Behavioural and Neural Correlates of Individual Differences in Episodic

Home , Transient global amnesia, Verbal memory

Behavioural and Neural Correlates of Individual

Differences in Episodic Memory

Daniela Jesseca Palombo

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy

Department of Psychology University of Toronto

Behavioural and Neural Correlates of Individual

Differences in Episodic Memory

Daniela Jesseca Palombo

Doctor of Philosophy

Department of Psychology University of Toronto

2013

Abstract

Episodic autobiographical memory (AM) refers to the real-life recollection of personal events that are contextually-bound to a particular time and place. Anecdotally, individuals differ widely in their ability to remember these types of experiences, yet little cognitive neuroscience research exists to support this idea. By contrast, there is a growing body of literature demonstrating that individual differences in episodic memory for laboratory experiences, intended to serve as a proxy for real life, are associated with brain-biomarkers. The present studies provide a starting point for exploring individual differences in the real-life expression of memory; which is more complex, multifaceted and has longer retention intervals, than laboratory memory (LM), thus allowing for the assessment of remote memory. While there are many factors that contribute to individual differences in episodic AM, the focus of this dissertation is on genetic influences. In particular, the KIBRA gene has been associated with episodic LM in a replicated genome- wide association study; T-carriers showing enhanced performance relative to individuals who lack this allele. The present series of studies explored the association of KIBRA with

episodic AM. An ancillary goal was to clarify the association between KIBRA and episodic LM. Accordingly, in Chapter 2, episodic LM and AM were probed in a large sample of healthy adults in relation to KIBRA genotype. Next, in a subset of participants, functional and structural magnetic resonance imaging (MRI) were used to explore

KIBRA-dependent differences in hippocampal and medial temporal lobe (MTL) blood- oxygen-level-dependent (BOLD) response (Chapter 3) and structure (Chapter 4). Akin to prior research, Chapter 2 showed that KIBRA is associated with episodic LM, albeit not on all measures. Moreover, a KIBRA association was observed for remote aspects of episodic AM on some measures. Chapter 3 showed mainly greater hippocampal and

MTL cortex BOLD response in T-carriers relative to non-carriers during the recollection of both LM and AM. Moreover, during episodic AM, T-carriers showed a distinct pattern of hippocampal-neocortical connectivity. Finally, Chapter 4 showed that T-carriers have a larger cornu ammonis and dentate gyrus relative to non-carriers, suggesting a potential neural locus for the effects of KIBRA on episodic memory.

iii

Acknowledgments

Writing a thesis is an extremely challenging process. When I doubted myself, I found strength in all the people who supported me throughout this journey. At the very top of this list was my supervisor, Dr. Brian Levine. Before I began graduate school, I knew that my future supervisor was a brilliant and accomplished scientist. But it was only during graduate school that I discovered his warmth, patience and depth as a mentor. Brian has been fundamental in shaping my academic development and instilling confidence in me. I am extremely proud of the relationship I have maintained with Brian over the last few years and I am grateful to have had an advisor who cared so much about my success (and cautioned me repeatedly to get more sleep and eat less junk food).

I would next like to thank my thesis committee members, Drs. Tomáš Paus and Morris Moscovitch, for not only inspiring me with their passion for science, but for their constant guidance, encouragement, and theoretical contributions to the work presented here. I am also grateful to my external committee members, Drs. Mary Pat McAndrews and Morgan Barense for their constructive feedback on my thesis as well as my Appraiser, Dr. Donna Rose Addis for her useful and informative comments.

Next, I would like to thank my DNA Affect and Memory Project (DAMP) collaborators, Drs. Rebecca Todd, Adam Anderson, Daniel Müller, as well as Natalie Freeman, who were instrumental in the inception of these projects. Rebecca has not only been a mentor to me but also a friend who always puts a smile on my face. Robert Amaral and Rosanna Olsen (fondly remembered as “the work triangle”) also made a significant contribution to the work presented here. I extend special thanks to Dr. Malcolm Binns (“the stats whisperer”) for statistical consulting.

I will never forget the bonds I formed in the magical land of “Levinia.” I am extremely grateful to its members, both current and alumni, and affiliated Rotman colleagues, who have assisted me in many capacities along this journey: Anjali Beharelle, Namita Kumar, Wayne Khuu, Aggie Bacopulos, Priya Kumar, Amanda Robertson, Louis Renoult, Mike Armson, Marjorie Green (“the lab mom”), Carrie Esopenko, Signy Sheldon, Leann Lapp,

Patty Hsu, Alice Kim, and Ryan Gosling (our fictional lab member). In particular, I am especially grateful to one of my first lab friends, and now one of my dearest friends, Charlene O’Connor, who has always been there for me for every academic milestone I faced, and who represents the better half of “superdesk,” an enchanted place where we spent countless hours working, snacking, and laughing.

Before I began graduate school, I had the fortunate opportunity to work in the developmental psychobiology laboratory of Drs. Alison Fleming and Gary Kraemer. Under their mentorship, I successfully complete an undergraduate thesis, where I studied maternal behaviour in rats. My time in their laboratory was short, but the lessons I learned there stood the test of time. I thank both of them for preparing me well for the (long) road ahead and, in particular, for providing me with an appreciation and understanding of the fundamentals of research in non-human species.

Next, I would like to thank my family, my parents, Paola and Tony Palombo for love, support, encouragement, and my amazing sisters, Jolene Conti and Sandra Palombo, for always being there for me and helping me get through graduate school-related stress (and for reminding me about the importance of fashion even on a graduate student budget). I thank my life-long friend, Cristina Bianco, for laminating all of my special Ph.D-related achievements on colourful construction paper (she is a grade school teacher); my dear friend, Marlene Schmidt, for teaching me to take life “one cookie at a time”-advice I took very seriously; and her husband, my dear friend, Alvaro De La Torre, for teaching me drive and determination.

Finally, I would like to thank my husband (who I married less than three weeks before defending my thesis), and the love of my life, Juan Pablo de la Torre, for patience, support, laughter, and inspiration - and for teaching me computer programming and helping me format my thesis. J.P. has always encouraged me to follow my dreams and has shown me that wherever I go, he will be there by my side. I look forward to our next chapter in life - the Boston Chapter!

Table of Contents

Acknowledgments iv

Table of Contents vi

List of Tables viii

List of Figures x

List of Acronyms xiv

Chapter 1 General Introduction 1 1.1 What is Episodic Memory? 3 1.2 Methods of Investigation 6 1.3 The Extreme Case Approach 8 1.4 Neural Correlates of Individual Differences in Episodic LM: Neuroimaging and Genetics 11 1.4.1 Structural Imaging 11 1.4.2 Functional Imaging 13 1.4.3 Genetics 14 1.5 Individual Differences in AM 16 1.5.1 Personality and AM 17 1.5.2 Summary and Future Directions 19

Chapter 2 The effects of KIBRA Polymorphism on Episodic AM 21

2 Introduction 21 2.1 Method 25 2.1.1 Participants 25 2.1.2 Genotyping 26 2.1.3 Data Analyses 31 2.2 Results 31 2.2.2 Study 1: Episodic LM 32 2.2.3 Study 2: Episodic AM 33 2.3 Discussion 35

Chapter 3 The effects of KIBRA Polymorphism on BOLD Response during Episodic AM 40

3 Introduction 40 3.1 Method 43 3.1.1 Participants 43 3.1.2 Genotyping 44 3.1.3 Image Acquisition 44 3.1.4 Scanner Task 1: Recognition LM for Faces 45 3.1.5 Scanner Task 2: Episodic AM 46 3.1.6 Data Analysis 48 3.2 Results 52 3.2.1 In Scanner Behavioral Responses 52 3.2.2 Scanner Task 1: Recognition LM for Faces 52 3.2.3 Scanner Task 2: Episodic AM 53 3.2.4 Univariate Imaging Analysis 53 3.2.5 Multivariate Imaging Analysis 55 3.3 Discussion 56

Chapter 4 The effects of KIBRA Polymorphism on Hippocampal Sub-regions using High-resolution MRI 62

4 Introduction 62 4.1 Method 63 4.1.1 Participants 63 4.1.2 Genotyping 64 4.1.3 MRI Acquisition 64 4.1.4 Segmentation 64 4.2 Results 65 4.3 Discussion 66

Chapter 5 General Discussion 70

References 78

Tables and Figures 100

vii

List of Tables

Table 1. Number of participants is shown for each KIBRA genotype group (CC and TT/TC) for each of the tasks used in Chapter 2 (IAPS; International Affective Picture System).

Table 2. Means and standard deviations are shown for each KIBRA genotype group (CC and TT/TC) for the objects task (FA; false alarms) used in Chapter 2.

Table 3. Means and standard deviations are shown for each KIBRA genotype group (CC and TT/TC) for the International Affective Picture System (IAPS) task used in Chapter 2 (FA; false alarms).

Table 4. Means and standard deviations are shown for each KIBRA genotype group (CC and TT/TC) for each of the scanner tasks used in Chapter 3 (FA; false alarms). Reaction times are displayed in milliseconds.

Table 6. Cluster maxima from the spatiotemporal Seed Partial Least Squares analysis used in Chapter 3, comparing recent and remote autobiographical memory between KIBRA genotype groups (CC and TT/TC; L; left, R; right, BA; brodmann area, Lat; laterality, Anat; anatomy). Size indicates the number of contiguous voxels in the cluster.

DG/CA2/3, and subiculum were segmented (CA; cornu ammonis, DG; dentate gyrus, Sub;

viii

Subiculum). Within MTL cortex, perirhinal cortex (PRC), entorhinal cortex (ERC), and parahippocampal cortex (PHC) were segmented.

Table 8. Mean volumes and standard error of the mean (mm3; corrected for total brain volume) are shown for KIBRA genotype groups (CC and TT/TC) for each region for Chapter 4. Volumes are shown for the left (L) and right (R) hemispheres. Within the hippocampus, CA1, DG/CA2/3, and subiculum were segmented (CA; cornu ammonis, DG; dentate gyrus, Sub; Subiculum). Within MTL cortex, perirhinal cortex (PRC), entorhinal cortex (ERC), and parahippocampal cortex (PHC) were segmented. (*p < .05; +p <.10).

List of Figures

Figure 1. A schematic depicting factors that relate to individual differences in episodic memory capacity.

Figure 2. Corrected accuracy for each KIBRA genotype group (CC and TT/TC) on the International Affective Picture System (IAPS) task for neutral, negative, and positive conditions. Error bars represent standard error of the mean (*p < .05; +p < .10).

Figure 3. Performance on the Autobiographical Interview for each KIBRA genotype group (CC and TT/TC) for the neutral and negative events. A. Number of internal and external details produced. B. Examiner-assigned episodic richness scores. Error bars represent standard error of the mean (*p < .05; +p < .10).

Figure 4. Coronal slice through the medial temporal lobes (MTL) on T1-weighted images for one representative participant depicting the MTL mask used for the functional imaging analysis for Chapter 3.

Figure 5. Results of the conjunction analysis (TT/TC versus CC) for the episodic laboratory memory task (i.e., faces task) for the contrast remember versus know for Chapter 3. BOLD response is overlaid on an average of participants’ T1-weighted images.

Figure 6. Results of the conjunction analysis (TT/TC versus CC) for the autobiographical memory task for the contrast recent versus odd/even for Chapter 3. BOLD response is overlaid on an average of participants’ T1-weighted images (MTL; medial temporal lobes).

Figure 7. Results of the conjunction analysis (CC versus TT/TC) for the autobiographical memory task for the contrast recent versus odd/even Chapter 3. BOLD response is overlaid on an average of participants’ T1-weighted images (MTL; medial temporal lobes).

Figure 8. Results of the conjunction analysis (TT/TC versus CC) for the autobiographical memory task for the contrast remote versus odd/even for Chapter 3. BOLD response is overlaid on an average of participants’ T1-weighted images (MTL; medial temporal lobes).

Figure 9. Pattern of brain activity for the spatiotemporal Task Partial Least Squares analysis in each genotype group (CC and TT/TC) for Chapter 3 for the autobiographical memory conditions (recent, remote) versus the odd/even control task. A. Colored bars represent brain scores, which represent the extent to which each experimental condition relates to the differences in hemodynamic response in each group with 95% confidence intervals plotted, which denote the standard error of the bootstrap ratios (BSR). The brain scores for each memory condition are considered statistically reliable if the error bars do not cross 0. B. BSRs for the LV represented in A, depicted on coronal images from an average of participants’ T1-weighted images. Interpretation of the relationship between the bars and the direction of change in hemodynamic response in areas reliably associated with this LV requires consideration of the brain scores; whereas positive saliences (i.e., areas of activation that are displayed with warm colors, or regions with positive BSR) indicate areas that are relatively more active in conditions with positive brain scores, negative saliences indicate areas relatively more active in conditions with negative brain scores (i.e., areas of activation that are displayed with cool colors, or regions with negative BSR).

Figure 10. Pattern of results for the non-rotated Task Partial Least Squares analysis collapsed across groups for Chapter 3 for autobiographical memory versus the odd/even control task. A. Colored bars represent brain scores, which represent the extent to which each experimental condition relates to the differences in hemodynamic response in each group with 95% confidence intervals plotted, which denote the standard error of the bootstrap ratios (BSR). The brain scores for each memory condition are considered statistically reliable if the error bars do not cross 0. B. Sagittal images of bilateral hippocampal BSRs for the LV represented in A; the peaks were selected as seeds for the spatiotemporal hippocampal connectivity analysis. Interpretation of the relationship between the bars and the direction of change in hemodynamic response in the images

requires consideration of the brain scores; positive saliences (i.e., areas of activation that are displayed with warm colors, or regions with positive BSR) indicate areas that are relatively more active in conditions with positive brain scores (i.e., recent and remote autobiographical memory; L; left, R; right).

Figure 11. Pattern of results for the spatiotemporal Seed Partial Least Squares analysis (i.e., spatiotemporal hippocampal [HC] connectivity analysis) for each genotype group (CC and TT/TC) for Chapter 3 for the autobiographical memory conditions (recent, remote) for the left and right HC. A. Colored bars represent HC-brain correlations, which represent the extent to which each experimental condition relates to the differences in hemodynamic response in each group with 95% confidence intervals plotted, which denote the standard error of the bootstrap ratios (BSR). The correlations for each memory condition are considered statistically reliable if the error bars do not cross 0. B. BSRs for the LV represented in A, depicted on coronal images from an average of participants’ T1- weighted images. Interpretation of the relationship between the bars and the direction of change in hemodynamic response in areas reliably associated with this LV requires consideration of the brain scores: negative saliences indicate areas relatively more active in conditions with negative brain scores (i.e., areas of activation that are displayed with cool colors, or regions with negative BSR).

Figure 12. Sagittal plane of T1-weighted (1 mm3) images depicting correct prescription for acquisition of high-resolution images through the medial temporal lobes for Chapter 4.

Figure 13. Coronal plane of T2-weighted (0.4 x 0.4 mm) images depicting correct prescription of high-resolution images through the medial temporal lobes, which were used in Chapter 4.

Figure 14. T2-weighted (0.4 x 0.4 mm) images depicting hippocampal subfields.

xii

were drawn where a clear ‘C-shape’ was discernible, which included all of the body but extended into the most posterior head slices (the remaining head slices and the entire tail included all subfields; see e.g., Zeineh, 2000; Olsen 2009). Demarcation varied across the long axis (see Amaral and Insausti, 1990). Anteriorly, the lateral (superior) boundary of CA1was drawn by bisecting the most lateral undulation of hippocampus (i). Moving posteriorly, the CA1 was drawn 3/4 of the way up the lateral bend of hippocampus and its medial extension bisected the DG/CA2/3 regions (iii). In the most posterior slices of the body, CA1 was drawn 3/4 of the way up the lateral bend of the hippocampus, and its medial extension was drawn in line with the medial extent of the ‘tear-drop’ (iv) shaped DG/CA2/3. Regions extending superior and medial to the CA1 were taken as

DG/CA2/3 (Zeineh, Engel, & Bookheimer, 2000). Anteriorly, the medial portion of the subiculum extended until the elbow of the isthmus (v) and in more posterior slices the medial subicular border was drawn halfway down the bend of the isthmus (vi). Perirhinal cortex (PRC) entorhinal cortex, and parahippocampal cortex (PHC) were segmented according to Insausti et al. (1998) (R; right, S; superior, CA; cornu ammonis, DG; dentate gyrus).

xiii

List of Acronyms

AFNI: Analysis of Functional NeuroImages AI: Autobiographical Interview AM: autobiographical memory AMI: Autobiographical Memory Interview ANCOVA: analysis of covariance ANOVA: analysis of variance APOE: apolipoprotein E BDNF: brain-derived neurotrophic factor BOLD: blood oxygen level dependent BSR: bootstrap ratio C: cytosine CA: cornu ammonis COMT: catechol-O-methyl transferase CR: correct rejection DG: dentate gyrus ERC: entorhinal cortex FA: false alarm fMRI: functional magnetic resonance imaging HSAM: highly superior autobiographical memory IAPS: International Affective Picture System KIBRA: Kidney and Brain LM: laboratory memory LTP: long-term potentiation LV: latent variable Met: methionine MRI: magnetic resonance imaging MTL: medial temporal lobes PKCζ: protein kinase C zeta PLS: partial least squares PTSD: post-traumatic stress disorder PFC: prefrontal cortex PRC: perirhinal cortex PHC: parahippocampal cortex ROC: receiver operating characteristics ROI: region of interest RT: reaction time SNP: single nucleotide polymorphism T: thymine Val: valine

xiv 1

Chapter 1 General Introduction

In real-life contexts, the expression of human memory entails the recovery of temporal, spatial, perceptual, and mental state details associated with experiences from one’s past. In particular, episodic autobiographical memory (AM) refers to the recollection of specific personal events that are contextually-bound to a particular time and place (Tulving, 2002). Anecdotally, people differ widely in their ability to remember these types of experiences. This notion is well captured by a song, “I Remember It Well,” in the 1958 musical, Gigi where a couple humorously recalls the details of their first date quite differently. While such individual differences may appear self- evident, there has been very little experimental focus on this topic. The sensitivity of episodic AM to many disease states, psychopathology, and normal aging gives it fundamental importance as a topic of study. Moreover, episodic AM is critical to our sense of self as well as our ability to plan and predict future outcomes (Buckner & Carroll, 2007; Sheldon, McAndrews, & Moscovitch, 2011).

A growing body of literature has demonstrated large individual differences in laboratory measures of episodic memory (hereafter referred to as episodic laboratory memory; LM), which refers to the assessment of episodic memory in the context of experimentally-generated “events.” LM is intended to serve as a proxy for real life AM within a tightly-controlled laboratory setting. Yet, humans evolved to process real-life, not laboratory experiences. While LM studies strive towards ecological validity and provide an important benchmark for understanding episodic variability, they do not directly assess the multimodal sensory nature, affective and self- referential importance, and recall latency that is inherent to the real-life expression of memory (Conway, 2001). Accordingly, episodic memory performance can be subserved by somewhat distinct biological processes depending on characteristics of the task used to quantify performance, whether LM versus AM is probed. Because of its greater complexity, I argue that episodic AM provides a qualitatively distinct and perhaps richer phenotype for understanding individual differences in memory relative to a pared down laboratory task.

Notwithstanding the many factors that likely are associated with individual differences in episodic AM capacity, the primary goal of this thesis was to explore whether episodic AM has a

2 genetic basis. This question was addressed using a combination of behavioural and endophenotypic (i.e., mid-level biological phenotype) approaches. Several studies have identified genetic variations and molecular pathways that are associated with individual differences in episodic LM in humans. These studies also demonstrate that genotype affects variability in putative brain biomarkers (e.g., see Koppel & Goldberg, 2009). However, this research has not yet been extended to the study of more naturalistic memory.

Given its complexity, episodic AM is likely a polygenetic trait; however, as a starting point, this thesis focuses on one genetic marker. While this represents a very narrow window for examining this topic, a genetic effect on episodic AM provides “proof-of-principle” for the hypothesis that individual differences in episodic AM have a neural basis. The KIBRA (Kidney and Brain) gene has emerged as an important biological candidate. KIBRA demonstrated the most robust association with episodic LM performance in a replicated genome-wide association study (Papassotiropoulos, et al., 2006). Carriers of a thymine to cytosine (T C) single nucleotide polymorphism (SNP) in this gene show altered episodic memory performance; T-carriers (i.e., TT or TC) show enhanced performance relative to C-allele homozygotes, referred to as non- carriers. Moreover, this gene is highly expressed in the hippocampus (Johannsen, Duning, Pavenstadt, Kremerskothen, & Boeckers, 2008; Papassotiropoulos, et al., 2006), a region that is critical for episodic memory (Eichenbaum, Yonelinas, & Ranganath, 2007).

While the primary goal of this thesis was to investigate the role of KIBRA in naturalistic memory by testing AM, an ancillary goal was also to clarify some inconsistencies in the literature regarding KIBRA’s role in episodic LM (see below). Accordingly, this thesis was framed around three main questions: First, do the associations with KIBRA, previously observed in the literature, extend to remote episodic AM performance? To explore this topic, episodic AM, but also episodic LM, was assessed in a large sample of participants, where T-carriers of the KIBRA gene were compared to non-carriers (Chapter 2). The second goal was to investigate the neural correlates of this association. Accordingly, functional magnetic resonance imaging (fMRI) was used to explore KIBRA-related differences in hippocampal and medial temporal lobe (MTL) blood oxygen level dependent (BOLD) response and hippocampal whole-brain connectivity in association with episodic LM and AM performance (Chapter 3). The third goal was to examine whether there is a structural basis to account for any KIBRA-related differences observed in

3 episodic memory and BOLD response. This was accomplished by using high-resolution MRI to segment the hippocampus, at the subfield level, and MTL cortex (Chapter 4).1

Although this thesis focuses primarily on genetic associations, this introductory chapter takes a broader scope to review what is known to date about individual differences in episodic memory in healthy adults, with an emphasis on biological markers that include brain structure, hemodynamic response, and genetic factors. I will also briefly discuss how variability in episodic memory likely corresponds to individual differences in other domains, such as personality and psychopathology. Where available, I review studies on individual differences as assessed by naturalistic measures such as episodic AM. However, where gaps in the literature exist, I have included studies of episodic LM. To avoid confusion, I use the general term “episodic memory” when referring to the underlying mnemonic process, but use episodic LM or episodic AM when referring to the type of assessment method. While this review is much broader than the experimental work described in subsequent chapters, some of the research described here motivated the studies composing this thesis and establishes precedence for examining individual differences in episodic AM.

1.1 What is Episodic Memory?

Before turning to individual differences, this first section will provide a brief conceptual and empirical overview of episodic memory. In the next section I review relevant methodology. The conceptualization of real world episodic remembering is rooted in dual process theories of memory (Aggleton & Brown, 1999; Eichenbaum, et al., 2007; Yonelinas, 2002), which are historically derived from studies of LM. While dual process models may slightly differ conceptually, prominent theories posit that recollection enables recovery of qualitative, contextually-specific information that supports episodic remembering. Recollection is distinct from another form of remembering, familiarity, which enables retrieval of context-free information, supporting semantic memory (Tulving, 1972, 1983). Extensive support for this

1 Chapter 4 was accepted for publication on July 9th, 2013. Palombo D.J., Amaral R.S., Olsen R.K., Müller D.J., Todd R.M., Anderson A.K., Levine B. KIBRA Polymorphism Is Associated with Individual Differences in Hippocampal Subregions: Evidence from Anatomical Segmentation using High-Resolution MRI. Journal of Neuroscience, 33, 13088-13093.

4 distinction is apparent from behavioral, functional, anatomical, and animal and human lesion studies (Aggleton & Brown, 2006; Bowles, et al., 2007; Eichenbaum, et al., 2007; Vargha- Khadem, et al., 1997; Yonelinas, 2002), including evidence that recollection, but not familiarity, is mediated by the hippocampus (for an alternative viewpoint, see e.g., Squire, Wixted, & Clark, 2007).

Likewise, AM is not unitary; episodic AM is the ability to recollect multimodal and contextually specific details of personally relevant events (e.g., “I remember my first day of university”). It is distinct from semantic AM, which pertains to memory of factual information about oneself that is context-independent (e.g., “I attended the University of Toronto for five years”). This form of AM is experienced in the third person and does not entail a sense of oneself in a specific past time and place (Tulving, 1983).

The distinction between episodic and semantic forms of AM (Eslinger, 1998; Gilboa, et al., 2006; Kapur, 1999; Levine, 2004; McGuire, Paulesu, Frackowiak, & Frith, 1996; Moscovitch, et al., 2005; Svoboda, McKinnon, & Levine, 2006) is supported by a wealth of clinical data. For example, deficits in episodic, but not semantic, AM have been observed in patients with amnesia following hippocampal and MTL cortex damage (e.g., Gilboa, et al., 2006; Rosenbaum, et al., 2008), dementia (McKinnon et al., 2008), and in psychopathology, particularly depression (Lemogne, Piolino, Jouvent, Allilaire, & Fossati, 2006; Williams, et al., 2007). It should be noted that in studies of patient populations, while multiple systems are often affected, the hippocampus and MTL appear to be a key hub structures in the network.

Insofar as patient research has demonstrated that the hippocampus and MTL cortex are necessary for intact episodic AM, functional neuroimaging research of healthy individual has provided useful information about regions that are associated with episodic AM retrieval as well as information about how regions are functionally coupled during AM performance. Like lesion studies, fMRI studies of AM have highlighted the importance of the hippocampus/MTL system, but additionally have demonstrated a network of other brain regions supporting this process. This network extends over midline anterior and posterior nodes (Cabeza & St Jacques, 2007; Maguire, 2001; Svoboda, et al., 2006) including robust patterns of hippocampal-cortical connectivity (Buckner & Carroll, 2007; Burianova, McIntosh, & Grady, 2010; Levine, et al., 2004; Soderlund, Moscovitch, Kumar, Mandic, & Levine, 2012).

Decomposing this network, hippocampal engagement is associated with the recovery of context- specific episodic details; engagement of this region increases with the amount of contextual details recovered (Moscovitch, et al., 2005; Svoboda, et al., 2006). Anteromedial regions (i.e., medial prefrontal cortex; medial PFC) are involved in self-referential processes, which are inherent to AM (Northoff, et al., 2006). Notably, medial PFC and MTL engagement, particularly hippocampal, is greater for episodic than for semantic AM tasks (Levine, et al., 2004; Maguire, 2001; Maguire & Mummery, 1999; Maguire, Vargha-Khadem, & Mishkin, 2001). Midline posterior regions, particularly the cuneus/precuneus regions, are important for the recovery of visuospatial elements of recollection (Cavanna & Trimble, 2006). Finally, activation in more lateral PFC regions is sometimes observed in AM retrieval. This is thought to reflect more domain general, top-down modulation processes, including memory search and other strategic aspects of retrieval (Svoboda, et al., 2006); lesions to the PFC produce milder memory deficits relative to those observed following MTL damage (Wheeler, Stuss, & Tulving, 1997).

Since many research studies have used LM to probe episodic memory, it is important to contrast neural correlates of episodic LM versus AM. Because lesions to the hippocampus and MTL produce deficits on both types of tasks, it is often assumed that these tasks engage the same neural processes. However, while some neuroimaging research suggests commonalities in the way that LM and AM are processed in the brain (Burianova, et al., 2010), there are also notable differences (Cabeza, et al., 2004; Gilboa, 2004; McDermott, Szpunar, & Christ, 2009). For example, in an fMRI study directly comparing LM and AM using a photo paradigm, where participants either recollected photos taken by others but were previously encoded in the laboratory versus photos that reflected their encoded experiences in the real world, Cabeza and colleagues (2004) found that while both tasks engaged the regions described above (the “AM network”), many of these regions, particularly the hippocampus/MTL, medial PFC, precuneus and cuneus, were engaged to a greater extent during AM relative to LM. These findings are likely due to the multimodal and rich sensory nature of AM relative to the simplicity of laboratory list “microevents” (McDermott, et al., 2009). Likewise, in a meta-analysis documenting both approaches, Gilboa (2004) showed that whereas LM tends to engage the lateral PFC more, AM studies tend to show engagement of the medial PFC. This difference likely has to do with the extent to which LM engages monitoring and AM engages self- referential processing (Gilboa, 2004). Finally, while the amygdala is more commonly

6 documented in studies of AM, even when emotional events are not specifically probed (Daselaar, et al., 2008; Svoboda, et al., 2006), this region is rarely observed in studies of LM, unless emotional items are used, a finding likely attributable to the greater emotional content in AM relative to LM. Together, these findings demonstrate both qualitative and quantitative differences between LM and AM. Focusing on quantitative differences, the observation that AM engages some of these key neural structures more robustly lends support to the idea proposed earlier that AM stimuli provide a more sensitive phenotype for probing individual differences in episodic memory, an idea to which I will return after reviewing some of the key methodological approaches in AM and how they differ from that of LM.

1.2 Methods of Investigation

There are various well-validated interviews designed for assessing AM (Kopelman, Wilson, & Baddeley, 1989; Levine, Svoboda, Hay, Winocur, & Moscovitch, 2002; Piolino, Desgranges, Benali, & Eustache, 2002). Here I focus on the two most popular methods.

The Autobiographical Memory Interview (AMI; Kopelman, et al., 1989) utilizes two interviews, one each to assess episodic and semantic AM across the lifespan, the latter of which concerns personal factual information (e.g., name of schools attended, addresses, etc). This method has been used in a number of patient studies (e.g., Herfurth, Kasper, Schwarz, Stefan, & Pauli, 2010; Kopelman, Stanhope, & Kingsley, 1999; Viskontas, McAndrews, & Moscovitch, 2000; Warren & Haslam, 2007). However, these interviews are not matched in terms of difficulty or other psychometric characteristics.

Levine and colleagues (2002) have developed an alternative technique, the Autobiographical Interview (AI), for assessing AM using a protocol that measures episodic and semantic components of AM within a single event. In naturalistic contexts, episodic and semantic aspects of memory are intertwined during memory retrieval (Cabeza & St Jacques, 2007; Levine, 2004; Renoult, Davidson, Palombo, Moscovitch, & Levine, 2012). For example, when one remembers a particular sailing trip in Portugal (e.g., the smell of the sea, the sun shining, etc), one might naturally recall other contextual details (e.g., how long the entire vacation lasted, how many times one has been sailing before, etc). Importantly, with the AI, the separation of episodic and

7 semantic components of memory takes place at the time of scoring, which removes the burden of doing so from the participant during retrieval and leaves the narration undisturbed.

Retrieval support (e.g., clarifying instructions, questioning about contextual details) is provided to elicit supplementary details that are not recollected during free recall, an important manipulation for reducing individual differences in non-task specific factors, such as verbosity or discourse style. This method has been useful for quantifying the extent of memory loss in patient populations (McKinnon, et al., 2008; Steinvorth, Levine, & Corkin, 2005), in healthy aging (Levine, et al., 2002), and for examining individual differences (Palombo, Williams, Abdi, & Levine, 2012).

