Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

Hippocampal differentiation without recognition: An fMRI analysis of the contextual cueing task

Hippocampal differentiation without recognition: An fMRI analysis of the contextual cueing task Downloaded from learnmem.cshlp.org on November 4, 2021 - Published by Cold Spring Harbor Laboratory Press Brief Communication Hippocampal differentiation without recognition: An fMRI analysis of the contextual cueing task 1,3 1 2 2 Anthony J. Greene, William L. Gross, Catherine L. Elsinger, and Stephen M. Rao 1 2 Department of Psychology, University of Wisconsin, Milwaukee, Wisconsin, 53211, USA; Department of Neurology, Medical College of Wisconsin, Milwaukee, Wisconsin 53226, USA A central role of the hippocampus is to consolidate conscious forms of learning and memory, while performance on implicit tasks appears to depend upon other structures. Recently, considerable debate has emerged about whether hippocampal-dependent tasks necessarily entail task awareness. In the contextual cueing task, repetition facilitation is implicit, but impaired in patients with amnesia. Whether the hippocampus alone or other MTL structures are required is unclear. Event-related functional magnetic resonance imaging revealed hippocampal activity that differentiates novel from repeated arrays. This pattern of results was observed without recognition of the repeating arrays. This finding provides support for the claim that the hippocampus is involved in processes outside the domain of conscious learning and memory. array of distractors provides a unique context that determines the Hippocampal differentiation without recognition: target (rotated “T”) location (see Fig. 1). Normal memory partici- An fMRI analysis of the contextual cueing task pants show repetition facilitation even though they do not rec- It has long been understood that damage to the hippocampus ognize repeated arrays; amnesic patients with MTL damage do and medial temporal lobe structures (MTL) in humans dimin- not show repetition facilitation (Chun and Phelps 1999). Both ishes or eliminates the capacity to form new long-term episodic normal memory participants and amnesic patients show a prac- (autobiographical) and semantic (knowledge) memory (Milner tice-related reduction in reaction time (RT), but only normal 1972; Takashima et al. 2006). Together, episodic and semantic memory patients show additional RT facilitation for repeated ar- memory are termed declarative memory (Cohen et al. 1985) be- rays. Normal memory participants show near chance perfor- cause they both require conscious or deliberative access. Con- mance on a subsequent recognition task, indicating that con- versely, numerous studies demonstrate implicit forms of learning scious memory processes cannot account for the repetition fa- and memory that are not substantially affected by damage to the cilitation effects. This finding demonstrates repetition hippocampus or MTL (Keane et al. 1995; Stark and Squire 2000). facilitation, which is impaired in amnesia but is not attributable Implicit forms of learning and memory are demonstrated by ex- to declarative memory. perience-dependent changes in task performance and do not re- One problem with the interpretation of this finding is am- quire conscious recollection. Examples of implicit forms of learn- biguity over whether the hippocampus proper or other temporal ing and memory include perceptual priming or facilitation (rep- lobe structures are mediating repetition facilitation in normal etition leads to greater accuracy and shorter response latency), memory participants. The extent of hippocampal damage was procedural or skill learning, and simple forms of classical condi- assessed in the patient group, but not the extent of concomitant tioning. The most common interpretation of this evidence is that damage to other temporal lobe regions (Chun and Phelps 1999), distinct systems mediate declarative and implicit forms of learn- so the observed impairment