Green Tea Catechins and Vitamin E InhibitAngiogenesis of Human Microvascular Endothelial

Cells Through Suppression of IL-8 Production

Feng-Yao Tang and Mohsen Meydani

Abstract: Epidemiological and animal studies have indi-cated that consumption of green tea and high vitamin E in-take are associated with a reduced risk of developing certainforms of cancer. However, the inhibitory mechanism ofgreen tea catechins and vitamin E in angiogenesis, an im-portant process in tumor growth, has not been well estab-lished. In the present study, �-tocopherol and several majorcatechins of green tea (catechin, epicatechin, epicatechingallate, epigallocatechin, and epigallocatechin gallate)were tested for their ability to inhibit tube formation in vitrousing a model in which human microvascular endothelialcells were exposed to a constant rate of a physiologicallylow level of H2O2. In this model, the production of inter-leukin (IL)-8 by human microvascular endothelial cells at alow level of H2O2 was required for angiogenesis, as assessedby tube formation in three-dimensional gel in culture. Vita-min E (d-�-tocopherol, 40 �M) in the culture media signifi-cantly reduced IL-8 production and angiogenesis. Amongthe green tea catechins, epigallocatechin (0.5–1 �M) wasthe most effective in reducing IL-8 production and inhibitingangiogenesis. These results suggest that consumption ofgreen tea catechins or supplemental intake of vitamin E mayhave preventive effects on tumor development, mediated, atleast in part, through inhibition of angiogenesis via suppres-sion of IL-8 production.

Introduction

Angiogenesis, the formation of new blood vessels frompreexisting blood vessels, is involved in the pathogenesis ofmany chronic diseases, such as cancer, rheumatoid arthritis,and diabetic retinopathy (1–4). The process of angiogenesisincludes basement membrane degradation, endothelial pro-liferation and migration, extracellular matrix remodeling,and vascular tube formation (1). The development of newblood vessels is well controlled by angiogenic and antian-giogenic factors that are produced by normal and tumor cells(5). Several angiogenic factors have been characterized, in-

cluding vascular endothelial growth factor (VEGF), fibro-blast growth factor, epidermal growth factors, transforminggrowth factors, and platelet-derived endothelial cell growthfactor, as well as angiogenin, pleiotrophin, angiotrophin,and interleukin (IL)-1, IL-6, and IL-8 (5–7). On stimulationby angiogenic factors, endothelial cells in tumor blood ves-sels divide rapidly, forming new blood vessels within tumorparenchyma and providing nutrition and oxygen for malig-nant cells (1).

Several antiangiogenic factors have been identified,including interferon-� (8,9), inhibitors of matrix metallopro-teinases, inhibitors of plasminogen activators (10), endo-statin (11), and angiostatin (12,13). In physiologicalconditions, the proliferation of endothelial cells is tightlyregulated and self-limited (5). In certain pathological condi-tions, however, angiogenesis significantly increases andloses its capacity to regulate itself (14). From a clinical per-spective, the most important manifestation of pathologicalangiogenesis may be observed during the development ofsolid tumors (15).

Epidemiological studies and clinical trials have indicatedthat dietary antioxidants may decrease cancer risk. Green teacatechins and vitamin E, which have antioxidant properties,appear to be effective in reducing cancer risk (16–19). Aprospective cohort study in Japan indicated that regulargreen tea consumption reduces the incidence of cancer (16).Epidemiological studies have indicated that regular greentea consumption and high vitamin E intake are associatedwith a lower risk of certain forms of cancer, including sarco-mas, head and neck tumors, and intestinal cancers (16–20).Similar results have been found in vitro and in in vivo ani-mal models. Vitamin E has been shown to inhibit experi-mental carcinogenesis and tumor development (21).Consumption of green tea and vitamin E has been shown toinhibit tumor growth in animals (19,21,22). Catechins ofgreen tea inhibited the growth of a human lung cancer cellline (23). Recently, Cao and Cao (24) reported that the greentea catechin epigallocatechin gallate (EGCG) suppressedbovine endothelial cell proliferation in culture and that

NUTRITION AND CANCER, 41(1&2), 119–125Copyright © 2001, Lawrence Erlbaum Associates, Inc.

The authors are affiliated with the Vascular Biology Laboratory, JM USDA-Human Nutrition Research Center on Aging at Tufts University, Boston, MA02111.

drinking green tea suppressed VEGF-induced angiogenesisin a mouse corneal model (24).

