Influence of pre-analytical variables on VEGF gene expression and circulating protein concentrations.

BACKGROUND: The extended role of vascular endothelial growth factor (VEGF) in human pathophysiology led us to evaluate pre-analytical parameters possibly influencing its levels in peripheral blood and tissues. The effects on VEGF protein levels and mRNA expression were measured after storage delay (blood and tissue), use of different types of anticoagulants (blood), and after different numbers of freeze-thaw cycles (blood). METHODS: Blood from healthy donors was sampled simultaneously in ethylene diamine tetraacetic acid (EDTA), acid citrate dextrose (ACD-A), hirudin, and serum separation tubes. For each anticoagulant, VEGF was measured by enzyme-linked immunosorbent assay (ELISA) with different conditions of delay at 4°C before centrifugation (2 h, 4 h, or 48 h) and of different numbers of freeze-thaw cycles (1, 2, and 10). The transcripts coding for the VEGF165 isoform were quantified in peripheral blood mononuclear cells by RT-PCR. Muscle biopsy samples were frozen with delays of 15, 30, or 60 min after surgery. VEGF expression was quantified on immunofluorescence stained slides. RESULTS: The period of storage and the number of freeze-thaw cycles correlated with an increase in the levels of circulating VEGF (for each anticoagulant but not for serum) and its expression in PBMCs. VEGF expression measured from muscle biopsy sections was higher with freezing delays, with a peak at 30 and 60 min as compared to 15 min. CONCLUSIONS: The most reliable conditions for measuring both circulating VEGF and its gene expression are to reduce time between blood collection and centrifugation, and to avoid multiple freeze-thaw cycles. Serum collection tubes with no additive and no separator were less sensitive to the pre-analytical variations analyzed in this study. Freezing delay had a significant influence on VEGF protein expression in tissue samples.

Identification of cis- and trans-acting genetic variants explaining up to half the variation in circulating vascular endothelial growth factor levels.

RATIONALE: Vascular endothelial growth factor (VEGF) affects angiogenesis, atherosclerosis, and cancer. Although the heritability of circulating VEGF levels is high, little is known about its genetic underpinnings. OBJECTIVE: Our aim was to identify genetic variants associated with circulating VEGF levels, using an unbiased genome-wide approach, and to explore their functional significance with gene expression and pathway analysis. METHODS AND RESULTS: We undertook a genome-wide association study of serum VEGF levels in 3527 participants of the Framingham Heart Study, with preplanned replication in 1727 participants from 2 independent samples, the STANISLAS Family Study and the Prospective Investigation of the Vasculature in Uppsala Seniors study. One hundred forty single nucleotide polymorphism (SNPs) reached genome-wide significance (P<5×10(-8)). We found evidence of replication for the most significant associations in both replication datasets. In a conditional genome-wide association study, 4 SNPs mapping to 3 chromosomal regions were independently associated with circulating VEGF levels: rs6921438 and rs4416670 (6p21.1, P=6.11×10(-506) and P=1.47×10(-12)), rs6993770 (8q23.1, P=2.50×10(-16)), and rs10738760 (9p24.2, P=1.96×10(-34)). A genetic score including these 4 SNPs explained 48% of the heritability of serum VEGF levels. Six of the SNPs that reached genome-wide significance in the genome-wide association study were significantly associated with VEGF messenger RNA levels in peripheral blood mononuclear cells. Ingenuity pathway analyses showed found plausible biological links between VEGF and 2 novel genes in these loci (ZFPM2 and VLDLR). CONCLUSIONS: Genetic variants explaining up to half the heritability of serum VEGF levels were identified. These new insights provide important clues to the pathways regulating circulating VEGF levels.

Biological and genetic factors associated with ABCB1 and pregnane-X-receptor expressions in peripheral blood mononuclear cells in the STANISLAS cohort.

BACKGROUND: ABCB1 is a membrane transporter ubiquitously expressed particularly in peripheral blood mononuclear cells (PBMCs). Resistance to drugs is associated with genetic variations of its gene and with modulation of its expression through the pregnane-X-receptor (PXR) transcription factor. We have previously shown that ABCB1 polymorphisms were associated with blood lipid concentrations. METHODS: We wanted to investigate the variation factors and the genetic determinants of ABCB1 and PXR expressions in PBMCs, and their interrelationships with plasma lipid levels. ABCB1 and PXR mRNA were quantified by real-time quantitative RT-PCR in PBMCs of 42 men and 39 women. RESULTS: ABCB1 and PXR were both expressed in PBMCs of all individuals, but their expressions were not significantly correlated. ABCB1 mRNA was correlated with body mass index (BMI; p=0.01) and age (p=0.03). In women, lymphocyte count also correlated with ABCB1 transcripts (p<0.01). After adjustment for BMI, correlation with age disappears. PXR mRNA expression depends on gender with men expressing higher PXR levels (p=0.01). PXR expression also correlates with γ-glutamyltransferase (GGT; p=0.02), but this disappeared after adjustment. CONCLUSIONS: Neither ABCB1 nor PXR expressions correlate with ABCB1 gene variants. Finally, association between ABCB1 or PXR expression in PBMCs and lipid or apolipoprotein plasma concentrations were not significant in this subset of healthy subjects. These results should be confirmed in a larger population sample and extended to patients with various cardiovascular risk profiles.

Expression of inflammatory molecules and associations with BMI in children.

BACKGROUND: Adipose tissue secrets several adipokines that have been proposed to be enrolled in many inflammatory pathways. Our aim was to investigate the adipokine expression in adipose tissue and peripheral blood mononuclear cells (PBMCs) in children. MATERIALS AND METHODS: Thirty-one (17 males and 14 females) healthy children aged 10.9 +/- 1.8 years with a body mass index (BMI) of 19.3 +/- 3.5 kg m(-2) were enrolled. Adipokines (TNF-alpha, IL-6 and leptin) gene expression was quantified by real-time quantitative PCR in adipose tissue and PBMCs from the same children. Their serum levels were also measured. RESULTS: BMI was positively correlated with leptin gene expression in adipose tissue and with leptin serum levels (beta = 0.476, P = 0.006 and beta = 0.576, P = 0.003 respectively). Leptin's serum levels were positively correlated with leptin gene expression in adipose tissue (beta = 0.462, P = 0.02). Adipose tissue gene expression of leptin and TNF-alpha and serum leptin and TNF-alpha serum levels were positively correlated (beta = 0.752, P < 0.001, beta = 0.311 and P = 0.015 respectively). In PBMCs, a positive correlation between TNF-alpha and IL-6 expression was found (beta = 0.526, P = 0.042). CONCLUSION: We demonstrated powerful correlations of adipokines gene expression in adipose tissue and PBMCs in children, underlying that these molecules share common pathways related to childhood obesity.

Adipokine expression in adipose tissue and in peripheral blood mononuclear cells in children Correlation with BMI and fatty acid content.

OBJECTIVE: Adipose tissue contributes in energy, lipid homeostasis and inflammation, through the adipokines it releases. Our aim was to study the associations between adipose tissue (AT) fatty acid content, adipokines' expression in AT and PBMCs and BMI in children. METHODS: Thirty-one (17 male) healthy children aged 10.9+/-1.8years and of BMI 19.3+/-3.5kg/m(2) were enrolled. Adipokines (TNF-alpha, IL-6, leptin and visfatin) expression was quantified by real-time quantitative PCR in AT and PBMCs. Serum levels were measured by ELISA and fatty acids (FA) from AT by gas chromatography. RESULTS: BMI was correlated with monounsaturated fatty acids (MUFAs) (beta=0.339, p=0.043), arachidonic (AA) (beta=0.576, p< or =0.001) and eicosapentaenoic acids (EPA) (beta=0.404, p=0.004) and negatively with stearic acid (beta=-0.577, p< or =0.001). TNF-alpha and visfatin expression from PBMCs were positively correlated with MUFAs (beta=0.271, p=0.027 and beta=0.214, p=0.020, respectively), n-9 fatty acids (beta=0.313, p=0.010 and beta=0.269 and p=0.024, respectively) and with docosahexaenoic acid (DHA) (beta=0.429, p=0.004 and beta=0.484, p< or =0.001, respectively), while negatively with the ratio n-6/n-3 (beta=-0.490, p=0.007 and beta=-0.374, p=0.044). CONCLUSIONS: A series of FA molecules were correlated with children's BMI and with TNF-alpha and visfatin expression from PBMCs indicating that AT fatty acid content, might have a role, as a potential regulator of PBMCs inflammatory gene expression.

