MCF-7 cells

Annonacin (0.1 �M) time-dependently (30 min to 12 h)ncreased p21WAF1 but decreased p21WAF1 protein expressiont 24–48 h (Fig. 2A). In contrast, annonacin (0.1 �M) time-ependently increased p27kip1 protein expression at 1–6 hFig. 2A). Additionally, annonacin (0.1 �M) time-dependently1–48 h) decreased cyclin D1 and time-dependently (12–48 h)ecreased cdk4 protein expression, but not that of cyclin EFig. 2B).

.3. Effects of annonacin and 4-hydroxytamoxifen on cell survivalnd cyclin D1 or Bcl-2 protein expression in MCF-7 cells

Annonacin (0.1 �M), 4-hydroxytamoxifen (1 �M) and a com-ination of annonacin (0.1 �M) plus 4-hydroxytamoxifen (1 �M)ecreased cell survival to a similar degree at 24 h (Fig. 3A). However,he combination of annonacin (0.1 �M) plus 4-hydroxytamoxifen

acology 137 (2011) 1283– 1290

(1 �M) decreased cell survival more than annonacin (0.1 �M) didalone at 48 h (Fig. 3A).

The estimated half-effective doses from dose–responsecurves at 48 h were: 0% 4-hydroxytamoxifen (i.e. annonacin,ED50 = 0.31 �M), 50% 4-hydroxytamoxifen (ED50 = 0.45 �M), 91%4-hydroxytamoxifen (ED50 = 1.01 �M), 100% 4-hydroxytamoxifen(ED50 = 1.13 �M), respectively. Isobologram analysis showedthat the ED50 of the various combinations (50% and 91% 4-hydroxytamoxifen) coincided with the estimated concentrationaddition isobole (Fig. 3B). In other words, the combination effectsof annonacin and 4-hydroxytamoxifen were additive instead ofsynergistic (ED50 of the various combinations shifted to left) orantagonistic (ED50 of the various combinations shifted to right).

As shown in supplementary Fig. 3A, annonacin (0.01–2 �M)evenly decreased cyclin D1 and Bcl-2 protein expression at48 h. Additionally, annonacin (0.1 �M) time-dependently (1–48 h)decreased Bcl-2 protein expression (supplementary Fig. 3B).Annonacin (0.1 �M) decreased cyclin D1 protein expressionmore than 4-hydroxytamoxifen (1 �M) did at 48 h (Fig. 3C). Incontrast, annonacin (0.1 �M) and 4-hydroxytamoxifen (1 �M)decreased Bcl-2 protein expression to a similar degree at 48 h(Fig. 3C). Moreover, the combination of annonacin (0.1 �M) plus4-hydroxytamoxifen (1 �M) decreased Bcl-2 protein expressionmore than annonacin (0.1 �M) did alone at 48 h (Fig. 3C).

3.4. Effects of annonacin and 4-hydroxytamoxifen on ER˛transcriptional activity, ER˛ protein expression andphosphorylation in MCF-7 cells

As shown in Fig. 4A, annonacin (0.1 �M) and 4-hydroxytamoxifen (1 �M) decreased ER� transcriptional activityto the same degree at 24–48 h. Moreover, the combination ofannonacin (0.1 �M) plus 4-hydroxytamoxifen (1 �M) decreasedER� transcriptional activity more than annonacin (0.1 �M) didalone at 24–48 h (Fig. 4A).

As shown in supplementary Fig. 4A, 17�-estradiol increased ER�transcriptional activity whereas annonacin (0.1 �M) attenuatedER� transcriptional activity at 12–48 h. Additionally, annonacin(0.1 �M) decreased ER� protein expression at 24 h and decreasedER� protein serine 118 phosphorylation at 12 h (supplementaryFig. 4B). Annonacin dose-dependently (0.01–2 �M) decreasedER� protein expression at 48 h (Fig. 4B). In contrast, 4-hydroxytamoxifen (1 �M) did not affect ER� protein expression at48 h (Fig. 4C). Finally, the combination of annonacin (0.1 �M) plus4-hydroxytamoxifen (1 �M) decreased ER� protein expression toa similar degree as annonacin (0.1 �M) did alone at 48 h (Fig. 4C).

3.5. Time-dependent effects of annonacin on phosphorylation ofERK1/2, JNK and STAT3 protein in MCF-7 cells

Annonacin (0.1 �M) time-dependently (1–48 h) decreasedERK1/2 protein phosphorylation (Fig. 5A). Annonacin (0.1 �M) alsotime-dependently (1–48 h) decreased JNK protein phosphoryla-tion (Fig. 5B). Additionally, annonacin (0.1 �M) time-dependently(30 min to 48 h) decreased STAT3 protein phosphorylation (Fig. 5C).

3.6. Effects of annonacin on MCF-7 xenograft tumor size andexpression of ER˛, cyclin D1 and Bcl-2 proteins in nude mice

MCF-7 cells were grafted into the flanks of nude mice as xenografts.As shown in Fig. 6A, annonacin decreased tumor size at 7–22 days.As shown in Fig. 6B, annonacin attenuated the expression of ER�,cyclin D1 and Bcl-2 protein in the nude mice at 22 days.


Fig. 3. Combination effects of annonacin and 4-hydroxytamoxifen on cell survival and cyclin D1 and Bcl-2 protein expression in MCF-7 cells. MCF-7 cells (6500 cells/well)were plated and incubated in 96-well plates. Cells were fasted for 24 h before adding fresh medium (10% FBS) containing 0.1% DMSO or 0.1% ethanol. Cell survival wasmeasured by the MTT assay. (A) Effects of annonacin (A, 0.1 �M), 4-hydroxytamoxifen (T, 1 �M) or their combination (A + T) on cell survival at 24 h (lanes 4–6) or 48 h (lanes7–9). *P < 0.05 versus control. #P < 0.05 versus annonacin alone. (B) Isobologram analysis of various combinations (50% 4-hydroxytamoxifen and 91% 4-hydroxytamoxifen)of annonacin and 4-hydroxytamoxifen at 48 h. Dose combinations (dotted lines) achieving ED50 (dots) with standard errors together with the estimated isoboles are shown.The estimated concentration addition isobole is shown as the solid line connecting the 100% and 0% 4-hydroxytamoxifen ED50 dots. MCF-7 cells (2.5 × 105 cells/dish) werep measu( and

e

4

i2dtSat

a(mSgaau(

sap2

lated and incubated in 6 cm dishes. Expression of cyclin D1 and Bcl-2 protein was

A, 0.1 �M), 4-hydroxytamoxifen (T, 1 �M) or their combination (A + T) on cyclin D1xperiments.

. Discussion and conclusions

This study adds mechanistic insights to our previous find-ng that annonacin inhibits growth in MCF-7 cells (Yuan et al.,003).We found that annonacin induced cell death only at highoses (≥0.5 �M). In contrast, low-dose (0.1 �M) annonacin, likeamoxifen and fulvestrant (Cariou et al., 2000; Musgrove andutherland, 2009), induced cell-cycle-dependent (G0/G1) growthrrest concomitantly with the induction of p21WAF1 and p27kip andhe inhibition of cyclin D1 protein expression.

Cell cycle (G0/G1 transition) inhibitors p21WAF1 and p27kip

re implicated in the pathogenesis or prognosis of breast cancerCaldon et al., 2006). For example, the loss of p21WAF1 or p27kip

ay mediate tamoxifen resistance in breast cancer (Musgrove andutherland, 2009). Conversely, transfection of p21WAF1 or p27kip

enes inhibits MCF-7 cell growth (Jiang et al., 2009). Surprisingly,nnonacin inhibited both p21WAF1 and cyclin D1 protein expressiont 24–48 h. However, inhibition of p21WAF1 alone cannot atten-ate cyclin D1 inhibition-induced growth arrest in cancer cellsMasamha and Benbrook, 2009).

Our finding that annonacin inhibited cyclin D1 protein expres-
ion is compatible with the notion that cyclin D1 overexpression isssociated with poor prognosis and tamoxifen resistance in ER�-ositive breast cancer (Butt et al., 2008; Musgrove and Sutherland,009). Interestingly, cyclin D1 overexpression induces whereas
red by immunoblotting and normalized to that of GAPDH. (C) Effects of annonacinBcl-2 protein expression at 48 h. This figure is representative of three independent

cyclin D1 knockout inhibits breast cancer in mice (Butt et al., 2008).Similarly, cyclin D1 overexpression increases whereas cyclin D1inhibition decreases proliferation in MCF-7 cells (Grillo et al., 2006).

Annonacin, like tamoxifen and fulvestrant (Lam et al., 2009;Musgrove and Sutherland, 2009), also induced apoptosis whileinhibiting Bcl-2 protein expression. Note that cell survival, celldeath and cell cycle pathways are interconnected in cancers(Maddika et al., 2007). For example, apoptosis is a cell-cyclecheckpoint during cell injury and Bcl-2 is an anti-apoptotic pro-tein which also inhibits the G1/S checkpoint of the cell cycle(Maddika et al., 2007; Musgrove and Sutherland, 2009). Transfec-tion of the Bcl-2 gene inhibits whereas inhibition of the Bcl-2 geneenhances chemosensitivity of breast cancer cells (Emi et al., 2005).Moreover, Bcl-2 protein expression is associated with tamoxifenresistance in breast cancer patients (Musgrove and Sutherland,2009).

The observation that annonacin decreased cell survival inER�-positive MCF-7 cells, but not ER�-negative MDA-MB-231cells, suggests that ER� is required for the effects of annonacin.Interestingly, annonacin (ED50 = 0.31 �M) was more potent than4-hydroxytamoxfien (ED50 = 1.13 �M), while annonacin and 4-
hydroxytamoxifen were additive, in inhibiting cell survival.Similarly, annonacin was more potent than 4-hydroxytamoxifen,and annonacin and 4-hydroxytamoxifen were additive, in inhibit-ing ER� transcriptional activity.

1288 Y.-M. Ko et al. / Journal of Ethnopharmacology 137 (2011) 1283– 1290

Fig. 4. Effects of annonacin and 4-hydroxytamoxifen on ER� transcriptional activity and ER� protein expression in MCF-7 cells. MCF-7 cells (1.2 × 105 cells/well) were platedand incubated in 6-well plates. Cells were fasted for 24 h before adding fresh medium (M, 10% FBS) containing 0.1% DMSO (D) or 0.1% ethanol. (A) Effects of annonacin(0.1 �M), 4-hydroxytamoxifen (1 �M) or their combination on ER� transcriptional activity at 24 h or 48 h. Cells were transfected with 3X ERE-TATA luc with LipofectAMINE2000 and luciferase activities were assayed by integrating the total light emission over 10 s by using the Dynatech ML1000 luminometer. *P < 0.05 versus control. #P < 0.05versus annonacin alone. MCF-7 cells (2.5 × 105 cells/dish) were plated and incubated in 6 cm dishes. Cells were fasted for 24 h before adding fresh medium containing 0.1%DMSO, 0.1% ethanol, annonacin or 4-hydroxytamoxifen. Expression of ER� protein was measured by immunoblotting. (B) Dose-dependent effects of annonacin (A, 0.01-2 �M) on ER� protein expression at 48 h. Expression of ER� was normalized to that of �-tubulin. (C) Effects of annonacin (A, 0.1 �M), 4-hydroxytamoxifen (T, 1 �M) or theircombination (A + T) on ER� protein expression at 48 h. Expression of ER� was normalized to that of GAPDH. Both figures B and C are representative of three independentexperiments.

Fig. 5. Time-dependent effects of annonacin on phosphorylation of ERK1/2, JNK and STAT3 protein in MCF-7 cells. MCF-7 cells (2.5 × 105 cells/dish) were plated and incubatedin 6 cm dishes. Cells were fasted for 24 h before adding fresh medium (M, 10% FBS) containing 0.1% DMSO (D) or annonacin for 30 min to 48 h. Phosphorylation of ERK1/2,JNK and STAT3 protein was measured by immunoblotting and normalized to that of ERK1/2, JNK and STAT3. (A) Time-dependent effects of annnonacin (0.1 �M) on p-ERK1/2protein phosphorylation. (B) Time-dependent effects of annnonacin (0.1 �M) on p-JNK protein phosphorylation. (C) Time-dependent effects of annnonacin (0.1 �M) onp-STAT3 protein phosphorylation. This figure is representative of three independent experiments.


Fig. 6. Effects of annonacin on MCF-7 xenograft tumor size and expression of ER�, cyclin D1 and Bcl-2 protein in the nude mice. MCF-7 cells (5 × 106 cells in 200 �L PBS) weregrafted into the flanks of the nude mice as xenografts, and tumors were allowed to develop for 30 days. Afterwards, annonacin (50 mg/kg/day) or vehicle (25% polyethyleneg suredd for thm .

aFbHptaiidsttoi

MSto2ttai

iEI(

lycol) was administered intraperitoneally for 21 days. (A) Tumor volume was meaays). (B) Tumor was excised at 22 days and immunohistochemistry was performedean ± SEM of 5 mice in each group. *P < 0.05, **P < 0.01, ***P < 0.001 versus vehicle

The mode of action of annonacin can be studied by ER� bindingssay and ER� transcriptional activation assay (Dang et al., 2011).or example, SERM and SERD usually act by inhibiting ER� ligand-inding. We did not study the ER�-binding ability of annonacin.owever, we found that annonacin, like SERD, also decreases ER�rotein expression. Thus, annonacin, like SERD, may be useful inamoxifen resistance (Osborne and Schiff, 2011). We also found thatnnonacin decreased ERE-dependent ER� transcriptional activ-ty. Note that ERE is indispensable for ER�-induced phenotypesn breast cancer cells (Nott et al., 2009). Moreover, annonacinecreased ER� serine 118 phosphorylation, which is required forome ER�-induced effects (Duplessis et al., 2011). In contrast, bothamoxifen and fulvestrant increase ER� serine 118 phosphoryla-ion (Lipfert et al., 2006; Maggi, 2011). However, there are manyther possible modes of action whereby annonacin can inhibit ER�-nduced effects (Shapiro et al., 2011).

Annonacin, unlike tamoxifen (Ishii et al., 2008; Lam et al., 2009;usgrove and Sutherland, 2009), inactivated ERK1/2, JNK and

TAT3 in this study. Interestingly, inhibition of ERK1/2 attenuatesamoxifen resistance (Ghayad et al., 2010) and inhibition of JNKr STAT3 induces apoptosis in breast cancer cells (Kunigal et al.,009; Mingo-Sion et al., 2004). In summary, annonacin differs fromamoxifen in terms of ER� protein abundance and phosphoryla-ion, p-ERK1/2, p-JNK and p-STAT3. These observations may partlyccount for the finding that annonacin and tamoxifen were additiven inhibiting MCF-7 cell growth.

The in vitro effects of annonacin were corroborated by our find-
ngs that annonacin attenuated tumor size and the expression ofR�, cyclin D1 and Bcl-2 protein in MCF-7 cell-grafted nude mice.n this regard, cyclin D1 is overexpressed in 50% of breast cancersCaldon et al., 2006). Breast epithelial cell-specific overexpression
every 3–4 days using calipers at 0, 3, 7, 11, 15, 19 and 22 days (inset: tumor at 22e expression of ER�, cyclin D1 and Bcl-2 protein. The results were expressed as the

of cyclin D1 induces breast cancer in mice, while cyclin D1-null miceare resistant to oncogene-induced breast cancer (Butt et al., 2008).Moreover, inhibition of Bcl-2 in combination with chemotherapywas effective in some breast cancer patients in a clinical trial (Fatoet al., 2008).

In conclusion, annonacin induced cell-cycle-dependent growtharrest and induced apoptosis in ER�-related pathways (ERK1/2,JNK, STAT3, cyclin D1, Bcl-2, p21WAF1 and p27kip1) in MCF-7 cells.Annonacin and 4-hydroxytamoxifen were additive in inhibitinggrowth and ER� transcriptional activity. Moreover, annonacinattenuated MCF-7 xenograft tumor growth while inhibiting ER�,cyclin D1 and Bcl-2 protein expression in nude mice.

Acknowledgement

This work was supported by the National Science Council ofTaiwan (NSC-94-2321-B-037-006) to Lea-Yea Chuang.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.jep.2011.07.056.

References

Butt, A., et al., 2008. Cell cycle machinery: links with genesis and treatment of breast
cancer. Advances in Experimental Medicine and Biology 630, 189–205.
Caldon, E., et al., 2006. Cell cycle control in breast cancer cells. Journal of CellularBiochemistry 97, 261–274.

Cariou, S., et al., 2000. Down-regulation of p21WAF1/CIP1 or p27Kip1 abro-gates antiestrogen-mediated cell cycle arrest in human breast cancer cells.

http://dx.doi.org/10.1016/j.jep.2011.07.056

1 pharm

C

C

C

C

D

D

E

F

F

G

G

G

I

J

K


Proceedings of the National Academy of Sciences of the United States of America97, 9042–9046.

assileth, B., 2008. Complementary therapies, herbs, and other OTC agents. Oncology(Williston Park) 22, 1202.

hou, W.-W., et al., 2008. Arecoline-induced growth arrest and p21WAF1 expressionare dependent on p53 in rat hepatocytes. Toxicology 243, 1–10.

huang, L.-Y., et al., 2006. Role of nitric oxide in high glucose-induced mitogenicresponse in renal fibroblasts. Molecular Endocrinology 20, 2548–2558.

oothankandaswamy, V., et al., 2010. The alternative medicine pawpaw and its ace-togenin constituents suppress tumor angiogenesis via the HIF-1/VEGF pathway.Journal of Natural Products 73, 956–961.

ang, Z., et al., 2011. Comparison of chemical-induced transcriptional activation offish and human estrogen receptors: regulatory implications. Toxicology Letters201, 152–175.

uplessis, T.T., et al., 2011. Phosphorylation of estrogen receptor � at serine 118directs recruitment of promoter complexes and gene-specific transcription.Endocrinology 152, 2517–2526.

mi, M., et al., 2005. Targeted therapy against Bcl-2-related proteins in breast cancercells. Breast Cancer Research 7, R940–R952.

ato, R., et al., 2008. Generation of reactive oxygen species by mitochondrialcomplex I: implications in neurodegeneration. Neurochemical Research 33,2487–2501.

ox, E., et al., 2009. Novel actions of estrogen to promote proliferation: integrationof cytoplasmic and nuclear pathways. Steroids 74, 622–627.

hayad, S., et al., 2010. Endocrine resistance associated with activated ErbB sys-tem in breast cancer cells is reversed by inhibiting MAPK or PI3K/Akt signalingpathways. International Journal of Cancer 126, 545–562.

rillo, M., et al., 2006. Validation of cyclin D1/CDK4 as an anticancer drug targetin MCF-7 breast cancer cells: effect of regulated overexpression of cyclin D1and siRNA-mediated inhibition of endogenous cyclin D1 and CDK4 expression.Breast Cancer Research and Treatment 95, 185–194.

uh, J.-Y., et al., 2003. Beta-hydroxybutyrate–induced growth inhibition and col-lagen production in HK-2 cells are dependent on TGF-beta and Smad3. KidneyInternational 64, 2041–2051.

shii, Y., et al., 2008. Tamoxifen stimulates the growth of cyclin D1-overexpressingbreast cancer cells by promoting the activation of signal transducer and activatorof transcription 3. Cancer Research 68, 852–860.

iang, D.-D., et al., 2009. [Effects of transfected pIRES-p21waf1-p27kip1 gene on theduplication of centrosomes and the proliferation of breast cancer cell line MCF-7]. Zhonghua Yi Xue Za Zhi 89, 3295–3298.

nezevic, S.Z., et al., 2009. Utilizing R software package for dose–response studies:the concept and data analysis. Weed Technology 21, 840–848.

acology 137 (2011) 1283– 1290

Kunigal, S., et al., 2009. Stat3-siRNA induces Fas-mediated apoptosis in vitro andin vivo in breast cancer. International Journal of Oncology 34, 1209–1220.

Lam, L., et al., 2009. Tamoxifen and ICI 182,780 increase Bcl-2 levels and inhibitgrowth of breast carcinoma cells by modulating PI3K/AKT, ERK and IGF-1Rpathways independent of ERalpha. Breast Cancer Research and Treatment 118,605–621.

Liaw, C.-C., et al., 2002. New cytotoxic monotetrahydrofuran annonaceous aceto-genins from Annona muricata. Journal of Natural Products 65, 470–475.

Liaw, C.C., et al., 2010. Historic perspectives on Annonaceous acetogenins from thechemical bench to preclinical trials. Planta Medica 76, 1390–1404.

Lipfert, L., et al., 2006. Antagonist-induced, activation function-2-independent estro-gen receptor alpha phosphorylation. Molecular Endocrinology 20, 516–533.

Maddika, S., et al., 2007. Cell survival, cell death and cell cycle pathways are inter-connected: implications for cancer therapy. Drug Resistance Updates 10, 13–29.

Maggi, A., 2011. Liganded and unliganded activation of estrogen receptor and hor-mone replacement therapies. Biochimica et Biophysica Acta 1812, 1054–1060.

Masamha, C.P., Benbrook, D.M., 2009. Cyclin D1 degradation is sufficient to induceG1 cell cycle arrest despite constitutive expression of cyclin E2 in ovarian cancercells. Cancer Research 69, 6565–6572.

McLaughlin, J.L., 2008. Paw paw and cancer: annonaceous acetogenins from discov-ery to commercial products. Journal of Natural Products 71, 1311–1321.

Mingo-Sion, A., et al., 2004. Inhibition of JNK reduces G2/M transit independentof p53, leading to endoreduplication, decreased proliferation, and apoptosis inbreast cancer cells. Oncogene 23, 596–604.

Musgrove, E., Sutherland, R., 2009. Biological determinants of endocrine resistancein breast cancer. Nature Reviews. Cancer 9, 631–643.

Nott, S.L., et al., 2009. Genomic responses from the estrogen-responsive element-dependent signaling pathway mediated by estrogen receptor alpha are requiredto elicit cellular alterations. Journal of Biological Chemistry 284, 15277–15288.

Oberlies, N.H., et al., 1997. Structure−activity relationships of diverse annonaceousacetogenins against multidrug resistant human mammary adenocarcinoma(MCF-7/Adr) cells. Journal of Medicinal Chemistry 40, 2102–2106.

Osborne, C.K., Schiff, R., 2011. Mechanisms of endocrine resistance in breast cancer.Annual Review of Medicine 62, 233–247.

Sørensen, H., et al., 2007. An isobole-based statistical model and test for syner-gism/antagonism in binary mixture toxicity experiments. Environmental andEcological Statistics 14, 383–397.

Shapiro, D.J, et al., 2011. Small molecule inhibitors as probes for estrogen and andro-gen receptor action. Journal of Biological Chemistry 286, 4043–4048.

Yuan, S.-S., et al., 2003. Annonacin, a mono-tetrahydrofuran acetogenin, arrests can-cer cells at the G1 phase and causes cytotoxicity in a Bax- and caspase-3-relatedpathway. Life Sciences 72, 2853–2861.

Cancer Letters 323 (2012) 29–40

Contents lists available at SciVerse ScienceDirect

Cancer Letters

journal homepage: www.elsevier .com/locate /canlet

Graviola: A novel promising natural-derived drug that inhibits tumorigenicityand metastasis of pancreatic cancer cells in vitro and in vivo through alteringcell metabolism

María P. Torres a,b, Satyanarayana Rachagani a, Vinee Purohit b, Poomy Pandey c, Suhasini Joshi a,Erik D. Moore a, Sonny L. Johansson b,d, Pankaj K. Singh a,b, Apar K. Ganti e, Surinder K. Batra a,b,d,⇑a Department of Biochemistry and Molecular Biology, Omaha, NE 68198-5870, USAb Eppley Institute for Research in Cancer and Allied Diseases, Omaha, NE 68198-5870, USAc Department of Environmental, Agricultural & Occupational Health, Omaha, NE 68198-5870, USAd Department of Pathology and Microbiology, Omaha, NE 68198-5870, USAe Department of Internal Medicine VA Nebraska-Western Iowa Health Care System and University of Nebraska Medical Center, Omaha, NE 68198-5870, USA

a r t i c l e i n f o

Article history:Received 19 January 2012Received in revised form 25 February 2012Accepted 26 March 2012

Keywords:Pancreatic cancerTherapyCancer metabolismNatural product

0304-3835/$ - see front matter � 2012 Elsevier Irelanhttp://dx.doi.org/10.1016/j.canlet.2012.03.031

⇑ Corresponding author at: Department of InterWestern Iowa Health Care System and UniversityOmaha, NE 68198-5870, USA. Tel.: +1 402 559 5455;

E-mail address: [email protected] (S.K. Batra).

a b s t r a c t

Pancreatic tumors are resistant to conventional chemotherapies. The present study was aimed at evalu-ating the potential of a novel plant-derived product as a therapeutic agent for pancreatic cancer (PC). Theeffects of an extract from the tropical tree Annona Muricata, commonly known as Graviola, was evaluatedfor cytotoxicity, cell metabolism, cancer-associated protein/gene expression, tumorigenicity, and meta-static properties of PC cells. Our experiments revealed that Graviola induced necrosis of PC cells by inhib-iting cellular metabolism. The expression of molecules related to hypoxia and glycolysis in PC cells (i.e.HIF-1a, NF-jB, GLUT1, GLUT4, HKII, and LDHA) were downregulated in the presence of the extract. Invitro functional assays further confirmed the inhibition of tumorigenic properties of PC cells. Overall,the compounds that are naturally present in a Graviola extract inhibited multiple signaling pathways thatregulate metabolism, cell cycle, survival, and metastatic properties in PC cells. Collectively, alterations inthese parameters led to a decrease in tumorigenicity and metastasis of orthotopically implanted pancre-atic tumors, indicating promising characteristics of the natural product against this lethal disease.

� 2012 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

The overall 5-year survival rate for pancreatic cancer (PC) pa-tients was 5.5% for the period of 2001–2007, according to the Na-tional Cancer Institute (NCI), a statistic that has not variedsignificantly for over the last four decades [1]. In 2012, it is esti-mated that 43,920 new PC cases will be diagnosed and approxi-mately 85% of these (i.e. 37,390) will succumb to the disease [2].The main reason behind the poor prognosis of PC patients is theinsidious and sporadic nature of the disease, which is often pre-sented with no specific early clinical symptoms. By the time ofdiagnosis, PC is already in advanced stages (i.e. III and IV) and isresistant to conventional chemotherapy and radiotherapy [3].Interestingly, even patients diagnosed with stage I PC that havethe option to undergo surgery have a 5-year overall survival ofapproximately 20%, a clear indication of the general failure of cur-

d Ltd. All rights reserved.

nal Medicine VA Nebraska-of Nebraska Medical Center,fax: +1 402 559 6650.

rent standard treatments for each stage of PC [4,5]. What is evenmore alarming, are the statistics that predict a 55% increase inthe expected number of new PC cases by 2030 [6]. Thus, immediateprogress must be made in the prevention, early diagnosis, and sys-temic treatments against this lethal disease.

Gemcitabine has been the standard line of treatment for PCpatients for over a decade and is associated with a median patientsurvival of 5.4 months [7]. Over all these years, numerous clinicalefforts have been devoted to improve PC chemotherapy outcomes,but unfortunately no significant improvements have been re-ported apart from a clinical trial reported in May of 2011 [8]. Thisphase III clinical trial reported an improved overall survival ofPC patients treated with a four-drug chemotherapy regimen com-prising fluorouracil, leucovorin, irinotecan, and oxaliplatin (FOLF-IRINOX). Nevertheless, a major disadvantage of this noveltreatment was its related toxicity, which was noticeably highwhen compared to PC patients treated with gemcitabine alone.Therefore, novel, alternative PC therapeutics must not only im-prove the prognosis of PC patients but also minimize any possibletoxicity-related side effects that will interfere with the quality oflife of PC patients.

http://dx.doi.org/10.1016/j.canlet.2012.03.031




http://www.elsevier.com/locate/canlet

30 M.P. Torres et al. / Cancer Letters 323 (2012) 29–40

It is well known that an increased consumption of fruits andvegetables is associated with a reduced risk of most cancers,including PC [9]. For this reason, the potential of natural productsin PC therapies has been widely investigated [10]. While some ofthese compounds have undergone clinical testing (i.e. curcumin,genistein) and have demonstrated some activity against PC, thepoor bioavailability in patients minimizes their therapeutic effi-cacy. However, as compared with conventional chemotherapeuticdrugs, the major benefit of these therapies is the apparent lack oftoxicities to healthy tissues. These facts attracted our attention tofind alternative, natural-derived chemotherapeutic drugs in orderto improve the prognosis of PC patients.

