The effects of APOE-ε4 on the BOLD response

Abstract

In the last decade, functional magnetic resonance imaging (fMRI) has emerged as a tool to study changes in brain function associated with a genetic risk for Alzheimer's disease (AD), with a particular focus on the effects of the APOE-ε4 allele. This review compiles the existing literature concerning the effects of APOE genotype on the blood oxygen level dependent (BOLD) response, measured during task-based functional magnetic resonance imaging. While most studies report a significant difference in brain activity between carriers and noncarriers of the ε4 allele, there are inconsistencies in the direction and location of change. These inconsistencies were addressed by examining the effect of task, family history of Alzheimer's disease, and age on the relationship between APOE genotype and the BOLD response, but no clear pattern emerged. The review discusses the interpretation of BOLD differences between ε4 carriers and noncarriers, provides suggestions for future studies, and highlights important limitations of this type of research.

Introduction

Current projections indicate that by the year 2050, over 115 million people worldwide will suffer from dementia, with Alzheimer's disease (AD) accounting for approximately 75% (86 million) of these cases (Alzheimer's disease International, 2009). An important step in preparing for, or even preventing this healthcare crisis is to understand the biology of those individuals predisposed to developing AD. Possession of an APOE-ε4 allele is the best established genetic risk factor for sporadic AD (Rocchi et al., 2003). APOE codes for the production of apolipoprotein E (ApoE), and has 3 major isoforms, E2, E3, and E4, which are coded for by the ε2, ε3, and ε4 alleles, respectively (Zannis et al., 1981). In addition to increased risk, ε4 carriers develop AD at a younger age than individuals homozygous for the ε3 allele. This pattern is dose-dependent: relative risk increases, and age at onset decreases, with increasing copies of the ε4 allele (Farrer et al., 1997). In contrast, ε2 carriers may be protected against the disease (Corder et al., 1994, Myers et al., 1996). Although the role of apolipoprotein E in the pathogenesis of AD is yet to be fully elucidated, contributions of ApoE4 at virtually every step in the pathological mechanism have been recorded (Cedazo-Mínguez, 2007).

The combination of genetic information and neuroimaging has proven useful in studying changes in the brain related to a predisposition to AD (Small, 2002). Structural and metabolic differences in the brains of ε4 carriers relative to noncarriers have been detected with conventional magnetic resonance imaging (MRI) and positron emission tomography (PET), the details of which are reviewed elsewhere (Bookheimer and Burggren, 2009, Small, 2002). In the last decade, functional MRI (fMRI) has also emerged as a tool to study the effects of a genetic risk for AD. fMRI has the advantage of being able to detect subtle differences in brain function before permanent changes in gross anatomy occur, and without the use of radiation (Small, 2002).

Smith and colleagues published the first report to use fMRI to investigate the effect of a genetic risk for AD on brain function. They reported that, relative to those at low risk, high risk individuals (ε4 carriers with a family history of AD) demonstrated decreased blood oxygen level dependent (BOLD) activation during visual confrontation naming and a letter fluency task (Smith et al., 1999). The suggestion that a genetic risk for AD manifests as a change in brain function during a cognitive task led to a proliferation of studies using fMRI to investigate such a risk. Unfortunately, there has been great variability in the results of these studies, suggesting that the effect of APOE genotype on brain function, as measured with fMRI, is inconsistent. The apparent lack of a clear effect may be due to considerable heterogeneity in study design. For example, the variety of tasks used to study the effects of APOE genotype on the BOLD response may contribute to the inconsistencies (Burggren et al., 2002, Köhler et al., 2002). In particular, due to the early and disproportionate impairment of episodic memory in AD (Yankner et al., 2008), it is possible that APOE genotype may affect brain function in this cognitive domain differently than in others (Burggren et al., 2002). Furthermore, other risk factors for AD are known to influence BOLD activation, including a family history of dementia and age (Bassett et al., 2006, Cerf-Ducastel and Murphy, 2009). Between-study differences in all of these factors may be masking the true effect of APOE genotype on the BOLD response. The objectives of this review are to present a concise summary of the literature concerning APOE and fMRI, and to examine to what extent differences in the paradigm used, family history of the subjects, and age of the subjects may be contributing to the reported inconsistencies. The review will conclude with suggestions regarding future research in this field, and by discussing the limitations of this type of research.