Shono et al. (25) showed that treatment of human micro-vascular endothelial cells (HMVEC) with a single bolusdose of H2O2 (100–500 �M) resulted in microvascular tubeformation in culture. H2O2 is a source of reactive oxygenspecies (ROS), which are believed to be important in the de-velopment of cancer, diabetes, and rheumatoid arthritis (26,27). ROS are generated under various physiological andpathological conditions, including inflammation, ischemiaand reperfusion, sepsis, and ultraviolet irradiation (28). Al-though high levels of ROS are cytotoxic, low levels are nec-essary, inasmuch as they mediate some biological processes(29). For example, several transcription factors use H2O2 asa second messenger for their activation (30). Studies haveshown that antioxidants including �-tocopherol and poly-phenols can influence cellular redox status, which in turnmodulates signaling pathways involved in the regulation ofgene expression, including that of IL-8 (31–35).

Evidence indicates that H2O2 induction of HMVEC tubeformation is mediated by the increased IL-8 production (25).Earlier, we showed that IL-1-stimulated IL-8 production byhuman aortic endothelial cells is suppressed by supplement-ing cells with �-tocopherol (36). Therefore, we hypothe-sized that the suppression of angiogenesis by antioxidantssuch as vitamin E or green tea catechins is mediated in partby the inhibition of IL-8 production under oxidative stressconditions. Therefore, in the present study, we sought to in-vestigate the effect of five major green tea catechins withvarying degrees of antioxidant activity in comparison to vi-tamin E on IL-8 production and HMVEC tube formation inculture. Our results demonstrate for the first time that greentea catechins or vitamin E dose dependently inhibits oxida-tive stress-induced tube formation, in part through inhibitionof IL-8 production.

Materials and Methods

Reagents and Antibodies

MCDB-131 medium and amphotericin B were obtainedfrom Sigma Chemical (St. Louis, MO); phenol red-free me-dium 199 (M199), fetal bovine serum (FBS), L-glutamine,and penicillin/streptomycin from Life Technologies (Gai-thersburg, MD); HMVEC, recombinant human epidermalgrowth factor, bovine brain extract, and hydrocortisone fromClonetics (San Diego, CA); and IL-8, monoclonal anti-IL-8antibody, and VEGF from R & D Systems (Minneapolis,MN). Type I collagen was prepared from rat tail tendons.Green tea catechins were purchased from Funakoshi (To-kyo, Japan). d-�-Tocopherol was purchased from Fluka/Sigma (St. Louis, MO).

Cell Culture and Vascular Tube Formation

HMVEC were cultured in MCDB-131 medium, whichwas changed every 2 days. For the tube formation assay, a

confluent monolayer of HMVEC from Passages 6–9 was cul-tured in phenol red-free M199 on a three dimensional (3-D)gel. 3-D gels were prepared from rat tails (37) containingmainly type I collagen, the major constituent of the peri-capillary connective tissue. The type I collagen was diluted to3 mg/ml, and the pH was neutralized by addition of 1/10 ofthe volume of 10 × M199. Aliquots of 600 �l were added to12-well culture plates and incubated at 37�C until gelatini-zation occurred. HMVEC (105 cells) were seeded on the col-lagen-coated wells. After stimulation with H2O2 or VEGF for72 h (see below), HMVEC were fixed with 0.5 ml ofglutaraldehyde-paraformaldehyde mixture (2.5%) and stainedwith modified May-Gruenwald’s solution (0.25%) (25). Tubeformations on 3-D gel were visualized under a phase-contrastmicroscope (×200), and photomicrographs were documentedby a digital camera (Nikon, Tokyo, Japan). Recorded imageswere analyzed by the NIH Image analyzer program (Scion,Frederick, MD) using a personal computer for the numbersand the total length of tube formation. Tube formation wasdefined as straight cellular extensions joining two cells’masses at branch points. Eight random fields per well wereused for angiogenesis assessment.

Green Tea Catechins and d-�-TocopherolSupplementation

Green tea is the product of fresh tea leaves steamed ordried at elevated temperatures. The high temperatures resultin the tea’s polyphenolic compounds, including flavonols,remaining unoxidized. Catechins and flavonols constitute35–52% of green tea solid extract. We tested the five majorgreen tea catechins, including catechin, epicatechin, epi-catechin gallate, epigallocatechin (EGC), and EGCG. Wesupplemented cell culture medium with each green tea cate-chin up to 2 �M, the highest concentration that can beachieved in the plasma by drinking four cups of green tea(38,39). We also supplemented cells with 20–60 �M vitaminE. The normal range of vitamin E in human plasma is 20–30�M (40). The high concentration of 60 �M in plasma can beachieved by oral supplementation of 200–800 IU of vitaminE per day (41).

