Novel gene-by-environment interactions: APOB and NPC1L1 variants affect the relationship between dietary and total plasma cholesterol

Cardiovascular disease (CVD) is the leading cause of death in developed countries. Plasma cholesterol level is a key risk factor in CVD pathogenesis. Genetic and dietary variation both influence plasma cholesterol; however, little is known about dietary interactions with genetic variants influencing the absorption and transport of dietary cholesterol. We sought to determine whether gut expressed variants predicting plasma cholesterol differentially affected the relationship between dietary and plasma cholesterol levels in 1,128 subjects (772/356 in the discovery/replication cohorts, respectively). Four single nucleotide polymorphisms (SNPs) within three genes (APOB, CETP, and NPC1L1) were significantly associated with plasma cholesterol in the discovery cohort. These were subsequently evaluated for gene-by-environment (GxE) interactions with dietary cholesterol for the prediction of plasma cholesterol, with significant findings tested for replication. Novel GxE interactions were identified and replicated for two variants: rs1042034, an APOB Ser4338Asn missense SNP and rs2072183 (in males only), a synonymous NPC1L1 SNP in linkage disequilibrium with SNPs 5′ of NPC1L1. This study identifies the presence of novel GxE and gender interactions implying that differential gut absorption is the basis for the variant associations with plasma cholesterol. These GxE interactions may account for part of the “missing heritability” not accounted for by genetic associations. Cardiovascular disease (CVD) is the leading cause of death in developed countries. Plasma cholesterol level is a key risk factor in CVD pathogenesis. Genetic and dietary variation both influence plasma cholesterol; however, little is known about dietary interactions with genetic variants influencing the absorption and transport of dietary cholesterol. We sought to determine whether gut expressed variants predicting plasma cholesterol differentially affected the relationship between dietary and plasma cholesterol levels in 1,128 subjects (772/356 in the discovery/replication cohorts, respectively). Four single nucleotide polymorphisms (SNPs) within three genes (APOB, CETP, and NPC1L1) were significantly associated with plasma cholesterol in the discovery cohort. These were subsequently evaluated for gene-by-environment (GxE) interactions with dietary cholesterol for the prediction of plasma cholesterol, with significant findings tested for replication. Novel GxE interactions were identified and replicated for two variants: rs1042034, an APOB Ser4338Asn missense SNP and rs2072183 (in males only), a synonymous NPC1L1 SNP in linkage disequilibrium with SNPs 5′ of NPC1L1. This study identifies the presence of novel GxE and gender interactions implying that differential gut absorption is the basis for the variant associations with plasma cholesterol. These GxE interactions may account for part of the “missing heritability” not accounted for by genetic associations. Cardiovascular disease (CVD), including coronary heart disease and carotid artery disease (CAAD), is the leading cause of death in developed countries, accounting for approximately one-third of all deaths in adults over age 35 (1Roger V.L. Go A.S. Lloyd-Jones D.M. Benjamin E.J. Berry J.D. Borden W.B. Bravata D.M. Dai S. Ford E.S. Fox C.S. et al.Heart disease and stroke statistics–2012 update: a report from the American Heart Association.Circulation. 2012; 125: e2-e220Crossref PubMed Scopus (0) Google Scholar). One of the key risk factors for CVD is hypercholesterolemia (2Verschuren W.M. Jacobs D.R. Bloemberg B.P. Kromhout D. Menotti A. Aravanis C. Blackburn H. Buzina R. Dontas A.S. Fidanza F. et al.Serum total cholesterol and long-term coronary heart disease mortality in different cultures: Twenty-five-year follow-up of the seven countries study.JAMA. 1995; 274: 131-136Crossref PubMed Google Scholar), with epidemiologic studies demonstrating a graded relationship between total plasma cholesterol and CVD risk (3Wallis E.J. Ramsay L.E. Haq I.U. Ghahramani P. Jackson P.R. Yeo K.R. Yeo W.W. Coronary and cardiovascular risk estimation for primary prevention: validation of a new Sheffield table in the 1995 Scottish health survey population.BMJ. 2000; 320: 671-676Crossref PubMed Scopus (161) Google Scholar). Levels of total cholesterol and other lipids and lipoproteins are both heritable and targets for pharmacologic and lifestyle interventions (4Kathiresan S. Manning A.K. Demissie S. D舗Agostino R.B. Surti A. Guiducci C. Gianniny L. Burtt N.P. Melander O. Orho-Melander M. et al.A genome-wide association study for blood lipid phenotypes in the Framingham Heart Study.BMC Med. Genet. 