Brain Struct Funct DOI 10.1007/s00429-015-1084-x

ORIGINAL ARTICLE

Distinct hippocampal functional networks revealed by tractography-based parcellation

1 2,5 3 2,4 Areeba Adnan • Alexander Barnett • Massieh Moayedi • Cornelia McCormick • 2,5 2,5 Melanie Cohn • Mary Pat McAndrews

Received: 27 February 2015 / Accepted: 7 July 2015 Ó Springer-Verlag Berlin Heidelberg 2015

Abstract Recent research suggests the anterior and pos- parahippocampal, inferior temporal and fusiform gyri and terior hippocampus form part of two distinct functional the , predicted interindividual relational neural networks. Here we investigate the structural under- performance. These findings provide important support for pinnings of this functional connectivity difference using the integration of structural and functional connectivity in diffusion-weighted imaging-based parcellation. Using this understanding the networks underlying episodic technique, we substantiated that the hippocampus can be memory. parcellated into distinct anterior and posterior segments. These structurally defined segments did indeed show dif- Keywords DTI Á fMRI Á Hippocampus Á Recognition ferent patterns of resting state functional connectivity, in that memory Á Networks the anterior segment showed greater connectivity with temporal and , whereas the posterior segment was more highly connected to medial and lateral Introduction parietal cortex. Furthermore, we showed that the posterior hippocampal connectivity to memory processing regions, There is a growing consensus for the specialization of hip- including the dorsolateral , pocampal function along the anterior–posterior axis, arising out of differential involvement in large-scale brain networks (Poppenk et al. 2013). Functional neuroimaging has shown A. Adnan, A. Barnett and M. Moayedi contributed equally to this work and share first authorship. that the anterior and posterior hippocampi (antHC, postHC) are nodes within two distinct intrinsic networks: the antHC is Electronic supplementary material The online version of this functionally connected to perirhinal, lateral temporal, and article (doi:10.1007/s00429-015-1084-x) contains supplementary material, which is available to authorized users. ventromedial prefrontal cortices, and the postHC is con- nected to regions of the default mode network (DMN) & Areeba Adnan including medial and lateral parietal areas, medial prefrontal [email protected] cortex (PFC), and posterior cingulate cortices (PCC) (Kahn

1 et al. 2008; Poppenk and Moscovitch 2011; Libby et al. Department of Psychology, York University, Toronto, ON M3J 1P3, Canada 2012). This distinction has also found support in studies examining co-activation of regions during memory retrieval. 2 Krembil Center and Toronto Western Research Institute, University Health Network, Toronto, ON M5T 2S8, The postHC has been shown to be a key node of the recol- Canada lection memory network, a group of regions including the 3 Department of Neuroscience, Physiology and Pharmacology, parahippocampal (PHG), retrosplenial posterior cin- University College London, London WC1E 6BT, UK gulate cortex (rsPCC), and lateral parietal cortices that are 4 Centre for Developmental Cognitive Neuroscience, preferentially engaged when individuals are able to recall University College London, London WC1E 3BG, UK rich contextual details about events (e.g., Henson et al. 1999; 5 Department of Psychology, University of Toronto, Toronto, Eldridge et al. 2000; Yonelinas et al. 2005; Rugg and Vil- ON M5S 3G3, Canada berg, 2013). In contrast, the antHC is considered to be a node 123 Brain Struct Funct in the lateral temporal network, which includes the perirhi- investigate the cingulate, parietal, and temporoparietal nal/entorhinal and anterior lateral temporal cortices (Kahn junction, frontal polar cortices (Beckmann et al. 2009; et al. 2008; Libby et al. 2012; Poppenk and Moscovitch Johansen-Berg et al. 2004; Mars et al. 2011, 2012; 2011), subserving memory processes including familiarity Tomassini et al. 2007; Moayedi et al. 2014). Our analysis or the sense of prior occurrence in the absence of recall of involved four steps: (1) probabilistic tractography seeded other qualitative details (e.g., Yonelinas et al. 2005; Ran- from the hippocampus to the rest of the brain, (2) deter- ganath 2010). mination of the number of spatially consistent subregions Distinct structural connectivity profiles of the antHC and using a K-means clustering algorithm of the tractographic postHC could play an important role in supporting the data in each subject, (3) characterization of the functional distinct functional engagement of these regions that connectivity of these subregions, and (4) examination of underlie different memory processes (Honey et al. 2009). the relationship between functional connectivity of these Indeed, the antHC has direct connections with the amyg- subregions and memory performance on a verbal recogni- dala (Duvernoy 2005), and to the temporal pole, insula and tion memory task. ventromedial PFC via the uncinate fasciculus (Kier et al. 2004; Catenoix et al. 2011), while the postHC is connected Data acquisition to frontal and parietal neo-cortical regions via a polysy- naptic pathway involving the fornix projections to mam- All subjects provided informed consent to procedures millary bodies, anterior nucleus of the thalamus, and approved by the UHN Research Ethics Board. Diffusion- anterior cingulum (Duvernoy 2005; for review see Poppenk weighted images (DWI) were acquired for 15 healthy et al. 2013). While behavioral and functional neuroimaging subjects (5 women and 10 men; mean ± SD age: findings suggest the presence of functional subregions 34.33 ± 11.9 years) on a 3T Sigma MR System (GE along the longitudinal axis of the hippocampus, the struc- Medical Systems, Milwaukee). DWI were acquired along tural basis of such divisions has not been previously 25 non-colinear diffusion-encoding directions investigated in humans. Thus, we used diffusion-weighted (b = 1000 s/mm2). For each acquisition, one B0 imaging (DWI) to elucidate the structural connectivity of (b = 0 mm/s2) volume was acquired at the beginning of the hippocampus in vivo to assess whether there are dis- the run, and the parameters were as follows: cernible subregions within the hippocampus along the TR = 12,000 ms, FOV = 24 9 24 cm2, 128 9 128 longitudinal axis. We hypothesized that the functional matrix, 1.875 9 1.875 mm2 in-plane resolution, 3-mm heterogeneity attributed to the antHC and postHC arises thick, 46 axial slices. from differences in their structural connectivity. Also, a whole brain high-resolution anatomical scan was These structural connections can be estimated using also acquired for each subject using a 3D fast spoiled tractography (Johansen-Berg et al. 2004), which we applied gradient echo (FSPGR) sequence. The parameters were as to identify two hippocampal regions with unique white follows: flip angle: 12°,TR= 7.88 ms, TE = 3.08 ms, matter connectivity to the rest of the brain. We then 120 sagittal slices, FOV: 22 9 22 cm2, 256 9 256 matrix, investigated whether these structurally defined subregions 0.9 9 0.9 9 1 mm voxels, 1-mm thick. were differentially functionally connected to the rest of the Task-free rs-fMRI T2*-weighted functional MRI scans brain based on resting state fMRI (rs-fMRI). Finally, we with an echo-planar pulse imaging (EPI) sequence were examined the relationship between the intrinsic functional also acquired for every subject. The parameters were as connectivity of hippocampal segments with performance follows: TR = 2000 ms, TE = 25 ms, FOV: on a memory task. Specifically, we tested the hypothesis 24 9 24 cm2,649 64 matrix, 3.75 9 3.75 9 5 voxels, that connectivity of the postHC would predict relational 5-mm thick, 28–32 slices depending on head size, for 180 memory (Poppenk and Moscovitch 2011), which we volumes. Subjects were instructed to lie still, clear their operationalized as the performance advantage obtained on thoughts and ‘‘not to think about anything in particular,’’ a verbal recognition task when the study context was with their eyes open (the entire field of view was black). reinstated compared to when decisions were based on single items (Cohn et al. 2009a, b). Anatomical parcellation using DWI

