Phytosterols, which encompass plant sterols and stanols, are steroid compounds similar to cholesterol which occur naturally in plants and vary only in side chains and/or double bonds. Although almost structurally identical to cholesterol, plant sterols and stanols possess an extra methyl or ethyl group at the C24 position, as well as, in the case of stanols, a single rather than double bond at the C5 position of the molecule. Plant sterols, found principally in foods including nuts and legumes in human diets, are typically ingested at levels of less than 400 mg/day.
Consumption of plant sterols and stanols at levels over 1 g per day has been demonstrated to reduce circulating cholesterol levels, particularly low density lipoprotein-cholesterol (LDL-C) levels, in humans. Although interest in use of these molecules to control circulating lipid levels has existed since the 1950s (Pollak and Kritchevsky 1981), the past 10 years has seen a resurgence in use of plant sterols and stanols as dietary agents capable of modulating circulating lipid levels (Law 2000; Katan, Grundy et al. 2003; Berger, Jones et al. 2004; Abumweis, Barake et al. 2008). Specifically, studies conducted on cholesterol-lowering efficacy of plant sterols and stanols have concluded that LDL-C reductions of 8-15% are realizable with plant sterol or stanol intakes of 1-2 g per day. As such, plant sterols and stanols possess substantial opportunity as a functional food ingredient for the treatment and prevention of dyslipidemia and subsequent heart disease risk reduction. Accordingly, due to highly favorable efficacy, safety and hedonic property profiles of this class of molecules, several major food companies have commenced marketing branded products which include plant sterols and stanols (eg. BioBest Plant sterol yogurt and Unilever Becel Pro-Activ® Margarine in Canada). On the basis of the almost indisputable efficacy profile for these products, combined with excellent acceptability, flexibility of food delivery, and cost of raw product, plant sterols and stanols exist as one of the most successful functional food ingredient initiatives of the past decade.
Despite the repeated demonstration of efficacy of plant sterols and stanols as cholesterol-lowering agents in humans, the growing human trial data-base reveals that not all individuals respond to an equal degree to standard dosages of plant sterol/stanol materials (See Appendix I). Compliance, food matrixing, and dosage issues have been thought to play a role in the lack of consistent response to sterols and stanols across individuals within a population, however, it is now increasingly clear that independent of such factors, some individuals respond immediately and with a major shift in lipid profiles, while other individuals are much more resistant or completely insensitive to plant sterol/stanol challenges (Jakulj, Trip et al. 2005; Rudkowska, AbuMweis et al. 2008; Zhao, Houweling et al. 2008). It has been suggested that response to phytosterols may be linked to the amount of cholesterol produced endogenously rather than absorbed from the diet, with a recent retrospective clinical analysis showing significantly higher cholesterol synthesis in non-responders to plant sterols than responders (Rideout, Harding et al. 2010). This heterogeneity of responsiveness appears to be subject-specific, with individuals showing consistency of lipid level response to plant sterols across repeated challenges (See Appendix II).
The substantial range of responsiveness has at least two important ramifications. First, marked responsiveness differences across individuals indicate that particular aspects of the cholesterol metabolic pathway must be selectively altered by plant sterols. Identification of such areas will enable better understanding of the precise mechanisms of plant sterol action. Second, discerning what factors are responsible for these variations in response will permit tests to be developed to distinguish individuals for whom plant sterols would exist as useful dietary adjuncts, from those who may require either alternative dietary approaches, or a pharmaceutical regimen to effectively manage circulating lipid levels. As such, there is a pronounced need to understand the genetic and metabolic factors that explain the substantial degree of heterogeneity in responsiveness of individuals to plant sterols and stanols.