could be due to the loss of extra- ing and memory (Squire and Zola 1996; Cohen et al. 1997). Ac- hippocampal structures. More recently the contextual cueing cordingly, the hippocampus would be critically involved in task produced differential outcomes for hippocampus-only am- learning and memory if, and only if, conscious awareness of the nesics (CA fields and dentate gyrus only) compared with tempo- contingencies occurs (Clark and Squire 1998; Reed and Squire ral-lobe amnesics (broad damage to include most of the MTL, 1999; Manns and Squire 2001; Smith et al. 2006). virtually all of the hippocampus, as well as other temporal-lobe While there is no meaningful dispute that the hippocampus regions). The temporal-lobe amnesics showed no repetition fa- is required for conscious learning and memory formation, it may cilitation, while the hippocampus-only amnesics did show near serve a broader function. Several studies have been recently pub- normal repetition facilitation (Manns and Squire 2001). This sug- lished suggesting that the hippocampus is also implicated in cer- gests that the ability to implicitly apprehend specific cue-context tain implicit tasks (Chun and Phelps 1999; Ryan et al. 2000; relations may depend on extra-hippocampal structures and not Greene et al. 2006). However, for some of these tasks there is the hippocampus proper. Thus, the existing evidence does not controversy about whether they are indeed implicit tasks (Smith rule out the hypothesis that the hippocampus is required only for and Squire 2005; Smith et al. 2006; Greene 2007), while for oth- declarative tasks. Problematically, while the temporal-lobe am- ers there is controversy as to whether the task depends upon the nesics had nearly complete hippocampal loss, the hippocampus- hippocampus (Manns and Squire 2001). only amnesics averaged only ∼32% loss of hippocampal tissue In the contextual cueing task (Chun and Jiang 1998), con- with none greater than a 50% loss (Manns and Squire 2001). text-dependent target search is impaired in MTL amnesics but Among the hippocampus-only amnesics, if the damaged tissue does not depend upon recognition (Chun and Phelps 1999). An included critical pathways, the remaining hippocampal volume could be functionally ineffective (e.g., Gold and Squire 2005). On Corresponding author. the other hand, hippocampal damage can spare task-critical E-mail ag@uwm.edu; fax (414) 229-5619. pathways (e.g., Mayanagi et al. 2001), in which case the remain- Article is online at http://www.learnmem.org/cgi/doi/10.1101/lm.609807. ing hippocampal tissue could have mediated performance in Freely available online through the Learning & Memory Open Access option. 14:548–553 © 2007 Cold Spring Harbor Laboratory Press 548 Learning & Memory ISSN 1072-0502/07; www.learnmem.org Downloaded from learnmem.cshlp.org on November 4, 2021 - Published by Cold Spring Harbor Laboratory Press Hippocampal differentiation without awareness interscan interval (TR) of 2 sec. The scanning apparatus and methods for both functional and anatomical data acquisition were identical to those we have published elsewhere (Greene et al. 2006). Functional images were generated with AFNI software (Cox 1996). Each image time series was time-shifted and then spatially registered to reduce the effects of head motion. To pro- vide adequate trials for deconvolution, behavioral blocks (1–20) were imaged in sequences of four at a time, yielding five fMRI runs. The deconvolution analysis included regressors for array type (novel vs. repeated), run (1–5) and RT (continuous) to ex- tract hemodynamic responses for the 14-sec period post-stimulus onset. RT was included as a regressor in the analyses to identify Figure 1. Typical stimulus array. The task was to locate the rotated “T” from among the rotated L distractors. The position and color of the the extent to which changes in hippocampal hemodynamic re- distractors serves as a context that determines the location of the target. sponse could be attributed to