Seed region definition Materials and methods A hippocampal mask for each hemisphere was selected We investigated the structural connectivity of the hip- based on the Harvard-Oxford subcortical atlas (Desikan pocampi to determine whether there are subregions within et al. 2006), available in FSL v4.0 (Jenkinson et al. 2012). this structure with an approach that has been used to The probabilistic mask was then thresholded to ensure full 123 Brain Struct Funct

Fig. 1 Hippocampal masks used for tractography-based parcellation. The left hippocampal mask is shown in the top panel (green) and right hippocampal mask is shown in the bottom panel (red). All masks are displayed on the MNI152 brain coverage of the hippocampus while eliminating voxels Cluster-based parcellation outside the structure such as in the or the PHG. Based on this requirement, we thresholded at 62 % for each DWI data were preprocessed using tools from FDT, part of hemisphere (Fig. 1). The hippocampus mask was trans- FSL. Motion and eddy current correction were performed formed to individual space using the linear registration tool using affine registration of all volumes to a target b0 vol- (FLIRT) implemented in FSL, using 12 degrees of freedom ume—the second acquired volume. Probability density (Jenkinson et al. 2002), and visually checked for aberrant functions on up to 2 principal fiber directions were esti- registration (See Supplementary Fig. 1). All registrations mated at each voxel in the brain using the Bayesian Esti- were satisfactory upon visual inspection. mation of Diffusion Parameters obtained using sampling

123 Brain Struct Funct techniques toolbox (BEDPOSTX; Behrens et al. 2007) unwarped. The subject’s anatomical image was segmented implemented in FSL. and normalized to the T1-weighted MNI152, and the nor- We parcellated the left and right hippocampi using malization parameters were then written to the functional previously described methods (Beckmann et al. 2009; data to ensure alignment between anatomical and functional Johansen-Berg et al. 2004; Mars et al. 2011, 2012; data. Finally, fMRI data were smoothed using a Gaussian Moayedi et al. 2014) in the ‘‘ccops’’ toolbox implemented kernel of 8-mm full-width half-maximum (FWHM). in FSL. This type of analysis has been shown to be suc- Resting state analysis was performed using the Conn cessful with as few as 12 gradient directions (Behrens et al. toolbox (Whitfield-Gabrieli and Nieto-Castanon) as fol- 2007) and with as few as eight participants (Klein et al. lows. Using SPM8, an MNI152-T1-weighted image was 2007). For each subject, a connectivity matrix between segmented for gray matter, white matter, and cerebrospinal voxels in each hippocampus and all voxels of the whole fluid (CSF). The data were also temporally preprocessed to brain was computed (Johansen-Berg et al. 2004). The control for other potential confounds, and to restrict the matrices consist of rows indicating each hippocampal analysis to frequencies of interest (\0.1 Hz). These steps voxel, and columns representing each voxel in the rest of include using linear regression to remove potential sources the brain. The values in each matrix element represent the of noise, including estimated subject motion parameters (3 connectivity value (i.e., number of streamlines) of the translation components and 3 rotation components) and hippocampal voxel and the brain voxel. We then generated removal of BOLD covariates in white matter and CSF a symmetric cross-correlation matrix of dimensions equal areas with the anatomical component based noise correc- to the number of seed voxels in each hippocampus. The (i, tion method (aCompCor; Behzadi et al. 2007). Next, the j)th element within the matrix represents the correlation time series of voxels within the antHC and postHC (ROIs between the connectivity profile of voxel i and the con- provided by K-means clustering of DWI data at K = 2, nectivity profile voxel j. We then permuted the rows of the subtracting out overlapping voxels) for each hemisphere cross-correlation matrix using a K-means clustering seg- was averaged and correlated with the time series of every mentation algorithm, implemented in the ccops toolbox in other voxel in the brain. Correlation values were then FSL, for automated clustering to define K different clusters. converted to z scores with a Fisher transformation. We included a scaled Euclidian distance matrix to the To examine the functional connectivity fingerprints of cross-correlation matrix (Tomassini et al. 2007), imposing the antHC and postHC, we conducted second-level random a weak distance constraint of 0.2, as has been previously effect analyses of antHC and postHC functional connec- used (Beckmann et al. 2009; Mars et al. 2011, 2012; tivity. Statistical maps of connectivity were thresholded Tomassini et al. 2007). with an explicit mask to constrain our analysis to a priori The K-means clustering algorithm requires the experi- defined regions known to be connected to the antHC and menter to provide the number of clusters to be resolved. postHC (Andrews-Hanna et al. 2010; Kahn et al. 2008; We used an iterative approach (Beckmann et al. 2009; Poppenk and Moscovitch 2011).This mask was constructed Mars et al. 2011) in which the number of clusters that are using the Harvard-Oxford cortical and subcortical atlases searched for within the K-means clustering algorithm is (Harvard Center for Morphometric Analysis; http://www. gradually increased as long as it is possible to identify the cma.mgh.harvard.edu/fsl_atlas.html) and included bilateral same component clusters in each subject. In the present anterior and posterior cingulate cortex, the study, clustering was performed with three different values paracingulate gyrus, frontal orbital cortex, frontal pole, for K (K = 2, 3, and 4 clusters) in separate analyses. Our a subcallosal cortex, , , supra- priori hypothesis was that the hippocampus would divide marginal gyrus, lateral occipital cortex, precuneus, cuneal into 2 subregions, anterior and posterior segments, along cortex anterior division of the inferior temporal, middle the longitudinal axis. We also tested K = 3 and 4 to temporal and , temporal pole, ascertain the stability of the subregions in our participants. amygdala, thalamus, accumbens, anterior and posterior division of the , temporal fusiform rs-fMRI cortex, and hippocampi. These regions were thresholded at 20 % probability of being present in the normative data Data analysis provided by the atlas (See Supplementary Fig. 3). Statis- tical maps were thresholded at a height threshold of Preprocessing procedures were performed with Statistical z = 1.96 and an extent threshold of p \ 0.05, cluster cor- Parametric Mapping 8 (SPM8; http://www.fil.ion.ucl.ac.uk/ rected. Finally, using these masks, we contrasted the antHC spm/software/spm8), a toolbox running in MATLAB v8 12.0 connectivity against postHC connectivity, at a height (Mathworks). Functional images were co-registered to the threshold of z = 1.96 and an extent threshold of p \ 0.05), subject’s anatomical image, and spatially realigned and cluster corrected. 123 Brain Struct Funct

