Abstract
Background: The consumption and blood concentrations of lycopene are both positively and inversely associated with the risk of several chronic diseases. The inconsistences in lycopene disease association studies may stem from a lack of knowledge about the genetic variation in the synthesis, metabolism, and deposition of transport and binding proteins, which potentially influence serum lycopene concentrations.
Objective: We examined the association between variation across the genome and serum concentrations of lycopene in a multiethnic population.
Methods: Participants included African (n = 914), Hispanic (n = 464), and European (n = 1203) American postmenopausal women from the Women’s Health Initiative. We analyzed ∼7 million single nucleotide polymorphisms (SNPs). Linear regression models were used to assess associations between each SNP and serum concentrations (log transformed, continuous) of lycopene; we adjusted for age, body mass index, and population substructure. Models were run separately by ethnicity, and results were combined in a transethnic fixed-effects meta-analysis.
Results: In the meta-analysis, the scavenger receptor class B, member 1 (SCARB1) gene, which encodes for a cholesterol membrane transporter, was significantly associated with lycopene concentrations (rs1672879; P < 2.68 × 10−9). Each additional G allele resulted in a 12% decrease in lycopene concentrations for African Americans, 20% decrease for Hispanic Americans, and 9% decrease for European Americans. In addition, 2 regions were significantly associated with serum lycopene concentrations in African Americans: the slit homolog 3 gene (SLIT3), which serves as a molecular guidance cue in cellular migration, and the dehydrogenase/reductase (SDR family) member 2 (DHRS2) gene, which codes for an oxidoreductase that mitigates the breakdown of steroids.
Conclusions: We found 3 novel loci associated with serum lycopene concentrations, 2 of which were specific to African Americans. Future functional studies looking at these specific genes may provide insight into the metabolism and underlying function of lycopene in humans, which may further elucidate lycopene’s influence on disease risk and health. This trial was registered at clinicaltrials.gov as NCT00000611.
Keywords: lycopene, carotenoid metabolism, genetic variation, scavenger receptor class B type 1 protein, multiethnic, postmenopausal women, nutritional epidemiology
Introduction
Lycopene is 1 of >600 carotenoids. Carotenoids, a diverse group of lipid-soluble phytochemicals, protect plants from oxidative damage, serve as pigments to attract pollinating insects, and act as precursors to key plant hormones (1, 2). In the United States, one of the most widely consumed carotenoids is lycopene. Lycopene is most known for the characteristic red color it imparts to tomatoes and tomato-based products, watermelon, red bell peppers, pink grapefruit, and other red or pink fruits and vegetables (2).
Both the consumption and blood concentrations of lycopene are inversely associated with the risk of several chronic diseases, including cardiovascular disease, metabolic syndrome, and cancer (most notably prostate cancer) (3–13). Recently, however, the presumed role for lycopene in prostate cancer prevention is disputed (14–16). Inconsistencies in lycopene-disease association studies may partly stem from a lack of knowledge about mediating factors that influence serum lycopene.
Known predictors of serum lycopene include diet, dietary supplement use, BMI, smoking, and sex (17–20). These characteristics account for, at best, 25% of the variance in serum lycopene concentrations. In addition to dietary and lifestyle predictors, a variety of binding proteins and carrier molecules influence nutrient concentrations in complex ways (1, 10, 17). Lycopene is predominantly transported in LDLs, which also carry the bulk of cholesterol in the plasma; thus, plasma lycopene concentrations and total cholesterol are highly correlated (21). Genetic variation in the synthesis, metabolism, and deposition of these transport and binding proteins may influence serum lycopene (22–25), but this is an understudied area that needs further investigation, especially in ethnically diverse populations.
Accordingly, our objective for this work was to examine the association between variation across the genome and circulating concentrations of lycopene in African American, Hispanic American, and European American postmenopausal women from the Women’s Health Initiative (WHI)6.
Methods
Study population
The WHI is a long-term national health study that has focused on strategies for preventing heart disease, breast and colorectal cancers, and osteoporotic fractures in postmenopausal women (26). Briefly, the WHI was designed as a set of randomized controlled clinical trials (CTs) and an observational study (OS). The CT (n = 68,132) included 3 overlapping components: the hormone therapy trials (n = 27,347), dietary modification trial (n = 48,835), and calcium and vitamin D trial (n = 36,282). Eligible women could be randomly assigned into 1, 2, or all 3 of the CT components. Women who were ineligible or unwilling to join the CT were invited to join the OS (n = 93,676). Twenty core blood analytes (including lycopene and total cholesterol) were measured in a 6% random subsample of CT participants at baseline, years 1, 3, and 6, and in a 1% random subsample of OS participants at baseline. All WHI participants provided written informed consent, and institutional review board approval was obtained at each of the 40 WHI clinical centers and at the clinical coordinating center at the Fred Hutchinson Cancer Research Center.