In the laboratory, the capacity for re-experiencing an episodic event has been operationalized using various paradigms. Here, I briefly review two popular techniques for assessing recognition memory that are of particular relevance to this thesis. In the Remember/Know task (Tulving, 1985), participants are asked to classify items that they recognize as previously encountered in a study list as either recollected (“remember”) or familiar (“know”; see above description of recollection versus familiarity). This technique is widely used as it has strong face validity and is relatively simple in practice. One disadvantage of this technique is that it relies heavily on introspection, which may be difficult to implement, although this is more of an issue with patients. To circumvent this issue, some researchers have used a source memory paradigm, where recollection occurs when a participant can correctly ascribe the source of an item previously endorsed as old, for example, whether the item was heard or spoken at encoding. However, source memory judgments do not directly map onto remember judgments on the Remember/Know task as the latter may involve additional aspects of recollection that are not captured by the source attribute being probed. For example, someone may fail to recall that an item was presented verbally (i.e., a source memory failure) but may recollect thinking that the item was humorous (Levine, Freedman, Dawson, Black, & Stuss, 1999). Moreover, to the extent that source information is “unitized” with the encoded item, such as picturing a purple cat when the word cat is printed in purple, source memory can rely to a greater extent on familiarity processes (e.g., see Yonelinas, Aly, Wang, & Koen, 2010 for review).

Another popular method utilizes receiver operating characteristics (ROC) to examine the impact of response criterion on hits and false alarms (FA) to estimate the contributions of recollection

8 and familiarity to recognition memory performance (Yonelinas, 1994; 2002). More specifically, an ROC curve represents a plot of hits and FA at various levels of confidence, (ranging from “sure new” to “sure old”) in a manner similar to linear regression but with a curvilinear function. According to one view (Yonelinas, 2002, but see Wixted, 2007), recollection involves a threshold process (you either recollect something or you do not), marked by high confidence responses for old items, with very few high confidence FA responses, and is measured as the y- intercept on the curve. Familiarity involves signal detection where all items are familiar but old items are more familiar then new items. Familiarity is measured as the distance between the two distributions of familiarity values for old and new items, and the degree of curvilinearity on the ROC curve. This method is advantageous in that it does not require participants to make a subjective distinction between recollection and familiarity. However, it still requires introspection and also requires participants to use the full range of confidence responses in order to derive reliable equations; it is also sensitive to ceiling and floor effects. Despite differences, the ROC and Remember/Know methods show high convergence (Yonelinas, 2002). However, it is important to note that there is some dispute in the literature about whether recollection is a categorical process, as described above, or, whether, akin to familiarity, is a continuous process and that the methods described above (in particular, the ROC method) are hampered by a memory strength confound, where items endorsed as “recollected” are simply stronger memories. These ideas are reviewed in detail elsewhere (Wixted, 2007; Wixted, Mickes, & Squire, 2010).

1.3 The Extreme Case Approach

The studies reviewed earlier shed light on the neural systems that characterize episodic memory function across individuals; however, much less is known about differences between individuals. An important question that has received very little attention is to what extent AM capacity varies within the general population. As a first step, the understanding of individual differences in any cognitive function requires the study of those at the extreme ends of this continuum. Just as extremely high or low abilities on the spectrum of reading or mathematics has obvious practical significance, so do similarly extreme mnemonic capacities.

Recently, a handful of “highly superior autobiographical memory” (HSAM) cases were reported (Ally, Hussey, & Donahue, 2012; Leport, et al., 2012; Parker, Cahill, & McGaugh, 2006). These individuals possess an uncanny ability to recollect information from their past. Their memory is relentless and remarkably accurate. For example, when given randomly selected dates or event cues from their past, they can effortlessly recall a lot of detail about what happened, both publicly and personally, even if trivial, such as what they had for dinner. Individuals with HSAM do not practice and they do not use mnemonic aids. Likewise, they are neither savants nor calendar counters. While only a handful of cases have been reported in the literature, many more individuals with claims to this ability have since emerged (see Leport, et al., 2012).

Anatomical analyses in HSAM individuals show structural differences, relative to a comparison group, in a number of brain areas previously implicated in AM processes (Cabeza & St Jacques, 2007; Maguire, 2001; Svoboda, et al., 2006), including increased grey and white matter in the parahippocampal gyrus and lateral temporal lobes. HSAM participants also show increased volume in the uncinate fasciculus, an important white matter tract in the AM network, which mediates information flow between the ventral PFC and temporal lobes. Patient work suggests that this pathway is important for the conscious re-experiencing of the past (Levine, et al., 1998; Levine, et al., 1999; Levine, Svoboda, Turner, Mandic, & Mackey, 2009; Markowitsch, 1995). Moreover, as I will discuss later, recent research suggests that this white matter pathway is associated with individual differences in episodic processes in healthy individuals (Schott, et al., 2011).

A recent single case report of HSAM (referred to as "hyperthymnesia"; Ally, et al., 2012) also reported hypertrophy of the right amygdala (~20%) relative to controls, as well as greater amygdala-hippocampal coupling during resting state fMRI. However, this individual was also blind, which clouds the interpretation of this finding.

Together, these reports suggest that HSAM is associated with more robust communication between prefrontal and temporal lobes, although more research is needed to clarify the nature of these differences, particularly using functional imaging in a larger sample of HSAM individuals.

Turning to the other extreme, we recently investigated four healthy and high functioning individuals with a self-reported inability to vividly recollect personally experienced events,

10 corroborated by behavioural evidence of impaired AM retrieval (Palombo et al., in preparation). Their inability to re-experience events from the past is presented in the absence of detectable pathology or daily life handicap. Moreover, these individuals show complete preservation of the ability to learn and retain semantic information.

Region of interest (ROI) volumetric analyses of the whole hippocampus revealed a subtle reduction in tissue on the right side in the cases relative to comparison participants. During retrieval of episodic AM, these individuals showed reduced activation in midline AM regions as measured with fMRI, particularly in the medial PFC and precuneus (Palombo et al., in preparation). Reduced activation of the AM network in these individuals may be mediated by deficient hippocampal-neocortical connectivity (Soderlund, et al., 2012), although we did not observe reduced hippocampal activity per se.

These cases are important because they largely lay the groundwork for validating the concept of individual differences in episodic AM. Moreover, contrasting these two sets of extreme AM cases allows for speculation concerning the behavioural and neural correlates of normal variation in AM in the general population. In both groups, alterations in key AM nodes suggest that these regions may be important candidates for mediating individual differences in episodic AM. Turning to behaviour, while the impaired AM cases spend virtually none of their time ruminating about the past, the HSAM cases obsess about their past to the point of dysfunction (Parker, et al., 2006) in at least one case. On the other hand, the low AM cases self-report difficulties in other cognitive functions important for mnemonic processes, including visuospatial imagery. These elements suggest that individual differences in AM may relate to other neuropsychological patterns that are important to other cognitive processes, daily function, and quality of life. As I will show, some of these elements are consistent with a growing literature on individual differences in episodic memory that are relatively subtle.

More of these extreme cases are necessary to determine the etiology of this profile, but it is likely that genetic factors play an important role. Genetic influences on other neurocognitive skills, such as reading ability and impairment (see Raskind, Peter, Richards, Eckert, & Berninger, 2012 for review), and face recognition (Duchaine & Nakayama, 2006; McConachie, 1976) are well documented. The same likely holds true for episodic AM.

1.4 Neural Correlates of Individual Differences in Episodic LM: Neuroimaging and Genetics

At the less extreme ends of the spectrum, neuroimaging research on individual differences in memory has been largely restricted to episodic LM, and in particular, to neuropsychological measures, where rapidly accumulating evidence has demonstrated neural markers associated with variability in episodic ability. As this evidence provides a benchmark for the construct of episodic AM as a variable in individual differences, these studies will be reviewed below, after which, I turn to a small number of behavioural studies examining individual differences in episodic AM.

1.4.1 Structural Imaging

Mapping grey matter volume onto individual differences in episodic memory function has been the focus of many studies, with a particular emphasis on the hippocampus. However, these studies have produced contradictory results. In a meta-analysis of thirty-three published studies, Van Petten (2004) found a positive, albeit weak, association between hippocampal volume and episodic memory performance in older adults only, while studies of younger adults and children showed the opposite findings: lower hippocampal volume associated with greater episodic memory. Yet, more recent evidence suggests that overall hippocampal volume may not be a precise enough proxy of neuroanatomical variation. Recently, Poppenk and Moscovitch (2011) found that while absolute hippocampal volume showed no relationship with episodic memory in young adults, the posterior portion and, to a greater extent, the ratio of posterior to anterior hippocampal volume was highly predictive of episodic memory, a finding that was replicated in four separate data sets. In contrast, the anterior volume alone was negatively correlated with memory performance. Interestingly, the structure-function relationships observed were weakest when there was very little delay (30 seconds) between study and test versus longer delays of 20- 30 minutes, suggesting that consolidation processes may mediate this relationship.

Overall, these findings suggest that functional heterogeneity within the hippocampus should be considered in studies of individual differences in episodic memory. Indeed, recent advents in high-resolution neuroimaging have allowed for more fine-grained analyses of hippocampal

12 subfields in relation to individual differences. A recent high resolution fMRI study relating individual performance on a neuropsychological test (i.e., California Verbal Learning Test II; Delis, Kramer, Kaplan, & Ober, 1987) and integrity of hippocampal subfields found that while a combined cornu ammonis 3 (CA3) and dentate gyrus (DG) region was correlated with immediate recall, the CA1 region was correlated with delayed recall (Mueller, Chao, Berman, & Weiner, 2011). However, the sample in this study was a mixture of healthy individuals and those with mild cognitive impairment.

In addition to the hippocampus, the role of the MTL cortex has been the focus of some studies. One important candidate in this circuitry is the entorhinal cortex (ERC), which is the main source of information flow between the hippocampus and the neocortex (Witter & Amaral, 1991). In a study of healthy adults, shrinkage of ERC, but not the hippocampus, over a period of five years, predicted episodic memory decline (Rodrigue & Raz, 2004). Similarly, using a median split, Rosen and colleagues (2003) stratified healthy older adults into high and low episodic memory groups and compared volume in the hippocampus and ERC between the two groups. At 2-year follow-up, both of these regions were smaller in the low memory group. Interestingly, while ERC volume was the greatest predictor of immediate recall, hippocampal volume was the greatest predictor of delayed recall, suggesting these regions may be related to dissociable kinds of memory performance, although it is likely the contributions of these regions are even more nuanced.

There is also evidence that the strength of white matter connections contributes to individual variability in episodic memory performance. For example, the microstructural integrity of the fornix, a white matter tract that connects the hippocampus to the diencephalon, is related to individual differences in recollection, but not familiarity, in healthy participants (Rudebeck, et al., 2009), a finding that is consistent with clinical studies of patients with fornix lesions who show deficits in episodic memory (Gilboa, et al., 2006). Similar findings have been reported for the inferior longitudinal fasciculus (Fuentemilla, et al., 2009), a white matter path that runs through MTL structures, connecting temporal and occipital regions (Schmahmann, et al., 2007).

Schott and colleagues (2011) used combined diffusion tensor imaging and fMRI to investigate how individual differences in memory performance in healthy adults are related to connectivity between the MTL and the PFC, two important nodes in the episodic memory network. The

13 degree of white matter integrity, particularly within the uncinate faciculus, between active peaks in the PFC and these MTL sub-regions, measured at encoding, was positively correlated with recall performance. These findings are consistent with those from other studies showing that individuals who have a thicker ERC show greater PFC activation during episodic memory tasks (e.g., Braskie, Small, & Bookheimer, 2009; Rosen, et al., 2005).

1.4.2 Functional Imaging

Functional neuroimaging studies also show reliable individual differences in brain activation during episodic memory tasks (e.g., Miller, et al., 2002). A series of studies by Miller and colleagues highlight the complexity of brain-behaviour correlations; two individuals can achieve identical performance on a task, yet display quite distinct patterns of brain activity, with some individuals showing increased BOLD response in mostly PFC, and others showing predominantly parietal activation. Yet, this neural variability is highly consistent within a particular individual across testing sessions (even up to 6 months later) and mnemonic tasks, suggesting that it has a meaningful source and is not simply noise (Miller, et al., 2009; Miller & Van Horn, 2007; Miller, et al., 2002). Moreover, some of the variability observed has been linked to individual differences in encoding strategy, even in cases where performance between individuals does not differ (Kirchhoff & Buckner, 2006; Miller, Donovan, Bennett, Aminoff, & Mayer, 2012). For example, the tendency to engage in a visualizing versus verbalizing cognitive strategy when processing information has been shown to relate to individual differences in brain activity at both encoding (Kirchhoff & Buckner, 2006) and retrieval (Miller, et al., 2012).

Recent insight on individual differences in episodic memory has also been gained from measuring brain activity at rest, while the individual fixates on a crosshair or has his or her eyes closed. A specific set of interacting brain regions, labeled the “default network,” show greater BOLD response at rest, relative to task engagement (e.g., Raichle, et al., 2001). Given that these regions largely overlap with regions identified as part of the AM network, including the hippocampus and other midline regions, the brain’s default mode is thought to underlie flexible mental explorations of the self, which includes not only episodic re-experiencing but also future prospection (e.g., Buckner & Carroll, 2007). This notion has implications for understanding performance variability.

Wig and colleagues (2008) showed that the magnitude of activation in the hippocampus at rest, in comparison to task-related activity during a simple cognitive task, predicted episodic memory capacity (measured offline). Confirming these findings, three recent studies showed that interhemispheric hippocampal connectivity (Wang, Negreira, et al., 2010) and hippocampo- cortical connectivity (Tambini, Ketz, & Davachi, 2010; Wang, Laviolette, et al., 2010) at rest predicted episodic memory performance. In particular, Tambini et al. (2010) found that resting state functional connectivity is modified by recent episodic experiences. The authors compared functional coupling between the hippocampus and cortex during two resting state scans, one just prior to an associative encoding task and one immediately following the task. Hippocampal- cortical correlations were enhanced during the post-rest scan in comparison to pre-task baseline resting activity. The magnitude of these post-encoding correlations predicted individual differences in subsequent memory performance. These results suggest that resting state activity serves to strengthen memory for recent experiences and that individuals who have a greater propensity to engage in consolidation processes during quiet wakefulness are better at remembering information later. More research is needed to confirm these interpretations.

1.4.3 Genetics

Laboratory studies provide substantial evidence that memory has a genetic basis. Twin studies suggest a heritability range of .30 to .60, some of which include measures of episodic memory (Alarcon, Plomin, Fulker, Corley, & DeFries, 1998; Finkel, Pedersen, & McGue, 1995; McClearn, et al., 1997; Volk, McDermott, Roediger, & Todd, 2006). Several genetic polymorphisms that are specifically associated with episodic memory have recently been identified in humans (see Koppel & Goldberg, 2009). Below I will review three genes that have been most consistently implicated in episodic memory functioning, including a C  T single- nucleotide polymorphism (SNP) in the KIBRA gene, a Valine  Methionine (Val  Met) SNP in the brain-derived neurotrophic factor (BDNF) gene (Val66Met), and the ε4 variant of the apolipoprotein E (APOE) gene.

The KIBRA gene, which is discussed more extensively in forthcoming chapters, showed the highest association with episodic memory performance in a genome-wide association study, where carriers of a C  T substitution in this gene show enhanced episodic memory performance in comparison to non-carriers (Papassotiropoulos, et al., 2006). This effect has been

15 replicated in many studies (Almeida, et al., 2008; Bates, et al., 2009; Kauppi, Nilsson, Adolfsson, Eriksson, & Nyberg, 2011; Papassotiropoulos, et al., 2006; Preuschhof, et al., 2010; Schaper, Kolsch, Popp, Wagner, & Jessen, 2008; Vassos, et al., 2010; Yasuda, et al., 2010), although there have also been null findings in several studies (Bates, et al., 2009; Burgess, et al., 2011; Jacobsen, Picciotto, Heath, Mencl, & Gelernter, 2009; Need, et al., 2008; Sedille- Mostafaie, et al., 2012; Wersching, et al., 2011) or even opposite effects (Nacmias, et al., 2008; Wagner, et al., 2012).

While KIBRA is expressed in many brain regions, it is most highly expressed in the hippocampus and temporal lobes, particularly in the hippocampal subfields (Johannsen, et al., 2008; Papassotiropoulos, et al., 2006), which is discussed in more detail in Chapter 4. Moreover, as discussed in Chapter 3, KIBRA is associated with alterations in hippocampal BOLD response during episodic re-experiencing of laboratory material, although the direction of this allelic- specific effect has been discrepant across the two studies that have investigated it (Kauppi, et al., 2011; Papassotiropoulos, et al., 2006). Molecular research has identified KIBRA as a specific substrate of proteins implicated in long-term potentiation, which refers to the persistent enhancement of synaptic transmission associated with long-term memory storage (LTP; Kremerskothen, et al., 2003; Shema, Hazvi, Sacktor, & Dudai, 2009).

The BDNF gene is involved in cell survival and proliferation as well as hippocampal synaptic plasticity and LTP (Huang, et al., 1999; Lu, 2003). At the molecular level, the Val  Met substitution is associated with reduced secretion of the BDNF protein (Egan, et al., 2003). Studies have shown that carriers of the Met variant (versus Val homozygotes) perform more poorly on episodic memory tasks (Egan, et al., 2003; Goldberg, et al., 2008; Hariri, et al., 2003), although this finding was not replicated in a genome-wide association study (Papassotiropoulos, et al., 2006). Met carriers also show reduced neuronal integrity in the hippocampus (Egan, et al., 2003; but see Dennis, et al., 2011) and altered hippocampal activation during episodic retrieval (Dennis, et al., 2011; Hariri, et al., 2003), but see Hashimoto et al. (2008). Moreover BDNF is associated with a wide variety of functions, suggesting more global effects on brain structure and function (Toro, et al., 2009).

Finally, the ε4 allele APOE gene, which is responsible for lipoprotein metabolism, particularly cholesterol (Mahley, 1988), is associated with reduced performance on episodic memory tasks in

16 older adults (De Blasi, et al., 2009; Deary, et al., 2002; Helkala, et al., 1996; Nilsson, et al., 2006) and confers a risk factor for the development of late-onset Alzheimer’s disease (e.g., Bondi, Salmon, Galasko, Thomas, & Thal, 1999; Corder, et al., 1993), in which episodic memory loss is a defining feature. This allele is also associated with reduced hippocampal volume (Cohen, Small, Lalonde, Friz, & Sunderland, 2001; Plassman, et al., 1997), but see Adamson, et al. (2010), and reduced MTL BOLD response during mnemonic tasks (e.g., Kukolja, Thiel, Eggermann, Zerres, & Fink, 2010), but see Bassett, et al. (2006). While the APOE ε4 allele has been the focus of many association studies in older adults, few studies have investigated the role of the ε4 allele in individual differences in healthy young participants. Some effects of this gene can be attributed to the presence of preclinical Alzheimer’s disease and amyloid deposition participants in ε4 groups (Bennett, et al., 2005). Moreover, in contrast to findings from older adults, there is some data to suggest that the ε4 allele is advantageous, rather than deleterious, in younger individuals (Mondadori, et al., 2007).

The study of genetics and episodic memory is still in its infancy. There are many issues with replication, and effect sizes in these studies are small. Further, given the complexity of episodic memory, it is not surprising that this mnemonic ability has been associated with a number of other genes (e.g., HTR2A, CSTN2, etc), with additional genetic variants not yet identified.

As noted earlier, akin to many studies of individual differences in memory, genetic studies have largely been restricted to episodic LM, although two recent studies suggest a genetic basis for individual differences in traumatic aspects of AM (de Quervain, et al., 2012; de Quervain, et al., 2007). Specifically, variation in ADRA2B and PKCα has been associated with individual differences in the propensity for intrusive traumatic recollection in Rwandan genocide survivors. These findings have important implications for understanding how individual difference factors may present as risk factors for the development of psychopathology, including post-traumatic stress disorder (PTSD).

1.5 Individual Differences in AM

A growing corpus of literature suggests that there are robust biomarkers of individual differences in association with LM. Yet naturalistic mnemonic capacities, i.e., AM, may show additional variability that evade the constraints of laboratory testing, for example, due to ceiling effects in

17 neuropsychological measures. Such ceiling effects can dilute individual differences in memory (Uttl, 2005) and often conflate episodic and semantic processes (see Wheeler et al., 1997). As noted earlier, while these studies provide an important cornerstone for the study of individual differences in AM, laboratory analogues, in general, are also less perceptually complex, making it difficult to capture the full spectrum of phenomenological features of re-experiencing such as vividness, emotion, or personal relevance. Additionally, they lack the temporal extension of real- life AM, which has significant theoretical relevance to theories of remote memory and may be relevant to understanding individual variability. And, as noted, the neural correlates of laboratory and naturalistic episodic memory, while similar in some respects, are distinct in others (Cabeza, et al., 2004; Gilboa, Winocur, Grady, Hevenor, & Moscovitch, 2004; McDermott, et al., 2009). Together, these differences highlight the importance of examining individual differences in naturalistic and laboratory measures separately. There have been a limited number of studies of individual differences in episodic AM, and these have largely been restricted to examining behavioral variability. Below I review the relationship between AM, personality, and psychopathology.

1.5.1 Personality and AM

Both personality and AM are inherent to one’s self-identity and everyday functioning. Yet the relationship between personality and AM has not been well delineated, mainly because existing studies largely probe self-defining memories, which may not be representative of one’s general ability to recall AMs. Many, but not all, AM studies of individual differences typically use self- report questionnaires that probe specific events. Personality is also probed via self-report and is typically based on the five-factor model - neuroticism, extroversion, openness, agreeableness, and conscientiousness, which is a widely used taxonomy derived from extensive factor analytic research (Costa & McCrae, 1992). More specific traits underlie these broader factors, such as trait-anxiety and novelty-seeking.

Among the five personality factors, openness and its facets, i.e., the extent to which a person is intellectually curious, fantasy prone, creative, and independent-minded, is most consistently related to episodic components of AM (Rasmussen & Berntsen, 2010; Rubin Boals, & Berntsen, 2008; Rubin & Siegler, 2004). For example, Rubin and Sieger (2004) found that openness showed the strongest relationship with various component processes of AM, including sensory

18 imagery, emotion, reliving, and confidence. Similarly, one’s tendency to suppress emotion was associated with decreased phenomenological components of AM (D'Argembeau & Van der Linden, 2006). Individuals who are more open and fantasy-prone may reflect on their feelings more, which may lead to more rich reliving of and integration of memories into one’s life (Rasmussen & Berntsen, 2010).

The trait neuroticism, and some of its facets, relate to affective properties of memory, particularly negative affect. Neuroticism reflects a range of negative emotions and affective instability. Individuals that score higher in neuroticism tend to recall fewer positive memories and more negative personal memories, as well as a greater recollection of details associated with negative personal memories (Mayo, 1983; Ruiz-Caballero & Bermudez, 1995). Associations with the remaining big-five personality traits have been less consistent.

The study of stable personality traits and their relation to AM has relevance to understanding individual differences in vulnerability to psychopathology. For example, neuroticism is a risk factor for affective disorders, such as depression (Kendler, Gatz, Gardner, & Pedersen, 2006) or PTSD (McFarlane, 1989) which, in turn, are characterized by alterations in AM, including difficulty in retrieving specific AMs and retrieving overgeneralized memory instead (Williams, et al., 2007) and reduced first person AM re-experiencing (Lemogne, Piolino, Friszer, et al., 2006). These alterations in AM are also observed during remission from depression (Bergouignan, et al., 2008; Mackinger, Pachinger, Leibetseder, & Fartacek, 2000), suggesting that these may be trait, as opposed to state, phenomena. Moreover, reduced first person AM retrieval is also observed in healthy individuals who are not depressed but are high in “harm avoidance,” which is a risk factor for depression (Lemogne, et al., 2009). Hence, if there is a relationship between personality and AM, it is likely highly complex and perhaps reciprocal in nature (Conway & Pleydell-Pearce, 2000).

In addition to overgeneral AM, PTSD, in particular, is associated with additional alterations in AM, namely, the tendency to recall traumatic events as flashbacks, which are characterized by recurrent sensory images of the incident (American Psychiatric Association, 2000; van der Kolk & Fisler, 1995). Conway (2001) emphasizes the importance of sensory imagery to the recollective process of AM, as these images provide contextual information that is inherent to the specificity of an event. In healthy people, D’Argembeau & Van Linden (2006) previously

19 showed that individual differences in imagery capacity are positively associated with AM retrieval. Extrapolating from this finding, it is possible that greater imagery ability poses a risk factor for the propensity to have flashbacks of trauma, a topic for future research.

To summarize, studies involving personally relevant memory (i.e., episodic AM) have been restricted to behavioral studies, which show that individual differences in episodic AM are rooted in personality, which in turn has implications for psychopathology, where episodic AM alterations are observed in patients with depression and PTSD.

1.5.2 Summary and Future Directions

As demonstrated in this broad review, there are many factors that contribute to individual differences in episodic memory and many avenues to assess this topic (see Figure 1 for a summary). However, what is lacking is an extension of this research to real-world AM processes. In particular, future research on individual differences should examine the relationship between AM and brain biomarkers as this type of remembering may offer a rich phenotype for understanding individual differences.

Focusing on one factor - genetics - the topic under consideration in this thesis, some of the key questions that need to be answered are: Do the same genetic markers that are involved in episodic LM also translate to episodic AM? If so, can episodic AM provide us with more information about genetic influences on memory? In particular, this relates to individual differences in the temporal components of AM. Genetic linkages to “remote” aspects of memory have been documented in non-human animals (Cui, et al., 2004; Frankland, O'Brien, Ohno, Kirkwood, & Silva, 2001; Hayashi et al., 2004). Yet in humans, the genetic correlates of remote memory have not yet been investigated. Instead, most genetic studies have focused on episodic memory acquisition, with assessment of memory at short retention intervals from seconds to delays of minutes to hours. Hence, an open question - one that is addressed in this thesis - is how some of these brain biomarkers that have been identified using laboratory measures relate to more remote aspects of memory that can only be probed with AM stimuli. In the next three chapters, these questions are addressed in relation to the KIBRA gene. The strength in the approach is that it involves a comprehensive analysis of KIBRA’s effects at three different levels –behavioural, functional, and structural neuroimaging - to provide an in-depth understanding of

20 the association between this gene and individual differences in episodic memory.

Chapter 2 The effects of KIBRA Polymorphism on Episodic AM

2 Introduction

The present study sought to examine the effects of the KIBRA gene on episodic memory performance. In the first genome-wide association study of human memory, an association between KIBRA (CT substitution within the ninth intron; rs17070145) and episodic memory capacity was identified (Papassotiropoulos, et al., 2006). Carriers of at least one T allele (i.e., TT or TC), which has a ~32% frequency in Caucasians (database of single nucleotide polymorphisms), conferred an advantage in episodic remembering, relative to individuals lacking the T allele (i.e., homozygous CC genotype; non-carriers). In the first cohort, (N = 333), T- carriers showed an advantage in 5-minute and-24 hour delayed free recall of words. This finding was replicated in two independent samples (N= 424, N = 256). In the second cohort, there was a T-carrier advantage for 10-minute delayed free recall of pictures and, in the third cohort, there was a T-carrier advantage for 20-minute delayed free recall of words (Papassotiropoulos, et al., 2006). The effect sizes for these cohorts (Cohen’s d) ranged from .34-.70, indicating generally medium effect sizes.

Since the original report, this finding has subsequently been replicated in many independent studies of both young and older adults with a variety of sample sizes and measures, including free recall and recognition memory of verbal or visual stimuli (Almeida, et al., 2008; Bates, et al., 2009; Kauppi, et al., 2011; Papassotiropoulos, et al., 2006; Preuschhof, et al., 2010; Schaper, et al., 2008; Vassos, et al., 2010; Yasuda, et al., 2010), although there have also been several reports with null findings (Bates, et al., 2009; Burgess, et al., 2011; Jacobsen, et al., 2009; Need, et al., 2008; Sedille-Mostafaie, et al., 2012; Wersching, et al., 2011) or even the opposite effect (Nacmias, et al., 2008; Wagner, et al., 2012).

Findings from the initial behavioural report of three cohorts (Papassotiropoulos, et al., 2006) suggested a specific effect of KIBRA on delayed recall, but not on working memory or immediate recall. A similar finding was observed by Bates et al. (2009) in a large cohort of older adults (> 2000 participants). The initial interpretation was that KIBRA was not important for

22 processes related to early memory formation, but instead relates to the consolidation or delayed retention of information. Yet, other studies have observed performance differences at immediate but not delayed intervals (Kauppi, et al., 2011; Schaper, et al., 2008) or both retention intervals (Almeida, et al., 2008; Nacmias, et al., 2008; Vassos, et al., 2010), suggesting that the effects of KIBRA are more nuanced.

At the molecular level, KIBRA’s function is not well known. It is highly expressed in the brain, particularly within the hippocampus and temporal lobes (Johannsen, et al., 2008; Kremerskothen, et al., 2003; Papassotiropoulos, et al., 2006) and has been identified as a scaffold protein, with a somatodendritic neural distribution, enriched in the postsynaptic density (Johannsen, et al., 2008; Kremerskothen, et al., 2003). It is specifically involved in maintaining the organization of the cytoskeleton, the proteins that maintain the cell’s shape and internal motility. The cytoskeleton is important for sustaining the structural changes involved in LTP (Ruiz-Canada, et al., 2004). While the specific function of KIBRA has not been delineated, its cellular function has largely been inferred from non-human animal studies demonstrating that KIBRA knockout mice have reduced LTP, with resulting learning deficits (Makuch, et al., 2011). The same report demonstrated KIBRA’s involvement in regulating the trafficking of α-amino-3-hydroxyl-5- methyl-4-isoxazole-propionate receptor (i.e., AMPAR), the main excitatory receptor in the brain (Makuch, et al., 2011). Within cultured hippocampal cells, KIBRA co-localizes with postsynaptic proteins involved in plasticity, including dendrin (Kremerskothen, et al., 2003), synaptopodin (Duning, et al., 2008) and PKMζ, a brain specific isoform of PKCζ (Buther, Plaas, Barnekow, & Kremerskothen, 2004; Shema, et al., 2009; Yoshihama, Chida, & Ohno, 2012). PKMζ in particular has been implicated in the maintenance of the late phases of LTP (i.e., more than 6 hours) but not early or transient LTP (Pastalkova, et al., 2006; Serrano, Yao, & Sacktor, 2005). Since KIBRA is co-expressed with PKMζ within hippocampal neurons and binds to this protein in living cells (Yoshihama, et al., 2012), one hypothesis is that this gene affects memory consolidation processes via its interaction with PKMζ. Hence, it is possible that carriers of the T allele show an episodic memory advantage via enhancement of late LTP or other aspects of synaptic plasticity, thereby producing a stronger hippocampal memory trace.