Hippocampal functional connectivity correlations Furthermore, the order of the pairs was randomly shuffled with verbal memory task across participants. For the purpose of our study, we were interested in Next, we sought to test whether the patterns of functional investigating the relationship between intrinsic functional connectivity for the hippocampal subregions during rest connectivity and behavioral performance. To do so we were related to memory performance on different compo- included both sets of behavioral data acquired from the nents of the associative verbal memory task. The experi- fMRI and EEG sessions. Using data for this task, we cal- mental memory task was designed to evaluate relational culated an index for single item memory (SIM) which retrieval versus item recognition, by asking individuals to included all hits to single items (ratings 4, 5, and 6 for study adjective–noun pairs (e.g., happy mother) and make targets) corrected for false alarms by subtracting these recognition decisions on the noun (mother) versus a novel (ratings 4, 5, and 6 for lures). Note that this index includes distractor (e.g., baker), first without cues and second in the guesses (4 ratings). We then averaged this measure of SIM context of the same (happy) or different (grumpy) adjective across both testing sessions for each subject. Second, we cues. Using the Medical Research Council (MRC) psy- calculated a measure of relational memory, which was cholinguistic database, 188 adjectives and 248 nouns were based on the advantage typically seen for recognizing a selected that depict fictional people to create two alternate target item when its original encoding context is repro- versions of the task equated for word frequency (each to be duced, termed reinstatement memory (RIM). This measure presented during two separate testing sessions). For each included only items with high confidence ‘‘old’’ responses version, 90 adjective–noun pairs were presented for (rating 6) as we know these are more likely to reflect 4500 ms during the study phase and participants were recollection processes (Yonelinas et al. 2010). RIM was instructed to rate whether it was easy or hard to imagine calculated as the proportion of high confidence hits to such a character. Recognition testing took place approxi- targets when presented in their studied pairing (intact mately 2–5 min after the encoding session ended. This time pairs) versus when presented alone (single targets). Again, was used to repeat instructions for the participants to we averaged across both testing sessions for each partici- ensure accurate understanding of the task and recheck pant to obtain a RIM index. Note that while the reinstate- setup for EEG (electrode signals) and fMRI (setup of next ment effect is likely to capture the same memory processes sequence by MRI technician). Participants were first pre- as recollection, as both are sensitive to retrieval of context sented with a single noun and indicated whether the item (Cohn et al. 2009a, b), the SIM index cannot be considered was new or old using a 6-point Likert confidence rating a specific index of familiarity as both types of recognition scale (1 very sure new to 6 very sure old). This was fol- processes (recollection and familiarity) could contribute to lowed by a cued trial in which the same noun was pre- hits for this condition. sented with its studied adjective (intact pair) or a studied We sought to test whether intrinsic functional connec- adjective originally paired with another noun (rearranged tivity of the hippocampal subregions predicted memory pair when the test noun was previously studied and new– performance. Thus, we performed four second-level anal- old pair when the test noun was a lure). Again, participants yses correlating the functional connectivity of each subre- were required to make a judgment solely about the noun gion to the other voxels within the ROI mask with using the same 6-point Likert confidence rating scale, measures of memory performance: antHC connectivity irrespective of which adjective was paired with it. In total, with SIM and RIM and postHC connectivity with SIM and the test phase included 60 studied single nouns and 30 RIM. Age was entered as a covariate of no interest. This single lures, as well as 30 intact, 30 rearranged, and 30 old– produced maps relating subregion functional connectivity new pairs. to individual differences in each type of memory. The Each participant completed two versions of the task on threshold for determining statistical significance was set at two different testing sessions: one during an fMRI session, a height threshold of z = 1.96 with a cluster-corrected and one during an EEG session. Given the different time threshold of p \ 0.05 within our a priori mask. resolution of these techniques, the timing of the two ver- sions differed at test. Test item presentation was fixed to 5500 ms for the fMRI and was self-paced during the EEG, Results but in both modalities, each item was followed by a jittered fixation ranging from 500 to 4500 ms. The experiment was Tractography-based parcellation programmed using E-prime software (Psychology Software Tools Inc). Prior to each version of the task, participants The first goal was to structurally parcellate the hippocam- completed a practice study and test runs. For each session, pus using DWI. The two-cluster solution (K = 2) resulted different sets of 90 adjective–noun pairs were used. in a consistent anterior and posterior segment across all 123 Brain Struct Funct subjects. The K = 3 and K = 4 solutions revealed seg- between the bilateral postHC and a network including the mentation patterns that were much more variable across right precuneus, anterior cingulate cortex, superior frontal subjects (see Supplementary Fig. 2). Hence, we deemed gyrus, and frontal pole was positively correlated with RIM the parcellations obtained from K = 2 as acceptable, with (p B 0.05, corrected). In contrast, there were no significant the following center of gravity for each segment and cre- functional connectivity patterns with the antHC to RIM or ated an average segment for each across all participants (all of the postHC to SIM, underscoring the specificity of the co-ordinates in MNI space (X, Y, Z); right antHC: (24, 13, RIM-postHC network relationship. We also observed a -21); right postHC: (29, -26, -9); left antHC: (-22, pattern of connectivity associated with SIM performance -12, -20); left postHC: (27, 25, -12). The results of the and the antHC seed, an inverse correlation with connec- parcellation for K = 2, 3, and 4 averaged across the group tivity to regions including right frontal pole, middle frontal are shown in Fig. 2. gyrus, , and lateral occipital cortex After averaging, the anterior and posterior segment (p B 0.05, corrected). Table 2 and Fig. 4 provide details obtained from K = 2 were found to be overlapping by 49 for these findings. Behavioral data for the encoding and test voxels in the left hemisphere and 20 voxels in the right phase of the task are detailed in Supplementary Table 1 hemisphere. The overlapping region in each hemisphere and behavioral performance for the two memory indices, was subtracted to give non-overlapping anterior and pos- RIM and SIM, is detailed in Supplementary Table 1. terior clusters for both hemispheres and the resulting antHC and postHC segments for each hemisphere were used as seeds in our resting state connectivity analysis. Discussion

Hippocampal rs-fMRI connectivity The aims of this study were (1) to investigate whether the hippocampus can be subdivided along its longitudinal axis The second goal was to examine the resting state functional based on structural connectivity using DWI, (2) to char- connectivity of the hippocampal subregions based on DWI- acterize the subregion connectivities at rest, and (3) to based parcellation. Our analyses (Table 1; Fig. 3) demon- determine whether these are related to memory perfor- strated that the left antHC is functionally connected to the mance. To our knowledge, this is the first study to inves- right antHC and amygdala, and bilateral frontal pole, and tigate the structural basis of anterior and posterior divisions PHG (cluster corrected, p \ 0.05). The right antHC dis- of the human hippocampus and to delineate these subre- played a similar pattern of functional connectivity with the gions in a data-driven manner. We provide evidence for contralateral antHC, right PHG and orbitofrontal cortex reliable anterior–posterior hippocampal segmentation in a (OFC), and bilateral medial OFC and perigenual anterior sample of healthy adults. In accordance with other recent cingulate cortex (cluster corrected, p \ 0.05). studies using a priori, manually defined anterior and pos- The left postHC was functionally connected with the terior parcellation of the hippocampus (Poppenk and ipsilateral PHG and rsPCC, contralateral postHC and PHG, Moscovitch 2011; Vincent et al. 2008; Andrews-Hanna and bilateral thalamus (cluster corrected, p \ 0.05). Simi- et al. 2010; Kahn et al. 2008; Voets et al. 2012), we found larly, the right postHC showed significant connectivity that these subregions had distinct patterns of intrinsic with the right PHG, left postHC and PHG, and bilateral functional connectivity. Finally, we provide novel evidence rsPCC (cluster corrected, p \ 0.05). that not only do these subregions co-activate with distinct In a direct contrast of the connectivity patterns of the cortical regions during different memory tasks, but, cru- two subregions (Table 1; Fig. 3), we found greater con- cially, the posterior region is shown as a critical component nectivity for the antHC to bilateral amygdala, right tem- of a resting state network that can predict memory per- poral pole, and right OFC (cluster corrected, p \ 0.05). We formance. Altogether, our results provide a structural basis report greater postHC connectivity to the left dorsal PCC, for the functional division of the hippocampus, and provide rsPCC, and PHG and right , novel evidence that the subregion connectivities are related supramarginal gyrus lateral occipital cortex and frontal to distinct memory task performance. pole. Diffusion-based parcellation of the hippocampi rs-fMRI connectivity and memory performance Previous studies of anatomical connectivity in animals and Finally, the third goal of this study was to examine the functional connectivity in humans have identified differ- relationship between intrinsic functional connectivity of ences along the longitudinal axis of the hippocampus. In the hippocampal subregions with measures of memory rodents, the dorsal hippocampus (posterior analog to performance. We found that the functional connectivity humans) is more connected with visuospatial and other 123 Brain Struct Funct

Fig. 2 Group-level results of parcellation solutions for the bilateral shown in yellow, the middle hippocampus is shown in green, and the hippocampi shown on the MNI152 brain. Group maps (n = 15) of the postHC is shown in blue. The antHC had a COG at (-22, -13, -21) K-means clustering algorithm parcellations images were created on the left and at (-23, -12, -20) on the right; the middle based on areas of the hippocampus that showed overlap across hippocampus had a COG at (-27, -20, -16) on the left and at (29, subjects. The images are thresholded at 62 % (9/15) of all subjects for -18, -17) on the right; and the postHC had a COG at the left and right hippocampi. Views of the left (L) hippocampal (-27,-30,-8.34) on the left and at (29.5, -28.5, -8.5) on the right. parcellations are shown in sagittal view in the top panel. The right The K = 4 solution, the antHC is shown in yellow, the middle (R) hippocampal parcellations are shown in sagittal view in the hippocampus is shown in green, the middle-anterior hippocampus is middle panel and both are shown in coronal views in the bottom shown in pink, and the postHC is shown in blue. The antHC had a panel. Images are in radiological convention. For the K = 2 solution, COG at (-25.5, -12, -21) on the left, and at (25, -12, -20) on the the anterior hippocampus (antHC) is shown in yellow, and the right. The middle hippocampus had a COG at (-26, -20, -17) on the posterior hippocampus (postHC) is shown in blue. The left antHC had left, and the COG at (29, -19, -15) on the right. The middle-anterior a center of gravity (COG) at (-22, -12, -20) and the right antHC hippocampus had a COG at (-19, -13, -19) on the left and the COG had a COG at (24, 13, -21). The left postHC had a COG at (27, 25, at (23, -14, -19) on the right. The postHC had a COG at (-27, -31, -12), and the right postHC had a COG at (29, -26, -9). COG and all -7) on the left and the COG at (28, -30, -6) on the right co-ordinates are in MNI space. In the K = 3 solution, the antHC is

123 Brain Struct Funct

Table 1 Peak functional Anatomical region BA Hemisphere Peak MNI co-ordinate z score connectivity between brain and hippocampal parcels derived XYZ from K-means clustering Anterior hippocampal connectivity Left seed Hippocampus – L -26 -18 -18 6.45 Parahippocampus 28 L -20 -26 -20 5.30 Hippocampus – R 26 -14 -16 5.48 Amygdala – R 28 -6 -14 4.94 Parahippocampus 28/34 R 22 -14 -24 4.57 Frontal pole 10 R 4 58 -10 4.44 Frontal pole 10 L -854-6 3.60 Right seed Hippocampus – R 18 -12 -16 6.33 Hippocampus – L -26 -16 -24 5.91 Parahippocampus 28 R 30 -16 -24 5.66 Orbital frontal cortex 11 m/14r R 14 44 -22 4.23 Orbital frontal cortex 32/14 m L/R 0 50 -10 3.98 PG/ACC 32 R 2 58 -12 3.41 PG/ACC 32/14 m L -440-6 3.39 Posterior hippocampal connectivity Left seed Hippocampus – L -28 -20 -14 7.16 Parahippocampus 34 L -22 -30 -12 6.53 Hippocampus – R 22 -22 -18 5.96 Parahippocampus 34 R 22 -34 -20 5.14 29/30 L -2 -56 14 4.51 Thalamus – L -14 -34 2 4.08 Thalamus – R 16 -28 0 3.63 Right seed Retrosplenial cortex 29/30 L -16 -58 4 5.97 Hippocampus – R 26 -18 -14 5.82 Retrosplenial cortex 29/30 R 18 -54 14 5.47 Hippocampus – L -24 -22 -18 5.41 Parahippocampus – R 18 -28 -16 5.24 Parahippocampus – L -20 -28 -16 4.86 Posterior cingulate 23/31 L -8 -44 36 3.15 Anterior vs. posterior connectivity Anterior [ posterior Hippocampus – R 18 -12 -16 4.79 Amygdala – R 20 -4 -16 4.7 Temporal pole 38 R 42 12 -32 4.67 Orbital frontal cortex 14c R 20 34 -18 4.5 Amygdala – L -32 -4 -20 5.85 Hippocampus – L -22 -6 -22 5.27 Anterior \ posterior dPCC 23 L -4 -26 28 4.46 Frontal pole 46 R 56 36 10 4.33 Parahippocampus 34/27 L -18 -34 -4 4.28 Retrosplenial cortex 29/30 L -8 -46 8 4.25 Supramarginal gyrus 40 R 52 -40 42 4.28

123 Brain Struct Funct

Table 1 continued Anatomical region BA Hemisphere Peak MNI co-ordinate z score XY Z

Angular gyrus 7 R 46 -54 38 3.72 Sup parietal lobule 7 R 38 -50 42 3.3 ACC anterior cingulate cortex, BA , dPCC dorsal posterior cingulate cortex, L left, MNI Montreal Neurological Institute, PG paracingulate gyrus, R right, Sup superior

Left Right mPFC mPFC Amyg Thal rsPCC

Ant HC mPFC HC/PHG mPFC OFC HC/PHG HC/PHG TP X = 0 Y = -24 Z= -6 X = 6 Y = 10 Z = -20

rsPCC/PCu rsPCC/PCu Thal Thal Thal

Post HC HC/PHG rsPCC/PCu HC/PHG rsPCC/PCu X = -8 Y = -24 Z= 10 X = 10 Y = -28 Z = 8 [R Ant + L Ant] [R Post + L Post]