Literature Review
Cholesterol Lowering Actions of Plant Sterols and Stanols
Actions Within the Intestinal Lumen
Several mechanisms have been reported through which phytosterols are believed to function as cholesterol-lowering agents. Sterols and stanols appear to function equally, so this proposal will focus on plant sterols. It is generally accepted that these molecules decrease cholesterol concentrations by inhibition of cholesterol absorption via competition with cholesterol during the formation of micelles in the intestine or inhibiting cholesterol uptake by the enterocytes as micelles move from the intention lumen to the brush border, as has been capably reviewed by Trautwein et al.(Trautwein, Duchateau et al. 2003)
Actions Within the Enterocyte
Phytosterols also appear to hinder cholesterol esterification and incorporation into chylomicrons within the enterocytes (Child and Kuksis 1983; Ikeda and Sugano 1983). This mechanism suggests that phytosterols must be consumed simultaneously with dietary cholesterol in order to inhibit cholesterol absorption. On the other hand, Plat et al. (2000) have shown that consumption of plant stanols once a day with lunch results in a similar LDL-C reduction compared with consumption of a similar dose of plant stanols divided over three daily meals. In addition, injection of plant sterols has been shown to lower circulating cholesterol concentrations in hamsters (Vanstone, Raeini-Sarjaz et al. 2001). It is speculated from these studies that plant sterols may have unknown molecular targets within intestinal enterocytes to hinder cholesterol absorption. These actions appear to involve at least two specific transporter systems within the enterocyte, namely the Niemann-Pick C1 Like 1 and ATP Binding Cassette transporter proteins (Yu 2008; Calpe-Berdiel, Escola-Gil et al. 2009).
Role of Niemann-Pick C1 Like 1 (NPC1L1) and ATP Binding Cassette (ABC) Transporter Proteins in Cholesterol Absorption
The mechanisms by which cholesterol is transported within various subcellular pools of the enterocyte are only now being defined. However, several observations of disruptions of normal subcellular cholesterol trafficking by NPC1L1 (Davis Jr and Altmann 2009), ABCG5 and ABCG8 gene variants (Berge, Tian et al. 2000), suggest the existence of specific sterol transport and sorting pathways. The NPC1L1 gene was mapped on chromosome 7p13 and proposed to have ~50% amino acid homology to NPC1 (Davies, Levy et al. 2000). This protein also contains a sterol-sensing domain, which is present in other key regulators of cholesterol homeostasis including 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMG-CoA R) (Chin, Gil et al. 1984), SREBP cleavage-activating protein (Hua, Nohturfft et al. 1996). Phytosterols likely exert their cholesterol-lowering action through NPC1L1 transporter genes, as recent studies in mice show that the NPC1L1 transporter protein is involved in intestinal absorption of numerous sterols (Davis Jr and Altmann 2009). In addition to the NPC1L1 transporter protein, ATP-binding cassette transporter genes, ABCG5 and ABCG8, are involved in the intestinal transport of sterols as well as in their excretion in bile (Kidambi and Patel 2008). Therefore, it is also postulated that phytosterols may exert their cholesterol-lowering action by stimulating the ABCG5/ABCG8 complex, which subsequently results in increased cholesterol excretion.
Genetic Polymorphism of Cholesterol Absorption Transporters and other Cellular Lipid Regulators
It is now well established that several major and minor polymorphisms exist for NPC1L1 and ABCG5 and G8 (Kidambi and Patel 2008; Davis Jr and Altmann 2009). For NPC1L1, several variants have been reported to exist including R306C, I647N, R693C, E1308K and N387S (Cohen, Pertsemlidis et al. 2006). Recent data indicate that sequence variations contribute to functional variations associated with cholesterol absorption and LDL-C levels (Simon, Karnoub et al. 2005; Cohen, Pertsemlidis et al. 2006). In individuals identified as characteristically low cholesterol absorbers, DNA sequencing revealed 20 variants. It has been speculated that susceptibility to common diseases is actually a result of multiple rare alleles which produce a larger phenotypic effect. For cholesterol metabolism, certain rare variants may contribute to variations in plasma cholesterol levels and metabolic control, with the effect on the phenotype being more modest. Modifications in common apolipoprotein E (APOE) alleles combined with multiple rare alleles in NPC1L1 may contribute to the phenotype. Similarly, several common variants of the ABCG5 (Q604E, R50C and A478T) and G8 (D19H, Y54C, V632A and T400K) transporters exist (Berge, von Bergmann et al. 2002; Gylling, Hallikainen et al. 2004; Plat, Bragt et al. 2005).Variations in genotype have for the ABC transporter system, as for NPC1L1, been shown to reflect efficiencies of cholesterol absorption in humans. Polymorphisms in sterol synthesis enzymes, sterol transport proteins, cholesterol digestive enzymes, and hepatic cholesterol regulators are being identified which suggest that the role of genotypic variation may be extremely important in control of cholesterol metabolism, as well as on the efficacy of dietary factors such as plant sterols and stanols on this process.