Considering KIBRA phenotype in more detail, studies have been restricted to LM. As mentioned in the first chapter, while such paradigms afford experimental control and provide an important

23 benchmark for understanding genetic correlates of memory, they lack multi-faceted aspects of AM, such as self-relevance and emotionality of episodic memory as it operates in naturalistic contexts (see e.g., Cabeza & St Jacques, 2007 for discussion). In addition, given the constraints of LM (see Chapter 1), genetic studies have mainly focused on the assessment of memory bound to short retention intervals (i.e., immediate memory or delays of minutes to hours; with the exception of the trauma studies noted earlier; de Quervain, et al., 2012; de Quervain, et al., 2007) thereby leaving potential effects of KIBRA on more remote aspects of memory unaddressed. As highlighted in a review by Koppel & Goldberg (2009), assessment of memory beyond 24 hours may be even more sensitive to capturing genotype-dependent differences as they are less conflated with working memory process and less susceptible to ceiling effects.

Given these considerations, the main goal of the present chapter was to examine the role of KIBRA in naturalistic episodic processes, episodic AM. In doing so, the current experiment expands the temporal retention window for assessing memory retrieval beyond what may be feasible with a standard laboratory task. More specifically, participants were asked to recall specific events that were at least 3 months old, but no older than 3 years. While there is no agreement in the literature about what constitutes a recent versus remote AM, the aim was to choose a time period that qualified as remote AM but not so remote as to be semanticized, nor so recent as to be influenced by recency effects. Moreover, in contrast to previous research examining KIBRA effects that has been restricted to neutral stimuli, this study examined naturalistic memory, which often involves retrieval of emotional content (see Cabeza & St Jacques, 2007 for review). In fact, participants were specifically asked to provide memories that were both neutral and negative in valence. For practical reasons (i.e., scoring narratives), positive memories were not sampled. As the goal of the present study was to focus on episodic aspects of AM, the AI (Levine, et al., 2002; described in Chapter 1) was used. It was hypothesized that T- carriers would produce more episodic, but not non-episodic, details relative to non-carriers. Analyses of valence were exploratory in nature.

An ancillary goal of the present study was to clarify the role of KIBRA in episodic LM. As noted, many studies of KIBRA have been restricted to the use of neuropsychological measures of memory that often conflate episodic and semantic processes together and are susceptible to ceiling effects (Uttl, 2005). Hence, the specificity of KIBRA’s effects on episodic memory

24 remains somewhat unclear. To this end, I ventured outside of the commonly used neuropsychological measures employed in genetic studies and turned to laboratory tests of recognition memory that are also rooted in dual process theories of memory. These theories posit that when a person correctly recognizes an item as “old” they may do so because they recollect contextual details associated with the encoding context of the item (i.e., episodic memory) or they fail to recollect any details associated with the item but are able to recognize it on the basis of familiarity (which is closely related to semantic memory; Tulving, 1985; Yonelinas, 2001). As noted in Chapter 1, there is a wealth of research data supporting the notion that recollection and familiarity are distinct mnemonic processes, with only the former supported by the hippocampus (see Yonelinas, 2002 for review, although see Wixted et al., 2007 for an anternative point of view).

Participants were asked to provide confidence ratings for old and newly presented (lures) photographs of everyday objects. Quantitative measures of recollection and familiarity were derived using the ROC method described in Chapter 1 (Yonelinas, 1994; 2002). The particular task was chosen it is sensitive to individual differences in recollection,but not familiarity, in relation to MTL white matter integrity, specifically within the fornix (Rudebeck et al., 2009). Participants performed the recognition memory component of the task after a delay of 20- minutes akin to Rudebeck and colleagues’ (2009) original version of the task2. Moreover, I sought to replicate and expand upon previous work demonstrating an effect of KIBRA at that retention interval (Papassotiropoulos, et al., 2006). Although Papassotiropoulos et al. (2006) also showed an effect of KIBRA after a 5-minute delay, I did not include this retention interval because I wanted to minimize the contribution of working memory to task performance.

A second recognition memory task was administered to examine episodic LM at a delay of 1- week. Participants were asked to view images selected from the International Affective Picture System (IAPS; Lang, Bradley, & Cuthbert, 2008), which is a database of real-life photographs depicting more mundane (e.g., household furniture) or extreme conditions (e.g., a mutilated body). These images are used to elicit a range of emotions and were chosen as a more

2 An additional 1-week delay was also planned for these stimuli, but piloting showed that participants’ recognition memory scores were at floor.

25 ecologically valid stimuli set. Since stimuli varied in valence, I was able to examine the effects of KIBRA on emotionally-laden items akin to the AM paradigm. For this task, I used neutral, negative, and positive images. Procedures for obtaining measures of recollection and familiarity were similar to those of the previous task. Given that previous research has demonstrated an overall effect of KIBRA on item recognition (Preuschhof, et al., 2010), I hypothesized that T- carriers would have higher recognition memory scores on both LM tasks relative to non-carriers. Further, I expected that differences between groups would be larger for measures of recollection relative to familiarity (Preuschhof, et al., 2010). As noted before, analyses of valence were exploratory in nature.

Notwithstanding the differences between the experimental tasks, these three paradigms allowed me to explore a larger spectrum of retention intervals, ranging from 20-minutes to 3-years with 1-week in between. As noted, previous research has predominantly used very short retention intervals, typically only a few minutes, with the exception of one study that found KIBRA effects after a 24-hour delay (Papassotiropoulos, et al., 2006).

Finally, although early reports of KIBRA demonstrated no effects of this gene on working memory (Papassotiropoulos, et al., 2006), a recent meta-analysis suggests that KIBRA does play a role in working memory performance (Milnik et al., 2012). Accordingly, this type of memory was also assessed to ensure that potential KIBRA effects on episodic memory were not driven by individual differences in working memory performance.

2.1 Method

2.1.1 Participants

KIBRA genotype information was available for 282 young and healthy participants (191 female; 91 male) of Caucasian descent: 152 T-carriers (21.0 + 3.7 years old; 14.1 + 2.5 years of education and 130 non-carriers (21.4 + 4.9 years old; 14.1 + 2.5 years of education) recruited mainly from the University of Toronto, either for financial compensation of $40 or for credit in a first-year psychology course (with an additional $10 remuneration). However, due to attrition, not all participants completed each measure. As such, the sample sizes for each task are listed

26 separately and stratified by genotype in Table 13. The distribution of the T allele (~33%) was similar to that of prior studies (Almeida, et al., 2008; Papassotiropoulos, et al., 2006; Schaper, et al., 2008; Wersching, et al., 2011).

Inclusion criteria required that participants were between the ages of 18 and 40 and were free of significant head injury, stroke, epilepsy, brain surgery and learning disabilities. Although we did not assess psychopathology formally, a subset of participants self-reported a history of depression and/or anxiety (N = 45) in an online questionnaire (Palombo, et al., 2012). To avoid losing significant power, these participants were not excluded. Instead, depression/anxiety status was used as a covariate in all analyses (see below). This study was approved by the University of Toronto and Centre for Addiction and Mental Health Research Ethics Boards. All participants provided written informed consent, which was obtained online prior to coming into the laboratory. In the laboratory, participants completed a 2-3 hour battery of testing. In addition to the measures described here, participants also completed additional cognitive tasks and questionnaires that were designed to address other hypotheses; these tasks will not be discussed here.

2.1.2 Genotyping

Participants provided a saliva sample (~2 mL) in an Oragene OG-500 DNA kit (DNA Genotek, Ottawa, ON). Although I was interested in KIBRA effects, additional SNPs were genotyped to match KIBRA groups on the distribution of other SNPS that have been previously associated with episodic memory. Accordingly, five SNPs across four genes were genotyped using a TaqMan pre-designed assay: catechol-O-methyl transferase (COMT) Val158Met (rs4680); BDNF Val66Met (rs6265); KIBRA (rs17070145); and both the APOE 112 (rs429358) and 158 polymorphism (rs7412; LifeTechnologies, Burlington, ON). For each reaction, 20 ng genomic DNA was amplified and scaled to a total volume of 10 µL in an Applied Biosystems 2720 thermal cycler. Post-amplification products were analyzed on the ABI Prism 7500 Sequence Detection System using the allelic discrimination option and genotype calls were determined manually by comparison to six No Template Controls. For the APOE markers, the 112 and 158

3 At the time of this dissertation, not all of the memories for the AI were scored. The sample sizes reported here reflect only participants with scored memories.

27 genotypes were combined to determine participants’ APOE ε diplotype. Genotyping of 10% of samples were replicated for quality control with no discrepancies. Following previous studies, all individuals carrying the T-allele were combined (i.e., TT/TC versus CC).

2.1.2.1 Experimental Tasks 2.1.2.1.1 Objects Task

Stimuli were presented using E-prime Version 1.2 (Psychology Software Tools, Pittsburgh, PA). The task, hereafter referred to as the “objects task,” involved viewing grayscale photographs of everyday objects (Rudebeck, et al., 2009). At encoding, 120 photographs were presented one at a time for 10-seconds each. To facilitate maintenance of attention, participants were asked to decide whether each item could fit in a shoebox. Following a 20-minute delay period, filled with an unrelated task, participants were then presented with the encoding items intermixed with 120 lures (i.e., new items) - each presented for 15-seconds in a surprise recognition memory task. Participants were asked to judge whether they recognized the items by using a six-point confidence scale ranging from “sure new” to “sure old, where ratings 1-3 range from “sure new to “unsure new, while ratings 4-6 range from “unsure old” to “sure old.”

Recognized images, or hits, were operationalized as collapsed confidence ratings of 4-6 assigned to old items, and FA were represented by collapsed 4-6 confidence ratings assigned to new items. Memory performance was measured by corrected hits: number of FA subtracted from total hits. Hits and FA were also reported separately. Moreover, d’ scores were also calculated as a measure of sensitivity [z(H)-z(FA)]. Next, a dual process signal detection model was used (Yonelinas, 1994, 2001) in order to derive recollection and familiarity scores (see Chapter 1 for discussion), using a Microsoft solver add-in that implements a sum of squares algorithm to calculate scores (available at http://psychology.ucdavis.edu/labs/Yonelinas).

2.1.2.1.2 IAPS Task

Participants saw images from the IAPS (Lang, et al., 2008), presented using E-prime Version 1.2 (Psychology Software Tools, Pittsburgh, PA), hereafter referred to as the “IAPS task.” The photographs were selected based on categorization as negative, neutral or positive, which were based on previously normed ratings of arousal and valence. In each trial, participants saw a

28 fixation cross for 750-milliseconds, followed by each IAPS randomized image, and then by a 500-milliseconds fixation cross. Participants rated the image in terms of valence on a nine-point scale ranging from “very negative” to “very positive,” and arousal on a nine-point scale from “not at all arousing” to “very arousing.” Participants completed a total of 30 trials, 10 per valence category.

One week later, participants saw these images online in a surprise recognition memory task. The images were intermixed with a set of 30 new images (lures) matched to target old items in terms of valence and arousal. Participants were asked to judge whether they recognized the items by using a six-point confidence scale from “sure new” to “sure old.” As with the objects task, memory performance was measured by hits corrected for FA. This score was calculated for each memory type (negative, neutral, and positive). Since more than 50% of participants did not obtain FA, d’ could not be computed. Moreover, reliable estimates of recollection and familiarity could not be computed, as there were not enough trials within each type of confidence response to make resulting recollection and familiarity scores stable.

2.1.2.2 Task 2: Episodic AM 2.1.2.2.1 AI

As noted, the AI entails text-based analysis of transcribed autobiographical protocols. The AI was administered according to previously described methods (Levine et al., 2002) with some modifications. Prior to coming into the laboratory, participants were asked to select and date two events (one negative and one neutral) that were specific in terms of time and place, and took place within the last 3 years, excluding the last 3 months. Prior to testing, participants’ emailed their event titles to the experimenter. On the rare occasion that the event was not specific in time and place or involved repeated occurrences such as “my trip to Portugal” or “going to the bank,” participants were asked to send another event. Participant also provided subjective ratings for each event (“participant-assigned ratings”), which included the following: (1) how clearly can you visualize this event, ranging from “vague” to “highly vivid”; (2) please rate the degree of happiness or sadness associated with this event, ranging from “extremely sad” to “extremely happy”; (3) how personally important is this event to you now; and (4) how personally important was this event to you then, ranging from “not important” to “of great importance.”

During Free Recall of the event, participants spoke about the event extemporaneously until it was evident that they had reached a natural ending point. After an event was recalled, General Probes were used to clarify instructions and to encourage greater recall of event details. General probes were limited to non-specific statements or repetitions of the instructions. The Specific Probe phase consisted of a structured interview designed to elicit additional sensory, perceptual, and mental state details of the event.

Participants’ descriptions of the selected events were audio-recorded for later transcription and analysis. Each memory was segmented into informational bits or details. Each detail was then classified by scorers who had previously attained high reliability on a separate set of memories already scored by an expert scorer (for reliability procedures, see Levine et al., 2002) and who were blind to any information concerning the participant.

Details were classified as “internal” or episodic if they were related directly to the main event described, were specific to time and place, and conveyed a sense of episodic re-experiencing. To avoid individual differences among scorers in judgment of detail categorization, the AI scoring instructions stipulate that any detail that could reasonably be interpreted to reflect episodic re- experiencing be scored as “internal.” Otherwise, details were considered “external,” and consisted of semantic facts (factual information or extended events that did not require recollection of a specific time and place), autobiographical events tangential or unrelated to the main event, repetitions, or other metacognitive statements (“I can’t remember”) or editorializing (“It was the best of times”).

Internal details were separated into five categories: event (i.e., happenings or the unfolding of the story), place, time, perceptual, and emotion/thought. External details were separated into four categories: semantic (i.e., factual information or extended events), repetitions and other details, external event details, which were details pertaining to specific events other than the main defined internal event.

For each internal detail category, a rating of 0-3 was assigned by the examiner (“examiner- assigned ratings”) to describe the level of specificity in recall, with the exception of the episodic richness rating, corresponding to event details, scored on a scale of 0-6. An additional rating was assigned for time integration, scored on a scale from 0-3, for the degree to which events before

30 and after the main event was included. These ratings were summed to form an overall rating composite. The ratings composite provides partially overlapping information to the internal detail composite, as it takes into consideration quality of details in addition to quantity. Thus, for example, someone who provided a richly evocative event description would receive high ratings even if they did so in few words. Scores obtained from specific probe, cumulatively summed across free recall and general probe as well as cumulatively across all levels of probing, are reported.

Memories were also assigned ratings of “episodic richness” by the examiner. This reflects the extent to which the memory evokes a sense of re-experiencing. Detail categories were summed to form internal and external composites. For simplicity, only scores obtained from specific probe, cumulatively summed across all probing levels, are reported, although discrepancies in results before and after specific probing are reported.

2.1.2.3 Control Tasks 2.1.2.3.1 General Intelligence

To account for potential differences in general intelligence, which may influence cognitive performance, participants were also administered the verbal subtest of the Shipley Institute of Living Scale (Shipley, 1940), which is a paper and pencil 40-item vocabulary test that requires participants to choose the closest synonym from a choice of four listed words to that of a target word. Participants receive a point for each correct response; total scores out of 40 were computed for each participant. These scores are predictive of verbal IQ (Zachary, 1986).

2.1.2.3.2 Working Memory

To assess individual differences in working memory performance, participants performed a visuospatial K-estimate task, hereafter referred to as “K-estimate task.” Stimuli were presented using E-prime Version 1.2 (Psychology Software Tools, Pittsburgh, PA). Arrays of 1, 2, 3, 4, or 6 colored squares were presented for 150 milliseconds. Following a 1200 milliseconds delay, a single colored square was presented in one of the positions occupied by the previous stimuli. Participants were asked to determine whether it was the same color as the square that occupied that position previously. There were 30 trials of each array size for a total of 150 trials. For each

31 participant, a score was calculated based on proportions of hits and misses for each trial as a measure of visuospatial working memory performance. The 4- and 6-square array sizes were averaged together, although the results did not change when using an average of the full array.

2.1.3 Data Analyses

Potential differences between CT/TT and CC genotype groups in age, sex, education, depression, general intelligence, working memory, and the distribution of other genes were examined using Student’s t-tests and Chi Square (χ2). Remaining data were examined using analysis of variance (ANOVA). When needed, simple effects are reported for comparisons of memory conditions for analyses involving repeated measures variables. In all mixed design analyses, corrections for violations in sphericity for repeated measures were used if necessary.

I observed an effect of self-reported depression/anxiety on many of the measures (also see Palombo, et al., 2012), as well as a significant association between many of the episodic memory measures and working memory and general intelligence. Hence, self- reported depression/anxiety status, working memory performance, and general intelligence were entered as covariates in an Analysis of Covariance (ANCOVA). However, covarying did not change any of the results. Therefore, all results are reported without covariates. Probability values were one- tailed for hypothesis driven analyses and two-tailed for control measures, at a level of 0.05. Cohen’s d is reported as a measure of effect size for the statisically significant episodic memory tasks.

2.2 Results 2.2.1.1 Control Measures

Stratifying by KIBRA genotype, groups did not significantly differ in age, t (239.384) = - .88, p = .38; education, t (277) = .13 p = .90; or the distribution of sex, χ 2 = .25, p = .62; handedness, χ 2 = .11, p = .74; and self-reported depression/anxiety, χ 2 = .09, p = .77. Groups did not differ in the distribution of a subset of genes previously implicated in episodic memory processes: APOE ε4 alleles, χ 2 = .08, p = .78; BDNF Val66Met, χ 2 = .90, p = .34; and COMT Val158Met, χ 2 = .75, p = .39 (see Todd, Palombo, Levine, & Anderson, 2011 for review). Groups did not differ in

32 working memory performance (K-estimate task), t (279) = .86, p = .39; or general intelligence (Shipley Institute of Living Scale), t (279) = -.66, p = .51.

2.2.2 Study 1: Episodic LM 2.2.2.1 Objects Task

A mixed design ANOVA, with group as a between subjects factor and memory condition (hits, FA) as a within subjects factor, revealed that participants had significantly more hits compared to FA, F (1, 226) = 2890.28, p < .0001, yet groups did not significantly differ from each other, p = .96, (see Table 2) and there was no significant interaction, p = .60, indicating that both groups were equally successful at recognizing items as old. Likewise, groups did not significantly differ in terms of corrected accuracy, p = .30 (one-tailed) or d’, p = .28 (one-tailed). A direct group comparison for recollection, p = .25 (one tailed), and familiarity, p = .13, also failed to reveal any significant group differences (see Table 2).

2.2.2.1.1 IAPS Picture Task

Ratings. Two mixed effects ANOVAs, with group as a between subjects factor and valence condition (neutral, negative, positive) as within subjects factors, revealed a significant condition effect for participant ratings of valence, F (2, 550) = 579.52, p < .001, and of arousal, F (2, 550) = 1010.00, p < .001, and groups rated items similarly, with no significant group differences in valence, p = .95, and arousal ratings, p =.78, or interactions, p =.39, p =.40, (respectively). Simple effects showed that both negative, t (292) = 33.26, p < .0001, and positive items, t (292) = 24.84, p < .0001, were rated as more arousing than neutral items, and negative items were rated as more arousing than positive items, t (274) = 9.78, p < .0001, which confirmed the appropriateness of the IAPS norming. Moreover, positive items were rated as more positive than neutral, t (292) = 5.40, p < .0001, and negative items, t (292) = 36.39, p < .0001, and neutral items were rated as more positive than negative items, t (292) = 49.66, p < .0001.

Performance. A mixed design ANOVA, with group as a between subjects factor and valence condition as a within subjects factor, revealed a significant effect of valence for corrected accuracy, F (1, 518) = 14.15, p < .0001 (see Figure 2; Table 3) that was driven by a performance advantage for the negative, t (274) = 5.24, p < .0001, and neutral events, t (274) = 4.64, p <

.0001, versus the positive event. Contrary to expectation, the former conditions did not differ from each other, p = .71. There were no overall group differences, p = .13, and no significant interaction, p = .24.

A series of exploratory mixed design ANOVAs, with group as a between subjects factor and memory condition as a within subjects factor (hits, FA), were used to investigate each valence category separately. For neutral items, participants had significantly more hits compared to FA, F (1, 259) = 1834.50, p < .0001. While there was no significant main effect of group (p = .69), a significant interaction between condition and group was observed, F (1, 259) = 4.14, p < .004, While hits were higher in T-carriers relative to non-carriers, and FA were higher in non-carriers, relative to carriers, these effects were not statistically significant, p = .13; p = .14, respectively. However, a direct comparison of hits minus FAs revealed a significant difference between groups, t (259) = 2.03, p = .02, Cohen’s d = 0.25, with T-carriers demonstrating better performance relative to non-carriers. These results suggest that this performance advantage was driven by a combination of higher hits in T-carriers and higher FA in non-carriers.

For positive items, participants had significantly more hits relative to FA, F (1, 259) = 1409.05, p < .0001, yet groups did not significantly differ from each other, p = .50, and there was no significant interaction, p = .72. No effects were observed with a direct group comparison of hits, p = .76, and FAs, p = .50. Likewise, there was no group difference for corrected accuracy p = .36 (one-tailed).

Finally, for negative items, participants had significantly more hits relative to FA, F (1, 259) = 2461.64, p < .0001, although there was no significant interaction, p =.18, and no group effect, p =.64. While hits were higher in T-carriers relative to non-carriers, this effect was not significant, p = .19, and groups did not differ in FA, p = .56, although there was a marginally significant group difference in hits minus FA, t (259) = 1.34, p = .09 (one-tailed), Cohen’s d = 0.17.

2.2.3 Study 2: Episodic AM 2.2.3.1 AI

Participant-Assigned Ratings. A series of mixed design ANOVAs, with group as a between subjects factor and valence condition (negative, neutral) as a within subjects factor, were used to

34 examine participant-assigned ratings. In accordance with expectation, negative events were rated as significantly sadder than neutral events, F (1, 143) = 496.99, p < .0001, yet groups did not differ p = .40, and there was no significant interaction, p = .95. The negative event was also rated as significantly higher on the participant-assigned rating for the visualize scale, F (1, 143) = 60.47, p < .0001, relative to the neutral event. While there was no group effect, p = .43, a marginally significant interaction, F (1, 143) = 3.28, p = .072, indicated that T-carriers rated the negative, t (136) = 2.10, p = .052, (one-tailed), Cohen’s d = 0.36, but not the neutral event, p = .31(one-tailed), as significantly higher on the visualize scale relative to non-carriers. The negative event was retrospectively rated as significantly more important at the time of the event than the neutral event, F (1, 143) = 214.84, p < .0001, although groups did not significantly differ, p = .74, and there was no significant interaction, p = .37. Likewise, the negative event was rated as significantly more important, currently, relative to the neutral event, F (1, 143) = 65.43, p < .0001. While there was no significant group effect, p = .99, a significant interaction was observed between group and valence condition, F (1, 143) = 4.25, p = .04, although follow up comparisons were not significant for neither the neutral nor negative event, p = .11; p = .22, respectively.

AM Performance. As noted earlier, AI data were analyzed both before and after specific probing. As probing did not substantively alter the pattern of results, only the total scores are reported for the sake of simplicity, with exceptions where noted.

A mixed design ANOVA, with group as a between subjects factor and valence condition (negative, neutral) and detail type condition (internal, external) as within subjects factors, demonstrated that participants produced more details overall for negative relative to the neutral event, F (1, 155) = 17.89, p < .0001, and participants produced significantly more internal compared to external details, indicating that participants provided mainly episodic details during the AI, F (1, 155) = 685.37, p < .0001 (see Figure 3A). There was a marginally significant three- way interaction between valence condition, detail type condition and group, F (1, 155) = 3.13, p = .08. As demonstrated below, the interaction was present for the negative but not the neutral event and will be decomposed there.

For the neutral event, there was neither a significant effect of group, p = .63, nor an interaction between detail type condition and group, p = .67. Likewise, exploratory analyses revealed no

35 significant group differences when internal and external details were examined separately (p = .30, one-tailed; p = .83, respectively). There were no significant differences in the total examiner-assigned ratings, p = .18 (one-tailed), or examiner-assigned episodic richness ratings (p = .45, one-tailed).

On the other hand, for the negative event, there was a significant interaction between detail type condition and group, F (1, 163) = 4.98, p = .03, but no main effect of group (p = .54). Exploratory analyses revealed an increase in the number of internal details in T-carriers relative to non-carriers that did not reach significance, t (163) = 1.40, p = .08, (one-tailed), Cohen’s d = 0.22. There was also no significant difference between groups in terms of the number of external details (p = .45). There were no significant differences in examiner-assigned episodic richness ratings for the negative event (p = .46, one-tailed), while examiner-assigned episodic richness ratings (Figure 3B) showed significantly higher ratings for the negative event in T-carriers relative to non-carriers, t (159) = 2.14, p = .017 (one-tailed), Cohen’s d = 0.34.

2.3 Discussion

The main goal of this study was to investigate whether KIBRA is associated with naturalistic aspects of AM. This goal was motivated by the notion that episodic AM is more complex and multifaceted relative to LM and likely more ecologically valid; additionally, this study affords the opportunity to examine remote aspects of memory. Overall, the present study demonstrated an association between KIBRA and episodic memory, as measured by both LM and AM, but the effects were not observed uniformly across measures. A main finding from this study was that a KIBRA association was observed on episodic AM with a retention interval of 3 months, albeit only for negative episodic AM and the effect sizes were very small. Nonetheless, this finding bears importance to mapping genetic correlates of episodic AM. Additionally, this is among a small group of studies involving humans to examine genetic influences on remote memory (de Quervain, et al., 2012; de Quervain, et al., 2007) and the first report on KIBRA.

After decomposing this effect, a significant interaction between KIBRA genotype and detail type for the negative event emerged. That is, KIBRA T-carriers’ memories of negative events contained more episodic autobiographical specificity. Accordingly, examiner-assigned ratings of

36 episodic richness, as well as participant-assigned ratings of visualizing, were significantly higher in T-carriers.

Upon first impression, it was surprising that this effect was only observed for negative AM. Previous studies have found effects of KIBRA for neutral laboratory stimuli, as was the case for the IAPS task in the present study. It is possible that for the neutral AMs, participants chose overly mundane events as they were juxtaposed against a negative event in the instructions. If this is indeed the case, this could have resulted in a selection of events that were not as unique, including repeated events that made it harder to distinguish details between similar events (e.g., my first psychology lecture of the semester versus lectures that occurred other days), and introduced more noise in the neutral event condition. Anecdotally, participants had difficulty understanding the notion of a neutral event at instruction, whereas in previous studies I have not restricted participants in this manner when asking them to generate AMs. Indeed, I found that negative AMs reflected a more personally significant AM, as demonstrated by ratings of significance, which may have rendered them more sensitive to capturing genotype differences; however, it is unclear whether this was due to the issue described here or due to the valence manipulation, or a combination of both. Nonetheless, T-carriers also reported visualizing the negative event more, yet this was not significant for the neutral event. This further illustrates the possibility of a sensitivity effect, but also sheds light on a possible mechanism for the effects of KIBRA on episodic memory, although I did not query other senses. The reverse may also be true: greater visualization is the result of greater detail availability. It would be interesting to examine whether the effects of KIBRA on episodic memory are mediated by differences in connectivity between the hippocampus and areas that are important for visual imagery -a question I will address in the next chapter.

An ancillary goal was to clarify the association of KIBRA genotype with advanced measures of LM, as prior studies used only neuropsychological measures. Additionally, I examined LM at a longer retention interval than typically has been probed in the literature. Akin to previous work, I found evidence that carriers of the T-allele confer a mnemonic advantage over non-carriers for LM. Specifically, on the IAPS task, which was performed 1-week post-encoding, KIBRA T-allele carriers showed better recognition memory performance relative to non-carriers, particularly for neutral items. This finding corroborates a previous report showing that KIBRA is associated with

37 recognition of single words at 90-second delay (Preuschhof, et al., 2010), but extends this work to a longer retention interval. Interestingly, in that study, the effects were larger for associative recognition relative to single item recognition, although the additional degree of modulation by associative recognition was only marginal (Preuschhof, et al., 2010). Unfortunately, I was unable to specifically determine whether the influences were more specific to recollection, as oppose to familiarity, because I did not have enough items across confidence levels to reliably compute estimates of these processes. Hence, this data cannot specifically address whether this effect was driven by episodic processes exclusively.

In light of the AM findings, it was also surprising that the effects of KIBRA were less robust for negative items. The greater sensitivity of neutral relative to negative IAPS items may be attributable to ceiling effects in the latter. While performance on negative items was higher relative to the neutral items, this difference was not statistically significant. By contrast, on the AI, there was a more robust memory enhancement effect for negative items. Taken together, this is suggestive of different emotional modulation effects in the two tasks, as expected given the known differences between laboratory and real-life memory tasks (see Chapter 1 for discussion). Indeed, the present data failed to demonstrate statistically significant correlations between internal details on the AI and either the IAPS task or the objects task in either of the genotype groups.

Contrary to expectation, I did not observe any effects of KIBRA genotype for the objects task. The task was selected because it demonstrated sensitivity to individual differences in a previous report (Rudebeck et al., 2009). Specifically, Rudebeck and colleagues (2009) found that individual differences in recollection, but not familiarity, predicted white matter integrity of the fornix. Indeed, our recollection and familiarity values were in the same range as those reported by the authors.

Comparing the two LM tasks, recognition memory on the objects task was significantly correlated with recognition accuracy on the IAPS task at all valence types; however, the correlations were modest, suggesting that there were important differences between the tasks (discussed more below). For the objects task, participant had moderate hit rates (approximately 70-75%) but high FA rates (approximately 20-25%; see Table 2). Hence, it is possible that the task was too difficult, creating a floor effect. In particular, this task requires participants to study

38 a large number of items and then discriminate old items from an equally large number of lures. Moreover, the task appears towards the end of a 3-hour cognitive test battery, which may have reduced attentional capacity. Given that genetic effects are subtle, with very modest effect sizes, introducing this type of noise into a paradigm may have muted potential effects of KIBRA. As noted in the introduction, the effect sizes of individual genetic loci are small, accounting for only approximately 1% of the variance in performance (Ehret, 2010). By contrast, the IAPS task appeared earlier on in the battery of testing and there were significantly fewer trials at encoding (30 items); participants performed better overall with hits between 75-85% and FA rates less than 20%.

However, there were other important differences between the tasks that could have contributed to the results. The IAPS task assessed memory at 1-week, while the objects task employed a 20- minute delay. While the delay interval could account for the differences in genotype groups, as noted, other studies have demonstrated effects of KIBRA at shorter delays. However, it is possible that increasing the delay reduced the influence of working memory on performance, thereby leading to more pronounced differences, although co-varying working memory performance in our analyses did not change the results.

Interestingly, in T-carriers, recognition memory on the objects task was positively correlated with recognition memory on the IAPS task across all valence types (all ps < .05; rs = .19-.32), while in non-carriers a correlation with the objects task was only demonstrated for the negative IAPS items (p < .05, r = .22) suggesting a greater similarity in T-carriers in the approach to the two tasks relative to non-carriers.