Drug metabolizing enzymes and transporters mRNA in peripheral blood mononuclear cells of healthy subjects: biological variations and importance of pre-analytical steps.

Quantification in peripheral blood mononuclear cells of mRNA of drug metabolizing enzymes or drug targets could give interesting, new information in the field of pharmacogenomics and molecular mechanisms. However, for the interpretation of these data, it is necessary to know mRNA biological variations. In this review, we propose a strategy based on the production and interpretation of clinical chemistry reference values. We discuss the concept of reference values; the necessity to master pre-analytical variations of CYP and ABC transporters; the choice of the analytical methods and of the reference genes; and finally the biological variations themselves. In particular, we focus on the importance of considering homogeneity for age, sex, degree of adiposity, tobacco and alcohol intake, food habits, and drug consumption, including their inductive effects, at the phase of subject recruitment. All this information is useful to define the partition and exclusion factors to obtain mRNA reference limits.

Human formyl peptide receptor 1 (FPR1) c.32C>T SNP is associated with decreased soluble E-selectin levels.

AIMS: The human formyl peptide receptor (FPR) is a G protein-coupled chemoattractant receptor that is thought to mediate inflammatory responses. The FPR1 gene is highly polymorphic. In a recent study, the FPR1 c.32C>T SNP, resulting in the amino-acid substitution I11T, was reported to be significantly associated with C-reactive protein levels. Therefore, this study sought to determine if the impact of such a genetic variation extends to other clinical parameters associated with inflammation, including cytokines, adhesion molecules and inflammatory markers. MATERIALS & METHODS: This study was carried out on a subsample of 325 adults selected from the STANISLAS cohort study. The FPR1 c.32C>T SNP was genotyped using PCR amplification followed by restriction enzyme digestion. Anthropometric measurements and biochemical profiles were assessed for each individual. RESULTS: The allele frequencies of FPR1 c.32C>T were 0.74 for the 32C allele and 0.26 for the 32T allele. Genotype frequencies were 0.55 for C/C, 0.38 for C/T and 0.07 for T/T. After adjusting for age, sex, BMI, alcohol and cigarette consumption, oral contraceptive, antibiotics and anti-inflammatory drug use, statistical analysis (under a recessive model of inheritance) demonstrated that serum E-selectin levels were 68% lower in individuals homozygous for T/T than in those with C/T or C/C genotypes (p = 0.001). However, no significant correlations were found for C-reactive protein or the other 18 tested clinical parameters that were analyzed in this study. CONCLUSION: The FPR1 c.32C>T SNP may be associated with E-selectin levels in the French population. Although of importance, these findings need confirmation in larger studies.

Candidate gene microarray analysis in peripheral blood cells for studying hypertension/obesity.

AIMS: The gene expression of 182 cardiovascular candidate genes was measured in high quality groups of individuals (n = 20) by microarrays to determine whether a subset of genes would discriminate obese and hypertensive individuals, in spite of the existence of a close link between these two cardiovascular risk factors. MATERIALS & METHODS: The results were validated on the 20 subjects used for microarray analysis and on 62 additional individuals by real-time PCR. RESULTS: The first analysis, where patient groups were compared with healthy subjects, revealed 15 out of 182 genes that differed in hypertensive, obese or obesity-related hypertensive individuals. These genes were ALOX5, APOA2, SELL, RGS2, CD14, FPR1, CAMP, DEFA3, DEFA4, CBS, CHRM1, ICAM1, NR1H2, SCNN1B and TGFB1. A second analysis was carried out in which patient groups were compared with each other, demonstrating FPR1 and DEFA3 as being significant genes discriminating patient groups. Furthermore, an analysis stratified by sex revealed that, with the exception of DEFA3, there are no other common genes between men and women. DISCUSSION: We were able to indentify a number of interesting genes that distinguish patient and healthy subject groups as well as patient groups between them. CONCLUSION: In addition, it seems that gender plays an important role, at least for some of the genes we tested. These findings may have important implications in the screening and etiology of hypertension or obesity, and could further help to focus on these specific mRNAs as antisense therapy targets.

Leptin expression in Peripheral Blood Mononuclear Cells (PBMCs) is related with blood pressure variability.

BACKGROUND: Leptin is an adipokine initially considered as a molecule related exclusively to obesity but advances in research revealed its multiple roles in other physio-pathological mechanisms and particularly in the inflammatory ones. The aim of the present study was to demonstrate the presence of leptin in human Peripheral Blood Mononuclear Cells (PBMCs) and to quantify its mRNA in this type of tissue, closely related to inflammation. METHODS: Leptin mRNA was present in PBMCs of healthy individuals. Its expression was further studied in 83 individuals in relation to constitutional factors, anthropometric variables, blood pressure, lipid profile, glucose and markers of inflammation (C-reactive protein, lymphocyte count). RESULTS: Expression levels were significantly associated with systolic blood pressure (SBP) (p = 0.03) and diastolic blood pressure (DBP) (p = 0.003). Using a multiple regression analysis model, we showed that leptin mRNA levels explained 11% of the variation of SBP (p = 0.007) and of DBP (p = 0.003). These percentages remained at the same magnitude for SBP (9%) and for DBP (10%), after introducing BMI in the model. CONCLUSION: We report here for the first time, leptin expression in human PBMCs of healthy individuals. The associations found with blood pressure suggest a possible role of leptin in blood pressure regulation via PBMCs.

Visfatin, low-grade inflammation and body mass index (BMI).

OBJECTIVE: Visfatin is an adipokine with revealing roles in inflammatory mechanisms but its implication in inflammation related to excessive adiposity/obesity is not studied yet. Our aim was to investigate the relations of visfatin with inflammation markers and body mass index (BMI) in the peripheral blood mononuclear cells (PBMCs), a type of cells closely related to inflammatory mechanisms. DESIGN: Cross-sectional study, quantification of visfatin, TNF-alpha, IL-6 mRNA in PBMCs. PATIENTS: Eighty-three supposed healthy individuals from the STANISLAS cohort, belonging in three BMI categories: BMI < 25 kg/m(2) (lean), 25 kg/m(2) or= 30 kg/m(2) (obese). MEASUREMENTS: We measured visfatin gene expression (by real-time quantitative PCR), in relation to gene expression of the pro-inflammatory cytokines TNF-alpha, IL-6 in PBMCs and to anthropometric parameters (weight, BMI, waist : hip ratio), blood pressure, lipid profile, glucose and inflammatory markers (C-reactive protein, lymphocyte count). RESULTS: Visfatin expression in PBMCs was significantly associated with BMI in a negative way (r = -0.21, P = 0.05). Global anova analysis test for lean and over-weight/obese individuals showed a negative significant association between visfatin expression in PBMCs and BMI both for men and women (P = 0.05 and P = 0.01, respectively) and these associations remained significant after separating subjects in three groups (lean, overweight, obese) for men and women (P = 0.02 and P = 0.05, respectively). Correlation analysis between levels of expression of visfatin and TNF-alpha showed a significant positive linear association (r(2) = 0.27, P < 0.0001). CONCLUSION: These findings reveal a probable new role of visfatin in inflammation reflected in PBMCs, in the context of obesity.

Transcription factor and drug-metabolizing enzyme gene expression in lymphocytes from healthy human subjects.

We aimed to measure simultaneously the expression of drug-metabolizing enzymes (DME) and transcription factors (TF) with high importance in cardiovascular physiopathology in lymphocytes from healthy subjects. RNA was isolated from peripheral blood mononuclear cells (PBMC) of 20 subjects from the Stanislas Cohort. We used a microarray approach to measure 16 DME and 13 TF. Cytochromes P450 (P450s), including CYP2C19, CYP2C9, CYP2J2, CYP2D6, CYP1A1, CYP4F2, CYP4A11, CYP2E1, CYP11B2, CYP2C18, and CYP2A6, were expressed in all the subjects. CYP3A4 and CYP3A5 were not expressed. Glutathione S-transferases (GST) were expressed, but GSTM1 was seen only in some subjects. Pregnane X receptor (PXR), myocyte enhancer factor 2, vitamin D receptor, liver X receptor (LXR)-alpha, aryl hydrocarbon receptor (AHR), T-cell factor 7, constitutive androstane receptor, and aryl hydrocarbon receptor nuclear translocator (ARNT) were expressed in the majority of the subjects. Glucocorticoid receptor, peroxisome proliferator-activated receptor (PPAR)-gamma, and LXRbeta were expressed only in some individuals. PPARalpha mRNA was found in one subject only, and farnesoid X-activated receptor was not expressed. In addition, we found significant correlations between the expression of AHR, ARNT, and CYP1A1 and between PXR and P450 involved in leukotriene metabolism (CYP2C, CYP4F2, CYP4A11, CYP2J2, and CYP11B2). We describe here for the first time the presence of the majority of TF and DME in PBMC of healthy subjects without previous induction. The expression of these genes in lymphocytes could be a useful tool for further studying the physiological and pathological variations of DME and TF related to environment, to drug intake, and to cardiovascular metabolic cycles.

Peripheral blood mononuclear cells (PBMCs): a possible model for studying cardiovascular biology systems.

BACKGROUND: The inflammation system, alone or in relation to or interaction with other cardiovascular pathways, is suggested to be the central pathway in the development and progression of cardiovascular diseases. The aim of the present investigation was to propose a specific and informative model for exploring this hypothesis. METHODS: In a biological system approach, we studied the expression of 182 candidate cardiovascular genes in peripheral blood mononuclear cells (PBMCs), cells that provide specific information on the inflammatory pathway. We explored their expression in 20 individuals with or without risk factors (obesity, hypertension) for cardiovascular disease. RESULTS: We found that: 1) 166 among the 182 selected genes were expressed in at least one individual's PBMCs, some of them being detected for the first time in this tissue; 2) all pathways were represented by the majority of their genes selected; 3) genes were expressed at a level sufficient for further study of the inter-individual variations in their mRNA to determine their biological variation; and 4) 15 genes discriminated hypertensive from obese or controls. CONCLUSIONS: The results of the present investigation support our proposal of a promising novel strategy based on PBMC transcriptomic studies to elucidate the complexity of the cardiovascular system in relation to inflammation. Preliminary data support the usefulness of the PBMC model in hypertension/inflammation research.

Association between factor VII polymorphisms and blood pressure: the Stanislas Cohort.

The purpose this study was to determine whether Arg353Gln and -323Del/Ins polymorphisms of factor VII (FVII) are related to blood pressure levels and hypertension. Subjects were drawn from the Stanislas Cohort, a longitudinal, familial French cohort examined twice since 1994. The "blood pressure study" included 1342 subjects free of medication use that could affect blood pressure. The "hypertension study" included 645 normotensive and 77 hypertensive adult subjects. Association with hypertension was also studied in 547 hypertensives enrolled in a clinical trial and in 624 normotensives drawn from the Stanislas Cohort. In the "blood pressure study," parents with the 353Gln or -323Ins allele had lower blood pressures than did noncarriers at each examination, independent of covariates (0.01< or =P< or =0.05, except for diastolic blood pressure [DBP] at baseline, where P=0.103). Similarly significant relations were observed in their offspring (P< or =0.05, except for systolic blood pressure [SBP] at 5 years, where P=0.186). In a representative subgroup of 267 individuals, the -323Del/Ins polymorphism was significantly associated with plasma FVII levels in both parents and offspring (P<0.001). FVII levels in plasma were significantly correlated with SBP and DBP in parents but not in offspring. After inclusion of both FVII levels and the -223Del/Ins in the same model in parents, only FVII levels remained significantly associated with SBP and DBP. The "hypertension study" revealed that the 353Gln and -323Ins alleles were related to decreased risk (odds ratio [OR]=0.554, 95% confidence interval [CI], 0.362 to 0.848, and OR=0.475, 95% CI, 0.299 to 0.755, respectively). These results suggest that the FVII gene may be a susceptibility locus for hypertension.