Before the induction of angiogenesis by the oxidativestress of H2O2, HMVEC monolayers were incubated withdifferent doses (0, 0.5, 1, and 2 �M) of green tea catechinsfor 24 h. Green tea catechins were incorporated into FBS(the final concentration in the medium was 2%) for 30 minand mixed with phenol-red free M199. After incubation,cells were washed with phosphate-buffered saline before theinduction of angiogenesis with H2O2. d-�-Tocopherol at 20,40, and 60 �M was used to compare green tea catechins’ ef-fect on a known antioxidant. d-�-Tocopherol was incorpo-rated into FBS for 30 min and mixed with phenol-red freeM199. HMVEC monolayers were incubated with mediacontaining different doses of d-�-tocopherol (0, 20, 40, and60 �M) for 24 h. Before stimulation for angiogenesis, cellswere washed with phosphate-buffered saline.

120 Nutrition and Cancer 2001

Experimental Design

Control HMVEC monolayers and cells supplementedwith �-tocopherol or green tea catechins were stimulated forangiogenesis with H2O2 for 72 h. Earlier studies used veryhigh, bolus, and unphysiological levels of H2O2 (100–500�M) that cannot be achieved in vivo (25). Therefore, tobetter simulate the in vivo condition in our in vitro model,we used a glucose-glucose oxidase (G/GO) system (G: 5mM; GO: 0.25 mU/ml) to produce a constant pathophysio-logical level (�30 �M) of H2O2 during the stimulation pe-riod. This level simulates the in vivo level of H2O2 producedby inflammatory cells at the site of inflammation. The pro-duction of H2O2 in culture media was measured by the fluo-rescence method of Ruch et al. (42). VEGF (50 ng/ml) wasused as a positive control to induce angiogenesis.

To investigate the role of IL-8 on oxidative stress-in-duced angiogenesis, HMVEC monolayers and cells supple-mented with �-tocopherol or green tea catechins werestimulated with H2O2 (� 30 �M) for 20 h, and IL-8 releasedinto the supernates was measured by enzyme-linkedimmunosorbent assay. Cellular protein was measured by thebicinchoninic acid Lowry protein assay (Pierce, Rockford,IL). Angiogenesis was also investigated when exogenousIL-8 (20 ng/ml) was added to the culture media for 72 hor when anti-IL-8 monoclonal antibody (1 �g/ml) was co-administered with an H2O2-generating system for 72 h.

Statistics

Each experiment was conducted in triplicate and repeatedtwice. Data were analyzed by unpaired Student’s t-test andexpressed as means � SEM. Significant differences were de-termined at P < 0.05.

Results

Effects of H2O2 and IL-8 on Tube Formation

Exposure of HMVEC to constant low levels of H2O2

(� 30 �M) produced by the G/GO system for 72 h inducedtube formation (Fig. 1A). The addition of IL-8 (20 ng/ml) tothe culture medium also induced tube formation (Fig. 1A).The effect of H2O2 or IL-8 on tube formation was compara-ble to that of VEGF (50 ng/ml), which was used as a positivecontrol in this study (Fig. 1A). The addition of anti-IL-8monoclonal antibody (1 �g/ml) to the medium inhibitedH2O2-induced tube formation (Fig. 1B). These data indicatethat IL-8 is an important mediator of low-level H2O2-in-duced tube formation in HMVEC.

Effect of Vitamin E on Tube Formation and IL-8Production

After supplementation of HMVEC with increasing dosesof �-tocopherol (20, 40, and 60 �M) for 24 h, H2O2-inducedtube formation was reduced and reached a statistically sig-

nificant level at 40 �M (Fig. 2A). �-Tocopherol supple-mentation also reduced production of IL-8 by HMVEC in adose-dependent manner (Fig. 2B). These data demonstratethat vitamin E at supplemental levels suppresses H2O2-in-duced IL-8 production and tube formation in HMVEC inculture.

Effects of Green Tea Catechins on Tube Formation andIL-8 Production

Because IL-8 was also found to be an important mediatorof tube formation in the presence of a low level of H2O2 inthis study, we first tested the effect of green tea catechins onHMVEC production of IL-8 when cells were stimulated byH2O2. As shown in Table 1, four of the five prominentcatechins tested significantly suppressed IL-8 production byH2O2-stimulated HMVEC at 1 �M. EGC at 0.5 �M reducedIL-8 production significantly, representing the most effec-tive IL-8 inhibitor of the five catechins tested. Therefore, wetested the effect of EGC on H2O2-induced tube formation.

