| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Clinical Diabetes Unit, Division of Endocrinology and Diabetes, University Hospital (S.D., I.L., M.-C.B.M., R.W.J.), 1211 Geneva, Switzerland; and INSERM, U-525 (V.N., L.T.), 75005 Paris, France
Address all correspondence and requests for reprints to: Dr. Richard W. James, Clinical Diabetes Unit, Division of Endocrinology and Diabetes, University Hospital, 24 rue Micheli du Crest, 1211 Geneva 14, Switzerland. E-mail: . richard.james{at}hcuge.ch
Abstract
Polymorphisms of the gene for the antioxidant enzyme, paraoxonase-1 (PON1), have been identified as risk factors for coronary disease (CHD), notably in diabetic patients. The polymorphisms have also been linked with other diabetic complications. The present study analyzed glucose metabolism as a function of PON1 polymorphisms in young healthy nondiabetic men from families with premature CHD and matched controls. The L55M PON1 polymorphism was independently associated with the glucose response to an oral glucose tolerance test. LL homozygotes had significantly impaired glucose disposal (P = 0.0007) compared with (LM+MM) genotypes. It was particularly marked for subjects from high CHD risk families and differentiated them from matched controls (P = 0.049). The area under the glucose curve (P = 0.0036) and the time to peak glucose value (P = 0.026) were significantly higher in the LL carriers, whereas the insulin response was slower (P = 0.013). Insulin resistance did not differ between L55M genotypes. There was a trend for reduced pancreatic ß-cell function as measured by glucose-induced insulin secretion (LL vs. LM vs. MM, 20.26 vs. 23.74 vs. 25.60; P = 0.077). The frequency of the L55 allele decreased significantly (P = 0.028) across regions defining a north-south European axis. No significant differences for the glucose response or case-control populations were observed as a function of the PON1 Q192R polymorphism. The study demonstrates an association between PON1 gene polymorphisms and glucose metabolism. The L55M-glucose interaction differentiated offspring of high CHD risk families, suggesting that it may be of particular relevance for vascular disease and possibly other diabetic complications.
PARAOXONASE-1 (PON1) IS a serum enzyme that protects low density lipoproteins (LDL) from oxidative stress. Several polymorphisms have been identified in the coding (1, 2) and promoter (3) regions of the human PON1 gene and shown to influence serum activities and concentrations of the enzyme. In nondiabetic patients both coding region Q192R (or A192B) and L55M polymorphisms as well as promoter polymorphisms have emerged as independent risk factors for coronary disease (CHD) (4, 5, 6, 7). This is not a consistent observation, however, and other studies, focusing in particular on the Q192R site, have not identified the polymorphism as an independent risk factor (6, 8, 9, 10, 11). The polymorphisms have emerged more consistently as independent risk factors for CHD in diabetic patients (12, 13, 14, 15, 16, 17). In addition, paraoxonase has been linked to other diabetic complications, including retinopathy, nephropathy, and neuropathy (18, 19, 20, 21), although these studies are preliminary in nature (22) and require independent confirmation in larger populations.
The pathophysiological mechanism underlying the links between PON1 polymorphisms and risk of CHD are presumed to involve oxidized LDL, whose role in atherosclerotic disease has been extensively documented (23). Data from PON1 knockout mice support this extrapolation, as the absence of serum PON1 is associated with increased lipoprotein oxidation and greater lesion development (24, 25). A role for oxidative stress has also been evoked for the other diabetic complications, and in this context, diabetic patients are reportedly exposed to a greater degree of oxidative stress (26). However, the apparently more extensive associations between PON1 polymorphisms and diabetic complications led us to consider potential links between the enzyme and pathological processes characteristic of the diabetic state. A systematic evaluation of PON1 polymorphisms and metabolic abnormalities specific to diabetes has not been undertaken. The European Atherosclerosis Risk Study (EARS) (27), which focused on healthy, young adults with and without a family history of premature CHD, provided the opportunity for such an evaluation with the additional possibility of analyzing its relevance to an increased risk of CHD. We report an independent association between the PON1 gene polymorphism L55M and a modified oral glucose tolerance test (OGTT). The interaction between the PON1 L55M polymorphism and OGTT was able to differentiate healthy young adults from families with a history of premature CHD, who manifested a significantly greater response to the OGTT than those without a family history of premature CHD.
