Polymorphic gene markers of lipid metabolism are associated with diabetic nephropathy in patients with type 1 diabetes mellitus

In groups of type 1 diabetes mellitus patients with and without clinical signs of diabetic nephropathy (n = 62 and n = 68, respectively), a search was made for associations between diabetic nephropathy and the polymorphic marker epsilon2/epsilon3/epsilon4 of apolipoprotein E gene (APOE), I/D marker of apolipoprotein B gene (APOB), and Ser447Ter marker of lipoprotein lipase-encoding gene (LPL). The risk of diabetic nephropathy was higher in the carriers of allele epsilon3 and genotype epsilon3/epsilon3 of the polymorphic marker epsilon2/epsilon3/epsilon4 of APOE gene as well as in the carriers of allele 1 and APOB genotype/gene (OR = 2.08 and 2.16; 1.91 and 2.11, respectively). Conversely, the carriers of allele D showed a reduced risk of this complication (OR = 0.52). No significant differences in distribution of alleles and genotypes of the polymorphic marker Ser447Ter of LPL gene were found between the groups. Our results indicate that the genes encoding two major components of lipid metabolism are involved in the development of diabetic nephropathy in patients with type 1 diabetes mellitus.     

Association of polymorphous markers Pro72Arg and C(-594)CC OF TP53 gene with diabetic polyneuropathy in patients with type 1 diabetes mellitus living in Moscow

The aim of this study was the search of association of polymorphous markers Pro72Arg and C(-594)CC of TP53 gene with diabetic polyneuropathy (DPN) in patients with type 1 diabetes mellitus with or without clinical signs of DPN. We have found that polymorphous marker Pro72Arg of TP53 gene was associated with DPN in Russian patients with type 1 diabetes mellitus living in Moscow. The carriers of Arg allele and Arg/Arg genotype had higher risk of DPN development (OR = 1.96; CI = 1.32-2.90; and OR = 2.14; CI = 1.23-3.73; relatively). On the contrary, the carriage of Pro allele was associated with the lower risk of DPN development (OR = 0.51; CI = 0.34-0.76). We have not found any association of polymorphous marker C(-594)CC of TP53 gene with DPN in Russian patients with type 1 diabetes mellitus living in Moscow.     

Genomics of type I diabetes mellitus and its late complications

In ethnic Russians, MHC (HLA) was shown to be the major locus determining the predisposition to type 1 diabetes mellitus (T1DM). To map the regions linked to T1DM, families with concordant or discordant sib pairs were selected from the Russian population of Moscow. With these families, linkage to T1DM was demonstrated for CTLA4 (IDDM12, 2q32.1-q33), which codes for a T-cell surface antigen, and PDCD2 (IDDM8, 6q25-q27), which is homologous to the mouse programmed cell death activator gene. With polymorphic microsatellites, regions 3q21-q25 (IDDM9) and 10p12.2 (IDDM10) were also linked to T1DM. Case/control and family studies of the polymorphic markers from region 11p13 revealed a new T1DM-associated locus in the vicinity of the catalase gene (CAT); linkage to this locus was not reported earlier for other populations. Diabetic polyneuropathy (DPN) proved to be associated with single-nucleotide polymorphisms Ala(-9)Val (SOD2), Arg213Gly (SOD3), and T(-262)C (CAT) and with a polymorphic microsatellite of the NOS2 promoter. Hence oxidative stress, which results from hyperglycemia, increased mitochondrial production of superoxide radicals, and insufficient activities of antioxidative enzymes, was assumed to play an important part in DPN development in T1DM. Diabetic nephropathy (DN) showed no association with the antioxidative enzyme genes. However, the association was observed for the insertion/deletion (I/D) polymorphism of ACE and the ecNOS34a/4b polymorphism of NOS3, two genes involved in controlling vascular tonicity, and for the I/D polymorphism of APOB and the epsilon 2/epsilon 3/epsilon 4 polymorphism of APOE, two genes involved in lipid transport. In addition, polymorphic microsatellites of chromosome 3q21-q25 proved to be closely associated with DN. The tightest association was established for D3S1550, carriers of allele 12 or genotype 12/14 having high risk of DN (OR = 4.85 and 6.25, respectively). Region 3q21-q25 was assumed to contain a major gene determining DN development, while the other DN-associated genes mostly affect the progression of DN.     