differences in dwell time for novel and repeated items. Any hemodynamic effect is evident either as these patients. Whether contextual cueing is a hippocampal- a main effect of RT, a main effect of stimulus type (orthogonal to dependent task is, therefore, unresolved. RT), or as an interaction of RT by stimulus type. Area under the One approach that may clarify ambiguous neuropsychologi- curve (AUC) was calculated by summing the hemodynamic re- cal findings is to use functional imaging to elucidate the role of sponses at all timepoints. the hippocampus during performance of the contextual cueing Individual images were transformed into standard stereo- task in memory-normal participants. If repetition facilitation on taxic space (Talairach and Tournoux 1988) and blurred using a the contextual cueing task depends upon extra-hippocampal 4-mm FWHM Gaussian filter. GLM analyses were conducted for temporal lobe structures, we should observe distinct patterns of (1) the main effect of accuracy, (2) the main effect of RT, and (3) hemodynamic activity associated with novel and repeated items the interaction of accuracy and RT. The region of interest (ROI) within temporal-lobe regions but not within the hippocampus was anatomically defined as the hippocampus (Binder et al. proper. If repetition facilitation on the contextual cueing task 2005). The cluster threshold was 50 µL within the ROI and 200 requires the hippocampus, we should observe distinct patterns of µL for whole-brain analyses. Minimum cluster thresholds were hemodynamic activity for novel and repeated items within the established using Monte Carlo simulations with voxelwise hippocampus proper, which may or may not include extra- P < 0.005 and groupwise  = 0.05 (Cox 1996). hippocampal temporal-lobe regions. Our behavioral results replicate those found elsewhere for We conducted an event-related fMRI study using the con- normal-memory participants (Chun and Phelps 1999; Manns textual cueing task with 26 participants (Sex: 19 f, 7 m; Age: and Squire 2001). Participants’ accuracy for determining the di- range = 18–38, mean = 21.4, SD = 4.2). The behavioral methods rection of the targets was nearly perfect (mean = 97.51% correct). followed those published elsewhere (Chun and Jiang 1998; Chun Figure 2 shows the RT data. The main effect of array type (novel and Phelps 1999) for normal memory participants, with the ex- versus repeated) was significant (F = 12.19, P < 0.01), indicat- (1,16) ceptions that we used visual instead of auditory feedback and ing repetition facilitation (target location for repeated arrays was incorporated a random interstimulus interval (ISI). In each of 20 faster than for novel arrays). The main effect of block (1–20) was blocks, 12 repeated and 12 novel arrays were presented. Arrays significant (F = 7.43, P < 0.01), indicating a practice effect. (19,304) The interaction of array type and block did not reach significance consisted of one target (“T” intersection rotated 90° or 270°) and (F = 1.24, n.s.), indicating that facilitation did not signifi- 11 distractors (“L” intersection rotated either 0°,90°, 180°,or (19,304) cantly increase as the number of exposures to repeated items 270°). Prior to the scanning session, verbal instructions were increased. Importantly, the observed repetition facilitation did given, and participants had one block of training as practice. not depend upon recognition. On the recognition test, we found Items presented during practice were not repeated during the that participants recognized the arrays at chance (binomial experimental scan session. The task was to locate the target and indicate discovery using the right or left key of a button box to indicate the direction of the tail of the “T” target. Each trial consisted of an array displayed for 3 sec followed by 1 sec of visual feedback. Accuracy and RT were recorded for each trial and only correct responses were considered for further analysis. Dur- ing each block of trials, 12 fixation screens were