Subjects and Methods
Populations
The EARS II study and population have been described previously in detail (28, 29). The study design was based on EARS I (30), but included only male subjects, and all underwent an OGTT (29). Briefly, university students (18–28 yr old) were recruited from 14 centers representing 11 European countries. The index (case) group (n = 380) was made up of sons of men who had a documented myocardial infarction before the age of 55 yr. The control group (n = 395) was age-matched (selected from student registers for age closest to the index case) on a 1:1 basis with cases and was without a family history of premature vascular disease. Data were incomplete for 47 participants who were not included in the present analysis. Participants were allocated to 4 groups according to rates of ischemic heart disease mortality and geographical location: A, Baltic (Finland, Estonia); B, United Kingdom (England, Scotland, Northern Ireland); C, Middle Europe (Denmark, Germany, Belgium, Switzerland); and D, Southern Europe (Portugal, Italy, Spain, Greece). Anthropometric, lifestyle, and clinical data were previously reported (28, 29). Other data are given in Table 1
.
|
Blood sampling and handling as well as analytical methods for blood lipids, lipoproteins, and apolipoproteins were described previously (28, 29). An OGTT was performed for all participants (n = 775) using 75 g glucose, and glucose and insulin levels were measured every 30 min over 2 h (29).
Insulin and glucose levels, measured fasting and during the OGTT, were used to estimate insulin resistance by the homeostasis model assessment method (31) and pancreatic ß-cell function as the insulin secretory response defined by the 30-min increment in insulin relative to glucose (
insulin (30 min - 0 min)/ glucose (30 min - 0 min)) (32).
Genotyping
Genomic DNA was isolated from white blood cells by the salting-out procedure (33). The coding region PON1 polymorphisms at positions 192 (Q192R) and 55 (L55M) were analyzed by restriction isotyping as previously described (34). To minimize possible bias during genotyping, cases, controls, and regions were mixed during the analyses.
Statistical analyses
Data were analyzed using SAS statistical software (SAS Institute, Inc., Cary, NC). Departure of genotype distributions from Hardy-Weinberg equilibrium was tested in each region and in cases and controls by a 
2 test with 1 df. Allele frequency was estimated by gene counting, and comparison of allele frequencies between cases and controls, adjusted for region, and between regions, adjusted for case-control status, was performed using a Mantel-Haenszel statistic with 1 df. The association of genotype (2 df) with fasting and postprandial biochemical variables was tested by one-way ANOVA adjusted for age, center, and case/control status. The difference between PON55 (LL) vs. (LM+MM) carriers (1 df) was also tested. The homogeneity of genotype effects between cases and controls was systematically tested by introducing the interaction term in the model. The postprandial responses were characterized by 1) two-way ANOVA for repeated measures to test for the overall significance of postprandial measurements over time and between genotypes; 2) the area under the curve, calculated by the trapezoidal rule; 3) the peak, calculated as the highest value minus the fasting value; and 4) the time at peak (time at which the highest value was observed). Differences in the time at peak between genotypes were tested by the nonparametric Wilcoxon rank-sum test. For positively skewed distributions, a power transformation (ln or square root) was applied for tests, but untransformed values are given in the tables. P < 0.05 was considered significant.
Results
Frequencies of polymorphisms
There were slight deviations from Hardy-Weinberg equilibrium in controls from the Baltic region and United Kingdom cases for L55M, with an excess of observed heterozygotes (P < 0.05). As the effect was observed in two subgroups only, cases and controls that were not from the same region, and precautions were taken to minimize bias during genotype analysis, it suggests a probable sampling effect. There was a strong linkage disequilibrium between the L55M and Q192R polymorphisms (D' = -0.98; P < 0.001). Allele frequencies of the two polymorphisms are illustrated in Fig. 1. There were no significant differences in allele frequencies between cases and controls for either polymorphism (L55M, P = 0.62; Q192R, P = 0.44). However, there were significant differences in allele frequencies between regions (P = 0.028) for the L55M polymorphism. The frequency of the L allele decreased when passing from northern to southern Europe (Fig. 1).
|
These are shown (Table 1
) as a function of PON1 genotypes in the combined case/control groups (there was no significant heterogeneity between cases and controls). For the L55M polymorphism, there were trends for differences in concentrations of total, LDL, and high density lipoprotein cholesterol, as well as apolipoprotein B, but these did not reach statistical significance. They tended to reflect a less favorable profile in MM homozygotes. No trends were observed for PON/R192Q (data not shown).
OGTT
There were no significant differences in the OGTT responses when cases and control groups were compared globally regardless of PON1 genotypes. The results of the OGTT analyzed as a function of the L55M polymorphism are shown in Fig. 2A for the combined populations. There were no differences between genotypes for fasting glucose levels. However, in repeated time analysis differences between genotypes were highly significant, reflecting less efficient glucose disposal in LL homozygote carriers [LL vs. (LM+MM): time x genotype, P = 0.0007; differences at 60 min, P = 0.0037; differences at 90 min, P = 0.0007; differences at 120 min, P = 0.025]. This effect was seen in all regions (test of heterogeneity, P = 0.70). The OGTT responses for the case and control patients separately are illustrated in Fig. 2, B and C. These demonstrate heterogeneity in the response between cases and controls (repeated time analysis: time x status x genotype, P = 0.049) where the impaired response of LL homozygotes is exaggerated in the case subjects.
|
for the combined populations. There were no significant differences between genotypes and no heterogeneity of response curves between cases and controls.
