The cholesteryl ester transfer protein (CETP) is a major determinant of the high-density lipoprotein (HDL) variability in the general population. This protein facilitates the exchange of cholesteryl esters from HDLs to triglyceride (TG)–rich lipoproteins and low-density lipoproteins (LDLs) in exchange to TGs, decreasing the HDL plasma level.¹ In human, CETP is expressed predominantly in the liver, spleen and adipose tissue.² Detectable levels of CETP can be observed in the small intestine, adrenal glands, heart, kidneys and skeletal muscle.³ The human CETP gene is located on chromosome 16q21 and consists of 16 exons. Several single nucleotide polymorphisms (SNPs) in the CETP gene have been identified.The most widely studied CETP variant, denoted TaqIB (rs708272), is the silent base change affecting the 279th nucleotide in the first intron of the gene.⁴ This was the first genetic variation related to HDL-C plasma levels. The less common B2 allele (absence of the TaqI restriction site) has been associated with a lower CETP mass and with a higher HDL-C level compared to the more common B1 allele.⁵ Though this polymorphism has been shown to be consistently associated with HDL-C concentration, it does not directly influence CETP concentration or function. In Caucasian population the TaqIB polymorphism has been found to be in complete linkage with −629C/A (rs1800775), a polymorphism located in the promoter region of the CETP gene.⁶ This promoter polymorphism has been shown to influence CETP gene expression and this could account for the associations found between TaqIB polymorphism and plasma CETP mass and HDL-C concentration.⁶ The −629A allele has been associated with lower CETP mass and higher HDL-C than the -629C allele as demonstrated by 50% reduction in transcriptional activity in a reporter construct comprising the CETP promoter and a luciferase reporter.⁷ Since its first description, a consistent number of investigations has suggested an association between the TaqIB polymorphism, plasma HDL-C level and the risk of cardiovascular disease ^8,9 although controversy exists on this.¹⁰ SNPs in the CETP gene were also investigated for their association with atrial fibrillation (AF). Using the multifactor-dimensionality reduction method Asselbergs and coworkers showed, for the first time, that the B1B1 genotype was associated with a decreased HDL-C level and the development of AF in presence of albuminuria, elevated C-reactive protein, renal dysfunction and ischemic heart disease.¹¹ On the contrary, using a total of 2145 cases with AF and 4073 controls from Germany, Sinner and coworkers could not reliably replicate this association.¹² In both studies the promoter variant has not been investigated and the lipid profile was not considered.

The aim of the present study was to determine the prevalence of TaqIB polymorphism in a cohort coming from Salento, that is a Mediterranean country, with a racially homogeneous population of Caucasian origin and to investigate the association of the CETP /TaqIB polymorphism with plasma HDL-C level and the development of AF. We found that in our population the B2B2 genotype was associated with the risk of AF in females.

AF is the most common arrhythmia found in everyday clinical practice. The majority of patients with AF has underlying heart disease, such as valvular heart disease, hypertension, or left ventricular dysfunction. However, some patients develop AF in absence of any known risk factor. Family studies have revealed that gene mutations with a mendelian hereditary pattern underlie rare forms of AF. In addition, numerous reports have suggested associations between genetic polymorphisms and common forms of AF^13-16 but the identified variants were not often replicated in independent populations.

The CETP TaqIB polymorphism is very common in the population and seems to play an important role in cardiovascular disease.¹⁷ Although HDL cholesterol levels are higher about 10% in B2B2 genotype compared to B1B1, the association of this polymorphism with cardiovascular disease has not been established unequivocally. Some studies have shown a protective effect,^18,19 others have highlighted the association with adverse cardiovascular events²⁰ or have found no association.

Our study reported that the TaqIB2 allele and the concordant -629A allele of the CETP gene are associated with a higher incidence of AF in general population, particularly in females with increased TG levels (OR=5.14, 95% CI 1.57-16.82, p=0.0061). The role of the CETP TaqIB polymorphism on AF has not yet been unequivocally reported. Asselbergs and coworkers showed for the first time an association between CETP B1B1 genotype, decreased HDL-C level and AF.¹¹ On the contrary, a study done in 8141 subjects showed that the TaqIB polymorphism was not associated with an increased risk of AF disease.¹²

The strength of the present study lies in the selection of the participants based on their ethnic background, recruited carefully after evaluation of AF condition by patient’s history, serial electrocardiograms or ambulatory ECG monitoring, with AF onset occurring after 65 years of age in almost all the patients. It is important to note the difference in age between our population and those recruited by Asselbergs ¹¹ and Sinner.¹² More than 90% of our patients was over 65 years, while all the subjects enrolled in the other two studies were less than 65 years. We know that the prevalence and incidence of atrial fibrillation increase with age,²¹ so a potential misclassification of referent subjects that will later develop AF may bias the results of the two studies. Moreover, they didn’t stratify their population according to gender, whereas we observed the most significant data in the female subgroup. In addition, Sinner sample consisted predominantly of males (over 70%).

Although it has been recognized that inflammation may facilitate the development of AF,²² our cohort could represent a selected group of patients in whom AF is associated with an abnormal lipid metabolism rather than inflammation. In this case the TG/HDL-C ratio is a better indicator than other measurements as CRP and IL-6, and polymorphisms in genes involved in lipid metabolism, as CEPT, may have a role in the onset of the disease.

In our general population the frequency of B2 allele and B2B2 genotype were the same of other Caucasian populations²³ but the frequency of the B2B2 genotype (9%) in our female control subgroup was lower compared to those reported in different studies: in the Framingham Offspring study the B2B2 genotype frequency in females was 18.2%.²⁴ Park and coworkers in their Korean cohorts discovered that the B2B2 genotype frequency was 22.7%.²⁵

Moreover in our B2B2 females it was observed a modification of the HDL-C levels by TG. The B2B2 individuals tended to have higher HDL-C levels and significant lower TG levels and TG/HDL-C ratio compared to B1B1 and B1B2 subjects (Table 3). Conversely in the B2B2 AF females we found significant decreased HDL-C levels and a significant higher plasma TG level and TG/HDL-C ratio compared to B2B2 controls.

When we performed sequence analysis of the CETP promoter to identify the functional polymorphisms located 629 bp upstream to the ATG codon, we observed the presence of the -629AA genotype in all the 23 B2B2 females. Thus, the -629A allele of the promoter region and the B2 allele of the TaqIB polymorphism were concordant in our population and might account for the observed increased HDL-C level in control females and the decreased HDL-C level in AF female patients with a higher TG level.

Our idea is that in the presence of hypertriglyceridemia there is a modification of the relationship between CETP and HDL-C levels. Our findings are consistent with those of Tato²⁶ and Föger²⁷ who showed that among hypertriglyceridemic individuals HDL-C concentration was dependent by CETP levels, while the contribution of CETP was less in normotriglyceridemic subjects. So, while in normotriglyceridemia reverse cholesterol transport by HDL-cholesteryl-ester-selective-uptake pathway, a non-CETP pathway, is important, in the presence of hypertriglyceridemia the CETP pathway becomes more important. But lower circulating CETP levels in B2B2 subjects could reduce the generation of small pre-β-HDL particles that stimulate the cellular cholesterol efflux. Low CETP levels could therefore reduce the removal of cellular cholesterol. This protein also regulates the cholesterol traffic directly at cellular level, as macrophages present in atherosclerotic lesions produce CETP. Zheng²⁸ has shown that voltage-dependent ion channels and lipids in close proximity form functional units and no-phospholipidic molecules, as cholesterol, can change opening/closing state of such channels.

The increase in triglyceride levels after the menopausal transition could account for the data we obtained in the female subgroup. In fact a longitudinal study reported an increase of 16% in triglyceride values.²⁹ In NHANES the gap in triglyceride levels between men and women narrowed in the 50- to 59-year age group, and from 60 years onward women had higher levels than men.³⁰ This was confirmed in our population, in which emerged a statistically significant difference between men and women, with higher triglyceride levels in the female subgroup (1.37±0.59 vs 1.19±0.48, p=0.0212).

The low number of our sample may represent a limitation of the study we conducted. Despite this, our work was performed on patients from a well-defined geographical area and in these association studies the genetic background is particularly important. In addition, the selection mode plays a key role. In atrial fibrillation, whose prevalence and incidence increase with age, it may be important to select patients taking into account this parameter, as an age too low can lead to a potential misclassification of enrolled subjects, which may later develop arrhythmia, distorting the results of the study. However, to better define the associations observed in our work, further studies are required in larger populations, also belonging to other geographical areas.

The results of our study indicate significant association of TaqIB2 and the concordant -629A alleles of CETP gene with lower HDL-C levels, higher TG levels and TG/HDL-C ratio in a subset of postmenopausal females coming from Salento (Southern Italy).

This lipid profile, added to the negative effects caused by CETP lower plasmatic levels induced by polymorphism, such as pre-β-HDL particles reduction, reduced cellular cholesterol removal, cellular membrane fluidity modification, could have promoted the onset of the arrhythmia in our population

The financial support of the University of Salento (Progetti di Ricerca Scientifica d'Ateneo) is gratefully acknowledged.

None.