randomly inter- mixed with the 24 arrays—12 novel and 12 repeated—which introduced a random ISI necessary for deconvolution analysis of event-related designs. Immediately after scanning, participants were given a recognition test: The 12 repeated arrays were se- quenced randomly with 12 novel arrays, and participants were asked for each array if they recognized the array from the initial trials. The recognition task was not done in the scanner because there were an insufficient number of trials for deconvolution. The RT data for the contextual cueing task were analyzed with a two-way repeated measures ANOVA. The independent variables were (1) Array Type (novel or repeated), and (2) Block (1–20). For the recognition test, accuracy was assessed with a single-sample t-test to determine whether accuracy differed sig- nificantly from chance (0.50). For the fMRI analysis, we grouped the 20 blocks into five Figure 2. Reaction time data for repeated versus novel arrays. imaging runs, each containing 96 trials and 48 fixations, with an Means  SEM computed for each of 20 blocks. www.learnmem.org 549 Learning & Memory Downloaded from learnmem.cshlp.org on November 4, 2021 - Published by Cold Spring Harbor Laboratory Press Hippocampal differentiation without awareness www.learnmem.org 550 Learning & Memory Figure 3. The hippocampus differentiates novel from repeated arrays. (A) Main effect of array type: activity is decreased for repeated items. At left is the Impulse Response Function (IRF; mean percent signal change  SEM) for the largest of the activations (circled in yellow) in the left posterior hippocampus. The IRF shows a decrease in functional activity for repeated items relative to novel items and baseline. Maps of hemodynamic activity are shown on the right. Coordinates (Talairach and Tournoux 1988) and volumes are provided adjacent to each activation. The top left image provides the index (Y-coordinate) for the positions of three coronal slices for four distinct clusters. (B) The interaction of array type by reaction time: only repeated items show activation that is inversely correlated with reaction time. Faster reactions correspond to greater activity. At the left is an interaction plot (the area under the curve for the IRF) for the larger of the two activations. For graphing purposes, the continuous RT variable was split along its median. The functional activity shows that repeated arrays at faster RTs show increased hemodynamic activity compared with novel arrays and repeated arrays at slower RTs, which do not differ significantly from zero. Maps of hemodynamic activity are shown on the right. Downloaded from learnmem.cshlp.org on November 4, 2021 - Published by Cold Spring Harbor Laboratory Press Hippocampal differentiation without awareness chance = 0.50; Mean = 0.48; t = 0.69, n.s.). To further ad- msec; t <1). That recognition took more than twice as long as (15) dress the possibility that explicit recognition could influence tar- target location (see Fig. 2) further underscores that recognition is get-location RT, we assessed decision time on the recognition test unlikely to have contaminated implicit performance. for novel versus repeated arrays and found no significant difference Consistent with our predictions, we found hemodynamic (novel RT = 3026.4  145.1 msec; repeated RT = 3039.7  143.2 activation in the hippocampal ROIs that were differentially sen- sitive to array type. Two principle find- ings are of interest. First, we found a Table 1. Whole brain activations main effect of array type, such that hip- Region Hemisphere Volume (uL) xyz pocampal activation was less for re- peated than for novel arrays (see Fig. Main effect of stimulus type 3A). The observed functional deactiva- Frontal tion for repeated arrays is consistent Middle Frontal Gyrus L 2266 35.1 7.2 46.8 with findings wherein repeated encod- Medial Frontal Gyrus L 1139 5.6 3.5 51.9 ing results in decreased hippocampal ac- Medial Frontal Gyrus R 570 8.9 9 53.2 tivity (e.g., Zeineh