For this analysis we included participants with baseline serum lycopene and total cholesterol data from 2 WHI genetic studies: 1) the SNP Health Association Resource (SHARe) cohort and 2) the Genomics and Randomized Trials Network (GARNET). SHARe is a cohort study of 12,007 minority women (n = 8405 African Americans and n = 3602 Hispanic Americans) who participated either in 1 of the CTs or the OS; 1378 SHARe participants have baseline serum lycopene and total cholesterol measures (914 African Americans and 464 Hispanic Americans). GARNET is a case-control trial of 4416 European Americans who participated in the hormone therapy with myocardial infarction, stroke, venous thrombosis, diabetes, and matching controls; 1203 European Americans have baseline serum lycopene and total cholesterol measures.
Procedures
At WHI baseline clinic visits, participants completed standardized questionnaires for assessment of health history, age, race/ethnicity, and other health and lifestyle characteristics, including diet, dietary supplement use, cigarette smoking, and leisure physical activity. Height, weight, and waist and hip circumferences were measured on all participants with the use of a standard protocol. BMI was calculated as weight (kg)/height (m2).
Blood lycopene and cholesterol measures
A blood specimen was obtained from all WHI participants after a 12-h, overnight fast at the baseline visit. Serum and buffy coat were processed by a standard protocol and stored at −70°C in a central biorepository (Fisher Bioservices) until analysis (26).
Serum lycopene was measured at Medical Research Laboratories with the use of HPLC on the Hitachi 747 analyzer (27). The assay CV for lycopene with the use of 8% blinded duplicates in each assay batch was 9.4%. Plasma total cholesterol was assayed at the University of Minnesota Medical Center with the use of the cholesterol oxidate method on the Roche Modular P Chemistry analyzer. The assay CV for cholesterol with the use of 19% blinded duplicates in each assay batch was 1.4%.
Genotyping
SHARe.
DNA samples plus 2% (n = 188) blinded duplicate pairs were sent to Affymetrix Inc. for genotyping on the Genome-wide Human SNP Array 6.0 (909,622 SNPs); ∼1% of samples failed genotyping. Other sample exclusions included call rate < 95%, unexpected duplicates, genotype data on the Y chromosome, and relative pairs (parent-offspring, twins, siblings, and half-siblings; total n = 313). For relative pairs, we excluded the sample with the lowest call rate. Average concordance for 188 pairs of blinded duplicate samples was 99.8%.
SNPs were excluded if they had call rate < 95%, concordance for duplicates < 98%, or Hardy-Weinberg equilibrium P < 10−4.
GARNET.
DNA samples plus 1% (n = 35) blinded duplicate pairs were sent to the Broad Institute Genetic Analysis Platform for genotyping on the Illumina HumanOmni1-Quad v1–0 B SNP array (1,016,423 SNPs); ∼2.7% of samples failed genotyping. Other sample exclusions included call rate < 98%, unexpected duplicates, genotype data on the Y chromosome, and relative pairs (parent-offspring, twins, siblings, and half-siblings; total n = 12). For relative pairs, we excluded the sample with the lowest call rate. Participants with sex chromosome abnormalities were retained, but their X chromosome genotypes were filtered out of the analysis. Average concordance for 35 pairs of blinded duplicate samples was 99.8%.
SNPs were excluded if call rate < 97%, greater than zero discordant call in duplicate genotyping, number of sample trio inheritance errors > 1, BeadStudio metrics GenTrain score < 0.6 or cluster separation values < 0.4, or Hardy-Weinberg equilibrium P < 10−4.
Imputation
SHARe and GARNET were both imputed to the 1000 Genomes Project (1kGP). The X chromosome was not imputed. The 1kGP reference panel (version 3.20101123) consisted of 1092 samples (246 Africans, 181 Admixed Americans, 286 Asians, and 379 Europeans). The genome-wide association data were first grouped into sets comprising 2500 SNPs; neighboring SNP sets overlapped by 500 SNPs. Then all SNP sets were phased by using BEAGLE (28). SHARe was imputed to 1kGP by using minimac, a low-memory and computationally efficient implementation of the MaCH algorithm (29), and GARNET was imputed by using BEAGLE. For each participant, genotypes were reported as a dosage value (a continuous number between 0 and 2) that reflects the expected number of copies of an allele at that SNP conditional on the directly observed genotypes in both the subject and the phased haplotype assignments in the 1kGP samples. For this analysis we excluded poorly imputed SNPs (R2 < 0.5) and those SNPs with minor allele frequencies (MAFs) < 5%. This resulted in the analysis of 8,458,634 SNPs for African Americans, 6,389,009 SNPs for Hispanic Americans, and 6,305,487 SNPs for European Americans.
Genetic ancestry
SHARe.
Genetic ancestry was calculated separately for African and Hispanic Americans, using principal components analysis (PCA) as described by Patterson et al. (31). Here, we used 178,101 independent autosomal SNPs (with missing call rates < 5% and MAFs > 5%) that were common between our samples and the reference panels for the PCA. Our reference included 475 publically available samples from HapMap Yoruba from Ibadan, Nigeria, HapMap Utah residents with Northern and Western European ancestry, the Human Genome Diversity Project East Asian population, and the Human Genome Diversity Project Native American populations. These same samples were used to determine ancestral percentages by using Frappe (30); among African Americans samples (n = 57) with <10% African American ancestry were identified and excluded. We used the top 4 principal components to adjust for genetic ancestry in our analysis of African and Hispanic Americans.
