Advanced
Dose-dependent Effects of Dietary Folate on Aortic Relaxation and Hepatic C-reactive protein Levels in C57BL/6 Mice
Dose-dependent Effects of Dietary Folate on Aortic Relaxation and Hepatic C-reactive protein Levels in C57BL/6 Mice
Journal of Life Science. 2015. Aug, 25(8): 889-895
Copyright © 2015, Korean Society of Life Science
This is an Open-Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
  • Received : April 28, 2015
  • Accepted : August 18, 2015
  • Published : August 30, 2015
Download
PDF
e-PUB
PubReader
PPT
Export by style
Share
Article
Author
Metrics
Cited by
TagCloud
About the Authors
은희 공
산시다 하
혜란 장
머리엘라 짐
상운 최
이다니모슨 메
eh-kong@hanmail.net

Abstract
Endothelial dysfunction is an initial step in atherosclerosis. B vitamins (B6, B12, and folate) are important contributing factors to vascular homeostasis. Deficiencies in these B vitamins induce cardiovascular diseases by altering vascular homeostasis. Folate plays important roles in nitric oxide homeostasis in the endothelium. To determine the dose-dependent effect of dietary folate on atherosclerosis, we studied aortic relaxation and hepatic C-reactive protein (CRP) levels in C57BL/6 mice. In this study, a total of 54 male C57BL/6, 8-wk old mice were split into 2 dietary groups (control and Western style diet). Each diet group was divided into 3 subgroups according to dietary folate dosage (0.2, 2, and 8 mg/kg). After 18 months, the relaxation response seen in aortic rings from mice fed 0.2 or 2 mg folate/kg in both diet groups. However, the aortic relaxation response was not seen and no differences were observed in mice fed 8mg folate/kg in either diet group ( p <0.05). Hepatic CRP levels at all folate dosages (0.2, 2, 8 mg folate/kg) were higher in the groups fed a Western style diet than in mice fed a control diet ( p =0.035). CRP levels were lower in mice fed 0.2 mg folate/kg than in mice fed 2 or 8 mg folate/kg in both diet groups ( p <0.05). These results indicate that in C57BL/6 mice 0.2 mg folate/kg may be enough to prevent atherosclerosis by inducing the relaxation responses of the aorta and by reducing levels of hepatic CRP, regardless of dietary style.
Keywords
Introduction
The Western style diet is characterized by a high fat content, low level of fiber, and deficiencies in the B vitamins (B6, B12, and folate). These B vitamins are important contributing factors to vascular homeostasis. Deficiencies in these B vitamins are important contributing factors to the high prevalence of cardiovascular diseases by altering vascular homeostasis and stimulating inflammation [7] . Endothelial dysfunction is one of the initial steps in the development of atherosclerosis [18] . It is characterized by a reduced capacity for the production of and availability of nitric oxide (NO) [13] . Folate plays important roles in NO homeostasis of the endothelium [12] . A deficiency in folate in rats elevated liver lipids, protein oxidation, and vascular generation of superoxide radical (O 2 ), which led to decreased NO bioavailability to the vascular system [14 , 17] . Thus, the association between folate and endothelial cell function is of clinical importance. Nevertheless, as of today, few studies have investigated the effect of folate on endothelium-dependent relaxation. Therefore, we hypothesized that dietary folate would be associated with endothelium-dependent relaxation caused by the NO bioavailability. Isolated thoracic aortas are widely used to assess the endothelial ability to release NO [4] . It has also been shown that dietary folate reduces serum C-reactive protein (CRP) levels [14 , 17] . CRP is a systemic marker of inflammation and a predictor of cardiovascular disease (CVD). It is synthesized by the liver and participates in the pathogenesis of atherosclerosis through activation of endothelial cells [20] . Thus, we also hypothesized that dietary folate would decrease hepatic CRP levels. To determine the dose-dependent effect of dietary folate on atherosclerosis, we studied aortic relaxation and hepatic CRP levels in C57BL/6 mice.