Traditionally, the leaves from the tropical tree Annona Muricata,also known as Graviola or Soursop, have been used for a widerange of human diseases including inflammatory conditions, rheu-matism, neuralgia, diabetes, hypertension, insomnia, cystitis, para-sitic infections, and cancer [11]. The major bioactive componentsthat have been extracted from different parts of the plant areknown as Annonaceous acetogenins. These are derivatives of longchain (C35 or C37) fatty acids derived from the polyketide pathway[12] that are selectively toxic to cancer cells, including multidrug-resistant cancer cell lines [13–17]. Annonaceous acetogenins in-duce cytotoxicity by inhibiting the mitochondrial complex I, whichis involved in ATP synthesis [14]. As cancer cells have a higher de-mand for ATP than the normal cells, mitochondrial complex Iinhibitors have potential in cancer therapeutics.

A few in vivo studies involving A. Muricata have been reported.Among these, two reports have shown the ability of the leaf extractto regenerate pancreatic islet b cells in diabetic rats [18,19]. Thesestudies suggest an additional benefit of the natural product againstPC given that diabetes has been classified as a risk factor of themalignant disease [20]. More recently, one study analyzing theanti-tumor efficacy of A. Muricata was published [21]. The extracthad a direct anti-tumorigenic effect on breast cancer cells bydownregulating the expression of the epidermal growth factorreceptor (EGFR). Although this study demonstrates the potentialanti-tumorigenic properties of Graviola, the doses used in theexperimental design were not properly controlled. The mice werefed with the extract mixed in the diet and the exact amount in-gested by each animal could not be estimated accurately.

Although a few in vitro reports have shown the cytotoxic charac-teristics of Graviola against various cancer cell lines, including PCcells [12], the comprehensive in vivo effects and mechanistic scien-tific studies are still lacking. To our knowledge, the studies reportedherein are the first to indicate that Graviola extract has promisingcharacteristics for PC therapeutics. Comprehensive in vitro andin vivo studies in various PC cell lines revealed that this natural prod-uct has inhibited multiple signaling pathways that regulate metab-olism, cell cycle, survival, and metastatic properties of PC cells.

2. Materials and methods

2.1. Graviola extract

Graviola supplement capsules were purchased from Raintree (Carson City, NV).The capsules consisted of 100% pure, finely milled Graviola leaf/stem powder withno binders or fillers. The capsule contents were suspended in DMSO (100 mg/mL).After incubating for 5 min, the suspension was centrifuged and the supernatant (i.e.extract) was filtered to remove any remaining particles. Subsequent dilutions wereprepared in Dulbecco’s modification of Eagle’s medium (DMEM) supplementedwith 10% of fetal bovine serum (FBS). Stock solutions and respective dilutions werefreshly prepared prior to treatment.

2.2. Cell culture

The metastatic PC cell lines FG/COLO357 and CD18/HPAF were purchased fromthe American Type Culture Collection (ATCC). Before performing experiments, thePC cell lines were authenticated by short tandem repeat analysis. It was ensuredthat PC cells were used at fewer than 20 passages after purchase from ATCC. Cells

were cultured in DMEM medium supplemented with 10% FBS and antibiotics(100 lg/mL penicillin and 100 lg/mL streptomycin). The cells were maintained at37 �C and 5% CO2 in a humidified atmosphere.

2.3. Antibodies

The antibodies for phospho-ERK1/2, total ERK, phospho-Akt (Ser 473), total Akt,NF-jB, and Caspase-3 were purchased from Cell Signaling Technology (Danvers,MA). The antibodies for Cyclin-D1, phospho-FAK (Tyr 925), and total FAK were ob-tained from Santa Cruz Biotechnology (Santa Cruz, CA). The b-actin and b-Tubulinantibodies were obtained from Sigma Aldrich (St. Louis, MO), whereas the HIF-1aantibody was purchased from BD Biosciences (San Jose, CA). The MUC4 monoclonalantibody (8G7) used in these studies was developed by our group [22]. MMP9 anti-body was obtained from a hybridoma cell supernatant kindly provided byDr. Rakesh Singh at UNMC. The secondary antibodies used for western blot analyseswere the ECL™ anti-mouse and anti-rabbit IgG conjugated to horseradish peroxi-dase (GE healthcare, UK). Fluorescein isothiocyanate (FITC) conjugated-anti-mouseand Alexa Fluor conjugated anti-mouse antibodies were obtained from Invitrogen(Carlsbad, CA).

2.4. Cytotoxicity assay

To determine the cytotoxicity of Graviola extract on PC cells, 1 � 104 cells wereseeded per well on a 96-well plate in DMEM supplemented with 10% FBS and anti-biotics. After overnight incubation, different concentrations (10–200 lg/mL) of theextract were added into triplicate wells. After 48 h, the media was replaced withfresh media containing thiazolyl blue tetrazolium bromide (MTT) reagent (SigmaAldrich, St. Louis, MO). After 4 h incubation at 37 �C in 5% CO2 in humidified atmo-sphere, the media was replaced with 100 lL of DMSO and the corresponding cyto-toxicity values were calculated (k = 540 nm). The experiment was repeated at leastthree times.

2.5. Western blot analysis

For protein analysis, 0.5 � 106 of PC cells were seeded on each well of a six-wellplate in DMEM supplemented with 10% FBS and antibiotics. After overnight incuba-tion, fresh solutions of Graviola (0–200 lg/mL) were prepared and added to therespective wells. Cells incubated with the corresponding amount of DMSO presentin the highest concentrated solution of Graviola were used as a negative control(0 lg/mL). After 48 h of incubation with the extract, protein lysates were isolatedand prepared for western blot analysis, as previously described [23].

2.6. Real-time PCR

The transcripts levels of the glucose transporters GLUT1 and GLUT4, the glyco-lytic enzymes hexokinase II (HKII) and lactate dehydrogenase A (LDHA), and themucin glycoprotein MUC4 in PC cells were determined after treatment with Gravi-ola extract by real-time PCR. 0.5 � 106 cells were seeded in each well of a six-wellplate in complete media. After overnight incubation, fresh solutions of Graviolaextract (50 and 100 lg/mL) were prepared and cells were incubated for 48 h. Sub-sequently, cDNA was synthesized from purified RNA and real-time PCR was carriedout as has been described by previous studies [23]. The sequences of the gene-spe-cific primers used were:

GLUT1: F 50-GCCATGGAGCCCAGCAGCAA-30; R 50-CGGGGACTCTCGGGGCAGAA-30

GLUT4: F 50-GCCTGTGGCCACTGCTCCTG-30; R 50-GGGGTCTCTGGGCCGGGTAG-30

HKII: F 50-GTCATCCCCTTGTGTCAGAG-30; R 50-CTTCATTAGTGTCCCCATCCTG-30

LDHA: F 50-CCAGTGTGCCTGTATGGAGTG-30; R 50-GCACTCTCAACCACCTGCTTG-30

MUC4: F 50-GTGACCATGGAGGCCAGTG-30; R 50-TCATGCTCAGGTGTCCACAG-30

2.7. Glucose uptake

Glucose-uptake rate was assayed by utilizing [3H] 2-deoxyglucose ([3H] 2-DG).5 � 104 PC cells were seeded per well in a 24-well plate. 12 h later, the cells weretreated with Graviola extract (10 and 50 lg/mL) for 48 h. The cells were thenstarved for glucose for 2 h and incubated for 20 min with 2 lCi [3H] 2-DG. Subse-quently, cells were lysed with 1% SDS and the lysates were counted for [3H] by uti-lizing a scintillation counter. Cells treated with labeled and excess unlabeled 2-DGwere used as controls to set a baseline for non-specific [3H] uptake. The results werenormalized to the cell counts for treated and untreated groups. Glucose uptake wasnormalized with that of the control cells (0 lg/mL) and it is presented as the meanvalues ± standard error from experiments performed in triplicate.

2.8. ATP quantification

The CellTiter-Glo� Luminescent Cell Viability Assay (Promega, Madison, WI)was used to measure the ATP content in the cells. Briefly, 1 � 104 PC cells wereseeded in each well of an opaque 96-well plate. Cells were seeded for both ATP

M.P. Torres et al. / Cancer Letters 323 (2012) 29–40 31

quantification and protein concentration estimation. Starting the next day, the cellswere incubated with Graviola extract-containing media for 48 h. Subsequently, theinstructions of the manufacturer for ATP quantification were followed and lumines-cence was measured on a Synergy™Mx Luminescent Plate Reader (BioTek,Winooski, VT). Data is presented as the mean value for samples in triplicates, nor-malized with the protein content for each treatment, as determined by utilizing mi-cro-BCA protein estimation kit.

2.9. Detection and quantification of apoptosis and necrosis

To quantify the number of PC cells undergoing apoptosis and necrosis afterbeing incubated with Graviola extract, the Annexin-V-FLUOS staining kit (RocheDiagnostics, Indianapolis, IN) was used. PC cells were seeded and treated with Gra-viola extract as described above. After 48 h of treatment with Graviola extract, theinstructions of the manufacturer were followed for staining cells for flow cytometryanalysis. The experiment was repeated three times.


PC cells were synchronized at the G1/S phase using a double thymidine block.After seeding cells in 100 cm2 Petri dishes, thymidine (2 mM) was added for 12 h.After washing cells with serum-free media, the cells were released from thymidineblock by culturing in fresh medium containing 24 mM 2-deoxycytidine for 9 h.Then, cells were washed and incubated once more with thymidine (2 mM) for14 h. Subsequently, the cells were released from the second thymidine block andthe respective treatment prepared in complete media was added for 48 h. For cellcycle analysis, cells were trypsinized and washed with PBS after the duration ofthe treatment. Cells were then fixed in 70% ethanol at 4 �C for 1 h. After washing,cells were incubated with Telford reagent (EDTA, RNAse A, propidium iodide, TritonX-100 in PBS) at 4 �C and analyzed by flow cytometry on the next day.

2.11. Confocal microscopy

For confocal analysis, 2 � 105 PC cells were seeded on sterilized round glasscover slips. After overnight incubation, Graviola extract (0, 50 and 100 lg/mL)was added to the cells, followed by a 48 h incubation. For the detection of reactiveoxygen species (ROS), Graviola extract-treated PC cells were incubated with 1 lM20-70-Dichlorofluorescein diacetate (DCFH-DA) (Sigma Aldrich, St. Louis, MO) for15 min. After three washes with PBS, glass cover slips were mounted on glass slidesand visualized by confocal microscopy. For b-tubulin and MUC4 confocal analysis,details of the procedure are published elsewhere [23]. Finally, to visualize thearrangement of actin filaments in Graviola extract-treated cells, the cells werestained with fluorescent phallotoxins (Invitrogen, Carlsbad, CA). The instructionsof the manufacturer were followed for formaldehyde-fixed cells. Post-staining,the glass cover slips were mounted with Vectashield medium (Vector Laboratories,Burlingame, CA). LSM 510 microscope, a laser scanning confocal microscope (CarlZeiss GmbH, Thornwood, NY) was utilized to image the cells in the respective chan-nels at a magnification of 630�.

2.12. Wound healing assay

For wound healing assays, 3 � 106 of PC cells were seeded in 60 mm petri dishesin DMEM media supplemented with 10% FBS and antibiotics. After overnight incu-bation, an artificial wound was induced on 100% confluent PC cell monolayers usinga sterile pipette tip. Graviola extract-containing (0, 50, 100 lg/mL) media solutionswere then added to the respective treatment plate. Images (40�) were capturedimmediately after adding Graviola extract (0 h) and after 24 h of treatment, by alight microscope. The motility of the cells across the wound was visualized in eachtreatment group.

2.13. Motility assay

The effect of Graviola extract on the migration of PC cells was also analyzed by atranswell migration assay. FG/COLO357 cells (0.5 � 106) were suspended in Gravi-ola extract-containing (0–100 lg/mL) 1% FBS-DMEM media and seeded for 48 h in8 lm pore size polyethylene terephthalate (PET) membranes (Becton Dickinson,San Jose, CA). DMEM supplemented with 10% FBS was added at the bottom of eachwell and after 48 h of incubation, the cells that migrated to the bottom of the PETmembrane were stained with Diff-Quick cell staining kit (Dade Behring Inc., New-ark, DE). The number of migrated cells was quantified by performing cell counts of10 random fields at 100� magnification. The results are presented as the averagenumber of cells in one field.

2.14. In vivo tumorigenicity studies

The effect of Graviola extract on pancreatic tumor growth was evaluated onorthotopic tumor xenografts. 6–8 week old female athymic immunodeficient micewere purchased from the Animal Production Area of the NCI/Frederick Cancer Re-

search and Development Center (Frederick, MD). The mice were treated in accor-dance with the Institutional Animal Care and Use Committee (IACUC) guidelinesat UNMC and were housed in pathogen-free environment and were fed sterilewater and food ad libitum.

Over 90% viable luciferase-labeled CD18/HPAF cells transduced with retroviralparticles (Addgene, Cambridge, MA) were orthotopically injected into the head ofthe pancreas of immunodeficient mice. Details of the orthotopic implantation pro-cedure are described elsewhere [22,24]. After 1 week of tumor growth, oral gavagetreatment of PBS-suspended Graviola extract was given daily for 35 days. The dosesof Graviola extract for these studies were based on previous in vivo studies[18,19,25] and on the recommended dose for human consumption [11]. Treatmentgroups (N = 8) included: PBS only (0 mg/kg), 50 mg/kg, and 100 mg/kg Graviola ex-tract. Graviola extract was not dissolved in DMSO for these studies in order to dem-onstrate the benefit of the aqueous natural oral supplement in PC therapy.Nevertheless, the cytotoxic properties of the Graviola extract suspended in PBSwere corroborated beforehand (Supplementary data Fig. 1). In vivo IVIS 200 biopho-tonic imaging system (Caliper Life Sciences, Hopkinton, MA) was used to captureimages of pancreatic tumors within every 2 weeks during the course of treatmentwith Graviola extract. . Mice were sacrificed after 42 days of tumor growth and35 days of treatment with Graviola extract. Changes in tumor growth and sites ofmetastasis were evaluated in each treatment group. Body weights of mice weremeasured before the treatment.

2.15. Analysis of pancreatic tumor tissues

On the necropsy day, pancreatic tumors from the different treatment groupswere divided for protein and immunohistochemistry (IHC) analyses. The tumorswere immediately frozen under liquid nitrogen for protein analysis. To prepare tu-mor lysates, the tumors were then suspended on radioimmunoprecipitation (RIPA)buffer and sonicated for three cycles with a Branson digital sonifier� (60% ampli-tude, 10 s). After centrifuging the homogeneous suspension, the protein concentra-tion in each sample was estimated and respective solutions for western blotanalyses were prepared as previously described [23].

For histopathological and IHC analyses, the tumor tissues were fixed in 10% For-malin for 48 h. The tumors were embedded in paraffin and 5 lm sections were cutand stained with hematoxylin and eosin stains (H&E) and various antibodies (i.e.MMP9 and MUC4). Details of the procedure for IHC staining is described elsewhere[24]. The IHC and H&E stained slides were evaluated by pathologist at University ofNebraska Medical Center.


The JMP� Statistical Discovery Software (Cary, NC) was used to determine thestatistical significance within the treatment replicates in each experiment. A Stu-dent’s t-test was used to calculate the corresponding p-value. All p-values <0.05were considered statistically significant.

3. Results

3.1. Graviola extract induces cytotoxicity of pancreatic cancer cells

The PC cells FG/COLO357 and CD18/HPAF were incubated for48 h with different concentrations of Graviola extract. The resultsfrom the MTT cytotoxicity assay indicated a progressive decreasein cell viability with the successive increase in the concentrationsof the extract (Fig. 1A). After 48 h of treatment, the resultingIC50 of Graviola extract on FG/COLO357 and CD18/HPAF cellswas 200 and 73 lg/mL, respectively and the results indicated thatCD18/HPAF cell line is more sensitive to the Graviola extract thanthe FG/COLO357 cell line (Fig. 1B).

It is well known that the activation of the extracellular signal-regulated kinase (ERK) and the phosphatidylinositol 30kinase(PI3K/Akt) pathways play a crucial role in the proliferation and sur-vival of PC [26] and inhibition of these pathways leads to the inhi-bition of pancreatic tumor growth [27,28]. The present studyrevealed that treatment of PC cells with Graviola extract resultedin decreased activation of both ERK and Akt pathways in PC cells(Fig. 1C). Thus, the inhibition of these pathways is in agreementwith the decreased viability of PC cells treated with Graviolaextract.

Fig. 1. Effect of Graviola extract on pancreatic cancer cell viability. (A) MTT cytotoxicity assay for Graviola extract-treated PC cells. Cells were incubated with differentconcentrations of Graviola extract and corresponding DMSO controls for 48 h. Data represents the mean value of experiments performed in triplicate ± standard error ofmean. (⁄p-value < 0.0001, ⁄⁄0.0001 < p-value < 0.001, ⁄⁄⁄p-value = 0.007, compared to untreated cells); (B) Inverted microscope images (40�) of PC cells after treatment withGraviola extract for 48 h; (C) Western blot analysis of proteins involved in PC cell proliferation after 48 h treatment with Graviola extract. Protein lysates (30 lg) wereresolved by 10% SDS–PAGE. b-actin was used as the loading control. Each experiment was performed three times.


3.2. Pancreatic cancer cell metabolism is inhibited by Graviola extract

Previous studies have shown that major bioactive componentspresent in Graviola extract inhibit mitochondrial complex I[13–17], suggesting their direct involvement in cell metabolism.It has already been well-documented that cancer cells undergo ametabolic shift to adapt and survive under harsh environmentsby enhancing aerobic glycolysis [29,30]. Also, Akt activation leadsto glycolytic ATP generation in tumor cells [31]. Hence, the effectof Graviola extract on several stages of the glycolytic pathway inPC cells was analyzed.

The expression of HIF-1a, a critical regulator of aerobic glycol-ysis in cancer cells [32], was analyzed in PC cells after incubationwith Graviola extract (Fig. 2A). We observed reduced HIF-1aexpression in both PC cell lines, suggesting a direct effect of thisnatural product on the metabolism of PC cells. Likewise, it has beenpreviously reported that NF-jB upregulates the expression ofHIF-1a [33,34]. Not surprisingly, the expression levels of NF-jBwere also reduced in PC cells after being incubated with Graviolaextract (Fig. 2A).

Subsequently, the expression of the glucose transporters 1 and4 (GLUT1 and GLUT4), and the expression of the glycolytic enzymeshexokinase II (HKII) and lactate dehydrogenase A (LDHA), all ofwhich are upregulated by HIF1-a in cancer cells [32,35], were ana-lyzed in Graviola extract-treated PC cells by real-time PCR analysis(Fig 2B). Overall, the transcript levels of GLUT1, HKII, and LDHAwere significantly reduced in both PC cell lines when comparedto untreated cells (i.e. 60–87% downregulation).

Cancer cells have an increased expression of glucose transport-ers to enhance glucose uptake, which in turn increases the glyco-lytic rate for an enhanced ATP production that will ultimatelylead to an enhanced tumor growth [10]. Thus, based on the resultsdiscussed above, it was not surprising that PC cells treated withGraviola extract doses over 10 lg/mL had a decreased rate ofglucose uptake when compared to untreated cells (0 lg/mL)(Fig. 2C). Finally, to evaluate the energy outcome of the glycolyticpathway of PC cells, we measured ATP production in Graviolaextract-treated PC cells (Fig. 2D), and observed significant inhibi-tion by 42–47% and by 31–43% doses in FG/COLO357 and CD18/HPAF PC cells, respectively. Altogether, these results indicate that

Fig. 2. Effect of Graviola extract on the metabolism of pancreatic cancer cells. (A) Western blot analysis of HIF-1a and NF-jB expression in PC cells after treatment withGraviola extract. Protein lysates (30 lg) were resolved on 10% SDS–PAGE gels. b-actin was used as a loading control; (B) Real-time PCR-based measurement of transcriptlevels of glucose transporters 1 and 4 (GLUT1, GLUT4), hexokinase II (HKII), and lactate dehydrogenase A (LDHA) in PC cells after incubation with Graviola extract. Data ispresented as the average fold difference in gene expression for the gene of interest in Graviola extract-treated cells versus untreated cells (0 lg/mL) ± standard error of mean.The housekeeping gene b-actin was used as an internal control. (⁄0.01 < p-value < 0.05, ⁄⁄0.005 < p-value < 0.001, ⁄⁄⁄p-value < 0.005); (C) Measurement of glucose uptake in PCcells after treatment with Graviola extract. Radioactive counts of cells labeled with [3H]-2-deoxyglucose were normalized with controls (⁄⁄⁄p-value < 0.0001); (D) ATPquantification of PC cells after treatment with Graviola extract. A Luminescent Cell Viability assay was used to measure the ATP content in the cells. Data is presented as meanvalue from experiments performed in triplicates normalized with the protein content ± standard error of mean. (⁄p-value = 0.003, ⁄⁄p-value = 0.002, ⁄⁄⁄p-value = 0.0002) Datain the left panel is from FG/COLO357 cells, whereas data in the right panel is from CD18/HPAF cells.


Graviola extract impairs the metabolism of PC cells that will ulti-mately lead to decreased cell viability.

3.3. Graviola extract induces necrosis of pancreatic cancer cells

In order to evaluate the cytotoxic pathways induced by Graviolaextract. PC cells were stained with annexin-V and propidium iodide(PI) to measure the necrotic and apoptotic cell populations byperforming flow cytometry. While the necrotic cell population inboth PC cell lines increased significantly after incubation with

Graviola extract, the apoptotic cell population remained unchanged(Fig. 3A). Subsequently, the production of Graviola extract-inducedROS in FG/COLO357 and CD18/HPAF PC cells was confirmed by con-focal microscopy (Fig. 3B). Additionally, it was also observed thatcells incubated with Graviola extract have a gain in cell volume, acharacteristic of necrotic cell death.

In order to confirm that Graviola extract was not inducingapoptosis of PC cells, the levels of Caspase-3 expression were ana-lyzed by western blot analysis. The Caspase-3 expression valuesremained statistically unaltered by treatment with the extract,

Fig. 3. Analysis of cytotoxic mechanism of Graviola extract in pancreatic cancer cells. (A) Quantification of apoptotic and necrotic PC cells after treatment with Graviolaextract (Apoptotic cells Annexin V+/PI- staining; Necrotic cells Annexin V+/PI + staining). Data is presented as the mean value of the corresponding % cell population induplicate samples ± standard error of mean; (B) The production of reactive oxygen species (ROS) in PC cells after treatment with Graviola extract was determined afterincubating Graviola extract-treated PC cells with 20 ,70-Dichlorofluorescein diacetate (DCFH-DA). Cells were then analyzed by confocal microscopy. Scale bar represents20 lm; (C) Western blot analysis of Caspase-3 expression in PC cells after treatment with Graviola extract. Protein lysates (30 lg) were resolved on 10% SDS–PAGE gels.b-actin was used as a loading control; (D) Cell cycle analysis of FG/COLO357 PC cells after treatment with Graviola extract. Cells were synchronized in the G1/S phase bythymidine block before adding Graviola extract. The effect of Graviola extract on the distribution of cells in different phases of the cell cycle was analyzed by flow cytometry.The data is presented as the mean value of the corresponding % cell population in duplicate samples ± standard error of mean. Representative flow cytometry histograms ofcells treated with different concentrations of Graviola extract are shown. (⁄p-value = 0.0001; ⁄⁄p-value < 0.0001); (E) Western blot analysis of the expression of the cell cycle-related protein CyclinD1 in PC cells after being incubated with Graviola extract. Protein lysates (30 lg) were resolved in 10% SDS–PAGE gels. b-actin was used as the loadingcontrol. In (A), (B), and (C), data in the left panel is from FG/COLO357 cells, whereas data in the right panel is from CD18/HPAF cells.


suggesting that apoptotic pathways are not involved (Fig. 3C). Fur-thermore, the apoptotic cells population in Graviola extract-trea-ted cells was also analyzed by Telford staining, and the resultscorroborated the findings from AnnexinV/PI staining studies,where the number of apoptotic cell population did not vary afterbeing incubated with the natural compound (data not shown).

An analysis of the different phases of the cell cycle after treat-ment with Graviola extract demonstrated cell cycle arrest at G1phase (Fig. 3D). While the G1 cell population increased from 43%to 65%, the S phase decreased from 56% to 32% with increasing con-centrations of Graviola extract (0, 5, 100 lg/mL). To support theseresults, the expression of CyclinD1 in Graviola extract-treated PCcells was analyzed (Fig. 3E). In agreement with previous studiesindicating that a decreased CyclinD1 expression induces G0/G1 cellcycle arrest [36], Graviola extract-treated PC cells had also reducedexpression of the cell cycle regulatory protein.

3.4. Motility of pancreatic cancer cells decreases after treatment withGraviola extract

The effect of Graviola extract on the functional properties of PCcells was analyzed in vitro wound healing and migration assays(Fig. 4A and B). As it can be observed in the images from the woundhealing assays, PC cells treated with Graviola extract did not closethe wound even after 24 h, as opposed to untreated cells (0 lg/mL),indicating reduced motility of PC cells after treatment with Gravi-ola extract (Fig. 4A). Similarly, the migratory capacity of PC cellswas also reduced after treatment with Graviola extract, as evalu-ated by a transwell assay (Fig. 4B), suggesting that the naturalextract reduces the motility of PC cells.

The motility and migration of cancer cells is associated with therearrangements of the cortical actin and microtubules network[37,38]. Additionally, cellular ATP depletion has been associated

Fig. 4. Effect of Graviola extract in the motility, migration, and cytoskeleton of pancreatic cancer cells. (A) Wound healing assay of FG/COLO357 PC cells after treatment withGraviola extract. Microscope images (40�) of the artificially created wound in PC cells monolayer were taken before (0 h) and after adding Graviola extract (24 h); (B)Migration of FG/COLO357 PC cells after treatment with Graviola extract. The number of cells that migrated through the 8 lm pores of a polyethylene terephtalate (PET)membrane was quantified in 10 random fields. Data represent the mean value of migrating cells ± standard error of mean (⁄p-value = 0.0009; ⁄⁄p-value < 0.0001, compared tountreated control cells); (C) Actin filaments were analyzed by confocal microscopy by Rhodamine-anti-Phalloidin staining of FG/COLO357 cells after treatment with Graviolaextract. Nucleus was stained with DAPI. Scale bars represent 20 lm; (D) Microtubules were analyzed by confocal microscopy after FITC-anti-b Tubulin staining ofFG/COLO357 cells after treatment with Graviola extract. Nucleus was stained with DAPI. Scale bars represent 20 lm; (E) Expression of proteins related to migration/motilityof PC cells after treatment with Graviola extract. Protein lysates (30 lg) were resolved by 10% SDS–PAGE. b-actin was used as a loading control.


with reorganization of the actin cytoskeleton [39] and suppressionof the dynamics of microtubules is known to induce mitotic arrest[40]. Taking this into consideration, the cytoskeleton of Graviolaextract-treated PC cells was analyzed by confocal microscopy(Fig. 4C and D). The image results of phallotoxins (i.e. phalloidin)staining indicate a disruption of the cortical actin network and dis-solution of stress fibers in Graviola extract-treated PC cells(Fig. 4C). Similarly, a disruption of microtubules dynamics was evi-dent after b-tubulin staining of PC cells incubated with Graviola

extract (Fig. 4D). To further analyze the effect of Graviola extracton motility and migration of PC cells, the expression levels of thephosphorylated focal adhesion kinase (pFAK), which is involvedin mitogenic signaling and motility [41], and matrix metallopro-teinase 9 (MMP9), which targets many extracellular proteinsincluding adhesion molecules [42], were analyzed by western blotanalysis (Fig. 4E). In agreement with the experiments discussedabove, we observed that the expression levels of both pFAK andMMP9 were downregulated in Graviola extract-treated cells.