Finally, the tasks also differed in terms of the type of stimuli used, the instructions at encoding, and the valence of items. Hence, an additional possibility is that the IAPS pictures were more ecologically valid as they depicted real-life situations as opposed to single objects. Indeed, relational or associative tasks place greater demands on the hippocampus relative to single items (Davachi & Wagner, 2002). Although the IAPS task was not an associative task per se, because it depicted scenarios or scenes, it is possible that it placed heavier demands on binding of items and their context, thereby recruiting hippocampus-based processes to a greater extent (Diana, Yonelinas, & Ranganath, 2007; Olsen, Moses, Riggs, & Ryan, 2012). Despite these nuances, the effects of KIBRA genotype on recognition of IAPS pictures represent the longest reported delay

39 period in the literature on KIBRA and LM. An interesting question for future research is to explore how KIBRA interacts with the normal forgetting curve by employing one paradigm with several retention intervals.

Finally, there were no significant effects of KIBRA on working memory and general intelligence. Earlier reports suggested that KIBRA did not affect immediate memory or short term processes (Bates, et al., 2009; Papassotiropoulos, et al., 2006), although such conclusions have been challenged in more recent studies of working memory (e.g., Milnik et al., 2012). One reason for this discrepancy in the literature may relate to the type of working memory task employed. For instance, recent work suggests that the hippocampus and MTL support short term memory if there are associative demands (Olsen, et al., 2012). Our working memory task was not designed to have a significant associative component and cannot speak to this issue directly. To assess the generalization of KIBRA mnemonic effects to other associative tasks, future studies should manipulate this element of working memory.

Recent research has suggested that sex modifies the relationship between KIBRA and cognition with larger genotype effects in females, although these findings were for executive functioning (Wersching, et al., 2011). As this sample had an approximately 2:1 female ratio, reflecting the ratio in the psychology classes from which we drew participants, the present study was unable to address sex effects. However, these considerations motivated our decision to include only females in the subsequent chapter (Witte, et al., 2012).

In summary, the present study showed an association with KIBRA genotype and episodic memory, although the effects were nuanced. These effects likely represent a combination of the complexity of gene cognition interactions as well as limitations of the measures used in the present study. Despite this, this chapter demonstrates, for the first time, an effect of KIBRA on naturalistic memory that is at least as large as that for LM. These results have implications for understanding the role of the genotype in memory. The next step is to identify the functional neuroanatomy of these effects, which is the goal of the subsequent chapter.

Chapter 3 The effects of KIBRA Polymorphism on BOLD Response during Episodic AM

3 Introduction

The purpose of this chapter was to examine the relationship between KIBRA genotype and hippocampal/MTL BOLD response during episodic re-experiencing. Consistent with behavioral evidence, two recent fMRI studies of KIBRA demonstrate altered MTL/hippocampal activity in association with the CT polymorphism, yet conflicting results were observed. In the first study, Papassotiropoulos et al. (2006) selected a sub-population from their Swiss behavioral cohort (age range 19-25 years; N = 30) for fMRI. Genotype groups were matched on laboratory recall performance prior to scanning, yielding 15 T-carriers and 15 non-carriers. Matching participants by performance was intended to minimize performance-related activation differences. The disadvantage of this approach is that it also may minimize neural changes associated with genotype related performance differences (see Rasch et al., 2010 for discussion).

Papassotiropoulos et al. (2006) found that non-carriers of the T allele showed greater BOLD response in the MTL and hippocampus relative to T-carriers during retrieval of professions paired with faces at encoding, while no genotype differences in BOLD response were observed during a working memory task. Since performance was matched between genotype groups, the authors interpreted greater brain activity in non-carriers to be compensatory - more neural resources were necessary in non-carriers to achieve the same level of performance of T-carriers. A similar interpretation of inefficient hippocampal/MTL processing has been proposed in association with other genes (e.g., Buchmann, et al., 2008; Dennis, et al., 2011; Huentelman, et al., 2007).

However, a recent study (Kauppi, et al., 2011) of middle aged to older adults (age range 55-60) failed to replicate this finding using a paired-associate task involving a forced-choice retrieval procedure. Participants had to select the letter corresponding to the first letter of a name previously associated with a given face at encoding. Instead, the authors found enhanced hippocampal BOLD response in T-carriers relative to non-carriers, both before (N = 83) and after

41 post-hoc laboratory memory performance matching (N = 64). By contrast, no significant group differences in hippocampal BOLD response were observed in an even older cohort (age range 65-75; N = 113).

Contrasting the two studies, Kauppi et al. (2011) used a larger and older sample of a different ethnicity, relative to Papassotiropoulos et al. (2006). Moreover, as noted by Kauppi and colleagues, their associative memory task was likely more challenging, as the previous study involved repeating pairs across runs. Task difficulty has been proposed to modulate whether over- versus under- neural recruitment is observed in aging, such that an age-related compensatory response typically occurs when processing demands are low; however when a task is too difficult, the opposite pattern may be observed (Prvulovic, Van de Ven, Sack, Maurer, & Linden, 2005), although these interpretations are typically in reference to PFC activation in aging. With respect to KIBRA, with so many distinct characteristics of the two studies, it is unclear which of these can account for the discrepancy; hence the impact of KIBRA on MTL/hippocampal function during fMRI is not yet understood.

Despite this, both the foregoing neuroimaging studies of KIBRA employed laboratory analogues of episodic memory, as opposed to real world AM. Yet, the study by Papassotiropolous and collegues (2006) involved repetition of stimuli. Strictly speaking, episodic memory involves single events. The repetition of stimuli employed in their task may have also recruited non- episodic processes. Kauppi et al. (2011) employed a task with forced-choice recognition, where participants were primed with the first letter of the name, which also could be accomplished using non-episodic processes.

Given these considerations, the main goal of the present study was to examine the relationship between KIBRA genotype and hippocampal/MTL BOLD response during episodic retrieval. Participants were asked to re-experience episodes from their personal past, thereby extending this literature to episodic AM. However, given the discrepancy discussed above, I also sought to align the present study with foregoing imaging studies of KIBRA by using laboratory tasks with comparable stimuli to those used previously. Hence in the first experiment, I compared BOLD response, using fMRI, in young (see below) T-carriers and non-carriers during a LM measure of episodic memory retrieval. There were 15 T-carriers and 15 non-carriers, the age range and sample size being comparable to those used by Papassotiropolous and colleagues. Akin to the

42 prior fMRI reports, groups were performance-matched to rule out genotype-independent performance-related differences (see below).

For scanning, there were two tasks, a LM task and an AM task. As in both prior reports, the LM task entailed a recognition memory paradigm that uses faces, which previously has been shown to activate the MTL/hippocampus (McCormick, Moscovitch, Protzner, Huber, & McAndrews, 2010). As noted before, under the assumption that recognition memory tasks can be accomplished using two independent processes, recollection and familiarity, participants were asked to make remember and know judgments for previously recognized items in order to specifically isolate the contribution of episodic from non-episodic processes, respectively, for correctly recognized items (Tulving, 1985; see Chapter 1 and 2). I hypothesized that KIBRA would modulate hippocampal/MTL BOLD response when directly contrasting recollection versus familiarity, but, given the discrepancy between the prior studies, I did not have any specific predictions about the direction of the effect.

Beyond the hippocampus/MTL, numerous neuroimaging studies have identified a distributed network in which midline frontal, and posterior regions are also engaged in association with recollection of contextually specific episodic AM details (Cabeza & St Jacques, 2007; Maguire, 2001; Svoboda, et al., 2006). Some of these regions are also active for LM tasks, but, as noted in Chapter 1, some of the neural correlates of LM versus AM are partially separable (see Gilboa, 2004 for a more detailed discussion; McDermott, et al., 2009). In the second task, hippocampal and MTL activity was examined in relation to retrieval of naturalistic episodes. The same participants were asked to re-experience personally relevant AM events during scanning using a method that reliably activates the hippocampus/MTL (Soderlund, et al., 2012). Participants were asked to re-experience recent (i.e., within the last 6 months) and remote (~ 10 years old) events. Although these time periods represent a departure from Chapter 2 where I sampled memories that were neither recent nor remote, these time-periods are commonly used in the neuroimaging literature, where different patterns associated with recent and remote AM have been identified (e.g., Soderlund, et al., 2012).

Given the widely distributed nature of brain activity during episodic remembering, an ancillary goal of the present study was to employ functional connectivity analyses to investigate potential genotype-dependent differences in hippocampal-whole brain connectivity associated within

43 episodic memory, which may provide a richer understanding of the specific effects of KIBRA genotype on neural activity above and beyond what can be gleaned from region-specific analyses. Assessment of modulation of the episodic memory network by between or within- subject comparisons requires a multivariate approach with an emphasis on large-scale network interactions. For example, in a prior study using the same AM paradigm, the hippocampus was co-activated with regions that are part of the AM network across time periods. More recent memories showed a pattern of positive co-activation across the entire epoch, while for remote memories there was an initial negative hippocampal co-activation, followed then by positive co- activation with other brain regions, perhaps reflecting more sluggish memory recovery for older memories (Soderlund, et al., 2012). Hence, this analysis will allow me to examine potential effects of KIBRA on distinct temporal stages and time courses of episodic AM retrieval.

On the one hand, since the hippocampus is involved in episodic memory irrespective of remoteness, I would expect genotype-dependent differences for both recent and remote memories (Gilboa, et al., 2004; Soderlund, et al., 2012). On the other hand, since I did not control for vividness in the present study, which is known to modulate BOLD response in the hippocampus (e.g., Gilboa, et al., 2004; Sheldon & Levine, 2013; n.b. these studies also did not experimentally manipulate vividness), it is also possible that the remote time period represents less episodically-rich memories, which would implicate greater involvement of the hippocampus in recent memories and, accordingly, greater sensitivity to associations with KIBRA genotype.

3.1 Method

3.1.1 Participants

Thirty-six young and healthy Caucasian participants (all female) were selected from a pool of individuals recruited for the behavioral study described in Chapter 2. Six participants were excluded from fMRI, either due to technical issues or anatomical abnormalities. Hence, a remaining 30 participants were included in the study and stratified according to KIBRA genotype: [15 T-carriers (22.3 + 3.8 years old; 15.7 + 3.3 years of education and 15 non-carriers (20.4 + 2.9 years old; 14.1 + 1.8 years of education)]. TC and TT genotypes were grouped together for all analyses according to the previous imaging studies of KIBRA (Kauppi, et al., 2011; Papassotiropoulos, et al., 2006).

Prior to functional imaging, genotype groups were carefully matched (one-to-one matching between genotype groups with a standard deviation of .4) on recognition memory scores from the objects task that was administered in a prior session (see Chapter 2), t28 = .41, p = .68. In retrospect, it would have been optimal to match participants on the IAPS task where I observed group differences, but because I ran these studies in parallel, the results of the behavioral study were unknown when the fMRI study began. While it is acknowledged that performance differences attributable to genotype are unlikely to be detected in this sample size, the present results may nonetheless help to explain findings from other larger studies where performance effects are observed. Genotype groups did not differ in the distribution of genes previously implicated in episodic memory processes: COMT Val158Met, χ 2 = .68, p = .41, APOE ε4 alleles, χ 2 = .68, p = .41, BDNF Val66Met χ 2 = .00, p = 1.00, genotype polymorphisms (see Todd, et al., 2011 for review). Groups did not differ in age, t (28) = 1.56, p = .13, and education, t (28) = 1.72, p = .100. This study was approved by the Baycrest Research Ethics Board. All participants provided written informed consent.

3.1.2 Genotyping

Procedures for genotyping participants are described in Chapter 2.

3.1.3 Image Acquisition

Brain images were acquired on a whole body 3T (Siemens Magnetom Trio Tim, Numaris/4Syngo MR B13; Siemens, Germany) with a standard birdcage 32-channel "matrix" phased-array head coil. Participants were provided with headphones and earplugs to reduce noise from the scanner. To minimize head movement, participant’s heads were firmly placed in a vacuum pillow. Sensors were placed on participants' left big toe to monitor their heart rate and respiration. Functional imaging was performed to measure BOLD (Ogawa, Lee, Kay, & Tank, 1990), using T2*-weighted single-shot echo planar imaging (EPI) k-space trajectories optimized for sensitivity to the BOLD effect (TE/TR/flip angle = 30ms/2000 ms/70 degrees, 40 slices in axial oblique orientation, voxel size 3.1 x 3.1 x 3.5 mm, slice spacing = 0, 64 x 64 acquisition matrix, FOV=200 mm). Prior to acquisition of functional images, a volumetric anatomical MRI was performed, using a 3D T1-weighted MP-Rage sequence (TE/TR = 2.63 ms/2000 ms, 176 oblique axial slices, 256 x 192 acquisition matrix, voxel size = 1 mm3, FOV = 256 mm).

For later segmentation (see Chapter 4), a high resolution T2-weighted MRI was acquired in the coronal plane, perpendicular to the long axis of the hippocampus (TE/TR = 68 ms/3000, 22-28 oblique slices depending on individual head size, 512 x 512 acquisition matrix, voxel size of 0.43 x 0.43 x 3 mm, no skip, FOV=220mm). Appropriate prescription of slices was determined anteriorly by placing the first slice just prior to the appearance of the collateral sulcus; posteriorly, the last slice was placed just prior to the appearance of the full superior-inferior extent of the ventricles in the coronal plane, such that the cerebrospinal fluid just lateral to the hippocampus sweeps up and joins the cerebrospinal fluid in the lateral ventricles (see Figure 13). While the procedure for segmentation is described in Chapter 4, the segmented ROIs were also collapsed together and used to create a structural group mask for examining BOLD activity within the hippocampus and adjacent MTL cortices (discussed below). However, due to issues with acquisition, this was based only on a subset of participants whom were randomly selected from those for whom we had acquired viable high-resolution T2-weighted images (N = 10, 5 in each group).

During functional imaging, visual stimuli were presented using an LCD projector (NEC Model MT1065) with a 2.75-5” zoom lens (Navitar, Inc.) and an fMRI-compatible display screen built within the laboratory, mounted at the rear of the magnet bore. Button press responses were recorded using a Fiber-Optic Response Pad System placed in the right hand (Current Designs Inc., 4 buttons available).

3.1.4 Scanner Task 1: Recognition LM for Faces

Stimuli were presented using E-prime Version 1.2 (Psychology Software Tools, Pittsburgh, PA). Ninety-two black and white photographs of faces were used (age range: 25-35), with each photograph presented alone. The stimuli and methodology were adopted, with slight modification from another report (McCormick, et al., 2010). Faces were presented in frontal view with hair cropped out of the image and faces were presented on a white background. Prior to the scan, participants were familiarized with the task and completed a practice trial with an additional 15 faces presented at study and 25 faces (15 old, 10 new) presented at test. Of the 92 faces used in the experiment, 60 faces were used as targets, while the remaining 32 faces were used as lures. During scanning, stimuli were presented in two encoding/retrieval runs, with half the stimuli used in each run and presented in a randomized fashion. For both study and test, the

46 interstimulus intervals, during which a black crosshair appeared on a white background, were jittered randomly between 2-, 4-, and 6-seconds (average of 4-seconds), to avoid anticipation effects and obtain a better estimate of the hemodynamic response function.

In each study phase, 30 faces were presented and participants were asked to decide whether the person depicted in each of the photographs was likely to be a “sporty-type,” “party-goer,” “homebody,” or “intellectual.” This was done to facilitate contextual/relational encoding of the faces as MTL/hippocampal activity is greater for relational memory relative to memory for single items (e.g., see Mayes, Montaldi, & Migo, 2007). Each face was presented for 4-seconds, followed by a 2-seconds response screen, where participants indicated the personality type via button press, using fMRI compatible response boxes. Each response choice was displayed on the screen with the corresponding button choice.

Following the study phase (~ 2-minutes), the recognition memory phase began, consisting of 30 old faces and 16 new faces. Participants saw each photograph for 4-seconds, followed by a response screen during which they had to indicate, via button press, whether the face presented was “old” or “new”. For each old response, participants were then asked to determine whether they remembered anything specific about the encoding context, including, but not restricted to, the personality assigned to the person, or any mental/sensorial association that accompanied the initial presentation of the face. This type of response was deemed a “remember” response. If participants were not able to recall any specific information about the face, but recognized it as old through a feeling of familiarity or “knowing,” they were asked to instead select “know” (Tulving, 1985). Participants were given 3-seconds to make either response type, using fMRI compatible response boxes. Scanning did not begin until participants were clear on the distinction between remembering and knowing, and practiced making each type of response. All participants demonstrated an understanding of this distinction by describing the difference between remembering and knowing to the experimenter in their own words.

3.1.5 Scanner Task 2: Episodic AM

Participants were asked to retrieve AMs in the scanner, in the same scanning session and always following the foregoing laboratory task. Counterbalancing of the two tasks was avoided due to possible residual effects of the AM task on the LM task. Stimuli were presented using E-prime

Version 1.2 (Psychology Software Tools, Pittsburgh, PA). A few days prior to scanning, participants were asked to generate, date, and title 20 events that were specific in time and place (e.g., “My first time sailing in Portugal”) from two time periods: recent (i.e., within the last 6 months) and remote (i.e., 5-15 years ago), making a total of 10 memories per time period. Participants emailed their event choices to the experimenter, which was reviewed for specificity. In the rare event that more general events (e.g., “My vacation in Portugal”) or repeated events (e.g., “going to the bank”) were chosen, participants were asked to generate new specific events. Similar procedures have been used previously (Soderlund, et al., 2012). Participants also performed an odd/even number judgment (hereafter referred to as “odd/even”) as a control condition as this has been shown to effectively reduce MTL activation (Stark & Squire, 2001).

These conditions were randomly distributed across 2 experimental runs, such that approximately 10 AM trials, randomized between recent or remote, and 5 odd/even trials, were presented within each run. Participants were familiarized with the task and completed a practice trial prior to scanning. All stimuli were presented in black text on a white background. Participants saw an AM event title for 16 seconds, preceded by a 2-seconds fixation and a 4-seconds reminder cue (i.e., “Autobiographical Memory”). During presentation of the event title, participants were asked to re-experience the event as vividly as possible by trying to recall thoughts, feelings, and visual images from the event. After 16-seconds, participants were given 6-seconds to rate the amount of re-experiencing, using a scale ranging from 1 to 8, using fMRI compatible response boxes.

For odd/even, participants saw eight numbers, one at a time for 1,900-milliseconds, with a 100- milliseconds interstimulus interval, preceded by a 2-seconds fixation and a 4-second cue (i.e., “Odd/Even?”). Participants were instructed to determine whether the number presented was odd or even in the absence of an overt response. Following presentation of the last number, participants rated the amount of re-experiencing of any AM, as described above. This rating served as a manipulation check to determine the degree of extraneous mnemonic contamination during odd/even judgments.

3.1.6 Data Analysis

3.1.6.1 Behavioral Data

Chi-square analysis was used to test for potential differences in the distribution of sex and other genes. All behavioral responses made in the scanner were analyzed with mixed-design ANOVAs or t-tests. When needed, simple effects are reported for comparisons of memory conditions for analyses involving repeated measures variables. In all mixed design analyses, corrections for violations in sphericity for repeated measures were used if necessary. Since groups were performance matched prior to scanning, these analyses served as a manipulation check as we did not expect genotype-dependent behavioural differences.

For the episodic laboratory task, genotype groups were compared on the proportion of hits and FA using a mixed design ANOVA, to ensure that groups were performing the task correctly and at the same level, as the performance matching procedure requires. To examine sensitivity, d’ was calculated. Since the main contrast of interest was to compare genotype groups on episodic memory performance, hits were examined based on the proportion of old items that were subsequently categorized as recollected (remember response) and familiar (know response), each corrected for FA within those response categories and compared groups using a mixed design ANOVA. Finally, these analyses were repeated for reaction time (RT). Separate d’ indices could not be computed for recollection versus familiarity because more than 50% of participants did not FA, particularly for remember responses.

For the AM task, genotype groups were compared on ratings generated in the scanner (recent/remote and odd/even) using a mixed design ANOVA and follow-up post hoc analyses to compare rating types across AM conditions.

3.1.6.2 Functional Imaging Preprocessing and Analyses

Analysis of Functional NeuroImages software (AFNI; Cox & Hyde, 1997) was used to pre- process the neuroimaging data for both tasks. Time series were first corrected for physiological motion (i.e., heart rate and respiration) via linear filtering. Next, data were corrected for slice- timing effects by aligning each slice to the time offset of the first slice using the AFNI program ‘3dTshift’ using a Fourier method. The data were then spatially co-registered to correct for head

49 motion, using the 3dvolreg program in AFNI, which aligns the data to a reference scan chosen from within one of the runs. A 3-D Fourier transform interpolation was used.

3.1.6.2.1 Univariate

Univariate analyses were used to examine hypothesis-driven hippocampal/MTL BOLD response, whereas multivariate analyses were used to examine whole brain task effects and connectivity. For univariate analyses, the data were normalized temporally, and then deconvolved (AFNI 3dDeconvolve plugin). T-statistics were calculated for each participant (at the individual level), contrasting each experimental condition to a baseline, which consisted of all non-event time points (e.g., fixation). Next, contrasts were set up for each individual (e.g., recent vs. odd/even). These activation maps were then extracted for each individual and spatially transformed to Talairach space (Cox & Hyde, 1997; Talairach & Tournoux, 1988) and spatially smoothed using a Gaussian filter with a full width at half maximum value of 6.0 mm to minimize individual anatomical variability. All data were resampled to voxels of 2 mm3. Finally, group analyses were performed using 3dANOVA3, which consisted of a voxelwise, mixed effects (conditions ﬁxed, participants random) ANOVA, treating experimental conditions as a within subjects factor.

Given the a priori interest in examining potential genotype-dependent group differences in the hippocampus and MTL, this analysis was restricted to this ROI. Specifically, a structural mask was applied to the functional images at the group level, which encompassed the hippocampus and parahippocampal gyrus (i.e., ERC, perirhinal; PRC and parahippocampal gyrus; PHC). The mask was derived from the high-resolution T2-weighed images; manually segmentation of the MTL was conducted in participants’ native space using FSLview (v3.0.2). Segmentation was performed blind to group status and based on established procedures (e.g., Olsen, et al., 2009), which are described in more detail in Chapter 4. Although segmentation was done at the level of hippocampal subfields for the structural analyses presented in Chapter 4, for functional imaging analyses, I averaged all the ROI together to form a unified MTL mask. To apply the ROIs to the functional imaging data, a series of steps were performed: First, for each participant, the T2- weighted image was resampled (2 mm3) and aligned to their T1-weighted image; these co- registration parameters were then applied to each ROI. Next, the normalization parameters used for Talairaching each participants’ T1-weighted image during preprocessing were applied to the ROI. Finally, the ROIs were averaged across participants. As noted earlier, since high-resolution

50 images were not available for all participants who were part of the functional imaging study, the mask was based on an average of only subset of participants (N=10; 5 T-carriers; see Figure 4 for an example participants’ mask). Although this presents a limitation, this mask was found to be a much better fit to our anatomical data relative to the hippocampal and parahippocampal gyrus masks provided in AFNI.

In order to ensure that group differences were specifically associated with task-related effects, a conjunction analysis approach was used, following MTL masking. First, within each group, the main contrasts of interest were examined, using a p < .05, uncorrected for multiple comparisons within the MTL ROI. For the episodic laboratory task, the main contrast of interest was remember versus know. For the AM task, we contrasted each AM condition (recent, remote) versus odd/even (see below). Next, these results were used as an inclusive map to determine group differences in activity, which was also thresholded at p < .05, uncorrected for multiple comparisons within the MTL ROI. In other words, MTL activity had to show both a task effect and a group effect. The conjoint probability for this conjunction was p = .0025, with the caveat that these contrasts are not completely independent. This approach has been used in previous neuroimaging genetic studies (e.g., Dennis, et al., 2011). Moreover, this p-value was deemed appropriate given that this was a hypothesis driven analysis and given that the hippocampal/MTL region typically has lower amplitude hemodynamic responses relative to other brain areas. To demonstrate condition effects, first contrasts of interest are reported, collapsed across genotype groups (p < .005 uncorrected with a minimum cluster size of 10). For all comparisons, the minimum cluster radius was set to 6.6 (i.e., two clusters need to be separated by a minimum of 6.6 mm to be classified as unique). The cluster size was thresholded to a minimum of 150 ml3.

3.1.6.2.2 Multivariate

For multivariate analyses, after preprocessing, the data for both tasks (LM and AM) were spatially transformed to Talairach space (Cox & Hyde, 1997; Talairach & Tournoux, 1988) and were resampled to voxels of 4mm3. The data were spatially smoothed using a Gaussian filter with a full width at half maximum value of 8.0 mm. Analyses were performed by partial least squares (PLS) analysis (Krishnan, Williams, McIntosh, & Abdi, 2011; McIntosh, Bookstein, Haxby, & Grady, 1996; McIntosh & Lobaugh, 2004). Generally, PLS assess the relationship between patterns of whole brain activity and one or more other variables, such experimental

51 condition (i.e., Task PLS) or behavior (i.e., Behavior PLS) or a particular voxel in the brain (i.e., Seed PLS). These relationships are expressed as mutually orthogonal singular vectors or latent variables (LVs) that represent similarities and differences in patterns of activation in relation to the selected variables of interest (e.g., condition). The LVs are computed using singular value decomposition and are analogous to eigenvectors in principal components analysis. The reliability of an LV was assessed using permutation testing, which was run 500 times. Permutation testing involves re-sampling without replacement to reassign the order of conditions within each participant; PLS is recalculated for each newly ordered sample to determine the probability of each LV occurring by chance.

An LV was considered statistically reliable if the probability of the single value for the LV for a given permutation was less than 0.05. Each voxel in the brain has a “salience,” which is a particular weight on each LV; this value can be positive or negative, depending on the nature of its relationship to the pattern described by that LV. The stability of each voxel’s contribution to the LV was assessed using a bootstrap estimation of the salience standard errors with 100 resamplings (i.e., bootstrap ratio; BSR), which involves re-sampling of participants with replacement for each voxel, while maintaining the assignment of conditions for each participant and rerunning the PLS following each re-sampling (Sampson, Streissguth, Barr, & Bookstein, 1989). In PLS, it is not necessary to correct for multiple comparisons because whole brain statistical assessment is performed in a single analytic step.

First, a whole brain Task PLS was used to examine potential differences between conditions and groups. For AM, since the task was a slow event related design (i.e., long experimental epochs), spatiotemporal PLS was used (McIntosh & Lobaugh, 2004), which identifies time-varying (i.e., across the length of an epoch) brain patterns as a function of condition and group. For this analysis, a voxel salience was considered reliable when the BSR was above 6 (p < 0.00001) and contained a minimum size of 10 voxels; the threshold was used given the robust nature of the pattern (see results).

Functional connectivity (i.e., spatiotemporal Seed PLS) was next performed to examine patterns of co-activation between the hippocampus and the rest of the brain; a correlation between hippocampal activity and patterns of brain activation was computed across individuals within each condition, thereby producing within-task seed-brain correlations for each of the memory

52 conditions. For this analysis, a voxel salience was considered reliable when the BSR was above 6 (p < 0.00001) and contained a minimum size of 10 voxels. To obtain a seed ROI from the hippocampus, unbiased with respect to group, mnemonic condition, time, or temporal lag, a supplementary non-rotated Task PLS was first computed, (e.g., AM versus odd/even). From this analysis, a seed was extracted from the left and right hippocampus with a neighborhood voxel range = 2 (e.g., Grady, McIntosh, Beig, & Craik, 2001). This analysis investigated whether regions that correlated most strongly with the seed differed between memory conditions or groups.

3.2 Results 3.2.1 In Scanner Behavioral Responses

Table 4 shows performance and RT for the episodic LM task and in- scanner ratings for the episodic AM task. In accord with the pre-scan performance matching procedure, there were no significant genotype-dependent behavioral differences or trends for any measures.

3.2.2 Scanner Task 1: Recognition LM for Faces

A mixed design ANOVA, with group as a between subjects variable and memory condition as a within subjects variable (hits, FA), revealed that participants had significantly more hits relative to FA, F (1, 128) = 871.33, p < .0001, yet genotype groups did not significantly differ from each other, p = .92 (see Table 4) and there was no significant interaction, p = 1.00, indicating that both groups were equally successful at recognizing items as old. A comparison of RT for these conditions revealed no significant condition effects, p = .70, nor group differences, p = .43. Groups also did not differ in sensitivity, d’; p = .96. A mixed design ANOVA, with group as a between subjects variable and memory condition as a within subjects variable (remember, know), revealed that the proportion of remember responses was significantly greater than that of know responses, F (1, 128) = 9.9, p = .004, yet groups did not significantly differ, p = 1.00 and there was no significant interaction, p =.85. Likewise, although there were significantly more FA for know responses than for remember, F (1, 128) = 41.84, p < .0001, groups did not significantly differ, p = .93, and there was no significant interaction, p = .98; the group

53 differences remained non-significant even when accounting for FA within remember and know responses via subtraction, p = .84.

Participants were slightly faster at making remember responses relative to know responses, F (1, 27) = 4.15, p = .052, although this was marginally significant (n.b., one participant’s RT data were missing for this analysis due to a technical issue [scanner blackout]; behavioral responses were recorded manually yet RT data was not saved), with no significant group differences, p = .26, and no signiticant interaction, p = .82.

3.2.3 Scanner Task 2: Episodic AM

A Mixed design ANOVA was computed with group as a between subjects variable and memory condition (recent, remote, odd/even) as a within subjects variable. Null genotype-dependent differences were observed, p = .79, for in-scanner ratings of re-experiencing (see Table 4). By contrast, as expected, there was a main effect of condition, F (1.68, 46.98) = 139.38, p < .0001 (Greenhouse Giesser corrected); both recent, p <.0001, and remote, p <.0001, memories were rated as more vividdly re-experienced than the odd/even control condition, confirming a dissociation between the experimental and control conditions, while recent memories were rated as more vividly re-experienced than remote memories, p <.0001 (see Table 4). As noted earlier, although previous studies have experimentally equated vividness of re-experience ratings across time periods, in the present study I intentionally did not manipulate vividness as I was interested in potential genotype effects over naturally occurring temporal stages of AM.

3.2.4 Univariate Imaging Analysis 3.2.4.1 Scanner Task 1: Recognition LM for Faces

As a manipulation check to ensure overall condition effects were present in the contrast of interest, I first examined remember contrasted against know, collapsed across groups. As expected, hippocampus/MTL activity was significantly greater, bilaterally, for remember relative to know (left local maxima: -30, -36, -10, 720 mm3; hippocampal/MTL cortex, right local maxima: 26, -16, -4, 328 mm3; hippocampal). There were no MTL areas significantly more active for know relative to remember.

Comparing genotype groups for this contrast for the conjunction analysis, there was significantly greater activity in the T-carriers relative to the non-carriers (right local maxima: 32, -20, -20, 640 mm3; hippocampal/MTL cortex, 34, -12, -36, 592 mm3; MTL cortex; see Figure 5). The opposite contrast (i.e., CC versus TT/TC) revealed no significant differences in BOLD response.

3.2.4.2 Scanner Task 2: Episodic AM

Contrasting each AM condition to odd/even, collapsed across groups, revealed significantly greater BOLD response in the MTL cortex bilaterally for recent (left local maxima: -22 -40 -12, 13816 mm3, hippocampus/MTL cortex, right local maxima: 0 -36 -2, 13572 mm3; hippocampus/MTL cortex) and remote conditions (left local maxima: -32 -10 -28, 12880 mm3; MTL; -24 -40 -12, 360 mm3; hippocampus/MTL cortex, right local maxima: 20 -20 -14, 12800 mm3; hippocampus/MTL cortex). The opposite contrasts (i.e., odd/even vs recent; odd/even vs remote) revealed no significant differences in BOLD response.