2007; 8 (Suppl 1): S17Crossref PubMed Scopus (245) Google Scholar). One of the primary methods of lifestyle intervention to alter lipid levels is through dietary changes, such as decreasing total fat intake (5Varady K.A. Jones P.J. Combination diet and exercise interventions for the treatment of dyslipidemia: an effective preliminary strategy to lower cholesterol levels?.J. Nutr. 2005; 135: 1829-1835Crossref PubMed Scopus (109) Google Scholar, 6Vogel J.H. Bolling S.F. Costello R.B. Guarneri E.M. Krucoff M.W. Longhurst J.C. Olshansky B. Pelletier K.R. Tracy C.M. Vogel R.A. et al.Integrating complementary medicine into cardiovascular medicine. A report of the American College of Cardiology Foundation Task Force on Clinical Expert Consensus Documents (Writing Committee to Develop an Expert Consensus Document on Complementary and Integrative Medicine).J. Am. Coll. Cardiol. 2005; 46: 184-221Crossref PubMed Scopus (110) Google Scholar). Genome-wide association studies (GWASs) have implicated numerous genetic loci for association with total cholesterol levels. A recent meta-analysis by Teslovich et al. (7Teslovich T.M. Musunuru K. Smith A.V. Edmondson A.C. Stylianou I.M. Koseki M. Pirruccello J.P. Ripatti S. Chasman D.I. Willer C.J. et al.Biological, clinical and population relevance of 95 loci for blood lipids.Nature. 2010; 466: 707-713Crossref PubMed Scopus (2781) Google Scholar) of over 100,000 individuals of European ancestry confirmed numerous previous GWAS findings and identified novel variants and loci for prediction of total lipid phenotypes, including total cholesterol levels. While genetic evidence is accumulating for many of these loci, less is known about the specific mechanisms through which these genetic variants affect cholesterol levels. We hypothesized that one pathway through which these genetic variants could alter cholesterol homeostasis is through differential absorption or processing of dietary cholesterol, thereby influencing the potential benefit of lifestyle changes in the prevention of CVD. Utilizing the Carotid Lesion Epidemiology and Risk (CLEAR) CAAD case-control cohort, we investigated whether there was any evidence of single nucleotide polymorphisim (SNP)-dietary cholesterol interactions on prediction of total cholesterol levels, to better understand the potential influence of the genetic variation on the handling of dietary cholesterol and the potential downstream pathogenesis of CVD. Institutional Review Boards at the University of Washington, Virginia Mason Medical Center, and Veterans Affairs Puget Sound Health Care approved the study. Written informed consent was obtained from all participants. The study population for these analyses consisted of 1,128 subjects (772 subjects in the discovery subset and 356 subjects in the replication subset) from the previously described CLEAR study (8Jarvik G.P. Hatsukami C. R. S. M. D. et not the linkage disequilibrium PubMed Scopus Google Scholar, G.P. Hatsukami is a better of disease is or 2000; PubMed Scopus Google Scholar, R. Hatsukami P. of identified two loci that to risk for carotid artery Health PubMed Scopus Google Scholar, G.P. polymorphisms in the not 2012; PubMed Google Scholar, et cholesterol 2012; PubMed Scopus Google Scholar, D.R. E.J. G.P. Novel and genetic of and PubMed Scopus Google Scholar), both diet and genetic The discovery Scholar) the replication was for SNPs of as in the methods This CAAD and subjects of other phenotypes, including carotid artery by as as coronary artery and artery subjects were to of in cohort. was confirmed with three M. P. of population 2000; PubMed Scopus Google Scholar). including and were from and as as was by a was on an or or was CLEAR total cholesterol the of or the to for the are in of the CLEAR CLEAR CLEAR CLEAR case-control total cholesterol, total cholesterol, total cholesterol was through lipid levels for the subjects lipid levels total cholesterol was through lipid levels for the subjects lipid levels table in a new were to the developed by The Health at The survey the of intake over the previous of of and the of and including the and of of and were to the of specific were to of Health and and analyses to of total total cholesterol, and from dietary cholesterol