Fig. 5. Evaluation of Graviola Extract in pancreatic cancer orthotopic xenograft model. (A) Pancreatic tumor weight results after treatment with Graviola extract.CD18/HPAF-Luciferase cells were injected orthotopically in the pancreas of athymic nude mice. After 1 week of tumor growth, oral gavage treatment of PBS-suspendedGraviola extract was given daily for 35 days (N = 8). Data is presented as box plots of the mean tumor weight of mice in each treatment group. (⁄p-value = 0.006; ⁄⁄p-value = 0.0008, compared to tumors of PBS-treated mice); (B) Major sites of metastasis in each treatment group. Results are presented as number of animals havingmetastasis out of total number of animals per group. Statistical analysis was done comparing Graviola extract-treated mice with untreated mice (0 mg/kg Graviola extract);(C) In vivo biophotonic imaging of pancreatic tumors during the course of treatment with Graviola extract. Representative IVIS images of mice from different treatment groupsare shown (D) Hematoxylin and Eosin (H&E) staining of paraffin embedded pancreatic tumors. Images on the right (20�) are magnified areas from the images located at theleft (10�). Yellow arrows in H&E sections represent necrotic areas in tumors from mice treated with Graviola extract.


3.5. Graviola extract inhibits tumor growth and metastasis ofpancreatic cancer cells

Based on the results obtained from in vitro experiments, Gravi-ola extract has promising properties to be incorporated in PC ther-apeutics. Nevertheless, these anti-tumorigenic properties requirefurther validation through in vivo experiments. In order to evaluatethe therapeutic potential of Graviola extract, a more realistic

situation for administering the extract was mimicked. It is recom-mended that Graviola extract supplement must be taken on a reg-ular basis [11], and therefore, it was decided that the extract mustbe administered by oral gavage after suspending contents of thecapsule in aqueous solution instead of dissolving it in DMSO. Priorto evaluating the anti-tumorigenic properties of aqueous Graviolaextract suspension by in vivo experiments, pertinent in vitro exper-iments corroborating the cytotoxic potential of the aqueous

Fig. 6. Immunohistochemical analyses of pancreatic tumors after treatment with Graviola extract. (A) Immunohistochemical staining of MMP9 in paraffin-embeddedpancreatic tumors. Representative images (20�) of tumors from different treatment groups are shown with the average composite score shown at the right. Data fromexperiments performed in triplicates is presented as the mean value of the composite score of tumors ± standard error of mean. MMP9 expression in pancreatic tumors wasalso assessed by western blot analysis. Homogenized protein tumor lysates (30 lg) were resolved by 10% SDS–PAGE. b-actin was used as a loading control. (B)Immunohistochemistry staining of MUC4 in paraffin-embedded pancreatic tumors. Representative images (200�) of tumors from different treatment groups are shown withthe average composite score shown at the right. Data from experiments performed in triplicate is presented as the mean value of the composite score of tumors ± standarderror of mean. MUC4 expression in pancreatic tumors was also assessed by western blot analysis. Homogenized protein tumor lysates (30 lg) were resolved by 2% agarosegels. b-actin was used as the loading control.


suspensions on PC cells were completed beforehand (Supplemen-tary Fig. 1).

For tumorigenic studies, CD18/HPAF cells expressing luciferasewere orthotopically injected into the pancreas of athymic mice.After 1 week, in vivo biophotonic imaging confirmed tumor growthin all animals and the treatment regimen was initiated. The tumorgrowth during the treatment was monitored by imaging every2 weeks. After 35 days of treatment, the animals were euthanizedand the pancreatic tumors were removed and weighed. Althoughpancreatic tumors were not completely eradicated, the resultsindicate that tumor growth decreased significantly in Graviola ex-tract-treated mice in comparison to the control group (Fig. 5A).Specifically, the tumor growth inhibition in mice treated with adose of 50 mg/kg Graviola extract was 59.8% (p-value = 0.0008)whereas in mice treated with 100 mg/kg Graviola extract the inhi-bition was 50.3% (p-value = 0.006), indicating the efficacy of thenatural product in PC regression. The metastatic lesions in eachmouse were evaluated in various vital organs including the liver,

spleen, mesenteric lymph nodes (LN), small and large intestines,peritoneum, diaphragm, and ovaries (Fig. 5B). Although all themetastatic lesions were reduced in Graviola extract-treated micein comparison to the untreated control mice, the incidence ofmetastasis in the liver, mesenteric LN, and ovaries was significantlyreduced (p-values 6 0.02). Representative biophotonic tumorimages illustrate the tumor growth across the different groups dur-ing the course of the treatment (Fig. 5C).

Further, tumors were evaluated by H&E (Fig. 5D) and IHC stain-ing (Fig. 6). The H&E stained tumor sections showed necrotic cellsin 20–50% of the pancreatic tumor tissues from Graviola extract-treated mice as compared with tumors from the control mice.These results further strengthen the results from in vitro experi-ments, which demonstrate that Graviola extract-mediated reduc-tion in PC cell viability was through the induction of necrosis.

The tumor lysates and paraffin embedded pancreatic tumorswere also evaluated by IHC for the expression of MMP9 (Fig. 6A)and MUC4 (Fig. 6B). In agreement with in vitro data, the levels of


MMP9 were reduced in tumors from Graviola extract-treated micecompared to the untreated controls. As the expression of MMP9has been related to invasion and metastasis, the reduced levels ofthe protein in Graviola extract-treated tumors substantiate ourfindings of reduced metastatic sites in these mice.

Previous studies performed by our group have established thecorrelation of the expression of mucin4 (MUC4) glycoprotein withprogression and metastasis of PC [24,43–45]. Therefore, we wereparticularly interested in evaluating the effect of Graviola extracton the expression of MUC4 in PC cells and pancreatic tumors. Invitro experiments demonstrated a significant downregulation inMUC4 expression, both at the translational (SupplementaryFigs. 1C and 2A and B) and transcriptional levels (SupplementaryFig. 2C) in Graviola extract treated PC cells. Similarly, the expres-sion of the MUC4 was reduced in pancreatic tumors from micetreated with Graviola extract as compared to the untreated mice(Fig. 6B), further supporting our findings of reduced tumor growthand metastasis after treatment with Graviola extract.

4. Discussion

Little or no progress has been accomplished in PC treatmentover the last 40 years. Novel therapeutics against this lethal malig-nancy must inhibit several pathways that promote survival, pro-gression, and metastasis of PC cells. Based on the fact that cancercells are mainly dependent on the glycolytic pathway for ATPproduction, glucose deprivation by anti-glycolytic drugs can in-duce cancer cell death [46], a pathway that can be targeted andexplored in PC therapies [47].

Natural products have been investigated in PC therapeutics overseveral decades, but to date none has been incorporated in routinechemotherapies [10]. Traditionally, the leaves from Graviola ( A.Muricata) have been used for a wide range of human diseasesincluding cancer [11]. The present study is the first to demonstratethat Graviola extract reduces the viability of PC cells and tumors byinducing necrosis and cell cycle arrest, and by inhibiting PC cellmotility (i.e. cytoskeleton rearrangement), migration, and metabo-lism. Overall, in vitro experiments revealed that the compoundspresent in the natural extract inhibited several pathways involvedin PC cell proliferation and metabolism, simultaneously. Such inhi-bitions ultimately led to a decrease in tumor growth and metasta-sis in orthotopically transplanted pancreatic tumor-bearing mice.

In PC patients, an increased metabolic activity and glucoseconcentration of malignant tumors has been linked to pancreatictumor aggressiveness [47]. Additionally, the presence of hypoxiain PC has been associated with tumor growth and metastasis[48,49]. Indeed, the presence of hypoxic environment has beenlinked to the oncogenic and metabolic transformation (i.e. glycoly-sis) of PC cells that results in resistance to conventional cancertherapeutics [48,50]. More specifically, it has been suggested thathypoxia can induce resistance to gemcitabine through the activa-tion of PI3K/Akt/NF-jB and MAPK/ERK pathways [51], which arealso related to PC progression and survival. The activation of bothof these signaling pathways was evaluated in PC cells after treat-ment with Graviola extract and it was found that the extractsuppressed phosphorylation of the key molecules involved in thesepathways, which correlated with reduced viability of PC cells. Sub-sequently, the expression of HIF-1a, the major transcription factoractivated under hypoxic conditions, and its ensuing downstreameffects on PC cell metabolism were analyzed in Graviola extract-treated cells. The results indicated the natural product inhibitedPC cell metabolism by inhibiting the expression of HIF-1a,NF-jB, glucose transporters (i.e. GLUT1, GLUT4), and glycolyticenzymes (i.e. HKII, LDHA), all of which lead to the reduction of glu-cose uptake and ATP production by PC cells.

The overall downregulation of PC cell metabolism induced byGraviola extract resulted in PC cell death and necrosis. In agreementwith previous studies of ATP reduction, the metabolic and therapeu-tic stress induced by Graviola extract led to an acute ATP depletion,which is accompanied by increased intracellular ROS, ultimatelyleading to necrosis [52–54]. While necrotic agents have not beenconsidered beneficial in cancer therapies due to induction of localinflammation, the process itself can lead to the activation of the in-nate immune system capable of initiating anti-tumor immunity[52]. It makes it imperative to evaluate the effect of a necrosis-induc-ing product such as Graviola extract in an immune competent host.In this regard, we plan to evaluate the effect of the natural product onthe progression of pancreatic adenocarcinoma in the KrasG12DPdx1-Cre spontaneous animal model, where the effect on the immune sys-tem can be evaluated [55,56]. In order to evaluate the potential ofGraviola extract in preventing PC progression, we plan to supple-ment the diet of KrasG12DPdx1-Cre mice with Graviola extract afterthe mice start developing pancreatic intraepithelial neoplastic (Pa-nIN) lesions. The effective concentrations of Graviola metabolitesafter oral absorption and effects on the immune system will be mea-sured as well. Additional experiments will be carried out to evaluatethe potential of a combination therapy of Graviola extract with thestandard chemotherapeutic drug Gemcitabine. With the results dis-cussed in the present study, it is expected that minimum doses of thechemotherapeutic drug will be needed to eradicate the malignantdisease.

The major bioactive compounds identified in A. Muricata havebeen classified as Annonaceous acetogenins, which inhibit mito-chondrial complex I that leads to a decreased ATP production[13–17]. Although the natural extract capsules used in these stud-ies contained numerous compounds, the presence of Annonaceousacetogenins was evident by the depletion of ATP production in PCcells after being incubated with Graviola extract. Bioactivity-guided fractionation for the identification of potent bioactive (i.e.anti-tumorigenic) compounds that are present in the Graviola ex-tract is currently being investigated. We are also ensuring thatcytotoxic effects are specific to tumorigenic cells only, by includingthe non-transformed immortalized pancreatic epithelial cell lineHPNE, which is derived from pancreatic duct (data not shown).

Pancreatic tumors develop from a complex interplay of numer-ous signaling pathways and Graviola extract has shown promisinganti-tumorigenic characteristics by targeting some of these path-ways all at once. Although novel glycolytic inhibitors, such asGraviola extract, may have broad therapeutic applications [57],inhibition of glycolysis alone may not be sufficient to eradicate tu-mor cells completely. Perhaps the use of alternative medicine, liketaking Graviola capsules on a regular basis, should still be considereda supplement, not a replacement for standard therapies. Currently,in vitro studies evaluating the potential of the natural product incombination with chemotherapeutic drugs are being conducted.

Acknowledgements

The invaluable technical support from Kavita Mallya is greatlyappreciated. We would like to give special thanks to UNMC profes-sors: Dr. Michel Ouellette for kindly providing CD18/HPAF-Lucifer-ase and HPNE cells, Dr. Shilpa Buch for allowing us to use theLuminescence plate reader, Dr. Vimla Band for allowing us to usethe microscope to image tumor H&E and IHC sections, and Dr. SteveCaplan for assisting with the analysis of confocal images and provid-ing us the b-Tubulin antibody. We also thank Janice A. Tayor andJames R. Talaska of the Confocal Laser Scanning Microscope CoreFacility at UNMC, Victoria B. Smith and Megan Michalak of theUNMC Cell Analysis Core Facility, and Ms. Kristi Berger, the EppleyCancer Center for editing this manuscript. We are also very gratefulfor the expertise and involvement of Drs. Amarnath Natarajan and


Abijah Nyong from the Chemistry department at UNMC in the bio-activity-guided fractionation of the Graviola extract. The authorsof this work are supported by Grants from the National Institutesof Health: NIH-NCI Cancer Biology Training Grant UNMCT32CA009479 (MPT), R01 CA78590, U01EDRN CA111294, R01CA131944, R01 CA133774, R01 CA 138791, P50 SPORE CA127297and U54 CA160163. AKG is supported by VA Career DevelopmentAward.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.canlet.2012.03.031.

References

[1] SEER Stat Fact Sheets:Pancreas. National Cancer Institute. 10-28-2011.[2] R. Siegel, D. Naishadham, A. Jemal, Cancer statistics, CA Cancer J. Clin. 62

(2012) 10–29.[3] S. Chakraborty, M.J. Baine, A.R. Sasson, S.K. Batra, Current status of molecular

markers for early detection of sporadic pancreatic cancer, Biochim. Biophys.Acta 1815 (2011) 44–64.

[4] K.Y. Bilimoria, D.J. Bentrem, C.Y. Ko, A.K. Stewart, D.P. Winchester, M.S.Talamonti, National failure to operate on early stage pancreatic cancer, Ann.Surg. 246 (2007) 173–180.

[5] K.Y. Bilimoria, D.J. Bentrem, C.Y. Ko, J.S. Tomlinson, A.K. Stewart, D.P.Winchester, M.S. Talamonti, Multimodality therapy for pancreatic cancer inthe US: utilization, outcomes, and the effect of hospital volume, Cancer 110(2007) 1227–1234.

[6] B.D. Smith, G.L. Smith, A. Hurria, G.N. Hortobagyi, T.A. Buchholz, Future ofcancer incidence in the United States: burdens upon an aging, changing nation,J. Clin. Oncol. 27 (2009) 2758–2765.

[7] H.A. Burris III, M.J. Moore, J. Andersen, M.R. Green, M.L. Rothenberg, M.R.Modiano, M.C. Cripps, R.K. Portenoy, A.M. Storniolo, P. Tarassoff, R. Nelson, F.A.Dorr, C.D. Stephens, D.D. Von Hoff, Improvements in survival and clinicalbenefit with gemcitabine as first-line therapy for patients with advancedpancreas cancer: a randomized trial, J. Clin. Oncol. 15 (1997) 2403–2413.

[8] T. Conroy, F. Desseigne, M. Ychou, O. Bouche, R. Guimbaud, Y. Becouarn, A.Adenis, J.L. Raoul, S. Gourgou-Bourgade, C. de la Fouchardiere, J. Bennouna, J.B.Bachet, F. Khemissa-Akouz, D. Pere-Verge, C. Delbaldo, E. Assenat, B. Chauffert,P. Michel, C. Montoto-Grillot, M. Ducreux, FOLFIRINOX versus gemcitabine formetastatic pancreatic cancer, N. Engl. J. Med. 364 (2011) 1817–1825.

[9] R.J. Jansen, D.P. Robinson, R.Z. Stolzenberg-Solomon, W.R. Bamlet, M. deAndrade, A.L. Oberg, T.J. Hammer, K.G. Rabe, K.E. Anderson, J.E. Olson, R. Sinha,G.M. Petersen, Fruit and vegetable consumption is inversely associated withhaving pancreatic cancer, Cancer Causes Control 22 (2011) 1613–1625.

[10] S.D. Stan, S.V. Singh, R.E. Brand, Chemoprevention strategies for pancreaticcancer, Nat. Rev. Gastroenterol. Hepatol. 7 (2010) 347–356.

[11] L. Taylor. Technical Data Report for Graviola: Annona Muricata. Sage Press, Inc.Herbal Secrets of the Rainforest second ed., 2002.

[12] G.S. Kim, L. Zeng, F. Alali, L.L. Rogers, F.E. Wu, S. Sastrodihardjo, J.L. McLaughlin,Muricoreacin and murihexocin C, mono-tetrahydrofuran acetogenins, fromthe leaves of Annona muricata, Phytochemistry 49 (1998) 565–571.

[13] N.H. Oberlies, J.L. Jones, T.H. Corbett, S.S. Fotopoulos, J.L. McLaughlin, Tumorcell growth inhibition by several annonaceous acetogenins in an in vitro diskdiffusion assay, Cancer Lett. 96 (1995) 55–62.

[14] J.L. McLaughlin, Paw paw and cancer: annonaceous acetogenins fromdiscovery to commercial products, J. Nat. Prod. 71 (2008) 1311–1321.

[15] J.R. Tormo, I. Royo, T. Gallardo, M.C. Zafra-Polo, P. Hernandez, D. Cortes, F.Pelaez, In vitro antitumor structure–activity relationships of threo/trans/threomono-tetrahydrofuranic acetogenins: correlations with their inhibition ofmitochondrial complex I, Oncol. Res. 14 (2003) 147–154.

[16] F.R. Chang, Y.C. Wu, Novel cytotoxic annonaceous acetogenins from Annonamuricata, J. Nat. Prod. 64 (2001) 925–931.

[17] C.C. Liaw, F.R. Chang, C.Y. Lin, C.J. Chou, H.F. Chiu, M.J. Wu, Y.C. Wu, Newcytotoxic monotetrahydrofuran annonaceous acetogenins from Annonamuricata, J. Nat. Prod. 65 (2002) 470–475.

[18] S.O. Adewole, E.A. Caxton-Martins, Morphological changes and hypoglycemiceffects of Annona Muricata Linn. (Annonaceae) leaf aqueous extract onpancreatic B-cells of streptozotocin-treated diabetic rats, African J. Biomed.Res. 9 (2006) 173–180.

[19] D.O. Adeyemi, O.A. Komolafe, S.O. Adewole, E.M. Obuotor, A.A. Abiodum, T.K.Adenowo, Histomorphological and morphometric studies of the pancreaticislet cells of diabetic rats treated with extracts of Annona Muricata, FoliaMorphol. 69 (2010) 92–100.

[20] J.T. Magruder, D. Elahi, D.K. Andersen, Diabetes and pancreatic cancer: chickenor egg?, Pancreas 40 (2011) 339–351

[21] Y. Dai, S. Hogan, E.M. Schmelz, Y.H. Ju, C. Canning, K. Zhou, Selective growthinhibition of human breast cancer cells by graviola fruit extract in vitro and

in vivo involving downregulation of EGFR expression, Nutr. Cancer 63 (2011)795–801.

[22] N. Moniaux, G.C. Varshney, S.C. Chauhan, M.C. Copin, M. Jain, U.A. Wittel, M.Andrianifahanana, J.P. Aubert, S.K. Batra, Generation and characterization ofanti-MUC4 monoclonal antibodies reactive with normal and cancer cells inhumans, J. Histochem. Cytochem. 52 (2004) 253–261.

[23] M.P. Torres, M.P. Ponnusamy, S. Chakraborty, L.M. Smith, S. Das, H.A. Arafat,S.K. Batra, Effects of thymoquinone in the expression of mucin 4 in pancreaticcancer cells: implications for the development of novel cancer therapies, Mol.Cancer Ther. 9 (2010) 1419–1431.

[24] A.P. Singh, N. Moniaux, S.C. Chauhan, J.L. Meza, S.K. Batra, Inhibition of MUC4expression suppresses pancreatic tumor cell growth and metastasis, CancerRes. 64 (2004) 622–630.

[25] O.V. de Sousa, G.D. Vieira, R.G. de Jesus, C.H. Yamamoto, M.S. Alves,Antinociceptive and anti-inflammatory activities of the ethanol extract ofAnnona muricata L. leaves in animal models, Int. J. Mol. Sci. 11 (2010) 2067–2078.

[26] T. Seufferlein, Novel protein kinases in pancreatic cell growth and cancer, Int. J.Gastrointest. Cancer 31 (2002) 15–21.

[27] Q. Chang, M.S. Chapman, J.N. Miner, D.W. Hedley, Antitumour activity of apotent MEK inhibitor RDEA119/BAY 869766 combined with rapamycin inhuman orthotopic primary pancreatic cancer xenografts, BMC. Cancer 10(2010) 515.

[28] W.T. Wei, H. Chen, Z.L. Ni, H.B. Liu, H.F. Tong, L. Fan, A. Liu, M.X. Qiu, D.L. Liu,H.C. Guo, Z.H. Wang, S.Z. Lin, Antitumor and apoptosis-promoting properties ofemodin, an anthraquinone derivative from Rheum officinale Baill, againstpancreatic cancer in mice via inhibition of Akt activation, Int. J. Oncol. 39(2011) 1381–1390.

[29] R.J. DeBerardinis, J.J. Lum, G. Hatzivassiliou, C.B. Thompson, The biology ofcancer: metabolic reprogramming fuels cell growth and proliferation, CellMetab. 7 (2008) 11–20.

[30] G. Kroemer, J. Pouyssegur, Tumor cell metabolism: cancer’s Achilles’ heel,Cancer Cell 13 (2008) 472–482.

[31] R.A. Cairns, I.S. Harris, T.W. Mak, Regulation of cancer cell metabolism, Nat.Rev. Cancer 11 (2011) 85–95.

[32] H. Lu, R.A. Forbes, A. Verma, Hypoxia-inducible factor 1 activation by aerobicglycolysis implicates the Warburg effect in carcinogenesis, J. Biol. Chem. 277(2002) 23111–23115.

[33] S.F. Fitzpatrick, M.M. Tambuwala, U. Bruning, B. Schaible, C.C. Scholz, A. Byrne,A. O’Connor, W.M. Gallagher, C.R. Lenihan, J.F. Garvey, K. Howell, P.G. Fallon,E.P. Cummins, C.T. Taylor, An intact canonical NF-kappaB pathway is requiredfor inflammatory gene expression in response to hypoxia, J. Immunol. 186(2011) 1091–1096.

[34] S.Y. Nam, Y.S. Ko, J. Jung, J. Yoon, Y.H. Kim, Y.J. Choi, J.W. Park, M.S. Chang, W.H.Kim, B.L. Lee, A hypoxia-dependent upregulation of hypoxia-inducible factor-1by nuclear factor-kappaB promotes gastric tumour growth and angiogenesis,Br. J. Cancer 104 (2011) 166–174.

[35] Q. Ke, M. Costa, Hypoxia-inducible factor-1 (HIF-1), Mol. Pharmacol. 70 (2006)1469–1480.

[36] C.P. Masamha, D.M. Benbrook, Cyclin D1 degradation is sufficient to induce G1cell cycle arrest despite constitutive expression of cyclin E2 in ovarian cancercells, Cancer Res. 69 (2009) 6565–6572.

[37] C.C. Cunningham, Actin structural proteins in cell motility, Cancer Met. Rev. 11(1992) 69–77.

[38] I. Kaverina, A. Straube, Regulation of cell migration by dynamic microtubules,Semin. Cell Dev. Biol 22 (2011) 968–974.

[39] R. Bacallao, A. Garfinkel, S. Monke, G. Zampighi, L.J. Mandel, ATP depletion: anovel method to study junctional properties in epithelial tissues. I.Rearrangement of the actin cytoskeleton, J. Cell Sci. 107 (Pt 12) (1994)3301–3313.

[40] M.A. Jordan, S.B. Horwitz, S. Lobert, J.J. Correia, Exploring the mechanisms ofaction of the novel microtubule inhibitor vinflunine, Semin. Oncol. 35 (2008)S6–S12.

[41] X. Zhao, J.L. Guan, Focal adhesion kinase and its signaling pathways in cellmigration and angiogenesis, Adv. Drug Deliv. Rev. 63 (2011) 610–615.

[42] L.J. McCawley, L.M. Matrisian, Matrix metalloproteinases: they’re not just formatrix anymore!, Curr Opin. Cell Biol. 13 (2001) 534–540.

[43] S. Bafna, S. Kaur, N. Momi, S.K. Batra, Pancreatic cancer cells resistance togemcitabine: the role of MUC4 mucin, Br. J. Cancer 101 (2009) 1155–1161.

[44] P. Chaturvedi, A.P. Singh, N. Moniaux, S. Senapati, S. Chakraborty, J.L. Meza, S.K.Batra, MUC4 mucin potentiates pancreatic tumor cell proliferation, survival,and invasive properties and interferes with its interaction to extracellularmatrix proteins, Mol. Cancer. Res. 5 (2007) 309–320.

[45] M.J. Swartz, S.K. Batra, G.C. Varshney, M.A. Hollingsworth, C.J. Yeo, J.L.Cameron, R.E. Wilentz, R.H. Hruban, P. Argani, MUC4 expression increasesprogressively in pancreatic intraepithelial neoplasia, Am. J. Clin. Pathol. 117(2002) 791–796.

[46] M.N. El, A. Caro-Maldonado, S. Ramirez-Peinado, C. Munoz-Pinedo, Sugar-freeapproaches to cancer cell killing, Oncogene 30 (2011) 253–264.

[47] G. Komar, S. Kauhanen, K. Liukko, M. Seppanen, S. Kajander, J. Ovaska, P.Nuutila, H. Minn, Decreased blood flow with increased metabolic activity: anovel sign of pancreatic tumor aggressiveness, Clin. Cancer Res. 15 (2009)5511–5517.

[48] S. Vasseur, R. Tomasini, R. Tournaire, J.L. Iovanna, Hypoxia induced tumormetabolic switch contributes to pancreatic cancer aggressiveness, Cancers 2(2010) 2138–2152.




[49] J.P. Duffy, G. Eibl, H.A. Reber, O.J. Hines, Influence of hypoxia andneoangiogenesis on the growth of pancreatic cancer, Mol. Cancer 2 (2003)12.

[50] K. Yokoi, I.J. Fidler, Hypoxia increases resistance of human pancreatic cancercells to apoptosis induced by gemcitabine, Clin. Cancer Res. 10 (2004) 2299–2306.

[51] E.Y. Chen, N.M. Mazure, J.A. Cooper, A.J. Giaccia, Hypoxia activates a platelet-derived growth factor receptor/phosphatidylinositol 3-kinase/Akt pathwaythat results in glycogen synthase kinase-3 inactivation, Cancer Res. 61 (2001)2429–2433.

[52] R.K. Amaravadi, C.B. Thompson, The roles of therapy-induced autophagy andnecrosis in cancer treatment, Clin. Cancer Res. 13 (2007) 7271–7279.

[53] Y. Eguchi, S. Shimizu, Y. Tsujimoto, Intracellular ATP levels determine celldeath fate by apoptosis or necrosis, Cancer Res. 57 (1997) 1835–1840.

[54] M. Leist, B. Single, A.F. Castoldi, S. Kuhnle, P. Nicotera, Intracellular adenosinetriphosphate (ATP) concentration: a switch in the decision between apoptosisand necrosis, J. Exp. Med. 185 (1997) 1481–1486.

[55] N. Festjens, T. Vanden Berghe, P. Vandenabeele, Necrosis, a well-orchestratedform of cell demise: signalling cascades, important mediators and concomitantimmune response, Biochim. Biophys. Acta 1757 (2006) 1371–1387.

[56] P. Golstein, G. Kroemer, Cell death by necrosis: towards a molecular definition,Trends Biochem. Sci. 32 (2007) 37–43.

[57] H. Pelicano, D.S. Martin, R.H. Xu, P. Huang, Glycolysis inhibition for anticancertreatment, Oncogene 25 (2006) 4633–4646.

International Journal of Applied Science and Technology Vol. 2 No. 1; January 2012

157

THE BREAST OF ANTICANCER FROM LEAF EXTRACT OF ANNONA MURICATA AGAINTS

CELL LINE IN T47D

Eka Prasasti Nur Rachmani

Tuti Sri Suhesti

Department of Pharmacy

Jenderal Soedirman University

Indonesia

Retno Widiastuti

Aditiyono

Department of Medicine

Jenderal Soedirman University

Indonesia

Abstract

The breast cancer is the second leading cause of death in women after cervical cancer. The soursop (Annona

muricata) is a traditional medicinal plant which is empirically by the people of Indonesia are used for anti-inflammatory and anti-tumor. This study aims to determine the cytotoxic effects from extracts of leaves of soursop

and fraction results in cancer cells T47D. The research was carried out by extraction using ethanol and

fractionation by column chromatography method that used various solvents were n-hexane, chloroform, ethyl

acetate and methanol. Cytotoxic test performed by the method of MTT assay and apoptosis tests performed by the method of Double Stainning. The parameters obtained from the cytotoxic test was IC50 values, ie values that

produce inhibitory concentrations of cancer cells by 50%. Apoptosis assay results are analyzed in a qualitative

description. The results showed that the ethanol extracts of leaves of the soursop has a cytotoxic activity with IC50 values of 17.149 µg / mL. The results of the four fractions obtained by fractionation and the fraction F3 were the

fraction that has the best cytotoxic activity with IC50 values of 30.112 µg / mL. Apoptosis assay results showed

that the fraction F3 were able to induce apoptosis of cells.