The association of the <Emphasis Type="Italic">TP53</Emphasis> polymorphisms Pro72Arg and C(−594)CC with diabetic polyneuropathy in Russian Muscovites with type 1 diabetes mellitus

The allele and genotype frequency distributions of the Pro72Arg and C(?594)CC polymorphisms of the TP53 gene were studied in type 1 diabetes mellitus (T1DM) patients, ethnic Russians from Moscow, with a T1DM record of no more than 5 years and diabetic polyneuropathy (DPN) or a T1DM record of more than 10 years but without DPN. The Pro72Arg polymorphism was associated with DPN, a higher risk of DPN being determined by allele Arg (OR = 1.96, CI 1.32?2.90) and genotype Arg/Arg (OR = 2.14, CI 1.23?3.73). Allele Pro was associated with a lower risk of DPN (OR = 0.51, CI 0.34?0.76). No association with DPN was observed for the C(?594)CC polymorphism.     

Association of the <Emphasis Type="Italic">POLG1</Emphasis> T(−365)C, <Emphasis Type="Italic">ANT1</Emphasis> G(−25)A,and <Emphasis Type="Italic">PEO1</Emphasis> G(−605)T polymorphisms with diabetic polyneuropathy in patients with Type 1 diabetes mellitus

This study is dedicated to a search for the association of the polymorphic markers T(?365)C of the POLG1 gene G(?25)A of the ANT1 gene and G(?605)T of the PEO1 gene with diabetic polyneuropathy (DPN) in Type 1 diabetes mellitus (DM1) patients. All patients were ethnic Russian Moscow residents, with DM1 records of no more than 5 years and DPN or DM1 records of more than 10 years but without DPN. We found that the polymorphic marker T(?365)C of POLG1 was associated with DPN in Russian patients with DM1. The carriers of the C allele and the CC genotype had a higher risk of DPN development (OR = 1.62; CI = 1.11–2.38; and OR = 1.76; CI = 0.99–3.13; respectively). In contrast, the T allele carrier status and the TT genotype were associated with a lower DPN risk (OR = 0.62, CI = 0.42–0.90; and OR = 0.61; CI = 0.35–1.07; respectively). We found no association of the polymorphic markers G(?25)A of ANT1 or G(?605)T of PEO1 with DPN in Russian DM1 patients living in Moscow.     

Genetic association between the apolipoprotein E (ApoE) gene alleles and various forms of Alzheimer's disease]

Allele epsilon 4 of the apolipoprotein E (APOE) gene is associated with higher risk of Alzheimer's disease (AD) in many, though not all, ethnic groups. The APOE allele and genotype frequency distributions were studied in 207 AD patients without cerebrovascular disorders, 62 AD patients with cerebrovascular disorders (combined AD), and 206 control individuals (ethnic Russians from the Russian population). The frequency of allele epsilon 4 in patients with early-onset and late-onset AD was three times higher than in control individuals (p < 0.000001). Compared with control people, patients with cerebrovascular disorders displayed a twofold higher frequency of allele epsilon 4; the difference between the two groups was significant (p = 0.0019). Relative risk of AD in carriers of allele epsilon 4 was five times higher than in carriers of alleles epsilon 2 and epsilon 3 (p < 0.000001). Allele epsilon 2 had a protective effect with respect to AD onset until 65 years of age (p = 0.015). Thus, APOE allele epsilon 4 proved to be a universal factor of early-onset, late-onset, and combined AD in ethnic Russians from Russia.     

Search for the association of polymorphic markers for genes coding for antioxidant defense enzymes, with development of diabetic polyneuropathies in patients with type 1 diabetes mellitus

The allele and genotype frequency distributions of polymorphic markers of genes coding for antioxidant enzymes were compared for type 1 diabetes mellitus patients with or without diabetic polyneuropathy (DPN). The groups (total 180 patients) had nonoverlapping (polar) phenotypes. Group DPN+ included 86 patients with DPN and diabetic record no more than 5 years. Control group DPN- included patients without DPN and diabetic record of at least 10 years. Comparative analysis with Fisher's exact test revealed a significant difference in allele and genotype frequency distributions of the T(-262)C polymorphic marker of the CAT gene. Polymorphic markers C1167T of the CAT gene, Pro/Leu of the GPX1 gene, 0/+ of the GSTT1 gene, and 0/+ of the GSTM1 gene showed no significant difference in allele or genotype frequency distribution. On this evidence, these markers were not associated with DPN in the sample examined.     

