Schizophrenia is a complex disorder characterized by disturbed patterns of thought, feelings, and behavior with a lifetime risk of 1%.1 Several family, twin, and adoption studies have demonstrated an important genetic contribution in its etiology.

In the last several years, a number of studies have suggested that the apolipoprotein E (APOE) gene is associated with this disorder. Apolipoprotein E is a 34-kDa protein encoded by a gene on human chromosome 19q13.2. The APOE gene has three introns and four exons; it is polymorphic, with three common alleles designated ε2, ε3, and ε4 coding for three common isoforms which differ by one or two amino acids. Changes of two single nucleotides in the APOE gene at positions 334 and 472 together with the combinations 334T/472T, 334T/472C, and 334C/472C constitute the known isoform-specific APOE alleles: ε2, ε3, and ε4, respectively.2

APOE is synthesized and secreted in the liver, brain, and skin tissues. In the CNS, APOE is synthesized in astrocytes and oligodendrocytes; it is thought to play a role in neural growth and repair.3,4

Recently, APOE-ε4 has been implicated as a risk factor for Alzheimer′s disease5,6 and has been correlated with an earlier age of onset of this illness.7 Moreover, an association between APOE and cognitive decline in elderly adults has been reported.8 Evidence suggests that the APOE genotype may influence the phenotype of several neuropsychiatric illnesses when combined with other factors in a variety of disorders.9

In 1995, Harrington et al.10 reported an increased ε4 allele frequency of the APOE gene in 62 patients with schizophrenia. However, our study11 together with other reports demonstrated no association among APOE genotypes and schizophrenia.11—15 Since 1995, growing interest on this issue prompted a variety of studies in Caucasian and Asian populations dealing with APOE polymorphism and schizophrenia, though no conclusive outcomes were attained.16 To date, no association has been established between the APOE gene and this mental disorder in the Mexican population.

On the other hand, in 1998 Artiga et al.17 found allelic polymorphism in the transcriptional regulatory region of the APOE gene. This region is very important because it can have a functional repercussion mediated by the regulation of APOE transcription. Nevertheless, only a few reports have analyzed the association of this regulatory region and schizophrenia.

In this work, we determined the association of the genotypes APOE and APOE-219G/T promoter polymorphisms with schizophrenia in sibling-pair families.

Participants

Sixty sibling pairs with schizophrenia from families living in Mexico City, Mexico, participated in this study. These families fulfilled the inclusion criteria of having at least two siblings affected with this disorder and descending from Mexican parents and grandparents. All patients participating in the study gave written consent to this effect. All patients were evaluated by a trained psychiatrist using the Diagnostic Interview for Genetic Studies (DIGS), validated in Spanish. The sample in this study forms part of a multicenter linkage project.18

APOE Genotyping

Genomic DNA was extracted from peripheral blood leukocytes using a modified version of the protocol from Lahiri and Nurnberger19 and amplified using the polymerase chain reaction assay reported by Wenham et al.20 The polymerase chain reaction products were digested with CfoI and the fragments separated by electrophoresis on a 10% polyacrilamide gel, stained with ethidium bromide, and visualized under ultraviolet illumination.

APOE-219G/T (rs405509) Genotyping

The final volume of the polymerase chain reaction reaction was 5 μL and consisted of 20 ng of genomic DNA, 2.5 μL of TaqMan Master Mix, and 0.125 μL of 20× Assay made to order. The amplification was performed in 96-well plates using the TaqMan Universal Thermal Cycling Protocol. After the polymerase chain reaction endpoint was reached, fluorescence intensity was measured with the 7500 real time polymerase chain reaction system with SDS v2.1 software (Applied Biosystems). An allelic discrimination was performed resulting in the clear identification of three genotypes for APOE-219G/T polymorphism. All genotyping was performed blind to patients’ outcome. As a quality control in our genotyping analyses, we used random blind duplicates.

Statistical Methods

The difference in allele frequencies between parents and affected siblings was compared and analyzed by the statistical package APL software v.1.1 (available at http/www.chg.duke.edu/research/software.html).21 This program is based on the difference between the observed number of copies of a specific allele in affected siblings and the expected number of copies conditional on parental genotypes under the null hypothesis, which proposes no association or linkage in nuclear families. This program adjusts to the characteristics of our sample since it can analyze families with up to three affected siblings and arbitrary numbers of unaffected siblings. The simulations generated for the APL program have shown that association in the presence of linkage test (APL) is more powerful than family-based association test (FBAT) and pedigree disequilibrium test (PDT) methods.21

Given that disease loci can have different effects on males and females, we performed separate testing for males and females, because APL may be more powerful and informative for sex-specific effects.21 We explored the relationship between APOE alleles and age of onset of schizophrenia. APL analysis was performed using separate tests for early-onset (≤18 years old) and late-onset families.21 Bonferroni’s correction was applied for multiple testing (level of significance was set to α<0.05/2=0.025).