Supplementing HMVEC with 0.5 �M EGC suppressedH2O2-induced tube formation by 67% (Fig. 3A). IncreasingEGC concentration in the medium to 1 or 2 �M further sup-

Vol. 41, Nos. 1&2 121

Figure 1. Tube formation induced by pathophysiological levels of H2O2,interleukin-8 (IL-8), and vascular endothelial growth factor (VEGF). A:confluent human microvascular endothelial cells (HMVEC) were stimu-lated by H2O2 (~30 �M), IL-8 (20 ng/ml), or VEGF (50 ng/ml) for 72 h, andtotal length of tube formation was determined by image analyzer program.B: to study role of IL-8, anti-IL-8 antibody (anti-IL-8 Ab, 1 �g/ml) wascoadministrated with H2O2 to confluent HMVEC for 72 h, and total lengthof tube formation was determined. Values are means � SEM of 8 randomlyselected fields in each culture dish, each performed in triplicate and re-peated twice. *, P < 0.05 compared with control.

pressed tube formation, but the results were not significantlydifferent from those observed with 0.5 �M EGC. Productionof IL-8 by HMVEC was significantly reduced with EGCsupplementation in a dose-dependent manner (Fig. 3B).Representative photomicrographs of tube formation incontrol and H2O2-treated HMVEC are shown in Fig. 4, Aand B, respectively. Figure 4C represents H2O2-stimulatedHMVEC supplemented with 2 �M EGC. Because the cellculture medium contained low levels of FBS, which con-tains residual amounts of growth factors, low levels of tubeformation were also observed in control and EGC-supple-mented cells (Fig. 4, A and C, respectively).

Role of IL-8 in Oxidative Stress-Induced TubeFormation

To demonstrate that EGC inhibits tube formation by re-ducing IL-8 production, IL-8 was administered with theH2O2-generating system in EGC-supplemented HMVEC.As shown in Fig. 5, addition of exogenous IL-8 reversed theinhibitory effect of EGC on oxidative stress-induced tubeformation. This finding suggests that the inhibitory effect ofEGC on oxidative stress-induced tube formation occurs viasuppression of endogenous IL-8 production.

122 Nutrition and Cancer 2001

Figure 2. Inhibitory effect of vitamin E on oxidative stress-induced vascu-lar tube formation (A) and IL-8 production (B). Confluent HMVEC were in-cubated with 0, 20, 40, and 60 �M d-�-tocopherol (�-toc) for 24 h at 37�C.HMVEC were then stimulated by H2O2 (~30 �M) for 20 h until measure-ment of IL-8 or for 72 h until measurement of tube formation. Tube forma-tion values are means � SEM of 8 randomly selected fields in each culturedish, each performed in triplicate and repeated twice; IL-8 production valuesare means � SEM in each culture dish, each performed in triplicate and re-peated twice. *, P < 0.05 compared with control; †, P < 0.05 compared withH2O2.

Table 1. Inhibitory Effect of Green Tea Catechins on Oxidative Stress-Induced IL-8 Productiona,b

IL-8, pg/�g protein

C EC ECG EGC EGCG

Control 2.10 � 0.42 2.35 � 0.36 2.11 � 0.42 2.26 � 0.31 2.35 � 0.36H2O2 3.71 � 0.5* 4.62 � 1.61* 4.36 � 0.34* 3.43 � 0.12* 4.62 � 1.61*Catechin (0.5 �M) + H2O2 3.58 � 0.20 4.31 � 0.33 4.48 � 0.21 2.88 � 0.31† 3.8 � 0.86Catechin (1 �M) + H2O2 2.87 � 0.29† 2.37 � 0.73† 2.91 � 0.45† 1.88 � 0.45† 2.95 � 1.23†

a: Values are means � SE. Experiments were performed in triplicate and repeated twice. Confluent human microvascular endothelial cells were incubatedwith 0, 0.5, or 1 �M catechin (C), epicatechin (EC), epicatechin gallate (ECG), epigallocatechin (EGC), or epigallocatechin gallate (EGCG) at 37�C for 24h. Cells were then stimulated with H2O2 (~30 �M) for 20 h, and interleukin-8 (IL-8) was measured by enzyme-linked immunosorbent assay.

b: Statistical significance is as follows: *, P < 0.05 compared with control; †, P < 0.05 compared with H2O2.

Figure 3. Inhibitory effect of epigallocatechin (EGC) on oxidative stress-induced vascular tube formation (A) and IL-8 production (B) in HMVEC.Confluent HMVEC were incubated with 0, 0.5, 1, or 2 �M EGC for 24 h at37�C. HMVEC were then stimulated by H2O2 (~30 �M) for 20 h untilmeasurement of IL-8 or for 72 h until measurement of tube formation. Tubeformation values are means � SEM of 8 randomly selected fields in eachculture dish, each performed in triplicate; IL-8 production values are means� SEM in each culture dish, each performed in triplicate and repeated twice.*, P < 0.05 compared with control; †, P < 0.05 compared with H2O2.