Other important genes regulating lipid metabolism which could be involved in responsiveness to plant sterols include scavenger receptor-BI (SR-BI), cholesterol ester transfer protein (CETP), HMG-CoA R, APOE, carboxyl ester lipase (CEL), and proprotein convertase subtilisin/kexin type 9 (PCSK9). The A350A SR-BI polymorphism was found significantly associated with atherogenic versus non-atherogenic lipid profiles (Morabia, Ross et al. 2004). CETP polymorphisms such as I405V have been associated with plasma cholesterol concentration (Lottenberg, Nunes et al. 2003). SNPs in the HMG-CoA R gene (V638I) and APOE (E4, E2) were found to be associated with LDL-C lowering effect of pravastatin (Thompson, Man et al. 2005). Similarly, rare alleles of CEL (Bengtsson-Ellmark, Nilsson et al. 2004) and PCSK9 (Polisecki, Peter et al. 2008) are associated with plasma LDL-cholesterol concentrations. However, it is not known whether SR-BI, CETP, HMG-CoA R, APOE, CEL, and PCSK9 gene polymorphisms are associated with the response of LDL-C to plant sterols.
Evidence for Within-Individual Consistency in Sterol and Stanol-Induced
Cholesterol Lowering across Individuals
Notable across studies examining the response of plasma cholesterol levels to the consumption of plant sterols and stanols has been the extreme heterogeneity of responsiveness despite control of compliance and consistency in dosage level. In individuals provided 1.8 g sterols for 4 weeks, the mean reduction of LDL-C ranged from -6 to -33% across tertiles (Jones, Ntanios et al. 1999). This range of response is apparently not due to methodological variability, as we have demonstrated a good degree of within-subject consistency (R2 = 0.7168, p<0.01) in the responsiveness of total circulating levels in volunteers to plant sterol consumption over two separate periods of 1 month, compared with a control diet (See Appendix II). Here, all diets were provided under controlled conditions with plant sterols given under supervision. These data suggest that it is likely that the high degree of the heterogeneity in responsiveness of circulating LDL-C to plant sterols includes a strong genetic component.
Cholesterol Synthesis and Phytosterol Response
Hepatic cholesterol synthesis is integral to the maintenance of whole-body cholesterol homeostasis and is regulated by multiple dietary factors (Jones, Pappu et al. 1996; Jones 1997; Sundram, French et al. 2003). It has been suggested that high basal hepatic cholesterol synthesis may confer protection against diet-induced hypercholesterolemia by creating a cholesterol buffering capacity in which dietary cholesterol can maximally reduce cholesterol biosynthesis through negative feedback inhibition (Ness and Chambers 2000). In agreement with this hypothesis, a recent retrospective clinical analysis suggests that high basal cholesterol synthesis may reduce the effectiveness of cholesterol lowering by plant sterols. This study found that individuals with low basal cholesterol synthesis, determined by deuterium incorporation, had clinically significant reductions in LDL-C (-15.2 ± 1.0 %) while those with relatively high basal cholesterol synthesis were unresponsive to PS therapy, displaying no reduction in LDL-C following PS consumption (3.7 ± 1.1 %). Similarly, several studies have observed a lower LDL-C lowering response to PS therapy in subjects with high basal sterol synthesis marker concentrations that provides an indirect reflection of whole-body cholesterol synthesis (Gylling, Radhakrishnan et al. 1997; Gylling, Puska et al. 1999; Carr, Krogstrand et al. 2009). After pre-screening for high and low plasma lathosterol/campesterol ratios, which should reflect the balance between cholesterol synthesis and absorption, greater LDL-C reductions have been observed in a small group of subjects (n=8) with plasma marker ratios indicative of low basal cholesterol synthesis/high cholesterol absorption, compared with subjects with marker ratios reflecting high basal synthesis/low cholesterol absorption in response to plant stanol consumption (-13.8 vs +4.2%) (Thuluva, Igel et al. 2005).
Accordingly, it is important to assess and identify genetic factors which can explain this variability and assist in identifying a priori individuals for which plant sterol or stanol treatment would be an appropriate therapeutic strategy, assuming an adequate safety profile for such products.