Key words: breast cancer, T47D cell line, soursop (Annona muricata), cytotoxic, apoptosis.

INTRODUCTION

The breast cancer is the cancer with the highest incidence in Indonesia in the year of 2005, that is amounting to 39.23% of all cancer patients (MOH, 2007). The number of patients and the number of deaths were caused by

cancer that is continued to increase, must be accompanied by curative efforts. Cancer treatment is medically still

caused by problems because of its side effects are great.

The plant that is empirically trusted by societies to have anticancer properties are the leaves of the soursop (Annona muricata Linn.). Based on chemotaxonomy approach, some plant family Annonaceae that have been

studied have anticancer activity. The results of the plant family Annonaceae have been carried out. A. Montana

contains monotetrahydrofuranic acetogenins which have toxicity to liver cancer in Hep G2 cells (Liaw, et.al,

2005). The seeds of A. crassiflora have high antioxidant activity (Roesler, 2007). A. squamosa containing ribosome-inactivating protein (RIP), an immunotoxin for the treatment of cancer (Sismindari 1998). Based on

studies chemotaxonomy plants that have close kinship likely contain similar compounds (Princess, 2008) so that

data from previous studies had showed that the plant Annona muricata (soursop) is a potent anticancer Annonaceae family.

MATERIALS AND METHODS

a. Collection of plant material

The plant was collected from the side of Purwokerto, Indonesia. The collection was made in July. The plant was identified in the Laboratorium of Taxonomy, Department of Biology, Jenderal Soedirman University, Indonesia.

© Centre for Promoting Ideas, USA www.ijastnet .com

158

b. Preparation of simplicia

The material taken is soursop leaf. The material is washed with running water, then dried with dioven at 60 ± 1ºC.

Simplicia soursop leaf then powdered using by grinder.

c. Extraction and fractionation of soursop leaf

The powder was extracted using maceration with ethanol for 3 x 24 hours. Once filtered, the filtrate was evaporated to obtain ethanol extract of leaves of soursop. Ethanol extract and then fractionated by column

chromatography, respectively, using solvent were n-hexane, chloroform, ethyl acetate and methanol. The results

obtained by fractionation of fractions F1, F2, F3, and F4. Each extract tested cytotoxicity against T47D cells.

d. Preparation of stock solutions of test material

The soursop leaf extract is weighed 5 mg, followed by retrieval of DMSO to 5 ml (stock solution concentration of 1 mg / ml) and stored as stock solutions for subsequent use in research. Cytotoxic concentration of extract to a test

carried out by using the dilution medium. Tamoxifen concentrations obtained by dilution with medium. As a

control solvent, used 2% DMSO (v/v), ie the highest concentration of DMSO in the test compound.

e. Cytotoxic test with soursop leaf extract on T47D breast cancer cells by MTT assay

i. Preparation of RPMI 1640 medium.

ii. Activation, culturing, and harvesting cells T47D. iii. Preparation of soursop leaf extract and tamoxifen.

iv. Cytotoxic activity assay using 96 wells culture medium.

Cytotoxic test performed using 96 wells microcultures. The number of wells is divided into 8 lines (A, B, C, D, E, F, G, H). Each row contained 12 wells (no. 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12). Microcultures charging scheme can

be seen in Figure 1.

Microcultures is filled with the following steps:

Each of the columns mikroplate pitting of the 1-3, 9-11, AE line, column and row G 1-7 filled T47D cell suspension of 1.5 x 104 cells dissolved in culture medium RPMI 1640. Cells is then incubated for 24 hours in 5%

CO2 incubator at 37°C. After incubation, wells with columns 1-3, rows AE and medium, with a concentration of

100; 50; 25; 12.5; 6.25 ug / ml. Columns 9-11 AE line tamoxifen plus 100 mL, with a concentration of 300; 150;

75; 37.5; 18.75 ug / ml. Line G plus 100 mL RPMI medium without treatment (control). H lines 1-7 column filled with 100 mL RPMI (control medium). Microcultures were then stayed on incubation for 24 hours in 5% CO2

incubator at 37°C. After that, the media removed, any sinks coupled with 100 mL of new media and 10 mL MTT

reagent (10μl/100μl per wells), then incubated for 4-6 h in 5% CO2 incubator at 37 ° C. After that, add 100μl of sodium dodecyl sulfate (SDS) 10% in HCl 0.01%. Then mikroplate rocking at room temperature for 5 minutes.

After that, wrap with aluminum foil microcultures, incubation at room temperature overnight. Mikroplate is then

read using an ELISA at a wavelength of 595 nm (Lippman, 2004).

f. Apoptosis using the method of double stainning ethidium bromide-acridine orange

Apoptosis tested in ethanol extract. Apoptosis test performed using microcultures 24 wells. The number of wells is divided into four lines (A, B, C, D). Each row contained six wells (no. 1, 2, 3, 4, 5, 6). Microcultures charging

scheme can be seen in Figure 2. The Test carried out by preparing mikroplate with 24 wells and cover slip, slip

cover inserted into the wells using tweezers. 1000 mL cell suspension pipetted on to the cover slip that had been

inserted into the wells in row B No. 2, 3 and 4, then the cells were observed under a microscope to see the cell distribution. Tues incubated in an incubator overnight, taken mikroplate with 24 wells which already contain

cells, culture media and then discarded. Cells in the wells were washed with PBS each 500 mL, and then removed

from the wells with PBS Pasteur pipette and then gently inserted soursop leaf ethanol extract with IC50 concentrations of ½ as much as 1000 mL in column B2. In the same way, with concentrations of tamoxifen IC50, a

half as much as 1000 mL in column B3 and control cells (without treatment) in column B4 then on incubation for

10 hours. Media from wells removed by pasteur pipette mikroplate slowly at the wall. Cells in wells were washed with 500 mL PBS. PBS from the wells is slowly removed. Furthermore, the cover slip placed on the object glass,

then dropped into 10 mL reagent mixture of ethidium bromide-acridine orange on the cover slip and then

observed under a microscope flouresen.


159

Cells that showed green berfluoresens living cells, and cells that berfuoresens red indicates dead cells. Red

berfluoresens whole cells showed cell necrosis and cells that showed fragmented cells undergoing apoptosis.

RESULTS AND DISCUSSION

A. Determination of Plant

The determination of the soursop leaf plants performed at the Laboratory of Applied Biology Faculty of PKA used a reference book Flora of Java, Volume I (Bakhuizen van De Backer and Brink, 1963). The determination

made to avoid mistakes of the plants to be used in the study. The determination results was explained that plants

used in the study are Annona muricata Linn.

B. Preparation Simplicia

The soursop leaves obtained from Purwokerto Indonesia collected in June 2011. The leaves are washed with

running water to remove dirt or dust attached to the leaf. Leaves a clean cut into small pieces to speed up the

drying process. Drying process using the oven with a temperature of 60ºC, to ensure that no damage occurs an unstable compound by heating (Gunawan and Mulyani, 2004). Simplicia was dried and then carried out

pollination to increase the surface area of particles, so that the extraction process can be more effective and easier

solvent in attracting compounds contained in the cell.

C. Preparation of Extracts

Soursop leaf powder of 250 g performed maceration using ethanol solvent during 3x24 hours. The selection method is chosen in addition to being easy, simple and expected to reduce the risk of damage to the content of the

compounds so it is a suitable method used in the study. The extract obtained was 37.16 g which is mean the

rendemen about 14,86%.

D. Cytotoxic Test Ethanol Extract

Cytotoxicity test is a qualitative and quantitative tests to determine how cell death. The method used to see cytotoxic effects of ethanol extract of leaves of the soursop on T47D breast cancer cells is the MTT assay. The

principle of the MTT assay is a spectroscopic method is by determining the absorbance value of formazan. MTT

will be absorbed into the cell and entered into the system of cell respiration in mitochondria. The action of the enzyme active mitochondria in cells was metabolize tetrazolium salts, resulting in termination of tetrazolium ring

by dehydrogenase enzymes which lead to tetrazolium formazan transformed into water-insoluble but soluble in

SDS 10% and the purple coloured. Formazan formed is colored purple will be proportionate to the number of

living cells (Pebriana et al., 2008). Cells that die dissolved in water and remain yellow because the mitochondria of cells that die are not respiration tetrazolium ring is disconnected so it can not reduce MTT reagent to formazan

and the color is still yellow.

The observations made by microscopic showed that the number of formazan formed in control wells with media more than the formazan formed in the wells treated test compound. This suggests that the treatment of ethanol

extract of leaves of the soursop on T47D breast cancer cells can lead to death. Cells that are dead will not be

affected by the MTT reagent. Characteristic morphology of living cells is round with a protected cell wall that shines and stuck to the bottom plate, while the dark-colored cells that die and are not attached to the base plate.

After addition of MTT and incubated for 4 hours of diving, added SDS in 10% HCl. The reason the use of SDS

10% as it can dissolve the formazan crystals and the results of MTT reaction did not cause precipitation. After

settling for a night, then used an ELISA reader to determine absorbance values. 595nm wavelength is used because it is the maximum wavelength in order to obtain a sensitive and specific measurements. Absorbance

value of each test compound can be seen in Table 1.

The results of Table 1 showed that the higher of the concentration of test compound, is the lower absorbance values. This may imply that the test compound has a potency in inhibiting or killing the T47D cells. The stronger

intensity of the color purple is obtained the greater the absorbance. The graph can be seen in Figure 3. From the

graph above showed that the percentage inhibition of T47D cell growth is increased with increasing concentrations of test compound. Absorbance data obtained, is used to calculate IC50 values. IC50 value indicates

the value of concentration that can inhibit proliferation of T47D cancer cells by 50%. IC50 value of ethanol extract

is 17.149 µg/mL which is indicates that the concentration 17.149 µg/mL, ethanol extract inhibit proliferation of

T47D cancer cells by 50%.


160

Figure 4 shows the IC50 value of each test compound, shows that the ethanol extract and tamoxifen has cytotoxic

effects due to both the test compound IC50 value was 17.149 µg/mL and 13.38 µg/mL. An extract is said to have

cytotoxic activity if the IC50 value of less than 1000 µg/mL after 24 hours contact time (Meiyanto et al., 2008).

The smaller the IC50 value of a test compound the more toxic compound.

E. Apoptosis Test

Apoptosis is programmed cell death mechanism that is important in multicellular organisms to maintain

equilibrium. Tests performed to determine the mechanism of apoptotic cell death is through the mechanism of apoptosis. In T47D cells that were given the test compound indicates that berflouresens orange cells, whereas

cells that are not given the test compound indicates green berfloures cells as shown in figure 5. The apoptosis

assay results by the method of double stainning above show that cancer cells are treated T47D ethanol extract of

leaves of soursop (A) and tamoxifen (B) some berflouresens orange. This indicates that the test compound can induce apoptosis. In control cells berflouresens still look bright green oval which means the cells do not undergo

apoptosis because the cells live only absorb acridine orange.

F. Separation of Ethanol Extracts of Active Fraction by Column Chromatography

Fractionation of the compounds contained in the ethanol extract was conducted by column chromatography. In

this method previously conducted to determine the orientation of the stationary phase to be used. After doing

orientation, the best of the stationary phase is obtained by silica gel GF 254 0.2-0.5 mm. The mobile phase used in the coloum chromatography ethanol extract was successively hexane, chloroform, ethyl acetate and methanol,

each of which produces fractions of hexane, chloroform, ethyl acetate and methanol fractions. Each fraction was

then performed tests such as the cytotoxic ethanol extract.

G. Soursop Leaf Fraction Cytotoxic Test

Performed the same test method to test cytotoxic cytotoxic ethanol extract of leaves on soursop. Absorbance value

of each fraction can be seen in Table 3. The results in Table 3 showed that the higher the concentration of test

compound have the lower absorbance values. The four factions have differences IC50 value. The percentage inhibition of T47D cell growth increased with increasing concentrations of test compound as shown in the graph

6. Absorbance data obtained, is used to calculate IC50 values. The smaller value of IC50 have the greater potential

cytotoxic against to T47D cell lines. The fractions that have a better cytotoxic potency, respectively consecutive are hexane fraction, chloroform, methanol and etyl acetate fraction with IC50 are 143.077; 120.718; 44.987 and

31.268 µg/mL. Etyl acetate fraction has the best potency of cytotoxic among other fractions against to T47D cell

lines. Figure 7 shows the IC50 values of each fraction. Ethyl acetate fraction has the smallest IC50 value, which means having the greatest cytotoxic effect than the three other fractions.

CONCLUSIONS

Ethanol extract of leaves of soursop (Annona muricata) has a cytotoxic activity in T47D breast cancer cell lines

with IC50 of 17.149 µg/mL and can induce apoptosis. Etyl acetate fraction has the best potency of cytotoxic

among other fractions against to T47D breast cancer cell lines with value of IC50 was 31.268 µg/mL.

Acknowledgements

The authors acknowledge the financial support provided by University of Jenderal Soedirman, Indonesia for

Institutional Research (No. 1947/H.23.9/PN/2011).

REFERENCES

Anomin, 2008. Plant resources of South-East Asia 8 (PROSEA) : Vegetables, Indonesia, p.152-153.

Brodsky, I., Crilley, P., and Terzian, A.E.L.1990. Cancer Chemotherapy in Basic Pharmacology in Medicine, Ed

3rd

, Eds. DiPalma, J.R., pp. 549-565, Mc Graw-Hill Publishing Company, Singapore. Cummings, J.H. and Bingham,S.A.1998. Diet and the Prevention of Cancer. Bri Med J.10:,1213-7

Ferreira, L., Perestrelo, R., Camara, J.S., 2009, Comparative analysis of the volatile fraction from Annona

cherimola Mill. cultivars by solid-phase microextraction and gas chromatography–quadrupole mass

spectrometry detection, Talanta 77 : 1087–1096 Gonzales, I.A., Agiirre, K.K., Roaro, L.M., 2008, Anticarcinogenicandgenotoxic effectsproducedby acetogenins

isolated from Annona muricata, Toxicology Letters, 180 : 32–246


161

Gunawan, D., dan Mulyani, S., 2004, Natural Herbal (Farmakognosi), Penebar Swadaya, Jakarta.

Health Departement. 2007. Indonesia Health Profile 2005. Jakarta, Indonesia

Liaw, C.C., Chang, F.R., Chen, S.L., Wu, C.C., Lee, K.H., 2005, Novel cytotoxic monotetrahydrofuranic

Annonaceous acetogenins from Annona montana, Bioorganic & Medicinal Chemistry 13 : 4767–4776 Lima, L.A.R., Pimenta, L.P.S., Amelia, M., 2010, Acetogenins from Annona cornifolia and their antioxidant

capacity, Food Chemistry 122: 1129–1138

Lumongga, F., 2008, Apoptosis, Departement of Patology and Anatomy, Medicine Faculty, Sumatera Utara University, Medan, Indonesia.

McPerson,K., Steel, C.M., and Dixon, J.M. 2000. ABC of Breast Diseases. Breast Cancer Epidemiology, Risk

Factor, and Genetics. Bri Med J.321: Sept 9 ;624-8 Mosmann, T.1983. Rapid Colorimetric Assay for Cellular Growth and Survival: Application to Proliferation and

Cytotoxicity Assays. J Immunol Meth 65: 55-63.

Padma, P., Pramod, N.P., Thyagarajan, S.P., Khosa, R.L., 1998, Effect of The Extract of Annona muricata and

Petunia nyctaginiflora on Herpes Simplex Virus, Journal of Ethnopharmacology, 61: 81-83. Putri, F.F ., 2008. Prascreening Test Activity of Anticancer From Extract of n-heksana Pandanus conoideus Lamk

Using Brine Shrimp Lethality Test (BST). Airlangga University Library. ADLN digital collection.

Surabaya. http://www.adln.lib.unair.ac.id., Accesed 20 December 2009. Rachmani

a, Eka Prasasti N., 2010, Review :Soursoup (Annona muricata) : Content of Compounds and Biology

Activity. National Seminar of Health, Jenderal Soedirman University, Indonesia.

Rachmanib, Eka Prasasti N., Suhesti, T.S., 2010, Potency Antioxidant of Soursoup Leaves (Annona muricata l.).

Rang, H.P., Dale, M.M., Ritter, J.M., and Moore, P.K. 2003. Cancer Chemotherapy in Pharmacology. Ed 5th, Eds.

Hunter, L. pp. 693-707, Churchill Livingstone Elsevier Science Limited, United Kingdom.

Ribble, D., Goldstein, N.B., Norris, D.A., Shellman, Y.G., 2005, A Simple Technique for Quantifying Apoptosis

in 96 Well Plates, BMC Biotechno, 5. Rocha, L.G., Almeidab, J.R.G.S., Mace, R.O., 2005, A review of natural products with antileishmanial activity,

Phytomedicine, 12 : 514–535

Rode, H.D., Eisel, D., Frost, I. 2004. Apoptosis, Cell Death and Cell Proliferation Manual. 3rd ed. Roche Applied

Science, USA.

Roesler, R., Catharino, R.R., Malta, L.G., 2007, Antioxidant activity of Annona crassiflora: Characterization of

major components by electrospray ionization mass spectrometry, Food Chemistry 104 : 1048–1054

Sismindari, Hussana, A., Mubarika, S., 1998, Squensing of DNA Superkoil Double Helix with In Vitro Method From Annona Squamosa L.,

Majalah farmasi indonesia, 9: 4-8

Vogel,V.G. 2000. Breast Cancer Prevention: A Review of Current Evidence.CA Cancer J Clin; 50:156-70 Wele, A., Zhang, Y., Caux, C., 2004, Annomuricatin C, a novel cyclohexapeptide

Figure 1. Microcultures charging scheme for the cytotoxic test Descriptions:

EK = ethanol extract, T = Tamoxifen, K = control cells (cancer cells and medium), M = medium

1 2 3 4 5 6 7 8 9 10 11 12

A EK EK EK T T T

B EK EK EK T T T

C EK EK EK T T T

D EK EK EK T T T

E EK EK EK T T T

F

G K K K K K K K

H M M M M M M M

http://www.adln.lib.unair.ac.id/


162

1 2 3 4 5

6

A

B S T K

C

D

Figure 2. Charging scheme to test apoptosis microcultures

Descriptions: S: Ethanol extract of leaves of soursop (Annona muricata)

Q: Tamoxifen

C: Control cells (T47D cells and media)

Table 1. The mean absorbance, percentage inhibition of T47D cells and IC50 values from ethanol extract of

leaves of A.muricata

Test

Materials

Concentration

(µg/mL)

Mean

absorbance

Living Cells

(%)

Retardation

(%)

IC50

(µg/mL)

Ethanol

extract

A.muricata

500

250

125

62.5 31.25

0.161

0.129

0.141

0.138 0.457

7.71

2.34

4.36

3.80 57.28

92.29

97.66

95.64

96.20 42.72

17.149

Tamoxifen 50 0.253 0 100

13.38

25 0.262 0.19 99.81

12.5 0.889 71.62 28.38

6.25 1.070 92.29 7.71

3.125 0.167 91.95 8.05

Figure 3. Graph showing the relationship between the percentage inhibition of T47D cells with various

concentrations of test compound

0

50

100

150

3.1

25

6.25

12.5 25 50

31

.25

62.5

125

250

500

% in

hib

itor

Concentration µg/mL

Ethanol

Tamoxifen


163

Figure 4. IC50 values of each test compound

Figure 5. Apoptosis assay results soursop leaf ethanol extract of T47D cancer cells. Ethanol extract of

leaves of soursop (A), tamoxifen (B), control cells (C). Red arrows indicate cells undergoing apoptosis and

blue arrows indicate the cells not undergoing apoptosis.

0

10

20

Etanol

Tamoxifen

1

3

1

7

B A

C


164

Table 3. The mean absorbance, percentage inhibition of T47D cells and IC50 values after administration of

test compound

Fraction Consentratio

n (µg/mL)

Mean Absorbances Living Cells (%) Retardation (%) IC50 (µg/mL)

Hexan Fraction

500 0.263 0.23 99.77

143.077 250 0.266 0.65 99.35

125 0.987 82.75 17.25

62.5 1.098 95.48 4.52

31.25 1.107 96.47 3.53

Chloroform Fraction

500 0.264 0.42 99.58

120.718 250 0.296 4.03 95.97

125 0.838 65.81 34.19

62.5 0.992 83.40 16.60

31.25 1.098 95.44 4.56

Etyl Acetat Fraction 500

250

125

62.5

31.25

0.292

0.269

0.261

0.283

0.856

3.53

0.99

0

2.55

67.90

96.47

99.01

100

97.45

32.10

31.268

Methanol Fraction 500

250

125

62.5

31.25

0.279

0.245

0.264

0.579

0.886

2.15

0

0.38

32.25

71.32

97.95

100

99.62

63.75

28.68

44.987

Tamoxifen 50

25

12.5

6.25

3.125

0.253

0.262

0.889

1.070

0.167

0

0.19

71.62

92.29

91.95

100

99.81

28.38

7.71

8.05

13.38

Figure 6. Graph showing the relationship between the percentage inhibition of T47D cells with various

concentrations of test compound

Figure 7. IC50 values of each test compound

0

50

100

150

3.12

5

6.25

12.5 25 50

31.2

5

62.5

125

250

500

% I

nh

ibit

or

Concentration µg/mL

Faction N Hexane

Faction Chloroform

Faction Ethyl acetate

Faction Methanol

Tamoxifen

0

50

100

150

IC50

(µg/

mL

)

test compound

Cytotoxicity IC50

Faction N Hexane

Faction Chloroform

Faction Ethyl acetate

Faction Methanol

Tamoxifen

3

1

1

2

1

1

3

4

5

1

4

3

ORIGINAL PAPER

Three new megastigmanes from the leaves of Annona muricata

Ayano Matsushige • Katsuyoshi Matsunami •

Yaichiro Kotake • Hideaki Otsuka •

Shigeru Ohta

Received: 13 July 2011 / Accepted: 10 August 2011 / Published online: 4 September 2011

� The Japanese Society of Pharmacognosy and Springer 2011

Abstract Three new megastigmanes (1–3), named

annoionols A and B (1, 2) and annoionoside (3), were

isolated from the leaves of Annona muricata L. (Annonaceae)

together with 14 known compounds (4–17). Among the

known compounds, annoionol C (4) was isolated from a

natural source for the first time. The structures of all

compounds were elucidated by spectroscopic and chemical

analyses.

Keywords Annona muricata � Annonaceae �Megastigmane � Annoionoside � Annoionol

Introduction

Megastigmanes and their glycosides are a currently

expanding class of compounds. In our continuing studies

on sub-tropical plants collected on Okinawa, we have

phytochemically investigated the leaves of Annona muri-

cata L. (Annonaceae) in this paper. The plant is an ever-

green tree of medium height, and is found in the Americas,

Africa and Southeast Asia. The edible fruit of this plant is

well-known as ‘‘soursop’’ and is cultivated widely in

tropical and subtropical areas nowadays. However, in the

Caribbean, consumption of this fruit is suggested to have a

connection to an atypical form of Parkinson’s disease [1].

The present study describes the isolation and structural

elucidation of three new megastigmanes (1–3) together

with 14 known compounds (4–17).

Results and discussion

Air-dried leaves of A. muricata were extracted with MeOH

three times by maceration. The combined MeOH extract was

evaporated and partitioned with n-hexane, CHCl3, EtOAc

and 1-BuOH successively, to give n-hexane, CHCl3, EtOAc

and 1-BuOH soluble fractions, respectively. The CHCl3 and

EtOAc-soluble fractions were combined because of the

similarity of their TLC patterns. The residue of the 1-BuOH-

soluble fraction and the combined CHCl3- and EtOAc-sol-

uble fractions were subjected to various kinds of column

chromatography to yield 17 compounds (1–17) (Fig. 1).

Annoionol A (1) was obtained as a colorless amorphous

powder and its molecular formula was determined to be

C13H26O3 from its high-resolution electrospray-ionization

time-of-flight mass spectrum (HR-ESI-TOF-MS) (m/z =

253.1775 [M ? Na]?). The IR absorption at 3382 cm-1

indicated the presence of a hydroxyl group. The 1H-NMR

spectrum exhibited the signals ascribable to two singlet (dH

0.88 and 0.94) and two doublet [dH 1.07 (d) and 1.15 (d)]

methyls, and two oxygenated methines (dH 3.49, 3.65 and

2.78) (Table 1). The 13C-NMR and DEPT spectra indicated

the presence of 13 carbon signals comprising those of four

methyls, three methylenes, five methines, of which three

were oxygenated, and one quaternary carbon (Table 1).

The proton (dH 2.78) was assigned to that on the highly

deshielded oxygenated methine carbon (dC 82.4) by the

HMQC spectrum. These atypically counter-shifted proton

A. Matsushige � Y. Kotake � S. Ohta

Department of Xenobiotic Metabolism and Molecular

Toxicology, Graduate School of Biomedical Sciences,

Hiroshima University, 1-2-3 Kasumi, Minami-ku,

Hiroshima 734-8553, Japan

K. Matsunami � H. Otsuka (&)

Department of Pharmacognosy, Graduate School of Biomedical

Sciences, Hiroshima University, 1-2-3 Kasumi,

Minami-ku, Hiroshima 734-8553, Japan

e-mail: [email protected]

123

J Nat Med (2012) 66:284–291

DOI 10.1007/s11418-011-0583-1

and carbon were also observed in the case of elaeocarp-

ionoside [2] and fruticosides A and B [3]. Considering the

above together with one degree of unsaturation, a

megastigmane skeleton was formulated for 1. The 1H–1H

correlation spectroscopy (COSY) spectrum revealed a

proton coupling framework from H-2 to H3-10 together

with the connection of H3-13 to H-5, which revealed the

planar structure of 1 shown in Fig. 1. The heteronuclear

multiple bond correlation (HMBC) spectrum confirmed

this structure (Fig. 2a). The relative configuration of 1 was

determined by considering the coupling constants (Table 1)

and phase-sensitive (PS) nuclear Overhauser enhancement

spectroscopy (NOESY) spectral data. The axial coupling of

H-2ax [dH 1.21 (1H, dd, J = 13, 12 Hz)], H-3 [dH 3.49

(1H, ddd, J = 12, 9, 5 Hz)], H-4 [dH 2.78 (1H, dd, J = 10,

9 Hz)] and H-6 [dH 0.71 (1H, ddd, J = 11, 4, 2 Hz)], and

the NOE correlations around the six-membered ring por-

tion are in good agreement with the relative configuration

shown in Fig. 2b. Finally, the absolute configuration of 1

was determined by the modified Mosher’s method (Fig. 2c)

[4]. The structure of annoionol A (1) was therefore eluci-

dated to be (3R,4R,5S,6S,9R)-megastigma-3,4,9-triol.