Apolipoprotein B, apolipoprotein E, and angiotensin-converting enzyme polymorphisms in 2 Italian populations at different risk for coronary artery disease and comparison of allele frequencies among European populations

Polymorphisms at the apolipoprotein B (APOB XbaI, EcoRI, insertion-deletion), apolipoprotein E (APOE), and angiotensin-converting enzyme (ACE) loci are thought to be involved in susceptibility to coronary artery disease (CAD) and myocardial infarction. The aim of this study was to determine whether the allele distribution of the APOB, APOE, and ACE polymorphisms is different in 2 Italian regions with higher (northern Italy) and lower (Sardinia) CAD occurrence. The frequencies of the APOB and APOE alleles that are considered CAD risk factors were higher in northern Italy (APOB X- = 0.655; APOB R- = 0.198; APOB insertion = 0.757; APOE*4 = 0.110) than in Sardinia (APOB X- = 0.568; APOB R- = 0.159; APOB insertion = 0.680; APOE*4 = 0.052), although only APOE allele frequencies differed significantly (p = 0.001). ACE deletion allele frequencies in the 2 geographic areas showed an opposite pattern (northern Italy = 0.658; Sardinia = 0.721). Furthermore, we investigated the impact of APOB and APOE polymorphisms on interindividual variation in total cholesterol level in the 2 Italian samples, which differ in dietary habits. Only APOE phenotypes showed different mean levels of total cholesterol; the association was significant only in northern Italy (p = 0.04), where continental dietary habits and higher mean cholesterol levels prevail. These results support the suggestion that the cholesterol increasing effect of APOE*4 is environmentally mediated. Analysis of allele distributions among European populations, with remarkable differences in CAD prevalence, revealed a constant positive relationship between APOE*4 allele frequency and CAD incidence. The highest frequencies of APOB X- and R- were observed in Finland, where the incidence of CAD is high, and there is a partial agreement between APOB R- frequency and CAD occurrence across Europe, while APOB insertion and ACE deletion alleles are evenly distributed among European populations.     

Effects of APOE, APOB and LDLR variants on serum lipids and lack of association with xanthelasma in individuals from Southeastern Brazil

Xanthelasma might be a clinical manifestation of dyslipidemia, a recognized risk factor for coronary artery disease. We investigated the association of apolipoprotein E (APOE HhaI), apolipoprotein B (APOB XbaI and Ins/Del) and LDL receptor (LDLR AvaII and HincII) gene polymorphisms with lipid profiles in 100 Brazilians with xanthelasma and 100 controls. Allele frequencies were similar in both groups. APOE, APOB and LDLR genotypes were not correlated with differences in the serum lipid profile. In individuals with xanthelasma, the APOB D allele was associated with less chance of having increased LDL-cholesterol (O.R. = 0.16, CI95% = 0.03-0.94, p = 0.042). In the control group, the APOB X+ allele was associated with less chance of having both increased total cholesterol (O.R. = 0.16, CI95% = 0.03-0.78, p = 0.023) and increased LDL-cholesterol (O.R. = 0.10, CI95% = 0.02-0.60, p = 0.012). Moreover, there was a significantly higher frequency of control individuals (68%) with elevated serum triglyceride levels, compared to patients (48%, p = 0.008). On the other hand, triglyceride levels in controls also seemed to be influenced by all other gene polymorphisms studied, an effect that might be enhanced by environmental factors.     

LRP-associated protein gene (LRPAP1) and susceptibility to degenerative dementia

Difference in genetic makeup of lipid metabolizing proteins may increase susceptibility to dementia. The aim of this study was to investigate variations in two lipid metabolizing genes, low-density lipoprotein receptor-related protein-associated protein gene (LRPAP) and apolipoprotein E (APOE), in vascular and degenerative dementias. This hospital-based association study included 147 patients with dementia and 163 age-matched controls. Genotyping for LRPAP1 intron 5 insertion/deletion and APOEHhaI polymorphism was carried out by polymerase chain reaction (PCR)/PCR-restriction fragment length polymorphism method. The frequency of DD genotype and *D allele of LRPAP gene was high in degenerative dementias compared with controls (76.6% vs. 44.8% and 86.4% vs. 69.3%, respectively) suggesting increased susceptibility [P < 0.0001, odds ratio (OR) =3.76 for DD genotype and P < 0.0001, OR = 2.80 for D allele]. Vascular dementia patients showed statistically insignificant but increasing trend for allele I (40.0% vs. 30.7%). We also observed significant association of APOEepsilon4 allele with degenerative dementias (P < 0.0001, OR =5.35). Our study suggests that LRPAP1-D and APOE E4 alleles significantly increase the susceptibility to degenerative dementias.     

