Effect of Polymorphisms of the Mthfr and Apoe Genes on Susceptibility to Diabetes and Severity of Diabetic Retinopathy in Brazilian Patients

Diabetes mellitus (DM) is a highly prevalent complex genetic disorder. There has been a worldwide effort in the identification of susceptibility genes for DM and its complications, and the 5-10-methylene-tetrahydrofolate reductase (MTHFR) and apolipoprotein-E (APOE) genes have been considered good candidate susceptibility genes to this condition. The objectives of the present study were to determine if the 677T MTHFR and ε2/ε3/ε4 APOE alleles are risk factors for DM and for severity of diabetic retinopathy (DR). A total of 248 individuals were studied: 107 healthy individuals and 141 diabetic patients (46 with type 1 diabetes and 95 with type 2 diabetes), who also had DR (81 with non-proliferative DR and 60 with proliferative DR). The poly-morphisms were analyzed by PCR followed by digestion with restriction enzyme or the single-nucleotide primer extension method. No evidence of association between the 677TT genotype of MTHFR gene and DM [cases: TT = 10/95 (10.6%); controls: TT = 14/107 (13%)] or with severity of DR was observed [cases: TT = 5/60 (8.5%); controls: TT = 9/81 (11.1%); P > 0.05]. We also did not find evidence of an association between APOE alleles and proliferative DR (ε2, ε3 and ε4 in cases: 9, 76, and 15%, and in controls: 5, 88, and 12%, respectively) but the carriers of ε2 allele were more frequent among patients with type 2 DM and DR than in controls [cases: 15/95 (15.8%); controls: 7/ 107 (6.5%); P < 0.05]. Therefore, our results suggest that the ε2 allele/ APOE might be a risk factor for diabetes in the Brazilian population.


Introduction
Diabetes is a highly prevalent complex disorder characterized by increased blood glucose levels.An autoimmune process destroys ß-cells in type 1 diabetes mellitus (DM1), whereas a combination of genetic and environmental factors brings about ßcell failure in type 2 DM (DM2), resulting in impaired action and secretion of insulin (1).The prevalence of DM in the population aged 30-69 years was 7.6% in the Brazilian Multicenter Study (2), which is comparable to European (8.8%) and USA (7.3%) data (2).The leading causes of morbidity and mortality in diabetic patients are the microand macrovascular complications (3).
Diabetic retinopathy (DR) is one of the major microvascular complications of diabetes.Clinical studies on subjects with DM have revealed substantial variation in the onset and severity of DR that is not fully explained by the known risk factors such as duration of DM and level of glycemic control.Furthermore, the risk of severe DR among the siblings of affected individuals is substantially increased (4,5), suggesting that genetic influences are operating in DR.
The identification of susceptibility genes associated with complex disorders has been a difficult task and case-control studies have been one of the approaches most frequently used.A large number of candidate genes have been examined in case-control studies, and a few of these studies have been carried out in more than one population (6).It is also of interest to identify genes involved in the severity of diabetes complications.Genes that play a role in vascular homeostasis, such as 5-10-methylenetetrahydrofolate reductase (MTHFR) and apolipoprotein-E (APOE), have been considered to be good candidates both for susceptibility to diabetes and for severity of the disease and its complications (7).
The c.677C>T polymorphism in the MTHFR gene, which encodes an enzyme involved in remethylation of homocysteine to methionine, leads to increased thermolability and impaired enzymatic activity.It is the most common genetic determinant of moderate hyperhomocystinemia, a risk factor for vascular disease.The 677T/MTHFR allele has also been associated with nephropathy and DR (8,9); however, the validity of these associations is still controversial (10)(11)(12).
The most studied functional polymorphisms of APOE are the ε2, ε3 and ε4 alleles, which differ by single-amino acid substitu-tions.The ε3 allele is the most frequent, representing 74 to 86% of all alleles, the frequency of the ε4 allele is approximately 7 to 16% and the frequency of the ε2 allele is 6.3 to 12% in European and American Caucasian populations (13)(14)(15).The allelic variation at the APOE locus contributes to lipid profile variation and has been suggested to be a risk factor for the development of micro-and macrovascular complications in diabetic patients in several studies, but not in all populations (16).There are, however, very few studies of this polymorphic system in DR, particularly for DM2 (17)(18)(19).
At-risk susceptibility genes might play a role in causing the disease depending on the population, and the molecular characterization of these genes is of the utmost importance in every population.Thus, the purposes of the present study were to determine if the 677T allele in the MTHFR gene and the ε2, ε3, ε4 alleles in the APOE gene play a role in the predisposition to DM and in the severity of DR in the Brazilian population.