A direct comparison of recent and remote memory conditions revealed significant differences with greater BOLD response for the recent relative to the remote condition (left local maxima: - 20 -38 -10, 208 mm3; MTL cortex). This was not surprising given that the former condition was rated as significantly more vivid and, as noted, previous work has suggested that vividness modulates the degree of hippocampal activity observed during AM retrieval (e.g., see Gilboa, et al., 2004). There were no clusters that were significantly more active for the remote condition. These conditions were treated separately in genotype analyses.

Comparing genotype groups, there was greater BOLD response in T-allele carriers relative to the non-carriers, bilaterally, for recent versus odd/even (left local maxima: -36 -20 -28, 568 mm3, MTL cortex; -26 -34 -4, 296 mm3; hippocampus; right local maxima: 26 -6 -32, 344 mm3, MTL cortex; see Figure 6). Interestingly, the opposite contrast also revealed a cluster that was significantly more active in non-carriers relative to T-carriers (left local maxima: -26 -24 -22, 528 mm3, MTL cortex; see Figure 7). For the remote time period, there was greater BOLD response in T-carriers relative to non-carriers (left local maxima: -32 -10 -28, 368 mm3, MTL cortex; -24 -40 0, 360 mm3; hippocampus/ MTL cortex). There were no areas of greater BOLD response in non-carriers relative to T-carriers (see Figure 8).

3.2.5 Multivariate Imaging Analysis 3.2.5.1 Experiment 1: Recognition memory

Contrary to expectation, there were no significant LVs that separated groups for the recognition memory LM task for the contrast remember versus know. This analysis identified two non- significant LVs, accounting for 52.10% and 47.90% of the cross-block covariance, p = .53, p = .59. Since these LVs were not statistically reliable they will not be discussed further.

3.2.5.2 Experiment 2: Episodic AM

In the first analysis, a whole-brain mean-centered Task PLS was run, examining the effects of conditions and groups across temporal lag (see methods). This revealed one significant LV, accounting for 77.84% of the cross-block covariance, p = 0.000. As shown in Figure 9A, this LV dissociated the AM conditions from the odd/even task in both groups. Focusing on overall condition effects, Figure 9B (which depicts every second lag) and Table 5 show that the pattern of BOLD response showed significantly greater activity in AM conditions versus odd/even and was widely more distributed within the AM network (Svoboda, et al., 2006). Across lags, increased BOLD activity was observed bilaterally in the precuneus. Beginning around lag 2 and extending into later lags, medial PFC activity was observed, which was mainly left lateralized, as well as middle temporal gyrus, and the cerebellum, bilaterally. Bilateral hippocampal/MTL activity was observed during middle lags only (i.e., lags 2-5). Finally, right lateral PFC regions, extending across dorsal and ventral portions across most lags (beginning around lag 2) and then restricted to dorsolateral PFC for later lags (around lag 6). While this pattern was more robust for the recent condition, this effect was modulated by genotype, as this difference was evident for the non-carrier group but the error bars overlapped for the T-carrier group. Finally, odd/even was associated with increased BOLD response in the superior temporal gyrus but only for middle lags.

Next, given the a priori interest in the hippocampus, a seed ROI was extracted from an unbiased (see methods) supplementary non-rotated Task PLS (i.e., specified contrast; AM versus odd/even, p = 0.000; see Figure 10A and 10B) The hippocampus was included in a cluster that extended from the parahippocampal gyrus in both hemispheres, with the most reliable voxels

56 identified in the hippocampi (left: -24 -32 -8, BSR = 6.741; right: 28 -24 -16, BSR = 6.898; see Figure 10B), providing seed voxels for the connectivity analysis (spatiotemporal Seed PLS).

The spatiotemporal Seed PLS analysis revealed 1 significant LV (58.72 of cross-block covariance explained), p = 0.000. As shown in Figure 11A, this LV was characterized by a distinct pattern of bilateral hippocampal connectivity in T-carriers for recent and remote AM; for non-carriers, this pattern was observed only for remote AM. The pattern of connectivity was widely distributed within the core AM network, as shown in Table 6 (and Figure 11B, which depicts every second lag). Across most lags, the hippocampus was functionally connected with lateral temporal regions, particularly right middle temporal gyrus as well as the cerebellum. During earlier lags (lags 2-4), hippocampal connectivity was observed with posterior regions, particularly the left precuneus and cuneus. During middle and late lags, beginning at lag 3, hippocampal connectivity was observed with the right superior (medial) PFC and the MTL, mostly parahippocampal gyrus. Finally, during middle lags, hippocampal connectivity was observed in the thalamus, bilaterally (lags 4 and 5) and in left basal ganglia regions (lags 3 and 4), and persisted into later lags as part of a MTL cluster. These regions were observed at late lags also but the peak was located in the MTL.

To ensure that the observed differences were not due to increased noise from the chosen seed in the non-carriers, I extracted the percent signal change from the seed locations, and performed a mixed effects ANOVA, with group as a between subjects variable and conditions of time period (recent, remote) and laterality (left, right) as within subjects variables. The signal change was similar across groups, p = .346, and there were no significant higher order interactions, all ps > .264.

3.3 Discussion

The present study investigated hippocampal/MTL cortex BOLD response in relation to a naturally occurring SNP of the KIBRA gene. Two previous imaging studies have reported conflicting patterns of hippocampal activity during episodic re-experiencing of LM: while Papassotiropoulos and colleagues (2006) reported enhanced hippocampal/MTL activity in non- carriers of the T allele with matched performance, more recently, Kauppi et al. (2011) reported

57 the opposite pattern: increased hippocampal/MTL activity in T-carriers in performance matched- and unmatched- groups.

I first attempted to clarify the discrepancy in the literature by examining KIBRA in relation to hippocampal/MTL BOLD response during re-experiencing of laboratory memories (Task 1). In accordance with the foregoing studies, I also matched participants’ performance across groups. Greater BOLD response was observed in T-carriers relative to non-carriers in the right hippocampus/MTL cortex when contrasting items endorsed as old that were recollected to old items that were familiar (McCormick, et al., 2010). These findings are consistent with the notion that enhanced episodic memory in T-carriers is associated with enhanced hippocampal/MTL processing during recollection.

The present results are in accordance with Kauppi and colleagues (2011) but are largely in opposition to those of Papasotiropolous and colleagues (2006). It is difficult to ascertain the reasons for the discrepancy among the studies. In the present study, we sought to match the methodology of Papasotiropolous and colleagues (2006) as closely as possible; indeed the current sample size and age range were very similar, while Kauppi and colleagues (2011) used older cohorts in their study, making it unlikely that age or sample size were contributing factors.

Examining task characteristics, in the present report, as in both prior studies, face stimuli were used to examine episodic processes, but the contrast of remember versus know may represent a more “process-pure” measure of episodic memory relative to these studies because it specifically compares items correctly recognized as old that were accompanied by conscious recollection with items that were also correctly recognized as old in the absence of recollection, while the prior studies did not distinguish episodic from non-episodic processes in this manner. While this is an important consideration, it does not explain why the present results would be different from Papasotiropolous and colleagues (2006), but similar to that of Kauppi et al. (2011). As noted in the introduction, Papasotiropolous and colleagues used an associative memory paradigm, where participants were asked to associate a name with an occupation, with each repeated 3 times (across 3 runs). This may made the task easier (performance means are not reported) and additionally recruited non-episodic processes such as working memory and semantic processes. At retrieval, participants were asked to indicate the subordinate category via button press for each face. Although it is unclear from their methodology precisely how they differentiated

58 recollected from non-recollected items, if correct subordinate was used as a measure of episodic memory, then this could lead to the exclusion of trials in which participants recollected the face but did remember the correct occupation (see Chapter 1 for discussion). Moreover, using fewer occupations may have made the task easier and required less pattern separation, while Kauppi et al. (2011) used trial unique face-name pairs. The present task utilized associating the face with one of four personality types, which was determined by the participant. Episodic recollection, however, was not restricted to correctly recollecting the personality type. Instead, a remember response constituted recollecting any sensorial or mental state details associated with recognizing the item. Hence, it is possible that the results obtained in the present study are related to task difficulty; lighter loads leads to significant increases in BOLD response in non-carriers, perhaps reflecting greater recruitment of resources; however, they can no longer do so when the task is too difficult, perhaps due to inefficient encoding/retrieval processes.

In addition to examining localized BOLD response, an ancillary goal was to use functional connectivity to expand the LM results. However, there were no significant LVs that differentiated groups. This could reflect the subtlety of the manipulation where remember versus know reflects differences at the hippocampus and MTL cortex and not at the network level. Yet, there are studies that suggest that remember versus know differences are reflected in larger scale neural networks (Wheeler & Buckner, 2004; Skinner & Fernandes, 2007).

The local effects from the LM task were extended to personally relevant, episodic AM events with a longer retention interval with a goal of assessing KIBRA-associations on naturalistic memory. KIBRA T-carriers also showed greater BOLD activity in the anterior and posterior portions of the hippocampus/MTL relative to non-carriers for both time periods, with strikingly similar localization across time periods within the left hippocampus. These results are consistent with those of Kauppi et al. (2011). Since the AM task involved re-experiencing of much more remote autobiographical experiences, relative to LM, this represents a novel contribution to the literature.

A second finding from the AM task is that for recent events there was also a significant cluster in the MTL cortex that was more active in non-carriers relative to T-carriers, although this cluster was not within the hippocampus itself but in the left parahippocampal gyrus (Brodmann area 35) around the PRC, which may reflect greater reliance on familiarity processes in non-carriers,

59 although no behavioral differences were observed per se. Whether this reflects a qualitative difference in mechanism is difficult to ascertain as both groups activated the MTL cortex. Considering the findings of Papassotiropolous and colleagues (2006), the cluster observed in non-carriers in their study was right lateralized with peak maxima in the hippocampus, while, in the present study, it was located on the left side in the parahippocampal gyrus.

While analyses of regional activity can reveal group differences in brain regions engaged during a task, functional connectivity analyses can demonstrate how regions co-vary together during a task, indicating functional coupling or synchronization of neural networks. This approach demonstrated a significant group by time period interaction, such that, for recent events, T- carriers demonstrated a pattern of hippocampal connectivity during AM re-experiencing but this pattern was not observed in non-carriers. On the other hand, for the remote time period, a similar pattern of hippocampal connectivity was observed between groups. These findings are consistent with the Task PLS results, where non-carriers showed a significant difference between recent and remote memories, while T-carriers did not.

For T-carriers, for recent events, and for both groups, for remote events, hippocampal connectivity was observed in midline anterior and posterior regions previously associated with episodic AM, including, for most lags, the medial PFC, a key region in the AM network (Svoboda, et al., 2006) that plays a prominent role in processes related to the self (Northoff, et al., 2006), and is essential to the mental time travel (i.e., autonoetic consciousness) inherent to re-experiencing events (Tulving, 2002). Indeed, this region is preferentially active for episodic over semantic AM (Levine, et al., 2004).

Another area that showed reliable hippocampal connectivity, particularly during earlier lags, was the precuneus, an important region in the AM network (Svoboda, et al., 2006), with connections to medial PFC (Cavanna & Trimble, 2006), previously associated with visuospatial imagery and self-referential processes (Cavanna & Trimble, 2006; Fletcher, et al., 1995). Accordingly, the precuneus is thought to mediate recall of visual spatial imagery associated with past events and links them to mental representations of the self (Andreasen, et al., 1995). Interestingly, we found these two regions to be less active in a group of healthy individuals who subjectively report reduced capacity to re-experience events across the lifespan (Palombo et al., in preparation).

This LV was also characterized by hippocampal connectivity with the parahippocampal gyrus, which was also observed for most lags. This is not surprising given that previous reports, as well as our univariate analysis show increased activity in T-carriers, not only within the hippocampus, but also the adjacent MTL cortices. Overall, this pattern of connectivity, prominent only in T- carriers for recent events, may reflect a qualitatively distinct mechanism of retrieval, such that T- carriers may uniquely co-activate areas that involve self-referential or imagery-based processes for recollection of past events. This may be due to group differences in the underlying anatomy, an issue partially addressed in the next chapter.

Turning to remote events, a similar pattern of functional connectivity was observed between groups; both groups reliably contributed to the pattern observed. Considering that the recent condition was rated as significantly more vivid relative to the remote memory condition in both groups this is not surprising. Indeed, it is well known that episodic memories become less vivid over time (see Piolino, Desgranges, & Eustache, 2009 for review). Hence, it is possible that the recent condition was a less noisy measure of episodic memory, while remote memories may have been more semanticized (Conway, Gardiner, Perfect, Anderson, & Cohen, 1997) for some participants and thus less likely to produce distinct patterns in hippocampal connectivity (also see Conway, et al., 1997; Gilboa, et al., 2004; Piolino, et al., 2009; Piolino, Lamidey, Desgranges, & Eustache, 2007). In the present study, while I purposefully did not manipulate vividness, it would be useful to repeat the experiment, while carefully equating memory vividness across time periods to address this issue more specifically. Indeed, while Gilboa and colleagues (2004) did not manipulate vividness per se, they found that hippocampal BOLD response was related to the degree of vividness associated with memory as cued by family photographs presented in the scanner irrespective of remoteness. Moreover, using a prospective cueing method, where participants were presented with recordings made directly after experiencing an event, Sheldon and Levine (2013) recently showed that when remote memories (i.e., memories 1.5 years old) are stratified according to participants’ vividness ratings, remote memories are highly similar to recent memories (~1 month old) in hippocampal connectivity but distinct from remote memories that are low in vividness. However, the authors additionally showed a pattern of hippocampal connectivity that was specific to remote memories, irrespective of vividness, suggesting that vividness cannot entirely account for memory-age related differences in brain activity. A recent magnetoencephalography study is also in accordance with

61 these fMRI studies; the authors found that the magnitude of MTL phase synchrony (theta oscillations) with midline regions (i.e., the precuneus and medial PFC) during episodic AM retrieval predicted the degree of subjectively rated visual imagery (Fuentemilla, Barnes, Duzel, & Levine, 2013)

There are some limitations of the present study. First, while the sample was matched in size to Papassotiropolous et al. (2006), it was nonetheless small, and smaller than that of Kauppi et al (2011). Moreover, unlike the foregoing studies, which used both sexes, I used an all female sample, which is more homogeneous, yet an obvious weakness with respect to generalization to male participants. On the other hand, Papassotiropolous et al. had had a 1:2 male to female ratio in each group, while Kauppi et al. had more balanced groups but only in their matched sample; their unmatched sample had a larger proportion of females in the T-carrier group. An additional limitation is that I was unable to parcel out precisely where group differences in BOLD response occurred within the MTL. Higher resolution images would allow one to localize these effects with more precision. Finally, an open question, addressed in the next chapter, is whether enhanced hippocampal activity and hippocampal-neocortical activity in T-carriers is mediated by structural differences between T-carriers and non-carriers.

Chapter 4 The effects of KIBRA Polymorphism on Hippocampal Sub-regions using High-resolution MRI4

4 Introduction

The goal of this chapter was to determine whether the KIBRA CT polymorphism is also associated with volumetric differences in the human hippocampus. KIBRA expression is high in the hippocampus in both humans and rodents (Johannsen, et al., 2008; Papassotiropoulos, et al., 2006; Yoshihama, Hirai, Ohtsuka, & Chida, 2009). Moreover, as described in Chapter 3, KIBRA- dependent differences in BOLD response in the hippocampus/MTL have been observed during episodic remembering in prior work (Kauppi, et al., 2011; Papassotiropoulos, et al., 2006), although these studies actually yielded opposite results (i.e., increased activation in non-carriers versus increased activation in T-carriers, respectively). Further, as described in Chapter 3, I also observed KIBRA-dependent differences in hippocampal/MTL BOLD response in support of Kauppi et al. (2011).

Yet, null genotype differences in whole hippocampal and MTL cortex volume have been observed in one report that used both manual and automated MRI segmentation (Papassotiropoulos, et al., 2006). However, the hippocampus consists of histologically heterogeneous sub-regions (i.e., CA fields 1-3, DG, subiculum) with distinct patterns of connectivity and cellular structure (e.g., Amaral & Insausti, 1990; Duvernoy, 2005).

Papassotiropoulos et al. (2006) demonstrated that KIBRA is most highly expressed in the CA1 and DG in rodents; likewise Johannsen (2008) observed intense KIBRA expression in DG granule cells and CA pyramidal cells in rodents (also see Yoshihama, et al., 2009), raising the possibility that whole-hippocampal volume may be less sensitive marker to assess KIBRA- dependent structural variability in humans. To assess regional differences in the effects of KIBRA genotype on hippocampal structure, we conducted MRI-based volumetric analyses of

4 A version of this Chapter was accepted for publication in the Journal of Neuroscience. Only slight modifications were made (particularly to the introduction and discussion) to the present chapter to integrate it with the remaining chapters.

63 hippocampal and adjacent MTL cortical sub-regions. I hypothesized that KIBRA-dependent differences in hippocampal volume would be localized within the hippocampal CA fields and DG, paralleling the expression patterns observed in rodents.

4.1 Method

4.1.1 Participants Thirty-two healthy Caucasian individuals, stratified by KIBRA genotype (T-carriers and non-carriers), were recruited from a larger study (Chapter 2). Genotype groups were matched on genetic characteristics previously implicated in episodic memory: APOE ε4 alleles, BDNF Val66Met, and COMT Val158Met. Prior to structural imaging, genotype groups were carefully matched (one-to-one matching between genotype groups with a standard deviation of .4) on recognition memory scores from the objects task that was administered in a prior session (see Chapter 2). Due to equipment failure, imaging data were unavailable for four non-carriers. Thus the sample included 18 T-carriers (22.2 ± 3.7 years old; 15.5 ± 2.8 years of education; 4 male) and 14 non-carriers (20.3 + 3.0 years old; 13.9 + 1.6 years of education; 3 male). This did not substantively affect the balance between genotype groups on the aforementioned variables: APOE ε4 alleles, p = .57; BDNF Val66Met, p = .77; or COMT Val158Met, p = .96; chi-square tests. Genotype groups also did not significantly differ in recognition memory scores, p = .78. While it is acknowledged that performance differences attributable to genotype are unlikely to be detected in this sample size, the present results may nonetheless help to explain findings from other larger studies where performance effects are observed.

Given previous research that sex modifies the relationship between KIBRA and cognition (e.g., Wersching, et al., 2011), sex was included as a covariate in all analyses, although this did not change the pattern of results. There was no significant difference in age between groups, p = .13. There was a marginal group difference in education, F (1, 29) = 3.65, p = .07. As education was not found to be associated with the volume of any ROIs within genotype group, it was not included as a covariate. This study was approved by the Baycrest Research Ethics Boards. Participants provided written informed consent and were compensated $50 for the MRI.

4.1.2 Genotyping

Procedures for genotyping participants are described in Chapter 2.

4.1.3 MRI Acquisition

As noted in Chapter 3, structural images were acquired using a 3T Siemens Trio scanner. For segmentation, high-resolution T2-weighted images were acquired in an oblique-coronal plane; slices were arranged perpendicular to the long axis of the hippocampus (TE/TR = 68 ms/3000, 22-28 oblique-coronal slices depending on head size, 512 x 512 acquisition matrix, voxel size = 0.43 x 0.43 x 3 mm, no skip, FOV = 220 mm). The first slice was placed slightly anterior of the collateral sulcus; the last slice was placed just prior to the full superior-inferior extent of the ventricles (see Figures 12-14). To confirm the placement of slices according to these boundaries, a whole-brain anatomical MRI was first acquired, using a 3D T1-weighted MP-RAGE (TE/TR = 2.63 ms/2000 ms, 176 oblique-axial slices, 256x192 acquisition matrix, voxel size = 1 mm3, FOV = 256 mm). The T1-weighted images were also used to obtain a measure of total brain volume (see below).

4.1.4 Segmentation

ROI segmentation was completed in participants’ native space (coronal plane; see Figure 15). Akin to most previous studies (Chen, Olsen, Preston, Glover, & Wagner, 2011; Mueller, et al., 2007; Olsen, et al., 2009; Olsen, et al., 2013) sub-regions were not segmented across the entire long axis of the hippocampus. Instead, segmentation was completed only within the middle section, where the dark bands separating hippocampal layers are clear. This corresponded mainly to the body of the hippocampus, however, the most posterior head slices were also demarcated, where the subfields can also be differentiated (Carr, Rissman, & Wagner, 2010; Chen et al., 2011; for an alternative approach strictly excluding head slices see e.g., Mueller et al. 2007).

CA1, subiculum, and DG/CA2/3, were segmented with the latter grouped into a single ROI, as these regions cannot be reliably delineated at this resolution. Combined ROIs (i.e., without subfield demarcation) were used to define the remaining head and tail hippocampal slices, which afforded a measure of full hippocampal volume. Outside the hippocampus, the MTL cortices were segmented: ERC, PRC, and PHC.

Segmentation was performed in FSLview (v3.1.2) by a single rater, blind to group status (Figure 15). Segmentation was guided by standard anatomical guidelines (e.g., Amaral & Insausti, 1990; Duvernoy, 2005; Insausti, et al., 1998) following the procedures described for MTL sub-region segmentation for hi-resolution fMRI by Olsen et al. (2009); which are similar to those used to assess sub-region volumetrics (Mueller et al. 2007; Olsen, et al., 2013).

Intrarater and interrater reliability were established by comparing segmentations of five participants’ brains segmented twice by the same rater (repeated with a 2-6 mo interval) and to that of a second rater. Reliability was assessed using the Dice overlap metric (Dice, 1945), which was computed for each ROI within each hemisphere (0 = no overlap; 1 = perfect overlap; see Table 7). It is computed by taking the overlap volume and multiplying it 2. Next this value is divided by the sum of the individual volumes (2[overlap of A and B]/A+B). Dice values were comparable to those reported elsewhere (e.g., Bonnici, et al., 2012). Total brain volume estimates were acquired from the T1-weighted images using an adapted version of the ANIMAL algorithm (Collins, Holmes, Peters, & Evans, 1995). Genotype groups differed marginally in total brain volume, F (1, 29) = 3.00, p = .09, non-carriers had a larger total brain volume relative to T- carriers. Total brain volume was accounted for in each ROI using a regression-based technique; each ROI was regressed on total brain volume (collapsed across groups) and the residual value (i.e., the structure’s actual size minus its predicted value based on the individual’s total brain volume) was accounted for in each ROI for each individual (Arndt, Cohen, Alliger, Swayze, & Andreasen, 1991). All analyses reported below were performed on total brain volume -adjusted ROI values.

4.2 Results

Group differences in MTL sub-region volumes (total brain volume-adjusted) were assessed with three mixed-design ANCOVAs (i.e., whole unsegmented hippocampus; segmented hippocampus; segmented MTL cortex), and post hoc tests. ANCOVAs included KIBRA genotype as a between-subjects factor and laterality and ROI as within-subjects factors (ROI modeled only for the latter two ANCOVAs). All models included sex as a covariate, as noted above, and a correction for violation of sphericity was applied for repeated measures. Since laterality did not interact with genotype (p > .10 for all comparisons), it was not interpreted.

Partial eta squared (hp2) was used as an estimate of effect size for significant effects. First, to complement methodology used by Papassotiropoulos and colleagues (2006), I conducted an ANCOVA on the full hippocampus (i.e., collapsing across all hippocampal ROIs including the head and tail within each hemisphere). Contrary to their findings, this analysis revealed larger hippocampal volume in T-carriers relative to non-carriers, F (1, 29) = 4.70, p = .04, hp2 = .14. Next, I conducted an ANCOVA on the segmented hippocampus, with ROI as a within subjects factor (i.e., CA1 and DG/CA2/3, subiculum). This analysis revealed a significant ROI x Group interaction, F (1.97, 49.87) = 4.63, p = .01, (Huynh-Feldt correction), and the main effect of KIBRA genotype was also significant, F (1, 29) = 5.35, p = .03, hp2 = .16 (see Table 8). Post hoc ANCOVAs for each ROI were performed to explore the nature of this interaction. T-carriers had 2 2 a larger CA1, F (1,29) = 8.51, p < .007, hp = .23, and DG/CA2/3, F (1, 29) = 4.21, p = .049, hp =.13. No group differences were found for subiculum, p = .20; indeed, non-carriers had a numerically larger volume relative to T-carriers in this region). Finally, I examined the effects of KIBRA genotype on MTL cortex volume, with ROI as a within-subjects factor (i.e. PRC, ERC, PHC), to determine the specificity of the observed effects. This analysis did not reveal group differences (p = .24) or interaction (p = .43).5

4.3 Discussion

This chapter investigated the effects of KIBRA genotype on hippocampal and MTL volume using structural imaging. I provide preliminary novel evidence that the KIBRA T-allele is specifically associated with larger CA1 and DG/CA2/3 volume in young adults, suggesting a putative neural mechanism for the effects of KIBRA genotype on episodic memory reported in the literature (Milnik et al., 2012) and those observed in Chapter 2. In terms of sub-region specificity, rodent work suggests functional dissociation within the hippocampus, with CA1 implicated in late retrieval and consolidation processes, while DG/CA2/3 is implicated in encoding and early retrieval (see e.g., O'Reilly & Rudy, 2001; Rolls & Kesner, 2006 for review). Similar

5 An earlier version of this dissertation included a female only sample. Analyses with the subset of females did not substantively alter the pattern on results observed in the present chapter, which includes both sexes.

67 differentiation has been demonstrated in healthy older adults (Yassa, et al., 2011) and clinical populations (Kerchner, et al., 2012; Mueller, et al., 2011; Mueller, et al., 2012). However, because these studies examined older adults and patients, it is not known whether they extend to young, healthy individuals.

Interestingly, Papassotiropoulos and colleagues (2006) observed effects of KIBRA on delayed memory in all three cohorts (ranging from 5 minutes to 24 hours), yet no effects on immediate recall, aligning the effect with the functioning of CA1. A similar pattern was observed by Bates et al. (2009) in the largest sampled KIBRA study (>2000 subjects). The interpretation was that KIBRA is not important for processes related to early memory formation, but instead relates to consolidation or delayed retention. These findings corroborate well with the present observation that KIBRA-related differences were evident in this subfield and with prior reports demonstrating high levels of expression in CA1 (Johannsen, et al., 2008; Papassotiropoulos, et al., 2006). Yet, other studies have observed KIBRA-related performance differences at immediate recall only (e.g., Schaper, et al., 2008) or both retention intervals (e.g., Vassos, et al., 2010), suggesting the effects of KIBRA on memory are nuanced.

There were no significant effects of KIBRA genotype on MTL cortex volume, suggesting that the effects of KIBRA genotype on episodic memory previously reported in the literature may be driven by differences in the hippocampus proper. However, fMRI studies have also implicated a major role of the MTL cortex in episodic memory. For example, PRC and PHC are held to encode item and context information, respectively, while the hippocampus integrates this information (Diana, et al., 2007). Since contextual information is critical for recollection, PHC is thought to be especially important for this process. Moreover, Chapter 3 of the present thesis as well as the fMRI studies of KIBRA (Kauppi, et al., 2011; Papassotiropoulos, et al., 2006) observed group differences in activation that extended into the MTL cortices, although the specific localization is unclear.

To address this issue, I performed exploratory post-hoc analyses, which revealed no group differences for PRC, p = .44, or ERC, p = .84, while PHC volume was slightly larger in T- carriers relative to non-carriers, with a marginally significant group difference, F (1, 29) = 3.77, p = .06, hp2 = .12. This subtle difference in volume may either reflect direct effects of KIBRA on

PHC or a downstream neuroplastic effect resulting from KIBRA’s effects on the hippocampus and associated mnemonic advantage in T-carriers observed previously. While Papassotiropolous and colleagues (2006) reported highest levels of KIBRA expression was in the hippocampus and the temporal lobes (encompassing the entire temporal lobe and hippocampus) in humans, the specific localization of expression within the temporal lobes was not examined.

An important caveat is that the current protocol does not differentiate between DG and

CA2/3. Hence, it is possible that I diluted specific effects of one of these sub-regions. Akin to previous investigations (e.g., Das, et al., 2012; Mueller, et al., 2007; Olsen, et al., 2009; Zeineh, et al., 2000) we employed relatively thicker slices resulting in anisotropic voxels, motivated, in part, by the hippocampal atlases used for landmark demarcation that employ ~2-4 mm thick slices (e.g., Amaral & Insausti, 1990; Duvernoy, 2005). Moreover, the layered structure of the hippocampus is best appreciated coronally, where the resolution was high. Nonetheless, given the resolution, the current segmentation scheme was limited to the hippocampal body and most posterior head slices. Accordingly, future studies employing higher resolution to segment across the entire hippocampal long axis are needed to assess the specificity of the KIBRA-sub-region association reported here and confirm these preliminary findings. Complementary data from human histological studies of KIBRA are also required.

Given differences in MTL segmentation reported in the literature, methodological factors cannot be ruled out as contributing to the results. Indeed, I observed KIBRA effects on whole hippocampus, while Papassotiropolous et al. (2006) failed to observe effects at a similar resolution. Likewise, the sub-region specific findings reported here may be due to the segmentation scheme adopted and, therefore, may not be observed with a different labeling approach used by other researchers. Additionally, the null effect of KIBRA genotype on subiculum volume in the present study may have been related to the difficulty in demarcating this structure, where reliability was low relative to the other regions, a finding reported elsewhere (e.g., Mueller et al., 2007; Bonnici et al., 2012). Finally, while subfield protocols rely heavily on the use of atlases to determine the placement of landmarks, for reliability, arbitrary landmarks are also sometimes employed as a “best guess” in situations where certain borders are less clear, representing a trade off between validity and reliability. This represents a challenge in this field that may be aided by future

69 studies mapping sub-region segmentation from postmortem (ex-vivo) MRI onto that of in- vivo imaging.

Another limitation is that the sample was predominantly female. Previous research has shown that sex modifies the relationship between KIBRA and cognition with larger genotype effects in females (e.g., Wersching, et al., 2011). Given the small number of males, I was not able to address this issue, although covarying sex in the analysis did not alter the pattern of results.

KIBRA-related differences were observed in the absence of behavioral differences, as groups were deliberately performance-matched. Likewise, the aforementioned two previous fMRI studies have reported KIBRA-related differences in hippocampal and MTL cortex BOLD response, in the absence of performance differences (due to matching; Kauppi, et al., 2011; Papassotiropoulos, et al., 2006). The purpose of performance matching is to ensure that neural differences are not driven by behavioral effects per se (Rasch, Papassotiropoulos, & de Quervain, 2010). As noted by Kauppi et al. (2011), this approach does not necessarily eliminate all behavioral differences; accordingly, the genotype-dependent differences in brain structure or BOLD response may reflect qualitative differences in the integrity of the memory trace (e.g., deeper encoding, greater vividness), which are not necessarily captured by the particular task employed for performance matching.

