Caffeine and Caffeic Acid Inhibit Growth and Modify Estrogen Receptor and Insulin-Like Growth Factor I Receptor Levels in Human Breast Cancer Ann H

Published OnlineFirst February 17, 2015; DOI: 10.1158/1078-0432.CCR-14-1748

Cancer Therapy: Clinical Clinical Cancer Research Caffeine and Caffeic Acid Inhibit Growth and Modify Estrogen Receptor and Insulin-like Growth Factor I Receptor Levels in Human Breast Cancer Ann H. Rosendahl1, Claire M. Perks2, Li Zeng2, Andrea Markkula1, Maria Simonsson1, Carsten Rose3, Christian Ingvar4, Jeff M.P. Holly2, and Helena Jernstrom€ 1

Abstract

Purpose: Epidemiologic studies indicate that dietary factors, 0.018), compared with patients with low consumption (1 cup/ such as coffee, may influence breast cancer and modulate hor- day). Moderate to high consumption was associated with lower þ mone receptor status. The purpose of this translational study was risk for breast cancer events in tamoxifen-treated patients with ER to investigate how coffee may affect breast cancer growth in tumors (adjusted HR, 0.51; 95% confidence interval, 0.26–0.97). þ relation to estrogen receptor-a (ER) status. Caffeine and caffeic acid suppressed the growth of ER (P 0.01) Experimental Design: The influence of coffee consumption on and ER (P 0.03) cells. Caffeine significantly reduced ER and þ patient and tumor characteristics and disease-free survival was cyclin D1 abundance in ER cells. Caffeine also reduced the assessed in a population-based cohort of 1,090 patients with insulin-like growth factor-I receptor (IGFIR) and pAkt levels in þ invasive primary breast cancer in Sweden. Cellular and molecular both ER and ER cells. Together, these effects resulted in impaired effects by the coffee constituents caffeine and caffeic acid were cell-cycle progression and enhanced cell death. þ evaluated in ER (MCF-7) and ER (MDA-MB-231) breast cancer Conclusion: The clinical and experimental findings demon- cells. strate various anticancer properties of caffeine and caffeic acid þ Results: Moderate (2–4 cups/day) to high (5 cups/day) against both ER and ER breast cancer that may sensitize tumor coffee intake was associated with smaller invasive primary tumors cells to tamoxifen and reduce breast cancer growth. Clin Cancer Res; þ (Ptrend ¼ 0.013) and lower proportion of ER tumors (Ptrend ¼ 21(8); 1–11. 2015 AACR.

Introduction gression is the loss of the estrogen receptor-a (ER), which is associated with a more aggressive tumor phenotype and the loss Breast cancer is the major cancer affecting women with a of sensitivity to endocrine therapies such as tamoxifen (3). ER can lifetime risk of more than 10% in the general population in stimulate cellular proliferation via progesterone receptor (PgR) Western countries. The prevalence is, however, much lower in induction and transcriptional upregulation of cyclin D1, leading Asia and many economically developing countries (1). However, to enhanced cell-cycle progression, whereas ER loss has been Asian immigrants to the United States acquire the prevalence of associated with enhanced motility and metastatic abilities of that country within one generation, indicating a strong effect of breast cancer cells. Although loss of ER activity and expression environmental exposures. Although the majority of primary has critical consequences for breast cancers, it generally occurs breast cancer is estrogen-driven and often responsive to endocrine without evidence of ER gene mutation or deletion. Alterations in treatment, many patients progress to estrogen-insensitive disease ER levels can be regulated at transcriptional or posttranscriptional with poor prognosis (2, 3). A critical determinant of such pro- stages, or through epigenetic mechanisms (4). Given that the prevalence of breast cancer varies between different geographic locations, potential modifiable environmental, dietary, and life- 1Lund University and Ska ne University Hospital, Department of Clinical style determinants may also influence the ER status. Sciences, Lund, Division of Oncology and Pathology, Lund, Sweden. The insulin-like growth factor (IGF) family has been identified 2University of Bristol, School of Clinical Sciences, IGFs and Metabolic Endocrinology Group, Southmead Hospital, Bristol, United Kingdom. as an additional driver of breast cancer. The mitogenic and 3Lund University, CREATE Health and Department of Immunotech- antiapoptotic signals are mediated via the IGF type I receptor nology, Medicon Village, Lund, Sweden. 4Lund University and Skane (IGFIR) and its downstream targets MAPK–ERK or PI3K–Akt. The University Hospital, Department of Clinical Sciences, Lund, Division of ER and IGFIR signaling systems are tightly linked and subject to Surgery, Lund, Sweden. feedback cross-talk. For example, estrogen and IGFs coregulate Note: Supplementary data for this article are available at Clinical Cancer several genes that affect breast cancer outcome (5). In addition, Research Online (http://clincancerres.aacrjournals.org/). hyperactive IGFIR–PI3K–Akt signaling has been implicated to Corresponding Author: Ann H. Rosendahl, Lund University and Skane University contribute in the acquisition of resistance to endocrine therapy Hospital, Barngatan 2B, SE-221 85 Lund, Sweden. Phone: 46-46-17-7567; Fax: during breast cancer progression (6). Estrogen increases IGFIR 46-46-14-7327; E-mail: [email protected] mRNA and antiestrogens, like tamoxifen, can reduce serum IGFI doi: 10.1158/1078-0432.CCR-14-1748 levels and IGF-mediated growth (7, 8). Conversely, blocking IGF 2015 American Association for Cancer Research. actions can inhibit estrogen-mediated growth (9). Similar to ER

www.aacrjournals.org OF1

Rosendahl et al.