intake from the was as the for with SNP in study. were from analyses intake was not between and or of a total of This survey two diet approximately and of an health 135: PubMed Scopus Google Scholar). the intake of dietary total lipid levels M. D. M. of for total fat intake plasma lipid levels as PubMed Scopus Google Scholar, of of intake by fat and diet in a population of 135: PubMed Scopus Google Scholar). were on methods were to determine the levels of total cholesterol methods for of PubMed Scopus Google Scholar, methods for of PubMed Scopus Google Scholar). We total cholesterol levels in We these on individuals in the CLEAR that lipid and of and that total cholesterol by R. Hatsukami P. of identified two loci that to risk for carotid artery Health PubMed Scopus Google Scholar). these previously we total cholesterol levels in within the CLEAR by cholesterol levels by A of the and total cholesterol levels is in These from a subset of the CLEAR are with in a recent meta-analysis of subjects on B. L. A. M. A. R. C. The of cholesterol with in at risk of meta-analysis of from 2012; PubMed Scopus Google Scholar). SNPs for the subjects the discovery were the Scholar). for individuals in 356 subjects the replication were by for the SNPs in with in individuals in SNPs were with at or cholesterol interactions on total cholesterol levels in the CLEAR from SNP to SNP identified in by et al. and with with APOB is in of of APOB and is not part of the and is in the of of APOB and is within the in the is in of of APOB and is not part of the and is in the of of APOB and is within the in the with NPC1L1 5′ nucleotide variant missense synonymous SNP to SNP identified in by et al. and is in of of APOB and is not part of the and is in the of of APOB and is within the in the table in a new 5′ nucleotide variant missense synonymous to the to we for gene-by-environment (GxE) interactions to identified as of total cholesterol by Teslovich et al. for total cholesterol in expressed in the or and in cholesterol handling is by SNP in and Teslovich et al. (7Teslovich T.M. Musunuru K. Smith A.V. Edmondson A.C. Stylianou I.M. Koseki M. Pirruccello J.P. Ripatti S. Chasman D.I. Willer C.J. et al.Biological, clinical and population relevance of 95 loci for blood lipids.Nature. 2010; 466: 707-713Crossref PubMed Scopus (2781) Google Scholar) identified SNPs at genetic loci that were significant at the genome-wide level for association with total cholesterol levels SNPs are to as We identified SNPs from to for SNPs that were in linkage disequilibrium C.J. a for and of SNPs PubMed Scopus Google Scholar) in the with SNPs by Teslovich et al. We identified such SNPs in for of identified to the of APOB SNPs and relevance to cholesterol absorption and total cholesterol, we on one APOB SNP that was to associated with variation in and While the of SNP is not in Teslovich and with the for the total cholesterol which was This APOB SNP was not in with the other identified APOB SNP We and H. a for and PubMed Scopus Google Scholar) to the of these loci to genes not expressed in the or with of through the total European Scholar). We genes for seven SNPs from genes CETP, and were We tested these seven SNPs for on total cholesterol in the discovery cohort. Four SNPs from one from and NPC1L1) were significant at the level and were for GxE SNPs for which a GxE was with a we analyses by CAAD case-control and We other SNPs in with these SNPs to determine better the SNP SNP associations with total cholesterol levels and GxE in the CLEAR SNP to primary SNP the specific loci identified in by Teslovich et al. and SNPs in in in the CLEAR from the CLEAR study cohort, and from the of an predicting total SNP are in the as by Teslovich et al. and with the of SNP of is of was for total cholesterol with two APOB SNPs are not in with other two APOB SNPs are not in with other 5′ variant missense synonymous SNP to primary SNP the specific loci identified in by Teslovich et al. and SNPs in in the CLEAR from the CLEAR study cohort, and from the of an predicting total SNP are in the as by Teslovich et al. and with the of SNP of is of was for total cholesterol These two APOB SNPs are not in with other table in a new 5′ variant missense synonymous analyses were the were in an for both dietary and total cholesterol were to three from the of of Scholar, of for Scopus Google Scholar). The for dietary cholesterol was by a to to for the of on total cholesterol, we to total cholesterol levels for and thereby a of total