GARNET.
Genetic ancestry was calculated by using PCA as described by Patterson et al. (31). Here, we used 112,304 independent autosomal SNPs (with missing call rates < 5% and MAFs > 5%) that were common between our samples and the reference panels for PCA. Our reference included 270 publically available samples from the following HapMap populations: Utah residents with Northern and Western European ancestry, Yoruba from Ibadan, Nigeria, Japanese from Tokyo, Japan, and Han Chinese from Beijing, China. We used the top 4 principal components to adjust for genetic ancestry in our analysis of European Americans.
Statistical analysis
To avoid the possibility of extreme lycopene values affecting our results, we used Winsorized means of lycopene and total cholesterol values. Therefore, instead of dropping outliers that are more extreme than the first and 99th quantile, we replaced these values by the value at the first and 99th quantile, respectively. In addition, we applied the natural logarithmic transformation to improve the normality of the biomarker distributions.
Linear regression models were used to assess associations between each SNP and serum concentrations (log-transformed, continuous) of lycopene. Genotype was modeled as an additive effect, with the genotype dosage values used as the primary predictor of interest. A 1 degree-of-freedom likelihood ratio test was used to assess statistical significance. A significance threshold of P < 5 × 10−8 was used to define a single result as genome-wide significant. The baseline model covariates included the 4 principal components of ancestry, BMI (continuous), and baseline age (continuous). We ran an additional model, which included baseline model covariates and total serum cholesterol (log-transformed, continuous). This model examined the association of genetic variants with serum lycopene independent of serum total cholesterol concentrations. We did not adjust for smoking status because only a small proportion of the population smoked (6–11%). We examined African Americans, Hispanic Americans, and European Americans both separately and combined (using a fixed-effects meta-analysis). Analyses were conducted with R 3.0 and METASOFT (for meta-analysis) (32).
Results
Baseline characteristics of study participants are given in Table 1. European American women were on average older than African American and Hispanic American women (P < 0.001). African American women had the largest BMI (P < 0.001). Among all ethnic groups, Hispanic Americans had the highest serum lycopene concentrations, whereas African Americans had the lowest (P < 0.01 and P < 0.05, respectively). European Americans had the highest total cholesterol concentrations compared with all other ethnic groups (P < 0.001). No other differences were observed between the 3 subpopulations, except that the proportions of each population enrolled in various parts of the WHI clinical trials varied.
TABLE 1.
Baseline characteristics of study population1
| Characteristic | African American (n = 914) | Hispanic American (n = 464) | European American (n = 1203) |
| Age groups, n (%) | |||
| 50–54 y | 176 (19.3) | 110 (23.7) | 84 (8.5) |
| 55–59 y | 203 (22.2) | 103 (22.2) | 188 (19.1) |
| 60–69 y | 397 (43.4) | 215 (46.3) | 456 (46.3) |
| 70–79 y | 138 (15.1) | 36 (7.8) | 257 (26.1) |
| WHI enrollment, n (%) | |||
| Observational study | 74 (8.1) | 70 (15.1) | NA |
| Clinical trials | 840 (91.9) | 394 (84.9) | 985 (100.0) |
| Hormone trials, n (%) | |||
| Estrogen alone | 132 (14.4) | 57 (12.3) | 199 (20.2) |
| Estrogen alone placebo | 151 (16.5) | 69 (14.9) | 206 (20.9) |
| Estrogen + progesterone | 107 (11.7) | 74 (15.9) | 307 (31.2) |
| Estrogen + progesterone placebo | 109 (11.9) | 85 (18.3) | 273 (27.7) |
| Dietary modification, n (%) | |||
| Intervention | 220 (24.1) | 91 (19.6) | NA |
| Comparison | 347 (38.0) | 123 (26.5) | NA |
| Calcium + vitamin D, n (%) | |||
| Intervention | 213 (23.3) | 118 (25.4) | NA |
| Placebo | 236 (25.8) | 112 (24.1) | NA |
| Case status, n (%) | |||
| Myocardial infarction | NA | NA | 30 (3.0) |
| Stroke | NA | NA | 18 (1.8) |
| Venous thrombosis | NA | NA | 25 (2.5) |
| Diabetes | NA | NA | 58 (5.9) |
| Controls | NA | NA | 864 (87.7) |
| Lifestyle and diet | |||
| Current smoker, n (%) | 102 (11.2) | 31 (6.7) | 107 (10.9) |
| BMI, kg/m2 | 31.8 ± 6.3 | 29.4 ± 5.2 | 28.8 ± 6.0 |
| Serum lycopene, μg/mL | 0.39 ± 0.2 | 0.44 ± 0.2 | 0.41 ± 0.2 |
| Serum cholesterol, mg/dL | 220 ± 40.5 | 220 ± 36.6 | 228 ± 37.6 |
Data are presented as means ± SDs unless noted otherwise. NA, not applicable; WHI, Women's Health Initiative.
The results of the fixed-effects meta-analysis of the genome-wide associations for serum lycopene in the baseline model are presented in Table 2. By combining the 3 subpopulations, 3 SNPs on chromosome 12 located near the scavenger receptor class B member 1 (SCARB1) gene met our prespecified criteria for genome-wide significance (Table 2). The strongest signal in this region was rs1672879 (P < 2.68 × 10−9). Here, each additional G allele resulted in a 12% decrease in lycopene concentrations for African Americans (MAF = 0.41), a 20% decrease for Hispanic Americans (MAF = 0.28), and a 9% decrease for European Americans (MAF = 0.03). In the model in which we adjusted for total serum cholesterol, rs1672879 remained significantly associated (P < 4.09 × 10−8) with a similar effect size (β = −0.12); no additional regions were significant in the adjusted model.
TABLE 2.
SNPs in the SCARB1 region associated with serum lycopene in transethnic fixed-effects meta-analysis1
| Meta-analysis |
African American |
Hispanic American |
European American |
||||||||||||||
| SNP | A1/A2 | P | β | P-het | I2 | P | β | CAF | MAF | P | β | CAF | MAF | P | β | CAF | MAF |
| rs1672879* | G/T | 2.68 × 10−9 | −0.14 | 0.25 | 28.3 | 7.85 × 10−5 | −0.12 | 0.59 | 0.41 | 6.26 × 10−6 | −0.20 | 0.77 | 0.23 | 0.14 | −0.09 | 0.97 | 0.03 |
| rs701107 | A/G | 2.43 × 10−8 | −0.13 | 0.30 | 17.2 | 2.59 × 10−4 | −0.12 | 0.60 | 0.40 | 2.04 × 10−5 | −0.19 | 0.77 | 0.23 | 0.15 | −0.09 | 0.97 | 0.03 |
| rs838861 | T/C | 2.32 × 10−8 | −0.14 | 0.19 | 40.4 | 7.29 × 10−4 | −0.11 | 0.67 | 0.33 | 4.65 × 10−6 | −0.20 | 0.78 | 0.22 | 0.14 | −0.09 | 0.97 | 0.03 |
Baseline model was further adjusted for age, BMI, and first 4 principal components of ancestry. *SNP with the most significant association in the meta-analysis with serum lycopene for this region. A1/A2, coded/noncoded allele; CAF, coded allele (A1) frequency; I2, index of the degree of heterogeneity; MAF, minor allele frequency; P-het, P values for heterogeneity, indicating whether effect sizes are homogeneous across ancestry samples; SCARB1, scavenger receptor class B member 1; SNP, single nucleotide polymorphism.
The results of the genome-wide associations for serum lycopene in African Americans are presented in Table 3. In Hispanic and European Americans no SNPs met the criteria for genome-wide significance in either the baseline or adjusted models; hence, results are not shown. Two regions were significantly associated with serum lycopene concentrations in African Americans: 4 SNPs in the slit homolog 3 (SLIT3) gene on chromosome 5 and the 3 SNPs in the dehydrogenase/reductase (SDR family) member 2 (DHRS2) gene on chromosome 14 (Table 3). The most statistically significant signal in the SLIT3 region was rs78219687 (P < 9.87 × 10−9; β = 0.39; MAF = 0.05), and the most statistically significant signal in the DHRS2 region was rs74036811 (P < 9.55 × 10−9; β = 0.38; MAF = 0.05). In the model adjusted for total cholesterol, the DHRS2 region remained significantly associated with serum lycopene concentrations in African Americans with a similar effect size (rs74036811; P < 4.48 × 10−9; β = 0.38). The results from the adjusted model also suggested an association between the SLIT3 region and lycopene with a similar effect size (rs78219687; P < 1.51 × 10−7; β = 0.34).
TABLE 3.
SNPs in the SLIT3 and DHRS2 regions associated with serum lycopene in African Americans1
| Nearest gene | Chr | SNP | A1/A2 | P | β | CAF | MAF |
| SLIT3 | 5 | rs78219687* | G/A | 9.87 × 10−9 | 0.39 | 0.95 | 0.05 |
| rs114819789 | G/A | 3.44 × 10−8 | 0.41 | 0.95 | 0.05 | ||
| rs74659379 | G/T | 3.42 × 10−8 | 0.41 | 0.95 | 0.05 | ||
| rs114731073 | A/T | 3.40 × 10−8 | 0.41 | 0.95 | 0.05 | ||
| DHRS2 | 14 | rs74036811* | C/G | 9.54 × 10−9 | 0.38 | 0.95 | 0.05 |
| rs17095390 | G/C | 9.55 × 10−8 | 0.38 | 0.95 | 0.05 | ||
| rs17095435 | C/T | 2.22 × 10−8 | 0.39 | 0.95 | 0.05 |
Baseline model was further adjusted for age, BMI, and first 4 principal components of ancestry. *SNP with the most significant association in African Americans with serum lycopene for this region. A1/A2, coded/non-coded allele; CAF, coded allele (A1) frequency; Chr, chromosome number; DHRS2, dehydrogenase/reductase (SCR family) member 2; MAF, minor allele frequency; SLIT3, slit homolog 3; SNP, single nucleotide polymorphism.