Safety of Use of Plant Sterols and Stanols
It is estimated that well over 2,400 subjects have participated in clinical trials involving supplementation of plant sterols and stanols with dosages up to 10 g per day, with no noted adverse events (Korpela, Tuomilehto et al. 2006; Tuomilehto, Tikkanen et al. 2009). Additionally, Cytellin, a drug primarily consisting of sitosterol was prescribed for more than 20 years with an excellent record of safety (Lichtenstein and Deckelbaum 2001).There is likewise suggestion that ancestral intakes of plant sterols may have been substantially higher than present day consumption levels. Available sterol and stanol esters are now listed as GRAS (Generally Recognized as Safe) foods by the FDA for the general population (FDA 2011). The only group to caution against consumption of plant sterol functional foods are those who have a highly rare autosomal recessive disorder termed phytosterolemia, also known as sitosterolemia, as consumption of plant sterols may lead to significant increases in the blood (Lichtenstein and Deckelbaum 2001). A sufficient body of evidence currently exists indicating that intakes of phytosterols at levels of 1 to 2 g per day do not significantly affect other biological parameters, while having the ability to significantly lower LDL-C (Katan, Grundy et al. 2003; Abumweis, Barake et al. 2008; Demonty, Ras et al. 2009). Thus, phytosterols can be regarded as being safe.
Research Objectives and Hypotheses
Specific Objectives
The overarching objective of the present research program is to identify a genetic basis for heterogeneity in responsiveness of lipids to plant sterol use, and to identify which regulators of cholesterol metabolism associate with the genetic factors identified. The long-term goal is to predetermine who will, and will not, respond to plant sterols as functional food ingredients. Specific objectives include:
Characterize a continuum of responsiveness of plasma LDL-C levels to plant sterol consumption in a cohort of individuals with hypercholesterolemia who are high or low cholesterol synthesizers as estimated by plasma lathosterol levels.
Identify genotypic traits for a host of lipid regulator systems including, but not limited to, NPC1L1, ABCG5, ABCG8, SR-BI, CETP, HMG CoA R, APOE, CEL, and PCSK9 that associate with the degree of plasma LDL-C responsiveness to plant sterol administration.
Identify haplotype traits that associate with expressed phenotypes in terms of basal kinetic parameters for cholesterol, as well as the responsiveness of these kinetic parameters to plant sterol intervention.
Examine the expression level of cholesterol and phytosterol responsive genes to plant sterol consumption and associate identified genotypic traits with gene and protein expression patterns in peripheral blood mononuclear cells.
Hypotheses Null hypotheses to be tested include:
Responsiveness of lipid lowering efficacy to plant sterol consumption does not differ between hyperlipidemic low and high cholesterol synthesizing individuals.
There will be no association identified between genotypic trait, responsiveness of lipid lowering efficacy, and expression pattern of sterol responsive genes to plant sterol consumption in hyperlipidemic individuals.
There will be no associations identified between different haplotype traits and basal kinetic parameters for cholesterol, or their responsiveness to plant sterol consumption in hyperlipidemic individuals.
Clinical Intervention Trial
Protocol
The clinical trial will consist of two phases of 28 days, during which subjects, pre-identified as high synthesizers and low synthesizers, according to their screening level of lathosterol to cholesterol, will consume a single meal a day for each phase under supervision. The two phases will be separated by a washout period of 4 weeks, during which subjects will consume their habitual diets. The 2 phases will include:
- Control phase: 11 g of Becel margarine contain no plant sterols.
- Plant sterol phase: 25 g of Becel Pro.Activ, containing 2.0 g/day plant sterols
Sample Size and Power
The difference in fractional synthesis rate (FSR) in % pools per day between responders and non-responders in previous plant sterol trials as an estimate of treatment effect size to calculate required sample size.
To calculate the size of the clinical trial the following formula and values will be used:
α set a 0.05 to limit the chance of Type 1 error, β set to 0.05 (power=0.95) to limit type 2 error, therefore Power index ( PI) =3.60
σ is 2.148 (%pool/Day) the standard deviation in FSR on control diet in 113 individuals from 3 previous plant sterol trials (Jones lab, unpublished data) .
μr and μnr are 4.34 (%pool/Day) and 5.48 (%pool/Day), the mean FSR values for the data set of plant sterol responders and non-responders respectively in the 113 individuals (Jones lab, unpublished data).