cholesterol. of total cholesterol from were as the to for interactions between the SNPs tested and dietary cholesterol intake through SNPs were an genetic as and interactions were to was we a and on replication to of between dietary and total cholesterol levels was the The in the the for total cholesterol level for a of dietary cholesterol. These were obtained of total cholesterol dietary such that and have the and and cholesterol were to the of total cholesterol by the of to in of the and total cholesterol for the are in The was of 1,128 subjects with a age of of which were with CVD the discovery the replication subset consisted of the subjects with which we for SNPs that were significant in GxE discovery The discovery and replication 356 as the replication subset was of subjects the discovery subset was The with carotid subjects with carotid and with the discovery there were a significantly of with the replication subset The of and were and The of were significantly in the discovery with the replication subset and total cholesterol were in the cohort, with a and The of total cholesterol by not by CAAD case-control or and respectively). The SNP is described in and SNPs from genes CETP, and were of the seven SNPs by Teslovich et al. (7Teslovich T.M. Musunuru K. Smith A.V. Edmondson A.C. Stylianou I.M. Koseki M. Pirruccello J.P. Ripatti S. Chasman D.I. Willer C.J. et al.Biological, clinical and population relevance of 95 loci for blood lipids.Nature. 2010; 466: 707-713Crossref PubMed Scopus (2781) Google Scholar) and for the in the of the the genes are in The two SNPs within APOB are not in with other the and NPC1L1 SNPs by the that were in with the SNPs identified by Teslovich et al. and within the discovery subset of CLEAR APOB SNP and with NPC1L1 SNP the association between the seven SNPs and total cholesterol, we for and Four SNPs three genes (APOB, CETP, and NPC1L1) were significant at the level for association with total cholesterol levels not to the SNP to the GxE and these SNPs were on associations with total cholesterol, we not a to the of these SNPs for these SNPs with on total cholesterol were to GxE to We to the of GxE interactions the of total cholesterol for and the between the of the SNPs and dietary cholesterol intake on the of total cholesterol levels of the SNP interactions with dietary cholesterol levels on the of total cholesterol in the discovery subset of CLEAR are in the SNPs an APOB missense and an NPC1L1 synonymous rs2072183 GxE interactions with dietary cholesterol that were significant at the level for prediction of total cholesterol. These two SNPs and SNPs in were for and replication to the to interactions with CAAD case-control or we analyses by CAAD case-control and These analyses not in or of for the SNP dietary cholesterol, or the dietary for the APOB SNPs in the discovery subset of the of NPC1L1 SNP in males and were different and the GxE was not significant for the in males and in and the GxE in for males We the of the SNP on total cholesterol by gender and that NPC1L1 SNP total cholesterol in males and not in in cohort. We tested for a in the prediction of total cholesterol and not significant however, we identified a report of gender in the L. of rs2072183 SNP and lipid levels in the and Health 2012; PubMed Scopus Google Scholar). we a in the analyses and that the cholesterol was which is in with in the CLEAR the and report of interactions and the between NPC1L1 we a to a replication analyses for these NPC1L1 of the of the two and and SNPs in interactions with dietary cholesterol levels on total cholesterol within the and analyses are in The APOB SNP GxE in the as the GxE of NPC1L1 SNP in males NPC1L1 SNP rs2072183 a replicated of the two SNPs identified other SNPs in that were in cohort. whether any of these were at predicting interactions with dietary cholesterol on the of total cholesterol levels, we missense with or we rs2072183 synonymous with in the for GxE analyses both the discovery and replication APOB SNP was significant of the two NPC1L1 SNPs was significant rs2072183 We an to the of total cholesterol to GxE The APOB SNP accounted for approximately of total cholesterol the NPC1L1 SNP was of the in total cholesterol levels in a subset of the cohort. the potential of other cardiovascular we analyses total and CAAD case-control were and to the and for analyses of the SNP and cholesterol for APOB SNP and NPC1L1 SNP on the of total plasma cholesterol. all three in with the both cholesterol interactions