Identifying a putative neural mechanism for KIBRA’s effects on human memory may have clinical utility. Recent research suggests that the hippocampal subfields are differentially associated with various neurological conditions. For example, CA1 is an early target site of pathology in Alzheimer’s disease (e.g., Kerchner, et al., 2012). Moreover, a recent meta- analysis also suggests an association between KIBRA and risk of Alzheimer’s disease (Burgess et al., 2011) although the findings have been somewhat mixed with respect to the direction of the allelic-specific effect (also see Hayashi et al., 2010). Future studies are needed to explore further the complex relationship between KIBRA, cognition, and disease- related neuropathology within the hippocampus. It would also be useful to examine hippocampal volume longitudinally, to determine if the effects of KIBRA accumulate with age.

Chapter 5 General Discussion

The goal of this dissertation was to explore factors that influence individual differences in real- world episodic remembering. In Chapter 1, I provided a broad overview of the literature, which demonstrated evidence that there are large individual differences in behavioral and brain biomarkers of episodic memory as defined by LM tasks. However, despite the importance of LM as a proxy to understanding individual differences in real-life mnemonic capacity, humans evolved to process much richer experiences than what can be captured by laboratory event. As such, using episodic AM as a phenotype more closely approximates the biologically relevant mnemonic functions expressed in these linkages than do laboratory tasks. Yet, evidence in support of the construct of episodic AM as an individual difference variable is far less developed experimentally, despite the fact that these differences are self-evident in daily life. As detailed in Chapter 1, evidence for individual differences in episodic AM is limited to behavioral studies of AM phenomenology, personality and psychopathology, as well as from a small number of case studies of healthy people with extreme AM profiles. These cases provide preliminary support that individual differences in AM have a neural basis. This large gap in the literature motivated the present dissertation.

Considering some of the neural factors that may relate to AM variability, this dissertation narrowly focused on the role of genetics, as there has been rapidly accumulating evidence suggesting that individual differences in episodic LM are associated with SNPs in several genes. Assuming that the laboratory tests assess biologically meaningful mnemonic capacities, I suggested that these are viable candidate genetic markers to probe the origins of individual differences in AM as a starting point. To explore this hypothesis, I focused exclusively on one SNP of the KIBRA gene, as this demonstrated the most robust association with episodic LM in previous research (Papassotiropoulos, et al., 2006) and as such, is a candidate for providing “proof-of-principle” for the hypothesis that episodic AM has a genetic basis.

In the experimental chapters, I explored the effects of the KIBRA gene on episodic AM at multiple levels of analysis. However, it was important also to clarify the role of KIBRA in episoidic LM as there are inconsistent behavioral and neuroimaging results. Moreover, it would

71 also be difficult to interpret differences (or lack thereof) in episodic AM in the present sample in the absence of laboratory measures as previous research has focused exclusively on this type of stimuli.

At the behavioral level, the experimental work described in Chapter 2 yielded several important findings. First, I showed, for the first time, that KIBRA influences individual differences in remote episodic AM. Since previous research had only examined retention intervals no longer than 24-hours, the observation that KIBRA influences episodic AM at a retention interval of at least 3-months represents an important contribution to the literature as it suggests that similar biological markers implicated in more proximate memory performance (as assed by LM) are involved in more remote aspects of memory. However, the effects observed were modest and the sample size was relatively small.

A somewhat surprising finding was that this effect was observed only negatively valenced AMs. This result may be driven by the greater sensitivity of the negative AM, which showed a typical memory enhancement effect in all participants and was also rated as more personally significant relative to the neutral event. Moreover, the negative event was rated as easier to visualize relative to the neutral event, a finding that was more pronounced in T-carriers relative to non-carriers. This finding is particularly interesting in light of the fMRI results from Chapter 3, where T- carriers showed a pattern of connectivity during AM re-experiencing (discussed in more detail below) for recent and remote AMs that was only present for remote AMs in non-carriers The hippocampus demonstrating more robust coupling with areas important for visual imagery, particularly the precuneus (Cavanna & Trimble, 2006). Similarly, in a case series of individuals with extremely low episodic AM capacity, we previously observed reduced hippocampal- precuneal coupling during AM experiencing using a very similar paradigm (Palombo et al., submitted). Together, these data are somewhat suggestive of a putative pathway that may be an important biological marker for individual differences in episodic AM. In particular, it is possible that T-carriers show a mnemonic episodic advantage via increased visual processing of experiences. Indeed, Rubin and colleagues have demonstrated that visual imagery is one of the highest phenomenological predictors of recollection (Rubin, Schrauf, & Greenberg, 2003; Rubin & Siegler, 2004).

More broadly, only two other studies have examined the role of genetics in AM (de Quervain, et al., 2012; de Quervain, et al., 2007). As noted earlier, these studies have identified genetic loci that are involved in individual differences in traumatic remembering and, as such, represent important possible candidates for understanding the role of genetics in susceptibility to trauma- related disorders such as PTSD. Notwithstanding obvious differences in the nature of the memories sampled in those studies relative to the present study (traumatic AM versus negative AM), it is interesting to speculate that KIBRA’s influence on AM also hold similar implications. However, it is difficult to disentangle the contribution of valence versus significance in the present study, an issue that is difficult to circumvent in AM studies as negative experiences tend to be highly significant.

The results from the behavioral study of AM did not map directly onto the LM results. While KIBRA influenced 1-week recognition memory for laboratory items, with T-carriers demonstrating a mnemonic advantage relative to non-carriers, this effect was observed for neutral items only, with only a marginal effect of KIBRA on negative items. However, because the LM task did not yield the expected memory enhancement effect typically observed in the literature (Kensinger & Corkin, 2003), it is unclear how to interpret these results. Nonetheless, these data lend support to the idea that laboratory and AM tasks can dissociate in their effects. This is an important general consideration for studies investigating individual differences in episodic memory. For example, in studies of individuals with highly superior memory, there is a dissociation between their performance on LM tasks of episodic memory, which is relatively normal, and their performance on naturalistic measures, which is highly superior and extremely accurate (Leport, et al., 2012; Parker, et al., 2006).

In subsequent chapters, I asked where the effects of KIBRA on episodic memory are localized in the brain. Such an approach is useful in that biological phenotypes, such as BOLD response, are considered somewhat more proximate to a gene’s function, relative to behavioral paradigms, and, therefore, can be more sensitive to determining the effects of a gene, although BOLD response is also an indirect measure of neural activity. Consequently, far fewer participants are typically needed to demonstrate genotype-dependent differences. On the other hand, such an approach still requires a task that will reliably evoke neural activity in association with the function of interest making task sensitivity of paramount importance.

To examine the effects of KIBRA on BOLD response in the hippocampus and MTL cortex, I employed performance matching in a subset of participants from the behavioral cohort. More specifically, participants were matched on the objects task from Chapter 1 as previous research had shown that KIBRA influences recognition memory for single items (Preuschhof, et al., 2010). One caveat worth addressing is that participants did not in fact differ on that task in the larger behavioral cohort. This raises the question of whether this performance matching achieved its desired goal. As a manipulation check, I examined performance on all scanner tasks and demonstrated null group differences and no group-level trends in the data, although, given the sample size, significant group differences would be unlikely. Hence, it is unclear how this issue affected the pattern of results.

At the ROI level, a general pattern of greater BOLD activity in T-carriers relative to non-carriers emerged in the hippocampus and adjacent MTL cortices during episodic re-experiencing of laboratory items (McCormick, et al., 2010) and AM (Soderlund, et al., 2012). Interestingly, while all of the contrasts showed clusters of greater BOLD response in T-allele carriers relative to non-carriers, in both the hippocampus and adjacent MTL cortices, there was also an MTL cortex cluster that was more active in non-carriers for recent memories, a cluster that was located in the PRC. This may represent a qualitatively different neural mechanism for non-carriers, perhaps a greater reliance on semantic processes, which are thought to be supported by some MTL cortical regions, (Vargha-Khadem, et al., 1997), but it is difficult to draw these conclusions given the restricted anatomical resolution in functional imaging and the use of blurring.

T-carriers also showed a distinct pattern of hippocampal-neocortical connectivity during AM re- experiencing. More specifically, for recent memories, T-carriers showed hippocampal connectivity within the AM network (Svoboda, et al., 2006), yet groups both groups showed this pattern for remote memories. Because episodic memories become semanticized over time, the remote memories may have been represented by a qualitatively different memory system that was not sensitive to KIBRA effects. This may have been due to the sampling method used, as recent and remote memories were not equated in terms of vividness. Indeed, studies have shown sensitivity of the hippocampus to vividness (Daselaar et al., 2008; Gilboa et al., 2004; Addis, Moscovitch, Crawley, & McAndrews, 2004; Sheldon & Levine, 2013).

Finally, in Chapter 4, I attempted to address the localization of KIBRA’s effects using high- resolution structural imaging. Based on expression studies in both humans and non-human species, which show that KIBRA is most highly expressed in the CA fields and the DG, I expected that the effects of KIBRA on hippocampal volume would be localized in those sub- regions. Sub-regions of the hippocampus and adjacent MTL cortices were segmented on T2- weighted images. T-allele carriers had a larger volume of the CA1 and the DG/CA2/3, while other sub-regions were unaffected. These novel findings provide evidence for a more putative neural mechanism for the effects of KIBRA on episodic memory performance previously reported by other researchers and, as such, provide more clues about KIBRA’s molecular function. Although there has been little work relating subfield integrity to naturalistic mnemonic ability, one recent study implicated the CA1 region in episodic AM. Participants with transient global amnesia, which results in focal and acute lesions to the CA1 region, showed a deficit in the retrieval of AM across the lifespan (Bartsch, Dohring, Rohr, Jansen, & Deuschl, 2011), suggesting a putative role of this region in episodic AM. This work and the present data raise the exciting possibility for identifying a relationship between AM performance and brain structure at the level of the subfields. In particular it would be interesting to examine different facets of AM (spatial, temporal, affective, self-relevance, temporal proximity, etc) in relation to hippocampal subfield integrity (Bonnici, Chadwick, & Maguire, 2013). For example, Bonnici et al. (2013) utilized pattern classification algorithms for fMRI data to show that while both recent and remote AMs were represented in the anterior portion of the hippocampus, remote memories were represented only within the posterior hippocampus and specifically within the CA3 and DG subfields.

In addition, in the present thesis, because PHC volume was numerically higher in T-carriers relative to non-carriers at the trend level, the present work is not only suggestive of a KIBRA influences on adjacent MTL cortex sub-regions, but more broadly provides an example of how genetics can be used as a platform for understanding hippocampal and MTL cortex function. There has been great interest in understanding the functional neuroanatomy within the MTL and whether there is a division of labor within this system (Squire, Stark, & Clark, 2004; Vargha- Khadem, et al., 1997). To inform this topic further, future research should specifically target whether distinct molecular processes are expressed both within the sub-regions of hippocampus and within parts of the MTL cortex and how this impacts behavioral performance. Hence, while the questions in this dissertation were framed in the context of KIBRA, at a more general level,

75 the structural results implicate the subfields are important biological candidates for studying individual differences. Such individual differences may be mediated by molecular processes that likely involve several genetic candidates.

Given the findings in Chapters 3 and 4, an interesting follow up study would be to examine BOLD response within the MTL sub-regions examined in Chapter 4 using high-resolution functional imaging to determine more precisely where the differences in activation originate from. This type of approach has been used in other studies examining regional activity (Bakker, et al., 2012; Bonnici, et al., 2013) and with network analyses (Libby, Ekstrom, Ragland, & Ranganath, 2012). Turning back to the Bonnici et al. (2013) findings, pattern analysis techniques would be useful for further parceling out the effects of KIBRA observed in the fMRI study in Chapter 3, that is, to determine whether groups differ along the long axis of the hippocampus with respect to the distribution of recent and remote AMs and within subfields. Given the findings of Chapter 4, an obvious candidate for genotype differences would be in the CA fields and DG.

While in the present study, there were ROIs segmented on high-resolution images for many participants, given the resolution of the functional data, and given that there were not labels available for every participant, I was not confident about using these as masks to determine the localization of BOLD response in Chapter 3. Moreover, many of the functional results were localized in either the head or tail region where the subfields are not typically labeled at the resolution employed here. Nonetheless, the findings of this Chapter once again highlight the importance of not only the hippocampus proper, but also the MTL cortices in KIBRA-related function. As noted, a trend was observed within the PHC region, where T-carriers showed larger volume of this region relative to non-carriers. Considering the extra-hippocampal MTL cortex BOLD response observed here, future studies should consider how the rest of the MTL contributes to KIBRA function. In particular, the importance of PHC in episodic memory has been noted elsewhere (Diana, Yonelinas & Ranganath, 2007). As noted in Chapter 4, while Papassotiropoulos and colleagues (2006) found greater KIBRA expression in the temporal lobes, they did not address which parts of the temporal lobes were driving these effects. Understanding what MTL regions are affected by KIBRA genotype have important implications for further

76 specifying the role of this gene in declarative memory function but also has implications for theories of MTL function.

Beyond the importance of replication, a future approach is to understand the role of gene x gene interactions. Other researchers have examined compounded effects of several genes on episodic memory performance (de Quervain & Papassotiropoulos, 2006). In particular, two studies have suggested that CLSTN2 (encoding synaptic protein calsyntenin 2) interacts with KIBRA to affect episodic memory performance (Papassotiropoulos, et al., 2006; Preuschhof, et al., 2010). Moreover, while many studies have focused on the role of rs17070145, there are other SNPS on the KIBRA gene that may be of importance. Since KIBRA is an intron, its effects are likely influenced by other causal SNPS that have yet to be identified.

A final issue relates to the clinical importance of KIBRA. Its role in episodic memory and hippocampal and MTL function begs the question as to whether this gene is relevant to Alzheimer’s disease. KIBRA has been examined in the context of Alzheimer’s disease in a few studies and the findings are mixed; while some studies have shown increased risk of Alzheimer’s disease in non-carriers, one study demonstrated the opposite and another study showed null effects. A recent meta-analysis of all these studies (Burgess et al., 2011), comprising more than 8000 people, suggests a protective effect of the T-allele. Moreover, recent research suggests that the hippocampal subfields are differentially associated with various neurological conditions. For example, CA1 is an early target site of pathology in Alzheimer’s disease (e.g., Kerchner, et al., 2012). Given the subfield specificity in the present study, future studies should explore further the complex relationship between KIBRA, cognition, and disease-related neuropathology within the hippocampal sub-regions. In particular, it would be useful to determine if the link between this gene and Alzheimer’s disease are independent of the effects of KIBRA on episodic memory. Moreover, it will be important to map the developmental trajectory of the effects of KIBRA on episodic memory and brain structure/function to determine if the effects accumulate with age.

In conclusion, the present data provide novel evidence in support of a genetic basis of AM. Humans evolved with the remarkable capacity to encode and subsequently “relive” autobiographical experiences. Until recently, there was little high quality neuroscience research on episodic AM. It is already understood that impaired AM is a hallmark sign of memory disorders, including amnesia and Alzheimer’s disease. Episodic AM is further relevant to

77 psychiatric disorders, especially depression and PTSD. In addition to the relevance to disease, understanding how individual differences in AM operate in non-diseased individuals could have major implications for the understanding of social and self-related mnemonic function.

References

Adamson, M. M., Landy, K. M., Duong, S., Fox-Bosetti, S., Ashford, J. W., Murphy, G. M., et al. (2010). Apolipoprotein E epsilon4 influences on episodic recall and brain structures in aging pilots. Neurobiology of Aging, 31, 1059-1063. doi: 10.1016/j.neurobiolaging.2008.07.017

Addis, D. R., Moscovitch, M., Crawley, A. P., & McAndrews, M. P. (2004). Recollective qualities modulate hippocampal activation during autobiographical memory retrieval. Hippocampus, 14, 752-762. doi: 10.1002/hipo.10215

Aggleton, J. P., & Brown, M. W. (1999). Episodic memory, amnesia, and the hippocampal- anterior thalamic axis. Behavioral and Brain Sciences, 22, 425-444; discussion 444-489. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/11301518

Aggleton, J. P., & Brown, M. W. (2006). Interleaving brain systems for episodic and recognition memory. Trends in Cognitive Sciences, 10, 455-463. doi: doi:10.1016/j.tics.2006.08.003

Alarcon, M., Plomin, R., Fulker, D. W., Corley, R., & DeFries, J. C. (1998). Multivariate path analysis of specific cognitive abilities data at 12 years of age in the Colorado Adoption Project. Behavior Genetics, 28, 255-264. doi: 10.1023/A:1021667213066

Ally, B. A., Hussey, E. P., & Donahue, M. J. (2012). A case of hyperthymesia: rethinking the role of the amygdala in autobiographical memory. Neurocase, 19, 166-81.doi: 10.1080/13554794.2011.654225

Almeida, O. P., Schwab, S. G., Lautenschlager, N. T., Morar, B., Greenop, K. R., Flicker, L., et al. (2008). KIBRA genetic polymorphism influences episodic memory in later life, but does not increase the risk of mild cognitive impairment. Journal of Cellular and Molecular Medicine, 12, 1672-1676. doi: JCMM229 [pii]10.1111/j.1582- 4934.2008.00229.x

Amaral, D. G., & Insausti, R. (1990). Hippocampal formation. In The human Nervous System, G. Paxinos (ed.). San Diego: Academic Press, pp. 711-755.

Andreasen, N. C., O'Leary, D. S., Cizadlo, T., Arndt, S., Rezai, K., Watkins, G. L., et al. (1995). Remembering the past: two facets of episodic memory explored with positron emission tomography. The American Journal of Psychiatry, 152, 1576-1585. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/7485619

Arndt, S., Cohen, G., Alliger, R. J., Swayze, V. W., 2nd, & Andreasen, N. C. (1991). Problems with ratio and proportion measures of imaged cerebral structures. Psychiatry Research, 40, 79-89. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/1946842

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). doi:10.1176/appi.books.9780890423349

Bakker, A., Krauss, G. L., Albert, M. S., Speck, C. L., Jones, L. R., Stark, C. E., et al. (2012). Reduction of hippocampal hyperactivity improves cognition in amnestic mild cognitive impairment. Neuron, 74, 467-474. doi: S0896-6273(12)00325-X [pii] 10.1016/j.neuron.2012.03.023

Bartsch, T., Dohring, J., Rohr, A., Jansen, O., & Deuschl, G. (2011). CA1 neurons in the human hippocampus are critical for autobiographical memory, mental time travel, and autonoetic consciousness. Proceedings of the National Academy of Sciences of the United States of America, 108, 17562-17567. doi: 10.1073/pnas.1110266108

Bassett, S. S., Yousem, D. M., Cristinzio, C., Kusevic, I., Yassa, M. A., Caffo, B. S., et al. (2006). Familial risk for Alzheimer's disease alters fMRI activation patterns. Brain: a Journal of Neurology, 129, 1229-1239. doi: 10.1093/brain/awl089

Bates, T. C., Price, J. F., Harris, S. E., Marioni, R. E., Fowkes, F. G., Stewart, M. C., et al. (2009). Association of KIBRA and memory. Neuroscience Letters, 458, 140-143. doi: 10.1016/j.neulet.2009.04.050

Bennett, D. A., Schneider, J. A., Wilson, R. S., Bienias, J. L., Berry-Kravis, E., & Arnold, S. E. (2005). Amyloid mediates the association of apolipoprotein E e4 allele to cognitive function in older people. Journal of Neurology, Neurosurgery, and Psychiatry, 76, 1194- 1199. doi: 76/9/1194 [pii]10.1136/jnnp.2004.054445

Bergouignan, L., Lemogne, C., Foucher, A., Longin, E., Vistoli, D., Allilaire, J. F., et al. (2008). Field perspective deficit for positive memories characterizes autobiographical memory in euthymic depressed patients. Behaviour Research and Therapy, 46, 322-333. doi: 10.1016/j.brat.2007.12.007

Bondi, M. W., Salmon, D. P., Galasko, D., Thomas, R. G., & Thal, L. J. (1999). Neuropsychological function and apolipoprotein E genotype in the preclinical detection of Alzheimer's disease. Psychology and Aging, 14, 295-303. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10403716

Bonnici, H. M., Chadwick, M. J., Kumaran, D., Hassabis, D., Weiskopf, N., & Maguire, E. A. (2012). Multi-voxel pattern analysis in human hippocampal subfields. Frontiers in Human Neuroscience, 6, 290. doi: 10.3389/fnhum.2012.00290

Bonnici, H. M., Chadwick, M. J., & Maguire, E. A. (2013). Representations of recent and remote autobiographical memories in hippocampal subfields. Hippocampus. doi: 10.1002/hipo.22155

Bowles, B., Crupi, C., Mirsattari, S. M., Pigott, S. E., Parrent, A. G., Pruessner, J. C., et al. (2007). Impaired familiarity with preserved recollection after anterior temporal-lobe resection that spares the hippocampus. Proceedings of the National Academy of Sciences of the United States of America, 104, 16382-16387. doi: 0705273104 [pii]10.1073/pnas.0705273104

Braskie, M. N., Small, G. W., & Bookheimer, S. Y. (2009). Entorhinal cortex structure and functional MRI response during an associative verbal memory task. Human Brain Mapping, 30, 3981-3992. doi: 10.1002/hbm.20823

Buchmann, A., Mondadori, C. R., Hanggi, J., Aerni, A., Vrticka, P., Luechinger, R., et al. (2008). Prion protein M129V polymorphism affects retrieval-related brain activity. Neuropsychologia, 46, 2389-2402. doi: 10.1016/j.neuropsychologia.2008.03.002

Buckner, R. L., & Carroll, D. C. (2007). Self-projection and the brain. Trends in Cognitive Sciences, 11, 49-57. doi: 10.1016/j.tics.2006.11.004

Burgess, J. D., Pedraza, O., Graff-Radford, N. R., Hirpa, M., Zou, F., Miles, R., et al. (2011). Association of common KIBRA variants with episodic memory and AD risk. Neurobiology of Aging, 32, 557 e551-559. doi: 10.1016/j.neurobiolaging.2010.11.004

Burianova, H., McIntosh, A. R., & Grady, C. L. (2010). A common functional brain network for autobiographical, episodic, and semantic memory retrieval. Neuroimage, 49, 865-874. doi: S1053-8119(09)00978-1 [pii]10.1016/j.neuroimage.2009.08.066

Buther, K., Plaas, C., Barnekow, A., & Kremerskothen, J. (2004). KIBRA is a novel substrate for protein kinase Czeta. Biochemical and Biophysical Research Communications, 317, 703- 707. doi: 10.1016/j.bbrc.2004.03.107

Cabeza, R., Prince, S. E., Daselaar, S. M., Greenberg, D. L., Budde, M., Dolcos, F., et al. (2004). Brain activity during episodic retrieval of autobiographical and laboratory events: an fMRI study using a novel photo paradigm. Journal of cognitive neuroscience, 16, 1583- 1594. doi:10.1162/0898929042568578

Cabeza, R., & St Jacques, P. (2007). Functional neuroimaging of autobiographical memory. Trends in Cognitive Sciences, 11, 219-227. doi: 10.1016/j.tics.2007.02.005

Carr, V. A., Rissman, J., & Wagner, A. D. (2010). Imaging the human medial temporal lobe with high-resolution fMRI. Neuron, 65, 298-308. doi: S0896-6273(09)01037-X [pii]10.1016/j.neuron.2009.12.022

Cavanna, A. E., & Trimble, M. R. (2006). The precuneus: a review of its functional anatomy and behavioural correlates. Brain: a Journal of Neurology, 129, 564-583. doi: 10.1093/brain/awl004

Chen, J., Olsen, R. K., Preston, A. R., Glover, G. H., & Wagner, A. D. (2011). Associative retrieval processes in the human medial temporal lobe: hippocampal retrieval success and CA1 mismatch detection. Learning and Memory, 18, 523-528. doi: 18/8/523 [pii]10.1101/lm.2135211

Cohen, R. M., Small, C., Lalonde, F., Friz, J., & Sunderland, T. (2001). Effect of apolipoprotein E genotype on hippocampal volume loss in aging healthy women. Neurology, 57, 2223- 2228. doi: 10.1212/WNL.57.12.2223

Collins, D. L., Holmes, C. J., Peters, T. M., & Evans, A. C. (1995). Automatic 3-D model-based neuroanatomical segmentation. Human Brain Mapping, 3, 190-208. doi: 10.1002/hbm.460030304

Conway, M. A. (2001). Sensory-perceptual episodic memory and its context: autobiographical memory. Philosophical transactions of the Royal Society of London. Series B, Biological sciences, 356, 1375-1384. doi: 10.1098/rstb.2001.0940

Conway, M. A., Gardiner, J. M., Perfect, T. J., Anderson, S. J., & Cohen, G. M. (1997). Changes in memory awareness during learning: the acquisition of knowledge by psychology undergraduates. Journal of Experimental Psychology. General, 126, 393-413. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/?term=9407649

Conway, M. A., & Pleydell-Pearce, C. W. (2000). The construction of autobiographical memories in the self-memory system. Psychological Review, 107, 261-288. doi: 10.1037/0033-295X.107.2.261

Corder, E. H., Saunders, A. M., Strittmatter, W. J., Schmechel, D. E., Gaskell, P. C., Small, G. W., et al. (1993). Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer's disease in late onset families. Science, 261, 921-923. doi: 10.1126/science.8346443

Costa, P. T., & McCrae, R. R. (1992). Revised NEO Personality lnventory (NEO-PI-R) and NEO Five-Factor Inventory (NEO-FFI) professional manual. Odessa, FL: Psychological Assessment Resources, Inc.

Cox, R. W., & Hyde, J. S. (1997). Software tools for analysis and visualization of fMRI data. NMR in Biomedicine, 10, 171-178. doi: 10.1002/(SICI)1099- 1492(199706/08)10:4/5<171::AID-NBM453>3.0.CO;2-L

Cui, Z., Wang, H., Tan, Y., Zaia, K. A., Zhang, S., & Tsien, J. Z. (2004). Inducible and reversible NR1 knockout reveals crucial role of the NMDA receptor in preserving remote memories in the brain. Neuron, 41, 781-793. doi: 10.1016/S0896-6273(04)00072-8

D'Argembeau, A., & Van der Linden, M. (2006). Individual differences in the phenomenology of mental time travel: The effect of vivid visual imagery and emotion regulation strategies. Consciousness and Cognition, 15, 342-350. doi: 10.1016/j.concog.2005.09.001

Das, S. R., Avants, B. B., Pluta, J., Wang, H., Suh, J. W., Weiner, M. W., et al. (2012). Measuring longitudinal change in the hippocampal formation from in vivo high- resolution T2-weighted MRI. Neuroimage, 60, 1266-1279. doi: S1053-8119(12)00115-2 [pii]10.1016/j.neuroimage.2012.01.098

Daselaar, S. M., Rice, H. J., Greenberg, D. L., Cabeza, R., LaBar, K. S., & Rubin, D. C. (2008). The spatiotemporal dynamics of autobiographical memory: neural correlates of recall, emotional intensity, and reliving. Cerebral Cortex, 18, 217-229. doi: 10.1093/cercor/bhm048

Davachi, L., & Wagner, A. D. (2002). Hippocampal contributions to episodic encoding: insights from relational and item-based learning. Journal of Neurophysiology, 88, 982-990. doi: 10.1152/jn.00046.2002

De Blasi, S., Montesanto, A., Martino, C., Dato, S., De Rango, F., Bruni, A. C., et al. (2009). APOE polymorphism affects episodic memory among non demented elderly subjects. Experimental Gerontology, 44, 224-227. doi: S0531-5565(08)00365-3 [pii]10.1016/j.exger.2008.11.005 de Quervain, D. J., Kolassa, I. T., Ackermann, S., Aerni, A., Boesiger, P., Demougin, P., et al. (2012). PKCalpha is genetically linked to memory capacity in healthy subjects and to risk for posttraumatic stress disorder in genocide survivors. Proceedings of the National Academy of Sciences of the United States of America, 109, 8746-8751. doi: 10.1073/pnas.1200857109 de Quervain, D. J., Kolassa, I. T., Ertl, V., Onyut, P. L., Neuner, F., Elbert, T., et al. (2007). A deletion variant of the alpha2b-adrenoceptor is related to emotional memory in Europeans and Africans. Nature Neuroscience, 10, 1137-1139. doi: nn1945 [pii]10.1038/nn1945 de Quervain, D. J., & Papassotiropoulos, A. (2006). Identification of a genetic cluster influencing memory performance and hippocampal activity in humans. Proceedings of the National Academy of Sciences of the United States of America, 103, 4270-4274. doi: 0510212103 [pii]10.1073/pnas.0510212103

Deary, I. J., Whiteman, M. C., Pattie, A., Starr, J. M., Hayward, C., Wright, A. F., et al. (2002). Cognitive change and the APOE epsilon 4 allele. Nature, 418, 932. doi: 10.1038/418932a418932a [pii]

Delis, D. C., Kramer, J. H., Kaplan, E., & Ober, B. A. (1987). California Verbal Learning: Adult version. Manual. San Antonio, TX: Psychological Corporation.

Dennis, N. A., Cabeza, R., Need, A. C., Waters-Metenier, S., Goldstein, D. B., & LaBar, K. S. (2011). Brain-derived neurotrophic factor val66met polymorphism and hippocampal activation during episodic encoding and retrieval tasks. Hippocampus, 21, 980-989. doi: 10.1002/hipo.20809

Diana, R. A., Yonelinas, A. P., & Ranganath, C. (2007). Imaging recollection and familiarity in the medial temporal lobe: a three-component model. Trends in Cognitive Sciences, 11, 379-386. doi: 10.1016/j.tics.2007.08.001

Dice, L. (1945). Measures of the amount of ecologic association between species. Ecology 26. Retrieved from http://www.jstor.org/stable/1932409

Duchaine, B. C., & Nakayama, K. (2006). Developmental prosopagnosia: a window to content- specific face processing. Current Opinion in Neurobiology, 16, 166-173. doi: S0959- 4388(06)00028-6 [pii]10.1016/j.conb.2006.03.003

Duning, K., Schurek, E. M., Schluter, M., Bayer, M., Reinhardt, H. C., Schwab, A., et al. (2008). KIBRA modulates directional migration of podocytes. Journal of the American Society of Nephrology : JASN, 19, 1891-1903. doi: 10.1681/ASN.2007080916

Duvernoy, H. (2005). Functional anatomy, vascularization and serial sections with MRI. In: The human hippocampus, Ed 3, p 232. Berlin: Springer.