The Leu554Phe polymorphism in the E-selectin gene is associated with blood pressure in overweight people.

BACKGROUND: Associations between circulating concentrations of E-selectin, blood pressure and obesity, and between the Leu554Phe (L/F554) polymorphism and blood pressure have been documented. OBJECTIVES: To investigate how the E-selectin L/F554 polymorphism is involved in longitudinal blood pressure changes, and how this polymorphism interacts with body mass index (BMI) on blood pressure. DESIGN AND PARTICIPANTS: For this study, 478 men and 546 women were selected from the Stanislas cohort, a French longitudinal study of volunteers for a free health check-up. These individuals underwent two examinations (t(0) and t(+5)) and were not taking medication that can affect blood pressure. RESULTS: At t(0), no relationship was observed between L/F554 polymorphism and blood pressure. However at t(+5), systolic blood pressure (SBP) was greater in individuals carrying the F allele, and the L/F554 polymorphism was associated with SBP in interaction with BMI (P < 0.001 in men and P < 0.05 for women). There was a steeper increase in SBP with BMI greater than 25 kg/m2 in carriers of the F allele than in LL homozygotes. Similar results were observed for diastolic blood pressure in men (P = 0.0103). CONCLUSION: These results suggest a BMI-specific effect of L/F554 polymorphism of the E-selectin gene on blood pressure, and strengthen the hypothesis that E-selectin is implicated in hypertension.

Association between Gly241Arg ICAM-1 gene polymorphism and serum sICAM-1 concentration in the Stanislas cohort.

Intracellular adhesion molecule-1 (ICAM-1), a cellular adhesion molecule that mediates the interaction of activated endothelial cells with leukocytes, is involved in various inflammatory and cardiovascular disorders. The relation between these markers and genetic polymorphism, however, remains to be elucidated. The aim of this study is to estimate the effect of a single-base polymorphism at codon 241 in exon 4 of ICAM-1 gene on serum sICAM-1 concentration in a healthy population, taking into account other biological determinants of sICAM-1 level. Serum sICAM-1 levels and the G/R241 polymorphism of the ICAM-1 gene were measured in a large healthy population consisting of 413 children aged 6-21 years and 363 adults aged 38-55 years extracted from the Stanislas cohort. The R241 allele was significantly associated with lower sICAM-1 levels and explained 3.4 and 1.9% of the sICAM-1 variability in children and adults, respectively. A codominant pattern contributed better to the model after adjustment for covariates as the RR homozygote effect was higher than that of the GR heterozygote. Moreover, significant independent associations were found between sICAM-1 and smoking, insulin resistance index (HOMA IR), interleukin-6 level, and alkaline phosphatase and aspartate aminotransferase activities. In conclusion, this study revealed a significant association between the G/R241 ICAM-1 polymorphism and serum sICAM-1 levels, probably due to the impairment in binding of ICAM-1 to leukocyte integrin Mac-1 protein.

Myeloperoxidase G-463A polymorphism and Alzheimer's disease in the ApoEurope study.

Alzheimer's disease (AD) is the most common cause of dementia in the elderly. Epidemiological and molecular genetic studies have shown the existence of several genes associated with increased risk of AD, the major genetic susceptibility locus coding for apolipoprotein E (apoE). A polymorphism in the myeloperoxidase gene (MPO) has previously been associated with AD susceptibility. However, results in the literature are controversial and seem to be dependent on several factors such as gender, apoE polymorphism or the genetic structure of the population. We investigated MPO G-463A and apoE polymorphism in 265 cases and 246 controls from the ApoEurope Study. In females, we found a significant association between MPO genotype and AD (P=0.034), GG genotype frequency being lower in cases (52.4%) as compared to controls (64.2%). In men, there was no significant effect of MPO polymorphism. No interaction was found between MPO polymorphism and apoE epsilon 4 allele. In conclusion, the G-463A polymorphism of MPO was statistically associated with AD in a gender-specific manner. However, given the low significance of P value we suggest no causal effect of the MPO gene in AD, as also evidenced in a recent meta-analysis. Our results support the hypothesis of a possible linkage disequilibrium between the MPO G-463A gene polymorphism and another functional variant involved in AD.