Materials and Methods
- Animals
This protocol was reviewed and approved by the Institutional Animal Care and Use Committee of Tufts University (IACUC approval No. 51000-067-02S). A total of 54 C57BL/6 male, 6-wk old mice were purchased from Harlan Teklad (Madison, WI, USA) and housed individually in cages with wire bottoms to prevent coprophagia. On arrival, all mice were fed a control (C) diet for 2 wks ( Table 1 ). The Western style (W) diet contained a high level of fat (20% by weight), and low levels of fiber and methyl-donor nutrients (folic acid, methionine and choline) ( Table 1 ). The mice fed 0.2 mg folate/kg represented the low level of folate commonly found in a Western style human diet. The mice fed 8 mg folate/kg represented the level of folate consumed by a Western human population receiving more than the necessary amount of folate through flour fortification, multivitamins, and fortified cereals
Composition of the control (C) and Western style (W) diets
PPT Slide
Lager Image
Dietary groups: C0.2 (0.2 mg folate/kg control diet), C2 (2 mg folate/kg control diet), C8 (8 mg folate/kg control diet), W0.2 (0.2 mg folate/kg Western style diet), W2 (2 mg folate/kg Western style diet), and W8 (8 mg folate/kg Western style diet). a The estimate of caloric content was based on the standard physiological fuel values for protein, fat, and carbohydrate of 4, 9 and 4, respectively.
At 8 wks age, mice were randomly divided into two dietary groups: C and W diet ( Fig. 1 ). The mice in the C diet group were randomly divided into three dietary subgroups according to folate dosage (n=9/subgroup). One subgroup was fed the C diet containing 0.2 mg folate/kg (C0.2). The second subgroup was fed 2 mg folate/kg (C2), and the third group was fed 8 mg folate/kg (C8) with the C diet. The mice in W group were also randomly divided into three subgroups (n=9/subgroup). The one subgroup consumed 0.2 mg folate/kg (W0.2), another subgroup 2 mg folate/kg (W2), while the third subgroup was fed 8 mg folate/kg (W8) with the W diet. At the end of the 72 wks of dietary treatment, the mice were sacrificed by CO 2 exposure.
PPT Slide
Lager Image
Experimental design of the control (C) and Western-style (W) diets.
- Measurement of aortic relaxation
The thoracic aortas of the mice were rapidly excised and immediately placed in Krebs buffer (119 mmol/l NaCl, 4.7 mmol/l KCl, 24 mmol/l NaHCO 3 , 1.2 mmol/l KH 2 PO 4 , 0.0023 mmol/l EDTA, 1.2 mmol/l MgCl 2 , 11 mmol/l glucose, and 2.5 mmol/l CaCl 2 [pH 7.4]). Loose connective tissues in the adventitia of the thoracic aorta were carefully removed without damage to the endothelial surface. The aorta was cut into cylindrical rings at 2.5 mm intervals from the origin of the left subclavian artery in the direction of the descending thoracic aorta. Aortic rings were mounted on two L-shaped, stainless steel arms of a multi chamber myograph (DMT-USA. Inc., Atlanta, GA, USA). In each chamber, one arm of the myograph was fixed to a forced displacement transducer for continuous recording of the isometric tension, and the other arm was attached to a calibrating device.
Experiments with the mounted aortic rings were performed in an incubating chamber bath containing Krebs buffer with a continuous aeration of oxygen at 37℃ according to the methods of Zhu et al [22] . The aortic segments were allowed to equilibrate in Krebs buffer for 1 hr prior to the start of the experiment. The optimal resting tension in all aortic rings of mice was adjusted to 9 millinewton (mN). Following the equilibration period, the contractile capacity of each aortic ring was examined by applying 1.25 mol/l KCl (Sigma Chemical Co., Saint Louis, MO, USA) and then applied increasing concentrations of phenylephrine (Sigma) (Phe, 0.01-10 μmol/l). The aortic rings were rinsed with Krebs buffer 3 times for 10 min each. After stabilization, the aortic rings were stimulated with 0.3 μmol/l Phe. After contraction caused by Phe, relaxation responses were recorded by applying the increased concentrations of acetylcholine (Sigma) (Ach, 0.01-10 μmol/l), an endothelial-dependent vasodilator. The relaxation responses were expressed as a percentage of the initial pre-contraction level induced by Phe.
- Measurement of hepatic CRP
The livers of the mice were removed, snap-frozen in liquid nitrogen, and stored at −80℃ until analysis. The liver tissues were homogenized in lysis buffer [50 mmol/l Tris-HCl, 10 mmol/l EDTA, 1% Tween 20, 1 mmol/l phenylmethylsulfonyl fluoride, 10 μmol/l leupeptin (pH 7.4), and protease inhibitor] using a PowerGen 125 homogenizer (Fisher Scientific, Pittsburgh, PA, USA) [2 , 9] and incubated on ice for 2 hr. The homogenates were centrifuged at 8000 × g at 4℃ for 30 min. The supernatants were collected for ELISA. The total protein levels were measured using the Pierce BCA TM protein assay kit (Thermo Scientific, Rockford, IL, USA). Hepatic CRP levels were measured by a sandwich-type ELISA (R&D Systems). The commercial antibodies used were rhCRP (R&D Systems), hmpCRP (monoclonal, R&D Systems), and GAM-HRP conjugate (Bio-Rad). A specific anti-human CRP antibody was coated onto microtiter plates at 4℃ and left overnight. Blocking solution was applied for 2 hr at room temperature to block the remaining protein-binding sites in the coated wells, followed by washing (2 times, 1 min each). The diluted supernatant liquid (appropriately 200 μg protein) and anti-human CRP Ab used as the standard were added in triplicate into each well of the plates. After washing (4 times, 1 min each), diluted detection Ab was added into the wells and incubated for 2 h at room temperature. Plates were washed (4 times, 1 min each) and incubated with horseradish peroxidase-conjugated goat anti-mouse IgG (1:1000 dilution, Bio-Rad) for 1 hr at room temperature. Finally, after washing (6 times, 1 min each), the substrate solution was mixed with 30% hydrogen peroxide (Sigma) and 2,2′-azino-di-[3-ethyl-benzothiazoline-6 sulfonic acid] (Abcam Inc., Cambridge, MA, USA) and was added for color development for 1 hr at room temperature. Absorbance was measured at 416 nm on an ELISA microplate reader (FLx 808, Bio-Tek Inc., Winooski, VT, USA). The minimum detectable CRP concentration in the assay was 1.56 μg/l.
- Statistical analysis
Results are presented as mean ± SEM. Vascular responses (Phe contractions and Ach % relaxations) and hepatic CRP levels were compared among dietary subgroups (C0.2, C2, C8, W0.2, W2, and W8). Data were analyzed by Kruskal Wallis test. A relevant post-hoc analysis was performed using a Tukey’s honestly significant difference (HSD) test. A P -value <0.05 was considered statistically significant. SPSS for Windows (version 20.0, SPSS Inc., Chicago, IL, USA) was used for data analyses.
Results
- Relaxation responses of aortic rings
The Ach-induced % relaxations of aortic rings from mice fed either the C or W diet are shown in Fig. 2 . Aortic rings from proximal segments of thoracic aortas were subjected to contraction using Phe (0.3 μmol/l). This was followed by the application of increasing concentrations of Ach (0.01-10 μmol/l). The Ach-induced % relaxation response was seen in aortic rings from mice fed 0.2 or 2 mg folate/kg in both diet groups. However, it was not seen in aortic rings from mice fed 8mg folate/kg in either diet group. The Ach-induced % relaxation did not show a concentration- dependent decrease in aortic rings from mice fed the C8 diet, compared with aortas from mice fed other C diets ( p =0.042) ( Fig. 2A ). The relaxation response did not show in aortic rings from mice fed the W8 diet, compared to those of aortas from mice fed 2 or 8 mg folate/kg with the W diet ( p =0.045) ( Fig. 2B ).
PPT Slide
Lager Image
Acetylcholine (Ach)-induced % relaxation of aortic rings from mice fed control (C) or Western-style (W) diets for 18 mo. Aortic rings from proximal segments of the thoracic aorta were subjected to contraction using phenylephrine (Phe, 0.3 μmol/l). This was followed by the application of increasing concentrations of Ach (0.01-10 μmol/l). The Ach-induced % relaxation response was seen in aortic rings from mice fed 0.2 or 2 mg folate/kg in both diet groups. However, it was not seen in aortic rings from mice fed 8mg folate/kg in either diet group. (A) Ach-induced % relaxation was small in the aortic rings of mice fed the C8 diet, compared to that of aortas from mice fed other C diets (p=0.042). (B) It was also less in the aortic rings of mice fed the W8 diet, compared to that of aortas from mice fed 2 or 8 mg folate/kg on the W diet (p=0.045). Dietary groups: C0.2 (0.2 mg folate/kg control diet), C2 (2 mg folate/kg control diet), C8 (8 mg folate/kg control diet), W0.2 (0.2 mg folate/kg Western style diet), W2 (2 mg folate/kg Western style diet), and W8 (8 mg folate/kg Western style diet). n=9/diet group. Values are mean ± SEM. * p<0.05 was obtained by Kruskal Wallis test.
