Influence of PON1 Polymorphisms on the Association between Serum Paraoxonase 1 and Homocysteinemia in a General Population

Homocysteine (Hcy)-induced vascular impairment may be partially mediated by the production of Hcy thiolactone (HTL). This compound acylates side-chain lysine groups in proteins and alters protein structure and function. HTL is formed under conditions of high Hcy resulting from insufficient remethylation of Hcy to methionine; however, HTL is not a reliable marker of plasma total Hcy (tHcy). In healthy volunteers, it contributes only 0.14%–0.28% of tHcy, has a half-life of 1 h, and is below the detection limit in approximately one half of volunteers (1). Paraoxonase 1 (PON1) is a hydrolase associated with HDL that is thought to degrade lipid peroxides and HTL (2). Decreased PON1 activity has been associated with atherosclerosis (3). Hepatic expression of the PON1 gene is down-regulated in hyperhomocysteinemic mice (4); it is plausible, therefore, that the proatherogenic effects of Hcy may involve diminished serum PON1 activity, leading to impaired antioxidant function and decreased capacity to degrade HTL. In support of this hypothesis, an inverse relationship between PON1 and Hcy in hospital patients was reported recently (5). However, this report did not provide any information regarding patient type or the influence of confounding factors. We investigated the association between tHcy and PON1 in a general population, correcting for the effects of the following confounding factors: HDL concentration, body mass index, sex, and PON1192, PON155, and PON1−107 genotypes. We studied a representative sample (200 women and 209 men; age range, 19–75 years) of the general population. Details of this population have been reported previously (6). Serum PON1 activity and concentration, HDL concentration, and PON1 polymorphisms were determined as described previously (1)(7)(8). We measured tHcy by fluorescence polarization immunoassay (Abbott Laboratories) and used log-transformations of tHcy and PON1 to approach normality for statistical tests requiring gaussian distributions. We determined the Spearman correlation coefficients for the bivariate associations between tHcy and PON1 and used the Student unpaired t-test to assess differences in mean PON1 activity and concentration in individuals in the highest tHcy tertile compared with the rest of the population. We used multiple linear regression analysis with adjustment for confounding factors to assess the relationships between these variables. Results are presented as the mean (geometric mean in the case of Hcy) and 95% confidence interval (95% CI). We observed the following means (95% CI) in our population: tHcy, 8.9 (8.7–9.2) μmol/L; HDL, 1.53 (1.49–1.57) mmol/L; PON1 activity, 411 (396–426) U/L; PON1 concentration, 96.5 (88.8–104.2) mg/L. Bivariate analysis showed no significant relationship between tHcy and PON1 activity (r = −0.089; P = 0.07) or between tHcy and PON1 concentration (r = 0.075; P = 0.15). The correlation was unchanged after exclusion of 12 participants with tHcy greater than the mean + 2 SD (r = −0.089; P = 0.08). However, persons in the highest tHcy tertile had significantly lower PON1 activity than the remaining participants [388 (362–414) vs 422 (404–441) U/L; P <0.05]. There was no significant difference between PON1 concentrations observed according to tHcy tertile [92 (80–103) vs 99 (89–109) mg/L]. The multiple linear regression models are summarized in Table 11 . tHcy had no significant or independent influence on either PON1 activity or concentration. Thus, although PON1 activity was lower in the highest tHcy tertile of our participants, the association between tHcy and PON1 activity was not significant when we adjusted for confounders, especially the PON192 genotype. We consider it essential, therefore, to adjust for PON1 genotypes when investigating the influences of other factors on PON1. Importantly, we studied a general population, whereas the previous report (5) was performed with hospital patients, in whom several diseases could have affected the results. Our population clearly had a substantially lower mean tHcy (8.9 μmol/L) than that of the hospital patients in the previous report (∼15.5 μmol/L) (5). Moreover, we measured the activity of PON1 with paraoxon as substrate, whereas the previous study (5) used phenylacetate as a substrate. From our study, which provides information regarding the relationship between tHcy concentrations and PON1 in a general population, we conclude that a negative effect of tHcy on PON1 activity might be seen only in situations with increased tHcy. Effect of tHcy and other confounding factors on PON1 activity and concentration.1 Multiple linear regression analysis. Models are adjusted for body mass index and sex. The table shows the percentage changes in PON1 activity and concentration for the indicated predictor. In the case of PON1 polymorphisms, the table shows the effect of being homozygotic. P <0.001. P <0.05. dNS, not significant. Effect of tHcy and other confounding factors on PON1 activity and concentration.1 Multiple linear regression analysis. Models are adjusted for body mass index and sex. The table shows the percentage changes in PON1 activity and concentration for the indicated predictor. In the case of PON1 polymorphisms, the table shows the effect of being homozygotic. P <0.001. P <0.05. dNS, not significant. This work was supported by Grants 02/0430 and 00/0954 from Fondo de Investigación Sanitaria and by the Red de Centros de Metabolismo y Nutrición (C03/08). Dr. Ferré is a researcher from the Juan de la Cierva program (Madrid, Spain).