OFC AG/SMG FP Amyg OFC rsPCC/PCu HC/PHG TP TP TP Y = 14 X = 50 X = -8 X = 30 Z= -12

-5 -1.96 score 1.96 5

Fig. 3 Hippocampal subregion resting state functional connectivity. differences (p \ 0.05, corrected) in the connectivity of both anterior Top panel Functional connectivity of the left and right anterior versus both posterior hippocampi parcels. All images are displayed in hippocampus. Both the left and right are connected to the lateral radiological convention. AG angular gyrus, Amyg amygdala, Ant temporal network and midline structures related to the default mode anterior, FP frontal pole, HC hippocampus, L left, mPFC medial network. Middle panel Functional connectivity of the left and right prefrontal cortex, OFC orbital frontal cortex, PCu precuneus, PHG posterior hippocampus. Both the left and right are connected to the parahippocampal gyrus, Post posterior, R right, rsPCC retrosplenial nodes of the recollection memory network and midline structures cortex, SMG supramarginal gyrus, Thal thalamus, and TP temporal related to the default mode network. Bottom Statistically significant pole sensory areas, whereas the ventral hippocampus (anterior Additionally, in humans, high-resolution MRI has analog to humans) is more connected with limbic areas demonstrated differences in subfield volume along the (Fanselow and Dong 2010; Bannerman et al. 2004; Segal anterior–posterior axis, suggesting the presence of two et al. 2010; Czerniawski et al. 2009; Royer et al. 2010). subregions. The antHC comprises the largest part of CA1-3 Both animal and human studies have found that the and the smallest part of the (Malykhin et al. perirhinal cortices and PHG differ in their anatomical 2010). Furthermore, both antHC and postHC have unique connectivity with the hippocampus along a ventral/dorsal pathways to distributed cortical and subcortical regions. and anterior/posterior axis (Suzuki and Amaral 1994; There is an evidence of anatomical connectivity of the Burwell and Amaral 1998; Libby et al. 2012). antHC with the fusiform cortex (Duvernoy 2005), temporal

123 Brain Struct Funct

Table 2 Peak correlations Anatomical Region BA Hemisphere Peak MNI Co-ordinate z score between hippocampal functional connectivity and XYZ memory performance for item and reinstatement memory Anterior hippocampal connectivity Item memory 8 R 34 6 56 -4.27 Lat occipital cortex 7/39 R 42 -68 40 -3.99 Supramarginal gyrus 40 R 46 -42 52 -3.43 Frontal pole 9 R 36 38 36 -2.64 Posterior hippocampal connectivity Reinstatement memory Frontal pole 9 R 36 44 34 4.72 Sup frontal gyrus 32 L -4 28 44 3.9 Precuneus 30 L -4 -78 36 3.87 Lat Frontal pole 10 L -30 62 -6 3.58

pole (Kier et al. 2004), insula, and ventromedial PFC antHC - Item Memory (Catenoix et al. 2011). The postHC, in contrast, has anatomical projections primarily to the PHG and via the fornix to diencephalic and medial forebrain regions, from which communication with other cortical regions occurs via polysynaptic pathways, (Duvernoy 2005). Our findings mirror these differences in anatomical connectivity between the antHC and postHC by demonstrating a reliable parcellation of the hippocampus into these segments based Y = -68X = 42 Z = 40 on its anatomical connectivity. Z score Having stressed the differences between antHC and postHC connectivity to the rest of the brain, it is important -5 -1.96 to keep in mind that there are also prominent longitudinal intrahippocampal fibers connecting both segments (Sloviter postHC - Reinstatement Memory and Lømo 2012; Yang et al. 2014). These intrahippocam- pal connections might facilitate the integration of infor- mation stemming from antHC and postHC subregions (Patel et al. 2013; Strange et al. 2014).

Resting state functional connectivity of the anterior and posterior hippocampi

Z = 34 X = -4 X = -32 The second aim of this study was to investigate whether the data-driven anatomical parcellation of the hippocampus Z score translated into discernable intrinsic functional connectivity 1.96 5 patterns. Our findings that the antHC and postHC are nodes in two partially distinct resting state networks related to Fig. 4 Hippocampal resting state functional connectivity and mem- memory are in line with previous findings that used a priori ory performance. Top panel Anterior hippocampus (antHC) connec- manual segmentation (Poppenk and Moscovitch 2011; tivity to the frontal pole, middle frontal gyrus, supramarginal gyrus, Poppenk et al. 2013). Direct contrasts of the resting state and lateral occipital cortex is negatively correlated with item memory activity of the hippocampal subregions revealed the antHC (p \ 0.05, corrected). Bottom panel Posterior hippocampus (postHC) connectivity to precuneus, anterior cingulate cortex, superior frontal was more functionally connected with limbic areas, gyrus, and frontal pole was positively correlated with reinstatement including the amygdala, OFC, and temporal pole. This memory (p \ 0.05, corrected) network has been related to semantic processing and social

123 Brain Struct Funct and emotional cues (Rogers et al. 2006; Cohn et al. 2014). recently reported that resting state connectivity between In contrast, the postHC had greater functional connectivity posterior hippocampus and posterior cingulate differed with the PHG, rsPCC, superior , and supra- between patients with epilepsy and controls, marginal gyrus. Notably, our report of postHC-PHG and and furthermore that reductions in this connectivity were postHC-PCC connectivity is consistent with other reports associated with impairments on clinical memory measures. (Libby et al. 2012; Ward et al. 2014). Examining all nodes in the DMN, which shows strong overlap with the recollection network, we found that con- Correlations between memory performance nectivity between the hippocampus and posterior neo-cor- and resting state functional connectivity tical regions was also predictive of better memory performance in temporal lobe epilepsy (McCormick et al. Two different processes support recognition memory: 2014). In a study on autobiographical retrieval in healthy recollection and familiarity (e.g., Mandler 1980; Yonelinas adults, we found enhanced connectivity between the 2002). In an experimental context, recollection occurs postHC and posterior cortical regions when individuals when a recognition test item elicits retrieval of qualitative were retrieving perceptual details of events as compared and relational information about the study episode. with generating from verbal cues (McCormick Familiarity supports a sense of past occurrence, but usually et al. 2013b). Thus, we hypothesized that RIM performance lacks any contextual information about the original epi- in healthy adults would be positively correlated with the sode. There is compelling evidence in studies of both functional connectivity of postHC to the recollection encoding (e.g., Ranganath et al. 2004; Uncapher and Rugg memory network (Rugg and Vilberg 2013). It is important 2005; Otten 2007; Davachi et al. 2003; Duarte et al. 2011; to note that a positive result would support the notion that Kensinger and Schacter 2006) and retrieval (e.g., Diana RIM performance on a temporally remote test would be et al. 2010; Eldridge et al. 2000; Yonelinas et al. 2005; indicated by the strength of intrinsic functional correlations Montaldi et al. 2006; Cohn et al. 2009b) that the hip- amongst these network constituents, and thus this network pocampus is activated to a greater extent during recollec- connectivity predicts interindividual memory performance. tion processes than those associated with familiarity. While We found support for this hypothesis as we observed that we did not directly contrast recognition decisions based on increased functional connectivity between RIM perfor- recollection and familiarity judgments, we used a measure mance and connectivity between bilateral postHC and of item recognition (hits to a target item presented alone) nodes of the recollection memory network and other rele- and one that shows a benefit due to context reinstatement vant brain regions, including the right frontal pole, left (highly confident hits to targets presented in the study lateral frontal pole, and midline regions such as , context compared to presented alone). Thus, while both precuneus, , and anterior cingulate recollection and familiarity could contribute to the former cortex. Furthermore, we highlight the potential utility of measure (SIM), the reinstatement effect (RIM) is arguably using resting state functional connectivity as a predictor of much more aligned with recollection. Indeed, we have task performance in healthy adults as has been shown previously shown context reinstatement to be sensitive to elsewhere in etiologies such as epilepsy impacting the hippocampal damage and to enhance recollection and medial temporal lobes (McCormick et al. 2013a). hippocampal activation (Cohn et al. 2009a, b). The antHC and associated lateral temporal network has Most of the data relating to hippocampal engagement been shown to be involved in familiarity-based recognition and different memory tasks/conditions come from task- (Ranganath and Ritchey 2012). As we did not include a based studies of activation of these regions and interpre- direct measure of familiarity-based decisions, but rather tations regarding memory ‘networks’ are typically based SIM, which can involve both familiarity and recollection, on reliable co-activation of these regions rather than direct we did not have an a priori hypothesis regarding possible assessment of connectivity patterns (for a comprehensive correlations involving the antHC and other brain regions. review, see Rugg and Vilberg 2013). Furthermore, there To our surprise, we found a negative correlation between has been relatively little focus on anterior versus posterior SIM performance and antHC functional connectivity with engagement in this literature, although a few observations right supramarginal gyrus, lateral occipital cortex, and right support the hypothesis that the posterior segment plays a middle frontal gyrus and frontal pole. As the correlation is greater role in recollection. Several studies reported a negative, it implies that across subjects, stronger connec- correlation between a static ‘offline’ measure (hippocampal tivity is associated with weaker item memory, a pattern that segment volume) and memory performance (recollection is difficult to interpret. These regions do not align with the and spatial memory) in healthy adults suggesting that the lateral temporal network that has been proposed to be posterior hippocampus is the critical region (Poppenk and important in other episodic and conceptual memory oper- Moscovitch 2011; Maguire et al. 2000). Voets et al. (2014) ations (Poppenk et al. 2013; Ranganath and Ritchey 2012). 123 Brain Struct Funct