This calculation yields a required sample size of 94 individuals.
Trials looking at genetic variations such a single nucleotide polymorphisms (SNPs) benefit from larger trial population, which allow a better chance of having individuals with rare mutations and a greater chance of finding significant associations between particular SNPs and trial outcomes. A sample size of n=120 individuals, (60 Responders and 60 Non-responders) will be recruited for the trial, this exceeds the samples size required for detecting a physiologically significant difference in FSR between responders and non-responders to plant sterol supplementation at a power of 0.95 and will allowing for greater genetic variation in the study participants.
Subject Selection
Males and female subjects, aged 35-75 years, with body mass indices (BMI) under 40 kg/m2 and plasma LDL-C levels between 3.0-6.5 mmol/L will be recruited using flyers, newspaper, radio, email and targeted internet (Facebook) advertisements in the greater Winnipeg MB and Washington DC areas. Volunteers will undergo historical and physical examination to exclude those with diabetes mellitus, kidney disease, or liver disease, smokers or people who report consuming a large amount of alcohol. Moreover, volunteers cannot have taken any medication known to affect lipid metabolism (cholestyramine, colestipol, niacin, clofibrate, gemfibrozil, probucol, HMG CoA R inhibitors, high dose dietary supplements, fish oil capsules or plant sterol) for at least the previous 2 months. At the first visit, 10 ml of blood will be obtained from volunteers in order to screen for blood glucose, creatinine, aspartate aminotransferase (AST), alanine aminotransferase (ALT), Lactate dehydrogenase (LDH), Alkaline Phosphatase (ALKP) and lipid profiles that include cholesterol, triglyceride, LDL-C and HDL-C. Only hypercholesterolemic, but otherwise healthy, subjects will further screened for plasma sterol levels. In these otherwise healthy individuals, plasma sterol levels will be measure by GC-MS and the ratio of lathosterol to cholesterol will be used as an estimate of cholesterol synthesis. The top and bottom 15% of eligible participants according to lathosterol to cholesterol ratio will be considered high and low cholesterol synthesizers and be accepted into the trial.
Our goal is to screen 400 subjects for the study, (200 at each of the two sites), yielding an approximately 30 high and 30 low cholesterol synthesizers at each of the two sites (n=120 total participants).
Study diets will optimize compliance both for macronutrient and micronutrient as well as for plant sterol intake, as per our previous clinical trials. One meal will be prepared each day for every subject within the metabolic kitchen of the RCCRU and the BHNRC. A 7-day rotating menu cycle will be used. Subjects will consume a daily evening meal on site and under supervision. Becel (Unilever, Canada) or Becel Pro.Activ (Unilever, Canada) will be incorporated into the meals as appropriate for each phase. Dietary instructions will recommend not to consuming excess (> 2 drinks daily) alcohol or coffee. Since plant sterols possess no taste or mouth-feel characteristics, it will be possible to maintain the human participants blinded to which treatment they are receiving. During the study period body weight will be monitored so that the baseline weight does not fluctuate by more than 5% body weight. If any of the subjects gain or lose weight during the first week, meal size adjustments will be made.
Blood Collection
Twelve-hour fasting blood samples (20 ml) will be collected on days 1, 2, 24, 25, 26, 27 and 28 of the trial phases. Blood samples will be centrifuged for 20 min at 520 X g and the separated aliquots will be frozen at each site until analysis.
Analysis of Cholesterol Absorption, and Synthesis
Ninety-six hours before the end of each phase, subjects will ingest 75 mg [3, 4-13C]-cholesterol, followed by deuterium water (D2O) will be given at a dose of 0.7g/kg body water (estimated at 60% of total body weight). These two isotopes are stable tracers, thus pose no radiation hazard and can be safely administered to human subjects at any life-cycle stage. The 13C-cholesterol will be dissolved in 5 g of warmed margarine, and consumed on a slice of toast. The D2O will be administered orally following the consumption of 13C-cholesterol margarine and toast. A fasted blood sample will be taken at baseline on day 24 prior to isotope administration, as well as fasting samples on days 25, 26, 27, and 28 to monitor enrichment levels of 13C-cholesterol in plasma total cholesterol. The enrichment of plasma cholesterol by 13C will be used to access cholesterol absorption. The change in D enrichment within red blood cell (RBC) free cholesterol will be determined as an index of synthesis over days 24 and 25. Subjects will consume their test meals as usual over these days.