et al. 2003). A second Inferior Frontal Gyrus L 492 43.5 7.6 20.5 Insula L 443 42.2 28.4 18.7 area of activation showed a significant Precentral Gyrus R 369 19.9 22.6 59.9 RT by array type interaction (see Fig. 3B), Insula L 344 33.3 19.5 16 such that for repeated arrays, hippocam- Superior Frontal Gyrus L 242 15.7 26.1 42.3 pal activation increased as RT decreased; Cingulate that is, hippocampal activity was ob- Cingulate Gyrus L 222 6.1 16.9 30.1 Parietal served corresponding to repetition facili- Postcentral Gyrus L 942 34 28.8 46.5 tation. This activation is comparable to Temporal findings from the explicit memory lit- Superior Temporal Gyrus L 244 48.9 25.7 2.1 erature demonstrating greater hippo- Middle Temporal Gyrus L 229 44.6 68.1 13.7 campal activity during successful re- Middle Temporal Gyrus L 219 39.9 68 25.5 trieval (Davachi and Wagner 2002; Parahippocampal Gyrus L 455 21.7 40.9 4.4 Parahippocampal Gyrus R 348 16.5 43.3 4.2 Strange et al. 2005; Greene et al. 2006). Occipital Thus, the hippocampus differentiates re- Middle Occipital Gyrus L 284 42.6 69.4 4.9 peated from novel arrays in the absence Subcortical of conscious recognition. While a com- Culmen L 615 8.7 44.1 4.7 parable division of labor between encod- Caudate L 337 9 6.3 6.8 Thalamus R 285 10.6 18.5 0.1 ing and retrieval is frequently observed Lentiform Nucleus L 226 29 19.8 0.4 in explicit hippocampal tasks, the re- spective hippocampal subregions in- Interaction of RT by stimulus type volved vary considerably by task and Frontal there is little consensus about how to in- Insula L 447 35.7 15.5 15.9 terpret these differences (Gabrieli et al. Medial Frontal Gyrus L 438 6.5 3.9 52.3 1997; Dolan and Fletcher 1999; Greicius Insula L 303 42.8 28.4 18.7 et al. 2003; Eldridge et al. 2005). How- Precentral Gyrus L 289 37.1 9.5 54.6 Medial Frontal Gyrus R 269 8.7 10.1 52.9 ever, in this task, involvement of poste- Superior Frontal Gyrus L 231 15.3 25.3 42.2 rior hippocampal regions for both acti- Middle Frontal Gyrus L 219 29.7 12.8 45.4 vations may be due to the spatial nature Temporal of the task (e.g., Maguire et al. 2003; Parahippocampal Gyrus L 230 23.8 39.1 5.7 Goel et al. 2004). Anatomical connec- Subcortical tions between the posterior hippocam- Culmen L 372 9.4 44.8 6.7 pus, posterior parahippocampal cortices, Main effect of RT and parietal systems (Munoz and Insau- Frontal sti 2005) may constitute a network for Medial Frontal Gyrus R 10349 22.7 0.3 45.5 the acquisition and expression of spatial Medial Frontal Gyrus L 7975 22.3 1.9 43.6 relational learning (Manns and Eichen- Insula R 2969 37.3 13.5 1.5 baum 2006; van Asselen et al. 2006). Insula L 1273 32.5 13.3 4.6 Importantly, the hippocampal acti- Middle Frontal Gyrus R 1204 30.8 34.5 24.7 vations are not attributable to differen- Parietal Inferior Parietal Lobule L 21159 24.5 59.7 38.4 tial dwell time for novel and repeated Inferior Parietal Lobule R 17415 21.9 58.7 43.1 arrays: (1) there was no main effect of RT Occipital within the hippocampus; (2) the main Middle Occipital Gyrus L 9019 35.1 62.8 10.2 effect of array type was orthogonal to Cuneus L 1355 6.5 69 6.3 RT; and (3) in the interaction, only re- Cuneus L 694 1.9 89.7 7.1 Middle Occipital Gyrus R 478 38.5 68.5 7.8 peated arrays showed greater activation Middle Occipital Gyrus R 273 31.3 81.6 16.9 and only at faster RTs. Middle Occipital Gyrus R 248 43.8 55.3 5.7 Whole-brain tables for factors in Subcortical the general linear model (GLM) are Culmen R 696 19.8 55.5 9.7 shown in Table 1. Note that several tem- poral lobe regions to include the MTL Region is defined as center of mass. Coordinates represent distance in millimeters from anterior com- are involved in both the main effect of missure: x right(+)/left(); y anterior(+)/posterior(); z superior(+)/inferior(). Individual voxel prob- ability < 0.005, minimum cluster size > 200 µL. array type and in the interaction of array www.learnmem.org 551 Learning & Memory Downloaded from learnmem.cshlp.org on November 4, 