Other features of the OGTT response are given in Table 2. The area under the glucose curve was significantly different between genotypes notably when comparing LL with other carriers (P = 0.0036). Further adjustment for fasting values did not change the result. The time to glucose peak was also significantly different between LL homozygotes and other carriers (P = 0.026; Table 2). Corresponding values for the Q192R polymorphism showed no consistent differences between genotypes (results not given).
|
. The insulin response (time to insulin peak) was slower in LL homozygotes (P = 0.013). For the Q192R site, there were no significant genotype differences for the various insulin measures. Insulin resistance and secretion
Given the indications of an impaired response to the OGTT, a final analysis examined insulin resistance, as measured by the HOMA method and the insulin secretory response as an indication of pancreatic ß-cell function (Table 3). For the L55M polymorphism, there was a trend to reduced insulin secretion for LL homozygotes [LL vs. (LM+MM), P = 0.077]. No differences between L55M genotypes were evident for HOMA insulin resistance (Table 3). There were no associations between insulin resistance or secretion and Q192R genotype (results not given).
|
The present study has demonstrated an association between coding region PON1 gene polymorphisms and modified glucose metabolism. Carriers homozygous for the L allele of the L55M polymorphism showed delayed glucose disposal compared with other genotypes. The effect was essentially limited to subjects from families with a history of premature coronary disease and differentiated such subjects from matched controls who were without a family history of premature CHD. The study thus suggests links between PON1 gene polymorphisms, impaired glucose metabolism, and risk of vascular disease.
The delayed disposal of glucose associated with the L55M polymorphism would appear to implicate ß-cell function rather than insulin resistance. There was a trend to lower insulin secretion as a function of the L55M genotypes, albeit of borderline significance (P = 0.07). ß-cell function was analyzed as the 30 min increment in plasma insulin relative to glucose. This measure has been shown to predict the development of type II diabetes in independent studies (35, 36). In contrast, no significant differences or trends in insulin sensitivity between L55M genotypes were observed using the HOMA model assessment (31). The latter has been validated, notably against the glucose clamp procedure, as a reliable measure of insulin resistance (37).
It is remarkable that the effect of the PON1 polymorphism on the response to a glucose load could differentiate high risk families in such a young, apparently healthy population. In this context, it should be noted that the Q192R polymorphism, with which the 55 site is in strong linkage disequilibrium, was not associated with glucose metabolism. Given that offspring studies have a decreased power to detect associations of polymorphisms with disease, it implies that the L55M-glucose metabolism link may be of particular relevance to a predisposition to vascular disease. Conversely, a potential advantage of such a young population is that gene-disease interactions may be detectable before they are masked by confounding factors later in life. Another intriguing observation is the north-south fall in the frequency of the L allele, a decrease that parallels the decrease in coronary disease incidence (38). A similar observation was made for apolipoprotein E polymorphisms in the EARS I population (39) with the frequency of the higher risk apolipoprotein E4 allele also decreasing significantly on a north-south axis. It suggests that the L55M polymorphism may be one of the genetic factors contributing to the north-south CHD gradient in Europe.
The association between the L55 allele and modified OGTT was evident in cases only. One potential explanation is that the modified OGTT is due to other polymorphisms with which the L55M mutation is in linkage disequilibrium. The region of human chromosome 7 containing the PON gene family is a susceptibility locus for diabetes in Pima Indians (40), whereas genes implicated in glucose metabolism and susceptibility to diabetic complications map to chromosome 7 (41, 42, 43, 44). This possibility appears unlikely, however, in the absence of a modified OGTT in LL carriers of the control group. The data are more consistent with an interaction between the L55 allele and other genetic or environmental factors associated with the case group, an interaction that influences glucose metabolism. Given the proposed function of PON1, it raises the question of a potential role for oxidative stress. There are data in the literature indicating an effect of oxidative stress on pancreatic ß-cells where antioxidant treatment beneficially influenced function (45). Lipid oxidation products have also been implicated, as they were shown to inhibit glucose-induced insulin secretion from rat pancreatic islets (46). Other studies have reported that glucose disposal may be impaired by oxidative stress (47, 48). These observations are consistent with a potential role for the antioxidant function of PON1. In this context, it has been suggested that the L isoform of PON1 protects less well from oxidative stress than the M isoform (49). There are two interesting extrapolations of the considerations outlined above. Pancreatic ß-cell dysfunction predicts type 2 diabetes, which is a powerful risk factor for coronary disease. The PON1 L55M polymorphism may provide one link between diabetes and heart disease. Secondly, PON1 polymorphisms have been somewhat inconsistently linked to the risk of coronary disease in population studies. One interpretation of our data are that interaction with other factors, genetic or environmental, may be necessary for PON1 polymorphisms to manifest fully their impact on coronary risk. Although speculative at present, these extrapolations can serve as hypotheses for future studies.
In conclusion, the present study establishes an association between the PON1 polymorphism L55M and modified glucose metabolism. An impaired response to an OGTT was noted for LL homozygotes in offspring from high risk families. It served to differentiate them from offspring of control families. It suggests that the L55M-glucose metabolism link may be of particular relevance to increased risk of vascular disease, and possibly other diabetic complications, associated with the PON1 gene.
Acknowledgments
EARS II Project Leader: D. St. J. O’Reilly (United Kingdom); EARS II Project Management Group: F. Cambien (France), G. De Backer (Belgium), D. St. J. O’Reilly (United Kingdom), M. Rosseneu (Belgium), J. Shepherd (United Kingdom), and L. Tiret (France); The EARS II Group Collaborating Centers and their associated investigators: Austria: H. J. Menzel, Institute for Medical Biology and Genetics, University of Innsbruck, laboratory; Belgium: G. De Backer and S. De Henauw, Department of Public Health, University of Ghent, recruitment center; Belgium: M. Rosseneu, Laboratorium voor Lipoproteïne Chemie/Vakgroep Biochemie, University of Ghent, laboratory; Denmark: O. Faergeman and C. Gerdes, Medical Department I, Aarhus Amtssygehus, Aarhus, recruitment center; Estonia: M. Saava and K. Aasvee, Department of Nutrition and Metabolism, Estonian Institute of Cardiology, Tallinn, recruitment center; Finland: C. Ehnholm*, R. Elovainio**, and J. Peräsalo, *National Public Health Institute and **The Finnish Student’s Health Service, Helsinki, recruitment center; Finland: Y. A. Kesäniemi*, M. J. Savolainen*, and P. Palomaa**, *Department of Internal Medicine and Biocenter, Oulu, and **The Finnish Student’s Health Service, University of Oulu, recruitment center; France: L. Tiret, V. Nicaud, and O. Poirier, INSERM, U-525, Paris, EARS data center, laboratory; France: S. Visvikis, Center de Médecine Préventive, Nancy, laboratory; France: J. C. Fruchart, J. Dallongeville, Service de Recherche sur les Lipoprotéines et l’Athérosclérose (SERLIA), INSERM, U-325, Institut Pasteur, Lille, laboratory; Germany: U. Beisiegel, C. Dingler, Medizinische Klinik Universitäts-Krankenhaus Eppendorf, Hamburg, recruitment center and laboratory; Greece: G. Tsitouris, N. Papageorgakis, and G. Kolovou, Department of Cardiology, Evangelismos Hospital, Athens, recruitment center; Italy: E. Farinaro, Department of Medical Preventive Sciences, University Frederico II of Naples, recruitment center; The Netherlands: L. M. Havekes, IVVO-TNO Health Research, Gaubius Institute, Leiden, laboratory; Portugal: M. J. Halpern, J. Canena, Instituto Superior de Ciencas da Saude, Lisbon, recruitment center; Spain: L. Masana, J. Ribalta, A. Jammoul, and A. LaVille, Unitat Recerca Lipids, University Rovira i Virgili, Reus, recruitment center and laboratory; Switzerland: F. Gutzwiller and B. Martin, Institute of Social and Preventive Medicine, University of Zurich, recruitment center and laboratory; United Kingdom: D. St. J. O’Reilly, M. Murphy, Institute of Biochemistry, Royal Infirmary, Glasgow, recruitment center and laboratory; United Kingdom: S. E. Humphries, P. J. Talmud, V. Gudnason, and R. M. Fisher, University College London School of Medicine, London, laboratory; United Kingdom: D. Stansbie, A. P. Day, and M. Edgar, Department of Chemical Pathology, Royal Infirmary, Bristol, recruitment center and laboratory; United Kingdom: F. Kee* and A. Evans**, *Northern Health and Social Services Board and **Department of Epidemiology and Public Health, Queen’s University of Belfast, Belfast, recruitment center.
Footnotes
This work was supported by a grant from the Swiss National Research Foundation (to R.W.J.).
Abbreviations: CHD, Coronary disease; EARS, European Atherosclerosis Risk Study; LDL, low density lipoprotein; OGTT, oral glucose tolerance test; PON1, paraoxonase-1.
Received August 8, 2001.
Accepted December 5, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  J. L. San Millan, M. Corton, G. Villuendas, J. Sancho, B. Peral, and H. F. Escobar-Morreale Association of the Polycystic Ovary Syndrome with Genomic Variants Related to Insulin Resistance, Type 2 Diabetes Mellitus, and Obesity J. Clin. Endocrinol. Metab., June 1, 2004; 89(6): 2640 - 2646. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