Discussion
Our genome-wide analysis found 3 novel regions associated with circulating serum lycopene concentrations. Two of these regions were ethnically specific to African Americans. This marks the first time such an analysis was done in a multiethnic study population.
The SCARB1 region significantly associated with serum lycopene in the meta-analysis. We found similar effect sizes for this variant across the 3 ethnic groups; the smaller MAF in European Americans likely explains the weaker signal observed in this population. The overall consistency for this locus adds support for this finding through built-in replication.
The SCARB1 gene encodes for a cholesterol membrane transporter called scavenger receptor class B type I (SR-BI). In vitro, SR-BI mediates the absorption of lycopene and cholesterol and other fat-soluble vitamins (33–36). Candidate gene studies have found variants in SCARB1 to associate with serum concentrations of vitamin E and provitamin A carotenoids but not lycopene (37, 38). To our knowledge this is the first genome-wide association study to observe a significant association between this region and serum lycopene. Variants in this gene likely influence the ability of SR-BI to absorb lycopene at the intestinal level and/or to modify plasma lipid concentrations such as LDLs, which predominately transport lycopene (22, 33). Additional consequences of altered SCARB1 function include modified platelet function, decreased adrenal steroidogenesis, changes in lipid profiles, and insulin sensitivity to dietary fat (39–41).
Variants in the SLIT3 and DHRS2 regions were significantly associated with lycopene concentrations in African Americans in the unadjusted model; after adjustment for total cholesterol only the variants in the DHRS2 region remained significant.
The variants found significant in these regions were monomorphic in Hispanic and European Americans, which is why we observed an association only in African Americans. Suggestion of African American-specific effects was noted before. In a study by Arab et al. (18) they found, despite stronger validity in reported energy intakes for African Americans than for whites, that the 24-h dietary recall correlated less with serum lycopene in African Americans than in whites. The researchers speculated that one reason for this finding could be racially related genetic variants that influence circulating concentrations. Perhaps the African American-specific effects we observed in the SLIT3 and DHRS2 regions partially explain the variance in serum lycopene concentrations unaccounted for by diet.
The SLIT3 gene serves as a molecular guidance cue in cellular migration, and its function may be mediated by interaction with roundabout homolog receptors. This gene is associated with birth defects and multiple types of cancer, including melanoma, thyroid cancer, and lung cancer (42–45). The DHRS2 region codes for a NAD/NADP-dependent oxidoreductase that mitigates the breakdown of steroids, retinoids, prostaglandins, and xenobiotics; one could hypothesize that this protein also aids in the metabolism of lycopene. In addition, DHRS2 plays a noncatalytic role in the regulation of the cell cycle and apoptosis (46).
A previous study found the β-carotene 15,15-monooxygenase 1 (BCMO1) gene, a key enzyme in provitamin A carotenoid metabolism, to associate with serum lycopene concentrations in European Americans (25). However, our analysis and a more recently conducted study by Hendrickson et al. (24) did not observe an association with this locus and serum lycopene concentrations in European Americans. In addition, we did not observe this association in either African or Hispanic Americans in our study. Earlier research notes that BCMO1 is not thought to catalyze lycopene or other non-provitamin A carotenoids because of their molecular structure (47, 48); however, a more recent in vitro study surprisingly found that BCMO1 catalyzed the oxidative cleavage of lycopene with a catalytic efficiency similar to that of β-carotene (49). Follow-up and replication in a larger transethnic study sample should further investigate the association between the BCMO1 region and lycopene concentrations.
Limitations of our study merit consideration. We had a limited sample size; however, despite this we found significant associations after adjusting for multiple testing. Our study population contained only women; therefore, sex differences in the association between the genetic variant and serum lycopene concentrations could not be examined.
In conclusion, we were able to examine associations between genetic variants and serum lycopene concentrations in an ethnically diverse study sample. This is the first time that this association was studied in an African American or Hispanic American population. Here, we found 3 novel loci associated with serum lycopene concentrations, 2 of which were specific to African Americans. Future studies, with ethnically diverse populations, should attempt to replicate these findings. Given the replication of findings, functional studies looking at these specific genes may help provide insight into the metabolism and underlying function of lycopene in humans, which may help further elucidate lycopene’s influence on disease risk and health.
Acknowledgments
MLN was responsible for project conception; NZ, JL, and CD analyzed the data; NZ wrote the manuscript; CK, UP, and MLN contributed to the edits of the final manuscript; and NZ had primary responsibility for the final content. All authors read and approved the final manuscript.