2021 - Published by Cold Spring Harbor Laboratory Press Hippocampal differentiation without awareness type by RT. While the present experiment was designed to test (McEchron and Disterhoft 1997). Whereas the present findings the involvement of the hippocampus proper in the contextual tend to argue against declarative memory as the core function of cueing task, most treatments of declarative memory assert that the hippocampus, future research is needed to determine the neither the hippocampus nor surrounding MTL regions are in- relative merit of binding and context-driven models. volved in implicit tasks (Squire and Zola 1996; Squire et al. 2004); on the other hand, processes involved in familiarity may involve Acknowledgments certain MTL regions, but not the hippocampus (for review, see This study was supported in part by a grant from the National Eichenbaum et al. 2007). Institute of Aging (R01 AG022304) to S.R., and the National Cen- The purpose of this experiment was to provide converging ter for Research Resources, National Institutes of Health (M01- evidence that the hippocampus is involved in implicit contex- RR00058) to the Medical College of Wisconsin General Research Center, and the W.M. Keck Foundation. tual learning. We found that the hippocampal hemodynamic response differentiated novel from repeated items despite the fact References that participants could not differentiate the repeated items in a recognition task. Our findings are consistent with neuropsycho- Binder, J.R., Bellgowan, P.S., Hammeke, T.A., Possing, E.T., and Frost, J.A. 2005. A comparison of two fMRI protocols for eliciting logical findings suggesting that the hippocampus plays a poten- hippocampal activation. Epilepsia 46: 1061–1070. tially important role in this implicit task (Chun and Phelps Chun, M.M. and Jiang, Y. 1998. Contextual cueing: Implicit learning 1999). To our knowledge, no study has found task-differential and memory of visual context guides spatial attention. Cognit. hippocampal activation in a perceptual priming task. Psychol. 36: 28–71. Chun, M.M. and Phelps, E.A. 1999. Memory deficits for implicit This finding also helps establish an important and necessary contextual information in amnesic subjects with hippocampal bridge between amnesic studies (which are limited by small damage. Nat. Neurosci. 2: 844–847. samples and substantial variation in focus and extent of damage) Clark, R.E. and Squire, L.R. 1998. Classical conditioning and brain and functional imaging studies (limited to showing that an area systems: The role of awareness. Science 280: 77–81. Cohen, N.J., Eichenbaum, H., Deacedo, B.S., and Corkin, S. 1985. is involved in a particular process, but not necessarily critical for Different memory systems underlying acquisition of procedural and its performance). Where possible, converging evidence between declarative knowledge. Ann. N. Y. Acad. Sci. 444: 54–71. neuropsychological and functional imaging methodologies pro- Cohen, N.J., Poldrack, R.A., and Eichenbaum, H. 1997. Memory for vides stronger evidence than either can provide alone. items and memory for relations in the procedural/declarative memory framework. Memory 5: 131–178. According to the declarative memory model, context- Cox, R.W. 1996. AFNI: Software foranalysis and visualization of dependent tasks, such as contextual cueing, involve the hippo- functional megnetic resonance neuroimages. Comput. Biomed. Res. campus because the contingencies are sufficiently complex that 29: 162–173. they require reorganization that may only take place as a delib- Davachi, L. and Wagner, A.D. 2002. Hippocampal contributions to episodic encoding: Insights from relational and item-based learning. erate act of memory manipulation (Clark and Squire 1998; Reed J. Neurophysiol. 