significant NPC1L1 APOB The of the of dietary cholesterol intake and the of APOB SNP on total cholesterol, in the are in The was associated with a in total cholesterol levels with dietary cholesterol NPC1L1 SNP the were associated with a less in total cholesterol the males in the CLEAR analyses not the and at the of total cholesterol for the of these two we a for APOB SNP the is associated with a in total cholesterol, the is for NPC1L1 SNP cholesterol level is the and risk factor for CVD (2Verschuren W.M. Jacobs D.R. Bloemberg B.P. Kromhout D. Menotti A. Aravanis C. Blackburn H. Buzina R. Dontas A.S. Fidanza F. et al.Serum total cholesterol and long-term coronary heart disease mortality in different cultures: Twenty-five-year follow-up of the seven countries study.JAMA. 1995; 274: 131-136Crossref PubMed Google Scholar, E.J. Ramsay L.E. Haq I.U. Ghahramani P. Jackson P.R. Yeo K.R. Yeo W.W. Coronary and cardiovascular risk estimation for primary prevention: validation of a new Sheffield table in the 1995 Scottish health survey population.BMJ. 2000; 320: 671-676Crossref PubMed Scopus (161) Google Scholar). A GWAS identified over 95 loci for blood lipid phenotypes (4Kathiresan S. Manning A.K. Demissie S. D舗Agostino R.B. Surti A. Guiducci C. Gianniny L. Burtt N.P. Melander O. Orho-Melander M. et al.A genome-wide association study for blood lipid phenotypes in the Framingham Heart Study.BMC Med. Genet. 2007; 8 (Suppl 1): S17Crossref PubMed Scopus (245) Google Scholar, T.M. Musunuru K. Smith A.V. Edmondson A.C. Stylianou I.M. Koseki M. Pirruccello J.P. Ripatti S. Chasman D.I. Willer C.J. et al.Biological, clinical and population relevance of 95 loci for blood lipids.Nature. 2010; 466: 707-713Crossref PubMed Scopus (2781) Google Scholar). the variants identified for total cholesterol accounted for of (7Teslovich T.M. Musunuru K. Smith A.V. Edmondson A.C. Stylianou I.M. Koseki M. Pirruccello J.P. Ripatti S. Chasman D.I. Willer C.J. et al.Biological, clinical and population relevance of 95 loci for blood lipids.Nature. 2010; 466: 707-713Crossref PubMed Scopus (2781) Google Scholar), the of such as to variants O. E.S. The of genetic interactions 2012; PubMed Scopus Google Scholar), genetic both between loci and GxE have to of O. E.S. The of genetic interactions 2012; PubMed Scopus Google Scholar). of genes with dietary intake have identified for D.R. A.C. K. M. O. et interactions and and in the Health Genet. PubMed Scopus Google Scholar, J.H. et and genetic risk of Med. 2012; PubMed Scopus Google Scholar). in study we have a novel and replicated that variants at APOB in males and and NPC1L1 in males GxE interactions with dietary cholesterol intake in the prediction of total cholesterol. NPC1L1 to have an with gender as with dietary cholesterol predicting total cholesterol, in These replicated GxE interactions may to in the absorption or processing of dietary cholesterol. the to which are on A recent study that A.C. E.M. intake by age in and in 2012; PubMed Scopus Google Scholar). While the hypothesized that is to in and is a is that of the are to genetic factors in We are not that in in the relationship between dietary and total cholesterol have previously is one of the primary lipoproteins of and all including and are two of which is in the and which is in the K. A.V. of two an with the and an a in both and PubMed Scopus Google Scholar) and is a for the in or are in hypercholesterolemia evidence for of lipoproteins to as a cause of primary PubMed Scopus Google Scholar). is a of in the to transport lipid to the and is in the absorption of lipids J.P. of of a new from PubMed Scopus Google Scholar). is through a single of a to at nucleotide that in a at K. A.V. of two an with the and an a in both and PubMed Scopus Google Scholar). associated with are in the M. A. F. of in 2000; PubMed Scopus Google Scholar). and not and dietary fat on in the of in a in the of L. of over in and fat an of and PubMed Scopus Google Scholar). one potential pathway the SNPs identified in study could affect the relationship between dietary and total cholesterol could at the of absorption or in the follow-up is to the specific mechanisms by which these SNPs influence the relationship of dietary to total cholesterol levels. NPC1L1 is in the of the and in the of M. A. M. et is for cholesterol PubMed Scopus Google Scholar) and is in the of cholesterol absorption by with J.H. L. L. M. The of NPC1L1 cholesterol and in cholesterol PubMed Scopus Google Scholar). studies of NPC1L1 in a significant