Blood Lipid Analysis
Serum samples will be analyzed in duplicate for lipid profiles including total cholesterol, LDL-C, HDL-C and triglyceride levels using an automated analyzer (Roche Diagnostic, Indianapolis, IN) in Dr Baer's laboratory. Analyses will be performed against reference standards authenticated by the Center for Disease Control (Atlanta, GA). In addition, plasma alpha- and β-tocopherol and total carotenoid levels will be analyzed by high performance liquid chromatography utilizing a reverse phase column, based on established procedures, within Dr Baer's laboratory.
Plasma Plant Sterol Concentration Assessment
Plasma samples will be used to quantify concentrations of blood sterols and sterol precursors. Briefly, plasma samples will be saponified with methanolic KOH solution. Sterols will then be extracted twice with petroleum ether. Extracted sterols will be derivatized using tri-methylsylation (TMS) procedures. Authenticated internal standards will be added to the TMS derivatized samples and sterol analysis will be carried out by gas chromatography mass-spectrometry (GC-MS). Campesterol, sitosterol, campestanol, sitostanol, cholestanol and dihydrobrassicasterol concentrations will be assessed, as well as lanosterol, desmosterol and lathosterol levels which are indirect indicators of cholesterol synthesis. Circulating campesterol and sitosterol levels have been used as indices of cholesterol aborption previously, however, intervention with plant sterols introduces error in this assay.
Determination of Cholesterol Absorption and Synthesis
Free cholesterol extracted from RBCs will be used to determine 13C- cholesterol enrichment according to established procedures in Dr. Jones' laboratory. Red blood cells contain almost exclusively free cholesterol representing the central rapidly exchangeable M1 pool. Briefly, lipids will be extracted from RBC in duplicate. The 13C enrichments of free cholesterol within the lipid extracts will be measured by differential IRMS using a GC-combustion IRMS system (Finnigan, Delta5). Enrichments are expressed relative to PDB limestone standard of the National Bureau of Standards (NBS). Linearity and gain of response of the IRMS instrument are assessed using a reference tank CO2 and National Bureau of Standards of known isotopic enrichment. Precision of measurement expressed as coefficient of variation for replicate 13C enrichment analyses has been shown to be 0.08 del (parts per thousand relative to NBS -PDB standards).
The 13C enrichment in 24, 48, 72 hr RBC free cholesterol, relative to baseline (day 25, t=0), samples will be utilized to compare changes in cholesterol absorption across and between plant sterol and placebo phases, as well as across responders and non-responders.
Cholesterol biosynthesis will be determined as the rate of incorporation of D from body water into RBC membrane free cholesterol over the period between 0 (day 25) and 24 hr at the end of each treatment according to established procedures in Dr. Jones' laboratory. D enrichment will be measured in RBC free cholesterol and plasma water. To determine plasma cholesterol D enrichment, total RBC lipids will be separated, pyrolyzed and analysed for D content using a TCEA- pyrolysis/IRMS (Finnigan, Delta 5). For D, enrichments are expressed relative to Standard Mean Ocean Water (SMOW) and a series of NBS standards of known enrichment analysed concurrently on each day of measurement to correct for any variations in linearity of gain of response of the IRMS. Precision of measurement expressed as coefficient of variation for replicate D enrichment analyses have been shown to be 2.3 del (parts per thousand relative to SMOW standard).
Fractional synthesis rate (FSR) is taken to represent the RBC free cholesterol D enrichment values relative to the corresponding mean plasma water sample enrichment after correcting for the fraction of protons that accept during short term synthesis. The FSR represents that fraction of the cholesterol pool that is synthesized in 24 hr.
Genotyping
Polymorphisms will be analyzed by PCR amplification and sequencing of gene regions containing previously published SNPs. Genomic DNA will be extracted from white blood cells by column based solid phase DNA extraction. Polymorphic regions of the NPC1L1, ABCG5, ABCG8, SR-BI, CETP, HMG CoA R, APOE, CEL and PCSK9 mutation genes (as indicated in Appendix III) will be amplified by PCR and sequenced using BDT-Sanger sequencing.
Haplotype Analysis
The expectation-maximization algorithm will be used to estimate the maximum-likelihood haplotype frequencies from multilocus genotypic data without known gametic phase using Arlequin software, Version 2.00. All subjects with one or more missing genotypes will be excluded for haplotype construction. The haplotypes that will be assigned unambiguously to subjects will be further analyzed. The linkage disequilibrium (LD) between polymorphisms will be similarly calculated with the same software.
Expression Analyses
In order to examine the expression level of cholesterol responsive genes to plant sterol consumption and associate identified genotypic traits with gene and protein expression patterns, mRNA and protein expression analyses will be performed in peripheral blood mononuclear cells according to established procedures in Dr Jones' laboratory. Target genes will include SR-BI, CETP, HMG-CoA R APOE, CEL, and PCSK9.
For mRNA expression experiments, total RNA will be isolated from peripheral blood mononuclear cells using TRIzol reagent (Invitrogen Canada Inc., Burlington, ON). RNA concentration and integrity will be determined with spectrophotometry (260 nm) and agarose gel electrophoresis, respectively. Prior to RT-PCR, RNA will be treated with DNase (Invitrogen) to remove potential contaminating genomic DNA and trace proteins. Primers for amplification of target and housekeeping genes will be developed using Primer 3 software. To examine the gene expression of target genes, quantitative real time reverse transcription polymerase-chain reaction (RT-PCR) will be performed in a one-step reaction using Quantitect SYBR Green RT-PCR kit (Qiagen Inc., Mississauga, ON, Canada) with a 7500 Real Time PCR System (Applied Bioscience) with commercially available primer sequences. Target gene expression will be normalized against that of b-actin (housekeeping control) and relative gene expression will be determined using the ∆∆Ct method. All PCR products will be run on a 1% agarose-gel to verify the specificity of amplification.
For immunoblotting experiments, protein extracts from peripheral blood mononuclear cells will be boiled in 2´SDS loading buffer (500 mM β-mercaptoethanol, 2% SDS, 0.1% phenol blue, 10% glycerol, 250 mM tris-HCL, pH 6.8) for 5 min. Aliquots (20 mg protein), will be subjected to 10% SDS-PAGE and transferred to polyvinylidene difluoride membrane (Millipore, Billerica, MA). The membranes will be blocked at room temperature for 1 h with 6% non-fat dry milk powder dissolved in 1´TBS (0.15 M NaCl, l25 mM tris-HCl, pH 7.4) and then incubated at 4°C with commercial monoclonal antibodies directed against target proteins. For all target proteins, b-actin (Biovision, Inc., Mountain View, CA) will be used as an internal loading control. Following the overnight incubation, membranes will be washed (6 ´ 10 min) in 1 - TBS with 0.1% Tween-20 and incubated at room temperature for 1 h with secondary anti-rabbit HRP-conjugated IgG (Promega Corporation, Madison, WI) or anti-mouse HRP-conjugated IgG (Promega). Blots will be visualized using the enhanced chemiluminescence detection system (Sigma).
Statistical Analysis
Data will be analyzed by SAS software (version 8.0 SAS Institute Inc., NC). The effects of plant sterol treatment on plasma lipid profiles, sterol trafficking, and gene/protein expression patterns, will be analyzed using a general linear model (GLM) procedure. Tests of normality will be conducted. If the distribution of responsiveness is not normally distributed, the data will be transformed logarithmically or with other appropriate transformations. Pearson correlation coefficient tests will be used to determine the relation between the degree of plasma lipid profiles and lipid kinetic parameters responsiveness to plant sterol. Genetic variants in the high and low responders to plant sterol administration will be compared by using Fisher's exact test. Plasma levels of LDL-C will be compared in carriers and non-carriers by using the Wilcoxon rank-sum test with age and gender as covariates.