88: 982–990. and Squire 1999; Manns and Squire 2001). By this account, con- Dolan, R.J. and Fletcher, P.F. 1999. Encoding and retrieval in human text-dependent learning is a subset of declarative memory, which medial temporal lobes: An empirical investigation using functional necessarily requires the hippocampus. Conversely, it has been magnetic resonance imaging (fMRI). Hippocampus 9: 25–34. Eichenbaum, H. 2006. Remembering: Functional organization of the argued that episodic and semantic memory are specific instances declarative memory system. Curr. Biol. 16: R643–R645. of context-dependent learning (Ryan et al. 2000; Greene et al. Eichenbaum, H. and Cohen, N.J. 2001. From conditioning to conscious 2006; Manns and Eichenbaum 2006; Greene 2007). Accordingly, recollection: Memory systems of the brain. Oxford University Press, New declarative memory is encoded as associations to context and the York. Eichenbaum, H., Yonelinas, A.P., and Ranganath, C. 2007. The medial necessary context for recognition and recall is provided by re- temporal lobe and recognition memory. Annu. Rev. Neurosci. 30: trieval cues (Nyberg et al. 1996; Rudy et al. 2002). A third hy- 123–152. pothesis is that the hippocampus is involved in binding nonpre- Eldridge, L.L., Engel, S.A., Zeineh, M.M., Bookheimer, S.Y., and disposed associations. That is, some types of association forma- Knowlton, B.J. 2005. A dissociation of encoding and retrieval processes in the human hippocampus. J. Neurosci. 25: 3280–3286. tion require synaptic connections where existing connectivity is Gabrieli, J.D., Brewer, J.B., Desmond, J.E., and Glover, G.H. 1997. inadequate. Some of these instances occur where the contextual Separate neural bases of two fundamental memory processes in the associations require a complex network (Chun and Phelps 1999; human medial temporal lobe. Science 276: 264–266. Reed and Squire 1999; Ryan et al. 2000; Manns and Squire 2001; Goel, V., Makale, M., and Grafman, J. 2004. The hippocampal system mediates logical reasoning about familiar spatial environments. J. Davachi and Wagner 2002; Greene et al. 2006) and some involve Cogn. Neurosci. 16: 654–664. multiple cortical systems that span distant regions of cortex Gold, J.J. and Squire, L.R. 2005. Quantifying medial temporal lobe (McEchron and Disterhoft 1997; Davachi and Wagner 2002; Ei- damage in memory-impaired patients. Hippocampus 15: 79–85. chenbaum 2006). According to this view, the large proportion of Greene, A.J. 2007. Human hippocampal-dependent tasks: Is awareness necessary or sufficient? Hippocampus 17: 429–433. observed hippocampal-dependent tasks that are declarative oc- Greene, A.J., Gross, W.L., Elsinger, C., and Rao, S.M. 2006. An fMRI curs because declarative memory is generally both context de- analysis of the human hippocampus: Inference, context and task pendent and involves numerous cortical regions (Eichenbaum awareness. J. Cogn. Neurosci. 18: 1156–1173. and Cohen 2001). Greicius, M.D., Krasnow, B., Boyett-Anderson, J.M., Eliez, S., Schatzberg, A.F., Reiss, A.L., and Menon, V. 2003. Regional analysis of The present results, taken together with neuropsychological hippocampal activation during memory encoding and retrieval: findings (Chun and Phelps 1999), provide converging evidence fMRI study. Hippocampus 13: 164–174. to suggest hippocampal involvement in a context-dependent Keane, M.M., Gabrieli, J.D., Mapstone, H.C., Johnson, K.A., and Corkin, task that does not rely on declarative memory. One possibility is S. 1995. Double dissociation of memory capacities after bilateral occipital-lobe or medial temporal-lobe lesions. Brain 118: 1129–1148. that declarative memory is a subset of context-dependent learn- Knowlton, B.J. and Squire, L.R. 1996. Artificial grammar learning ing and memory, and accordingly, contextual learning is closer depends on implicit acquisition of both abstract and exemplar-specific to the central role of the hippocampus. Problematically, learning information. J. Exp. Psychol. Learn. Mem. Cogn. 22: 169–181. tasks that one might argue are contextual in nature may be in- Maguire, E.A., Spiers, H.J., Good, C.D., Hartley, T., Frackowiak, R.S., and Burgess, N. 2003. Navigation expertise and the human dependent of the hippocampus (Knowlton and Squire 1996). hippocampus: A structural brain imaging analysis. Hippocampus Other tasks such as trace conditioning are hippocampal depen- 13: 250–259. dent but have no greater claim to context dependency than com- Manns, J.R. and Eichenbaum, H. 2006. Evolution of declarative parable nonhippocampal-dependent conditioning tasks memory. Hippocampus 16: 795–808. www.learnmem.org 552 Learning & Memory Downloaded from learnmem.cshlp.org on November 4, 2021 - Published by Cold Spring Harbor Laboratory Press Hippocampal differentiation without awareness Manns, J.R. and Squire, L.R. 2001. Perceptual learning, awareness, and hippocampus-dependent memory. J. Neurosci. 26: 11304–11312. the hippocampus. Hippocampus 11: 776–782. Squire, L.R. and Zola, S.M. 1996. Structure and function of declarative Mayanagi, Y., Watanabe, E., Nagahori, Y., and Nankai, M. 2001. and nondeclarative memory systems. Proc. Natl. Acad. Sci. Psychiatric and neuropsychological problems in epilepsy surgery: 93: 13515–13522. Analysis of 100 cases that underwent surgery. Epilepsia 42 (Suppl Squire, L.R., Stark, C.E., and Clark, R.E. 2004. The medial temporal lobe. 6): 19–23. Annu. Rev. Neurosci. 27: 279–306. McEchron, M.D. and Disterhoft, J.F. 1997. Sequence of single neuron Stark, C.E. and Squire, L.R. 2000. Recognition memory and familiarity changes in CA1 hippocampus of rabbits during acquisition of trace judgments in severe amnesia: No evidence for a contribution of eyeblink conditioned responses. J. Neurophysiol. 78: 1030–1044. repetition priming. Behav. Neurosci. 114: 459–467. Milner, B. 1972. Disorders of learning and memory after temporal lobe Strange, B.A., Duggins, A., Penny, W., Dolan, R.J., and Friston, K.J. 2005. lesions in man. Clin. Neurosurg. 19: 421–446. Information theory, novelty and hippocampal responses: Munoz, M. and Insausti, R. 2005. Cortical efferents of the entorhinal Unpredicted or unpredictable? Neural Netw. 18: 225–230. cortex and the adjacent parahippocampal region in the monkey Takashima, A., Petersson, K.M., Rutters, F., Tendolkar, I., Jensen, O., (Macaca fascicularis). Eur. J. Neurosci. 22: 1368–1388. Zwarts, M.J., McNaughton, B.L., and Fernandez, G. 2006. Declarative Nyberg, L., McIntosh, A.R., Houle, S., Nilsson, L.G., and Tulving, E. memory consolidation in humans: A prospective functional 1996. Activation of medial temporal structures during episodic magnetic resonance imaging study. Proc. Natl. Acad. Sci. memory retrieval. Nature 380: 715–717. 103: 756–761. Reed, J.M. and Squire, L.R. 1999. Impaired transverse patterning in Talairach, J. and Tournoux, P. 1988. Co-planar stereotaxic atlas of the human amnesia is a special case of impaired memory for two-choice human brain. Thieme, New York. discrimination tasks. Behav. Neurosci. 113: 3–9. van Asselen, M., Kessels, R.P., Neggers, S.F., Kappelle, L.J., Frijns, C.J., Rudy, J.W., Barrientos, R.M., and O’Reilly, R.C. 2002. Hippocampal and Postma, A. 2006. Brain areas involved in spatial working formation supports conditioning to memory of a context. Behav. memory. Neuropsychologia 44: 1185–1194. Neurosci. 116: 530–538. Zeineh, M.M., Engel, S.A., Thompson, P.M., and Bookheimer, S.Y. 2003. Ryan, J.D., Althoff, R.R., Whitlow, S., and Cohen, N.J. 2000. Amnesia is Dynamics of the hippocampus during encoding and retrieval of a deficit in relational memory. Psychol. Sci. 11: 454–461. face-name pairs. Science 299: 577–580. Smith, C.N. and Squire, L.R. 2005. Declarative memory, awareness, and transitive inference. J. Neurosci. 25: 10138–10146. Smith, C.N., Hopkins, R.O., and Squire, L.R. 2006. Experience-dependent eye movements, awareness, and Received April 17, 2007; accepted in revised form June 25, 2007. www.learnmem.org 553 Learning & Memory Downloaded from learnmem.cshlp.org on November 4, 2021 - Published by Cold Spring Harbor Laboratory Press Hippocampal differentiation without recognition: An fMRI analysis of the contextual cueing task Anthony J. Greene, William L. Gross, Catherine L. Elsinger, et al. