Genetic variation in the adiponectin receptor 2 (ADIPOR2) gene is associated with coronary artery disease and increased ADIPOR2 expression in peripheral monocytes

Our study analysis consisted of 68 patients from the Greek population with cardiovascular risk factors, who were screened for the existence of chronic stable CAD. All individuals underwent elective coronary angiography. Case subjects (n = 40) were patients who had angiographic evidence of stenosis > 50% in at least one major coronary artery (CAD). Control subjects (n = 28) were people without coronary stenosis at angiography (non-CAD). Subjects with acute myocardial infarction, systemic inflammatory diseases, malignancies, renal failure (creatinin > 1.5 mg/dl), heart failure and severe obesity with body mass index (BMI) > 35 were excluded from our study.

All patients gave their written informed consent and the study protocol was approved by the Scientific and Ethics Committee of Attikon University General Hospital. All patients were of a stable weight and had been on a normal isocaloric diet with normal physical activity during the previous four months. None of the patients were taking thiazolidinedione medication. Waist and hip circumferences were measured and the waist to hip ratio (WHR) was calculated. BMI was calculated as the ratio of weight (Kg) to height (m²). All patients were subjected to Intima-Media Thickness (IMT) assessment in common carotids and in carotid bulbs as an index of atherosclerosis, using B-mode ultrasound imaging (Vivid 7 General Electric Horten, Norway), as previously described [34] and Flow Mediated Dilatation (FMD) assessment of the brachial artery as an index of endothelial dysfunction [35].

The coronary angiography was followed by a 2-hour oral glucose tolerance test (OGTT) two weeks later. Insulin resistance status in the fasting state was calculated by the homeostasis model assessment index (HOMA) [36], whilst post-glucose insulin sensitivity was calculated by the Matsuda index [37]. Fasting blood was drawn at the beginning of the OGTT and plasma was kept at -80°C for measurement of adiponectin levels. Fasting triglycerides, as well as total and high-density lipoprotein (HDL) cholesterol were determined by an ILAB Analyser (Instrumentation Laboratory SpA, Italy). Plasma glucose levels at fasting and during the OGTT were analysed on a Falcor 300 chemical analyser (Menarini diagnostics, Italy) and their corresponding insulin levels were measured by IRMA (INSI-CTK, Dia Sorin, Italy).

Fasted blood was collected from all individuals and genomic DNA was extracted from the buffy coat according to the manufacturer's instructions (Flexigene DNA Kit, Qiagen, Chatsworth, CA, USA). The quality and the integrity of the DNA samples was determined by standard ethidium-bromide agarose gels and quantification was performed spectrophotometrically (UV/VIS Spectrometer Lambda Bio10, Perkin Elmer, USA). The following common single nucleotide polymorphisms were examined lying either in the promoter [rs10773980 C/T and rs1029629 C/T], the intron 5 [rs767870 A/G], the exon 6 [rs16928751 G/A], the exon 7 [rs9805042 C/T and I290I C/A] and the exon 8 [rs12342 C/T and rs1044771 C/T] of ADIPOR2 (Fig. 1).

The Allele Specific Primer Extension assay and xMAP technology were used to genotype our samples. This technique is a solution-based sequence-specific enzymatic reaction that determines the target SNPs. It involves the incorporation of a specific capture sequence, named Allele Specific Primer Extension (ASPE) that carries the major or the minor nucleotide of the SNP at its 3' -end [38, 39]. Therefore, two ASPE primers are needed for every SNP in order to resolve each possible allele pair. ASPE primers also carry a tag-sequence which hybridises to a complementary sequence called anti-tag, in the surface of the polystyrene microspheres (xMAP microspheres), which are internally dyed with fluorophores.

Various PCR products encompassing the SNPs of interest, ranging in size from 353 to 1180 bp, were generated through standard PCR reactions from human genomic DNA, using Platinum Taq Polymerase (Invitrogen, Carlsbad, CA, USA) in a GeneAmp 9600 thermocycler (Perkin Elmer Inc, Massachusetts, USA). An aliquot from the PCR reactions was run on 1.5% agarose gels and visualised with ethidium bromide staining. Product lengths were estimated using a standard molecular weight marker (ΦX174-HaeIII digest, NEB, UK) (Fig 2A and 2B) and subsequently the PCR products were treated with exo-SAP mix (2:1) (Exonuclease I:Shrimp Alkaline Phosphatase, USB Europe GmbH, Germany) at 37°C for 30 minutes in order to remove unconsumed dNTPs and primers remaining in the mixture, followed by inactivation at 80°C for 15 minutes. Consequently the treated PCR products were combined with ASPE primers and extension took place with the use of biotin-14-dCTP and Platinum Tsp DNA Polymerase (Invitrogen, Carlsbad, CA, USA). Extension occurs only if the 3' nucleotide of the ASPE primer is complementary to the template DNA. The conditions for the ASPE reaction were as follows: denaturation at 96°C for 2 minutes, amplification for 40 cycles at 95°C for 30 seconds, 54°C for 30 seconds and 74°C for 60 seconds, followed by one cycle at 4°C. The subsequent hybridisation of the anti-tag microsphere to the biotinilated ASPE-tag PCR product took place at 37°C for 60 minutes, after a short denaturation cycle at 95°C for 90 seconds.