Polymorphic markers of the NO synthase genes and genetic predisposition to diabetic polyneuropathy in patients with type 1 diabetes mellitus

The allele and genotype frequencies of polymorphic markers of NOS1, NOS2 and NOS3 genes, encoding three types of NO synthases, were compared in type 1 diabetes patients with and without diabetic polyneuropathty. 180 type 1 diabetes patients (T1DM) of Russian or Eastern Slavonic origin, living in Moscow city, were divided into two groups using non-overlapping (polar) phenotypes. 86 patients had overt DPN and T1DM duration in this group was less than 5 years (DPN+ group) and 94 patients had no clinical DPN and T1DM duration was more than 10 years (DPN- group). We have not found the significant differences of allele and genotype frequencies of polymorphic markers (CA)n of NOS1 gene, (CCTTT)n of NOS2 gene, ecNOS4a/4b and Glu298Asp of NOS3 gene that indicates that all these markers are not associated with diabetic polyneuropathty. Only in the case of (CCTTT)n marker of NOS2 gene we have found a tendency for the association of 14 allele with DPN development. The carriers of this allele have the lower risk of DPN in T1DM.     

Association of polymorphic markers of the lipid metabolism genes with diabetic polyneuropathy in type 1 diabetes mellitus

A possible association of the polymorphic markers 2/3/4 of the apolipoprotein E gene (APOE) and I /D of the apolipoprotein B gene (APOB) with diabetic polyneuropathy (DPN) was analyzed in patients with type 1 diabetes mellitus (T1DM) with (N=86) or without (N=94) clinical signs of DPN. The two groups did not differ significantly in allele and genotype frequencies of the 2/3/4 polymorphic marker of the APOE gene. Analysis of the allele and genotype frequency distributions of the I/D polymorphic marker of the APOB gene showed that risk of DPN is higher in carriers of allele I or genotype I/I (OR=1.66 and 2.01, respectively) and lower in carriers of allele D (OR=0.60). The results implicate the APOB gene, which codes for one of the major components of the lipid metabolism system, in DPN development in patients with T1DM.__________Translated from Molekulyarnaya Biologiya, Vol. 39, No. 2, 2005, pp. 230–234.Original Russian Text Copyright © 2005 by Voronko, Yakunina, Strokov, Lavrova, Nosikov.     

Association of the SOD2 Ala(–9)Val and SOD3 Arg213Gly Polymorphisms with Diabetic Polyneuropathy in Diabetes Mellitus Type 1

Single-nucleotide polymorphisms of the genes for mitochondrial (SOD2) and extracellular (SOD3) superoxide dismutases were tested for association with diabetic polyneuropathy (DPN) in diabetes mellitus (DM) type 1. Patients (N = 180) were divided into two groups with nonoverlapping (polar) phenotypes. Group DPN+ included 86 individuals with DPN and DM type 1 record of no more than 5 years. Group DPN– included 94 patients with DM type 1 record of more than 10 years but without clinical signs of DPN. Fisher's exact test revealed significant differences in allele and genotype frequencies for the two groups. Higher frequencies of SOD2 allele Val and genotype Val/Val and of SOD3 allele Arg and genotype Arg/Arg were established for group DPN+. On this evidence, SOD2 and SOD3 were associated with DPN in DM type 1.     

Genetic influences on lipid metabolism trait variability within the Stanislas Cohort.