Footnotes
Abbreviations used: BCMO1, 15,15′-monooxygenase 1; CT, controlled clinical trials; DHRS2, dehydrogenase/reductase (SDR family) member 2; GARNET, Genomics and Randomized Trials Network; MAF, minor allele frequency; OS, observational study; PCA, principal components analysis; SHARe, SNP Health Association Resource; SCARB1, scavenger receptor class B, member 1; SLIT3, slit homolog 3; SNP, single nucleotide polymorphism; SR-BI, scavenger receptor class B type 1 protein; WHI, Women's Health Initiative; 1kGP, 1000 Genomes Project.
References
- 1.Ford NA, Erdman JW., Jr Are lycopene metabolites metabolically active? Acta Biochim Pol 2012;59:1–4. [PubMed] [Google Scholar]
- 2.Murphy MM, Barraj LM, Herman D, Bi X, Cheatham R, Randolph RK. Phytonutrient intake by adults in the United States in relation to fruit and vegetable consumption. J Acad Nutr Diet 2012;112:222–9. [DOI] [PubMed] [Google Scholar]
- 3.Hu F, Wang Yi B, Zhang W, Liang J, Lin C, Li D, Wang F, Pang D, Zhao Y. Carotenoids and breast cancer risk: a meta-analysis and meta-regression. Breast Cancer Res Treat 2012;131:239–53. [DOI] [PubMed] [Google Scholar]
- 4.Eliassen AH, Hendrickson SJ, Brinton LA, Buring JE, Campos H, Dai Q, Dorgan JF, Franke AA, Gao YT, Goodman MT, et al. Circulating carotenoids and risk of breast cancer: pooled analysis of eight prospective studies. J Natl Cancer Inst 2012;104:1905–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Aune D, Chan DS, Vieira AR, Navarro Rosenblatt DA, Vieira R, Greenwood DC, Norat T. Dietary compared with blood concentrations of carotenoids and breast cancer risk: a systematic review and meta-analysis of prospective studies. Am J Clin Nutr 2012;96:356–73. [DOI] [PubMed] [Google Scholar]
- 6.Ollberding NJ, Maskarinec G, Conroy SM, Morimoto Y, Franke AA, Cooney RV, Wilkens LR, Le Marchand L, Goodman MT, Hernandez BY, et al. Prediagnostic circulating carotenoid levels and the risk of non-Hodgkin lymphoma: the Multiethnic Cohort. Blood 2012;119:5817–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Shardell MD, Alley DE, Hicks GE, El-Kamary SS, Miller RR, Semba RD, Ferrucci L. Low-serum carotenoid concentrations and carotenoid interactions predict mortality in US adults: the Third National Health and Nutrition Examination Survey. Nutr Res 2011;31:178–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Slattery ML, Lundgreen A, Welbourn B, Wolff RK, Corcoran C. Oxidative balance and colon and rectal cancer: interaction of lifestyle factors and genes. Mutat Res 2012;734:30–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Thies F, Masson LF, Rudd A, Vaughan N, Tsang C, Brittenden J, Simpson WG, Duthie S, Horgan GW, Duthie G. Effect of a tomato-rich diet on markers of cardiovascular disease risk in moderately overweight, disease-free, middle-aged adults: a randomized controlled trial. Am J Clin Nutr 2012;95:1013–22. [DOI] [PubMed] [Google Scholar]
- 10.Wang XD. Lycopene metabolism and its biological significance. Am J Clin Nutr 2012;96:1214S–22S. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Aydemir G, Kasiri Y, Birta E, Beke G, Garcia AL, Bartok EM, Ruhl R. Lycopene-derived bioactive retinoic acid receptors/retinoid-X receptors-activating metabolites may be relevant for lycopene's anti-cancer potential. Mol Nutr Food Res 2013;57:739–47. [DOI] [PubMed] [Google Scholar]
- 12.Ghavipour M, Saedisomeolia A, Djalali M, Sotoudeh G, Eshraghyan MR, Moghadam AM, Wood LG. Tomato juice consumption reduces systemic inflammation in overweight and obese females. Br J Nutr 2013;109:2031–5. [DOI] [PubMed] [Google Scholar]
- 13.Palozza P, Simone R, Catalano A, Parrone N, Monego G, Ranelletti FO. Lycopene regulation of cholesterol synthesis and efflux in human macrophages. J Nutr Biochem 2011;22:971–8. [DOI] [PubMed] [Google Scholar]
- 14.Peters U, Leitzmann MF, Chatterjee N, Wang Y, Albanes D, Gelmann EP, Friesen MD, Riboli E, Hayes RB. Serum lycopene, other carotenoids, and prostate cancer risk: a nested case-control study in the prostate, lung, colorectal, and ovarian cancer screening trial. Cancer Epidemiol Biomarkers Prev 2007;16:962–8. [DOI] [PubMed] [Google Scholar]