Finally, a fluorescent-labelled reporter molecule (Streptavidin-R-phycoerythrin, 1 mg/ml, Molecular Probes, Invitrogen, Carlsbad, CA, USA) binds to the biotinilated hybrid ASPE microsphere product. This coupling of the reporter molecule quantifies the biomolecular interaction that occurs at the microsphere surface. The reaction can finally be analysed on a Luminex 200 instrument (Luminex Corp, TX, USA). Precision fluidics aligns the microspheres in single file and passes them through the lasers, one at a time. One laser excites the reporter molecule bound to the microsphere surface and the second laser excites the microsphere fluorophores. Reactions are measured with fluorescent intensity and reported in real time. The fluorescent signals generated for each microsphere bead population were used to assess the genotype in each sample. The median fluorescent intensities (MFI) were used to calculate allelic rations for each allele as follows:

To be homozygous for a particular allele, the allelic ratio must be at least 0.75. For heterozygous samples each allele must have a ratio between 0.25 and 0.75. An allele with a ratio of 0.25 or less is considered negative i.e. not present. Representative samples results are shown in figure 3A.

Peripheral monocytes were isolated from mononuclear cells using a Magnetic Cell-Sorting technique and the CD14 Human MicroBeads (Miltenyi Biotec, Gladbach, Germany). The purity of the magnetically isolated CD14⁺ monocytes was determined after staining the cells with a CD14 PE-CD45 FITC antibody (BD Biosciences, San Jose, CA, USA).

Total RNA was extracted from CD14⁺ monocytes using the Tripure Isolation Reagent (Roche Diagnostics GmbH, Mannheim, Germany). Reverse transcription of 1 μg of total RNA was performed in all samples with the Transcriptor First Strand cDNA Synthesis Kit using random hexamer primers, according to the manufacturer's instructions (Roche Diagnostics GmbH, Mannheim, Germany).

A quantitative real-time PCR using fluorescent-labelled hybridisation probes was developed for detecting relative ADIPOR2 mRNA levels, in human CD14⁺ monocytes, using the calibrator-normalised standard curve and the LightCycler Relative Quantification 1.0.1 Software, in a spectrofluorometric thermal cycler (LightCycler, ROCHE, Manheim, Germany), as previously described [40]. The human housekeeping gene of β-actin was used as standard for normalisation. Hybridisation specific primers and fluorescent-labelled hybridisation specific probes for the genes were designed and manufactured by the TIM-MOLBIOL (Berlin, Germany). Hybridisation primers and probes for the target and reference genes were as follows:

Primers were chosen to lie between exons whenever possible and the analysis was performed with the programme OLIGO 6.0 (Molecular Biology Insights, Cascade, USA).

For quantifying the relative expression levels of ADIPOR2 mRNA in the various samples, we ran the target (AdipoR2) and the reference (β-actin) gene of each sample along with the calibrator cDNA, using the Second Derivative Maximum Method with the Arithmetic baseline adjustment for the determination of the various crossing points (Fig 3B). Samples were run in duplicates or triplicates and results are expressed in arbitrary units.

Surface adiponectin receptor ADIPOR2 was determined after staining the isolated cells ('buffy'coat) with a suitable antibody (Alpha Diagnostic International, San Antonio, USA). Since the antibody was not fluorochrome conjugated, it was labelled with the Zenon™ Alexa Fluor^® 488 Rabbit IgG labelling kit (Invitrogen, Carlsbad, CA, USA). Cells were incubated for 30 minutes with Alexa Fluor^® 488-conjugated immunoglobulin in a ratio 1 × 10⁶cells μg^-1 of immunoglobulin under mild constant shaking. The monocyte fraction was simultaneously stained with anti-CD14-PE monoclonal antibody (BD Biosciences, San Jose, CA, USA). The specificity of the antibody used was evaluated by staining cells with isotype control suitable for each antisera and the blockage of Fc-receptors prior to staining.