Feasibility, Facilities, and Limitations
The RCFFN and BHNRC facilities have complementary equipment and scientific expertise to complete all aspects of this proposed research. RCFFN houses analytical laboratories, molecular biology laboratories, food safety laboratories, cell culture chambers and the Richardson Centre Clinical Research Unit (RCCRU). RCFFN possesses core molecular biology laboratories dedicated for the nutrigenomics research. Molecular biology laboratories at RCFFN are equipped with a wide range of molecular biology equipment including thermocyclers, electrophoretic systems, gel scanners, real time PCR systems, genetic analyzers, and a microarray scanning system, as well as genetics software including Arlequin. This nutrigenomics laboratory will provide us with the capability to conduct research, not only to identify new target genes of nutrients, but also to follow through with the functional analysis of these new genes regulated by nutrients. Carbon-13 and D tracer incorporation methodologies are routinely carried out within the Stable Isotope Laboratory (SIL) at the Richardson Centre which opened mid- 2006 with $10M in research analytical and bio-processing now fully commissioned. The Richardson Centre houses a state-of-the-art on-line Finnigan GC-combustion/pyrolysis-IRMS, as well as a GC-SIMMS for checking tracer enrichments. The presence of the new on-line IRMS system, together with lower level of invasiveness and the lower cost of isotopes, form the basis for the rationale to measure cholesterol kinetics in individuals in the feeding trial both after placebo and after plant sterol interventions. The throughput for D and 13C enriched samples using the GC-combustion/pyrolysis-IRMS is tenfold greater than the off-line system utilized previously. Technicians and trainees are familiar with clinical and laboratory procedures. Methodologically, all other procedures are highly feasible. Genetics technique consulting, as well as additional clinical capacity, is available through Dr. Peter Eck, who is also at the Richardson Center.
The Beltsville Human Nutrition Research Center contains a new state-of-the-art facility specifically designed for conducting dietary intervention studies. The facility, opened in 2003, includes a dining facility capable of providing tightly controlled diets (all food weighed for all meals and snacks) for up to 120 subjects each day. The facility has over 1800 sq ft of space for food storage, and a 2000 sq ft food preparation area. It is staffed by 6 full-time employees and up to 20 temporary food service workers. In addition, there is over 3,000 sq ft of space for collecting, handling, and storing biological specimens. This space includes phlebotomy rooms, sample processing rooms, examination rooms, and includes three room-sized calorimeters, one exercise calorimeter, three metabolic carts, a Hologic Lunar 4500A dual energy x-ray absorptiometer, a Life Measurement Inc., Bod Pod, a Tanita bioelectrical impedance analyzer, five Dynamap blood measure machines, two AtCor Medical SphygmoCor tonometers, and an Itamar Medical EndoPAT 2000. Specialized laboratory equipment includes a DSX automated plate reader for high-throughput ELISA analyses, a Leco CN analyzer, a Dade-Berhing clinical chemistry analyzer, a Beckman ACL 1000 coagulation analyzer, a Parr calorimeter, a Beckman Optima XL 100K ultracentrifuge, and an Europa isotope ratio mass spectrometer. The laboratory also has an extensive chromatography capability including a Shimadzu HPLC with UV/Vis and refractive index detector, a HP 5890 Series II with autosampler 7673 and HP 3365 ChemStation, one Agilent 6890 GC-MS, one Agilent quadropole LC MSD, one Agilent ion trap 1100 LC-MSD, one HP 1050 HPLC with PDA detection, one Agilent 1100 HPLC with PDA detection, and one Agilent 1100 HPLC with coularray detection.
The greatest limitation for this trial is the limited number of participants. n=120 is sufficient for seeing response and non-response in plasma lipids during plant sterol supplementation, however, it may not be sufficient to find any associations between plant sterol response and the single nucleotide polymorphisms which may or may not be present in the population.
Anticipated Significance
The results of the present investigations will lead to improvement in our understanding of the heterogeneity in response of lipids to plant sterol consumption. Given the wide variety of functional food products containing plant sterols presently available in and being developed for the Canadian marketplace, there is a critical knowledge gap as to why such products work in some individuals but not in others. The present research is anticipated to result in identification of genetic traits which will predict which individuals do and do not respond to plant sterols. It is likely that such traits may exist for more than a single gene studied, resulting in the potential to establish a genetic profiling of responders and non-responders. How these traits associate with differences in processes that control the flux of cholesterol, and lipoprotein particles that contain cholesterol, will also be elucidated from this study. These mechanistic explanations will be essential in creating the link between lipid level responsiveness and genetic control of processes that drive the degree of such responses. In terms of clinical applications, the present study will likely lead to the development of screening assays to assess the likelihood of efficacy of plant sterols as a cholesterol lowering agent within any given individual. The availability of such a nutrigenomic test will allow for more effective personalization of functional food use which will maximize health benefits for consumers.
