BRAIN COACTIVITY: PROSPECTIVE MEMORY, APOE AND AGE Candidate no. 87533
Differences in brain coativity in prospective memory associated with varying APOE allele combinations and age: An event-related fMRI experiment.
Candidate Number: 87533
Supervisor: Jennifer Rusted
Word Count: 5, 956 (inc. footnotes)
Psychology BSc
School of Psychology, University of Sussex
May 2013
Acknowledgements
This study was funded by a BBSRC grant to Jennifer Rusted (BB/H000518/1) who supervised this project. Thanks to Torsten Ruest for the data collection and Simon Evans for study design and image preprocessing. The writer of this report was responsible for the PPI analysis – with guidance from Simon Evans – and completion of this manuscript.
Abstract
The apolipoprotein (APOE) ε4 allele is theorised to cause accelerated cognitive aging because of its antagonistic pleiotropic properties: the allele correlates with increased cognitive performance in young adulthood and earlier onset of AD in older adulthood. To better understand how APOE affects cognitive aging, the current study investigated differences in brain coactivation as a function of genotype (ε4+s vs. ε4-s) and age (middle age vs. younger adult) in addition to trial-type (ongoing trials vs. prospective memory (PM) trials). The study primarily focused on middle-aged participants who completed a computerised-PM task whilst in the fMRI scanner. Secondarily, a younger cohort was added to examine the genotype-specific effects of age on coactivation. The current study predicted there would be an increase in bilateral frontal coactivation as a function of age in ε4+s based on theories of normal cognitive aging and the antagonistic pleiotropic theory. The most significant finding was the genotype-specific effect of age: increased bilateral coactivation with the lInfFrontal (seed) region for middle-aged ε4+s compared to younger ε4+s when masked by the contrast, middle-aged ε4-s > younger ε4-s. For future research, it would be interesting to conduct a longitudinal study that focused on coactivation differences as a function of age, genotype and AD-conversion to better understand how coactivation differences predict AD.
Keywords: APOE ε4, antagonistic pleiotropy, prospective memory, coactivation
1. Introduction
1.1 APOE ε4, The Brain and Alzheimer’s Disease
Since the early 1980s—when beta-amyloid (Aβ) plaques and neurofibrillary tangles were first linked to the Alzheimer’s Disease (AD) pathology— there has been an ongoing debate regarding what causes AD (Selkoe, 2011). The conversation continues because no genetic factor, biomarker or environmental factor has been unequivocally linked to the cause of AD.
In the mid-1990s, it was discovered that harbouring at least one apolipoprotein (APOE) ε4 allele, located on chromosome 19, increased the risk of AD and negatively correlated with the age of onset (Selkoe, 2011). In support of this discovery, Corder et al. (1993) found that 64% of patients within a sporadic AD sample and 80% within a familial sample had at least one ε4 allele, whilst only 14% of the general population have the allele (Bertram, McQueen, Mullin, Blacker, & Tanzi, 2007). Therefore, though harbouring the ε4 allele is not the direct cause of AD, the allele’s presence is correlated with AD[1].
The ε4 allele’s overrepresentation in the AD population has correlated with the presence of amyloid beta (Aβ) plaques (Fillipinni, 2011; Mann, 1991), neuronal atrophy (Buttini et al., 1999), and a loss of cortical choline acetyltransferase (ChAT) activity in key areas of the brain (Poirier et al., 1995). In terms of ε4’s relation to Aβ plaques, Aβ deposition is one of the first visible signs of the pathological process of AD, and APOE ε4 is known to inhibit the normal clearing process of Aβ plaques in the brain (Fillipinni, 2011; Mann, 1991). In addition, Buttini et al. (1999) discovered that mice harbouring at least one ε4 allele experienced increased dendritic and synaptic loss with age (i.e., increased cerebral atrophy).
1.2. APOE ε4 as an Antagonistic Pleiotropic
Curiously, on average, younger adults with at least one ε4 allele have higher IQ scores and achieve higher levels of education than their ε3 and ε2 counterparts (Hubacek et al., 2001; Yu, Lin, Chen, Hong, & Tsai, 2000). More specifically, young ε4+s (participants with at least one ε4 allele) have been found to have a domain-specific advantage on frontal tasks (decision making, prospective memory performance, and verbal fluency) (Marchant, King, Tabet & Rusted, 2010). In turn, APOE ε4 is considered antagonistic pleiotropic (i.e., a gene which has different effects on evolutionary fitness at different ages): APOE ε4 is correlated with increased cognitive performance in younger adults and decreased cognitive performance in older adults (Marchant et al.,2010).
Marchant et al. (2010) speculated that younger ε4+s’ increased cognitive performance caused accelerated cognitive aging/earlier age-related decline in older ε4+s’, which may be an additional explanation for the allele’s overrepresentation in the AD population. The antagonistic pleiotropy theory is supported by the metabolic function of APOE ε4 discussed earlier: high Aβ levels correlate with increased neuronal and synaptic activity in younger adults and reduced functional brain connectivity in healthy older adults (Wei, 2010).
1.3. The Effects of APOE ε4 on Cognition and Cognitive Aging
Today, because APOE ε4 is the second leading risk factor for developing AD, after age, there has been motivation to determine how ε4 affects specific cognitive processes affected by AD and cognitive aging (i.e., prospective memory, visual attention, and memory encoding) throughout the lifespan (Rocchi, Pellegrini, Siciliano and Murri, 2003). Determining how the ε4 allele affects cognitive processes throughout the lifespan will help researchers more accurately understand how the ε4 allele facilitates cognitive aging and the development of AD in old age.
The ε4 allele appears to impede visual attention in middle-age. Greenwood, Sunderland, Friz and Prasuraman (2000) discovered that the effect of cue-validity (i.e., the cost of having an invalid cue) on reaction time is greatest for middle-age ε4+ participants compared to ε4-s (participants with homozygous ε3). Thus, the researchers concluded that the possession of an ε4 allele in middle-age is associated with changes in attention processing similar to that of someone with mild-AD (Greenwood et al., 2000).
In terms of how ε4 affects memory (i.e., encoding) with age, Filippini (2011) discovered that aging was associated with decreased brain activity in ε4+s and increased brain activity in ε4-s. Furthermore, the over-activity of brain function initially found in young ε4+s was found to be disproportionately reduced with age even before the onset of measurable memory impairment (Filippini, 2011; Mondadori et al., 2006). A caveat in Filippini’s (2011) study is the broader age-range in the “older adults” group (50-78: 28 years) compared to the “younger adults” group (20-35: 15 years). Thus, in light of current age-related theories of compensation and dedifferentiation (to be discussed in the next section), it is possible that that the large “older adult” age-range diluted a point of increased activation in ε4+s that occurred earlier on in their lifespan (i.e., middle-age) than non-carriers (Marchant et al., 2010).
1.4. Theories of Normal Cognitive Aging
The three most prominent theories of normal cognitive aging include the Posterior-Anterior Shift in Aging model (PASA), the dedifferentiation model of cognitive aging, and the Hemispheric Asymmetry Reduction in Older Adults model (HAROLD). These all involve forms of increased brain activation with age as compensation for neuronal loss and/or an inability to recruit specialised brain regions (Cabeza, 2001; Davis, Dennis, Daselaar, Fleck & Cabeza, 2008; Park & Reuter-Lorenz, 2009).
1.4.1. The Posterior-Anterior Shift in Aging Model (PASA). The PASA model contends that cognitive aging is associated with a reduction in posterior activity (e.g., in the occipital lobe) and an increase in anterior activity (e.g., in the frontal lobe) as a compensatory mechanism of cognitive aging (Davis et al., 2008; Grady et al., 1994). Parker and Reuter-Lorenz (2009) argued that this increase in frontal activation, as a function of age, is a compensatory mechanism in response to neuronal changes caused by declining neural structure and function.
1.4.2. The Dedifferentiation Model. In opposition to PASA—and all other compensatory models— the dedifferentiation model suggests that when older adults carry out certain cognitive processes, they engage more generalized neuronal processes as a consequence of cognitive aging (Han, Bangen, & Bondi, 2009). In contrast to Han et al. (2009), Cabeza (2001) argued that compensation and dedifferentiation theories are not necessarily incompatible because dedifferentiation (combining neuronal pathways) can be thought to compensate for cognitive decline associated with cognitive aging.
1.4.3. The Hemispheric Asymmetry Reduction in Older Adults Model (HAROLD). An example of how compensation and dedifferentiation models can be compatible is found in the HAROLD model (Cabeza, 2001). This model suggests that brain activity tends to be less lateralized in older adults compared to younger adults during memory tasks. Thus, the HAROLD effect can be explained by compensation models (i.e., bihemispheric involvement may help counteract age-related neurocognitive decline) and dedifferentiation models alike (i.e., the loss of lateralisation reflects a difficulty in recruiting specialized neural mechanisms).
1.5. Prospective Memory
The current work specifically focused on how genotype and age affected prospective memory (PM)—an individual’s ability to create, rehearse and carry out an intended action—because of its relevance to everyday life and independent living (Burgess, Gonen-Yaacovi, & Volle, 2011; Burgess, Scott & Frith, 2003; Marchant et al., 2010; Rusted, Ruest & Gray 2011). In fact, a decline in PM is one of the first major complaints older adults with memory loss experience (Luo & Craik 2008). In addition, prospective memory is frontally mediated and thus useful for determining the extent to which harbouring an APOE ε4 allele accelerates age-related processes, which would be evidence for accelerated cognitive aging.