Discussion

In this study, we used an in vitro model of angiogenesis todemonstrate that �-tocopherol and green tea catechins, par-ticularly EGC, inhibit H2O2-induced tube formation byHMVEC via suppression of IL-8 production. These findingsprovide mechanistic insights into the effects that supple-mentation with vitamin E and green tea might have on can-cer prevention through suppression of angiogenesis. Ourfindings on green tea catechins’ inhibition of tube formationby HMVEC in culture are in line with a reduced risk of can-cer with drinking green tea in human and experimental stud-ies (16,43,44), in particular, the recent report demonstratingthe inhibition of angiogenesis by drinking green tea extractin the mouse corneal model of angiogenesis (24). Our obser-vation on �-tocopherol’s inhibition of H2O2-induced tubeformation by HMVEC on a 3-D gel is novel and suggests anadditional mechanism by which this vitamin may reduce therisk of certain forms of cancer (17–21).

Green tea catechins and vitamin E, in addition to beingcapable of scavenging oxygen free radicals (31,45), mightinfluence tumor development by modulating the productionof mediators such as IL-8, as indicated by our data. IL-8plays a role in the formation of new capillaries during tumorgrowth. Studies have indicated that IL-8 may be an impor-tant modulator of angiogenesis-related diseases, includingcancer and arthritis (46–48). A high bolus dose of H2O2 hasbeen reported to be angiogenic by increasing IL-8 produc-tion (25). Endogenous IL-8 has an autocrine or a paracrineeffect on the proliferation and migration of endothelial cellsand the development of new capillaries. IL-8 is produced byother cells as well, including lymphocytes, monocytes, andcancer cells (49,50). Indeed, expression of IL-8 by humanmelanoma cells correlates with their metastatic potential invivo (47). Moreover, human recombinant IL-8 was shownto be potently angiogenic when implanted in rat corneas(46). Our data clearly demonstrate that IL-8 production byHMVEC was increased when cells were exposed to a lowbut constant level of H2O2 in culture and support the notionthat IL-8 is an important factor in angiogenesis. The increasein IL-8 production by H2O2 in HMVEC in the present studywas associated with an increase in tube formation. This wasfurther supported by an increase in tube formation when re-combinant IL-8 was added into the HMVEC culture withoutH2O2 stimulation (Fig. 1A). Furthermore, the addition ofanti-IL-8 antibody to HMVEC culture, which was stimu-lated by H2O2 (Fig. 1B), inhibited tube formation. Thus ourdata clearly demonstrate that production of IL-8 is requiredfor tube formation in this in vitro model of angiogenesis.

To best simulate the in vivo condition of angiogenesis,we selected several criteria for our in vitro system, including1) the cell culture matrix we used was a preparation of 3-Dgels with type I collagen, the predominant collagen in thepericapillary connective tissue (37); 2) for the induction ofoxidative stress, in contrast to an earlier study where thebolus dose of H2O2 was used (25), we used the G/GO systemto generate a constant level of H2O2; 3) we also used apathophysiological level (� 30 �M), the level that would be

Vol. 41, Nos. 1&2 123

Figure 4. Photomicrographs of angiogenesis in HMVEC. A: control HMVEC without H2O2 stimulation. B: HMVEC stimulated with H2O2 for 72 h (note in-creased level of vascular tube formation). C: HMVEC supplemented with 2 �M EGC and stimulated with H2O2 (note reduced level of vascular tube formation).Arrows, vascular tube formation.

Figure 5. EGC-inhibited H2O2-induced angiogenesis occurs through sup-pression of endogenous IL-8 production. Confluent HMVEC were incu-bated with 2 �M EGC for 24 h at 37�C. HMVEC were then stimulated byH2O2 (~30 �M) in presence or absence of anti-IL-8 antibody for 72 h.Values are means � SEM of 8 randomly selected fields in each culture dish,each performed in triplicate and repeated twice. *, P < 0.05 compared withcontrol; †, P < 0.05 compared with H2O2.

found during chronic inflammation (51), whereas the earlierstudy used very high levels (100–500 �M) of H2O2, a condi-tion not representative of an in vivo situation for angio-genesis and tumor development (25); 4) we supplementedcells with media containing levels of green tea catechinscomparable to those found in plasma after the consumptionof four cups of green tea per day (38,39); 5) the cell mediawere supplemented with the levels of �-tocopherol compa-rable to plasma levels that are achievable after supplemen-tation with vitamin E at 200–800 IU on a regular basis(40,41). Inclusion of these criteria provided a more realisticin vitro model to investigate the effects of green tea cate-chins and vitamin E on angiogenesis.

Our data provide the first direct evidence of an anti-angiogenic effect of vitamin E and green tea catechins viasuppression of IL-8. Vitamin E reduced production of IL-8by H2O2-stimulated HMVEC in a dose-response manner.This effect was observed when HMVEC were supplementedwith 20–60 �M �-tocopherol in their culture media. VitaminE suppression of tube formation was significant when cellswere supplemented with 40 or 60 �M, the levels comparableto those found in plasma with vitamin E supplementation of200–800 IU/day.