Future studies should investigate the role of intrinsic Andrews-Hanna JR, Reidler JS, Sepulcre J, Poulin R, Buckner RL structural and functional connectivity of the antHC with (2010) Functional-anatomic fractionation of the brain’s default network. Neuron 65:550–562. doi:10.1016/j.neuron.2010.02.005 familiarity-based memory performance. Bannerman DM, Rawlins JNP, McHugh SB, Deacon RMJ, Yee BK, Bast T, Zhang WN, Pothuizen HHJ, Feldon J (2004) Regional dissoci- ations within the hippocampus—memory and anxiety. Neurosci Conclusion Biobehav R 28:273–283. doi:10.1016/j.neubiorev.2004.03.004 Beckmann CF, DeLuca M, Devlin JT, Smith SM (2005) Investiga- tions into resting-state connectivity using independent compo- Our study is, to the best of our knowledge, the first to use a nent analysis. Philos T Roy Soc B 360:1001–1013. doi:10.1098/ DWI-based parcellation to examine the existence of sub- rstb.2005.1634 regions suggested both by both animal and human studies Beckmann M, Johansen-Berg H, Rushworth MF (2009) Connectivity- based parcellation of human cingulate cortex and its relation to of hippocampal function. We present novel structural evi- functional specialization. J Neurosci 29:1175–1190. doi:10. dence for the existence of an anterior and posterior sub- 1523/JNEUROSCI.3328-08.2009 region within the hippocampus based on its structural Behrens TEJ, Berg HJ, Jbabdi S, Rushworth MFS, Woolrich MW connectivity profile. Furthermore, we also provide con- (2007) Probabilistic diffusion tractography with multiple fibre orientations: what can we gain? Neuroimage 34:144–155. verging support for the notion that these structurally doi:10.1016/j.neuroimage.2006.09.018 defined subregions are also functionally different—the Behzadi Y, Restom K, Liau J, Liu TT (2007) A component based antHC and postHC are nodes in two functional networks. noise correction method (CompCor) for BOLD and perfusion These data suggest that the functional connectivity of the based fMRI. Neuroimage 37:90–101. doi:10.1016/j.neuroimage. 2007.04.042 hippocampus relies on a structural substrate for these Burwell RD, Amaral DG (1998) Perirhinal and postrhinal cortices of connections. Previous evidence from our group supports the rat: interconnectivity and connections with the entorhinal this concept, demonstrating that functional connectivity is cortex. J Comp Neurol 391:293–321 disrupted with hippocampal damage in medial temporal Catenoix H, Magnin M, Mauguiere F, Ryvlin P (2011) Evoked potential study of hippocampal efferent projections in the human lobe epilepsy (McCormick et al. 2013a), and lesion studies brain. Clin Neurophysiol 122:2488–2497. doi:10.1016/j.clinph. by others provide further support (O’Reilly et al. 2013). As 2011.05.007 highlighted by Neubert and colleagues (2014), both struc- Clower DM, West RA, Lynch JC, Strick PL (2001) The inferior tural and functional connectivity have limitations, but parietal lobule is the target of output from the superior colliculus, hippocampus, and cerebellum. J Neurosci 21:6283–6291 combining the two methods provides a powerful tool to Cohn M, McAndrews MP, Moscovitch M (2009a) Associative reinstate- elucidate brain mechanisms of behavior. In the current ment: a novel approach to assessing associative memory in patients study, we believe that using structural connectivity to with unilateral temporal lobe excisions. Neuropsychologia delineate an anatomical region provides a powerful tool to 47:2989–2994. doi:10.1016/j.neuropsychologia.2009.06.029 Cohn M, Moscovitch M, Lahat A, McAndrews MP (2009b) then relate functional measures of that region to behavior. Recollection versus strength as the primary determinant of hippocampal engagement at retrieval. Proc Natl Acad Sci USA Acknowledgments The authors would like to thank Irene Gian- 106:22451–22455. doi:10.1073/pnas.0908651106 noylis, Keith Ta and Eugene Hlasny for their help in fMRI data Cohn M, St-Laurent M, Barnett A, McAndrews MP (2014) Social collection. The authors declare no competing financial interests. This inference deficits in temporal lobe epilepsy and lobectomy: risk work was supported by Grants from the Canadian Institute of Health factors and neural substrates. Soc Cogn Affect Neur. doi: 10. Research (CIHR; Grant Number 97891 to M.P.M; CIHR fellowship 1093/scan/nsu101. Accessed on 25 July 2014 to M.M) and the James S. McDonnell Foundation (JSMF # 22002055 Czerniawski J, Yoon T, Otto T (2009) Dissociating space and trace in to M.P.M). dorsal and ventral hippocampus. Hippocampus 19:20–32. doi:10. 1002/hipo.20469 Compliance with ethical standards Davachi L, Mitchell JP, Wagner AD (2003) Multiple routes to memory: distinct medial temporal lobe processes build item and The authors declare no potential conflict of interest. The institutions source memories. Proc Natl Acad Sci USA 100:2157–2162. ethics board approved the presented research and informed consent doi:10.1073/pnas.0337195100 was received from all participants. Desikan RS, Se´gonne F, Fischl B, Quinn BT, Dickerson BC, Blacker D et al (2006) An automated labeling system for subdividing the human on MRI scans into gyral based regions of interest. Neuroimage 31:968–980 Diana RA, Yonelinas AP, Ranganath C (2010) Medial temporal lobe References activity during source retrieval reflects information type, not memory strength. J Cognitive Neurosci 22:1808–1818. doi:10. Aggleton JP (2012) Multiple anatomical systems embedded within 1162/jocn.2009.21335 the primate medial temporal lobe: implications for hippocampal Duarte A, Henson RN, Graham KS (2011) Stimulus content and the function. Neurosci Biobehav R 36:1579–1596. doi:10.1016/j. neural correlates of source memory. Brain Res 1373:110–123. neubiorev.2011.09.005 doi:10.1016/j.brainres.2010.11.086 Aggleton JP, Wright NF, Vann SD, Saunders RC (2012) Medial temporal Duvernoy HM (2005) The human hippocampus: functional anatomy, lobe projections to the retrosplenial cortex of the macaque monkey. vascularization and serial sections with MRI. Springer, New Hippocampus 22:1883–1900. doi:10.1002/hipo.22024 York 123 Brain Struct Funct