The contribution of 17 polymorphisms within 13 candidate genes on lipid trait variability was investigated by a multiplex assay in 772 men and 780 women coming for a health checkup examination. The studied genes were APOE, APOB, APOC3, CETP, LPL, PON, MTHFR, FGB, GpIIIa, SELE, ACE, and AGT. We found that APOB-Thr71Ile, APOE-(112/158), APOC3-1100C/T, and SELE-98G/T polymorphisms had a significant effect on lipid traits (P < or = 0.001 to P < or = 0.01). Genetic effects accounted for 3.5-5.7% of variation in apolipoprotein B (apoB)-related traits among men, and for 5.7-9.0% among women. The contribution of APOE polymorphism on apoB-related traits variability was two to three times more important in women than in men. We found suggestive evidence for interactive effects between genetics and age, smoking status, and oral contraceptives. Increase of LDL-cholesterol and apoB concentrations with age was stronger among the epsilon4 carriers in women, and apolipoprotein A-I (apoA-I) concentration decreased with age in epsilon4 male carriers. The effect of epsilon2 allele on LDL-cholesterol was more important in the oral contraceptive users. In nonsmokers only, the APOC3-1100C allele in women was related to lower apoB-related traits concentrations, and in men to higher apoA-I and HDL-cholesterol concentrations. In conclusion, this work, in addition to the reinforcement of the already known associations between APOB, APOE, and APOC3 genes and lipids, leads to new perspectives in the complex relationships among genes and environmental factors. The newly observed relationships between E-selectine gene and lipid concentrations support the hypotheses of multiple metabolic pathways contributing to the complexity of lipids variability.     

Relations of APOE promoter polymorphisms to LDL cholesterol and markers of subclinical atherosclerosis in young adults

The common apolipoprotein E (apoE) gene (APOE) epsilon2/epsilon3/epsilon4 polymorphism explains part of serum lipid variation, and polymorphisms in the APOE promoter region have been proposed to participate in the regulation of serum lipid levels within the most common APOE epsilon3/epsilon3 genotype group. We determined APOE -219G/T and +113G/C promoter genotypes and estimated APOE haplotypes in 525 participants of the Cardiovascular Risk in Young Finns Study. We studied the associations of the APOE promoter polymorphisms and their haplotypes with cross-sectional and longitudinal serum lipid and apolipoprotein concentrations as well as with flow-mediated dilatation (FMD), carotid artery compliance (CAC), and intima-media thickness (IMT) within the APOE epsilon3/epsilon3 carriers. We found no significant association between the APOE promoter genotypes and serum lipids [low density lipoprotein-cholesterol (LDL-C), HDL-C, and triglycerides], apolipoproteins (apoA-I and apoB), or brachial artery FMD, CAC, or carotid IMT in either men or women. In longitudinal analyses in males, the carriers of heterozygous genotypes (-219G/T or +113G/C) and, furthermore, carriers of the -219T/+113C/epsilon3 haplotype had significantly higher LDL-C and total cholesterol concentrations throughout the 21 year follow-up period compared with homozygous G allele carriers or noncarriers of the -219T/+113C/epsilon3 haplotype. Such associations were not found in females. In summary, the APOE promoter polymorphisms -219G/T and +113G/C as well as their haplotype are associated with longitudinal changes in LDL-C and total cholesterol concentrations in young Finnish males but do not seem to be major determinants for FMD, CAC, or carotid IMT in males or females.     

Association of the SOD2 Ala(-9)Val and SOD3 Arg213Gly polymorphisms with diabetic polyneuropathy in patients with diabetes mellitus type 1

Single-nucleotide polymorphisms of the genes for mitochondrial (SOD2) and extracellular (SOD3) superoxide dismutases were tested for association with diabetic polyneuropathy (DPN) in diabetes mellitus (DM) type 1. Patients (n = 180) were divided into two groups with nonoverlapping (polar) phenotypes. Group DPN+ included 86 individuals with DPN and DM type 1 record of no more than 5 years. Group DPN-included 94 patients with DM type 1 record of more than 10 years but without clinical signs of DPN. Fisher's exact test revealed significant differences in allele and genotype frequencies for the two groups. Higher frequencies of SOD2 allele Val and genotype Val/Val and of SOD3 allele Arg and genotype Arg/Arg were established for group DPN+. On this evidence, SOD2 and SOD3 were associated with DPN in DM type 1.     