- Hepatic CRP
Hepatic CRP levels from mice fed the C or W diets are shown in Fig. 3 . Hepatic CRP levels were higher in the groups fed the W diet at all folate dosages (0.2, 2, and 8 mg folate/kg) than in groups fed the C diet ( p =0.035). Levels were lower in mice fed the C0.2 diet than in mice fed 2 or 8 mg folate/kg on the C diet ( p =0.041). Levels were also lower in mice fed the W0.2 diet than in mice fed other W diets ( p =0.030).
PPT Slide
Lager Image
Hepatic C-reactive protein (CRP) levels in mice fed control (C) or Western style (W) diets for 18 mo. Mice fed a W diet, irrespective of folate dose (0.2, 2, 8 mg folate/kg), had increases their hepatic CRP levels compared to mice that consumed a C diet (p=0.035). CRP was lower in mice fed C0.2 than in mice in the 2 or 8 mg folate/kg C diet groups (p=0.041). CRP was also lower in mice fed W0.2 than in mice fed other W diets (p=0.030). Dietary groups: C0.2 (0.2 mg folate/kg control diet), C2 (2 mg folate/kg control diet), C8 (8 mg folate/kg control diet), W0.2 (0.2 mg folate/kg Western style diet), W2 (2 mg folate/kg Western style diet), and W8 (8 mg folate/kg Western style diet). n=9/diet group. Values are mean ± SEM. * p<0.05 was obtained by Kruskal Wallis test.
Discussion
Aortas from mice fed 0.2 or 2 mg folate/kg along with either a W or C diet showed Ach-% relaxation responses in this study. We also found that hepatic CRP levels were low in mice fed 0.2 mg folate/kg along with either the W or C diet.
Relaxation of the normal rabbit aorta in response to Ach is mediated in large part by NO [6] . The presence of sufficient amounts of the active metabolites of folate, i.e., 5-methyltetrahydrofolate (5-MTHF), is necessary for activation of NO synthase (eNOS) to produce NO [16] . However, a longterm overdose of folate may lead to eNOS uncoupling from its substrate, which results in decreased NO formation [21] . Our observation that a long-term overdose of folate (C8 and W8 diets) has a negative impact on vascular relaxation is in accordance with the above notion. In support of this concept, Lentz and colleagues reported that long-term B vitamin supplementation did not prevent endothelial dysfunction and atherosclerosis in monkeys fed an atherogenic diet [11] . In placebo-controlled, randomized clinical trials, healthy individuals who consumed folate supplements for 1 to 2 years showed no improvement in the stiffness of the common carotid artery [19] .
CRP is synthesized by the liver and secreted into the circulation in response to acute changes in diet, toxins, drugs, and inflammatory diseases [3] . CRP has been reported to impair endothelial vasoreactivity by inhibiting NO production through down-regulating eNOS activity [10] . This was suggested to occur through the uncoupling of eNOS via increasing reactive oxygen species, which leads to decreased dimerization and phosphorylation of eNOS at Ser-117 [15] . In the present study, we measured CRP levels in the liver where it is mainly produced. Regardless of folate dosage in the diet, the levels of hepatic CRP in mice fed W diets were higher than in mice fed C diets. A greater amount of adipose tissue contributes to systemic inflammation and high levels of CRP. This association between being overweight and having high levels of CRP has been reported in humans [20] . In this study, the higher levels of hepatic CRP may have been related to higher body weights (BW) and larger amounts of adipose tissue in mice fed the W diet (BW=51.1±2.1 g), compared to mice fed the C diet (BW=46.8±1.82 g).
To our knowledge, the association of hepatic CRP in humans with dietary folate dosage has not been reported. Therefore, this is the first time that reduced hepatic CRP levels have been reported to be associated with a low folate dose (0.2 mg folate/kg). Our novel observation will need to be confirmed in further studies.