Egan, M. F., Kojima, M., Callicott, J. H., Goldberg, T. E., Kolachana, B. S., Bertolino, A., et al. (2003). The BDNF val66met polymorphism affects activity-dependent secretion of BDNF and human memory and hippocampal function. Cell, 112, 257-269. doi: 0270- 6474/03/236690-05$15.00/0

Ehret, G. B. (2010). Genome-wide association studies: contribution of genomics to understanding blood pressure and essential hypertension. Current Hypertension Reports, 12, 17-25. doi: 10.1007/s11906-009-0086-6

Eichenbaum, H., Yonelinas, A. P., & Ranganath, C. (2007). The medial temporal lobe and recognition memory. Annual Review of Neuroscience, 30, 123-152. doi: 10.1146/annurev.neuro.30.051606.094328

Eslinger, P. J. (1998). Autobiographical memory after temporal lobe lesions. Neurocase, 4, 481- 495. doi:10.1080/13554799808410641

Finkel, D., Pedersen, N., & McGue, M. (1995). Genetic influences on memory performance in adulthood: comparison of Minnesota and Swedish twin data. Psychology and Aging, 10, 437-446. doi: 10.1037/0882-7974.10.3.437

Fletcher, P. C., Frith, C. D., Baker, S. C., Shallice, T., Frackowiak, R. S., & Dolan, R. J. (1995). The mind's eye-precuneus activation in memory-related imagery. Neuroimage, 2, 195- 200. doi: S1053-8119(85)71025-7 [pii]10.1006/nimg.1995.1025

Frankland, P. W., O'Brien, C., Ohno, M., Kirkwood, A., & Silva, A. J. (2001). Alpha-CaMKII- dependent plasticity in the cortex is required for permanent memory. Nature, 411, 309- 313. doi: 10.1038/35077089

Fuentemilla, L., Barnes, G. R., Duzel, E., & Levine, B. (2013). Theta oscillations orchestrate medial temporal lobe and neocortex in remembering autobiographical memories. Neuroimage. doi: S1053-8119(13)00889-6[pii]10.1016/j.neuroimage.2013.08.029

Fuentemilla, L., Camara, E., Munte, T. F., Kramer, U. M., Cunillera, T., Marco-Pallares, J., et al. (2009). Individual differences in true and false memory retrieval are related to white matter brain microstructure. Journal of Neuroscience, 29, 8698-8703. doi: 29/27/8698 [pii]10.1523/JNEUROSCI.5270-08.2009

Gilboa, A. (2004). Autobiographical and episodic memory--one and the same? Evidence from prefrontal activation in neuroimaging studies. Neuropsychologia, 42, 1336-1349. doi: 10.1016/j.neuropsychologia.2004.02.014

Gilboa, A., Winocur, G., Grady, C. L., Hevenor, S. J., & Moscovitch, M. (2004). Remembering our past: functional neuroanatomy of recollection of recent and very remote personal events. Cerebral Cortex, 14, 1214-1225. doi: 10.1093/cercor/bhh082

Gilboa, A., Winocur, G., Rosenbaum, R. S., Poreh, A., Gao, F., Black, S. E., et al. (2006). Hippocampal contributions to recollection in retrograde and anterograde amnesia. Hippocampus, 16, 966-980. doi: 10.1002/hipo.20226

Goldberg, T. E., Iudicello, J., Russo, C., Elvevag, B., Straub, R., Egan, M. F., et al. (2008). BDNF Val66Met polymorphism significantly affects d' in verbal recognition memory at short and long delays. Biological Psychology, 77, 20-24. doi: S0301-0511(07)00145-7 [pii]10.1016/j.biopsycho.2007.08.009

Grady, C. L., McIntosh, A. R., Beig, S., & Craik, F. I. (2001). An examination of the effects of stimulus type, encoding task, and functional connectivity on the role of right prefrontal cortex in recognition memory. Neuroimage, 14, 556-571. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/11506530

Hariri, A. R., Goldberg, T. E., Mattay, V. S., Kolachana, B. S., Callicott, J. H., Egan, M. F., et al. (2003). Brain-derived neurotrophic factor val66met polymorphism affects human memory-related hippocampal activity and predicts memory performance. Journal of Neuroscience, 23, 6690-6694. doi: 23/17/6690 [pii]

Hashimoto, R., Moriguchi, Y., Yamashita, F., Mori, T., Nemoto, K., Okada, T., et al. (2008). Dose-dependent effect of the Val66Met polymorphism of the brain-derived neurotrophic factor gene on memory-related hippocampal activity. Neuroscience Research, 61, 360- 367. doi: 10.1016/j.neures.2008.04.003

Hayashi, M. L., Choi, S. Y., Rao, B. S., Jung, H. Y., Lee, H. K., Zhang, D., et al. (2004). Altered cortical synaptic morphology and impaired memory consolidation in forebrain- specific dominant-negative PAK transgenic mice. Neuron, 42, 773-787. doi: 10.1016/j.neuron.2004.05.003

Hayashi, N., Kazui, H., Kamino, K., Tokunaga, H., Takaya, M., Yokokoji, M., et al. (2010). KIBRA genetic polymorphism influences episodic memory in Alzheimer's disease, but does not show association with disease in a Japanese cohort. Dementia and Geriatric Cognitive Disorders, 30, 302-308. doi: 000320482 [pii]10.1159/000320482

Helkala, E. L., Koivisto, K., Hanninen, T., Vanhanen, M., Kervinen, K., Kuusisto, J., et al. (1996). Memory functions in human subjects with different apolipoprotein E phenotypes during a 3-year population-based follow-up study. Neuroscience Letters, 204, 177-180. doi: 10.1016/0304-3940(96)12348-X

Herfurth, K., Kasper, B., Schwarz, M., Stefan, H., & Pauli, E. (2010). Autobiographical memory in temporal lobe epilepsy: role of hippocampal and temporal lateral structures. Epilepsy Behavior, 19, 365-371. doi: S1525-5050(10)00513-5 [pii]10.1016/j.yebeh.2010.07.012

Huang, Z. J., Kirkwood, A., Pizzorusso, T., Porciatti, V., Morales, B., Bear, M. F., et al. (1999). BDNF regulates the maturation of inhibition and the critical period of plasticity in mouse visual cortex. Cell, 98, 739-755. doi: S0092-8674(00)81509-3 [pii]

Huentelman, M. J., Papassotiropoulos, A., Craig, D. W., Hoerndli, F. J., Pearson, J. V., Huynh, K. D., et al. (2007). Calmodulin-binding transcription activator 1 (CAMTA1) alleles predispose human episodic memory performance. Human Molecular Genetics, 16, 1469- 1477. doi: 10.1093/hmg/ddm097

Insausti, R., Juottonen, K., Soininen, H., Insausti, A. M., Partanen, K., Vainio, P., et al. (1998). MR volumetric analysis of the human entorhinal, perirhinal, and temporopolar cortices. AJNR. American Journal of Neuroradiology, 19, 659-671.

Jacobsen, L. K., Picciotto, M. R., Heath, C. J., Mencl, W. E., & Gelernter, J. (2009). Allelic variation of calsyntenin 2 (CLSTN2) modulates the impact of developmental tobacco smoke exposure on mnemonic processing in adolescents. Biological Psychiatry, 65, 671- 679. doi: S0006-3223(08)01311-5 [pii]10.1016/j.biopsych.2008.10.024

Johannsen, S., Duning, K., Pavenstadt, H., Kremerskothen, J., & Boeckers, T. M. (2008). Temporal-spatial expression and novel biochemical properties of the memory-related protein KIBRA. Neuroscience, 155, 1165-1173. doi: 10.1016/j.neuroscience.2008.06.054

Kapur, N. (1999). Syndromes of retrograde amnesia: a conceptual and empirical synthesis. Psychological Bulletin, 125, 800-825. Retrieved at http://psycnet.apa.org/index.cfm?fa=buy.optionToBuy&id=1999-01567-009

Kauppi, K., Nilsson, L. G., Adolfsson, R., Eriksson, E., & Nyberg, L. (2011). KIBRA polymorphism is related to enhanced memory and elevated hippocampal processing. Journal of Neuroscience, 31, 14218-14222. doi: 10.1523/JNEUROSCI.3292-11.2011

Kendler, K. S., Gatz, M., Gardner, C. O., & Pedersen, N. L. (2006). Personality and major depression: a Swedish longitudinal, population-based twin study. Archives of General Psychiatry, 63, 1113-1120. doi: 10.1001/archpsyc.63.10.1113

Kensinger, E. A., & Corkin, S. (2003). Memory enhancement for emotional words: are emotional words more vividly remembered than neutral words? Memory and Cognition, 31, 1169-1180. doi: http://www.ncbi.nlm.nih.gov/pubmed/15058678

Kerchner, G. A., Deutsch, G. K., Zeineh, M., Dougherty, R. F., Saranathan, M., & Rutt, B. K. (2012). Hippocampal CA1 apical neuropil atrophy and memory performance in Alzheimer's disease. Neuroimage, 63, 194-202. doi: S1053-8119(12)00664-7 [pii]10.1016/j.neuroimage.2012.06.048

Kirchhoff, B. A., & Buckner, R. L. (2006). Functional-anatomic correlates of individual differences in memory. Neuron, 51, 263-274. doi: 10.1016/j.neuron.2006.06.006

Kopelman, M. D., Stanhope, N., & Kingsley, D. (1999). Retrograde amnesia in patients with diencephalic, temporal lobe or frontal lesions. Neuropsychologia, 37, 939-958. Retrieved at http://www.ncbi.nlm.nih.gov/pubmed/10426519

Kopelman, M. D., Wilson, B. A., & Baddeley, A. D. (1989). The autobiographical memory interview: a new assessment of autobiographical and personal semantic memory in amnesic patients. Journal of Clinical and Experimental Neuropsychology, 11, 724-744. doi: 10.1080/01688638908400928

Koppel, J., & Goldberg, T. (2009). The genetics of episodic memory. Cognitive Neuropsychiatry, 14, 356-376. doi: 10.1080/13546800902990438

Kremerskothen, J., Plaas, C., Buther, K., Finger, I., Veltel, S., Matanis, T., et al. (2003). Characterization of KIBRA, a novel WW domain-containing protein. Biochemical and Biophysical Research Communications, 300, 862-867. doi: S0006291X02029455 [pii]

Krishnan, A., Williams, L. J., McIntosh, A. R., & Abdi, H. (2011). Partial Least Squares (PLS) methods for neuroimaging: a tutorial and review. Neuroimage, 56, 455-475. doi: S1053- 8119(10)01007-4 [pii]10.1016/j.neuroimage.2010.07.034

Kukolja, J., Thiel, C. M., Eggermann, T., Zerres, K., & Fink, G. R. (2010). Medial temporal lobe dysfunction during encoding and retrieval of episodic memory in non-demented APOE epsilon4 carriers. Neuroscience, 168, 487-497. doi: 10.1016/j.neuroscience.2010.03.044

Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (2008). International affective picture system (IAPS): Affective ratings of pictures and instruction manual. Technical Report A-8. Gainesville, Florida: University of Florida.

Lemogne, C., Bergouignan, L., Boni, C., Gorwood, P., Pelissolo, A., & Fossati, P. (2009). Genetics and personality affect visual perspective in autobiographical memory. Consciousness and Cognition, 18, 823-830. doi: 10.1016/j.concog.2009.04.002

Lemogne, C., Piolino, P., Friszer, S., Claret, A., Girault, N., Jouvent, R., et al. (2006). Episodic autobiographical memory in depression: Specificity, autonoetic consciousness, and self- perspective. Consciousness and Cognition, 15, 258-268. doi: 10.1016/j.concog.2005.07.005

Lemogne, C., Piolino, P., Jouvent, R., Allilaire, J. F., & Fossati, P. (2006). [Episodic autobiographical memory in depression: a review]. L'Encephale, 32, 781-788.

Leport, A. K., Mattfeld, A. T., Dickinson-Anson, H., Fallon, J. H., Stark, C. E., Kruggel, F., et al. (2012). Behavioral and neuroanatomical investigation of Highly Superior Autobiographical Memory (HSAM). Neurobiology of Learning and Memory, 98, 78-92. doi: 10.1016/j.nlm.2012.05.002

Levine, B. (2004). Autobiographical memory and the self in time: brain lesion effects, functional neuroanatomy, and lifespan development. Brain and Cognition, 55, 54-68. doi: 10.1016/S0278-2626(03)00280-X

Levine, B., Black, S. E., Cabeza, R., Sinden, M., McIntosh, A. R., Toth, J. P., et al. (1998). Episodic memory and the self in a case of isolated retrograde amnesia. Brain: a Journal of Neurology, 12, 1951-1973.

Levine, B., Freedman, M., Dawson, D., Black, S., & Stuss, D. T. (1999). Ventral frontal contribution to self-regulation: Convergence of episodic memory and inhibition. Neurocase, 5, 263-275. doi:10.1080/13554799908402731

Levine, B., Svoboda, E., Hay, J. F., Winocur, G., & Moscovitch, M. (2002). Aging and autobiographical memory: dissociating episodic from semantic retrieval. Psychology and Aging, 17, 677-689. doi: 10.1037/0882-7974.17.4.677

Levine, B., Svoboda, E., Turner, G. R., Mandic, M., & Mackey, A. (2009). Behavioral and functional neuroanatomical correlates of anterograde autobiographical memory in isolated retrograde amnesic patient M.L. Neuropsychologia, 47, 2188-2196. doi: 10.1016/j.neuropsychologia.2008.12.026

Levine, B., Turner, G. R., Tisserand, D., Hevenor, S. J., Graham, S. J., & McIntosh, A. R. (2004). The functional neuroanatomy of episodic and semantic autobiographical remembering: a prospective functional MRI study. Journal of Cognitive Neuroscience, 16, 1633-1646. doi: 10.1162/0898929042568587

Libby, L. A., Ekstrom, A. D., Ragland, J. D., & Ranganath, C. (2012). Differential connectivity of perirhinal and parahippocampal cortices within human hippocampal subregions revealed by high-resolution functional imaging. Journal of Neuroscience, 32, 6550-6560. doi: 32/19/6550 [pii]10.1523/JNEUROSCI.3711-11.2012

Lu, B. (2003). BDNF and activity-dependent synaptic modulation. Learning and Memory, 10, 86-98. doi: 10.1101/lm.54603

Mackinger, H. F., Pachinger, M. M., Leibetseder, M. M., & Fartacek, R. R. (2000). Autobiographical memories in women remitted from major depression. Journal of Abnormal Psychology, 109, 331-334. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10895571

Maguire, E. A. (2001). Neuroimaging studies of autobiographical event memory. Philosophical transactions of the Royal Society of London. Series B, Biological Sciences, 356, 1441- 1451. doi: 10.1098/rstb.2001.0944

Maguire, E. A., & Mummery, C. J. (1999). Differential modulation of a common memory retrieval network revealed by positron emission tomography. Hippocampus, 9, 54-61. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10088900

Maguire, E. A., Vargha-Khadem, F., & Mishkin, M. (2001). The effects of bilateral hippocampal damage on fMRI regional activations and interactions during memory retrieval. Brain: a Journal of Neurology, 124, 1156-1170. doi: 10.1093/brain/124.6.1156

Mahley, R. W. (1988). Apolipoprotein E: cholesterol transport protein with expanding role in cell biology. Science, 240, 622-630. doi: 10.1126/science.3283935

Makuch, L., Volk, L., Anggono, V., Johnson, R. C., Yu, Y., Duning, K., et al. (2011). Regulation of AMPA receptor function by the human memory-associated gene KIBRA. Neuron, 71, 1022-1029. doi: 10.1016/j.neuron.2011.08.017

Markowitsch, H. J. (1995). Which brain regions are critically involved in the retrieval of old episodic memory? Brain research. Brain Research Reviews, 21, 117-127. doi: 10.1016/0165-0173(95)00007-0

Mayes, A., Montaldi, D., & Migo, E. (2007). Associative memory and the medial temporal lobes. Trends in Cognitive Sciences, 11, 126-135. doi: 10.1016/j.tics.2006.12.003

Mayo, P. R. (1983). Personality traits and the retrieval of positive and negative memories. Personality and Individual Differences, 4, 465-471. doi: 10.1016/0191-8869(83)90076-4

McClearn, G. E., Johansson, B., Berg, S., Pedersen, N. L., Ahern, F., Petrill, S. A., et al. (1997). Substantial genetic influence on cognitive abilities in twins 80 or more years old. Science, 276, 1560-1563. doi: 10.1126/science.276.5318.1560

McConachie, H. R. (1976). Developmental prosopagnosia. A single case report. Cortex, 12, 76- 82. Retrieved at http://www.ncbi.nlm.nih.gov/pubmed/1261287

McCormick, C., Moscovitch, M., Protzner, A. B., Huber, C. G., & McAndrews, M. P. (2010). Hippocampal-neocortical networks differ during encoding and retrieval of relational memory: functional and effective connectivity analyses. Neuropsychologia, 48, 3272- 3281. doi: 10.1016/j.neuropsychologia.2010.07.010

McDermott, K. B., Szpunar, K. K., & Christ, S. E. (2009). Laboratory-based and autobiographical retrieval tasks differ substantially in their neural substrates. Neuropsychologia, 47, 2290-2298. doi: 10.1016/j.neuropsychologia.2008.12.025

McFarlane, A. C. (1989). The aetiology of post-traumatic morbidity: predisposing, precipitating and perpetuating factors. The British Journal of Psychiatry: the journal of Mental Science, 154, 221-228. doi: 10.1192/bjp.154.2.221

McGuire, P. K., Paulesu, E., Frackowiak, R. S., & Frith, C. D. (1996). Brain activity during stimulus independent thought. Neuroreport, 7, 2095-2099. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8930966

McIntosh, A. R., Bookstein, F. L., Haxby, J. V., & Grady, C. L. (1996). Spatial pattern analysis of functional brain images using partial least squares. Neuroimage, 3, 143-157. doi: S1053-8119(96)90016-6 [pii]10.1006/nimg.1996.0016

McIntosh, A. R., & Lobaugh, N. J. (2004). Partial least squares analysis of neuroimaging data: applications and advances. Neuroimage, 23 Suppl 1, S250-263. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/15501095

McKinnon, M. C., Nica, E. I., Sengdy, P., Kovacevic, N., Moscovitch, M., Freedman, M., et al. (2008). Autobiographical memory and patterns of brain atrophy in frontotemporal lobar degeneration. Journal of Cognitive Neuroscience, 20, 1839-1853. doi: 10.1162/jocn.2008.20126

Miller, M. B., Donovan, C. L., Bennett, C. M., Aminoff, E. M., & Mayer, R. E. (2012). Individual differences in cognitive style and strategy predict similarities in the patterns of

brain activity between individuals. Neuroimage, 59, 83-93. doi: 10.1016/j.neuroimage.2011.05.060

Miller, M. B., Donovan, C. L., Van Horn, J. D., German, E., Sokol-Hessner, P., & Wolford, G. L. (2009). Unique and persistent individual patterns of brain activity across different memory retrieval tasks. Neuroimage, 48, 625-635. doi: 10.1016/j.neuroimage.2009.06.033

Miller, M. B., & Van Horn, J. D. (2007). Individual variability in brain activations associated with episodic retrieval: a role for large-scale databases. Int Journal of Psychophysiology, 63, 205-213. doi: S0167-8760(06)00108-5 [pii]10.1016/j.ijpsycho.2006.03.019

Miller, M. B., Van Horn, J. D., Wolford, G. L., Handy, T. C., Valsangkar-Smyth, M., Inati, S., et al. (2002). Extensive individual differences in brain activations associated with episodic retrieval are reliable over time. Journal of Cognitive Neuroscience, 14, 1200-1214. doi: 10.1162/089892902760807203

Mondadori, C. R., de Quervain, D. J., Buchmann, A., Mustovic, H., Wollmer, M. A., Schmidt, C. F., et al. (2007). Better memory and neural efficiency in young apolipoprotein E epsilon4 carriers. Cerebral Cortex, 17, 1934-1947. doi: 10.1093/cercor/bhl103

Moscovitch, M., Rosenbaum, R. S., Gilboa, A., Addis, D. R., Westmacott, R., Grady, C., et al. (2005). Functional neuroanatomy of remote episodic, semantic and spatial memory: a unified account based on multiple trace theory. Journal of Anatomy, 207, 35-66. doi: 10.1111/j.1469-7580.2005.00421.x

Mueller, S. G., Chao, L. L., Berman, B., & Weiner, M. W. (2011). Evidence for functional specialization of hippocampal subfields detected by MR subfield volumetry on high resolution images at 4 T. Neuroimage, 56, 851-857. doi: 10.1016/j.neuroimage.2011.03.028

Mueller, S. G., Laxer, K. D., Scanlon, C., Garcia, P., McMullen, W. J., Loring, D. W., et al. (2012). Different structural correlates for verbal memory impairment in temporal lobe epilepsy with and without mesial temporal lobe sclerosis. Human Brain Mapping, 33, 489-499. doi: 10.1002/hbm.21226

Mueller, S. G., Stables, L., Du, A. T., Schuff, N., Truran, D., Cashdollar, N., et al. (2007). Measurement of hippocampal subfields and age-related changes with high resolution MRI at 4T. Neurobiology of Aging, 28, 719-726. doi: S0197-4580(06)00091-1 [pii]10.1016/j.neurobiolaging.2006.03.007

Nacmias, B., Bessi, V., Bagnoli, S., Tedde, A., Cellini, E., Piccini, C., et al. (2008). KIBRA gene variants are associated with episodic memory performance in subjective memory complaints. Neuroscience Letters, 436, 145-147. doi: 10.1016/j.neulet.2008.03.008

Need, A. C., Attix, D. K., McEvoy, J. M., Cirulli, E. T., Linney, K. N., Wagoner, A. P., et al. (2008). Failure to replicate effect of Kibra on human memory in two large cohorts of European origin. American Journal of Medical Genetics. Part B, Neuropsychiatric

genetics: the official publication of the International Society of Psychiatric Genetics, 147B, 667-668. doi: 10.1002/ajmg.b.30658

Nilsson, L. G., Adolfsson, R., Backman, L., Cruts, M., Nyberg, L., Small, B. J., et al. (2006). The influence of APOE status on episodic and semantic memory: data from a population- based study. Neuropsychology, 20, 645-657. doi: 2006-20657-003 [pii]10.1037/0894- 4105.20.6.645

Northoff, G., Heinzel, A., de Greck, M., Bermpohl, F., Dobrowolny, H., & Panksepp, J. (2006). Self-referential processing in our brain--a meta-analysis of imaging studies on the self. Neuroimage, 31, 440-457. doi: 10.1016/j.neuroimage.2005.12.002

O'Reilly, R. C., & Rudy, J. W. (2001). Conjunctive representations in learning and memory: principles of cortical and hippocampal function. Psychological Review, 108, 311-345. doi: 10.1037/0033-295x.108.2.311

Ogawa, S., Lee, T. M., Kay, A. R., & Tank, D. W. (1990). Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proceedings of the National Academy of Sciences of the United States of America, 87, 9868-9872. Retrieved from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC55275/

Olsen, R. K., Moses, S. N., Riggs, L., & Ryan, J. D. (2012). The hippocampus supports multiple cognitive processes through relational binding and comparison. Frontiers in Human Neuroscience, 6, 146. doi: 10.3389/fnhum.2012.00146

Olsen, R. K., Nichols, E. A., Chen, J., Hunt, J. F., Glover, G. H., Gabrieli, J. D., et al. (2009). Performance-related sustained and anticipatory activity in human medial temporal lobe during delayed match-to-sample. Journal of Neuroscience, 29, 11880-11890. doi: 10.1523/JNEUROSCI.2245-09.2009

Olsen, R. K., Palombo, D. J., Rabin, J. S., Levine, B., Ryan, J. D., & Rosenbaum, R. S. (2013). Volumetric analysis of medial temporal lobe subregions in developmental amnesia using high-resolution magnetic resonance imaging. Hippocampus. doi: 10.1002/hipo.22153

Palombo, D. J., Williams, L. J., Abdi, H., & Levine, B. (2012). The survey of autobiographical memory (SAM): A novel measure of trait mnemonics in everyday life. Cortex. doi: 10.1016/j.cortex.2012.08.023

Papassotiropoulos, A., Stephan, D. A., Huentelman, M. J., Hoerndli, F. J., Craig, D. W., Pearson, J. V., et al. (2006). Common Kibra alleles are associated with human memory performance. Science, 314, 475-478. doi: 10.1126/science.1129837

Parker, E. S., Cahill, L., & McGaugh, J. L. (2006). A case of unusual autobiographical remembering. Neurocase, 12, 35-49. doi: 10.1080/13554790500473680

Pastalkova, E., Serrano, P., Pinkhasova, D., Wallace, E., Fenton, A. A., & Sacktor, T. C. (2006). Storage of spatial information by the maintenance mechanism of LTP. Science, 313, 1141-1144. doi: 10.1126/science.1128657

Piolino, P., Desgranges, B., Benali, K., & Eustache, F. (2002). Episodic and semantic remote autobiographical memory in ageing. Memory, 10, 239-257. doi: 10.1080/09658210143000353

Piolino, P., Desgranges, B., & Eustache, F. (2009). Episodic autobiographical memories over the course of time: cognitive, neuropsychological and neuroimaging findings. Neuropsychologia, 47, 2314-2329. doi: 10.1016/j.neuropsychologia.2009.01.020

Piolino, P., Lamidey, V., Desgranges, B., & Eustache, F. (2007). The semantic and episodic subcomponents of famous person knowledge: dissociation in healthy subjects. Neuropsychology, 21, 122-135. doi: 10.1037/0894-4105.21.1.122

Plassman, B. L., Welsh-Bohmer, K. A., Bigler, E. D., Johnson, S. C., Anderson, C. V., Helms, M. J., et al. (1997). Apolipoprotein E epsilon 4 allele and hippocampal volume in twins with normal cognition. Neurology, 48, 985-989. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9109888

Poppenk, J., & Moscovitch, M. (2011). A hippocampal marker of recollection memory ability among healthy young adults: contributions of posterior and anterior segments. Neuron, 72, 931-937. doi: 10.1016/j.neuron.2011.10.014

Preuschhof, C., Heekeren, H. R., Li, S. C., Sander, T., Lindenberger, U., & Backman, L. (2010). KIBRA and CLSTN2 polymorphisms exert interactive effects on human episodic memory. Neuropsychologia, 48, 402-408. doi: 10.1016/j.neuropsychologia.2009.09.031

Prvulovic, D., Van de Ven, V., Sack, A. T., Maurer, K., & Linden, D. E. (2005). Functional activation imaging in aging and dementia. Psychiatry Research, 140, 97-113. doi: 10.1016/j.pscychresns.2005.06.006

Raichle, M. E., MacLeod, A. M., Snyder, A. Z., Powers, W. J., Gusnard, D. A., & Shulman, G. L. (2001). A default mode of brain function. Proceedings of the National Academy of Sciences of the United States of America, 98, 676-682. doi: 10.1073/pnas.98.2.676

Rasch, B., Papassotiropoulos, A., & de Quervain, D. (2010). Imaging genetics of cognitive functions: Focus on episodic memory. Neuroimage. doi: S1053-8119(10)00005-4 [pii]10.1016/j.neuroimage.2010.01.001

Raskind, W. H., Peter, B., Richards, T., Eckert, M. M., & Berninger, V. W. (2012). The genetics of reading disabilities: from phenotypes to candidate genes. Frontiers in Psychology, 3, 601. doi: 10.3389/fpsyg.2012.00601

Rasmussen, A. S., & Berntsen, D. (2010). Personality traits and autobiographical memory: Openness is positively related to the experience and usage of recollections. Memory, 18, 774-786. doi: 10.1080/09658211.2010.514270

Renoult, L., Davidson, P. S., Palombo, D. J., Moscovitch, M., & Levine, B. (2012). Personal semantics: at the crossroads of semantic and episodic memory. Trends in Cognitive Sciences, 16, 550-558. doi: 10.1016/j.tics.2012.09.003

Rodrigue, K. M., & Raz, N. (2004). Shrinkage of the entorhinal cortex over five years predicts memory performance in healthy adults. Journal of Neuroscience, 24, 956-963. doi: 10.1523/JNEUROSCI.4166-03.2004

Rolls, E. T., & Kesner, R. P. (2006). A computational theory of hippocampal function, and empirical tests of the theory. Progress in Neurobiology, 79, 1-48. doi: 10.1016/j.pneurobio.2006.04.005

Rosen, A. C., Gabrieli, J. D., Stoub, T., Prull, M. W., O'Hara, R., Yesavage, J., et al. (2005). Relating medial temporal lobe volume to frontal fMRI activation for memory encoding in older adults. Cortex, 41, 595-602. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/16042035

Rosen, A. C., Prull, M. W., Gabrieli, J. D., Stoub, T., O'Hara, R., Friedman, L., et al. (2003). Differential associations between entorhinal and hippocampal volumes and memory performance in older adults. Behavioral Neuroscience, 117, 1150-1160. doi: 10.1037/0735-7044.117.6.1150

Rosenbaum, R. S., Moscovitch, M., Foster, J. K., Schnyer, D. M., Gao, F., Kovacevic, N., et al. (2008). Patterns of autobiographical memory loss in medial-temporal lobe amnesic patients. Journal of Cognitive Neuroscience, 20, 1490-1506. doi: 10.1162/jocn.2008.20105

Rubin, D. C., Boals, A., & Berntsen, D. (2008). Memory in posttraumatic stress disorder: properties of voluntary and involuntary, traumatic and nontraumatic autobiographical memories in people with and without posttraumatic stress disorder symptoms. Journal of Experimental Psychology. General, 137, 591-614. doi: 10.1037/a0013165

Rubin, D. C., Schrauf, R. W., & Greenberg, D. L. (2003). Belief and recollection of autobiographical memories. Memory and Cognition, 31, 887-901. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/14651297

Rubin, D. C., & Siegler, I. C. (2004). Facets of personality and the phenomenology of autobiographical memory. Applied Cognitive Psychology, 18, 913-930. doi: 10.1002/acp.1038

Rudebeck, S. R., Scholz, J., Millington, R., Rohenkohl, G., Johansen-Berg, H., & Lee, A. C. (2009). Fornix microstructure correlates with recollection but not familiarity memory. Journal of Neuroscience, 29, 14987-14992. doi: 29/47/14987 [pii]10.1523/JNEUROSCI.4707-09.2009

Ruiz-Caballero, J. A., & Bermudez, J. (1995). Neuroticism, mood, and retrieval of negative personal memories. The Journal of General Psychology, 122, 29-35. doi: 10.1080/00221309.1995.9921219

Ruiz-Canada, C., Ashley, J., Moeckel-Cole, S., Drier, E., Yin, J., & Budnik, V. (2004). New synaptic bouton formation is disrupted by misregulation of microtubule stability in aPKC mutants. Neuron, 42, 567-580. doi: 10.1016/s0896-6273(04)00255-7

Sampson, P. D., Streissguth, A. P., Barr, H. M., & Bookstein, F. L. (1989). Neurobehavioral effects of prenatal alcohol: Part II. Partial least squares analysis. Neurotoxicology and teratology, 11, 477-491. doi:10.1016/0892-0362(89)90025-1

Schaper, K., Kolsch, H., Popp, J., Wagner, M., & Jessen, F. (2008). KIBRA gene variants are associated with episodic memory in healthy elderly. Neurobiology of Aging, 29, 1123- 1125. doi: 10.1016/j.neurobiolaging.2007.02.001

Schmahmann, J. D., Pandya, D. N., Wang, R., Dai, G., D'Arceuil, H. E., de Crespigny, A. J., et al. (2007). Association fibre pathways of the brain: parallel observations from diffusion spectrum imaging and autoradiography. Brain, 130, 630-653. doi: 10.1093/brain/awl359

Schott, B. H., Niklas, C., Kaufmann, J., Bodammer, N. C., Machts, J., Schutze, H., et al. (2011). Fiber density between rhinal cortex and activated ventrolateral prefrontal regions predicts episodic memory performance in humans. Proceedings of the National Academy of Sciences of the United States of America, 108, 5408-5413. doi: 10.1073/pnas.1013287108

Sedille-Mostafaie, N., Sebesta, C., Huber, K. R., Zehetmayer, S., Jungwirth, S., Tragl, K. H., et al. (2012). The role of memory-related gene polymorphisms, KIBRA and CLSTN2, on replicate memory assessment in the elderly. Journal of Neural Transmission, 119, 77-80. doi: 10.1007/s00702-011-0667-9

Serrano, P., Yao, Y., & Sacktor, T. C. (2005). Persistent phosphorylation by protein kinase Mzeta maintains late-phase long-term potentiation. The Journal of Neuroscience, 25, 1979-1984. doi: 10.1523/JNEUROSCI.5132-04.2005

Sheldon, S., & Levine, B. (2013). Same as it ever was: Vividness modulates the similarities and differences between the neural networks that support retrieving remote and recent autobiographical memories. Neuroimage. doi: S1053-8119(13)00728-3 [pii]10.1016/j.neuroimage.2013.06.082

Sheldon, S., McAndrews, M. P., & Moscovitch, M. (2011). Episodic memory processes mediated by the medial temporal lobes contribute to open-ended problem solving. Neuropsychologia, 49, 2439-2447. doi: 10.1016/j.neuropsychologia.2011.04.021

Shema, R., Hazvi, S., Sacktor, T. C., & Dudai, Y. (2009). Boundary conditions for the maintenance of memory by PKMzeta in neocortex. Learning and Memory, 16, 122-128. doi: 16/2/122 [pii]10.1101/lm.1183309

Shipley, W. C. (1940). A self-administering scale for measuring intellectual impairment and deterioration. The Journal of Psychology, 371-377. doi: 10.1080/00223980.1940.9917704

Skinner, E. I., & Fernandes, M. A. (2007). Neural correlates of recollection and familiarity: a review of neuroimaging and patient data. Neuropsychologia, 45, 2163-2179. doi:10.1016/j.neuropsychologia.2007.03.007

Soderlund, H., Moscovitch, M., Kumar, N., Mandic, M., & Levine, B. (2012). As time goes by: Hippocampal connectivity changes with remoteness of autobiographical memory retrieval. Hippocampus, 22, 670-679. doi: 10.1002/hipo.20927

Squire, L. R., Stark, C. E., & Clark, R. E. (2004). The medial temporal lobe. Annual Review of Neuroscience, 27, 279-306. doi: 10.1146/annurev.neuro.27.070203.144130

Squire, L. R., Wixted, J. T., & Clark, R. E. (2007). Recognition memory and the medial temporal lobe: a new perspective. Nature Review Neuroscience, 8, 872-883. doi:10.1038/nrn2154

Stark, C. E., & Squire, L. R. (2001). When zero is not zero: the problem of ambiguous baseline conditions in fMRI. Proceedings of the National Academy of Sciences of the United States of America, 98, 12760-12766. doi: 10.1073/pnas.221462998

Steinvorth, S., Levine, B., & Corkin, S. (2005). Medial temporal lobe structures are needed to re- experience remote autobiographical memories: evidence from H.M. and W.R. Neuropsychologia, 43, 479-496. doi: 10.1016/j.neuropsychologia.2005.01.001

Svoboda, E., McKinnon, M. C., & Levine, B. (2006). The functional neuroanatomy of autobiographical memory: a meta-analysis. Neuropsychologia, 44, 2189-2208. doi: 10.1016/j.neuropsychologia.2006.05.023

Talairach, J., & Tournoux, P. (1988). Co-Planar Stereotactic Atlas of the Human Brain. Thieme, Stuttgart/New York.