ACE I/D polymorphism in Alzheimer’s disease

Angiotensin-converting enzyme (ACE) has been reported to show altered activity in patients with neurological diseases. The recent studies found that a 287 bp insertion/deletion (I/D) polymorphism of the ACE gene may be associated with susceptibility to Alzheimer’s disease (AD) but the results have been heterogenous between studies in Europe. In the present study we examined for the first time the association of ACE I/D polymorphism along with APOE genotype in 70 sporadic AD and 126 control subjects in Slovak Caucasians (Central Europe). An increased risk for AD was observed in subjects with at least one APOE*E4 allele (OR=3.99, 95% CI=1.97–8.08). No significant differences for the genotype distribution or the allele frequency were revealed comparing controls and patients for ACE gene. Gene-gene interaction analysis showed increase of the risk to develop AD in subjects carrying both the ACE DD genotype and the APOE*E4 allele (OR=10.32, 95% C.I. 2.67–39.81).     

Association of Polymorphic Markers of the Antioxidant Enzyme Genes with Diabetic Polyneuropathy in Type 1 Diabetes Mellitus

The allele and genotype frequency distributions of polymorphic markers of genes coding for antioxidant enzymes were compared for type 1 diabetes mellitus patients with or without diabetic polyneuropathy (DPN). The groups (total 180 patients) had nonoverlapping (polar) phenotypes. Group DPN+ included 86 patients with DPN and diabetic record no more than 5 years. Control group DPN– included patients without DPN and diabetic record of at least 10 years. Comparative analysis with Fisher's exact test revealed a significant difference in allele and genotype frequency distributions of the (–262) polymorphic marker of the CAT gene. Polymorphic markers C1167T of the CAT gene, Pro/Leu of the GPX1 gene, 0/+ of the GSTT1 gene, and 0/+ of the GSTM1 gene showed no significant difference in allele or genotype frequency distribution. On this evidence, these markers were not associated with DPN in the sample examined.     

Polymorphic Gene Markers of Lipid Metabolism Are Associated with Diabetic Nephropathy in Patients with Type 1 Diabetes Mellitus

In groups of type 1 diabetes mellitus patients with and without clinical signs of diabetic nephropathy (n = 62 and 68, respectively), a search was made for associations between diabetic nephropathy and the polymorphic marker ε2/ε3/ε4 of apolipoprotein E gene (APOE), I/D marker of apolipoprotein B gene (APOB), and Ser447Ter marker of lipoprotein lipase-encoding gene (LPL). The risk of diabetic nephropathy was higher in the carriers of allele ε3 and genotype ε3/ε3 of the polymorphic marker ε2/ε 3/ε4 of APOE gene as well as in the carriers of allele I and APOB genotype I/I gene (OR = 2.08 and 2.16; 1.91 and 2.11, respectively). Conversely, the carriers of allele D showed a reduced risk of this complication (OR = 0.52). No significant differences in distribution of alleles and genotypes of the polymorphic marker Ser447Ter of LPL gene were found between the groups. Our results indicate that the genes encoding two major components of lipid metabolism are involved in the development of diabetic nephropathy in patients with type 1 diabetes mellitus.__________Translated from Genetika, Vol. 41, No. 7, 2005, pp. 931–937.Original Russian Text Copyright © 2005 by Yakunina, Shestakova, Voron’ko, Vikulova, Savost’yanov, Chugunova, Shamkhalova, Dedov, Nosikov.     

GAB2 alleles modify Alzheimer's risk in APOE epsilon4 carriers

The apolipoprotein E (APOE) epsilon4 allele is the best established genetic risk factor for late-onset Alzheimer's disease (LOAD). We conducted genome-wide surveys of 502,627 single-nucleotide polymorphisms (SNPs) to characterize and confirm other LOAD susceptibility genes. In epsilon4 carriers from neuropathologically verified discovery, neuropathologically verified replication, and clinically characterized replication cohorts of 1411 cases and controls, LOAD was associated with six SNPs from the GRB-associated binding protein 2 (GAB2) gene and a common haplotype encompassing the entire GAB2 gene. SNP rs2373115 (p = 9 x 10(-11)) was associated with an odds ratio of 4.06 (confidence interval 2.81-14.69), which interacts with APOE epsilon4 to further modify risk. GAB2 was overexpressed in pathologically vulnerable neurons; the Gab2 protein was detected in neurons, tangle-bearing neurons, and dystrophic neuritis; and interference with GAB2 gene expression increased tau phosphorylation. Our findings suggest that GAB2 modifies LOAD risk in APOE epsilon4 carriers and influences Alzheimer's neuropathology.