The present study had several limitations. We were not able to measure the vasocontracting or vasodilating factors from the endothelium. In addition to NO, the vascular endothelium is known to release other vasodilator substances, such as prostacyclin and endothelium-derived hyperpolarizing factor [1 , 5 , 6] . Thus, we cannot exclude the possibility that an increase in the release of contracting factors from the endothelium or an increase in the sensitivity of vascular smooth muscle cells to endothelium-derived contracting factors may have also occurred. The essential mechanisms behind this discrepancy need to be explored further. Moreover, a prospective study with a more representative population is warranted in order to validate our results.
The present study showed that it is important to find the optimum concentration relative to the target. 0.2 mg folate/kg in C57BL/6 mice may be enough to prevent atherosclerosis by inducing the relaxation response of the aorta and reducing levels of hepatic CRP, regardless of dietary style.
Acknowledgements
Declaration of interestAll coauthors accept responsibility for the content of the manuscript. The authors declare that there are no conflicts of interest. This project was supported in part by USDA Contract #58-1950-7-707 and by the National Institute of Health Grant R01 AG025834 (SWC). The contents of this publication do not necessarily reflect the views or politics of the U.S. Department of Agriculture, nor does the mention of trade names, commercial products or organizations imply endorsement by the U.S. Government.
References
Brandes R. P. , Schmitz-Winnenthal F. H. , Feletou M. , Godecke A. , Huang P. L. , Vanhoutte P. M. , Fleming I. , Busse R. 2000 An endothelium-derived hyperpolarizing factor distinct from NO and prostacyclin is a major endothelium-dependent vasodilator in resistance vessels of wild-type and endothelial NO synthase knockout mice Proc. Natl. Acad. Sci. USA 97 9747 - 9752    DOI : 10.1073/pnas.97.17.9747
Breve J. J. , Drukarch B. , van Strien M. , van Dam A. M. 2008 Validated sandwich ELISA for the quantification of tissue transglutaminase in tissue homogenates and cell lysates of multiple species J. Immunol. Methods 332 142 - 150    DOI : 10.1016/j.jim.2008.01.005
Calabro P. , Willerson J. T. , Yeh E. T. 2003 Inflammatory cytokines stimulated C-reactive protein production by human coronary artery smooth muscle cells Circulation 108 1930 - 1932    DOI : 10.1161/01.CIR.0000096055.62724.C5
Choi M. S. , Jang Y. Y. , Lee W. S. , Song J. H. , Shin Y. K. 2003 Dietary ascorbate supplementation reduces oxidative tissue damage and expression of iNOS in the kidney of streptozotocin induced diabetic rats Kor. J. Physiol. Pharmacol. 7 39 - 45
Edith-Rodriguez J. , Resendiz-Albor A. A. , ArciniegaMartinez I. M. , Campos-Rodriguez R. , Hong E. , Huang F. , Villafana S. 2013 Effect of early diabetes on the expression of alpha-1 adrenergic receptors in aorta and carotidarteries of Wistar Kyoto and spontaneously hypertensive rats Clin. Exp. Hypertens. 35 389 - 395    DOI : 10.3109/10641963.2012.739233
Feletou M. , Kohler R. , Vanhoutte P. M. 2012 Nitric oxide: orchestrator of endothelium-dependent responses Ann. Med. 44 694 - 716    DOI : 10.3109/07853890.2011.585658
Fletcher R. H. , Fairfield K. M. 2002 Vitamins for chronic disease prevention in adults: clinical applications JAMA 287 3127 - 3129    DOI : 10.1001/jama.287.23.3127
Gentile M. , Panico S. , Rubba F. , Mattiello A. , Chiodini P. , Jossa F. , Marotta G. , Pauciullo P. , Rubba P. 2010 Obesity, overweight, and weight gain over adult life are main determinants of elevated hs-CRP in a cohort of Mediterranean women Eur. J. Clin. Nutr. 64 873 - 878    DOI : 10.1038/ejcn.2010.69