Compound 2 was obtained as a colorless amorphous

powder and its molecular formula was determined to

be C13H24O4 by HR-ESI-TOF-MS (m/z = 267.1562

[M ? Na]?). The 1H and 13C-NMR spectra were similar to

HOOH

OH

OR

OOH

OR

2 HO

OH

4OH

GlcO

OH

7

OHHO 8OGlc

C

H

O

OO

HORO

OH

HO

OH

OH1

9

O

OGlc

OGlc

O

O

HO

OH

OH

OR2

12

13

HGlc

R

3

5 HGlc

R

6

10 -MeR

11 -Meαβ

R1

14R1

15

R2

1617

HH

H

OHGlc(6)RhaGal(6)Rha

Glc(6)Rha

Glc2

Glc(6)Rha

3

4 5

69

11 12

13

10

Fig. 1 Structures of the isolated compounds

Table 1 1H-NMR spectral data for 1–4 (d in ppm, J in Hz, in CD3OD)

1 2 3 4

2 1.21 ax dd (13, 12) 1.43 eq dd (12, 5) 1.48 eq dd (13, 5) 1.10 ax dd (12, 12)

1.62 eq dd (13, 5) 1.81 ax dd (12, 12) 1.82 ax dd (13, 12) 1.65 eq ddd (12, 3, 2)

3 3.49 ddd (12, 9, 5) 3.60 ddd (12, 9, 5) 3.71 ddd (12, 9, 5) 3.72 m

4 2.78 dd (10, 9) 3.26 dd (11, 9) 3.36 dd (11, 9) 1.03 ax ddd (12, 12, 12)

– – – 2.12 eq m

5 1.37 m 1.77 dq (11, 7) 1.98 dq (11, 7) 1.45 m

6 0.71 ddd (11, 4, 2) – – 0.78 ddd (11, 5, 2)

7 1.08 m 5.55 dd (16, 1) 5.55 dd (16, 1) 1.09 m

1.56 m – – 1.58 m

8 1.44 m 5.70 dd (16, 6) 5.70 dd (16, 6) 1.40 m

1.51 m – – 1.50 m

9 3.65 m 4.30 qdd (6, 6, 1) 4.30 qdd (6, 6, 1) 3.66 qt (6, 6)

10 1.15 d (6) 1.25 d (6) 1.25 d (6) 1.15 d (6)

11 0.88 s 1.02 s 1.01 s 0.84 s

12 0.94 s 0.88 s 0.88 s 0.97 s

13 1.07 d (6) 0.95 d (7) 1.04 d (7) 3.42 dd (10, 7)

– – – 3.70 dd (10, 3)

10 4.311 d (8)

20 3.25 dd (9, 8)

30 3.37 dd (9, 9)

40 3.34 m

50 3.34 m

60 3.87 dd (12, 1)

3.67 dd (12, 5)

m Multiplet or overlapping signals

J Nat Med (2012) 66:284–291 285

123

those of 1 except for the appearance of an oxygen-bearing

quaternary carbon and a trans-double bond. The 1H–1H

COSY spectra revealed two proton coupling networks from

H-2 to H-13 through H-5, and from H-7 to H-10, which

revealed the planar structure of 2 shown in Fig. 1. The

HMBC spectrum also confirmed that the structure of 2 was

megastigman-7-ene-3,4,6,9-tetraol (Fig. 3). The coupling

constant of 16 Hz for H-7 and H-8 indicated the geometry

of the double bond to be the E-form. The relative config-

uration of 2 was determined by considering the coupling

constants around the six-membered ring portion and PS

NOESY spectral data. The axial coupling of H-2ax [dH

1.81 (1H, dd, J = 12, 12 Hz)], H-3 [dH 3.60 (1H, ddd,

J = 12, 9, 5 Hz)], H-4 [dH 3.26 (1H, dd, J = 11, 9 Hz)]

and H-5 [dH 1.77 (1H, dq, J = 11, 7 Hz)] revealed the

equatorial nature of 3-OH, 4-OH and 5-Me. The relative

configuration of C-6 was also determined to be as shown in

Fig. 2b from the NOE correlations between H-7 and H-5ax,

and between H-7 and H3-11. Finally, the absolute config-

uration of 2 was determined by the modified Mosher’s

method (Fig. 3). Annoionol B (2) was therefore elucidated

to be (3R,4R,5R,6R,7E,9R)-megastigman-7-ene-3,4,6,9-

tetraol.

Compound 3 was obtained as a colorless amor-

phous powder and its molecular formula was determined

to be C19H34O9 by HR-ESI-TOF-MS (m/z = 429.2090

[M ? Na]?). The 1H and 13C NMR spectra were closely

similar to those of 2. An anomeric proton and six oxygen-

ated carbon signals at dc 105.4, 78.1 (C 9 2), 75.5, 71.6 and

62.5 indicated the presence of glucopyranose. The HMBC

correlations between H-4 and C-10, and H-10 and C-4 con-

firmed the connectivity of glucose at C-4, and other two-

dimensional NMR analyses confirmed the planar structure

of 3 (Fig. 1). The coupling constant of the anomeric proton

(8 Hz) and chiro-optical HPLC analysis of the sugar frac-

tion following enzymatic hydrolysis of 3 revealed that the

configuration of glucose was of the D-series. The relative

configuration around the six-membered ring of 3 was

determined to be equatorial for 3-OH, 4-OH and 5-Me by

considering the axial–axial coupling constants of the cor-

responding protons (Table 1). The relative configuration of

C-6 was also determined to be as shown in Fig. 3 from the

NOE correlations between H-7 and H-5ax, and between H-7

and H3-11. The absolute stereochemistry at C-3 was tenta-

tively determined to be 3R by application of the glucosy-

lation-induced shift-trend rule [5]. Finally, the aglycone

liberated on enzymatic hydrolysis was then esterified to

afford (R)- and (S)-a-methoxy-a-(trifluoromethyl)phenyla-

cetic acid (MTPA) diesters. The 1H-NMR spectral data

of both the (R) and (S)-MTPA diesters were essentially

identical to those of the MTPA derivatives of 2. The

absolute structure of 3 was therefore elucidated to be

(3R,4R,5R,6R,7E,9R)-megastigman-7-ene-3,4,6,9-tetraol

4-O-b-D-glucopyranoside (Fig. 3).

a

b

c

Fig. 2 Absolute structure of 1. a COSY (bold line) and HMBC

(arrows) correlations. b Important NOESY correlations. c Analysis

by the modified Mosher’s method. The values are expressed as DdS-R

Table 2 13C-NMR spectral data for 1–4 (d in ppm, in CD3OD)

1 2 (=3a) 3 (d3–3a) 4

1 36.2 40.1 39.7 36.8

2 49.0 43.9 43.0 51.9

3 72.4 72.9 71.1 (-1.8) 67.5

4 82.4 78.7 90.3 (?11.6) 40.8

5 41.6 42.1 41.6 (-0.5) 42.7

6 52.7 80.1 80.1 48.5

7 26.7 133.7 133.3 26.1

8 42.4 135.6 135.8 42.1

9 69.2 69.2 69.1 69.1

10 23.4 24.2 24.2 23.4

11 21.5 25.5 25.4 21.4

12 30.9 25.2 25.1 31.3

13 16.8 12.1 12.1 66.1

10 105.4

20 75.5

30 78.1

40 71.6

50 78.1

60 62.5

m Multiplet or overlapping signals

286 J Nat Med (2012) 66:284–291

123

Compound 4 was obtained as a colorless amorphous

powder and its molecular formula was revealed to be the

same as that of 1, C13H26O3, by HR-ESI-TOF-MS

(m/z = 253.1770 [M ? Na]?) analysis. In the 13C-NMR

spectrum, two methylenes (dC 40.8 and 66.1), of which one

was oxygenated, were seen in 4 instead of the methyl (C-13

of 1) and oxygenated methine (C-4 of 1) groups observed

in 1. Therefore, the planar structure of 4 was assumed to be

a positional isomer of 1 having a hydroxy group at C-13

(Fig. 1). The relative stereostructure of 4 was elucidated

from the coupling constants and by two-dimensional NMR

analyses (Fig. 4). Finally, the absolute configuration of 4

was determined by the modified Mosher’s method (Fig. 4).

Thus, the structure of annoionol C (4) was elucidated to be

(3S,5R,6S,9R)-megastigman-3,9,13-triol. Annoionol C (4)

has already been reported as the aglycone of two related

glucosides, bridelionoside D [6] and rhusonoside A [7];

however, this was the first time that 4 had been isolated

from a natural source.

It is noteworthy that the several stereoisomers for

compounds 1–3 have already been isolated from various

plant sources [8–11]; however the megastigmanes having

(3R,4R)-diol moiety have rarely been found in nature. The

remaining known compounds (5–17) were identified by

comparison of the spectroscopic data with those reported in

the literature, as follows. Vomifoliol (5), [a]D26 ?176.6�

(c 0.31, MeOH) [12], roseoside (6), [a]D25 ?100.8� (c 0.54,

MeOH) [13], turpinionoside A (7), [a]D27 -37.4� (c 0.53,

MeOH), tR = 18.5 min under the same HPLC conditions

as described in the literature [14, 15], citroside A (8), [a]D27

-83.9� (c 0.73, MeOH) [16], blumenol C (9), [a]D26 ?49.6�

(c 0.22, CHCl3) [17], (?)-epiloliolide (10), [a]D26 ?22.3�

(c 0.12, CHCl3) [18], loliolide (11), [a]D23 -67.9� (c 0.88,

MeOH) [19], (1S,2S,4R)-trans-2-hydroxy-1,8-cineole b-D-

glucopyranoside (12), [a]D26 ?0.67� (c 0.48, MeOH) [20],

(Z)-3-hexenyl b-D-glucopyranoside (13), [a]D27 -20.5�

(c 0.25, MeOH) [21], rutin (14), [a]D27 -10.4� (c 9.4,

MeOH) [22], kaempferol 3-O-rutinoside (15), [a]D26 -11.1�

(c 0.42, MeOH) [23], kaempferol 3-O-robinobioside (16),

[a]D26 -63.1� (c 0.36, pyridine) [24], kaempferol 3-O-b-D-

(200-O-b-D-glucopyranosyl,600-O-a-L-rhamnopyranosyl)gluco-

pyranoside (17), [a]D27 -82.2� (c 0.63, MeOH) [25].

Compounds 1–4 were examined for 1,1-diphenyl-2-

picrylhydrazyl (DPPH) radical scavenging activity, and

also for tumor cell growth inhibitory activity toward A549

and SBC-3 by means of a MTT assay. However, these

compounds did not show any significant activity at

100 lM.

Experimental

General experimental procedures

Silica gel column chromatography (CC) was performed on

silica gel 60 (Merck, Darmstadt, Germany), and reversed-

phase [octadecyl silica gel (ODS)] open CC (RPCC) on

Cosmosil 75C18-OPN (Nacalai Tesque, Kyoto, Japan)

(U = 5 cm, L = 20 cm). HPLC was performed on ODS

(Cosmosil; Nacalai Tesque, Japan; U = 10 mm, L =

250 mm), and the eluate was monitored with a refractive

index monitor.

Optical rotations were measured on a JASCO P-1030

polarimeter. IR spectra were measured on a Horiba FT-710

Fourier transform infrared spectrophotometer. NMR spec-

tra were taken on a JEOL ECA 600 spectrometer at

600 MHz for 1H, and 150 MHz for 13C, respectively, with

tetramethylsilane as an internal standard. Positive-ion

Fig. 3 Absolute structures of 2 and 3. COSY (bold lines), HMBC

(arrows), and NOESY (arrows in 3D drawing) correlations were

indicated. The DdS–R values are expressed in ppm

Fig. 4 Absolute structure of 4. a Pivaloyl chloride/pyridine. b (R) or

(S)-MTPA, EDC, DMAP/CH2Cl2. COSY (bold line), HMBC

(arrows), and NOESY (arrows in 3D drawing) correlations are

indicated. The DdS–R values are expressed in ppm

J Nat Med (2012) 66:284–291 287

123

HR-ESI-TOF-MS was recorded on a Applied Biosystem

QSTAR XL spectrometer. A VersaMax (Molecular Devi-

ces) was used as a microplate reader.

Plant material

Leaves of A. muricata were collected in Yaeyama-gun,

Okinawa, Japan, in November 2004, and a voucher speci-

men was deposited in the Herbarium of the Department of

Pharmacognosy, Graduate School of Biomedical Sciences,

Hiroshima University (No. 04-AM-Okinawa-1105).

Extraction and isolation

Air-dried leaves of A. muricata (520 g) were extracted with

MeOH (2 L) three times by maceration. The MeOH

extracts were combined and evaporated to dryness to afford

a viscous gummy material (78.2 g). This residue was sus-

pended in 1.5 L of H2O, and then extracted with equal

volumes of n-hexane, CHCl3, EtOAc and 1-BuOH suc-

cessively to afford 13.4, 34.9, 1.1 and 5.0 g of fractions,

respectively. The remaining H2O layer was concentrated to

furnish an H2O-soluble fraction (23.9 g). The 1-BuOH

soluble fraction (5.0 g) was subjected to silica gel column

chromatography (U = 2.5 cm, L = 50 cm) with stepwise

gradient elution with increasing amounts of MeOH in

CHCl3 [CHCl3 (1 L), CHCl3-MeOH (20:1, 750 mL),

(10:1, 750 mL), (5:1, 750 mL), (3:1, 750 mL)], CHCl3:

MeOH:H2O = 15:6:1, 750 mL, and MeOH, 750 mL. The

combined residue (0.83 g) of the CHCl3-MeOH (10:1) and

(5:1) eluates obtained on silica gel CC was subsequently

subjected to RPCC with stepwise gradient elution with

increasing amounts of MeOH in H2O (30, 50, 70, 90 and

100% MeOH, 300 mL each containing 0.1% TFA). The

residue (470 mg) obtained from the 30% MeOH eluate was

purified by HPLC (ODS) with 13% CH3CN to afford 2

(12.7 mg), 6 (14.3 mg), 12 (4.8 mg) and 13 (7.1 mg) from

the peaks at 10.8, 18.2, 23.5 and 33.0 min (flow rate:

2.5 mL/min), respectively. The residue (320 mg) obtained

from the 50% MeOH eluate was purified by HPLC (ODS)

with 30% acetone to afford 4 (11.1 mg) from the peak at

8.1 min (flow rate: 2.8 mL/min).

The residue (1.43 g) of the CHCl3–MeOH (3:1) eluate

obtained on silica gel CC was subsequently subjected to

RPCC with stepwise gradient elution with increasing

amounts of MeOH in H2O (30, 50, 70, 90 and 100%

MeOH, 300 mL containing 0.1% TFA). The residue

(850 mg) obtained from the 30% MeOH eluate was puri-

fied by HPLC (ODS) with 13% CH3CN to afford 3

(64.3 mg), 7 (9.7 mg) and 8 (12.8 mg) from the peaks at

10.5, 14.2 and 18.1 min (flow rate: 2.5 mL/min), respec-

tively. The residue (620 mg) obtained from the 50%

MeOH eluate was purified by HPLC (ODS) with 30%

acetone to afford 15 (59.6 mg) from the peak at 14.9 min

(flow rate: 2.8 mL/min).

The residue (1.13 g) of the CHCl3:MeOH:H2O = 15:6:1

eluate obtained on silica gel CC was subsequently sub-

jected to RPCC with stepwise gradient elution with

increasing amounts of MeOH in H2O (30, 50, 70, 90 and

100% MeOH, 300 mL each containing 0.1% TFA). The

residue (400 mg) obtained from the 50% MeOH eluate was

purified by HPLC (ODS) with 20% CH3CN–0.1% TFA to

afford 14 (199 mg) from the peak at 11.2 min (flow rate:

3.0 mL/min).

The residue (1.05 g) of the MeOH eluate obtained on

silica gel CC was subsequently subjected to RPCC with

stepwise gradient elution with increasing amounts of

MeOH in H2O (30, 50, 70, 90 and 100% MeOH, 300 mL

each containing 0.1% TFA). The residue (660 mg)

obtained from the 50% MeOH eluate was purified by

HPLC (ODS) with 22% acetone to afford 17 (12.7 mg) and

14 (14.3 mg) from the peaks at 8.9 and 15.5 min (flow rate:

2.8 mL/min), respectively.

The CHCl3 and EtOAc-soluble fractions were combined

and subjected to silica gel column chromatography

(U = 5.8 cm, L = 38 cm) with stepwise gradient elution

with increasing amounts of MeOH in CHCl3 [CHCl3 (3 L),

CHCl3-MeOH (100:1, 3 L), (50:1, 3 L), (30:1, 3 L), (20:1,

3 L), (10:1, 3 L), (5:1, 3 L), (2:1, 3 L) and (MeOH, 3 L)].


obtained on silica gel CC was subsequently subjected to

RPCC with stepwise gradient elution with increasing

amounts of MeOH in H2O (30, 50, 70, 90 and 100%

MeOH, 300 mL each with 0.1% TFA). The residue

(70 mg) obtained from the 30% MeOH eluate was purified

by HPLC (ODS) with 20% CH3CN-0.1% TFA to afford 10

(4.0 mg) from the peak at 17.2 min (flow rate: 2.5 mL/

min). The residue (120 mg) obtained from the 50% MeOH

eluate was purified by HPLC (ODS) with 38% acetone to

afford 11 (8.8 mg) and 9 (5.0 mg) from the peaks at 8.4

and 12.1 min (flow rate: 2.5 mL/min), respectively.


obtained on silica gel CC of the CHCl3 and EtOAc soluble

fractions was subsequently subjected to RPCC with step-

wise gradient elution with increasing amounts of MeOH in

H2O (30, 50, 70, 90 and 100% MeOH, 300 mL each with

0.1% TFA). The residue (60 mg) obtained from the 30%


CH3CN to afford 5 (6.0 mg) from the peak at 10.8 min

(flow rate: 2.5 mL/min).


obtained on silica gel CC of the CHCl3 and EtOAc-soluble





288 J Nat Med (2012) 66:284–291

123


acetone–0.1% TFA to afford 1 (1.1 mg) from the peak at

9.6 min (flow rate: 2.8 mL/min).


obtained on silica gel CC of the CHCl3 and EtOAc-soluble






acetone to afford 16 (7.1 mg) and 15 (9.9 mg) from the

peaks at 10.5 and 12.2 min (flow rate: 2.8 mL/min),

respectively.

Annoionol A (1)

Amorphous powder; [a]D24 ?1.45� (c 0.11, MeOH); IR mmax

(film) cm-1: 3382, 2964, 2927, 1594, 1372, 1294, 1255,

1124, 1063; 1H- and 13C-NMR (CD3OD): Tables 1 and 2,

respectively; HR-ESI-TOF-MS (positive-ion mode)

m/z: 253.1775 [M ? Na]? (Calcd for C13H26O3Na:

253.1774).

Annoionol B (2)

Amorphous powder; [a]D24 -4.17� (c 1.27, MeOH); IR mmax

(film) cm-1: 3398, 2971, 2934, 1676, 1516, 1460, 1371,

1136, 1054; 1H and 13C NMR (CD3OD): Tables 1 and 2,

respectively; HR-ESI-TOF-MS (positive-ion mode) m/z:

267.1562 [M ? Na]? (Calcd for C13H24O4Na: 267.1566).

Annoionoside (3)


(film) cm-1: 3396, 2971, 2932, 1672, 1516, 1444, 1373,

1132, 1075, 1033; 1H and 13C NMR (CD3OD): Tables 1

and 2, respectively; HR-ESI-TOF-MS (positive-ion mode)

m/z: 429.2090 [M ? Na]? (Calcd for C19H34O9Na:

429.2095).

Annoionol C (4)


(film) cm-1: 3374, 2931, 1672, 1370, 1201, 1125, 1044; 1H

and 13C NMR (CD3OD): Tables 1 and 2, respectively;

HR-ESI-TOF-MS (positive-ion mode) m/z: 253.1770

[M ? Na]? (Calcd for C13H26O3Na: 253.1774).

Preparation of (R)- and (S)-MTPA diesters

(1a and 1b) of 1

A solution of 1 (0.5 mg) in 1 mL of dehydrated CH2Cl2 was

reacted with (R)-a-methoxy-a-(trifluoromethyl)phenylacetic

acid (MTPA) (25.2 mg) in the presence of 1-ethyl-3-(3-

dimethylaminopropyl)carbodiimide hydrochloride (EDC)

(23.5 mg) and 4-N,N0-dimethylaminopyridine (DMAP)

(12.6 mg), followed by standing at 35�C for 12 h. After the

addition of 1.0 mL each of H2O and CHCl3, the solution

was washed with 1 M HCl (1.0 mL), NaHCO3-saturated

H2O (1.0 mL) and saturated brine (1.0 mL) successively.

The organic layer was dried over Na2SO4 and then evapo-

rated under reduced pressure. The residue was purified by

preparative TLC [silica gel (0.25 mm thickness, applied for

18 cm and developed with CHCl3–(CH3)2CO (20:1) for

9 cm and eluted with CHCl3–MeOH (9:1)] to furnish a

diester, 1a (0.6 mg). Through a similar procedure, 1b

(0.5 mg) was prepared from 1 (0.5 mg) using (S)-MTPA

(28.2 mg), EDC (25.6 mg) and DMAP (11.3 mg).

(R)-MTPA diester, (1a): Amorphous powder; 1H-NMR

(CDCl3) d: 7.57–7.52 (4H, m, aromatic protons), 7.43–7.39

(6H, m, aromatic protons), 5.08 (1H, m, H-9), 5.02 (1H, m,

H-3), 3.56 (3H, br s, OMe), 3.53 (3H, br s, OMe), 3.17 (1H,

dd, J = 10, 10 Hz, H-4), 1.79 (1H, dd, J = 12, 4 Hz,

H-2eq), 1.64 (2H, m, H-8), 1.50 (1H, m, H-5), 1.49 (1H, m,

H-7a), 1.29 (3H, d, J = 6 Hz, H3-10), 1.27 (1H, dd,

J = 12, 12 Hz, H-2ax), 1.12 (1H, m, H-7b), 1.08 (3H, d,

J = 6 Hz, H3-13), 0.91 (3H, s, H3-12), 0.89 (3H, s, H3-11),

0.72 (1H, br d, J = 11 Hz, H-6); HR-ESI-TOF-MS (posi-

tive-ion mode) m/z: 685.2581 [M ? Na]? (Calcd for

C33H40O7F6Na: 685.2570).

(S)-MTPA diester, (1b): Amorphous powder; 1H-NMR



H-3), 3.57 (6H, br s, OMe), 3.09 (1H, dd, J = 10, 10 Hz,

H-4), 1.83 (1H, dd, J = 12, 4 Hz, H-2eq), 1.59 (2H, m,

H-8), 1.44 (1H, m, H-5), 1.36 (1H, m, H-7a), 1.34 (3H, d,

J = 6 Hz, H3-10), 1.32 (1H, dd, J = 12, 12 Hz, H-2ax),

1.04 (1H, m, H-7b), 0.99 (3H, d, J = 6 Hz, H3-13), 0.84

(3H, s, H3-12), 0.82 (3H, s, H3-11), 0.65 (1H, br d,

J = 11 Hz, H-6); HR-ESI-TOF-MS (positive-ion mode)

m/z: 685.2581 [M ? Na]? (Calcd for C33H40O7F6Na:

685.2570).


(2a and 2b) of 2

Through a similar procedure, (R)- and (S)-MTPA diesters,

2a (1.0 mg) and 2b (0.9 mg), were prepared from 2

(1.0 mg each) with the respective reagents, (R)- and

(S)-MTPA (13.6 and 12.6 mg), EDC (11.2 and 10.7 mg)

and DMAP (5.9 and 6.7 mg).

(R)-MTPA diester, (2a): Amorphous powder; 1H-NMR


(6H, m, aromatic protons), 5.74 (1H, d, J = 16 Hz, H-7),

5.69 (1H, dd, J = 16, 6 Hz, H-8), 5.61 (1H, qd, J = 6,

6 Hz, H-9), 5.12 (1H, ddd, J = 12, 10, 5 Hz, H-3), 3.61

J Nat Med (2012) 66:284–291 289

123

(1H, dd, J = 10, 10 Hz, H-4), 3.55 (3H, br s, OMe), 3.52

(3H, br s, OMe), 1.90 (1H, dq, J = 10, 7 Hz, H-5), 1.84

(1H, dd, J = 12, 12 Hz, H-2ax), 1.65 (1H, dd, J = 12,

5 Hz, H-2eq), 1.39 (3H, d, J = 6 Hz, H3-10), 1.08 (3H, s,

H3-11), 0.95 (3H, d, J = 7 Hz, H3-13), 0.812 (3H, s, H3-

12); HR-ESI-TOF-MS (positive-ion mode) m/z: 699.2362

[M ? Na]? (Calcd for C33H38O8F6Na: 699.2363).

(S)-MTPA diester, (2b): Amorphous powder; 1H-NMR


(6H, m, aromatic protons), 5.64–5.59 (2H, m, H-7 and 8),

5.62 (1H, qd, J = 6, 6 Hz, H-9), 5.10 (1H, ddd, J = 12, 10,

5 Hz, H-3), 3.54 (1H, dd, J = 10, 10 Hz, H-4), 3.58 (3H,

br s, OMe), 3.56 (3H, br s, OMe), 1.92 (1H, dd, J = 12,

12 Hz, H-2ax), 1.84 (1H, dq, J = 10, 7 Hz, H-5), 1.71 (1H,

dd, J = 12, 5 Hz, H-2eq), 1.43 (3H, d, J = 6 Hz, H3-10),

1.03 (3H, s, H3-11), 0.88 (3H, d, J = 7 Hz, H3-13), 0.811

(3H, s, H3-12); HR-ESI-TOF-MS (positive-ion mode)

m/z: 699.2368 [M ? Na]? (Calcd for C33H38O8F6Na:

699.2363).

Enzymatic hydrolysis of 3

Compound 3 (10.0 mg) was hydrolyzed with b-glucosi-

dase (5.5 mg) at 37�C in 1 mL of 20 mM acetate buffer

(pH 5.0) with reciprocal shaking for 12 h. The liberation

of glucose was monitored by TLC analysis (CHCl3:

MeOH:H2O, 15:6:1, Rf values, 3: 0.50, aglycone 1a: 0.76,

and glucose: 0.19). The reaction mixture was concentrated

and then subjected to preparative TLC (silica gel,

0.25 mm thickness, applied for 18 cm and developed with

CHCl3:MeOH:H2O, 15:6:1 for 9 cm and eluted with the

same solvent) to furnish an aglycone, 3a (5.5 mg). The

absolute configuration of the liberated glucose was

determined to be of the D-series from the positive optical

rotation sign and the retention time (8.7 min) on HPLC

analysis [JASCO OR-2090 Plus; Optical rotation detector,

Shodex Asahipak NH2P-50; U = 4.5 mm, L = 25 cm,

75% CH3CN aq., 1 mL/min] of the sugar-containing

fraction. Peak materials were identified by co-chroma-

tography with authentic D-glucose. The physicochemical

data for 3a including the optical rotation value were

essentially identical to those of 2. Finally, this identity

was confirmed by preparing MTPA diesters through a

similar procedure to that mentioned above, (R)- and (S)-

MTPA diesters, 3b (0.6 mg) and 3c (0.4 mg), being

prepared from 3a (0.7 mg each) with the respective

reagents, (R)- and (S)-MTPA (10.5 and 11.7 mg), EDC

(11.8 and 10.6 mg) and DMAP (11.1 and 10.3 mg).1H-NMR of 3b and 3c was essentially identical to that of

2a and 2b, respectively; HR-ESI-TOF-MS (positive-ion

mode) of 3b and 3c, m/z: 699.2361 and 699.2358

[M ? Na]?, respectively (Calcd for C33H38O8F6Na:

699.2363).

Pivaloylation of 4

A solution of 4 (2.4 mg) in 1.0 mL of dehydrated pyridine

was reacted with pivaloyl chloride (5 lL) on ice for 3.0 h

with stirring. After the addition of 0.5 mL of H2O, the

reaction mixture was concentrated under reduced pressure.

The residue was purified by preparative TLC (silica gel,

0.25 mm thickness, applied for 18 cm and developed with

CHCl3–MeOH (10:1) for 9 cm and eluted with CHCl3-

MeOH (2:1) to furnish a pivaloyl ester, 4a (1.8 mg). 4a:

Amorphous powder; 1H-NMR (CDCl3) d: 4.20 (1H, dd,

J = 11, 3 Hz, H-13a), 3.92 (1H, dd, J = 11, 6 Hz, H-13b),

3.79 (1H, dddd, J = 11, 11, 4, 4 Hz, H-3), 3.73 (1H, m,

H-9), 2.06 (1H, m, H-4 eq), 1.72 (1H, ddd, J = 12, 4,

2 Hz, H-2eq), 1.67 (1H, m, H-5), 1.60 (1H, m, H-7a), 1.46-

1.38 (2H, m, H2-8), 1.23 (9H, s, CH3), 1.18 (3H, d,

J = 6 Hz, H3-10), 1.13 (1H, dd, J = 12, 11 Hz, H-2ax),

1.12 (1H, ddd, J = 12, 12, 12 Hz, H-4ax), 1.09 (1H, m,

H-7b), 0.98 (3H, s, H3-12), 0.85 (1H, ddd, J = 11, 5, 3 Hz,

H-6), 0.84 (3H, s, H3-11); 13C-NMR (CDCl3) d: 178.6

(Me3CC = O–), 68.5 (C-9), 67.0 (C-13), 66.7 (C-3), 50.9

(C-2), 47.4 (C-6), 41.1 (C-8), 40.0 (C-4), 39.0

(Me3CC = O–), 38.6 (C-5), 35.9 (C-1), 30.6 (C-12), 27.3

(Me3CC = O–), 24.6 (C-7), 23.6 (C-10), 20.8 (C-11); HR-

ESI-TOF-MS (positive-ion mode) m/z: 337.2353

[M ? Na]? (Calcd for C18H34O4Na: 337.2349).