- 15.Zu K, Mucci L, Rosner BA, Clinton SK, Loda M, Stampfer MJ, Giovannucci E. Dietary lycopene, angiogenesis, and prostate cancer: a prospective study in the prostate-specific antigen era. J Natl Cancer Inst 2014;106:djt430. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kristal AR, Till C, Platz EA, Song X, King IB, Neuhouser ML, Ambrosone CB, Thompson IM. Serum lycopene concentration and prostate cancer risk: results from the Prostate Cancer Prevention Trial. Cancer Epidemiol Biomarkers Prev 2011;20:638–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Erdman JW., Jr How do nutritional and hormonal status modify the bioavailability, uptake, and distribution of different isomers of lycopene? J Nutr 2005;135:2046S–7S. [DOI] [PubMed] [Google Scholar]
- 18.Arab L, Cambou MC, Craft N, Wesseling-Perry K, Jardack P, Ang A. Racial differences in correlations between reported dietary intakes of carotenoids and their concentration biomarkers. Am J Clin Nutr 2011;93:1102–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Zaripheh S, Erdman JW., Jr Factors that influence the bioavailablity of xanthophylls. J Nutr 2002;132:531S–4S. [DOI] [PubMed] [Google Scholar]
- 20.Kabat GC, Kim MY, Sarto GE, Shikany JM, Rohan TE. Repeated measurements of serum carotenoid, retinol and tocopherol levels in relation to colorectal cancer risk in the Women's Health Initiative. Eur J Clin Nutr 2012;66:549–54. [DOI] [PubMed] [Google Scholar]
- 21.Palozza P, Catalano A, Simone RE, Mele MC, Cittadini A. Effect of lycopene and tomato products on cholesterol metabolism. Ann Nutr Metab 2012;61:126–34. [DOI] [PubMed] [Google Scholar]
- 22.Moran NE, Erdman JW, Jr, Clinton SK. Complex interactions between dietary and genetic factors impact lycopene metabolism and distribution. Arch Biochem Biophys 2013;539:171–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Wang TT, Edwards AJ, Clevidence BA. Strong and weak plasma response to dietary carotenoids identified by cluster analysis and linked to beta-carotene 15,15'-monooxygenase 1 single nucleotide polymorphisms. J Nutr Biochem 2013;24:1538–46. [DOI] [PubMed] [Google Scholar]
- 24.Hendrickson SJ, Hazra A, Chen C, Eliassen AH, Kraft P, Rosner BA, Willett WC. β-Carotene 15,15'-monooxygenase 1 single nucleotide polymorphisms in relation to plasma carotenoid and retinol concentrations in women of European descent. Am J Clin Nutr 2012;96:1379–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ferrucci L, Perry JR, Matteini A, Perola M, Tanaka T, Silander K, Rice N, Melzer D, Murray A, Cluett C, et al. Common variation in the beta-carotene 15,15'-monooxygenase 1 gene affects circulating levels of carotenoids: a genome-wide association study. Am J Hum Genet 2009;84:123–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Anderson GL, Manson J, Wallace R, Lund B, Hall D, Davis S, Shumaker S, Wang CY, Stein E, Prentice RL. Implementation of the Women's Health Initiative study design. Ann Epidemiol 2003;13:S5–17. [DOI] [PubMed] [Google Scholar]
- 27.Kaplan LA, Miller JA, Stein EA. Simultaneous measurement of serum retinol, tocopherols, carotenes, and carotenoids by high performance liquid chromatography. J Clin Lab Anal 1987;1:147–52. [Google Scholar]
- 28.Browning BL, Browning SR. A unified approach to genotype imputation and haplotype-phase inference for large data sets of trios and unrelated individuals. Am J Hum Genet 2009;84:210–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Howie B, Fuchsberger C, Stephens M, Marchini J, Abecasis GR. Fast and accurate genotype imputation in genome-wide association studies through pre-phasing. Nat Genet 2012;44:955–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Tang H, Peng J, Wang P, Risch NJ. Estimation of individual admixture: analytical and study design considerations. Genet Epidemiol 2005;28:289–301. [DOI] [PubMed] [Google Scholar]
- 31.Patterson N, Price AL, Reich D. Population structure and eigenanalysis. PLoS Genet 2006;2:e190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Han B, Eskin E. Random-effects model aimed at discovering associations in meta-analysis of genome-wide association studies. Am J Hum Genet 2011;88:586–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Moussa M, Landrier JF, Reboul E, Ghiringhelli O, Comera C, Collet X, Frohlich K, Bohm V, Borel P. Lycopene absorption in human intestinal cells and in mice involves scavenger receptor class B type I but not Niemann-Pick C1-like 1. J Nutr 2008;138:1432–6. [DOI] [PubMed] [Google Scholar]