The observed inhibition of H2O2-induced IL-8 productionand angiogenesis with vitamin E or EGC supplementationwas not due to their direct antioxidant effect on H2O2. In apreliminary experiment, we measured the level of H2O2 pro-duction by the G/GO system in the supernatant of cells sup-plemented with vitamin E or EGC. Even though the level ofH2O2 at the end of the incubation period was low (�15 �M),we did not find a dose-dependent reduction of H2O2 by in-creasing vitamin E (from 20 to 60 �M) or EGC (from 0.5 to2 �M) levels, whereas we found that IL-8 production andtube formation were reduced dose dependently by an in-crease in the level of these compounds in HMVEC (Figs. 2and 3). In addition, we have shown that the inhibition of IL-8production can be observed when human aortic endothelialcells are supplemented with vitamin E and stimulated with aproinflammatory cytokine such as IL-1, rather than H2O2

(36). Therefore, vitamin E and EGC inhibition of oxidativestress-induced tube formation is, in part, mediated throughthe suppression of IL-8 production, rather than a direct inter-action of these antioxidants with H2O2 or with the enzymesystem.

Thus vitamin E may inhibit tumor development not onlyby its antioxidant activity (which could prevent DNA dam-age) but also through the inhibition of angiogenesis via IL-8suppression. Therefore, the results of this study support theepidemiological data indicating an association of a low riskof certain forms of cancer with a high intake of vitamin E.

In the present study, all the green tea catechins reducedIL-8 production by H2O2-stimulated HMVEC. However,EGC was the most effective catechin tested. At 0.5 �M,EGC reduced H2O2-induced angiogenesis by 67%. This wascomparable to the effect of vitamin E (74% reduction) ontube formation when cells were supplemented with 40 �Mvitamin E. Thus, on a molar basis, EGC appears to be more

effective than vitamin E. EGC at 2 �M totally abolishedH2O2-induced tube formation (Fig. 3A). The addition of ex-ogenous recombinant IL-8 to the stimulated cells fully ab-rogated the effect of EGC, further demonstrating thatinhibition of IL-8 is one mechanism by which EGC inhibitsH2O2-induced tube formation (Fig. 5). These data furthersupport the notion that EGC’s inhibition of tube formation isnot mediated solely through its antioxidant effect and elimi-nation of H2O2, but also by suppression of IL-8 production.

Our data provide mechanistic insights for the effect ofgreen tea and vitamin E on the reduction of certain forms ofcancer. Although we have not yet investigated the effects ofa mixture of green tea catechins or a mixture of green teacatechins with vitamin E, evidence from this study indicatesthat a dietary regimen containing green tea catechins andsupplemented with vitamin E might be effective in suppress-ing IL-8 production and angiogenesis. Vitamin E and greentea are dietary components with numerous health benefitsand can be easily incorporated into a daily regimen. VitaminE at 40 �m in plasma is attainable with supplemental intakeof vitamin E (41,52). Furthermore, 0.5–1 �M EGC inplasma can be achieved by drinking four cups of green teaper day (38,39).

Acknowledgments and Notes

The authors thank Catarina Sacristan for technical assistance andStephanie Marco for preparation of the manuscript. This material is basedon work supported by the US Department of Agriculture under AgreementNo. 58-1950-9-001. Any opinions, findings, conclusions, or recommenda-tions expressed in this publication are those of the author(s) and do notnecessarily reflect the view of the US Department of Agriculture. Addresscorrespondence to M. Meydani, Vascular Biology Laboratory, JM USDAHuman Nutrition Research Center on Aging at Tufts University, 711Washington St., Boston, MA 02111. Phone: (617) 556-3126. FAX: (617)556-3224. E-mail: MMeydani@hnrc.tufts.edu.

Submitted 26 January 2001; accepted in final form 16 August 2001.

References

1. Folkman J and Shing Y: Angiogenesis. J Biol Chem 267, 10931–10934, 1992.

2. Pili R, Guo Y, Chang J, Nakanishi H, Martin GR, et al.: Alteredangiogenesis underlying age-dependent changes in tumor growth.JNCI 86, 1303–1314, 1994.

3. Folkman J: Angiogenesis in cancer, vascular, rheumatoid and otherdisease. Nat Med 1, 27–31, 1995.

4. Folkman J: Angiogenesis and breast cancer. J Clin Oncol 12, 441–443,1994.

5. Pepper MS: Positive and negative regulation of angiogenesis: from cellbiology to the clinic. Vasc Med 1, 259–266, 1996.

6. Leek RD, Harris AL, and Lewis CE: Cytokine networks in solid hu-man tumors: regulation of angiogenesis. J Leukoc Biol 56, 423–435,1994.

7. Jacques E, Christensen J, Mooney D, and Polverin P: Vascular endo-thelial growth factor (VEGF)-mediated angiogenesis is associated withenhanced endothelial cell survival and induction of Bcl-2 expression.Am J Pathol 154, 375–384, 1999.