(4b and 4c) of 4a

Through a similar procedure to that described above, (R)-

and (S)-MTPA diesters, 4b (0.73 mg) and 4c (0.82 mg),

were prepared from 4a (0.9 mg each) with the respective

reagents, (R)- and (S)-MTPA (29.2 and 23.1 mg), EDC

(22.3 and 21.7 mg) and DMAP (8.1 and 10.5 mg).

(R)-MTPA diester, (4b): Amorphous powder; 1H-NMR



H-9), 4.16 (1H, dd, J = 11, 3 Hz, H-13a), 3.93 (1H, dd,

J = 11, 6 Hz, H-13b), 3.54 (3H, br s, OMe), 3.52 (3H, br s,

OMe), 2.15 (1H, m, H-4 eq), 1.76 (2H, m, H-2eq and H-5),

1.61 (2H, m, H-8), 1.56 (1H, m, H-7a), 1.32 (1H, ddd,

J = 12, 12, 12 Hz, H-4ax), 1.25 (1H, m, H-2ax), 1.20 (9H, s,

Me3CC = O–), 1.08 (1H, m, H-7b), 0.86 (3H, s, H3-11), 0.92

(3H, s, H3-12); HR-ESI-TOF-MS (positive-ion mode) m/z:

769.3152 [M ? Na]? (Calcd for C38H48O8F6Na: 769.3145).

(S)-MTPA diester, (4c): Amorphous powder; 1H-NMR



H-9), 4.12 (1H, dd, J = 11, 3 Hz, H-13a), 3.86 (1H, dd,

J = 11, 6 Hz, H-13b), 3.57 (3H, br s, OMe), 3.55 (3H, br s,

OMe), 2.06 (1H, m, H-4 eq), 1.79 (1H, m, H-2 eq), 1.72

(1H, m, H-5), 1.56 (2H, m, H-8), 1.42 (1H, m, H-7a), 1.33

290 J Nat Med (2012) 66:284–291

123

(3H, d, J = 6 Hz, H3-10), 1.32 (1H, m, H-2ax), 1.19 (1H,

m, H-4ax), 1.18 (9H, s, Me3CC = O–), 1.00 (1H, m, H-7b),

0.84 (3H, s, H3-12), 0.79 (3H, s, H3-11), 0.79 (1H, m, H-6);

HR-ESI-TOF-MS (positive-ion mode) m/z: 769.3149

[M ? Na]? (Calcd for C38H48O8F6Na: 769.3145).

DPPH radical-scavenging assay

The reagents, (S)-(-)-6-hydroxy-2,5,7,8-tetramethylchro-

man-2-carboxylic acid (Trolox) and 2,2-diphenyl-1-picryl-

hydrazyl (DPPH), were purchased from Aldrich Chemical

Co., and the DPPH radical-scavenging activities of the

isolated compounds were examined according to the

method previously described [26].

Human cancer cell growth inhibition assay

Growth inhibitory activities were determined using human

promyelocytic leukemia cells (HL-60) and human small cell

lung cancer cells (SBC-3) by the 3-[4,5-dimethylthiazol-

2-yl]-2,5-diphenyltetrazolium bromide (MTT) method [27].

Acknowledgments The authors are grateful for the access to the

superconducting NMR instrument (JEOL ECA-600) and the Applied

Biosystem QSTAR XL system ESI (NanoSpray) mass spectrometer at

the Natural Science Center for Basic Research and Development

(N-BARD), Hiroshima University. This work was supported in part

by Grants-in-Aid from the Ministry of Education, Culture, Sports,

Science and Technology of Japan (Nos. 22590006 and 23590130),

and the Ministry of Health, Labour and Welfare. Thanks are also due

to the Research Foundation for Pharmaceutical Sciences and the

Takeda Science Foundation for their financial support.

References

1. Lannuzel A, Hoglinger GU, Champy P, Michel PP, Hirsch EC,

Ruberg M (2006) Is atypical parkinsonism in the Caribbean

caused by the consumption of Annonaceae? J Neural Transm

Suppl 70:153–157

2. Shitamoto J, Matsunami K, Otsuka H, Shinzato T, Takeda Y

(2010) Elaeocarpionoside, a megastigmane glucoside from the

leaves of Elaeocarpus japonicus Sieb. et Zucc. J Nat Med

64:104–108

3. Liu YP, Cai XH, Feng T, Li Y, Li XN, Luo XD (2011) Triterpene

and sterol derivatives from the roots of Breynia fruticosa. J Nat

Prod 74:1161–1168

4. Kusumi T, Ooi T, OhkuboY YabuuchiT (2006) The modified

Mosher’s method and the sulfoximine method. Bull Chem Soc

Jpn 79:965–980

5. Kasai R, Suzuno M, Asakawa J, Tanaka O (1977) Carbon-13

chemical shifts of isoprenoid-b-D-glucopyranosides and -b-D-

mannopyranosides. Stereochemical influences of aglycone alco-

hols. Tetrahedron Lett 18:175–178

6. Sueyoshi E, Liu H, Matsunami K, Otsuka H, Shinzato T,

Aramoto M, Takeda Y (2006) Bridelionosides A–F: megastig-

mane glucosides from Bridelia glauca f. balansae. Phytochem-

istry 67:2483–2493

7. Pasi S, Aligiannis N, Pratsinis H, Skaltsounis AL, Chinou LB

(2009) Rhusonoside A, a new megastigmane glycoside from

Rhus sylvestris, increases the function of osteoblastic MC3T3-E1

cells. Planta Med 75:163–167

8. Xu Y, Feng T, Cai XH, Luo XD (2009) A new C13-norisoprenoid

from leaves of Alstonia scholaris. Zhongguo Tianran Yaowu

7:21–23

9. Zhang Z, Zhen W, Ji YP, Zhao Y, Wang CG, Hu JF (2010)

Gynostemosides A–E, megastigmane glycosides from Gynos-temma pentaphyllum. Phytochemistry 71:693–700

10. Mei W, Dai H, Wu D (2006) Isolation and identification of

chemical constituents from Cephalomappa sinensis. Zhongguo

Yaowu Huaxue Zazhi 16:240–243

11. Perez C, Trujillo JM, Almonacid LN, Navarro E, Alonso SJ

(1996) New C13-norisoprenoids from Apollonias barbujana. Nat

Prod Lett 8:1–6

12. Kuang HX, Yang BY, Xia YG, Feng WS (2008) Chemical

constituents from the flowers of Datura metel L. Arch Pharm Res

31:1094–1097

13. Otsuka H, Yao M, Kamada K, Takeda Y (1995) Alangionosides

G–M: glycosides of megastigmane derivatives from the leaves of

Alangium premnifolium. Chem Pharm Bull 43:754–759

14. Kawakami S, Matsunami K, Otsuka H, Shinzato T, Takeda Y (2011)

Crotonionosides A–G: megastigmane glycosides from leaves of

Croton cascarilloides Raeuschel. Phytochemistry 72:147–153

15. Yu Q, Otsuka H, Hirata E, Shinzato T, Takeda Y (2002) Tur-

pinionosides A–E: megastigmane glucosides from leaves of

Turpinia ternata Nakai. Chem Pharm Bull 50:640–644

16. Umehara K, Hattori I, Miyase T, Ueno A, Hara S, Kageyama C

(1988) Studies on the constituents of leaves of Citrus unshiuMarcov. Chem Pharm Bull 36:5004–5008

17. Matsunami K, Otsuka H, Takeda Y (2010) Structural revisions of

blumenol C glucoside and byzantionoside B. Chem Pharm Bull

58:438–441

18. Mori K, Khlebnikov V (1993) Carotenoids and degraded

carotenoids, VIII – Synthesis of (?)-dihydroactinidiolide, (?)-

and (-)-actinidiolide, (?)- and (-)-loliolide as well as (?)- and

(-)-epiloliolide. Liebigs Ann Chem 1993: 77–82

19. Kim, Lee SK, Kim CS, Kim KS, Moon DC (2004) Phytochemical

constituents of Carpesium macrocephalum F(R). et S(AV). Arch

Pharm Res 27:1029–1033

20. Someya Y, Kobayashi A, Kubota K (2001) Isolation and identi-

fication of trans-2- and trans-3-hydroxy-1,8-cineole glucosides

from Alpinia galanga. Biosci Biotechnol Biochem 65:950–953

21. Mizutani K, Yuda M, Tanaka O, Saruwatari Y, Fuwa T, Jia MR,

Ling YK, Pu XF (1988) Chemical studies on the Chinese tradi-

tional medicine, dangshen. I. Isolation of (Z)-3- and (E)-2-hexe-

nyl b-D-glucosides. Chem Pharm Bull 36:2689–2690

22. Ly TN, Shimoyamada M, Yamauchi R (2006) Isolation and

characterization of rosmarinic acid oligomers in Celastrus hindsiiBenth leaves and their antioxidative activity. J Agric Food Chem

54:3786–3793

23. Beck MA, Haberlein H (1999) Flavonol glycosides from Esc-hscholtzia californica. Phytochemistry 50:329–332

24. Brasseur T, Angenot L (1986) Flavonol glycosides from leaves of

Strychnos variabilis. Phytochemistry 25:563–564

25. Fukui Y, Tanaka Y, Kusumi T, Iwashita T, Nomoto K (2003)

A rationale for the shift in colour towards blue in transgenic

carnation flowers expressing the flavonoid 30,50-hydroxylase

gene. Phytochemistry 63:15–23

26. Matsunami K, Takamori I, Shinzato T, Aramoto M, Kondo K,

Otsuka H, Takeda Y (2006) Radical-scavenging activities of new

megastigmane glucosides from Macaranga tanarius (L.) Mull.

-Arg. Chem Pharm Bull 54:1403–1407

27. Yen CT, Wu CC, Lee JC, Chen SL, Morris NSL, Hsieh PW, Wu

YC (2010) Cytotoxic N-(fluorenyl-9-methoxycarbonyl) (Fmoc)-

dipeptides: Structure–activity relationships and synergistic stud-

ies. Eur J Med Chem 45:2494–2502

J Nat Med (2012) 66:284–291 291

123

at SciVerse ScienceDirect

Tetrahedron 67 (2011) 9358e9367

Contents lists available

Tetrahedron

journal homepage: www.elsevier .com/locate/ tet

Heteroannelated (þ)-muricatacin mimics: synthesis, antiproliferative propertiesand structureeactivity relationships

Bojana Sre�co a, Goran Benedekovi�c a, Mirjana Popsavin a, Pavle Had�zi�c b, Vesna Koji�c c,Gordana Bogdanovi�c c, Vladimir Divjakovi�c d, Velimir Popsavin a,*

aDepartment of Chemistry, Biochemistry and Environmental Protection, Faculty of Sciences, University of Novi Sad, Trg Dositeja Obradovi�ca 3, 21000 Novi Sad, SerbiabGo�sa Institute, Milana Raki�ca 35, 11000 Belgrade, SerbiacOncology Institute of Vojvodina, Institutski put 4, 21204 Sremska Kamenica, SerbiadDepartment of Physics, Faculty of Sciences, University of Novi Sad, Trg Dositeja Obradovi�ca 4, 21000 Novi Sad, Serbia

a r t i c l e i n f o

Article history:Received 20 July 2011Received in revised form 6 September 2011Accepted 26 September 2011Available online 5 October 2011

Keywords:MuricatacinAnnonaceous acetogeninsHeteroannelated muricatacin mimicsIsostereAntitumour activitySAR

* Corresponding author. Tel.: þ381 21 485 27 68; faddress: [email protected] (V. Popsavin).

0040-4020/$ e see front matter � 2011 Elsevier Ltd.doi:10.1016/j.tet.2011.09.132

a b s t r a c t

Six new (þ)-muricatacin mimics bearing a furano-furanone core have been synthesized and their in vitroantiproliferative activity was evaluated against a panel of human tumour cell lines. A straightforwardtotal synthesis of (þ)-muricatacin (1) from D-xylose is disclosed providing a sample of 1 that served asa positive control in antitumour assays. All new compounds showed diverse antiproliferative effectsagainst human malignant cell lines, but were devoid of any significant cytotoxicity towards the normalfoetal lung fibroblasts (MRC-5). Additionally, the most of (þ)-muricatacin analogues show selective cy-totoxicities towards certain cancer cell lines, whereas only two of six analogues are broadly toxic againstall cell lines under evaluation. A SAR study reveals the structural features that may be beneficial for theantiproliferative activity of these lactones. These include the absolute stereochemistry, introduction ofa THF ring, interchange of the O8 ether functionality and the C8 methylene group in the side chain ofmuricatacin oxa analogues, as well as the one- or two-carbon homologation of the side chain in both3 and 6.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Muricatacin, a naturally occurring acetogenin derivative hasbeen isolated by McLaughlin and co-workers from the seeds ofAnnona muricata.1 The isolated material was found to be a mixtureof (þ)-(4S,5S)-5-hydroxyheptadecan-4-olide (1, Fig. 1) and its(�)-(4R,5R)-enantiomer (ent-1), with a slight predominance of thelater. Both (þ)- and (�)-muricatacin have demonstrated a remark-able antiproliferative activity towards several human tumour celllines. These findings have stimulated a significant interest in thesynthesis of this type of compounds. Accordingly, many synthesesof (þ)- and (�)-muricatacin and congeners from various precursorshave been reported.2e4 In addition, thesemolecules have been usedas starting materials for the synthesis of other complex biologicallyrelevant natural products.5 A number of muricatacin analogues andstereoisomers have also been synthesized,6 and some of themwereevaluated for their antitumour activity.6h,7,8

Previous studies in our laboratory revealed that a mimic of(�)-muricatacin in which a methylene group from the side chain

Fig. 1. Structures of (þ)-muricatacin (1), (�)-muricatacin (ent-1) and the relatedanalogues.

ax: þ381 21 454 065; e-mail

All rights reserved.


www.sciencedirect.com/science/journal/00404020

http://www.elsevier.com/locate/tet

http://dx.doi.org/10.1016/j.tet.2011.09.132



B. Sre�co et al. / Tetrahedron 67 (2011) 9358e9367 9359

has been replaced by an ether function (compound ent-2) exhibitedin vitro antitumour activity against several human cancer celllines.9 We have also found that introduction of a THF ring increasesthe activity of heteroannelated (�)-muricatacin mimic ent-3against the HeLa malignant cells, and that one-carbon homologa-tion of the side chain in ent-3 increases the activity of resultinghomologue ent-4 against the K562, HL-60 and HeLa cell lines.10

As part of our continuing efforts to further optimize the anti-tumour potency of leads of type ent-1, ent-2, ent-3 and ent-4, wehave investigated the corresponding opposite enantiomers, asconformationally constrained analogues of (þ)-muricatacin (1).The target compounds 3e8 were designed to restrict the rotationaround the C4eC5 and C5eC6 bonds through incorporation ofa condensed THF ring as shown in Fig. 1, providing a number ofheteroannelated (þ)-muricatacin mimics. Thus, compounds 3 and4 are the opposite enantiomers of ent-3 and ent-4, while the mol-ecule 5 represents a one-carbon higher homologue of 4. Analogues6e8 represent classical isosteres of 3e5 designed by replacement ofan ether function from the side chain with a methylene group.Analogues 3 and 6 represent one-carbon lower homologues of 4and 7, while lactones 5 and 8 represent one-carbon higher homo-logues of 4 and 7. In addition to the synthesis of analogues 3e8,a novel route to (þ)-muricatacin (1) was developed in order toprovide a sample of the lead that would serve as a positive controlin antitumour assays.

2. Results and discussion

2.1. Chemical synthesis

The synthesis of (þ)-muricatacin (1) is shown in Scheme 1. Thestarting hydroxy lactone 9 was prepared from D-xylose in eightsteps as reported earlier by us.9

Scheme 1. Reagents and conditions: (a) eight steps, 23.6%, Ref. 9; (b) I2, ImdH, Ph3P,MeCN, N2, 90 �C, 1.5 h, 96%; (c) dodec-1-ene, Grubbs catalyst II generation, CH2Cl2, rt,27.5 h, 61%; (d) H2, 10% Pd/C, MeOH, rt, 4 h, 90%.

Scheme 2. Reagents and conditions: (a) two steps, 50%, Ref. 12; (b) Meldrum’s acid,Et3N, DMF, 46e48 �C, 48 h, 82%; (c) H2, 10% Pd/C (0.1 equiv of Pd), abs EtOH, rt,105 min, 87%; (d) C9H19Br, Ag2O, AgOTf, Et2O, reflux, 30 h, 61%; (e) H2, 10% Pd/C, EtOH,rt, 18 h for 16, 96% of 3, 20 h for 17, 86% of 4, 3.5 h for 18, 83% of 5; (f) C10H21Br, Ag2O,AgOTf, Et2O, reflux, 31 h, 51%; (g) C11H23Br, Ag2O, AgOTf, Et2O, reflux, 48 h, 56%.

Reaction of 9 with iodine, triphenylphosphine and imidazole,according to the methodology developed by Garegg and Samuels-son,11 gave the corresponding terminal alkene 10 in 96% yield. Thecross metathesis reaction of 10 with dodec-1-ene (5 mol equiv) inthe presence of Grubbs second generation catalyst (10 mol %)afforded the corresponding disubstituted olefin 11 in 61% yield withexclusively E-selectivity (J6,7¼15.5 Hz). Catalytic hydrogenation of11 over 10% Pd/C inmethanol gave (þ)-muricatacin (1) in 90% yield.The physical data of thus obtained product 1 {mp 68e69 �C, [a]D

þ22.3 (c 0.43, CHCl3)}, were found to be in reasonable agreementwith those previously reported {lit.3 mp 68e70 �C, [a]D þ22.4(c 0.42, CHCl3); lit.4 mp 68e70 �C, [a]D þ19.6 (c 1.0, CHCl3)}.

Syntheses of conformationally restricted (þ)-muricatacin oxaanalogues 3e5 are outlined in Scheme 2.

The sequence started with the preparation of the protected pri-mary alcohol 15 from the known12

D-xylose derivative 13. Com-pound13was treatedwithMeldrum’s acid inDMF, in thepresenceofEt3N,whereupon theprotected lactone14wasobtained in82%yield.Compound 14 was previously prepared in our laboratory by Z-se-lectiveWittig olefination of lactol 13with Ph3P]CHCO2Me but onlyin 61%yield. Catalytic reduction of14 over 10%Pd/C (0.1 mol equivofPd) for 105 min at room temperature effected selective removal ofthe benzyl group from the primary position to afford the alcohol 15in 87% yield.13 Alcohol 15 readily reacted with an excess of nonylbromide and silver oxide in ether, in the presence of a catalyticamount of silver triflate, to give the corresponding 7-O-nonyl de-rivative 16 in 61% yield. The 5-O-benzyl protecting group from 16was removedbycatalytic hydrogenolysis over10%Pd/C inmethanol,to give a 96% yield of 3. Treatment of 15 with decyl bromide, underthe reaction conditions similar to those applied for the preparationof 16, gave the expected 7-O-decyl derivative 17 in 51% yield. In-termediate 17 was earlier synthesized in our laboratory in 18%overall yield from starting compound 12.14 This new synthesis of 4proceeds in five steps with 24% overall yield based on the samestarting compound. Hydrogenolytic removal of the benzyl etherprotective group in 17 under the same reaction conditions as re-portedearlier byus,14 afforded 4 in 86% yield.Moreover, byusing theundecyl bromide as an alkylation agent, compound 15 was firstconverted to the protected lactone 18 (56%) and finally to target 5(83%), after removal of the benzyl protecting group.

B. Sre�co et al. / Tetrahedron 67 (2011) 9358e93679360

Syntheses of (þ)-muricatacin mimics 6 and 7 are presented inScheme 3.

Fig. 2. ORTEP drawing of (a) analogue 7 and (b) analogue 8.

Scheme 3. Reagents and conditions: (a) three steps, 56%, Ref. 15a; (b) Meldrum’s acid,Et3N, DMF, 46e48 �C, 70 h, 43%; (c) I2, ImdH, Ph3P, MeCN, N2, 90 �C, 1.5 h, 93%; (d)undec-1-ene, Grubbs catalyst II generation, CH2Cl2, rt, 24 h, 68%; (e) dodec-1-ene,Grubbs catalyst II generation, CH2Cl2, Ar, rt, 68 h, 69%; (f) H2, 10% Pd/C, MeOH, rt, 3 h for22, 82% of 6, 4.5 h for 23, 57% of 7.

Scheme 4. Reagents and conditions: (a) four steps, 40%, Ref. 17; (b) C12H25MgBr, THF,reflux, 4 h, 56%; (c) 70% aq AcOH, reflux, 3.5 h, 57%; (d) Meldrum’s acid, Et3N, DMF,46 �C, 48 h, 66%.

3-O-Benzyl-D-glucose (19), readily available from D-glucose,15

was used as a convenient starting compound in this part of thework. Thus, compound 19 was allowed to react with Meldrum’sacid in DMF, in the presence of Et3N, to afford the expected lactone20 in 43% yield. Reaction of 20 with iodine, imidazole and triphe-nylphosphine gave the corresponding terminal olefin 21 in 93%yield. Preparation of (�)-21 has been recently disclosed by Kapit�anand Gracza.16 The authors have described this product as a colour-less oil, but we have isolated it in the form of colourless needles, mp62e63 �C. However, the 1H and 13C NMR data of thus obtainedfuranolactone 21 were in full agreement with reported values. Thecross metathesis between 21 and undec-1-ene or dodec-1-ene, inpresence of Grubbs second generation catalyst (10 mol %) producedthe desired long-chain olefins 22 (68%) and 23 (69%), respectively,both with exclusively E-selectivity (J7,8¼15.5 Hz). In the final step,the hydrogenation of double bond simultaneously with the re-moval of the benzyl ether protection in 22 and 23 were carried outsmoothly by hydrogenation over 10% Pd/C in MeOH, to afford thetarget molecules 6 (82%) and 7 (57%) as white solids.

The 3,5-anhydro-D-xylose derivative 24, readily available fromD-xylose,17 was used as a convenient starting material for the syn-thesis of target 8 (Scheme 4).

Treatment of 24with dodecylmagnesium bromide produced theexpected alcohol 25 in 56% yield. Hydrolytic removal of the cyclo-hexylidene protective group (7:3 H2O/AcOH) gave the expectedlactol 26 (57%), which was finally converted to the furanolactone 8(66%) after treatment with Meldrum’s acid and Et3N in DMF.

2.2. X-ray analysis

The complete structure and relative stereochemistry of 7 and 8were established by X-ray diffraction analysis.18 The absolute con-figurationwas determined by the use of D-xylose or D-glucose as theenantiomerically pure starting materials. A view of the molecularstructures is provided in Fig. 2a and b. Additional details of crys-tallographic data are given in Supplementary data.

2.3. In vitro antitumour activity

The human cancer cell lines used in this study represent severalcommon types of solid cancer and leukaemia. These include themyelogenous leukaemia (K562), promyelocytic leukaemia (HL-60),Jurkat T cell leukaemia, Burkitt’s lymphoma (Raji), colon carcinoma(HT-29), oestrogen receptor negative breast carcinoma (MDA-MB-231), and cervix carcinoma (HeLa) cells. The use of human foetallung fibroblasts (MRC-5) serves to evaluate the toxicity of the an-alogues towards normal cells. Cytotoxic activity was evaluated byusing the standard MTT assay, after exposure of cells to the testedcompounds for 72 h. (þ)-Muricatacin (1), the corresponding 7-oxaanalogue 2,19 and the commercial antitumour agent doxorubicin(DOX) were used as reference compounds.

According to the resulting IC50 values of the cytotoxic assay(Table 1), all synthesized compounds demonstrated diverse anti-proliferative effects against human malignant cell lines but weredevoid of any significant cytotoxicity towards the normal foetallung fibroblasts (MRC-5). Additionally, the most of (þ)-muricatacinanalogues show selective cytotoxicities towards certain cancer cell

Table 1In vitro cytotoxicity of (þ)-muricatacin (1), analogues 3e8 and DOX

Compd IC50a (mM)

K562 HL-60 Jurkat Raji HT-29 MDA-MB-231 HeLa MRC-5

1 0.03 0.06 0.14 1.32 >100 >100 1.09 >1002 1.01 17.36 1.16 >100 0.02 >100 >100 >1003 1.95 >100 25.46 13.25 1.69 >100 9.17 >1004 3.91 0.06 >100 0.99 >100 45.32 >100 >1005 7.71 1.64 2.03 5.33 35.71 5.58 10.19 >1006 1.55 5.31 2.54 0.95 >100 >100 1.01 >1007 2.02 4.44 3.87 1.01 >100 >100 0.22 >1008 0.36 0.89 0.05 1.23 15.45 3.56 2.24 >100DOX 0.25 0.92 0.03 2.98 0.15 0.09 0.07 0.10

a IC50 is the concentration of compound required to inhibit the cell growth by 50% compared to an untreated control. Values are means of three independent experiments.Coefficients of variation were less than 10%.


lines, whereas only 5 and 8 are broadly toxic against all cell linesunder evaluation.

Remarkably all analogues 3e8, as well as both parent com-pounds 1 and 2 exhibit potent in vitro anticancer activity towardsthe K562 cell line, with IC50 values in the low-micromolar range.The most active compound against this cell line was lead 1(IC50¼0.03 mM), being 8-fold more potent than the commercialantitumour agent doxorubicin. Only analogue 8 exhibited a sub-micromolar antiproliferative activity (IC50¼0.36 mM) against theK562 cells, being essentially as potent as the commercial anti-tumour agent doxorubicin.

Most of synthesized analogues demonstrated potent cytotoxic-ities against HL-60 cell line. The only exception is the analogue 3,which was inactive towards HL-60 cells, as well as 2 that showeda moderate cytotoxicity (IC50¼17.36 mM) in the same cell line. The7-O-decyl derivative 4 is the most potent cytotoxic agent in this cellline, showing exactly the same antiproliferative activity as thenatural product 1 (IC50¼0.06 mM).

Analogues 5e8 including parent compounds 1 and 2 exhibitedstrong antiproliferative activities against the Jurkat cells with IC50values in the range of 0.05e3.87 mM. Compound 4 was inactivetowards these cells, while analogue 3 showed a moderate cyto-toxicity (IC50¼25.46 mM) in the same cell line. The most activemolecule in the culture of Jurkat cells is analogue 8 that exhibitedover 2- and 20-fold higher potency than both control compounds 1and 2, respectively. In the same time, this molecule showed a sim-ilar activity as DOX in the same cell line.

Parent compound 2 was inactive against Raji cells. However,analogues 4, 6, 7 and 8 exhibited notable cytotoxic effects towardsthis cell line with IC50 values in the range of 0.95e1.23 mM. Thesemolecules are the most active compounds against the Raji cells,being essentially 2e3-fold more active than the commercial anti-tumour agent doxorubicin.

HT-29 cell line appears to be much less sensitive to the syn-thesized (þ)-muricatacin analogues, including the parent com-pound 1 that was inactive against these cells. However, 7-oxaanalogue 2 demonstrated the most potent activity against the HT-29 cells (IC50¼0.02 mM) being 7.5-fold more cytotoxic than thestandard antitumour agent doxorubicin.

MDA-MB-231 cells are even less sensitive to the synthesized(þ)-muricatacin analogues. Both parent molecules 1 and 2, as wellas analogues 3, 6 and 8 were inactive towards this cell line, whileanalogue 4 showed a weak cytotoxicity (IC50¼45.32 mM) in thesame cell line. The most active molecule in the culture of MDA-MB-231 cells is analogue 8 (IC50¼3.56 mM).

The majority of synthesized analogues exhibited notable anti-proliferative effects on HeLa cells, with IC50 values in the range of0.22e10.19 mM. Themost active compound against this cell linewas8 (IC50¼0.22 mM), being approximately 5-fold more potent than the

control compound 1. Lead 2 and analogue 4 were inactive againstthese cells.

2.4. SAR studies

In an early attempt to correlate the structures of muricatacinand congeners with their cytotoxic activities against KB and VEROcell lines, Figad�ere and co-workers7 have found that cytotoxicitywas dependent on the length of the alkyl chain. A shorter chaindramatically decreased the activity, whereas a longer chain did notinfluence to the activity. Introduction of unsaturation in the lactonering improved the activity of the analogues with a short chain buthad no effect onmuricatacin itself. When other functionalities werepresent, such as an oxo function, the activity was about the same asfor the parent compound. Addition of a tetrahydrofuran ring did notchange the activity of the analogues as long as the length of thealkyl chain was not changed. In the cases of the pyrrolidones (aza-muricatacins), the activity was either identical to that of mur-icatacin or even better, independent of the relative or absolutestereochemistry of the analogues.