anti-IGFIRb (C-20) was purchased from Santa Cruz Biotechnol- Translational Relevance Ser473 ogy; anti–phospho-Akt (#4060), anti-Akt (#9272), anti–Bcl- Despite improved diagnostic and treatment modalities, 2 (#2870), anti–Bcl-xL (#2764), anti-cleaved caspase-7 (#9491), breast cancer remains the leading cause of cancer-related death anti-cleaved PARP (#9541), anti-cyclin D1 (#2926), and anti– among women worldwide. An increased understanding of the glyceraldehyde-3-phosphate dehydrogenase (GAPDH, #3683) clinical and mechanistic effects of modifiable components of were purchased from Cell Signaling Technology Inc. our daily diet on breast cancer characteristics and prognosis could lead to better patient information for managing disease. Study population This study reveals that increasing coffee consumption is The present study is based on an extended cohort of 1,090 associated with significantly smaller invasive breast tumor women, ages 24 to 99 years, diagnosed with primary invasive þ sizes, a lower proportion of ER tumors, and improved dis- breast cancer between October 2002 and December 2012 at the þ ease-free survival among tamoxifen-treated women with ER Skane University Hospital (Lund, Sweden), without preoperative breast cancer. Mechanistically, this study shows that predom- treatment and with no other cancer history within the past 10 inantly caffeine, but also caffeic acid, mimics the actions of years. This is an ongoing cohort (BC Blood study) and is now antiestrogens and modifies major growth regulatory pathways updated to include almost twice as many women as recently ER/cyclin D1 and IGFIR/pAkt, resulting in impaired cell-cycle reported by our group (17). The inclusion and exclusion criteria progression and reduced cellular proliferation. Together, these were compliant to the REporting recommendations for tumor translational findings provide a more comprehensive under- MARKer prognostic study (REMARK) guidelines (18). During the standing of the clinical and mechanistic effects of these dietary preoperative visit, body measurements of weight, height, waist factors that can modify breast cancer progression. and hip circumference, and breast volume were collected (as previously described; refs. 19, 20). Extensive questionnaire data with information on reproductive factors, the use of exogenous hormones or other medications, lifestyle factors, including smok- activity, various anthropometric, lifestyle, and dietary factors can ing, alcohol, and coffee consumption, etc., were obtained from influence the IGF system with impact on cancer development and the cohort members at the preoperative visit. Information on progression. tumor size, lymph node status, histologic grade, proliferative In recent years, mechanisms involved in mediating cancer index (Ki67), and hormone receptor status (ER and PgR) were a chemopreventive effects of naturally derived plant polyphenols obtained through pathology reports and medical records. ER from green tea and coffee have been widely studied in several status was routinely determined until December 2009, using neoplasias, including breast and prostate cancer (10–13). How- antibody clone 1D5 (DAKO), and from 2010 using clone SP1 ever, there have been few translational studies on anticancer (Ventana Medical Systems), as previously described (17, 21). effects of coffee in relation to ER status in breast cancer. Prospec- Follow-up information of breast cancer recurrence or death was tive and retrospective epidemiologic studies show an association obtained from patient charts, pathology reports, the Regional between coffee consumption and a significantly reduced risk, Tumor Registry or the Population Registry. Written informed delayed onset, and reduced growth of breast cancer (14, 15). consents were obtained from all participating women and the Coffee beans are one of the richest dietary sources of the anti- study was approved by the local ethics committee at Lund oxidants caffeine and the plant phenolic compound caffeic acid, University (LU 75-02, LU 37-08, LU 658-09, LU 58-12, LU both of which may be mechanistically involved in the cancer- 379-12, and LU 227-13). suppressive properties of coffee. Various dietary and lifestyle factors are also known to influence the sensitivity to cancer Cell culture a treatment (10, 16). In line with this, our group recently showed The human breast cancer cell lines MCF-7, T47D (both ER - a in a subset of the present study cohort, that moderate to high positive) and MDA-MB-231 (ER ) were purchased from and coffee consumption was associated with improved disease-free validated by ATCC-LGC Standards and maintained in DMEM þ survival among ER tamoxifen-treated women (17). supplemented with 10% FBS and antibiotics (100 U/mL penicillin and 100 mg/mL streptomycin) in a humidified 5% CO2 The aim of the present translational study was to investigate the effect of coffee consumption on tumor characteristics and disease- atmosphere at 37 C. Tamoxifen-resistant (Tam-R) MCF-7 cells free survival in an extended cohort of women with breast cancer in were kindly supplied by Julia Gee (Cardiff University, UK) and in Sweden, and to establish underlying molecular mechanisms by were derived and maintained as outlined previously (22). For vitro which coffee constituents may affect breast cancer cell growth in experiments, cells were seeded in growth media for 24 hours relation to ER status. Understanding how exposure to compo- followed by transfer to phenol red- and serum-free media [SFM; nents of our daily diet can modify tumor growth or molecular DMEM:Ham's nutrient mix F-12 supplemented with BSA (0.2 adaptations that occur within a tumor could lead to better patient mg/mL), sodium bicarbonate (1.2 mg/mL), transferrin (0.01 information for managing disease. mg/mL), L-glutamine (2 mmol/L), and antibiotics as above] and dosed as specified below. Materials and Methods Cell proliferation assay Reagents MCF-7, MDA-MB-231, T47D, and Tam-R cells were exposed to All chemicals and reagents were purchased from Sigma-Aldrich caffeine (0–5 mmol/L) or caffeic acid (0–50 mmol/L) with or unless stated otherwise. Cell culture media, penicillin–strepto- without tamoxifen (1 mmol/L) for 48 hours in SFM. Cell prolif- mycin, and FBS were purchased from Invitrogen. For immuno- eration was determined following [3H]thymidine incorporation blotting, the anti-ERa antibody (1D5) was obtained from Abcam; as a measure of DNA synthesis for a period of 4 hours before the