124 Nutrition and Cancer 2001

8. Sidkey YA and Border EC: Inhibition of angiogenesis by interferons:effects of tumor- and lymphocyte-induced vascular responses. CancerRes 47, 5155–5161, 1987.

9. Majewski S, Szmurlo A, Marczak M, Jablonska S, and Bollag W: Syn-ergistic effect of retinoids and interferon-� on tumor-induced angio-genesis: anti-angiogenic effect on HPV-harboring tumor cell lines. IntJ Cancer 57, 81–85, 1994.

10. Fan TP, Jagger R, and Bicknell R: Controlling the vasculature: angio-genesis, anti-angiogenesis and vascular targeting therapy. TrendsPharmacol Sci 16, 57–66, 1995.

11. Dhanabal M, Ramchandran R, Volk R, Stillman IE, Lombardo M, etal.: Endostatin: yeast production, mutants, and antitumor effect in renalcell carcinoma. Cancer Res 59, 189–197, 1999.

12. Maucer HJ, Hanna NN, Beckett MA, Gorskki DH, Staba MJ, et al.:Combined effects of angiostatin and ionizing radiation in antitumortherapy. Nature 394, 287–291, 1998.

13. O’Reilly MS: Angiostatin: an endogenous inhibitor of angiogenesisand of tumor growth. EXS 79, 273–294, 1997.

14. Denekamp J: Vascular attack as a therapeutic strategy for cancer. Can-cer Metastasis Rev 9, 267–282, 1990.

15. Folkman J: Tumor angiogenesis. Adv Cancer Res 43, 175–203, 1985.16. Nakachi K, Suemasu K, Suga K, Takeo T, Imai K, et al.: Influence of

drinking green tea on breast cancer malignancy among Japanese pa-tients. Jpn J Cancer Res 89, 254–261, 1998.

17. Das S: Vitamin E in the genesis and prevention of cancer. Acta Oncol33, 615–619, 1994.

18. el Attar TM and Lin HS: Effect of vitamin C and vitamin E on prosta-glandin synthesis by fibroblasts and squamous carcinoma cells. Pros-taglandins Leukot Essent Fatty Acids 47, 253–257, 1992.

19. Fleshner N, Fair WR, Huryk R, and Heston WD: Vitamin E inhibits thehigh-fat diet promoted growth of established human prostate LNCaPtumors in nude mice. J Urol 161, 1651–1654, 1999.

20. Knekt P, Aromaa A, and Maatela J: Vitamin E and cancer prevention.Am J Clin Nutr 53, 283S–286S, 1991.

21. Shklar G and Schwartz JL: Vitamin E inhibits experimental carcino-genesis and tumour angiogenesis. Eur J Cancer 32, 114–119, 1996.

22. Mukhtar H, Katiyar SK, and Agarwal R: Green tea and skin-anti-carcinogenic effects. J Invest Dermatol 102, 3–7, 1994.

23. Okabe S, Suganuma M, Hayashi M, Sueoka E, Komori A, et al.: Mech-anisms of growth inhibition of human lung cancer cell line, PC-9, bytea polyphenols. Jpn J Cancer Res 88, 639–643, 1997.

24. Cao Y and Cao R: Angiogenesis inhibited by drinking tea. Nature 398,381, 1999.

25. Shono T, Ono M, Izumi H, Jimi S-I, Matsushima K, et al.: Involvementof the transcription factor NF-�B in tubular morphogenesis of humanmicrovascular endothelial cells by oxidative stress. Mol Cell Biol 16,4231–4239, 1996.

26. Dreher D and Junod AF: Role of oxygen free radicals in cancer devel-opment. Eur J Cancer 32A, 30–38, 1996.

27. Roth E: Oxygen free radicals and their clinical implication. Acta ChirHung 36, 302–305, 1997.

28. Pahl HL and Baeuerle PA: Oxygen and the control of gene expression.Bioessays 16, 497–502, 1994.

29. Natarajan V, Taher MM, Roehm B, Parinandi NL, Schmid HO, et al.:Activation of endothelial cell phospholipase D by hydrogen peroxideand fatty acid hydroperoxide. J Biol Chem 268, 930–937, 1993.

30. Schmidt KN, Amstad P, Cerutti P, and Baeuerle PA: Identification ofhydrogen peroxide as the relevant messenger in the activation pathwayof transcription factor NF-�B. Adv Exp Med Biol 387, 63–68, 1996.

31. Hanasaki S, Ogawa S, and Fukui S: The correlation between active ox-ygen scavenging and antioxidative effects of flavonoids. Free RadicBiol Med 16, 845–850, 1994.