Haplotype Construction

Haplotype analysis was performed with the APOE and -219G/T polymorphisms using the PHASE program (version 2.0.2). Finally the participants were categorized in distinct sharing haplotypes in best pairs.22

Significant differences in allelic sharing were found for allele ε3 in the general sample (χ²=6.24, p=0.01); this was mainly observed in female patients (χ²=8.33, p=0.003). For allele ε2 (p=0.04) and allele ε4 (p=0.09), no significant differences were observed. When the results of allelic sharing for APOE-219G/T were summarized, no differences between frequencies of the transmitted and nontransmitted allele were seen in the general sample. However, a suggested association was encountered in the male group (p=0.02). Finally, Table 1 presents the results of haplotype analysis. Higher frequencies of haplotypes APOE-ε3/APOE-219G were found in the general group (χ²=11.61, p=0.01). However, no differences were encountered in the female (p=0.18) and male groups (p=0.13).

Finally, APL analysis of age of onset (late and early groups) did not show significant differences for the allele transmission of APOE. However, we observed a lower allele transmission of APOE-ε2 in the early onset families (p=0.02). Analysis of APOE-219G/T did not show statistical differences associated with the age of onset (data not shown).

In this study, we investigated the association in the presence of linkage among APOE and APOE-219G/T polymorphisms and schizophrenia by assessing the frequencies of transmitted and nontransmitted haplotype combinations. We observed that subjects carrying the APOE-ε3/APOE-219G combination showed a higher frequency of this transmitted haplotype than noncarriers. An association in the presence of linkage for APOE ε3 and schizophrenia was found in female patients; the same happened for APOE -219G and this mental disorder in the male group.

To our knowledge this is the first study correlating the APOE gene and schizophrenia in sibling pairs. The APOE gene has been previously associated with schizophrenia in case-control association studies,9,10 with the outcome that this gene may modulate the age of onset of this disease.23 A gender specific association was established between female carriers of APOE-ε4 and an early onset of schizophrenia.24 However, all related reports in the literature on this issue were derived from population-based case-control studies, which can be mistaken for stratification problems. Family-based studies are robust for population stratification. Nevertheless, we found no family-based studies in the literature correlating APOE and schizophrenia. Therefore, we decided to test the APOE locus by the method of association in the presence of linkage (APL) analysis in a sample of Mexican families. As a result, we found that APOE-ε3 was associated with schizophrenia. In view of the increased probability of transmission of the APOE-ε3 from parents to affected offspring and a decreased transmission of alleles APOE-ε2 and -ε4, we conclude that APOE-ε3 may be an important susceptibility allele for schizophrenia. The analysis of the female results showed that this group exhibited a marked statistical significance for this allele. Our results suggest that in females the APOE-ε3 allele could contribute to the risk for schizophrenia. This sex-linked association is relevant, although further studies in which the number of families is increased are necessary. Interestingly, our findings that showed a lower allele transmission of APOE-ε2 in the total sample and in the analysis of early age of onset may suggest that APOE-ε2 protects against the development of schizophrenia. However, in view of the small sample size, these results must be taken with caution because the families were further subdivided by schizophrenia age-of-onset. The influence of APOE-ε2 has been investigated in neuropsychiatric disorders. It has been suggested that the possession of the APOE-ε2 is associated with late-onset Alzheimer’s disease,25 longevity in Down’s syndrome,26 protection from the development of dementia,27 and protection against early-onset schizophrenia,28 although the mechanism by which APOE participates in the different pathogenic processes remains to be elucidated.

The APOE promoter has been characterized and contains a number of important regulatory elements participating in gene transcription activity.17 APOE-219G/T polymorphism located on 5′-franking region has been associated with variations with gene transcriptional activity.17 A negative association between APOE-219G/T and schizophrenia was reported.29 Our analysis observed an association between APOE-219G and male schizophrenia patients (p=0.02).

The APOE-219G/T polymorphism has been shown to produce variations in APOE transcriptional activity in neural cells, where the G allele has been related to higher activity,30 suggesting that APOE could have a specific role in some pathological conditions.31,32

To our knowledge, there are no previous studies analyzing APOE-219G/T/APOE haplotypes in schizophrenia. Mexican schizophrenia patients showed a preferential transmission of APOE-ε3/APOE-219G haplotype. We did not find differences linked to gender because a limitation of our study was the small sample size. Therefore, future studies in a larger sample size are important to determine whether the APOE-ε3/APOE-219G haplotype plays an etiological role in the susceptibility to schizophrenia or is associated with gender differences.

Recently, a genome-wide scan in a Latino population from the U.S., Mexico, Guatemala, and Costa Rica included 99 sibling-pair families diagnosed with schizophrenia.18 The study reported significant evidence of linkage on chromosome 1pter-p36 and suggestive linkage on chromosome 18p11 and 5q32. However, it did not show linkage for the APOE region.18 The present study included only a Mexico City, Mexico, sample. This selection may decrease genetic heterogeneity which in turn may increase statistical power if one is studying a polymorphism in a particular population.33

Our results showed a preferential transmission of APOE-ε3/APOE-219G haplotype in sibling-pairs affected with schizophrenia. Also, we observed a lower transmission of APOE-ε2 allele in schizophrenia patients with an early age of onset, suggesting a protective effect in the development of schizophrenia in the Mexican population.

Future studies must analyze whether the association of this gene with schizophrenia can be replicated in larger samples of Mexican families suffering from this mental disorder.

Data and biomaterials used in this research report were collected by the International Neuro-Genetics Association of Spanish America and the United States (INGASU) and funded by a collaborative NIMH grant (Genetics of Schizophrenia in Latino Populations) to Dr. Michael Escamilla (MH60881) and to Dr. Ricardo Mendoza (MH60875).