Eldridge LL, Knowlton BJ, Furmanski CS, Bookheimer SY, Engel Malykhin NV, Lebel RM, Coupland NJ, Wilman AH, Carter R (2010) SA (2000) Remembering episodes: a selective role for the In vivo quantification of hippocampal subfields using 4.7 T fast hippocampus during retrieval. Nature Neurosci 3:1149–1152 spin echo imaging. Neuroimage 49:1224–1230. doi:10.1016/j. Fanselow MS, Dong HW (2010) Are the dorsal and ventral neuroimage.2009.09.042 hippocampus functionally distinct structures? Neuron 65:7–19. Mandler G (1980) Recognizing: the judgment of previous occurrence. doi:10.1016/j.neuron.2009.11.031 Psychol Rev 87:252 Henson RN, Rugg MD, Shallice T, Josephs O, Dolan RJ (1999) Mars RB et al (2011) Diffusion-weighted imaging tractography-based Recollection and familiarity in recognition memory: an event- parcellation of the human parietal cortex and comparison with related functional magnetic resonance imaging study. J Neurosci human and macaque resting-state functional connectivity. 19:3962–3972 J Neurosci 31:4087–4100. doi:10.1523/JNEUROSCI.5102-10. Honey CJ, Sporns O, Cammoun L, Gigandet X, Thiran JP, Meuli R, 2011 Hagmann P (2009) Predicting human resting-state functional Mars RB, Sallet J, Schu¨ffelgen U, Jbabdi S, Toni I, Rushworth MF connectivity from structural connectivity. Proc Natl Acad Sci (2012) Connectivity-based subdivisions of the human right USA 106:2035–2040. doi:10.1073/pnas.0811168106 ‘‘ area’’: evidence for different areas Ho¨rtnagl H, Berger ML, Sperk G, Pifl CH (1991) Regional participating in different cortical networks. Cereb Cortex heterogeneity in the distribution of neurotransmitter markers in 22:1894–1903. doi:10.1093/cercor/bhr268 the rat hippocampus. Neuroscience 45:261–272 McCormick C, Quraan M, Cohn M, Valiante TA, McAndrews MP Jenkinson M, Bannister P, Brady M, Smith S (2002) Improved (2013a) Default mode network connectivity indicates episodic optimization for the robust and accurate linear registration and memory capacity in mesial temporal lobe epilepsy. Epilepsia motion correction of brain images. Neuroimage 17:825–841 54:809–818. doi:10.1111/epi.12098 Jenkinson M, Beckmann CF, Behrens TE, Woolrich MW, Smith SM McCormick C, St-Laurent M, Ty A, Valiante TA, McAndrews MP (2012) Fsl. Neuroimage 62:782–790 (2013b) Functional and effective Hippocampal–Neocortical Johansen-Berg H, Behrens TEJ, Robson MD, Drobnjak I, Rushworth connectivity during construction and elaboration of autobio- MFS, Brady JM, Smith SM, Higham DJ, Matthews PM (2004) graphical memory retrieval. Cereb Cortex. doi:10.1093/cercor/ Changes in connectivity profiles define functionally distinct bht324. Accessed on 24 November 2013 regions in human medial frontal cortex. Proc Natl Acad Sci USA Moayedi M, Salomons TV, Dunlop KA, Downar J, Davis KD (2014) 101:13335–13340. doi:10.1073/pnas.0403743101 Connectivity-based parcellation of the human frontal polar Johnson JD, Suzuki M, Rugg MD (2013) Recollection, familiarity, cortex. Brain Struct Funct:1-14. doi:10.1007/s00429-014-0809-6 and content-sensitivity in lateral parietal cortex: a high-resolu- Mohedano-Moriano A et al (2007) Topographical and laminar tion fMRI study. Front Human Neurosci 7:219. doi:10.3389/ distribution of cortical input to the monkey . fnhum.2013.00219 J Anat 211:250–260. doi:10.1111/j.1469-7580.2007.00764.x Kahn I, Andrews-Hanna JR, Vincent JL, Snyder AZ, Buckner RL Mohedano-Moriano A et al (2008) Convergence of unimodal and (2008) Distinct cortical anatomy linked to subregions of the polymodal sensory input to the entorhinal cortex in the medial temporal lobe revealed by intrinsic functional connec- fascicularis monkey. Neuroscience 151:255–271. doi:10.1016/j. tivity. J Neurophysiol 100:129–139. doi:10.1152/jn.00077.2008 neuroscience.2007.09.074 Kensinger EA, Schacter DL (2006) Amygdala activity is associated Montaldi D, Spencer TJ, Roberts N, Mayes AR (2006) The neural with the successful encoding of item, but not source, information system that mediates familiarity memory. Hippocampus for positive and negative stimuli. J Neurosci 26:2564–2570. 16:504–520. doi:10.1002/hipo.20178 doi:10.1523/JNEUROSCI.5241-05.2006 Neubert FX, Mars RB, Thomas AG, Sallet J, Rushworth MF (2014) Kier EL, Staib LH, Davis LM, Bronen RA (2004) MR imaging of the Comparison of human ventral frontal cortex areas for cognitive temporal stem: anatomic dissection tractography of the uncinate control and language with areas in monkey frontal cortex. fasciculus, inferior occipitofrontal fasciculus, and Meyer’s loop Neuron 81:700–713. doi:10.1016/j.neuron.2013.11.012 of the optic radiation. Am J Neuroradiol 25:677–691 O’Reilly JX et al (2013) Causal effect of disconnection lesions on Klein JC, Behrens TE, Robson MD, Mackay CE, Higham DJ, interhemispheric functional connectivity in rhesus monkeys. Johansen-Berg H (2007) Connectivity-based parcellation of Proc Natl Acad Sci 110:13982–13987. doi:10.1073/pnas. human cortex using diffusion MRI: establishing reproducibility, 1305062110 validity and observer independence in BA 44/45 and SMA/pre- Otten LJ (2007) Fragments of a larger whole: retrieval cues constrain SMA. Neuroimage 34:204–211. doi:10.1016/j.neuroimage.2006. observed neural correlates of memory encoding. Cereb Cortex 08.022 17:2030–2038. doi:10.1093/cercor/bhl111 Kobayashi Y, Amaral DG (2003) Macaque monkey retrosplenial Patel J, Schomburg EW, Bere´nyi A, Fujisawa S, Buzsa´ki G (2013) cortex: II. Cortical afferents. J Comp Neurol 466:48–79. doi:10. Local Generation and Propagation of Ripples along the Sep- 1002/cne.10883 totemporal Axis of the Hippocampus. J Neurosci Ko¨hler C, Eriksson LG, Davies S, Chan-Palay V (1987) Co- 33:17029–17041. doi:10.1523/JNEUROSCI.2036-13.2013 localization of neuropeptide tyrosine and somatostatin Pitka¨nen A, Pikkarainen M, Nurminen N, Ylinen A (2000) Reciprocal immunoreactivity in neurons of individual subfields of the rat connections between the amygdala and the hippocampal forma- hippocampal region. Neurosci Lett 78:1–6 tion, , and postrhinal cortex in rat: a review. Ann Libby LA, Ekstrom AD, Ragland JD, Ranganath C (2012) Differen- N Y Acad Sci 911:369–391. doi:10.1111/j.1749-6632.2000. tial connectivity of perirhinal and parahippocampal cortices tb06738.x within human hippocampal subregions revealed by high-resolu- Poppenk J, Moscovitch M (2011) A hippocampal marker of tion functional imaging. J Neurosci 32:6550–6560. doi:10.1016/ recollection memory ability among healthy young adults: 0304-3940(87)90551-9 contributions of posterior and anterior segments. Neuron Maguire EA, Gadian DG, Johnsrude IS, Good CD, Ashburner J, 72:931–937. doi:10.1016/j.neuron.2011.10.014 Frackowiak RS, Frith CD (2000) Navigation-related structural Poppenk J, Evensmoen HR, Moscovitch M, Nadel L (2013) Long- change in the hippocampi of taxi drivers. Proc Natl Acad Sci axis specialization of the human hippocampus. Trends Cogn Sci USA 97:4398–4403. doi:10.1073/pnas.070039597 17:230–240. doi:10.1016/j.tics.2013.03.005