- 34.Reboul E, Goncalves A, Comera C, Bott R, Nowicki M, Landrier JF, Jourdheuil-Rahmani D, Dufour C, Collet X, Borel P. Vitamin D intestinal absorption is not a simple passive diffusion: evidences for involvement of cholesterol transporters. Mol Nutr Food Res 2011;55:691–702. [DOI] [PubMed] [Google Scholar]
- 35.van Bennekum A, Werder M, Thuahnai ST, Han CH, Duong P, Williams DL, Wettstein P, Schulthess G, Phillips MC, Hauser H. Class B scavenger receptor-mediated intestinal absorption of dietary beta-carotene and cholesterol. Biochemistry 2005;44:4517–25. [DOI] [PubMed] [Google Scholar]
- 36.Reboul E, Klein A, Bietrix F, Gleize B, Malezet-Desmoulins C, Schneider M, Margotat A, Lagrost L, Collet X, Borel P. Scavenger receptor class B type I (SR-BI) is involved in vitamin E transport across the enterocyte. J Biol Chem 2006;281:4739–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Borel P, Desmarchelier C, Nowicki M, Bott R, Morange S, Lesavre N. Interindividual variability of lutein bioavailability in healthy men: characterization, genetic variants involved, and relation with fasting plasma lutein concentration. Am J Clin Nutr 2014;100:168–75. [DOI] [PubMed] [Google Scholar]
- 38.Borel P, Lietz G, Goncalves A, Szabo de Edelenyi F, Lecompte S, Curtis P, Goumidi L, Caslake MJ, Miles EA, Packard C, et al. CD36 and SR-BI are involved in cellular uptake of provitamin A carotenoids by Caco-2 and HEK cells, and some of their genetic variants are associated with plasma concentrations of these micronutrients in humans. J Nutr 2013;143:448–56. [DOI] [PubMed] [Google Scholar]
- 39.Vergeer M, Korporaal SJ, Franssen R, Meurs I, Out R, Hovingh GK, Hoekstra M, Sierts JA, Dallinga-Thie GM, Motazacker MM, et al. Genetic variant of the scavenger receptor BI in humans. N Engl J Med 2011;364:136–45. [DOI] [PubMed] [Google Scholar]
- 40.Acton S, Osgood D, Donoghue M, Corella D, Pocovi M, Cenarro A, Mozas P, Keilty J, Squazzo S, Woolf EA, et al. Association of polymorphisms at the SR-BI gene locus with plasma lipid levels and body mass index in a white population. Arterioscler Thromb Vasc Biol 1999;19:1734–43. [DOI] [PubMed] [Google Scholar]
- 41.Pérez-Martinez P, Pérez-Jimenez F, Bellido C, Ordovas JM, Moreno JA, Marin C, Gómez P, Delgado-Lista J, Fuentes F, López-Miranda J. A polymorphism exon 1 variant at the locus of the scavenger receptor class B type I (SCARB1) gene is associated with differences in insulin sensitivity in healthy people during the consumption of an olive oil-rich diet. J Clin Endocrinol Metab 2005;90:2297–300. [DOI] [PubMed] [Google Scholar]
- 42.Davidson MR, Larsen JE, Yang IA, Hayward NK, Clarke BE, Duhig EE, Passmore LH, Bowman RV, Fong KM. MicroRNA-218 is deleted and downregulated in lung squamous cell carcinoma. PLoS ONE 2010;5:e12560. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Denk AE, Braig S, Schubert T, Bosserhoff AK. Slit3 inhibits activator protein 1-mediated migration of malignant melanoma cells. Int J Mol Med 2011;28:721–6. [DOI] [PubMed] [Google Scholar]
- 44.Dickinson RE, Dallol A, Bieche I, Krex D, Morton D, Maher ER, Latif F. Epigenetic inactivation of SLIT3 and SLIT1 genes in human cancers. Br J Cancer 2004;91:2071–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Guan H, Wei G, Wu J, Fang D, Liao Z, Xiao H, Li M, Li Y. Down-regulation of miR-218–2 and its host gene SLIT3 cooperate to promote invasion and progression of thyroid cancer. J Clin Endocrinol Metab 2013;98:E1334–44. [DOI] [PubMed] [Google Scholar]
- 46.Gabrielli F, Tofanelli S. Molecular and functional evolution of human DHRS2 and DHRS4 duplicated genes. Gene 2012;511:461–9. [DOI] [PubMed] [Google Scholar]
- 47.Kim YS, Oh DK. Substrate specificity of a recombinant chicken beta-carotene 15,15'-monooxygenase that converts beta-carotene into retinal. Biotechnol Lett 2009;31:403–8. [DOI] [PubMed] [Google Scholar]
- 48.Lindqvist A, Andersson S. Biochemical properties of purified recombinant human beta-carotene 15,15'-monooxygenase. J Biol Chem 2002;277:23942–8. [DOI] [PubMed] [Google Scholar]
- 49.dela Seña C, Narayanasamy S, Riedl KM, Curley RW, Jr, Schwartz SJ, Harrison EH. Substrate specificity of purified recombinant human beta-carotene 15,15'-oxygenase (BCO1). J Biol Chem 2013;288:37094–103. [DOI] [PMC free article] [PubMed] [Google Scholar]