Our previous findings9 observed with (�)-muricatacin de-rivatives have prompted us to execute a more comprehensive SARinvestigation of the analogues of both (þ)- and (�)-muricatacinseries. The first structural element considered was the absolutestereochemistry of analogues. The importance of this structuralfeature for the cytotoxic activities of these compounds was studiedby comparing the IC50 values of (þ)- (1) and (�)-muricatacin (ent-1), as well as of 3 pairs of analogues (2 and ent-2, 3 and ent-3, 4 andent-4), each of which contains exactly the same substituents anddiffers only in their absolute stereochemistry. As shown in Fig. 3a,the results indicate that, in most cases, the (�)-muricatacin mimicsshow a more potent cytotoxicity than the opposite enantiomers of(þ)-muricatacin series. A comparison of biological data of 1with 6,7, and 8 (Fig. 3b), revealed that introduction of a THF ring may onlyslightly affect the cytotoxic activities of the corresponding(þ)-muricatacin mimics (6e8). Interestingly, the effect is signifi-cantly more pronounced if the side chain of (þ)-muricatacin isextended for two carbon atoms. However, it is difficult to evaluateany trend conclusively in this case, as only three pairs of analogueswere available for comparison. As shown in Fig. 3c, hetero-annelation of lead 1 followed by replacement of the C7 atom withan ether group significantly decreases the activities of (þ)-mur-icatacin mimics (3e5). However, the same structural changes inlead ent-1 increases the cytotoxicities of the resulting (�)-mur-icatacin mimics (ent-3 and ent-4). As observed in (þ)-muricatacinseries the effect is significantly more pronounced in the analoguewith the extended side chain for one carbon atom (ent-4). As shownin Fig. 3d, interchange of the O8 ether functionality and C8 meth-ylene groups in the side chain of muricatacin oxa analogues results

Fig. 3. Contributions of selected structural features to the cytotoxic activities: (a) influence of absolute stereochemistry, (b) influence of an additional THF ring, (c) influence ofintroduction of a THF ring followed by CH2/O isosteric replacement, (d) influence of O/CH2 interchange, (e) influence of one- or two-carbon homologation of the side chain. TheIC50 values of two structures differing in one given position were compared, and the Δlog IC50 was subsequently calculated (Δlog IC50 is a difference between the IC50 values of ananalogue and the corresponding control compound). Positive Δlog IC50 values indicate a decrease of cytotoxic activity, whereas negative values show an increase in the activity uponthe structural modification being considered.


in a substantial increase in cytotoxic activity against most of the celllines tested in this study. The most pronounced activities wereobserved with analogues 2, 3 and 5. These findings indicated that-introduction of an oxygen atom in the side chain increases theantiproliferative activity of the analogues against most tumour celllines under evaluation. The final structural element considered wasthe length of the alkyl chain (Fig. 3e). Thus, one-or two-carbonhomologation of the side chain in 3 (to give analogues 4 and 5),as well as in 6 (to afford analogues 7 and 8) increases the activity ofthe resulting homologues against most of the cell lines tested.These results are in good agreement with previous findings that thelength of the side chain is crucial for antitumour activity of mur-icatacin analogues.7 However, the mechanism of action is still notunderstood. Since it has been proposed that acetogenins of Anno-naceae act as an inhibitor of complex I in the mitochondrial re-spiratory system,20 it is possible that muricatacin and analogues actvia an identical mechanism. The observed differences in the anti-proliferative potencies in respect to the cancer cell lines used maythen be explained by a specificity difference in the hosts’

mitochondrial complex I (NADH-ubiquinone oxidoreductase), butmore work is needed before any conclusion can be made.

3. Conclusion

In conclusion, we have developed a straightforward synthesis ofantitumour acetogenin derivative (þ)-muricatacin, as well as a di-vergent route to several new (þ)-muricatacin mimics (3e8) andevaluated them for in vitro cytotoxic activities against seven humantumour cell lines. All synthesized compounds demonstrated di-verse antiproliferative effects against human malignant cell linesbut were devoid of any significant cytotoxicity towards the normalfoetal lung fibroblasts (MRC-5). Additionally, the most of(þ)-muricatacin analogues show selective cytotoxicities towardscertain cancer cell lines, whereas only two of six are broadly toxicagainst all cell lines under evaluation. A SAR study reveals that thefollowing structural features are beneficial for the antiproliferativeactivity of these lactones: the absolute stereochemistry, presence ofan additional tetrahydrofuran ring, interchange of the O8 ether


functionality and C8 methylene groups in the side chain of mur-icatacin oxa analogues, as well as the one- and two-carbon ho-mologation of the side chain in both 3 and 6. It was found thatcytotoxicity is dependent on the length of the alkyl chain, wherebythe C2 homologation has the most pronounced effects on inhibitionof cells growth. The results of MTTassay alongwith the SAR analysisenabled us to identify the (þ)-muricatacin mimic 8 as the mostpromising antitumour agent, since it exhibited a potent anti-proliferative activity against all malignant cell lines under evalua-tion, but was completely inactive against the normal human cells(MRC-5). Hence, we believe that this approach may be of use in thesearch for new, more potent and selective anticancer agents de-rived from the natural product 1.

4. Experimental section

4.1. Chemistry

4.1.1. General methods. Melting points were determined on a B€uchi510 apparatus and were not corrected. Optical rotations were mea-sured on P 3002 (Kr€uss) and Autopol IV (Rudolph Research) polar-imeters at 24 �C. NMR spectra were recorded on a Bruker AC 250 Einstrument and chemical shifts are expressed in parts per milliondownfield from TMS. IR spectra were recorded with an FTIR Nexus670 spectrophotometer (Thermo-Nicolet). Low resolution massspectra (CI) were recorded on Finnigan-MAT 8230 and on an AgilentTechnologies HPLC/MS 3Q system (ESI), series 1200/6410. Highresolution mass spectra (ESI) of synthesized compounds were ac-quired on an Agilent Technologies 1200 series instrument equippedwith Zorbax Eclipse Plus C18 (100 mm�2.1 mm i.d. 1.8 mm) columnand DAD detector (190e450 nm) in combination with a 6210 time-of-flight LC/MS instrument (ESI) in the positive ion mode. Flashcolumn chromatography was performed using Kieselgel 60(0.040e0.063, E. Merck). All organic extracts were dried with an-hydrous Na2SO4. Organic solutions were concentrated in a rotaryevaporator under reduced pressure at a bath temperature below35 �C.

4.1.2. 5-O-Benzyl-2,3,6,7-tetradeoxy-L-threo-hept-6-enono-1,4-lactone (10). To a mixture of iodine (0.98 g, 3.84 mmol), imidazole(1.08 g, 15.94 mmol) and triphenylphosphine (1 g, 3.79 mmol) indry MeCN (6 mL) was added a solution of 9 (0.25 g, 0.95 mmol) indry MeCN (6 mL). The mixture was vigorously stirred at 90 �C (bathtemperature) for 1.5 h, in an atmosphere of N2, then evaporated andpurified by flash column chromatography (9:1/4:1/1:1 lightpetroleum/EtOAc), to yield unsaturated lactone 10 (0.21 g, 96%) asa pale yellow oil, [a]D þ51.5 (c 2.04, CHCl3), Rf¼0.63 (1:1 light pe-troleum/EtOAc). IR (neat): nmax 1777 (C]O). 1H NMR (250 MHz,CDCl3): d 1.96e2.26 (m, 2H, H-3), 2.33e2.64 (m, 2H, H-2), 3.84 (dd,1H, J4,5¼4.5 Hz, J5,6¼7.8 Hz, H-5), 4.39 (d, 1H, Jgem¼12.0 Hz, PhCH2),4.53 (ddd, 1H, J3,4¼6.1 Hz, J3,4¼7.9 Hz, J4,5¼4.5 Hz, H-4), 4.66 (d, 1H,Jgem¼12.0 Hz, PhCH2), 5.38 (d, 1H, J6,7a¼18.5 Hz, H-7a), 5.40 (d, 1H,J6,7b¼10.5 Hz, H-7b), 5.82 (m, 1H, H-6), 7.22e7.41 (m, 5H, Ph). 13CNMR (62.9 MHz, CDCl3): d 23.6 (C-2), 28.0 (C-3), 70.2 (PhCH2), 81.2and 81.24 (C-4 and C-5), 120.4 (C-7), 127.5, 128.2, 137.6 (Ph), 133.4(C-6), 177.2 (C-1). HRMS (ESI): m/z 233.1165 (MþþH), calcd forC14H17O3: 233.1172.

4.1.3. (E)-5-O-Benzyl-7-C-decyl-2,3,6,7-tetradeoxy-L-threo-hept-6-enono-1,4-lactone (11). To a stirred solution of olefin 10 (0.12 g,0.50 mmol) and dodec-1-ene (1.1 mL, 5.0 mmol) in dry CH2Cl2(2.3 mL) was added the second generation Grubbs catalyst (42 mg,0.05 mmol). The reaction mixture was stirred in an argon atmo-sphere for 27.5 h at room temperature. The solvent was evaporatedand the remaining crude product was purified by flash columnchromatography (19:1/9:1/4:1 light petroleum/EtOAc), to give 11

(0.12 g, 61%) as a bright yellow syrup, [a]D þ58.2 (c 1.03, CHCl3),Rf¼0.19 (19:1 light petroleum/EtOAc). IR (neat): nmax 1779 (C]O). 1HNMR (250 MHz, CDCl3): d 0.89 (t, 3H, J¼6.8 Hz, Me), 1.13e1.48 (m,16H, 8�CH2), 1.97e2.28 (m, 4H, CH2e3, CH2e8), 2.35e2.66 (m, 2H,CH2e2), 3.81 (dd, 1H, J4,5¼4.6 Hz, J5,6¼8.4 Hz, H-5), 4.38 and 4.66(2�d, 2H, Jgem¼12.0 Hz, PhCH2), 4.54 (m,1H, J3,4¼7.8 Hz, J3,4¼5.8 Hz,J4,5¼4.6 Hz, H-4), 5.44 (m, 1H, J5,6¼8.4 Hz, J6,7¼15.5 Hz, J6,8¼1.3 Hz,H-6), 5.78 (dt, 1H, J6,7¼15.5 Hz, J7,8¼6.6 Hz, H-7), 7.28e7.40 (m, 5H,Ph). 13C NMR (62.9 MHz, CDCl3): d 14.1 (Me), 22.6, 23.8, 23.9, 28.3,29.0, 29.1, 29.3, 29.4, 29.6, 31.9, 32.3, (9�CH2, C-2 and C-3), 69.9(PhCH2), 81.2, 81.8 (C-4 and C-5), 125.0 (C-6), 127.6, 127.7, 128.4 (Ph),138.1 (C-7),177.4 (C-1). HRMS (ESI):m/z 373.2728 (MþþH), calcd forC24H37O3: 373.2737.

4.1.4. (þ)-Muricatacin (1). To a stirred solution of 11 (59 mg,0.2 mmol) in MeOH (1.2 mL) was added 10% Pd/C (83 mg,0.08 mmol). The suspension was hydrogenated at room tempera-ture and normal pressure of H2 for 4 h, then filtered through a Celitepad, washed with 1:1 CH2Cl2/EtOAc and evaporated. Silica gel flashcolumn chromatography (7:3 light petroleum/EtOAc) of the residuegave pure 1 (39 mg, 90%) as a white solid that was recrystallizedfrom a mixture of Et2O/pentane to afford colourless needles, mp68e69 �C, [a]D þ22.3 (c 0.43, CHCl3), Rf¼0.21 (7:3 light petroleum/EtOAc); lit.3 mp 68e70 �C, [a]D þ22.4 (c 0.42, CHCl3); lit.4 mp68e70 �C, [a]D þ19.6 (c 1.0, CHCl3). IR (CHCl3): nmax 3360 and 2447(OH), 1742 (C]O). 1H NMR (250 MHz, CDCl3): d 0.87 (t, 3H,J¼6.7 Hz, Me),1.13e1.64 (m, 22H,11�CH2), 1.90 (d,1H, J5,OH¼5.9 Hz,OH), 2.01e2.33 (m, 2H, 2�H-3), 2.50 (dd, 1H, J2a,2b¼17.8 Hz,J2a,3¼9.0 Hz, H-2a), 2.63 (ddd, 1H, J2a,2b¼17.8 Hz, J3,2b¼9.5 Hz,J3,2b¼5.2 Hz, H-2b), 3.56 (m, 1H, J4,5¼4.6 Hz, H-5), 4.41 (td, 1H,J3,4¼7.4 Hz, J4,5¼4.6 Hz, H-4). 13C NMR (62.9 MHz, CDCl3): d 14.1(Me), 24.1 (C-3), 28.7 (C-2), 22.7, 25.4, 29.3, 29.5, 29.6, 29.63, 29.65,31.9 and 33.0 (9�CH2), 73.7 (C-5), 82.9 (C-4), 177.1 (C-1). HRMS(ESI): m/z 285.2420 (MþþH), calcd for C17H33O3: 285.2424.

4.1.5. 3,6-Anhydro-5,7-di-O-benzyl-2-deoxy-D-ido-heptono-1,4-lactone (14). To a solution of 13 (0.95 g, 2.89 mmol) in dry DMF(9.4 mL), was added anhydrous Et3N (0.81 mL, 5.81 mmol) andMeldrum’s acid (0.83 g, 5.78 mmol). The mixture was stirred for48 h at 46e48 �C and then evaporated. The residue was purified byflash column chromatography on silica gel (49:1 CH2Cl2/EtOAc) toafford pure 14 (0.84 g, 82%), as a colourless solid. Recrystallizationfrom a mixture of CH2Cl2/Et2O/light petroleum gave colourlessneedles, mp 90e91 �C, Rf¼0.50 (3:2 Et2O/toluene); lit.21 mp 90 �C.Spectroscopic data of thus prepared sample 14 matched thosepreviously reported by us.21

4.1.6. 3,6-Anhydro-5-O-benzyl-2-deoxy-D-ido-heptono-1,4-lactone(15). A solution of 14 (0.56 g, 1.58 mmol) in a mixture of EtOAc(4 mL) and abs EtOH (4 mL) was added to a stirred suspension of10% Pd/C (0.17 g, 0.16 mmol, 0.1 equiv Pd) in abs EtOH (8 mL), whichwas pre-saturated with H2 for 1 h. The suspension was hydroge-nated at room temperature and normal pressure of H2 for 105 min,then filtered through a Celite pad, washed with EtOH, and evapo-rated. Flash column chromatography (9:1 Et2O/light petroleum) ofthe residue gave pure 15 (0.36 g, 87%), as a colourless syrup, [a]Dþ4.3 (c 1.0, CHCl3), Rf¼0.31 (Et2O). IR (CHCl3): nmax 3467 (OH), 1789(C]O). 1H NMR (250 MHz, CDCl3): d 2.52 (br s, 1H, OH), 2.58e2.78(m, 2H, 2�H-2), 3.76 (dd, 1H, J6,7a¼4.3 Hz, J7a,7b¼12.0 Hz, H-7a),3.84 (dd, 1H, J6,7b¼5.1 Hz, J7a,7b¼12.0 Hz, H-7b), 4.17 (m, 1H,J5,6¼4.9 Hz, H-6), 4.25 (d, 1H, J5,6¼4.9 Hz, H-5), 4.56 and 4.71 (2�d,Jgem¼11.9 Hz, CH2Ph), 4.91e5.01 (m, 2H, H-3 and H-4), 7.26e7.42(m, 5H, Ph). 13C NMR (62.9 MHz, CDCl3): d 35.8 (C-2), 61.1 (C-7), 72.7(CH2Ph), 76.7 (C-3), 80.7 (C-6), 82.1 (C-5), 85.7 (C-4), 127.6, 128.2,128.6, 136.7 (Ph), 175.2 (C-1). LRMS (ESI): m/e 265 (MþþH), 529(2 MþþH). HRMS (ESI): m/z 265.1066 (MþþH), calcd for C14H17O5:


265.1070; m/z 287.0885 (MþþNa), calcd for C14H16NaO5: 287.0890;m/z 303.0625 (MþþK), calcd for C14H16KO5: 303.0629.

4.1.7. 3,6-Anhydro-5-O-benzyl-2-deoxy-7-O-nonyl-D-ido-heptono-1,4-lactone (16). To a solution of 15 (0.25 g, 0.95 mmol) in dry Et2O(5 mL) were added successively Ag2O (0.55 g, 2.36 mmol), AgOTf(61 mg, 0.24 mmol) and C9H19Br (0.45 mL, 2.35 mmol). The mix-ture was stirred under reflux for 30 h, then diluted with CH2Cl2(10 mL), filtered, and evaporated. The residue was purified ona column of flash silica (3:2 light petroleum/Et2O) to give pure 16(0.23 g, 61%), a colourless oil that crystallized from cooled (�10 �C)CH2Cl2/hexane as colourless needles, mp 31e32 �C, [a]D þ11.6 (c1.8, CHCl3), Rf¼0.33 (1:1 Et2O/light petroleum). IR (neat): nmax1790 (C]O). 1H NMR (250 MHz, CDCl3): d 0.88 (t, 3H, J¼6.7 Hz,Me), 1.18e1.49 (m, 12H, 6�CH2), 1.58 (m, 2H, CH2), 2.72 (d, 2H,2�H-2), 3.46 (m, 2H, OCH2CH2), 3.65 (d, 2H, J6,7¼5.4 Hz, H-7), 4.20(d, 1H, J5,6¼4.0 Hz, H-5), 4.25 (m, 1H, H-6), 4.60 and 4.70 (2�d, 2H,Jgem¼11.9 Hz, CH2Ph), 4.93 (d, 1H, J3,4¼4.8 Hz, H-4), 4.99 (m, 1H, H-3), 7.29e7.46 (m, 5H, Ph). 13C NMR (62.9 MHz, CDCl3): d 14.0 (Me),22.6, 26.0, 29.2, 29.4, 29.45, 29.5, 31.8 (7�CH2) 35.9 (C-2), 68.5(2�H-7), 71.7 (OCH2CH2), 72.6 (PhCH2), 76.7 (C-3), 79.5 (C-6), 81.3(C-5), 85.3 (C-4), 127.6, 128.1, 128.5, 137.0 (Ph), 175.3 (C-1). LRMS(ESI): m/z 391 (MþþH). Anal. Found: C, 70.47; H, 8.70. Calcd forC23H34O5: C, 70.74; H, 8.78.

4.1.8. 3,6-Anhydro-2-deoxy-7-O-nonyl-D-ido-heptono-1,4-lactone(3). A solution of 16 (0.11 g, 0.28 mmol) in MeOH (4 mL) was hy-drogenated over 10% Pd/C (21 mg, 0.02 mmol) at room temperatureand normal pressure of H2 for 18 h, then filtered through a Celitepad, washed with MeOH, and evaporated. The residue was purifiedby flash column chromatography (7:3 CH2Cl2/EtOAc), to afford pure3 (81 mg, 96%) as a colourless syrup oil that crystallized fromCH2Cl2/hexane, as transparent needles, mp 51e52 �C, [a]D þ29.7 (c1.9, CHCl3), Rf¼0.40 (7:3 CH2Cl2/EtOAc). IR (neat): nmax 3286 (OH),1776 (C]O). 1H NMR (250 MHz, CDCl3): d 0.85 (t, 3H, J¼6.8 Hz, Me),1.11e1.39 (m, 12H, 6�CH2), 1.57 (m, 2H, CH2), 2.65 (dd, 1H,J2a,2b¼18.7 Hz, J2a,3¼1.0 Hz, H-2a), 2.76 (dd, 1H, J2a,2b¼18.7 Hz,J2b,3¼5.4 Hz, H-2b), 3.49 (m, 2H, OCH2CH2), 3.85 (m, 2H, 2�H-7),4.10 (m, 1H, H-6), 4.28 (d, 1H, J5,OH¼3.8 Hz, OH), 4.50 (dd, 1H,J5,6¼3.1 Hz, J5,OH¼3.8 Hz, H-5), 4.86 (d, 1H, J3,4¼4.2 Hz, H-4), 5.01(ddd, 1H, J2a,3¼1.0 Hz, J2b,3¼5.4 Hz, J3,4¼4.2 Hz, H-3). 13C NMR(62.9 MHz, CDCl3): d 14.0 (Me), 22.6, 25.8, 29.1, 29.26, 29.3, 29.4,31.7 (7�CH2) 36.0 (C-2), 69.4 (C-7), 72.5 (OCH2CH2), 75.8 (C-5), 76.8(C-3), 78.5 (C-6), 88.2 (C-4), 175.5 (C-1). LRMS (ESI): m/z 301(MþþH), 601 (2 MþþH).

4.1.9. 3,6-Anhydro-5-O-benzyl-7-O-decyl-2-deoxy-D-ido-heptono-1,4-lactone (17). To a solution of 15 (0.18 g, 0.66 mmol) in dry Et2O(3.5 mL) were added successively Ag2O (0.39 g, 1.66 mmol), AgOTf(43 mg, 0.17 mmol) and C10H21Br (0.35 mL, 1.66 mmol). The mix-ture was heated under reflux for 31 h, then cooled to room tem-perature, dilutedwith CH2Cl2 (5 mL), filtered, and evaporated. Flashcolumn chromatography (4:1/7:3 hexane/Et2O) of the residuegave pure 17 (0.14 g, 51%) as a colourless oil, [a]D þ13.1 (c 1.0,CHCl3), Rf¼0.25 (1:1 hexane/Et2O). Anal. Found: C, 71.14; H, 8.72.Calcd for C24H36O5: C, 71.26; H, 8.97. Spectroscopic data of thusprepared sample 17 matched those previously reported by us.14

4.1.10. 3,6-Anhydro-7-O-decyl-2-deoxy-D-ido-heptono-1,4-lactone(4). A solution of 17 (0.15 g, 0.37 mmol) in EtOH (3 mL) was hy-drogenated over 10% Pd/C (76 mg, 0.07 mmol) at room temperatureand normal pressure of H2 for 20 h. The suspension was filteredthrough a Celite pad and washed with ether. The combined filtrateswere evaporated and the residue purified by flash column chro-matography (4:1 Et2O/hexane/Et2O), to afford pure 4 (0.1 g, 86%)as a colourless solid. Recrystallization from CH2Cl2/hexane gave an

analytical sample 4, as colourless needles, mp 59e60 �C, [a]D þ35.4(c 0.45, CHCl3), Rf¼0.24 (4:1 Et2O/hexane). Anal. Found: C, 64.66; H,9.86. Calcd for C17H30O5: C, 64.94; H, 9.62. Spectroscopic data ofthus prepared sample 4matched those previously reported by us.14

4.1.11. 3,6-Anhydro-5-O-benzyl-2-deoxy-7-O-undecyl-D-ido-hep-tono-1,4-lactone (18). A mixture of 15 (84 mg, 0.32 mmol), Ag2O(0.2 g, 0.85 mmol), AgOTf (17 mg, 0.07 mmol) and C11H23Br(0.18 mL, 0.81 mmol) in anhydrous Et2O (2 mL) was heated underreflux for 48 h. After themixture cooled to room temperature it wasdiluted with CH2Cl2 (5 mL), filtered, and evaporated. Flash columnchromatography (3:2 light petroleum/Et2O) of the residue gavepure lactone 18 (74 mg, 56%) as a colourless oil, [a]D þ10.9 (c 0.19,CHCl3), Rf¼0.23 (3:2 light petroleum/Et2O). IR (neat): nmax 1790(C]O). 1H NMR (250 MHz, CDCl3): d 0.89 (t, 3H, J¼6.8 Hz, Me),1.12e1.41 (m, 16H, 8�CH2), 1.58 (m, 2H, OCH2CH2CH2), 2.70 (m, 2H,2�H-2), 3.46 (m, 2H, OCH2CH2), 3.64 (d, 2H, J6,7¼5.3 Hz, H-7), 4.20(d, 1H, J5,6¼4.0 Hz, H-5), 4.24 (m, 1H, H-6), 4.59 and 4.70 (2�d, 2H,Jgem¼11.9 Hz, CH2Ph), 4.92 (d, 1H, J3,4¼4.7 Hz, H-4), 4.97 (m, 1H, H-3), 7.29e7.40 (m, 5H, Ph). 13C NMR (62.9 MHz, CDCl3): d 14.1 (Me),22.7, 26.1, 29.3, 29.4, 29.6, 31.9 (9�CH2), 36.0 (C-2), 68.5 (C-7), 71.8(OCH2CH2), 72.7 (CH2Ph), 76.8 (C-3), 79.6 (C-6), 81.4 (C-5), 85.5 (C-4), 127.7, 128.1, 128.6, 137.2 (Ph), 175.4 (C-1). HRMS (ESI): m/z441.2625 (MþþNa), calcd for C25H38NaO5: 441.2612; m/z 457.2370(MþþK), calcd for C25H38KO5: 457.2351.

4.1.12. 3,6-Anhydro-2-deoxy-7-O-undecyl-D-ido-heptono-1,4-lactone(5). A solution of benzyl ether 18 (59 mg, 0.14 mmol) in MeOH(1.5 mL) was hydrogenated over 10% Pd/C (30 mg, 0.03 mmol) atroom temperature and normal pressure of H2 for 3.5 h. The sus-pension was filtered through a Celite pad and washed with MeOH.The combined filtrates were evaporated and the residue purified byflash column chromatography (24:1 CH2Cl2/MeOH) to afford pure 5(38 mg, 83%) as a colourless solid. Recrystallization from CH2Cl2/hexane gave transparent needles, mp 60 �C, [a]D þ27.3 (c 0.98,CHCl3), Rf¼0.49 (19:1 CH2Cl2/MeOH). IR (neat): nmax 3484e3275(OH), 1790 (C]O). 1H NMR (250 MHz, CDCl3): d 0.87 (t, 3H,J¼6.7 Hz, Me), 1.12e1.39 (m, 16H, 8�CH2), 1.57 (m, 2H, CH2), 2.65(dd,1H, J2a,3¼1.2 Hz, J2a,2b¼18.6 Hz, H-2a), 2.76 (dd,1H, J2b,3¼5.3 Hz,J2a,2b¼18.7 Hz, H-2b), 3.51 (m, 2H, OCH2CH2), 3.84 (dd, 1H,J6,7a¼3.2 Hz, J7a,7b¼11.0 Hz, H-7a), 3.90 (dd, 1H, J6,7b¼3.3 Hz,J7a,7b¼11.2 Hz, H-7b), 4.11 (m, 1H, H-6), 4.52 (d, 1H, J5,6¼3.2 Hz, H-5), 4.86 (d, 1H, J3,4¼4.2 Hz, H-4), 5.02 (m, 1H, H-3). 13C NMR(62.9 MHz, CDCl3): d 14.0 (Me), 22.6, 25.9, 29.2, 29.3, 29.36, 29.4,29.5, 31.8 (9�CH2), 36.0 (C-2), 69.5 (C-7), 72.6 (OCH2CH2), 76.0 (C-5), 76.8 (C-3), 78.6 (C-6), 88.2 (C-4), 175.4 (C-1). HRMS (ESI): m/z329.2317 (MþþH), calcd for C18H33O5: 329.2322; m/z 351.2146(MþþNa), calcd for C18H32NaO5: 351.2142.