Two-colour flow cytometric analysis was performed on a BD FACSCalibur 4 colour flow cytometer (BD Biosciences) equipped with two air-cooled lasers: a 15 mW Argon laser (488 nm) and a Red Diode laser (632 nm). Data acquisition and analysis were performed using the BD CELL Quest Pro software (BD Biosciences). Results are expressed in mean fluorescence intensity arbitrary units (MFI AU).

Circulating levels of adiponectin were measured in the plasma of all subjects, using a high sensitivity multiplex assay (xMAP technology) and fluorescent-labelled microsphere beads (HCVD1-67AK Lincoplex kit, Millipore Corp., MA, USA), in a LUMINEX 200 instrument (Luminex Corp, TX, USA). Sensitivity of the assay was 56.0 pg/ml, intra-assay and inter-assay coefficients of variation were 9.2% and 15.9%, respectively. All samples were diluted 1:100 and measured in duplicate. Samples that could not be detected were measured again using a higher dilution.

Statistical analysis was performed using the SPSS version 14.0 software (Chicago IL, USA). Deviation of SNPs from the Hardy-Weinberg equilibrium was performed using the chi-square test and the Hardy-Weinberg equilibrium calculator [41]. If genotype frequencies for a particular SNP differ from those expected under equilibrium, the Hardy-Weinberg principle is being violated and is therefore not safe to try to associate this SNP with coronary artery disease. Distribution of alleles was compared using the χ² and the level of significance adopted was p < 0.05. The whole population odds ratio associated with genotypes was calculated by binary logistic regression analysis. Normal distribution was tested with the Kolmogorov-Smirnov test and logistic transformations of variances to achieve normal distribution were used when needed. The differences in various variables between genotypes were evaluated with the univariate ANOVA, adjusted for age, sex, BMI and HOMA. Kruskal-Wallis test was also used when homogeneity of variances were not met. Non-parametric Mann-Whitney and 2-independent samples t-tests were used accordingly for analysis. Data are expressed as means ± SEM.

The clinical and metabolic characteristics of the population studied (CAD and non-CAD) are shown in table 1. There were no significant differences between the two groups regarding age, weight, BMI, fasting and two-hour glucose and insulin levels, insulin resistance indices (HOMA and Matsuda), lipid levels (total cholesterol, LDL, HDL), systolic and diastolic blood pressure, endothelial dysfunction (FMD) and the circulating adiponectin (Table 1). However, CAD patients had higher WHR and IMT values compared to control subjects (Table 1). Furthermore, CAD subjects had higher ADIPOR2 mRNA levels, while they did not differ in ADIPOR2 protein expression, from peripheral CD14⁺monocytes, compared to controls.

The allele frequencies of the investigated polymorphisms of the ADIPOR2 are shown in table 2. Only rs767870 and rs1044771 polymorphisms were found to be in Hardy-Weinberg equilibrium, (p < 0.05, χ² < 3.83). Thus we excluded from further analysis, any polymorphism that deviated from Hardy-Weinberg equilibrium.

We subsequently investigated the distribution of genotypes for the rs767870 and rs1044771 polymorphisms between the two populations and their possible association with coronary artery disease (Table 3). There was a significant difference in the distribution of alleles of rs767870 polymorphism of ADIPOR2 between CAD and non-CAD individuals (p = 0.017).

No significant difference was found in the distribution of alleles for the rs1044771 polymorphism of ADIPOR2 between the two groups (Table 3).

Using binary logistic regression analysis, the risk for CAD for AG heterozygotes of the rs767870 polymorphism was 50% higher compared to AA homozygotes of the same polymorphism, although not to a significant degree (OR = 1.543 and 95% CI: 0.465-5.120) (Table 3). Furthermore, the marker rs767870 of ADIPOR2 was most significantly associated with waist circumference, IMT and FMD measurements in all individuals (Table 4). To be more specific, AA homozygotes and AG heterozygotes of the rs767870 had significantly higher IMTbulb and lower FMD values compared to GG homozygotes (1.62 ± 0.21 mm and 5.09 ± 0.76% vs 1.26 ± 0.07 mm and 7.62 ± 0.74%, p = 0.025 and p = 0.002, adjusted for age, sex, BMI, WHR and HOMA, respectively) (Table 4 and Fig.4A and 4B).

We then investigated the possibility that the association of rs767870 polymorphism with markers of endothelial dysfunction and atherosclerosis may be mediated through an effect on gene expression. In peripheral CD14⁺ monocytes, carriers of the major A allele (homozygotes and heterozygotes) of rs767870 polymorphism had higher levels of ADIPOR2 protein expression compared to homozygotes of the minor allele (69.17 ± 4.80 MFI and 92.03 ± 14.41 MFI vs 45.20 ± 7.70 MFI, p = 0.008, after adjustment for age, sex, WHR and HOMA) (Table 4 and Fig. 4C).