In their recent publication, Trabetti et al.1reported new data on the association of the OLR1 gene and acute myocardial infarction (AMI) or coronary artery disease (CAD), showing that they do not fully confirm the positive results of previous studies,2, 3, 4 even though they observed an association of one OLR1 polymorphism with CAD severity. They also comment that these discrepancies could be owing to an ascertainment difference, or to a population difference, or to the sample size reported in some of the published studies. In 2003,3 we identified and described a set of OLR1 single-nucleotide polymorphisms (SNPs) in complete linkage disequilibrium that were found to be associated with AMI in a series of Italian patients and controls carefully selected. All subjects enrolled in the study underwent coronary angiography and left ventriculography. Control subjects were without any angiographically demonstrable coronary lesions and no evidence of active myocardial ischemia. On this basis, ORL1 SNPs were considered an important risk factor in AMI.3

Difficulty in confirmation of the genetic association data has become a major impediment to progress in elucidating the basis of complex genetic disorders. This may relate to false-positive reports of association and uncertainty that originates from the unknown complexity of the overall genetic heterogeneity in the sampled populations. In addition, when genetic effects are modest in magnitude across the sample population with too many variables, such as race and ethnicity, replication may require very large sample size and appropriate selection of the control group.5 Accuracy in the definition of the phenotype of patients and the clinical (and even the subclinical status) of controls is, therefore, crucial in these studies. Sometimes risk factors may be significantly different between the groups of controls and patients, and this can affect the association analysis.

In a recent study, we demonstrated that ORL1 SNPs previously identified have a functional effect.6 In fact, two different OLR1 transcripts were isolated in the mRNA extracted from human monocyte-derived macrophages. One of these products corresponded to the full-length transcript, whereas the other lacked exon 5, and we termed it LOXIN. LOXIN was predicted to code a protein that lacks two-third of the lectin-like domain of LOX-1, a region that is important for oxLDL binding. Both isoforms were detected in several cell types, such as endothelial cells, fibroblasts, and smooth muscle cells, and tissues, including the heart, kidney, and brain.6 We were able to demonstrate that the OLR1/LOXIN mRNA ratio was 33% higher in human monocyte-derived macrophages of subjects homozygous for risk allele compared with homozygous for the nonrisk allele.6 This suggests that the SNPs may result in reduced expression of LOXIN. Correspondingly, a relative increase in the amount of LOXIN was found in cells derived from subjects homozygous for the nonrisk allele, suggesting that the expression level of LOXIN influence the incidence of AMI in humans and, at least in part, explain the risk resulting from polymorphisms in the human OLR1 gene.

Altogether, these data provide a biological plausibility of the OLR1 SNPs association studies and offer support for an important role for OLR1 in CAD and the evolution of AMI in certain patients.7 However, vagaries in patient presentation, background risk factors, and extent of atherosclerosis may all relate to the variable penetrance of a large set of alleles in different genes.