123 Brain Struct Funct

Ranganath C (2010) A unified framework for the functional Thompson CL et al (2008) Genomic anatomy of the hippocampus. organization of the medial temporal lobes and the phenomenol- Neuron 60:1010–1021. doi:10.1016/j.neuron.2008.12.008 ogy of episodic memory. Hippocampus. 20:1263–1290. doi:10. Tomassini V, Jbabdi S, Klein JC, Behrens TE, Pozzilli C, Matthews 1002/hipo.20852 PM, Rushworth MFS, Johansen-Berg H (2007) Diffusion- Ranganath C, Ritchey M (2012) Two cortical systems for memory- weighted imaging tractography-based parcellation of the human guided behaviour. Nat Rev Neurosci 13:713–726. doi:10.1038/ lateral identifies dorsal and ventral subregions nrn3338 with anatomical and functional specializations. J Neurosci Ranganath C, Johnson MK, D’Esposito M (2003) Prefrontal activity 27:10259–10269. doi:10.1523/JNEUROSCI.2144-07.2007 associated with working memory and episodic long-term mem- Uncapher MR, Rugg MD (2005) Encoding and the durability of ory. Neuropsychologia 41:378–389. doi:10.1016/S0028- episodic memory: a functional magnetic resonance imaging 3932(02)00169-0 study. J Neurosci 25:7260–7267. doi:10.1523/JNEUROSCI. Ranganath C, Yonelinas AP, Cohen MX, Dy CJ, Tom SM, 1641-05.2005 D’Esposito M (2004) Dissociable correlates of recollection and Vincent JL, Kahn I, Snyder AZ, Raichle ME, Buckner RL (2008) familiarity within the medial temporal lobes. Neuropsychologia Evidence for a frontoparietal control system revealed by intrinsic 42:2–13. doi:10.1016/j.neuropsychologia.2003.07.006 functional connectivity. J Neurophysiol 100:3328–3342. doi:10. Rockland KS, Van Hoesen GW (1999) Some temporal and parietal 1152/jn.90355.2008 cortical connections converge in CA1 of the primate hippocam- Voets NL, Beckmann CF, Cole DM, Hong S, Bernasconi A, pus. Cereb Cortex 9:232–237. doi:10.1093/cercor/9.3.232 Bernasconi N (2012) Structural substrates for resting network Rogers TT, Hocking J, Noppeney U, Mechelli A, Gorno-Tempini disruption in temporal lobe epilepsy. Brain 135:2350–2357. ML, Patterson K, Price CJ (2006) Anterior temporal cortex and doi:10.1093/brain/aws137 semantic memory: reconciling findings from neuropsychology Ward AM, Schultz AP, Huijbers W, Van Dijk KR, Hedden R, and functional imaging. Cogn Affect Behav Neurosci Sperling RA (2014) The parahippocampal gyrus links the 6:201–213. doi:10.3758/CABN.6.3.201 default-mode cortical network with the medial temporal lobe Royer S, Sirota A, Patel J, Buzsa´ki G (2010) Distinct representations memory system. Hum Brain Mapp 35:1061–1073. doi:10.1002/ and theta dynamics in dorsal and ventral hippocampus. J Neu- hbm.22234 rosci 30:1777–1787. doi:10.1523/JNEUROSCI.4681-09.2010 Whitfield-Gabrieli S, Nieto-Castanon A (2012) Conn: a functional Rugg MD, Vilberg KL (2013) Brain networks underlying episodic connectivity toolbox for correlated and anticorrelated brain memory retrieval. Curr Opin Neurobiol 23:255–260. doi:10. networks. Brain Conn 2:125–141. doi:10.1089/brain.2012.0073 1016/j.conb.2012.11.005 Witter MP, Van Hoesen GW, Amaral DG (1989) Topographical Segal M, Richter-Levin G, Maggio N (2010) Stress-induced dynamic organization of the entorhinal projection to the dentate gyrus of routing of hippocampal connectivity: a hypothesis. Hippocam- the monkey. J Neurosci 9:216–228 pus 20:1332–1338. doi:10.1002/hipo.20751 Wong JX, de Chastelaine M, Rugg MD (2013) Comparison of the Sloviter RS, Lømo T (2012) Updating the lamellar hypothesis of neural correlates of encoding item-item and item-context hippocampal organization. Front Neural Circuits 6:1–16. doi:10. associations. Front Hum Neurosci 7:436. doi:10.3389/fnhum. 3389/fncir.2012.00102 2013.00436 Small SA, Nava AS, Perera GM, DeLaPaz R, Mayeux R, Stern Y Yang S et al (2014) Interlamellar CA1 network in the hippocampus. (2001) Circuit mechanisms underlying memory encoding and Proc Natl Acad Sci 111:12919–12924. doi:10.1073/pnas. retrieval in the long axis of the . Nat 1405468111 Neurosci 4:442–449. doi:10.1038/86115 Yonelinas AP (2002) The nature of recollection and familiarity: a Strange BA, Witter MP, Lein ES, Moser EI (2014) Functional review of 30 years of research. J Mem Lang 46:441–517. doi:10. organization of the hippocampal longitudinal axis. Nat Rev 1006/jmla.2002.2864 Neurosci 15:655–669. doi:10.1038/nrn3785 Yonelinas AP, Otten LJ, Shaw KN, Rugg MD (2005) Separating the Summerfield JJ, Hassabis D, Maguire EA (2009) Cortical midline brain regions involved in recollection and familiarity in recog- involvement in autobiographical memory. Neuroimage nition memory. J Neurosci 25:3002–3008. doi:10.1523/JNEUR 44:1188–1200. doi:10.1016/j.neuroimage.2008.09.033 OSCI.5295-04.2005 Suzuki WA, Amaral DG (1994) Topographic organization of the Yonelinas AP, Aly M, Wang WC, Koen JD (2010) Recollection and reciprocal connections between the monkey entorhinal cortex familiarity: examining controversial assumptions and new and the perirhinal and parahippocampal cortices. J Neurosci directions. Hippocampus 20:1178–1194. doi:10.1002/hipo. 14:1856–1877 20864

123