OF2 Clin Cancer Res; 21(8) April 15, 2015 Clinical Cancer Research

Impact of Coffee on Breast Cancer

end of the treatment protocol, as described in ref. 23. Total cell means SE of triplicate wells. Densitometry measurements of number and viability was assessed by trypan blue exclusion using Western blot data are expressed as means SE of at least three a hemocytometer. individual repeats. Statistical analyses were performed using the unpaired Student t test with Excel 2007 software. A P value of Cell-cycle analysis <0.05 was considered statistically significant. Cells were dosed with caffeine (0–5 mmol/L) or caffeic acid (0– m 50 mol/L) in SFM for 48 hours. Floating and adherent cells were Results collected, fixed in 70% cold ethanol, and incubated with PI/RNase buffer (BD Biosciences). Approximately 2 104 events (cells) Associations between tumor size, ER status, and coffee were collected for each sample using a FACSCalibur (BD Bios- consumption among women with breast cancer ciences) dual laser, flow cytometer with CellQuest Pro Software To evaluate the impact of coffee consumption on patient and (BD Biosciences). Cell-cycle distribution was determined using tumor characteristics among women with breast cancer, coffee FlowJo Software (TreeStar Inc.). consumption was categorized into low (1 cup/day), moderate (2–4 cups/day), or high (5 cups/day), whereas no further Western immunoblotting definition of a cup was specified. Patient and tumor character- Cells were exposed to caffeine (0–5 mmol/L) or caffeic acid (0– istics according to coffee consumption for the 1,090 patients 50 mmol/L) in SFM for 72 hours after which cell lysates were included in the study are shown in Table 1. Women that were prepared and protein concentration determined, as described younger at their first full-term pregnancy showed a trend of previously (24). Protein samples (50–60 mg/lane) were resolved drinking moderate to high amounts of coffee daily. In addition, by SDSPAGE and transferred to nitrocellulose membranes. The the proportion of current smokers was higher with increasing membranes were blocked with 5% (w/v) milk in TBS Tween-20 coffee consumption, whereas the proportion of alcohol abstai- (TBST) and probed with antibodies to Bcl-2, Bcl-xL, cleaved ners was higher with decreasing coffee consumption. Interest- caspase-7, cleaved PARP, ERa, IGFIRb, pAkt, Akt, cyclin D1, or ingly, moderate to high coffee consumption was associated GAPDH as a loading control. Protein abundance was detected with significantly smaller invasive tumor sizes (Ptrend ¼ 0.013), þ with horseradish peroxidase–specific secondary antibodies and whereas concomitantly lower proportion of ER tumors and visualized by SuperSignal West Extended Duration Substrate conversely higher proportion of ER tumors (Ptrend ¼ 0.018). (Thermo Fisher Scientific) using the Alpha Innotech FluorChem After stratification according to ER status (Table 2), increasing FC2 imaging system and AlphaView version 3.0.3.0 software coffee consumption remained significantly associated with þ (ProteinSimple) or ImageJ software (NIH). smaller invasive tumor sizes among women with ER tumors (Ptrend ¼ 0.009), whereas no significant association was found Statistical analysis among ER tumors (Ptrend ¼ 0.512). However, a nonsignificant Clinical data. Baseline characteristics of patients with low (1 trend toward lower tumor proliferation measured by the Ki67 P ¼ cup/day), moderate (2–4 cups/day), or high (5 cups/day) coffee index ( trend 0.092) was observed with increasing coffee consumption were compared as specified by using the Jonc- intake among women with ER tumors. These results suggest kheere–Terpstra test for continuous variables (not normally dis- that coffee may have growth-suppressing properties on breast þ tributed) and linear-by-linear association test for categorical vari- cancer cells and that ER tumors may be more sensitive to the ables. In survival analyses, patients were followed from inclusion cellular effects by coffee. until first breast cancer recurrence, last follow-up, or death, þ whichever came first, before January 1, 2013. Of 1,090 women, Effects of caffeine and caffeic acid on ER and ER breast cancer 8 were excluded because of generalized disease within 3 months cell growth and viability of diagnosis, leaving 1,082 women included in the survival Given the clinical observations of smaller breast tumor sizes þ analyses. Disease-free survival was estimated using the Kaplan– and lower frequency of ER tumors with increasing coffee con- Meier method. The log-rank test was used for univariable survival sumption, we next examined how two major coffee constituents analysis of the risk of early breast cancer recurrence in relation to may affect breast cancer growth in relation to ER status at the coffee consumption. The influence of coffee consumption on cellular level. Increasing doses of caffeine significantly suppressed þ prognosis was also analyzed using multivariable Cox regression the proliferation and total cell number of both ER MCF-7 and to calculate HR with 95% confidence intervals (CI), adjusted for ER MDA-MB-231 (Fig. 1A) breast cancer cells, with a maximum potential confounding factors; age (<60 years), histologic grade 3 inhibition at 5 mmol/L, compared with control (P < 0.01). At 1 (yes/no), invasive tumor size (>20 mm, or muscular or skin mmol/L caffeine, the proliferation was reduced by 80% in MCF-7 involvement) and any axillary lymph node involvement (yes/ cells (P < 0.01) and by 40% in MDA-MB-231 cells (P ¼ 0.054). no). Because few patients had an invasive tumor size 51 mm, or Similarly, caffeic acid reduced both MCF-7 and MDA-MB-231 cell muscular or skin involvement, these patients were combined with growth, although to a lesser extent than caffeine (Fig. 1A). A the patients with invasive tumor sizes between 21 and 50 mm in concomitant increase in cell death was observed for MCF-7, but the multivariable analyses. All statistical analyses were performed not for MDA-MB-231 (Fig. 1A). The response to caffeine and using SPSS Statistics 19.0 software (IBM). A P value of <0.05 was caffeic acid were identical in the Tam-R cells as with the parental considered statistically significant. Because this was an explorato- MCF-7 cells (Fig. 1B). The caffeine-induced cell death in MCF-7 ry study, nominal P values are presented without adjustment for was associated with reduced prosurvival Bcl-xL levels whereas multiple testing. increased active caspase-7 and cleaved PARP, a signature event of apoptosis (Fig. 1C and Supplementary Fig. S1). These results Experimental data. A minimum of three independent repeats were demonstrate that caffeine, but also caffeic acid, suppress the performed for each experiment. Proliferation data are expressed as proliferation of human breast cancer cells, and suggest that

www.aacrjournals.org Clin Cancer Res; 21(8) April 15, 2015 OF3

Rosendahl et al.

Table 1. Patient and tumor characteristics of the 1,090 patients in relation to daily coffee consumption (cups/day) All Low (0–1) Moderate (2–4) High (5þ)

(n ¼ 1,090) (n ¼ 202) (n ¼ 668) (n ¼ 220) Ptrend Patient characteristics, median (IQR) Age at diagnosis, y 61.2 (52.1–68.1) 59.4 (48.6–67.8) 62.7 (55.0–69.3) 58.6 (49.5–65.6) >0.3a Weight (kg) 69.0 (62.0–78.0) 68.0 (60.0–74.6) 70.0 (62.0–79.0) 69.2 (61.0–79.0) >0.3a Missing 36 8 18 10 Height (m) 1.65 (1.62–1.70) 1.65 (1.61–1.69) 1.65 (1.62–1.70) 1.67 (1.63–1.70) >0.3a Missing 36 8 18 10 BMI (kg/m2) 25.1 (22.5–28.4) 24.5 (21.9–27.7) 25.2 (22.8–28.6) 25.0 (22.6–27.9) >0.3a Missing 38 8 20 10 Waist-to-hip ratio 0.86 (0.81–0.91) 0.86 (0.81–0.90) 0.86 (0.81–0.90) 0.87 (0.82–0.91) >0.3a Missing 53 13 23 17 Total breast volume (mL) 1,000 (650–1,500) 925 (550–1,463) 1,000 (700–1,600) 950 (600–1,400) >0.3a Missing 178 36 101 41 Age at ﬁrst full-term pregnancy (y) 25.0 (22.0–28.0) 26.0 (23.0–29.0) 24.0 (21.0–28.0) 24.0 (21.0–28.0) 0.009a Missing 140 33 85 22 Current smoker, n (%) 219 (20.1%) 18 (8.9%) 119 (17.8%) 82 (37.3%) <0.001b Missing 4 1 3 0 Alcohol abstainer, n (%) 116 (10.6%) 33 (16.3%) 70 (10.5%) 13 (5.9%) 0.001b Missing 3 0 2 1 Tumor characteristics, n (%) Invasive tumor size (pT) 1 785 (72.0%) 134 (66.3%) 485 (72.6%) 166 (75.5%) 0.013b 2 287 (26.3%) 60 (29.7%) 175 (26.2%) 52 (23.6%) 3 16 (1.5%) 8 (4.0%) 6 (0.9%) 2 (0.9%) 4 2 (0.2%) 0 2 (0.3%) 0 Node status 0 668 (61.4%) 116 (57.7%) 419 (62.8%) 133 (60.5%) >0.3b 1–3 326 (30.0%) 65 (32.3%) 192 (28.8%) 69 (31.4%) 4þ 94 (8.6%) 20 (10.0%) 56 (8.4%) 18 (8.2%) Missing (n)2110 Histologic grade 1 266 (24.4%) 43 (21.3%) 167 (25.0%) 56 (25.5%) >0.3b 2 540 (49.6%) 110 (54.5%) 324 (48.6%) 106 (48.2%) 3 283 (26.0%) 49 (24.3%) 176 (26.4%) 58 (26.4%) Missing (n)1010 Ki67 Negative/low (20%) 300 (63.0%) 63 (63.0%) 173 (62.0%) 64 (66.0%) >0.3b High (>20%) 176 (37.0%) 37 (37.0%) 106 (38.0%) 33 (34.0%) Missing (n) 614 102 389 123 Hormone receptor status ERþ 956 (87.9%) 185 (91.6%) 586 (88.1%) 185 (84.1%) 0.018b ER 131 (12.1%) 17 (8.4%) 79 (11.9%) 35 (15.9%) Missing (n)3030 PgRþ 774 (71.2%) 139 (68.8%) 483 (72.6%) 152 (69.1%) >0.3b PgR 313 (28.8%) 63 (31.2%) 182 (27.4%) 68 (30.9%) Missing (n)3030 ERþPgRþ 767 (70.6%) 138 (68.3%) 479 (72%) 150 (68.2%) >0.3b ERþPgR 189 (17.4%) 47 (23.3%) 107 (16.1%) 35 (15.9%) 0.052b ERPgR 124 (11.4%) 16 (7.9%) 75 (11.3%) 33 (15.0%) 0.022b ERPgRþ 7 (0.6%) 1 (0.5%) 4 (0.6%) 2 (0.9%) >0.3b Missing (n)3030 aThe Jonckheere–Terpstra test. bThe linear-by-linear association test.