Subjects
A total of 248 individuals were identified at Hospital das Clínicas, Faculdade de Medicina, São Paulo University (HC, FM, USP): 107 were control individuals (90 individuals of European-Brazilian descent and 17 individuals of African-Brazilian descent) with no DM or hypertension with a mean age of 66.11 ± 7.06 years and 141 were diabetic patients (46 with DM1 and 95 with DM2; 105 of European-Brazilian descent and 36 of African-Brazilian descent) with a mean age of 55.43 ± 15.33 years.The ethnic classification as European-Brazilian and African-Brazilian was based on phenotype (hair color, type and conformation of the nose and lips) and family history (20).The diagnosis of diabetes was based on clinical features, laboratory data and the guidelines of the Ameri-can Diabetes Association (21).Onset and duration of the disease were determined on the basis of the date of clinical diagnosis definition.Clinical and laboratory data were collected retrospectively and the average of the measurements made during 1997-2001 was considered.The main clinical features of the patients are summarized in Table 1.Most of patients were taking lipid-lowering drugs and therefore the lipid profile could not be used in genotype-phenotype correlation analysis.
Assessment of DR was performed by fundus ophthalmoscopy and biomicroscopy by a retinal specialist; fluorescein angiography was obtained in some cases when indicated.DR was classified according to the International Clinical Diabetic Retinopathy Disease Severity Scale (22) as: non-proliferative DR (NPDR) and proliferative DR (PDR).NPDR denotes signs of microaneurysms, intraretinal hemorrhage, exudates, macular edema, venous dilatation, soft exudates, peripheral ischemia, intraretinal microvascular abnormalities, and diffuse intraretinal hemorrhage.PDR denotes signs of neovascularization in the optic disc and elsewhere, vitreous hemorrhage, and fibrovascular proliferation.Of the 141 diabetic patients, 81 were non-PDR and 60 were PDR.Blood was drawn only after approval of the protocol by the local Ethics Committee.Each patient gave written informed consent.

Molecular and statistical analysis
DNA was isolated from whole peripheral blood samples by standard techniques (23).Genotyping of MTHFR gene polymorphism was carried out by PCR using primers and amplification conditions reported elsewhere (24), followed by a single-nucleotide primer extension assay (primer 5'-3': GCTGCGTGA TGATGAAATCG) analyzed by Megabace (Amersham-Biosciences Pharmacia Biotech Inc., Piscataway, NJ, USA).The different alleles of the APOE were detected according to published protocols (25).The distribution of genotypes and alleles was compared by the chi-square (χ²) or Fisher exact test.Hardy-Weinberg equilibrium was also verified in all samples.The level of significance adopted was 5%.The proportion of European-Brazilian and African-Brazilian descendants was the same in the subgroups analyzed.

Characterization of the sample
A total of 248 individuals were included in the study: 107 healthy individuals and 141 diabetic patients (46 with DM1 and 95 with DM2).All patients had DR (81 with NPDR and 60 with PDR).