Tambini, A., Ketz, N., & Davachi, L. (2010). Enhanced Brain Correlations during Rest Are Related to Memory for Recent Experiences. Neuron, 65, 280-290. doi: 10.1016/j.neuron.2010.01.001

Todd, R. M., Palombo, D. J., Levine, B., & Anderson, A. K. (2011). Genetic differences in emotionally enhanced memory. Neuropsychologia, 49, 734-744. doi: S0028- 3932(10)00478-1 [pii]10.1016/j.neuropsychologia.2010.11.010

Toro, R., Chupin, M., Garnero, L., Leonard, G., Perron, M., Pike, B., et al. (2009). Brain volumes and Val66Met polymorphism of the BDNF gene: local or global effects? Brain Struct Funct, 213, 501-509. doi: 10.1007/s00429-009-0203-y

Tulving, E. (1972) Episodic and semantic memory. In: Tulving, E. and Donaldson, W. (Eds.), Organization of Memory. Academic Press, New York, pp. 382–403.

Tulving, E. (1983). Elements of episodic memory. Oxford: Clarendon Press.

Tulving, E. (1985). Memory and consciousness. Canadian Psychologist, 25, 1-12. doi: 10.1037/h0080017

Tulving, E. (2002). Episodic memory: from mind to brain. Annual Review of Psychology, 53, 1- 25. doi: 10.1146/annurev.psych.53.100901.135114

Uttl, B. (2005). Measurement of individual differences: lessons from memory assessment in research and clinical practice. Psychological Science, 16, 460-467. doi: 10.1111/j.0956- 7976.2005.01557.x van der Kolk, B. A., & Fisler, R. (1995). Dissociation and the fragmentary nature of traumatic memories: overview and exploratory study. Journal of Traumatic Stress, 8, 505-525. doi: 10.1002/jts.2490080402

Van Petten, C. (2004). Relationship between hippocampal volume and memory ability in healthy individuals across the lifespan: review and meta-analysis. Neuropsychologia, 42, 1394- 1413. doi: 10.1016/j.neuropsychologia.2004.04.006

Vargha-Khadem, F., Gadian, D. G., Watkins, K. E., Connelly, A., Van Paesschen, W., & Mishkin, M. (1997). Differential effects of early hippocampal pathology on episodic and semantic memory. Science, 277, 376-380. doi: 10.1126/science.277.5324.376

Vassos, E., Bramon, E., Picchioni, M., Walshe, M., Filbey, F. M., Kravariti, E., et al. (2010). Evidence of association of KIBRA genotype with episodic memory in families of psychotic patients and controls. Journal of Psychiatric Research, 44, 795-798. doi: 10.1016/j.jpsychires.2010.01.012

Viskontas, I. V., McAndrews, M. P., & Moscovitch, M. (2000). Remote episodic memory deficits in patients with unilateral temporal lobe epilepsy and excisions. Journal of Neuroscience, 20, 5853-5857.

Volk, H. E., McDermott, K. B., Roediger, H. L., 3rd, & Todd, R. D. (2006). Genetic influences on free and cued recall in long-term memory tasks. Twin research and human genetics : the official journal of the International Society for Twin Studies, 9, 623-631. doi: 10.1375/183242706778553462

Wagner, A. K., Hatz, L. E., Scanlon, J. M., Niyonkuru, C., Miller, M. A., Ricker, J. H., et al. (2012). Association of KIBRA rs17070145 polymorphism and episodic memory in individuals with severe TBI. Brain Injury : [BI], 26, 1658-1669. doi: 10.3109/02699052.2012.700089

Wang, L., Laviolette, P., O'Keefe, K., Putcha, D., Bakkour, A., Van Dijk, K. R., et al. (2010). Intrinsic connectivity between the hippocampus and posteromedial cortex predicts memory performance in cognitively intact older individuals. Neuroimage, 51, 910-917. doi: S1053-8119(10)00214-4 [pii]10.1016/j.neuroimage.2010.02.046

Wang, L., Negreira, A., LaViolette, P., Bakkour, A., Sperling, R. A., & Dickerson, B. C. (2010). Intrinsic interhemispheric hippocampal functional connectivity predicts individual differences in memory performance ability. Hippocampus, 20, 345-351. doi: 10.1002/hipo.20771

Warren, Z., & Haslam, C. (2007). Overgeneral memory for public and autobiographical events in depression and schizophrenia. Cognitive Neuropsychiatry, 12, 301-321. doi: 779330582 [pii]10.1080/13546800601066142

Wersching, H., Guske, K., Hasenkamp, S., Hagedorn, C., Schiwek, S., Jansen, S., et al. (2011). Impact of common KIBRA allele on human cognitive functions. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology, 36, 1296-1304. doi: 10.1038/npp.2011.16

Wheeler, M. A., Stuss, D. T., & Tulving, E. (1997). Toward a theory of episodic memory: the frontal lobes and autonoetic consciousness. Psychological Bulletin, 121, 331-354. doi: 10.1037/0033-2909.121.3.331

Wheeler, M. E., & Buckner, R. L. (2004). Functional-anatomic correlates of remembering and knowing. Neuroimage, 21, 1337-1349. doi: 10.1016/j.neuroimage.2003.11.001S1053811903007213 [pii]

Wig, G. S., Grafton, S. T., Demos, K. E., Wolford, G. L., Petersen, S. E., & Kelley, W. M. (2008). Medial temporal lobe BOLD activity at rest predicts individual differences in memory ability in healthy young adults. Proceedings of the National Academy of Sciences of the United States of America, 105, 18555-18560. doi: 10.1073/pnas.0804546105

Williams, J. M., Barnhofer, T., Crane, C., Herman, D., Raes, F., Watkins, E., et al. (2007). Autobiographical memory specificity and emotional disorder. Psychological Bulletin, 133, 122-148. doi: 2006-23058-006 [pii]10.1037/0033-2909.133.1.122

Witte, A. V., Kurten, J., Jansen, S., Schirmacher, A., Brand, E., Sommer, J., et al. (2012). Interaction of BDNF and COMT polymorphisms on paired-associative stimulation- induced cortical plasticity. Journal of Neuroscience, 32, 4553-4561. doi: 10.1523/JNEUROSCI.6010-11.2012

Witter, M. P., & Amaral, D. G. (1991). Entorhinal cortex of the monkey: V. Projections to the dentate gyrus, hippocampus, and subicular complex. The Journal of Comparative Neurology, 307, 437-459. doi: 10.1002/cne.903070308

Wixted, J. T. (2007). Dual-process theory and signal-detection theory of recognition memory. Psychological Review, 114, 152-176. doi: 2006-23341-006 [pii]10.1037/0033- 295X.114.1.152

Wixted, J. T., Mickes, L., & Squire, L. R. (2010). Measuring recollection and familiarity in the medial temporal lobe. Hippocampus, 20, 1195-1205. doi: 10.1002/hipo.20854

Yassa, M. A., Lacy, J. W., Stark, S. M., Albert, M. S., Gallagher, M., & Stark, C. E. (2011). Pattern separation deficits associated with increased hippocampal CA3 and dentate gyrus activity in nondemented older adults. Hippocampus, 21, 968-979. doi: 10.1002/hipo.20808

Yasuda, Y., Hashimoto, R., Ohi, K., Fukumoto, M., Takamura, H., Iike, N., et al. (2010). Association study of KIBRA gene with memory performance in a Japanese population. The world journal of biological psychiatry : the official journal of the World Federation of Societies of Biological Psychiatry, 11, 852-857. doi: 10.3109/15622971003797258

Yonelinas, A. (2002). The Nature of Recollection and Familiarity: A Review of 30 Years of Research. Journal of Memory and Language, 46, 441-517. doi: 10.1006/jmla.2002.2864

Yonelinas, A. P. (1994). Receiver-operating characteristics in recognition memory: evidence for a dual-process model. Journal of Experimental Psychology. Learning, Memory, and Cognition, 20, 1341-1354. doi: 10.1037/0278-7393.20.6.1341

Yonelinas, A. P. (2001). Components of episodic memory: the contribution of recollection and familiarity. Philosophical transactions of the Royal Society of London. Series B, Biological sciences, 356, 1363-1374. doi: 10.1098/rstb.2001.0939

Yonelinas, A. P., Aly, M., Wang, W. C., & Koen, J. D. (2010). Recollection and familiarity: examining controversial assumptions and new directions. Hippocampus, 20, 1178-1194. doi: 10.1002/hipo.20864

Yoshihama, Y., Chida, K., & Ohno, S. (2012). The KIBRA-aPKC connection: A potential regulator of membrane trafficking and cell polarity. Communicative & Integrative Biology, 5, 146-151. doi: 10.4161/cib.18849

Yoshihama, Y., Hirai, T., Ohtsuka, T., & Chida, K. (2009). KIBRA Co-localizes with protein kinase Mzeta (PKMzeta) in the mouse hippocampus. Bioscience, Biotechnology, and Biochemistry, 73, 147-151. doi: JST.JSTAGE/bbb/80564 [pii]

Zachary, R. A. (1986). Shipley Institute of Living Scale: Revised Manual. Los Angeles: Western Psychological Services.

Zeineh, M. M., Engel, S. A., & Bookheimer, S. Y. (2000). Application of cortical unfolding techniques to functional MRI of the human hippocampal region. Neuroimage, 11, 668- 683. doi: 10.1006/nimg.2000.0561S1053-8119(00)90561-5 [pii]

Table 1. Number of participants is shown for each KIBRA genotype group (CC and TT/TC) for each of the tasks used in Chapter 2 (IAPS; International Affective Picture System).

Number of Participants CC TT/TC

Objects Task 107 121 IAPS Task 117 144 Neutral Autobiographical Memory 77 96 Negative Autobiographical Memory 77 88 K-estimate 129 152

Table 2. Means and standard deviations are shown for each KIBRA genotype group (CC and TT/TC) for the objects task (FA; false alarms) used in Chapter 2.

Objects Task

CC TT/TC

P (Hits) 0.72 (0.10) 0.72 (0.12) P (FA) 0.25 (0.09) 0.24 (0.09) d' 1.31 (0.37) 1.34 (0.44) Recollection 0.19 (0.19) 0.17 (0.20) Familiarity 1.12 (0.67) 1.24 (0.62)

100

Table 3. Means and standard deviations are shown for each KIBRA genotype group (CC and TT/TC) for the International Affective Picture System (IAPS) task used in Chapter 2 (FA; false alarms).

IAPS Task

CC TT/TC

Neutral P (Hits) 0.74 (0.20) 0.78 (0.19) P (FA) 0.10 (0.17) 0.08 (0.11)

Negative P (Hits) 0.82 (0.16) 0.85 (0.14) P (FA) 0.16 (0.18) 0.15 (0.15)

Positive P (Hits) 0.80 (0.18) 0.79 (0.16) P (FA) 0.20 (0.22) 0.18 (0.22)

101

Behavioural Measures in Scanner

CC TT/TC

Task 1 - Episodic Laboratory Memory Performance P (Hits) 0.81 (0.09) 0.80 (0.10) P (Recollection) 0.59 (0.16) 0.60 (0.17) P (Familiarity) 0.41 (0.16) 0.40 (0.17) P (FA) 0.17 (0.12) 0.16 (0.15) d' 2.12 (0.80) 2.13 (0.70)

Reaction Times Hits 679.58 (161.80) 658.55 (141.97) Recollection 746.44 (156.42) 666.78 (213.62) Familiarity 821.93 (319.71) 727.23 (142.50) FA 617.16 (310.41) 766.17 (373.91)

Task 2 - Episodic Autobiographical Memory Re-experiencing Ratings Recent 6.75 (0.52) 6.25 (0.99) Remote 4.64 (1.42) 5.06 (1.07) Odd/Even 2.09 (1.07) 1.99 (1.12)

102

Table 5. Cluster maxima from the spatiotemporal Task Partial Least Squares analysis used in Chapter 3, comparing autobiographical memory to the odd/even control task between KIBRA genotype groups (CC and TT/TC; L; left, R; right, BA; brodmann area, Lat; laterality, Anat; anatomy). Size indicates the number of contiguous voxels in the cluster. Spatiotemporal Task Partial Least Squares Lag Lat Anat BA Talairach coordinates Size Bootstrap ratio x y z Saliences/Bootstrap ratios 1 L Cingulate Gyrus 31 -12 -52 24 75 -9.07 2 L Precuneus 31 -8 -60 20 828 -17.83 L Middle Temporal Gyrus 39 -48 -76 24 166 -11.95 L Middle Temporal Gyrus 21 -56 -12 -8 41 -11.24 L Superior Frontal Gyrus 9 -8 52 36 294 -11.03 R Cerebellum (Culmen) 24 -28 -20 114 -10.74 L Inferior Frontal Gyrus 47 -44 28 -8 121 -10.24 L Parahippocampal Gyrus 36 -36 -28 -20 78 -10.11 R Middle Temporal Gyrus 19 44 -84 20 24 -9.98 R Middle Temporal Gyrus 21 60 -4 -12 28 -9.92 R Inferior Frontal Gyrus 47 36 32 -8 53 -8.82 L Middle Temporal Gyrus 22 -52 -44 0 16 -8.18 L Medial Frontal Gyrus 8 -16 28 40 15 -7.86 R Superior Temporal Gyrus 38 48 16 -24 13 -7.6 3 R Superior Temporal Gyrus 42 68 -28 20 11 8.78 L Precuneus 31 -4 -64 20 1742 -20.67 R Medial Frontal Gyrus 10 4 56 20 486 -13.61 L Superior Frontal Gyrus 8 -8 16 52 17 -11.48 L Middle Temporal Gyrus 21 -56 -16 -12 49 -11.2 R Middle Temporal Gyrus 21 52 0 -16 134 -10.71 L Parahippocampal Gyrus 28 -20 -20 -20 140 -10.64 L Inferior Frontal Gyrus 47 -52 28 -12 145 -9.85 L Middle Temporal Gyrus 21 -56 -40 -4 40 -9.55 L Cerebellum (Tonsil) 0 -52 -40 23 -8.24 L Middle Frontal Gyrus 6 -48 12 44 10 -8.21 4 R Superior Temporal Gyrus 42 64 -28 16 14 7.63 R Posterior Cingulate 30 12 -60 16 775 -16.94 L Medial Frontal Gyrus 11 0 40 -12 486 -13.42 L Middle Occipital Gyrus 19 -32 -92 16 180 -11.94 L Parahippocampal Gyrus 35 -20 -20 -12 132 -11.12 L Inferior Frontal Gyrus 47 -40 24 -16 124 -10.52

103

R Middle Temporal Gyrus 21 60 -4 -12 46 -9.17 L Superior Frontal Gyrus 8 -8 16 48 38 -8.94 L Fusiform Gyrus 19 -20 -84 -12 22 -8.63 L Middle Frontal Gyrus 6 -40 8 52 21 -8.32 L Middle Temporal Gyrus 22 -52 -40 0 19 -8.11 R Precuneus 19 32 -84 36 42 -8.01 R Lingual Gyrus 18 16 -84 -8 54 -7.6 L Middle Temporal Gyrus 21 -56 -16 -12 13 -7.36 R Cerebellum (Semi-Lunar) 28 -76 -36 13 -7.34 5 L Posterior Cingulate 30 -16 -56 8 447 -13.92 L Superior Occipital Gyrus 19 -36 -80 28 168 -12.58 L Medial Frontal Gyrus 8 -16 32 40 366 -11.61 R Superior Occipital Gyrus 19 40 -84 28 27 -9.45 R Parahippocampal Gyrus 35 20 -24 -16 46 -9.4 L Superior Frontal Gyrus 8 -8 16 52 21 -8.71 R Middle Temporal Gyrus 21 60 -4 -12 22 -8.19 R Cerebellum (Tonsil) 4 -56 -44 15 -7.93 L Cerebellum (Culmen) -16 -40 -12 42 -7.77 L Inferior Frontal Gyrus 45 -56 28 8 12 -7.77 L Middle Frontal Gyrus 6 -40 8 56 19 -7.36 R Cerebellum (Semi-Lunar) 32 -76 -36 10 -7.04 L Posterior Cingulate 30 -16 -60 12 319 -13.77 6 L Superior Occipital Gyrus 19 -36 -80 32 143 -10.94 R Superior Occipital Gyrus 19 40 -84 28 31 -10.66 L Medial Frontal Gyrus 10 0 56 -12 189 -10.17 L Superior Frontal Gyrus 6 -8 12 52 147 -9.38 R Cerebellum (Semi-Lunar) 32 -76 -36 33 -8.12 L Cerebellum (Culmen) -20 -36 -12 16 -7.82 R Cerebellum (Culmen) 24 -28 -20 20 -7.57 R Middle Temporal Gyrus 21 60 -4 -12 11 -6.78 7 L Posterior Cingulate 23 -4 -60 16 273 -13.64 L Angular Gyrus 39 -36 -76 32 103 -9.96 L Medial Frontal Gyrus 10 -4 60 8 121 -9.19 R Cerebellum (Tonsil) 8 -52 -44 14 -8.75 R Cerebellum (Pyramis) 40 -72 -32 17 -8.29 L Superior Frontal Gyrus 8 -24 28 48 61 -8.23 L Superior Frontal Gyrus 8 -8 16 52 15 -7.9 L Superior Frontal Gyrus 8 -8 52 40 20 -7.65 L Middle Frontal Gyrus 46 -56 24 24 13 -7.61

104

Table 6. Cluster maxima from the spatiotemporal seed partial least squares analysis used in Chapter 3, comparing recent and remote autobiographical memory between KIBRA genotype groups (CC and TT/TC; L; left, R; right, BA; brodmann area, Lat; laterality, Anat; anatomy). Size indicates the number of contiguous voxels in the cluster.

Spatiotemporal Seed Partial Least Squares Lag Lat Anat BA Talairach coordinates Size Bootstrap ratio x y z Negative Saliences/Bootstrap ratios 1 L Cerebellum (Culmen) -24 -36 -16 23 -11.24 2 L Cerebellum (Culmen) -16 -44 -20 332 -9.93 R Superior Occipital Gyrus 19 36 -76 24 20 -8.75 R Middle Temporal Gyrus 21 60 -16 -12 30 -7.4 L Precuneus 7 -16 -52 52 32 -6.91 3 L Cerebellum (Culmen) -24 -36 -16 420 -11.25 R Superior Frontal Gyrus 9 8 56 32 32 -10.74 L Precuneus 7 0 -36 44 319 -10.42 R Parahippocampal Gyrus 35 24 -24 -24 201 -10.27 L Cingulate Gyrus 24 -8 -4 36 83 -8.85 L Inferior Occipital Gyrus 19 -40 -72 0 215 -7.86 L Superior Temporal Gyrus 38 -44 16 -28 28 -7.31 R Superior Occipital Gyrus 19 40 -76 24 24 -7.16 R Superior Temporal Gyrus 13 56 -40 20 54 -7.13 L Middle Occipital Gyrus 19 -28 -76 20 43 -7.13 Lentiform Nucleus L (Putamen) -12 12 -4 32 -7.06 R Middle Temporal Gyrus 21 64 -16 -12 24 -6.91 L Superior Temporal Gyrus 22 -48 -4 0 39 -6.91 Lentiform Nucleus L (Putamen) -32 -20 0 24 -6.8 R Postcentral Gyrus 2 40 -24 44 31 -6.6 4 R Parahippocampal Gyrus 35 20 -24 -16 605 -14.24 L Parahippocampal Gyrus 35 -20 -12 -24 721 -12.65 L Caudate (Head) -4 20 4 23 -9.26 L Cuneus 18 -8 -100 8 39 -9.24 R Insula 13 44 -40 16 52 -9.1 R Superior Frontal Gyrus 9 8 56 32 31 -8.7 R Inferior Frontal Gyrus 45 56 12 20 26 -8.22

105

R Middle Temporal Gyrus 39 48 -68 20 33 -7.88 R Inferior Temporal Gyrus 20 60 -20 -16 31 -7.8 L Thalamus (Lateral Dorsal Nucleus) -12 -20 16 22 -7.31 L Middle Occipital Gyrus 19 -52 -68 -8 20 -6.69 5 R Cerebellum (Declive) 4 -68 -20 79 -10.91 R Parahippocampal Gyrus 35 20 -20 -8 128 -10.44 L Parahippocampal Gyrus 35 -16 -20 -12 121 -8.97 R Middle Temporal Gyrus 21 60 -20 -12 22 -8.48 R Superior Frontal Gyrus 9 8 56 32 24 -8.18 R Middle Temporal Gyrus 19 52 -64 16 34 -7.4 R Thalamus 8 -28 -4 82 -7.25 R Cerebellum (Pyramis) 20 -84 -32 28 -7.2 L Thalamus -4 -12 16 33 -6.62 L Inferior Temporal Gyrus 37 -52 -60 -8 23 -6.57 6 R Hippocampus 32 -8 -20 412 -11.13 R Middle Temporal Gyrus 21 60 -20 -8 36 -7.99 R Superior Frontal Gyrus 9 8 56 28 21 -7.61 L Middle Temporal Gyrus 22 -48 -48 0 44 -7.1 R Cerebellum (Pyramis) 20 -84 -32 21 -6.6 7 R Parahippocampal Gyrus 28 20 -20 -16 843 -13.29 L Cingulate Gyrus 24 -8 -4 32 25 -8.32 R Superior Frontal Gyrus 9 12 60 28 21 -7.58 L Middle Occipital Gyrus 37 -56 -64 -8 20 -7.03 R Superior Temporal Gyrus 21 56 -16 -4 40 -6.95 L Superior Temporal Gyrus 38 -44 4 -12 20 -6.59

106

Table 7. Dice reliability values for intrarater (intra) and interrater reliability (inter) for Chapter 4 for the hippocampus and medial temporal lobes (MTL). Reliability values are shown for the left (L) and right (R) hemispheres. Within the hippocampus, CA1, DG/CA2/3, and subiculum were segmented (CA; cornu ammonis, DG; dentate gyrus; Sub; Subiculum). Within MTL cortex, perirhinal cortex (PRC), entorhinal cortex (ERC), and parahippocampal cortex (PHC) were segmented.

Reliability

Intra Inter L R L R CA1 0.85 0.86 0.76 0.7 DG/CA2/3 0.89 0.9 0.84 0.81 Sub 0.8 0.77 0.7 0.67 PRC 0.84 0.88 0.73 0.77 ERC 0.83 0.85 0.69 0.72 PHC 0.92 0.91 0.78 0.83

107

MTL Subregions

Hippocampus

* * CA1 DG/CA2/3 Sub

CC TT/TC CC TT/TC CC TT/TC

1029 1099 L 594 (18) 659 (15) (35) (41) 505 (18) 476 (16) 1079 1201 R 595 (29) 692 (26) (37) (47) 491 (18) 470 (10)

MTL Cortices

PRC ERC PHC+

CC TT/TC CC TT/TC CC TT/TC

2904 3254 1100 1074 1972 2194 L (159) (254) (50) (32) (124) (80) 2611 2647 1035 1040 1898 2102 R (163) (170) (51) (44) (86) (59)

108

Figure 1. A schematic depicting factors that relate to individual differences in episodic memory capacity.

109

110

111

Figure 4. Coronal slice through the medial temporal lobes (MTL) on T1-weighted images for one representative participant depicting the MTL mask used for the functional imaging analysis for Chapter 3.

112

Figure 5. Results of the conjunction analysis (TT/TC versus CC) for the laboratory memory task (i.e., faces task) for the contrast remember versus know for Chapter 3. BOLD response is overlaid on an average of participants’ T1-weighted images (MTL; medial temporal lobes).

113

114

115

116

Figure 9. Pattern of brain activity for the spatiotemporal Task Partial Least Squares analysis in each genotype group (CC and TT/TC) for Chapter 3 for the autobiographical memory conditions (recent, remote) versus the odd/even control task. A. Colored bars represent brain scores, which represent the extent to which each experimental condition relates to the differences in hemodynamic response in each group with 95% confidence intervals plotted, which denote the standard error of the bootstrap ratios (BSR). The brain scores for each memory condition are considered statistically reliable if the error bars do not cross 0. B. BSRs for the LV represented in A, depicted on coronal images from an average of participants’ T1- weighted images. Interpretation of the relationship between the bars and the direction of change in hemodynamic response in areas reliably associated with this LV requires consideration of the brain scores; whereas positive saliences (i.e., areas of activation that are displayed with warm colors, or regions with positive BSR) indicate areas that are relatively more active in conditions with positive brain scores, negative saliences indicate areas relatively more active in conditions with negative brain scores (i.e., areas of activation that are displayed with cool colors, or regions with negative BSR).

117

118

Figure 11. Pattern of results for the spatiotemporal Seed Partial Least Squares analysis (i.e., spatiotemporal hippocampal [HC] connectivity analysis) for each genotype group (CC and TT/TC) for Chapter 3 for the autobiographical memory conditions (recent, remote) for the left and right HC. A. Colored bars represent HC- brain correlations, which represent the extent to which each experimental condition relates to the differences in hemodynamic response in each group with 95% confidence intervals plotted, which denote the standard error of the bootstrap ratios (BSR). The correlations for each memory condition are considered statistically reliable if the error bars do not cross 0. B. BSRs for the LV represented in A, depicted on coronal images from an average of participants’ T1-weighted images. Interpretation of the relationship between the bars and the direction of change in hemodynamic response in areas reliably associated with this LV requires consideration of the brain scores: negative saliences indicate areas relatively more active in conditions with negative brain scores (i.e., areas of activation that are displayed with cool colors, or regions with negative BSR).

119

Figure 12. Sagittal plane of T1-weighted (1 mm3) images depicting correct prescription for acquisition of high-resolution images through the medial temporal lobes for Chapter 4.

120

Figure 13. Coronal plane of T2-weighted (0.4 x 0.4 mm) images depicting correct prescription of high-resolution images through the medial temporal lobes, which were used in Chapter 4.

121

Figure 14. T2-weighted (0.4 x 0.4 mm) images depicting hippocampal subfields.

122

Figure 15. Slices from T2-weighted (0.4 x 0.4 mm) images through the medial temporal lobes (MTL) for one representative participant. The left panel depicts a 3D-rendering of the hippocampus; middle and right panels show coronal slices through MTL. Subfields were drawn where a clear ‘C-shape’ was discernible, which included all of the body but extended into the most posterior head slices (the remaining head slices and the entire tail included all subfields; see e.g., Zeineh, 2000; Olsen 2009). Demarcation varied across the long axis (see Amaral and Insausti, 1990). Anteriorly, the lateral (superior) boundary of CA1was drawn by bisecting the most lateral undulation of hippocampus (i). Moving posteriorly, the CA1 was drawn 3/4 of the way up the lateral bend of hippocampus and its medial extension bisected the DG/CA2/3 regions (iii). In the most posterior slices of the body, CA1 was drawn 3/4 of the way up the lateral bend of the hippocampus, and its medial extension was drawn in line with the medial extent of the ‘tear-drop’ (iv) shaped DG/CA2/3. Regions extending superior and medial to the CA1 were taken as DG/CA2/3 (Zeineh, Engel, & Bookheimer, 2000). Anteriorly, the medial portion of the subiculum extended until the elbow of the isthmus (v) and in more posterior slices the medial subicular border was drawn halfway down the bend of the isthmus (vi). Perirhinal cortex (PRC) entorhinal cortex, and parahippocampal cortex (PHC) were segmented according to Insausti et al. (1998) (R; right, S; superior, CA; cornu ammonis, DG; dentate gyrus).