in cholesterol absorption that was by M. A. M. et is for cholesterol PubMed Scopus Google Scholar, J.P. C. K. A. of NPC1L1 lipid transport and 2005; PubMed Scopus Google Scholar). in have the potential mechanisms through which variants affect cholesterol through of NPC1L1 or L. of the NPC1L1 variants identified from cholesterol PubMed Scopus Google Scholar). et al. L. of rs2072183 SNP and lipid levels in the and Health 2012; PubMed Scopus Google Scholar) a for NPC1L1 SNP rs2072183 association with total cholesterol levels in a of subjects and males was not significant in a of subjects the rs2072183 SNP in the subset was significant in not in the both the and as as in the rs2072183 was associated with an in total cholesterol in et al. not for we a in for NPC1L1 SNP with the SNP on total cholesterol significant in males and not and the an in cholesterol levels in males we have identified a potential for the of association with total cholesterol in differential absorption or processing of dietary cholesterol on NPC1L1 in which is not in genetic have other for SNPs in of identified in study in APOB and NPC1L1. APOB SNP was associated with a to L. genetic variants the to a 2010; PubMed Scopus Google Scholar). This is with the GxE we and on differential dietary cholesterol absorption or processing by an APOB variant in with two SNPs 5′ in the APOB not in with the SNPs of have associated with an risk of and A. P. et of genes in the lipid pathway and risk of and a case-control study in PubMed Scopus Google Scholar). association is not to between the of SNPs in to determine which is the genetic variant the associations. The APOB SNP is a missense in is in with a which in Ser4338Asn as as SNP While is is in the of total of a that The SNP a for however, any of the tested variants could the is in with seven SNPs in the which are less to the The NPC1L1 SNP rs2072183 is synonymous and less to the 5′ SNP in which was significant for GxE in both the discovery and replication these are in with an 5′ have identified a NPC1L1 missense which a differential to from with in the P. P. et variation in NPC1L1 and association with in to 2005; PubMed Scopus Google Scholar). These and other variants the SNPs identified in both APOB and NPC1L1 may the findings of GxE interactions in study. for findings may from SNPs in to that alter recent studies into loci that influence an of SNPs at and F. Smith J.D. S. C. et and of loci in primary Genet. PubMed Scopus Google Scholar), including APOB SNP which is in the of the of to the APOB is not to in in the and we the of SNP to relationship to the SNPs in cohort. recent of the of identified a to as to and in and S. B. S. R. A.K. et in factor 2012; PubMed Scopus Google Scholar, R. H. R. H. et of variation in 2012; PubMed Scopus Google Scholar, A. S.F. P.J. F. S. R. et of in the 2012; PubMed Scopus Google Scholar). have the 5′ to the K. R. D. L. N.P. S. K. of and factor at by PubMed Scopus Google Scholar) and H. R. J.C. M. R.A. interactions and for the factor 2010; PubMed Scopus Google Scholar). These in with the study findings of at and may as a for findings of cholesterol interactions for APOB and NPC1L1. studies to determine which of these both and are for the APOB and NPC1L1 GxE of the study a with dietary intake genetic and total lipid through of SNPs identified as significant for interactions with dietary cholesterol in the discovery subset of the cohort, we were to findings in an subset of the CLEAR study. the of which the of these and the to in GxE seven SNPs in studies could a of variation at loci for GxE interactions and the for NPC1L1. to we a to genetic loci to that are expressed in the the and and tested SNPs for GxE interactions were at significant for association with total cholesterol in these is that GxE interactions could is study identified and replicated novel GxE interactions between SNPs in APOB in both and in NPC1L1 in and dietary cholesterol intake in the prediction of total cholesterol replicated interactions may account for part of the of such as total cholesterol levels. is to the whether through or through changes, by which these or variants in alter dietary cholesterol absorption or processing and alter the relationship between dietary and total cholesterol. The the study participants. the for and Teslovich and for with carotid artery disease Carotid Lesion Epidemiology and Risk cardiovascular disease genome-wide association study gene-by-environment linkage disequilibrium single nucleotide total cholesterol