1.5.1. Prospective Memory and the Brain. Burgess et al. (2011) concluded that the rostral prefrontal cortex (rPFC), especially the lateral rostral PFC, Brodmann Area 10 (BA 10), plays a superordinate role during the many stages of PM. Thus, the researcher proposed the Gateway Hypothesis of Rostral PFC, which suggests that the main purpose of the rPFC is to control differences between attending to independent thought (inner mental life) and attending to the external world (stimulus-oriented attention). In addition, there has recently been evidence for a fronto-parietal role in PM which links into Burgess’s Gateway Hypothesis. More specifically, the fronto-parietal hypothesis suggests that parietal regions project to frontal areas to complete the more executive tasks associated with PM (Rusted et al., 2011; Simons et al., 2006).
1.5.2. Event-Based Prospective Memory (EBPM). There are two types of PM tasks that can be tested in the laboratory and that are susceptible to age-related decline: The event-based PM task (EBPM) and the time-based PM task (TBPM) (Luo & Craik, 2008; Henry, MacLeod, Phillips &, Crawford, 2004). This study focused on the EBPM task—wherein participants are asked to perform an action in response to an environmental cue. The specific EBPM task used in this study was embedded in a computer-based card-sort task (i.e., the ongoing task) first developed by Rusted, Sawyer, Jones, Trawley and Marchant (2009). Rusted et al. (2009) defined the task as attention-demanding and contended that PM detection engages general attention processes in addition to those under the control of the central executive system.
1.6. The Current Study
The current study looked at the effect of trial-type (PM trials vs. ongoing trials), genotype (ε4+s vs. ε4-s) and age (younger adults vs. middle-age adults) on coactivation in the brain (based on the results of an fMRI analysis) whilst participants completed Rusted et al.’s (2009) card-sort task in the scanner. Differences in coactivation as a function of the independent variables were measured by a Psychophysiological Interaction (PPI) Analysis. Three seed regions were chosen which have been consistently implicated in PM: BA 10, the Left Frontal Cortex and the Right Inferior Parietal Cortex (Burgess et al., 2011; Rusted et al., 2011).
The two major goals of this study were to find out (1) how differences in coactivation as a function of trial-type relate to current theories of PM and (2) how differences in coactivation as a function of genotype and genotype-specific age differences relate to current theories of cognitive aging.
In terms of cognitive aging theories, the results of our study will help us to better understand how harboring an ε4 allele links to theories of normal cognitive aging and the development of mild cognitive impairment (MCI) and/or mild AD. In contrast to Fillipinni (2011), this study focuses on middle-age (50-65) because it is concerned with neuronal changes as a function of age and genotype that specifically precede the onset of age-related memory decline. Additionally, the middle-age cohort is an under-researched group compared to older adults.
Previous research has already suggested that for certain memory tasks, middle-aged adults with an APOE ε4 allele demonstrated increased bilateral activation compared to non-carriers (Bondi et al., 2008; Bondi et al., 2006). This supports the HAROLD model and Marchant et al.’s (2010) contention that harbouring an ε4 allele is related to accelerated cognitive-aging (Cabeza, 2001).
1.6.1 Aims and Predictions. The primary aim of the current research was to look at the effect of trial-type and genotype on brain coactivity as measured by the Blood Oxygen Level Dependent (BOLD) response – an indirect measure of brain activity— in the middle-aged cohort and secondarily to include the younger cohort in order to examine genotype-specific age differences. Data from the younger adults have also been analysed elsewhere.
Based on past research, this study predicted that there would be differences in coactivation as a function of trial-type and genotype. In terms of trial-type, we predicted that BA 10 would coactivate more with other frontal regions during the PM trials compared to the ongoing trials. In support of the fronto-parietal hypothesis, we predicted that more fronto-parietal coactivation would occur during the PM trials compared to the ongoing trials. In terms of genotype, based on the PASA and HAROLD models of cognitive aging and the possibility that harbouring an ε4 allele accelerates cognitive aging, we hypothesised that there would be an increase in bilateral frontal coactivation in middle-aged ε4+s compared to middle-aged ε4-s.
In addition, this study predicted that there would be coactivation differences as a function of genotype-specific age differences. More specifically, we predicted that the increase in bilateral frontal coactivation as a function of age would be more significant in ε4+s compared to ε4-s because of ε4s role as an antagonistic pleiotropic.