þ ER tumors may be more sensitive to the growth inhibiting cell death. Although caffeic acid reduced MCF-7 and MDA-MB- properties of the coffee constituents. 231 cell densities (Fig. 2A and B), only minor modulations of the þ cell-cycle distribution by caffeic acid were observed in the ER Coffee constituents impair the cell-cycle progression in breast (Fig. 2A and C) as well as ER cells (Fig. 2B and D). cancer cells Having demonstrated that caffeine and caffeic acid suppressed Caffeine and caffeic acid alter hormone receptor levels and the proliferation and viability of breast cancer cells, we further downstream effectors in breast cancer cells assessed their effect on cell-cycle distribution. Increasing doses of The role of coffee constituents in modulating molecular þ caffeine resulted in impaired G1 to S transition in both ER MCF-7 mechanisms with impact on breast cancer cell growth in relation (P 0.003; Fig. 2A and C) and ER MDA-MB-231 (P 0.001; Fig. to ER status was further investigated. Exposure to caffeine signif- þ 2B and D) cells, with an accumulation of cells in G0–G1,a icantly reduced the ER abundance in ER MCF-7 cells. In fact, 5 subsequent reduction of cells in S and G2–M phases and enhanced mmol/L caffeine almost completely abolished the ER levels

OF4 Clin Cancer Res; 21(8) April 15, 2015 Clinical Cancer Research

Impact of Coffee on Breast Cancer

Table 2. Selected tumor characteristics of 1,087a patients stratiﬁed by ER status in relation to daily coffee consumption (cups/day) All Low (0–1) Moderate (2–4) High (5þ) b (n ¼ 1,087) (n ¼ 202) (n ¼ 665) (n ¼ 220) Ptrend ERþ tumors (n ¼ 956) Invasive tumor size (pT) 1 704 (73.6%) 124 (67.0%) 436 (74.4%) 144 (77.8%) 0.009 2 236 (24.7%) 54 (29.2%) 143 (24.4%) 39 (21.1%) 3 14 (1.5%) 7 (3.8%) 5 (0.8%) 2 (1.1%) 4 2 (0.2%) 0 2 (0.3%) 0 Ki67 Negative/low (20%) 292 (68.7%) 62 (67.4%) 171 (68.1%) 59 (72.0%) 0.527 High (>20%) 133 (31.3%) 30 (32.6%) 80 (31.9%) 23 (28.0%) Missing (n) 531 93 335 103 ER tumors (n ¼ 131) Invasive tumor size (pT) 1 79 (60.3%) 10 (58.8%) 47 (59.5%) 22 (62.9%) 0.512 2 50 (38.2%) 6 (35.3%) 31 (39.2%) 13 (37.1%) 3 2 (1.5%) 1 (5.9%) 1 (1.3%) 0 40000 Ki67 Negative/low (20%) 8 (15.7%) 1 (12.5%) 2 (7.1%) 5 (33.3%) 0.092 High (>20%) 43 (84.3%) 7 (87.5%) 26 (92.9%) 10 (66.7%) Missing (n)8095120 aInformation on ER status was missing for 3 patients. bThe linear-by-linear association test.

(Fig. 3A). Although an approximately 50% reduction of MCF-7 hazard of early breast cancer recurrence when drinking moderate proliferation was obtained by caffeic acid treatment, the effect on to high amounts of coffee compared with low coffee consump- ER abundance was less pronounced (Fig. 3A). The ER suppression tion, adjusted HR 0.51 (95% CI, 0.26–0.97; P ¼ 0.040). In was associated with a concomitant cyclin D1 decrease, and no contrast, there was no significant difference in disease-free survival þ detectable levels remained after 5 mmol/L caffeine exposure (Fig. in relation to coffee consumption among women with ER 3A). In contrast, caffeine exposure resulted in reexpression of ER tumors treated with aromatase inhibitors, chemotherapy, or and cyclin D1 in ER MDA-MB-231 cells (Fig. 3B). However, radiotherapy (all adjusted Ps 0.26). In vitro evaluation of the caffeine did not under these conditions reactivate ER protein effects on cell proliferation of the combined exposure of tamox- expression to the endogenous ER levels seen in untreated MCF- ifen and caffeine or caffeic acid showed that the response to the 7 cells. coffee constituents were maintained and additive to the inhibitory þ Caffeine exposure was further associated with a 70% and a 25% effects of tamoxifen in both ER MCF-7 and T47D breast cancer reduction in IGFIR abundance in MCF-7 (Fig. 3A) and MDA-MB- cells (data not shown). 231 cells (Fig. 3B), respectively, when compared with control. Increasing doses of caffeine resulted in a concomitant dose- dependent and complete inhibition of AktSer473 phosphorylation Discussion in both MCF-7 and MDA-MB-231 cells (Fig. 3). In line with the A growing body of literature indicates that modifiable dietary or lower magnitude in growth inhibition by caffeic acid, a 30% lifestyle factors, such as coffee, may have a protective role against reduction in IGFIR levels was observed for MCF-7, but not for the risk or progression of breast cancer (14, 15, 17). In the present MDA-MB-231 cells, although a 50% reduction in Akt phos- translational study, we demonstrate an association between phorylation was found for both cell types following caffeic acid increasing coffee consumption and significantly smaller invasive þ exposure. These results demonstrate that caffeine and caffeic acid tumor sizes, whereas concomitantly lower proportion of ER significantly alter ER and cyclin D1 levels, and suppress the IGFIR– tumors among women with breast cancer. Moreover, in the þ pAkt signaling pathway in both ER and ER breast cancer cells. present study, mechanistic in vitro data show that caffeine and þ caffeic acid attenuated the growth of predominantly ER , but also Coffee consumption and reduced risk of early breast cancer ER , human breast cancer cells by modulating ER and IGFIR levels recurrence with impact on downstream effectors and cell-cycle progression. Finally, impact by coffee consumption on breast cancer prog- Finally, the present study demonstrate that tamoxifen-treated þ nosis was evaluated. Patients were followed for up to 9 years women with ER tumors drinking 2 or more cups of coffee per (median 3.0 years, IQR 1.0–5.1). During the study period, 100 of day show a significantly reduced risk of early breast cancer 1,082 patients were diagnosed with breast cancer recurrence, of recurrence, compared with women with low daily coffee them 65 patients had distant metastases. In total 76 deaths consumption. occurred, of which 45 had reported breast cancer event before Several epidemiologic reports suggest that coffee consumption death. Coffee consumption alone was not associated with disease decreases breast cancer risk (14, 25, 26), whereas other studies prognosis overall or when stratifying according to ER status (all report no association between coffee intake and breast cancer risk adjusted Ps 0.22). However, the disease-free survival was (27). However, relatively few studies have examined the associ- significantly improved among tamoxifen-treated women with ation with risk or progression by ER status. Sweden is among the þ ER tumors drinking 2 or more cups of coffee per day (log-rank top coffee consuming countries worldwide, and less than 1% of P ¼ 0.013; Fig. 4). Among these women there was a 49% decreased the coffee consumed is decaffeinated (28). Coffee contains a

www.aacrjournals.org Clin Cancer Res; 21(8) April 15, 2015 OF5

Rosendahl et al.

Figure 1. Caffeine or caffeic acid reduces human breast cancer cell growth in vitro.ERþ (MCF-7), ER (MDA-MB-231), and Tam-R breast cancer cells were treated with the indicated concentrations of caffeine (Caff) or caffeic acid (CaA) in SFM for 48 (A and B) or 72 hours (C). Proliferation (A and B) was assessed using [3H]thymidine incorporation as a measurement of DNA synthesis. Total cell number and cell death were assessed by cell counting (A). The graphs represent the mean SE of three (A) or ﬁve (B) determinations; , P < 0.05; , P < 0.01; , P < 0.001, the Student t test. Total protein (C) was extracted and analyzed by immunoblotting for the expression of Bcl-2, Bcl-xL, cleaved caspase-7, and cleaved PARP. GAPDH is shown as loading control. One representative image from at least three independent experiments is shown.

mixture of bioactive components, and there are several plausible subset of the present cohort (17). When stratifying according to ER biologic pathways whereby coffee might alter breast cancer risk or status, the association between increasing coffee consumption þ progression. Caffeine and the polyphenol constituents have been and smaller invasive tumor size was only observed in ER tumors, suggested to contribute to its anticancer activities. However, suggesting that coffee suppresses breast tumor growth, and that þ þ various coffee constituents may differentially affect ER and ER tumors may be more sensitive to these effects. A previous ER breast cancer subtypes. The present findings demonstrate report demonstrated a protective effect of caffeine intake on both þ smaller invasive tumor sizes, whereas concomitant lower propor- ER and ER breast cancer risk in the Nurses' Health Study, þ þ tion of ER tumors and consequently higher proportion of ER although only significant for the ER group (25). Although at tumors with increasing coffee consumption. These findings from the same time, no association with caffeine intake and overall almost twice as many women, confirm the previous report from a breast cancer risk in the Women's Health Study was reported, the

OF6 Clin Cancer Res; 21(8) April 15, 2015 Clinical Cancer Research

Impact of Coffee on Breast Cancer

Figure 2. Cell-cycle distribution after exposure to caffeine or caffeic acid. Representative images showing the morphology and cell-cycle distribution of MCF-7 (A) and MDA-MB-231 cells (B) after 48-hour exposure to the indicated concentrations of caffeine (Caff) or caffeic acid (CaA). Floating and adherent cells were collected and subjected to cell-cycle analysis using ﬂow cytometry. Graph shows cell-cycle distribution of MCF-7 (C) and MDA-MB-231 cells (D) with the percentage of cells in sub-G1 (apoptotic), G0–G1,S,andG2–M phases, as the mean of at least three individual repeats.

þ proportion of ER tumors was decreased and the proportion ER JB6 cells, caffeine attenuated cell growth via induction of G0–G1 tumors increased (29). The shifted ER proportions are in line with cell-cycle arrest (32, 33). Furthermore, caffeine-mediated inhibi- both the clinical and experimental data in the present study. tion of metastatic potential have been demonstrated for melano- Several studies have described caffeine and caffeic acid as ma cells (34). Caffeine has also been shown to inhibit skin cancer anticarcinogenic agents in various cell culture and in vivo models in mice by enhancing UVB-induced apoptosis (35). The experi- (30, 31). For instance, in hepatocellular carcinoma and epidermal mental doses used in the present study are equivalent to previous

www.aacrjournals.org Clin Cancer Res; 21(8) April 15, 2015 OF7

Rosendahl et al.

OF8 Clin Cancer Res; 21(8) April 15, 2015 Clinical Cancer Research

Impact of Coffee on Breast Cancer

tamoxifen, inhibits estrogen-induced breast cancer growth by competitively binding ER, leading to inhibition of ER transcriptional activity and targeting ER to proteasomal degradation (37). In addition, a recent study described that a derivative of caffeic acid (CAPE) specifically bound and downregulated ER expression (38). þ Caffeine affected ER levels differently in ER versus ER breast cancer cell lines. Surprisingly, the caffeine and caffeic acid–induced growth inhibition of ER MDA-MB-231 cells was associated with a reexpression of ER protein abundance. MDA- MB-231 cells have densely methylated ER CpG islands and epigenetic repression via promoter methylation has been suggested to be the predominant mechanism for ER downregulation in ER tumors (39). Caffeic acid has been reported to influence epigenetic repression and to reactivate methylation- silenced genes (40). However, the modulation of ER levels in the present study appeared to involve other mechanisms as initial investigation of the methylation of the ER promoter using Combined Bisulphite Restriction Assay revealed no effect of the coffee constituents (data not shown). Further studies are needed to identify whether other epigenetic changes such as histone acetylation played a role or whether nonepigenetic Figure 4. mechanisms were involved in the caffeine and caffeic acid Impact by coffee consumption on disease-free survival among tamoxifen- induce reexpression of ER in ER cells. þ – treated women with ER breast cancer. Kaplan Meier estimates of disease- ER directly activates the cyclin D1 promoter, and has been free survival in relation to coffee consumption for 506 tamoxifen-treated women with ERþ breast cancer are shown (log-rank 1 df; P ¼ 0.013). The demonstrated to be capable of stimulating cell-cycle progression adjusted HR was 0.51 (95% CI, 0.24–0.97; P ¼ 0.040). The number of patients in the absence of peptide growth factors (41). In addition, cyclin at each follow-up is indicated. Because this is an ongoing study, the number of D1 can stimulate ER transcriptional activity in the absence of þ patients decreases with each follow-up. estrogen (42). The caffeine-mediated ER downregulation in ER cells was associated with a concomitant cyclin D1 reduction, in vitro – studies (30 33), although the higher doses are above the impaired G1 to S cell-cycle transition, reduced growth, and induc- physiologic achievable systemic doses. Translating experimental tion of apoptosis. This is potentially important as cyclin D1 gene in vitro doses to human in vivo settings frequently requires signif- amplification and protein overexpression is associated with icant dose adjustments due to pharmacokinetic and pharmaco- impaired tamoxifen response (43), and if coffee constituents dynamic parameters along with differences in acute in vitro modify cyclin D1 levels this may potentially augment tamoxifen exposure versus cumulative effects following long-term exposure. treatment response. However, cyclin D1 appears to have contrast- þ Our present results demonstrate that predominantly caffeine, and ing effects in ER and ER cells. Although overexpression of cyclin þ to a lesser extent caffeic acid, inhibited the proliferation of ER D1 may promote treatment resistance and a metastatic phenotype þ MCF-7 and ER MDA-MB-231 human breast cancer cells. In line in ER cells, it appears to decrease migration and mammosphere þ with our clinical observations, the ER breast cancer cells formation in ER breast cancer cells (43–45). Such contrasting appeared more sensitive to the growth inhibitory effects by the effects may explain the divergent regulation of ER and cyclin D1 þ coffee constituents compared with ER cells. Interestingly, the by caffeine in ER versus ER cells, both resulting in growth þ growth inhibitory property by caffeine on ER MCF-7 cells was inhibition. associated with a significant reduction in ER abundance. A mod- In addition to modulating ER and cyclin D1 levels, caffeine erate effect was also found for caffeic acid. Downregulation of ER treatment dose-dependently downregulated the growth and anti- function and expression with subsequent growth inhibition has apoptotic IGFIR levels with subsequent impaired AktSer473 phos- previously been demonstrated for green tea polyphenolic cate- phorylation in both MCF-7 and MDA-MB-231 cells, which may chenins (EGCC) and the structurally related plant flavonoid partially contribute to the reduced cell proliferation observed. wogonin (12, 13). In a similar manner to these dietary constitu- This is to our knowledge, the first study to show a direct effect of ents, caffeine mimicked the actions of the antiestrogen fulvestrant caffeine on IGFIR regulation. Previously, inhibition of IGFIR (ICI 182,780), which inhibits ER-dependent functions and signaling by green tea polyphenols as measured by reduced pAkt decreases ER expression, resulting in suppression of tumor cell and pErk1/2, has been shown to reduce prostate cancer progres- growth (36). Similarly, endoxifen, the bioactive metabolite of sion and invasion in the TRAMP mouse model (11). The IGF and

Figure 3. Altered hormone receptor levels and downstream effectors by coffee constituents. ERþ MCF-7 (A) and ER MDA-MB-231 cells (B) were treated with the indicated concentrations of caffeine (Caff) or caffeic acid (CaA) in SFM for 72 hours. Total protein was extracted and analyzed by immunoblotting for the expression of ER, cyclin D1, IGFIR, pAkt, and Akt. GAPDH is shown as loading control. Relative protein abundance was quantiﬁed by densitometry analysis and corrected for GAPDH (ER, cyclin D1, and IGFIR) or total Akt (pAkt). The graphs show modulation of relative protein abundance versus untreated control and represent the mean SE of at least three independent experiments; , P < 0.05; , P < 0.01; , P < 0.001, the Student t test.

www.aacrjournals.org Clin Cancer Res; 21(8) April 15, 2015 OF9

Rosendahl et al.

estrogen growth regulatory pathways are tightly linked and com- Authors' Contributions pensatory mechanisms by the IGFIR have been suggested to Conception and design: A.H. Rosendahl, C.M. Perks, C. Ingvar, H. Jernstrom€ mediate cell proliferation in tamoxifen-resistant tumors (46). Development of methodology: A.H. Rosendahl, C.M. Perks, L. Zeng, J.M.P. € Therapies targeting IGFIR may, thus, be of clinical benefit for Holly, H. Jernstrom Acquisition of data (provided animals, acquired and managed patients, treatment of such cancers. The association between increasing provided facilities, etc.): A.H. Rosendahl, L. Zeng, A. Markkula, M. Simonsson, coffee consumption and reduced risk of early breast cancer þ C. Ingvar, J.M.P. Holly recurrence among tamoxifen-treated women with ER tumors in Analysis and interpretation of data (e.g., statistical analysis, biostatistics, the present study may mechanistically be linked to the altered ER computational analysis): A.H. Rosendahl, L. Zeng, C. Rose, C. Ingvar, J.M.P. and impaired IGFIR signaling by enhanced exposure to the coffee Holly, H. Jernstrom€ constituents. This could sensitize the tumor cells to further inhib- Writing, review, and/or revision of the manuscript: A.H. Rosendahl, C.M. Perks, A. Markkula, M. Simonsson, C. Rose, C. Ingvar, J.M.P. Holly, H. Jernstrom€ itory effects by tamoxifen with improved clinical response rate. fi Administrative, technical, or material support (i.e., reporting or organizing These ndings are consistent with and support the previous report data, constructing databases): A.H. Rosendahl, A. Markkula, M. Simonsson, showing an association between increasing coffee consumption C. Ingvar, J.M.P. Holly, H. Jernstrom€ and significantly increased disease-free survival among tamoxi- Study supervision: A.H. Rosendahl, C. Ingvar, H. Jernstrom€ þ fen-treated women with ER tumors, in a subset of the present Other (PI of the BC Blood study): H. Jernstrom€ cohort (17). The present cell work demonstrates that caffeine and caffeic acid have inhibitory effects that are at least additive with the Acknowledgments effects of tamoxifen. It appears that the inhibitory effects of The authors thank the research nurses Maj-Britt Hedenblad, Karin Henriks- þ son, Anette Moller,€ Monika Meszaros, Anette Ahlin Gullers, Linda Ågren, and caffeine and caffeic acid on ER cells are maintained and not Jessica Åkesson. The authors also thank Erika Bageman, Maria Henningson, and diminished when the ER is either inactive or effectively blocked. Maria Hjertberg for data entry. Furthermore, it is possible that coffee consumption could alter the pharmacokinetics of tamoxifen via, for example, CYP1A2 or Grant Support CYP2C8 (17), but that would require examination in an appro- in vivo This study was supported by grants from the Swedish Cancer Society CAN priate model. 2011/497, the Swedish Research Council K2012-54X-22027-01-3 (PI, In summary, this study shows inhibitory effects by caffeine H. Jernstrom),€ the Faculty of Medicine at Lund University, the Unit of Nursing and caffeic acid on breast cancer cell growth. These results are in Research at Malmo€ University, the Faculty of Health and Society at Malmo€ line with previous epidemiologic reports demonstrating pro- University, the South Swedish Health Care Region (Region Skane ALF), the tective roles of coffee constituents and the risk or progression of Lund Hospital Fund, the Swedish Breast Cancer Group (BRO), the Mrs. Berta breast cancer. The present findings enhance the general under- Kamprad Foundation, the Crafoord Foundation, the Gunnar Nilsson Founda- fi tion, King Gustaf V's Jubilee Foundation, the Maja and Hjalmar Leander Fund, standing of how modi able factors present in our daily diet, and the Royal Physiographic Society in Lund. such as coffee constituents, may alter ER status, IGFIR levels, The costs of publication of this article were defrayed in part by the payment of þ and contribute to reduced growth of ER and ER breast cancer page charges. This article must therefore be hereby marked advertisement in cells. accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Disclosure of Potential Conﬂicts of Interest Received July 8, 2014; revised December 9, 2014; accepted February 2, 2015; No potential conﬂicts of interest were disclosed. published OnlineFirst February 17, 2015.

References 1. Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer 9. Lee AV, Weng CN, Jackson JG, Yee D. Activation of estrogen receptor- statistics. CA Cancer J Clin 2011;61:69–90. mediated gene transcription by IGF-I in human breast cancer cells. 2. Andersen J, Poulsen HS. Immunohistochemical estrogen receptor deter- J Endocrinol 1997;152:39–47. mination in parafﬁn-embedded tissue. Prediction of response to hormonal 10. Zeng L, Holly JM, Perks CM. Effects of physiological levels of the green tea treatment in advanced breast cancer. Cancer 1989;64:1901–8. extract epigallocatechin-3-gallate on breast cancer cells. Front Endocrinol 3. Clarke R, Leonessa F, Welch JN, Skaar TC. Cellular and molecular phar- 2014;5:61. macology of antiestrogen action and resistance. Pharmacol Rev 2001; 11. Adhami VM, Siddiqui IA, Ahmad N, Gupta S, Mukhtar H. Oral consump- 53:25–71. tion of green tea polyphenols inhibits insulin-like growth factor-I-induced 4. Yan L, Yang X, Davidson NE. Role of DNA methylation and histone signaling in an autochthonous mouse model of prostate cancer. Cancer Res acetylation in steroid receptor expression in breast cancer. J Mammary 2004;64:8715–22. Gland Biol Neoplasia 2001;6:183–92. 12. Chung H, Jung YM, Shin DH, Lee JY, Oh MY, Kim HJ, et al. Anticancer 5. Casa AJ, Potter AS, Malik S, Lazard Z, Kuiatse I, Kim HT, et al. Estrogen and effects of wogonin in both estrogen receptor-positive and -negative human insulin-like growth factor-I (IGF-I) independently downregulate critical breast cancer cell lines in vitro and in nude mice xenografts. Int J Cancer repressors of breast cancer growth. Breast Cancer Res Treat 2012;132:61–73. 2008;122:816–22. 6. Casa AJ, Dearth RK, Litzenburger BC, Lee AV, Cui X. The type I insulin-like 13. Farabegoli F, Barbi C, Lambertini E, Piva R. (-)-Epigallocatechin-3-gallate growth factor receptor pathway: a key player in cancer therapeutic resis- downregulates estrogen receptor alpha function in MCF-7 breast carcino- tance. Front Biosci 2008;13:3273–87. ma cells. Cancer Detect Prev 2007;31:499–504. 7. Lee AV, Jackson JG, Gooch JL, Hilsenbeck SG, Coronado-Heinsohn E, 14. Kotsopoulos J, Ghadirian P, El-Sohemy A, Lynch HT, Snyder C, Daly M, Osborne CK, et al. Enhancement of insulin-like growth factor signaling in et al. The CYP1A2 genotype modiﬁes the association between coffee human breast cancer: estrogen regulation of insulin receptor substrate-1 consumption and breast cancer risk among BRCA1 mutation carriers. expression in vitro and in vivo. Mol Endocrinol 1999;13:787–96. Cancer Epidemiol Biomarkers Prev 2007;16:912–6. 8. Freiss G, Rochefort H, Vignon F. Mechanisms of 4-hydroxytamoxifen anti- 15. Bageman E, Ingvar C, Rose C, Jernstrom H. Coffee consumption and growth factor activity in breast cancer cells: alterations of growth factor CYP1A21F genotype modify age at breast cancer diagnosis and estro- receptor binding sites and tyrosine kinase activity. Biochem Biophys Res gen receptor status. Cancer Epidemiol Biomarkers Prev 2008;17: Commun 1990;173:919–26. 895–901.

OF10 Clin Cancer Res; 21(8) April 15, 2015 Clinical Cancer Research

Impact of Coffee on Breast Cancer

16. Fleisher B, Unum J, Shao J, An G. Ingredients in fruit juices interact with 32. Okano J, Nagahara T, Matsumoto K, Murawaki Y. Caffeine inhibits the dasatinib through inhibition of BCRP: a new mechanism of beverage-drug proliferation of liver cancer cells and activates the MEK/ERK/EGFR signal- interaction. J Pharm Sci 2015;104:266–75. ling pathway. Basic Clin Pharmacol Toxicol 2008;102:543–51. 17. Simonsson M, Soderlind V, Henningson M, Hjertberg M, Rose C, Ingvar C, 33. Hashimoto T, He Z, Ma WY, Schmid PC, Bode AM, Yang CS, et al. Caffeine et al. Coffee prevents early events in tamoxifen-treated breast cancer inhibits cell proliferation by G0/G1 phase arrest in JB6 cells. Cancer Res patients and modulates hormone receptor status. Cancer Causes Control 2004;64:3344–9. 2013;24:929–40. 34. Gude RP, Menon LG, Rao SG. Effect of Caffeine, a xanthine derivative, in 18. McShane LM, Altman DG, Sauerbrei W, Taube SE, Gion M, Clark GM. the inhibition of experimental lung metastasis induced by B16F10 mel- Reporting recommendations for tumor marker prognostic studies anoma cells. J Exp Clin Cancer Res 2001;20:287–92. (REMARK). J Natl Cancer Inst 2005;97:1180–4. 35. Lu YP, Lou YR, Xie JG, Peng QY, Zhou S, Lin Y, et al. Caffeine and caffeine 19. Ringberg A, Bageman E, Rose C, Ingvar C, Jernstrom H. Of cup and bra size: sodium benzoate have a sunscreen effect, enhance UVB-induced apoptosis, reply to a prospective study of breast size and premenopausal breast cancer and inhibit UVB-induced skin carcinogenesis in SKH-1 mice. Carcinogen- incidence. Int J Cancer 2006;119:2242–3. esis 2007;28:199–206. 20. Markkula A, Bromee A, Henningson M, Hietala M, Ringberg A, Ingvar C, 36. Oliveira CA, Nie R, Carnes K, Franca LR, Prins GS, Saunders PT, et al. The et al. Given breast cancer, does breast size matter? Data from a prospective antiestrogen ICI 182,780 decreases the expression of estrogen receptor- breast cancer cohort. Cancer Causes Control 2012;23:1307–16. alpha but has no effect on estrogen receptor-beta and androgen receptor in 21. Jernstrom H, Bageman E, Rose C, Jonsson PE, Ingvar C. CYP2C8 and rat efferent ductules. Reprod Biol Endocrinol 2003;1:75. CYP2C9 polymorphisms in relation to tumour characteristics and early 37. Wu X, Hawse JR, Subramaniam M, Goetz MP, Ingle JN, Spelsberg TC. The breast cancer related events among 652 breast cancer patients. Br J Cancer tamoxifen metabolite, endoxifen, is a potent antiestrogen that targets 2009;101:1817–23. estrogen receptor alpha for degradation in breast cancer cells. Cancer Res 22. Morgan L, Gee J, Pumford S, Farrow L, Finlay P, Robertson J, et al. Elevated 2009;69:1722–7. Src kinase activity attenuates Tamoxifen response in vitro and is associated 38. Jung BI, Kim MS, Kim HA, Kim D, Yang J, Her S, et al. Caffeic acid phenethyl with poor prognosis clinically. Cancer Biol Ther 2009;8:1550–8. ester, a component of beehive propolis, is a novel selective estrogen 23. Clark MA, Perks CM, Winters ZE, Holly JM. DNA damage uncouples the receptor modulator. Phytother Res 2010;24:295–300. mitogenic response to IGF-I in MCF-7 malignant breast cancer cells by 39. Yang X, Phillips DL, Ferguson AT, Nelson WG, Herman JG, Davidson NE. switching the roles of PI3 kinase and p21WAF1/Cip1. Int J Cancer 2005; Synergistic activation of functional estrogen receptor (ER)-alpha by DNA 116:506–13. methyltransferase and histone deacetylase inhibition in human ER-alpha- 24. Rosendahl AH, Gundewar C, Said K, Karnevi E, Andersson R. Celecoxib negative breast cancer cells. Cancer Res 2001;61:7025–9. synergizes human pancreatic ductal adenocarcinoma cells to sorafenib- 40. Lee WJ, Zhu BT. Inhibition of DNA methylation by caffeic acid and induced growth inhibition. Pancreatology 2012;12:219–26. chlorogenic acid, two common catechol-containing coffee polyphenols. 25. Ganmaa D, Willett WC, Li TY, Feskanich D, van Dam RM, Lopez-Garcia E, Carcinogenesis 2006;27:269–77. et al. Coffee, tea, caffeine and risk of breast cancer: a 22-year follow-up. Int J 41. Sabbah M, Courilleau D, Mester J, Redeuilh G. Estrogen induction of the Cancer 2008;122:2071–6. cyclin D1 promoter: involvement of a cAMP response-like element. Proc 26. Li J, Seibold P, Chang-Claude J, Flesch-Janys D, Liu J, Czene K, et al. Coffee Natl Acad Sci U S A 1999;96:11217–22. consumption modiﬁes risk of estrogen-receptor negative breast cancer. 42. Zwijsen RM, Wientjens E, Klompmaker R, van der Sman J, Bernards R, Breast Cancer Res 2011;13:R49. Michalides RJ. CDK-independent activation of estrogen receptor by cyclin 27. Gierach GL, Freedman ND, Andaya A, Hollenbeck AR, Park Y, Schatzkin A, D1. Cell 1997;88:405–15. et al. Coffee intake and breast cancer risk in the NIH-AARP diet and health 43. Jirstrom K, Stendahl M, Ryden L, Kronblad A, Bendahl PO, Stal O, et al. study cohort. Int J Cancer 2012;131:452–60. Adverse effect of adjuvant tamoxifen in premenopausal breast cancer with 28. Federation EC. European Coffee Report 2012/2013. 2014; Available from: cyclin D1 gene ampliﬁcation. Cancer Res 2005;65:8009–16. http://www.ecf-coffee.org/images/European_Coffee_Report_2012- 44. KennyFS,HuiR,MusgroveEA,GeeJM,BlameyRW,NicholsonRI,etal. 13_European_chapter.pdf. Overexpression of cyclin D1 messenger RNA predicts for poor prognosis 29. Ishitani K, Lin J, Manson JE, Buring JE, Zhang SM. Caffeine consumption in estrogen receptor-positive breast cancer. Clin Cancer Res 1999;5: and the risk of breast cancer in a large prospective cohort of women. Arch 2069–76. Intern Med 2008;168:2022–31. 45. Lamb R, Lehn S, Rogerson L, Clarke RB, Landberg G. Cell-cycle regulators 30. Ravi D, Muniyappa H, Das KC. Caffeine inhibits UV-mediated NF-kappaB cyclin D1 and CDK4/6 have estrogen receptor-dependent divergent func- activation in A2058 melanoma cells: an ATM-PKCdelta-p38 MAPK-depen- tions in breast cancer migration and stem cell-like activity. Cell Cycle dent mechanism. Mol Cell Biochem 2008;308:193–200. 2013;12:2384–94. 31. Jung JE, Kim HS, Lee CS, Park DH, Kim YN, Lee MJ, et al. Caffeic acid and its 46. Knowlden JM, Hutcheson IR, Barrow D, Gee JM, Nicholson RI. Insulin-like synthetic derivative CADPE suppress tumor angiogenesis by blocking growth factor-I receptor signaling in tamoxifen-resistant breast cancer: a STAT3-mediated VEGF expression in human renal carcinoma cells. Carci- supporting role to the epidermal growth factor receptor. Endocrinology nogenesis 2007;28:1780–7. 2005;146:4609–18.

www.aacrjournals.org Clin Cancer Res; 21(8) April 15, 2015 OF11

Caffeine and Caffeic Acid Inhibit Growth and Modify Estrogen Receptor and Insulin-like Growth Factor I Receptor Levels in Human Breast Cancer

Ann H. Rosendahl, Claire M. Perks, Li Zeng, et al.

Clin Cancer Res Published OnlineFirst February 17, 2015.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-14-1748

Supplementary Access the most recent supplemental material at: Material http://clincancerres.aacrjournals.org/content/suppl/2015/02/18/1078-0432.CCR-14-1748.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/early/2015/03/24/1078-0432.CCR-14-1748. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.