32. Palmer HJ and Paulson KE: Reactive oxygen species and antioxidantsin signal transduction and gene expression. Nutr Rev 55, 353–361,1997.

33. Gerritsen M, Carley WW, Ranges GE, Shen C-P, Phan SA, et al.:Flavonoids inhibit cytokine-induced endothelial cell adhesion proteingene expression. Am J Pathol 147, 278–292, 1995.

34. Erl W, Weber C, Wardemann C, and Weber PC: �-Tocopheryl suc-cinate inhibits monocytic cell adhesion to endothelial cells by sup-pressing NF-�B mobilization. Am J Physiol 273, H634–H640, 1997.

35. Faruqi R, Motte C, and DiCorleto PE: �-Tocopherol inhibits agonist-induced monocytic cell adhesion to cultured human endothelial cells. JClin Invest 94, 592–600, 1994.

36. Wu D, Martin KR, Koga T, and Meydani M: Effect of vitamin E on hu-man aortic endothelial cell production of chemokines and adhesion tomonocytes. Atherosclerosis 147, 297–307, 1999.

37. Roberto M and Lelio O: Tumor-promoting phorbol esters induce an-giogenesis in vitro. Cell 42, 469–477, 1985.

38. Nakagawa K, Okuda S, and Miyazawa T: Dose-dependent incorpora-tion of tea catechins, (–)-epigallocatechin-3-gallate and (–)-epigallo-catechin, into human plasma. Biosci Biotechnol Biochem 61, 1981–1985, 1997.

39. Yang CS, Chen L, Lee M-J, Balentine D, Kuo MC, et al.: Blood andurine levels of tea catechins after ingestion of different amounts ofgreen tea by human volunteers. Cancer Epidemiol Biomarkers Prev 7,351–354, 1998.

40. Meydani M and Blumberg JB: Vitamin E. In Boston Nutrition StatusSurvey, Rosenberg I and Hartz S (eds). London: Smith-Gordon, 1992,pp 103–109.

41. Meydani SN, Meydani M, Blumberg JB, Leka LS, Siber G, et al.: Vita-min E supplementation enhances in vivo immune response in healthyelderly: a dose-response study. JAMA 277, 1380–1386, 1997.

42. Ruch W, Cooper P, and Baggiolini M: Assay of hydrogen peroxide bymacrophages and neutrophils with homovanillic acid and horseradishperoxidase. J Immunol Methods 63, 347–357, 1983.

43. Wang ZY, Agarwal R, Bickers DR, and Mukhtar H: Protection againstultraviolet B radiation-induced photocarcinogenesis in hairless miceby green tea polyphenols. Carcinogenesis 12, 1527–1530, 1991.

44. Wang ZY, Huang M-T, Ferraro T, Wong C-Q, Lou Y-R, et al.: Inhibi-tory effect of green tea in the drinking water on tumorigenesis by ultra-violet light and 12-O-tetradecanoylphorbol-13-acetate in the skin ofSKH-1 mice. Cancer Res 52, 1162–1170, 1992.

45. Herbaczynska-Cedro K, Kosiewicz-Wasek B, Cedro K, Wasek W,Panczenko-Kresowska B, et al.: Supplementation with vitamins C andE suppresses leukocyte oxygen free radical production in patients withmyocardial infarction. Eur Heart J 16, 1044–1049, 1995.

46. Koch AE, Polverini PJ, Kunkel SL, Harlow LA, DiPietro LA, et al.:Interleukin-8 as a macrophage-derived mediator of angiogenesis. Sci-ence 258, 1798–1801, 1992.

47. Bar-Eli M: Role of interleukin-8 in tumor growth and metastasis of hu-man melanoma. Pathobiology 67, 12–18, 1999.

48. Koch AE: Angiogenesis: implications for rheumatoid arthritis. Arthri-tis Rheum 4, 951–962, 1998.

49. Kruger S, Krasagakis K, Garbe C, and Diamantstein T: Production ofcytokines by human melanoma cells and melanocytes. Recent ResultsCancer Res 139, 155–168, 1995.

50. Oppenheim JJ, Zachariae CO, Mukaida N, and Matsushima K: Prop-erties of the novel proinflammatory supergene, intercrine cytokinefamily. Annu Rev Immunol 9, 617–648, 1991.

51. Kazmierczak M, Wysocki H, Wykretowicz A, and Minczykowski A:Estimation of hydrogen peroxide plasma levels in patients evaluatedfor coronary heart disease using dipyridamole infusion followed bySEPCT. Coron Artery Dis 6, 65–69, 1995.

52. Meydani SN, Barklund PM, Liu S, Meydani M, Miller RA, et al.: Vita-min E supplementation enhances cell-mediated immunity in healthyelderly subjects. Am J Clin Nutr 52, 557–563, 1990.

Vol. 41, Nos. 1&2 125