Genotype analysis
The distribution of the genotypes for the 677T>C/MTHFR and the ε2/ε3/ε4/APOE polymorphisms are summarized in Table 2.
Genotype frequencies were in Hardy-Weinberg equilibrium in both polymorphic systems (MTHFR: controls P = 0.31 and patients P = 0.57; APOE: controls P = 0.50 and patients P = 0.87).Homozygotes for allele ε2 and ε4 were not observed in this sample.
In order to evaluate the effect of the genotypes on the severity of retinopathy we pooled the patients with DM1 and DM2 since development of DR is considered to be a consequence of the hyperglycemic state rather than a result of the underlying cause of diabetes (26)(27)(28).
The allelic and genotypic frequencies for the ε2/ε3/ε4 polymorphisms of APOE were not significantly different according to severity of retinopathy (NPDR vs PDR: χ² = 3.50, d.f.= 3, P = 0.31).Due to the low frequency of the ε2/ε4 genotype, individuals were divided into a group of carriers of the ε2 allele (ε2/ε3 and ε2/ε4 genotypes) and a group of non-carriers of the ε2 allele (ε3/ε3 and ε3/ε4 genotypes) for statistical analysis.
In this comparison, an association between APOE allele ε2 and DM2 was observed with a borderline level of significance (P = 0.04, OR = 2.67: 1.0-6.88).

Discussion
In the present study, we did not find any evidence of association between the 677T MTHFR allele and DM or DR, in agreement with some reports of the literature (10,11), including one in the Brazilian population in which all DR patients were of European ancestry (12).We believe that ethnicity did not lead to false-negative results, since similar results were observed after exclusion of the African-Brazilian descendants both from the control and patient's samples (the frequency of genotype TT was 15% (14/90) in the control group and 10% (10/105) in the patient group).It is important to note that the frequency of homozygotes for the 677T allele observed in our group of diabetic patients was comparable with the TT genotype frequency previously estimated in other ethnically mixed healthy populations of São Paulo and Ceará (29,30).
Maeda et al. (31) analyzed their data taking into account glycemic control and suggested that 677T/MTHFR might be a susceptibility locus for DR in individuals with poor glycemic control.If this hypothesis is correct, a significantly higher frequency of the TT genotype should have been an obligatory finding in the sample reported here, since the majority of our patients (112/ 141) had high glycated hemoglobin (GHb; >8.5%).Therefore, our data further support that the 677T/MTHFR polymorphism is not a predictive marker for DM or DR.
The APOE gene has been extensively studied in several diseases around the world (32).In Brazil, it has been more explored in Alzheimer's disease (14,15) and we are not aware of any genetic studies of APOE in diabetes or its complications in our population.The genotypic and allelic frequencies of the APOE polymorphic system in our control sample were similar to those of the control population of São Paulo (15).In the present study, we did not find evidence of an association between APOE alleles and the severity of retinopathy in either DM1 or DM2, in agreement with other reports (17,18).
We also did not find an association between APOE and DM1 in this sample.Although the lack of association may have been due to the small sample size, it is very likely that these alleles do not play a major role in the genetic predisposition to this type of diabetes.On the other hand, we observed an increased frequency of the ε2 allele among patients with DM2.Considering that all the DM2 patients had retinopathy, we cannot yet generalize this finding to DM2 without complications.
This positive association of allele ε2/ APOE with DM2 is in agreement with data from the literature, which points to the involvement of allele ε2/APOE in the occurrence of DM2 (33) or in diabetic nephropathy (34)(35)(36).We could speculate that our positive results were due to the simultaneous occurrence of DM and nephropathy; however, the frequency of the ε2/APOE allele was similar (ε2 = 13/190 = 6.8%) in patients with and without nephropathy (ε2 = 7/112 = 6.1%), implying that the higher frequency of the ε2/APOE allele might be primarily related to DM2.A possible mechanism to explain this association is that the increased triglyceride levels observed in the presence of the ε2 allele (37,38) can be a predisposing factor to insulin resistance and consequently to DM2 (39).
In summary, our data confirm reports that the 677C>T/MTHFR allele does not play an important role in the severity of DR.In contrast, the ε2/APOE allele seems to be a predisposing factor to DM2 but not to severity of DR.This result should be confirmed in a larger sample and indicates the need to further investigate APOE in patients with DM2 in the Brazilian population.

Table 1 .
Main clinical and laboratory data of diabetic patients.

Table 2 .
Genotype and allele distribution of MTHFR and APOE polymorphisms in diabetic patients with retinopathy and controls.