4.1.13. 3,6-Anhydro-5-O-benzyl-2-deoxy-D-glycero-D-ido-octono-1,4-lactone (20). To a solution of 19 (0.25 g, 0.94 mmol) in dry DMF(2.5 mL) were added anhydrous Et3N (0.52 mL, 3.73 mmol) andMeldrum’s acid (0.55 g, 3.85 mmol). The resulting reaction mixturewas stirred at 46e48 �C for 70 h, and then evaporated. Purificationby flash column chromatography (4:1 EtOAc/light petroleum) gave20 (0.12 g, 43%) as a white solid. Recrystallization from a mixture ofCH2Cl2/hexane gave pure 20 as a colourless powder, mp104e105 �C, [a]D þ32.2 (c 0.5, EtOH); Rf¼0.2 (Et2O). IR (KBr): nmax

3444 (OH), 1784 (C]O). 1H NMR (250 MHz, CDCl3): d 2.64 (d, 1H,J2a,2b¼18.5 Hz, H-2a), 2.74 (dd, 1H, J2a,2b¼18.5 Hz, J3,2b¼4.5 Hz, H-2b), 3.66 (dd, 1H, J7,8a¼4.7 Hz, J8a,8b¼12.7 Hz, H-8a), 3.79 (dd, 1H,J7,8b¼2.7 Hz, J8a,8b¼12.7 Hz, H-8b), 3.94e4.04 (m, 2H, H-5 and H-7),4.35 (d, 1H, J¼2.5 Hz, H-6), 4.65 and 4.75 (2�d, 2H, Jgem¼11.7 Hz,PhCH2), 4.91e4.98 (m, 2H, H-3 and H-4), 7.31e7.43 (m, 5H, Ph). 13CNMR (62.9 MHz, CDCl3): d 36.0 (C-2), 64.2 (C-8), 69.2 (C-7), 72.9(PhCH2), 77.1 (C-3), 80.2 (C-5), 81.4 (C-6), 85.0 (C-4), 128.0, 128.5,


128.8, 136.8 (Ph), 175.4 (C-1). HRMS (ESI): m/z 317.0999 (MþþNa),calcd for C15H18NaO6: 317.0996.

4.1.14. 3,6-Anhydro-5-O-benzyl-2,7,8-trideoxy-D-ido-oct-7-enono-1,4-lactone (21). To a mixture of iodine (0.55 g, 2.18 mmol), imid-azole (0.26 g, 4.40 mmol) and Ph3P (0.56 g, 2.14 mmol) in dryMeCN(4.0 mL) was added a solution of 20 (0.16 g, 0.53 mmol) in dryMeCN (4 mL). The mixture was stirred at 90 �C (bath temperature)for 1.5 h, in an atmosphere of N2, and then evaporated. Flash col-umn chromatography (1:1 Et2O/light petroleum) of the residueyielded pure 21 (0.13 g, 93%) as awhite solid. Recrystallization fromCH2Cl2/hexane/Et2O afforded colourless needles, mp 62e63 �C, [a]Dþ11.2 (c 1.0, CHCl3), Rf¼0.29 (1:1 Et2O/light petroleum). IR (CHCl3):nmax 1789 (C]O). 1H NMR (250 MHz, CDCl3): d 2.66 (dd, 1H,J2a,2b¼18.9 Hz, J2a,3¼1.7 Hz, H-2a), 2.76 (dd, 1H, J2b,3¼4.8 Hz,J2a,2b¼18.9 Hz, H-2b), 4.14 (d, 1H, J5,6¼3.6 Hz, H-5), 4.51 (dd, 1H,J5,6¼3.7 Hz, J6,7¼7.0 Hz, H-6), 4.62 and 4.68 (2�d, 2H, Jgem¼12.1 Hz,CH2Ph), 4.80e5.00 (m, 2H, H-3 and H-4), 5.35 (dd,1H, J7,8a¼10.4 Hz,J8a,8b¼1.0 Hz, H-8a), 5.42 (dd, 1H, J7,8b¼17.3 Hz, J8a,8b¼1.0 Hz, H-8b),6.01 (ddd, 1H, J6,7¼7.0 Hz, J7,8a¼10.4 Hz, J7,8b¼17.3 Hz, H-7),7.27e7.42 (m, 5H, Ph). 13C NMR (62.9 MHz, CDCl3): d 35.9 (C-2), 72.7(CH2Ph), 76.4 (C-3), 81.9 (C-6), 82.6 (C-5), 85.8 (C-4), 119.4 (C-8),127.6, 128.0, 128.5, 137.0 (Ph), 132.0 (C-7), 175.3 (C-1). HRMS (ESI):m/z 261.1112 (MþþH), calcd for C15H17O4: 261.1121.

4.1.15. (E)-3,6-Anhydro-5-O-benzyl-2,7,8-trideoxy-8-C-nonyl-D-ido-oct-7-enono-1,4-lactone (22). Toa solutionof21 (79 mg, 0.30 mmol)in dry CH2Cl2 (1.6 mL)were addedundec-1-ene (0.5 mL, 2.43 mmol)and the second generation Grubbs catalyst (25 mg, 0.03 mmol). Themixture was stirred in an argon atmosphere for 24 h at room tem-perature. The solvent was removed under vacuum and the mixturepurified by flash column chromatography (CH2Cl2). Eluted first waspure 22 (80 mg, 68%), isolated as a colourless oil, [a]D þ7.5 (c 0.99,CHCl3), Rf¼0.35 (CH2Cl2). IR (neat): nmax 1790 (C]O). 1H NMR(250 MHz, CDCl3): d 0.9 (t, 3H, J¼7.0 Hz, Me), 1.06e1.49 (m, 14H,7�CH2), 2.10 (m, 2H, 2�H-9), 2.63 (d, 1H, J2a,2b¼17.8 Hz, H-2a), 2.75(dd, 1H, J2a,2b¼17.8 Hz, J2b,3¼4.7 Hz, H-2b), 4.07 (d, 1H, J5,6¼3.4 Hz,H-5), 4.46 (dd, 1H, J5,6¼3.4 Hz, J6,7¼7.7 Hz, H-6), 4.61 and 4.67 (2�d,2H, Jgem¼12.1 Hz, CH2Ph), 4.90e4.98 (m, 2H, H-3 and H-4), 5.66 (dd,1H, J6,7¼7.8 Hz, J7,8¼15.5 Hz, H-7), 5.85 (dt, 1H, J7,8¼15.5 Hz,J8,9¼6.5 Hz, H-8), 7.29e7.42 (m, 5H, Ph). 13C NMR (62.9 MHz, CDCl3):d 14.0 (Me), 22.6, 28.8, 29.1, 29.2, 29.4, 29.6, 31.8, 32.3 (8�CH2), 36.0(C-2), 72.6 (CH2Ph), 76.1 (C-3), 81.8 (C-5), 82.5 (C-6), 85.9 (C-4),123.2(C-7),127.5,127.9,128.4 (Ph),137.2 (C-8),175.4 (C-1). HRMS (ESI):m/z 387.2517 (MþþH), calcd for C24H35O4: 387.2530; m/z 425.2077(MþþK), calcd for C24H34KO4: 425.2089. Eluted second was un-changed starting compound 21 (0.019 g, 25%).

4.1.16. 3,6-Anhydro-2-deoxy-6-C-undecyl-D-ido-hexono-1,4-lactone(6). To a stirred solution of 22 (57 mg, 0.15 mmol) in dry MeOH(1.13 mL) was added 10% Pd/C (78 mg, 0.07 mmol). The suspensionwas hydrogenated at room temperature and normal pressure of H2for 3 h, then filtered through a Celite pad, washed with 1:1 CH2Cl2/EtOAc, and evaporated. Flash chromatography (7:3 Et2O/light pe-troleum) of the residue gave pure 6 (36 mg, 82%) as a white solid.Recrystallization from CH2Cl2/hexane gave colourless needles, mp74�75 �C, [a]D þ23.9 (c 0.51, CHCl3), Rf¼0.18 (7:3 Et2O/light petro-leum). IR (CHCl3): nmax 1779 (C]O), 3389 (OH). 1H NMR (250 MHz,CDCl3): d 0.87 (t, 3H, J¼6.9 Hz, Me), 1.16e1.72 (m, 20H, 10�CH2),2.32 (br s, 1H, OH), 2.62 (d, 1H, J2a,2b¼18.9 Hz, H-2a), 2.77 (dd, 1H,J2a,2b¼18.9 Hz, J2b,3¼5.9 Hz, H-2b), 3.91 (td, 1H, J5,6¼2.6 Hz,J6,7¼6.7 Hz, H-6), 4.27 (d,1H, J5,6¼2.6 Hz, H-5), 4.87e4.98 (m, 2H, H-3 and H-4). 13C NMR (62.9 MHz, CDCl3): d 14.1 (Me), 22.6, 26.1, 27.8,29.4, 29.5, 29.6, 29.62, 31.8 (10�CH2), 35.9 (C-2), 74.4 (C-5), 75.6 (C-3), 80.5 (C-6), 87.8 (C-4), 175.9 (C-1). HRMS (ESI): m/z 299.2213(MþþH), calcd for C17H31O4: 299.2217; m/z 316.2475 (MþþNH4),

calcd for C17H34NO4: 316.2482; m/z 337.1772 (MþþK), calcd forC17H30KO4: 337.1776.

4.1.17. (E)-3,6-Anhydro-5-O-benzyl-8-C-decyl-2,7,8-trideoxy-D-ido-oct-7-enono-1,4-lactone (23). To a solution of 21 (32 mg, 0.12 mmol)in dry CH2Cl2 (0.65 mL) was added dodec-1-ene (0.27 mL,1.22 mmol) and the second generation Grubbs catalyst (8 mg,0.01 mmol). Themixturewas stirred in anargonatmosphere for 68 hat room temperature. The solvent was removed in vacuum and themixture purified by flash column chromatography (CH2Cl2) to affordpure 23 (34 mg, 69%) as a pale yellow oil, [a]D þ11.2 (c 0.52, CHCl3),Rf¼0.38 (1:1 Et2O/light petroleum). IR (neat): nmax 1790 (C]O). 1HNMR (250 MHz, CDCl3): d 0.89 (t, 3H, J¼6.8 Hz, Me), 1.02e1.49 (m,16H, 8�CH2), 2.09 (m, 2H, 2�H-9), 2.64 (d, 1H, J2a,2b¼17.6 Hz, H-2a),2.74 (dd, 1H, J2b,3¼4.6 Hz, J2a,2b¼17.6 Hz, H-2b), 4.01 (d, 1H,J5,6¼3.4 Hz, H-5), 4.45 (dd,1H, J5,6¼3.4 Hz, J6,7¼7.7 Hz, H-6), 4.61 and4.67 (2�d, 2H, Jgem¼12.1 Hz, CH2Ph), 4.89e4.99 (m,2H,H-3 andH-4),5.65 (dd,1H, J6,7¼7.8 Hz, J7,8¼15.5 Hz, H-7), 5.84 (dt,1H, J7,8¼15.5 Hz,J8,9¼6.6 Hz, H-8), 7.29e7.43 (m, 5H, Ph). 13C NMR (62.9 MHz, CDCl3):d 14.0 (Me), 22.6, 28.8, 29.1, 29.2, 29.4, 29.5, 31.8, 32.8 (9�CH2), 36.0(C-2), 72.7 (CH2Ph), 76.2 (C-3), 81.9 (C-5), 82.6 (C-6), 85.9 (C-4),123.3(C-7), 127.6, 128.0, 128.4, 136.6 (Ph), 137.2 (C-8), 175.5 (C-1). HRMS(ESI):m/z383.2572 (MþþH�H2O), calcd for C25H35O3: 383.2581;m/z423.2486 (MþþNa), calcd for C25H36NaO4: 423.2506.

4.1.18. 3,6-Anhydro-2-deoxy-6-C-dodecyl-D-ido-hexono-1,4-lactone(7). To a stirred solution of 23 (54 mg, 0.13 mmol) in dry MeOH(1 mL) was added 10% Pd/C (61 mg, 0.06 mmol). The suspensionwas hydrogenated at room temperature and normal pressure of H2for 4.5 h, then filtered through a Celite pad, washed with 1:1CH2Cl2/EtOAc, and evaporated. Flash chromatography (3:2 Et2O/light petroleum) of the residue gave pure 7 (24 mg, 57%) as a whitesolid. Recrystallization from CH2Cl2/hexane gave colourless nee-dles, mp 88e89 �C, [a]D þ19.3 (c 0.52, CHCl3), Rf¼0.32 (1:1 Et2O/light petroleum). IR (neat): nmax 1779 (C]O), 3394 (OH). 1H NMR(250 MHz, CDCl3): d 0.88 (t, 3H, J¼6.7 Hz, Me), 1.12e1.73 (m, 22H,11�CH2), 2.34 (br s, 1H, OH), 2.62 (d, 1H, J2a,2b¼18.9 Hz, H-2a), 2.78(dd, 1H, J2a,2b¼18.9 Hz, J2b,3¼5.9 Hz, H-2b), 3.91 (td, 1H, J5,6¼2.6 Hz,J6,7¼6.8 Hz, H-6), 4.27 (d,1H, J5,6¼2.5 Hz, H-5), 4.87e4.99 (m, 2H, H-3 and H-4). 13C NMR (62.9 MHz, CDCl3): d 14.1 (Me), 22.6, 26.1, 27.8,29.3, 29.4, 29.5, 29.6, 29.62, 31.9 (11�CH2), 35.8 (C-2), 74.4 (C-5),75.6 (C-3), 80.5 (C-6), 87.8 (C-4), 175.8 (C-1). HRMS (ESI): m/z313.2373 (MþþH), calcd for C18H33O4: 313.2373; m/z 330.2634(MþþNH4), calcd for C18H36NO4: 330.2639.

4.1.19. 1,2-O-Cyclohexylidene-5-deoxy-5-C-dodecyl-a-D-xylo-pento-furanose (25). A crystal of iodine was added to a suspension ofmagnesium turnings (0.75 g, 30 mmol) in dry THF (15 mL) and thendodecylbromide (7.5 g, 30 mmol) was added while stirring, in oneportion at room temperature. The reaction started spontaneouslyand was completed after 1 h at reflux, whereupon the completedissolution of magnesiumwas observed. To this mixture was addeda solution of 24 (6.3 g, 30 mmol) in dry THF (15 mL) and the stirringunder reflux was continued for the next 4 h. The mixture wasquenched by the addition of 10% aq hydrochloric acid (100 mL) andproducts were extracted with light petroleum (3�50 mL). Thecombined organic layers were washed with 20% aq NaHCO3(50 mL), dried, discoloured with activated carbon and evaporated.Flash column chromatography (C6H6) of the residue (9.6 g) gavepure 25 (6.4 g, 56%) as a white waxy solid, [a]D �9.5 (c 1.1, CHCl3),Rf¼0.69 (1:1 Et2O/light petroleum). IR (neat): nmax 3409 (OH). 1HNMR (250 MHz, CDCl3): d 0.85 (t, 3H, J¼6.8 Hz, Me), 1.18e1.77 (m,34H,17�CH2), 2.54 (br s, 1H, OH), 4.00 (d,1H, J3,4¼2.4 Hz, H-3), 4.07(dt, 1H, J3,4¼2.2 Hz, J4,5¼6.7 Hz, H-4), 4.46 (d, 1H, J1,2¼3.8 Hz, H-2),5.86 (d, 1H, J1,2¼3.8 Hz, H-1). 13C NMR (CDCl3): d 14.0 (Me), 22.6,23.4, 23.8, 24.8, 26.0, 27.5, 29.3, 29.46, 29.5, 29.6, 29.7, 31.8, 35.5,


36.1 (17�CH2), 60.7 (C-7), 75.2 (C-3), 80.3 (C-4), 84.7 (C-2),103.6 (C-1), 111.9 (qC). LRMS (ESI): m/z 383 (MþþH), 382 (Mþ). HRMS (ESI):m/z 421.2711 (MþþK), calcd for C23H42KO4: 421.2715.

4.1.20. 5-Deoxy-5-C-dodecyl-D-xylo-pentofuranose (26). A solutionof 25 (0.26 g, 0.68 mmol) in 70% aq AcOH (10 mL) was stirred for3.5 h at reflux. After the mixture cooled to room temperature it wasconcentrated by co-distillation with toluene and the residue puri-fied by flash chromatography (EtOAc), to afford pure 26 (0.12 g,57%) as a colourless solid. Recrystallization from a mixture ofMeOH/H2O gave an analytical sample 26, as transparent needles,mp 115e117 �C, [a]D þ12.4 (c 0.50, MeOH) the initial [a]D value thatmutarotated toþ4.2 after equilibration for 71 h, Rf¼0.32 (EtOAc). IR(KBr): nmax 3381 (OH). 1H NMR (250 MHz, acetone-d6): d 0.81 (t, 6H,J¼6.8 Hz, Me a and b), 1.11e1.67 (m, 48H, 12�CH2 a and b),3.82e4.21 (m, 6H, H-2, H-3, H-4 a and b), 4.97 (s, 1H, H-1b), 5.30 (d,1H, J1,2¼3.7 Hz, H-1a). 13C NMR (62.9 MHz, methanol-d4): d 14.5(Me), 23.8, 27.1, 27.3, 30.0, 30.5, 30.76, 30.8, 30.9, 33.1 (13�CH2a and b), 77.2, 77.9, 78.4, 80.3, 82.6 and 83.4 (C-2, C-3 and C-4, a andb), 97.3 (C-1a), 104.0 (C-1b). HRMS (ESI): m/z 347.2444(MþHCOO�), calcd for C18H35O6: 347.2439.

4.1.21. 3,6-Anhydro-2-deoxy-6-C-tridecyl-D-ido-hexono-1,4-lactone(8). To a solution of 26 (76 mg, 0.25 mmol) in dry DMF (1 mL) wereadded Meldrum’s acid (80 mg, 0.56 mmol) and dry Et3N (0.07 mL,0.51 mmol). The mixture was stirred for 48 h at 46 �C and thenevaporated. The residue was purified by flash chromatography (4:1Et2O/light petroleum) to afford pure 8 (54 mg, 66%) as a colourlesssolid. Recrystallization fromCH2Cl2/hexane, gave colourless needles,mp 92 �C, [a]D þ16.2 (c 1.09, CHCl3), Rf¼0.30 (4:1 CH2Cl2/EtOAc). IR(CHCl3): nmax 3384 (OH), 1777 (C]O). 1H NMR (250 MHz, CDCl3):d0.83 (t, 3H, J¼6.7 Hz,Me),1.12e1.71 (m, 24H,12�CH2), 2.20 (br s,1H,OH), 2.62 (d, 1H, J2a,2b¼18.9 Hz, H-2a), 2.77 (dd, 1H, J2a,2b¼18.9 Hz,J2b,3¼5.8 Hz,H-2b), 3.91 (td,1H, J5,6¼2.7 Hz, J6,7¼6.8 Hz,H-6), 4.27 (d,1H, J5,6¼2.7 Hz, H-5), 4.88e4.98 (m, 2H, H-3 and H-4). 13C NMR(62.9 MHz, CDCl3): d 14.1 (Me), 22.6, 26.1, 27.8, 29.3, 29.4, 29.5, 29.6,31.9 (12�CH2), 35.9 (C-2), 74.3 (C-5), 75.6 (C-3), 80.5 (C-6), 87.8 (C-4),175.9 (C-1). HRMS (ESI): m/z 327.2525 (MþþH), calcd for C19H35O4:327.2530;m/z 349.2347 (MþþNa), calcd for C19H34NaO4: 349.2349;m/z 365.2084 (MþþK), calcd for C19H34KO4: 365.2089.

4.2. X-ray crystal structure determination

Single colourless crystals of compounds 7 and 8 were selectedand glued on glass fibres. Diffraction data were collected on anOxford Diffraction KM4 four-circle goniometer equipped withSapphire CCD detector. The crystal to detector distance was45.0 mm and a graphite monochromated MoKa (l¼0.71073�A) X-radiation was employed in both measurements. The frame widthof 1� in u, with 40 and 154 s were used to acquire each frame forboth 7 and 8. More than one hemisphere of three-dimensionaldata was collected in the measurement. The data were reducedusing the Oxford Diffraction program CrysAlisPro.22 A semi em-pirical absorption-correction based upon the intensities of equiv-alent reflections was applied, and the data were corrected forLorentz, polarization, and background effects. The structure wassolved by direct methods23 and the figures were drawn usingMercury v. 2.4.24 Refinements were based on F2 values and doneby full-matrix least-squares25 with all non-H atoms anisotropic.The positions of all non-H atoms were located by direct methods.The positions of hydrogen atoms were found from the inspectionof the difference Fourier maps. The final refinement includedatomic positional and displacement parameters for all non-Hatoms. At the final stage of the refinement H atoms were posi-tioned geometrically (OeH¼0.82 and CeH¼0.96e0.98�A) and re-fined using a riding model with fixed isotropic displacement

parameters. The crystal data and refinement parameters are listedin Table S1 in Supplementary data.

4.3. In vitro antitumour assay

Exponentially growing cells were harvested, counted by trypanblue exclusion and plated into 96-well microtitar plates (Costar) atoptimal seeding density of 104 (K562, HL-60, Jurkat and Raji) or5�103 (HT-29, MDA-MB-231, HeLa, and MRC-5) cells per well toassure logarithmic growth rate throughout the assay period. Anti-proliferative activity was evaluated by the tetrazolium colorimetricMTT assay, after exposure of cells to the tested compounds for 72 h,following the recently reported procedure.14

Acknowledgements

This work was supported by a research grant from the Ministryof Science and Technological Development of the Republic of Serbia(Grant No. 172006).

Supplementary data

More detailed description of the crystallographic results. Sup-plementary data associated with this article can be found in onlineversion, at doi:10.1016/j.tet.2011.09.132.

References and notes

1. Rieser, M. J.; Kozlowski, J. F.; Wood, K. V.; McLaughlin, J. L. Tetrahedron Lett.1991,32, 1137e1140.

2. (a) Srinivas, C.; Kumar, C. N. S. S. P.; Raju, B. C.; Rao, V. J. Helv. Chim. Acta 2011, 94,669e674; (b) Kumaraswamy, G.; Ramakrishna, D.; Santhakumar, K. Tetrahe-dron: Asymmetry 2010, 21, 544e548; (c) Muricia, M. C.; Navarro, C.; Moreno, A.;Cs�ak€y, A. G. Curr. Org. Chem. 2010, 14, 15e47 and references therein.

3. Ghosal, P.; Kumar, V.; Shaw, A. K. Carbohydr. Res. 2010, 345, 41e44.4. Yaragorla, S.; Muthyala, R. Arkivoc 2010, 10, 178e184.5. For a brief review see: Makabe, H. Biosci. Biotechnol. Biochem. 2007, 71,

2367e2374.6. (a) Navarro, C.; Moreno, A.; Cs�ak€y, A. G. J. Org. Chem. 2009, 74, 466e469; (b)

Konno, H.; Hiura, N.; Yanaru, M. Heterocycles 2002, 57, 1793e1797; (c) Andres, J.M.; de Elena, N.; Pedrosa, R.; Perez-Encabo, A. Tetrahedron: Asymmetry 2001, 12,1503e1509; (d) Pichon, M.; Jullian, J. C.; Figad�ere, B.; Cav�e, A. Tetrahedron Lett.1998, 39, 1755e1758; (e) Chang, S. W.; Hung, C. Y.; Liu, H. H.; Uang, B. J. Tet-rahedron: Asymmetry 1998, 9, 521e529; (f) Rassu, G.; Pinna, L.; Spanu, P.; Za-nardi, F.; Battistini, L.; Casiraghi, G. J. Org. Chem. 1997, 62, 4513e4517; (g)Gypser, A.; Peterek, M.; Scharf, H. D. J. Chem. Soc., Perkin Trans. 1 1997,1013e1016; (h) Baussanne, I.; Schwardt, O.; Royer, J.; Pichon, M.; Figad�ere, B.;Cav�e, A. Tetrahedron Lett. 1997, 38, 2259e2262; (i) Vanaar, M. P. M.; Thijs, L.;Zwanenburg, B. Tetrahedron 1995, 51, 11223e11234; (j) Saiah, M.; Bessodes, M.;Antonakis, K. Tetrahedron Lett. 1993, 34, 1597e1598.

7. Cav�e,A.;Chaboche,C.; Figad�ere, B.;Harmange, J. C.; Laurens,A.; Peyrat, J. F.; Pichon,M.; Szlosek, M.; Cotte-Lafitte, J.; Qu�ero, A. M. Eur. J. Med. Chem.1997, 32, 617e623.

8. (a) Tsai, S. H.; Hsieh, P. C.; Wei, L. L.; Chiu, H. F.; Wu, Y. C.; Wu, M. J. TetrahedronLett. 1999, 40, 1975e1976; (b) Yoon, S. H.; Moon, H. S.; Kang, S. K. Bull. KoreanChem. Soc. 1998, 19, 1016e1018; (c) Yoon, S. H.; Moon, H. S.; Hwang, S. K.; Choi,S. R.; Kang, S. K. Bioorg. Med. Chem. 1998, 6, 1043e1049; (d) Figad�ere, B.; Har-mange, J.-C.; Laurens, A.; Cav�e, A. Tetrahedron Lett. 1991, 32, 7539e7542.

9. Popsavin, V.; Krsti�c, I.; Popsavin, M.; Sre�co, B.; Benedekovi�c, G.; Koji�c, V.;Bogdanovi�c, G. Tetrahedron 2006, 62, 11044e11053.

10. Popsavin, V.; Sre�co, B.; Benedekovi�c, G.; Popsavin, M.; Francuz, J.; Koji�c, V.;Bogdanovi�c, G. Bioorg. Med. Chem. Lett. 2008, 18, 5182e5185.

11. Garegg, P. J.; Samuelsson, B. J. Synthesis 1979, 469e470.12. Popsavin, V.; Grabe�z, S.; Stojanovi�c, B.; Popsavin, M.; Pejanovi�c, V.; Miljkovi�c, D.

Carbohydr. Res. 1999, 321, 110e115.13. This part of the work was previously published as a preliminary communica-

tion: Popsavin, V.; Grabe�z, S.; Popsavin, M.; Krsti�c, I.; Koji�c, V.; Bogdanovi�c, G.;Divjakovi�c, V. Tetrahedron Lett. 2004, 45, 9409e9413.

14. Popsavin, V.; Sre�co, B.; Krsti�c, I.; Popsavin, M.; Koji�c, V.; Bogdanovi�c, G. Eur. J.Med. Chem. 2006, 41, 1217e1222.

15. (a) Karakawa, M.; Nakatsubo, F. Carbohydr. Res. 2002, 337, 951e954; (b) Bichard,F. J. C.; Wheatley, R. J.; Fleet, J. W. G. Tetrahedron: Asymmetry 1994, 5, 431e440.

16. Kapit�an, P.; Gracza, T. Tetrahedron: Asymmetry 2008, 19, 38e44.17. Had�zi�c, P.; Vukojevi�c, N.; Popsavin, M.; �Canadi, J. J. Serb. Chem. Soc. 2001, 66, 1e8.18. Crystallographic data (excluding structure factors) for the structures 7 and 8

have been deposited with the Cambridge Crystallographic Data Centre assupplementary publications number CCDC 832453 and 832454, respectively.Copies of the data can be obtained, free of charge, on application to CCDC, 12

http://dx.doi.org/doi:10.1016/j.tet.2011.09.132


Union Road, Cambridge CB2 1EZ, UK [fax: þ44 (0)1223 336033 or e-mail: [email protected]].

19. The synthesis of2waspreviously reportedbyus, alongwithpreliminary results ofits antiproliferative effects against K562, HL-60, Jurkat, HeLa and MCF-7 tumourcell lines, but after exposure of cells to the tested compound for 48 h (see Ref. 9).

20. Bermejo, A.; Figad�ere, B.; Zafra-Polo, M.-C.; Barrachina, I.; Estornell, E.; Cortes,B. L. Nat. Prod. Rep. 2005, 22, 269e303.

21. Popsavin, V.; Grabe�z, S.; Krsti�c, I.; Popsavin, M.; Djokovi�c, D. J. Serb. Chem. Soc.2003, 68, 795e804.

22. Data collection and processing software CrysAlisPro, v. 1; Oxford DiffractionLtd: Oxford, 2006.

23. Altornare, A.; Gascarano, G.; Giacovazzo, C.; Guagliardi, A. J. Appl. Crystallogr.1993, 26, 343e350.

24. Macrae, C. F.; Bruno, I. J.; Chisholm, J. A.; Edgington, P. R.; McCabe, P.; Pidcock,E.; Rodiguez-Monge, L.; Taylor, R.; Van de Streek, J.; Wood, P. A. J. Appl. Crys-tallogr. 2008, 41, 466e470.

25. Sheldrick, G. M. SHELXL 97, Program for Refinement of Crystal Structures; Uni-versity of G€ottingen: G€ottingen, 1997.




Disclaimer: The information provided is for informational purposes only and is not intended to replace the advice of your physician. Do not change or discontinue any medication or treatment without consulting your physician. The statements here are not intended to diagnose, treat, cure or prevent any disease.

hoaf-caim · 2018-01-16 · graviola, soursop, guanabana, guyabano . soursop leaves come from...

Documents