Hsu T. C. , Huang C. Y. , Chiang S. Y. , Lai W. X. , Tsai C. H. , Tzang B. S. 2008 Transglutaminase inhibitor cystamine alleviates the abnormality in liver from NZB/W F1 mice Eur. J. Pharmacol. 579 382 - 389    DOI : 10.1016/j.ejphar.2007.10.059
Jialal I. , Verma S. , Devaraj S. 2009 Inhibition of endothelial nitric oxide synthase by C-reactive protein: clinical relevance Clin Chem. 55 206 - 208
Lentz S. R. , Piegors D. J. , Malinow M. R. , Heistad D. D. 2001 Supplementation of atherogenic diet with B vitamins does not prevent atherosclerosis or vascular dysfunction in monkeys Circulation 103 1006 - 1011    DOI : 10.1161/01.CIR.103.7.1006
Libby P. , Ridker P. M. , Maseri A. 2002 Inflammation and atherosclerosis Circulation 105 1135 - 1143    DOI : 10.1161/hc0902.104353
Moens A. L. , Vrints C. J. , Claeys M. J. , Timmermans J. P. , Champion H. C. , Kass D. A. 2008 Mechanisms and potential therapeutic targets for folic acid in cardiovascular disease Am. J. Physiol. Heart Circ. Physiol. 294 H1971 - 1977    DOI : 10.1152/ajpheart.91503.2007
Rasouli M. L. , Nasir K. , Blumenthal R. S. , Park R. , Aziz D. C. , Budoff M. J. 2005 Plasma homocysteine predicts progression of atherosclerosis Atherosclerosis 181 159 - 165    DOI : 10.1016/j.atherosclerosis.2005.01.001
Singh U. , Devaraj S. , Vasquez-Vivar J. , Jialal I. 2007 C-reactive protein decreases endothelial nitric oxide synthase activity via uncoupling J. Mol. Cell. Cardiol. 43 780 - 791    DOI : 10.1016/j.yjmcc.2007.08.015
Stroes E. S. , van Faassen E. E. , Yo M. , Martasek P. , Boer P. , Govers R. , Rabelink T. J. 2000 Folic acid reverts dysfunction of endothelial nitric oxide synthase Circ. Res. 86 1129 - 1134    DOI : 10.1161/01.RES.86.11.1129
Symons J. D. , Rutledge J. C. , Simonsen U. , Pattathu R. A. 2006 Vascular dysfunction produced by hyperhomocysteinemia is more severe in the presence of low folate Am. J. Physiol. Heart Circ. Physiol. 290 H181 - 191
Symons J. D. , Zaid U. B. , Athanassious C. N. , Mullick A. E. , Lentz S. R. , Rutledge J. C. 2006 Influence of folate on arterial permeability and stiffness in the absence or presence of hyperhomocysteinemia Arterioscler. Thromb. Vasc. Biol. 26 814 - 818    DOI : 10.1161/01.ATV.0000204408.01416.16
van Dijk R. A. , Rauwerda J. A. , Steyn M. , Twisk J. W. , Stehouwer C. D. 2001 Long-term homocysteine-lowering treatment with folic acid plus pyridoxine is associated with decreased blood pressure but not with improved brachial artery endothelium-dependent vasodilation or carotid artery stiffness: a 2-year, randomized, placebo-controlled trial Arterioscler. Thromb. Vasc Biol. 21 2072 - 2079    DOI : 10.1161/hq1201.100223
Venugopal S. K. , Devaraj S. , Yuhanna I. , Shaul P. , Jialal I. 2002 Demonstration that C-reactive protein decreases eNOS expression and bioactivity in human aortic endothelial cells Circulation 106 1439 - 1441    DOI : 10.1161/01.CIR.0000033116.22237.F9
Wilmink H. W. , Stroes E. S. , Erkelens W. D. , Gerritsen W. B. , Wever R. , Banga J. D. , Rabelink T. J. 2000 Influence of folic acid on postprandial endothelial dysfunction Arterioscler. Thromb. Vasc. Biol. 20 185 - 188    DOI : 10.1161/01.ATV.20.1.185
Zhu Y. , Bian Z. , Lu P. , Karas R. H. , Bao L. , Cox D. , Hodgin J. , Shaul P. W. , Thoren P. , Smithies O. , Gustafsson J. A. , Mendelsohn M. E. 2002 Abnormal vascular function and hypertension in mice deficient in estrogen receptor beta Science 295 505 - 508    DOI : 10.1126/science.1065250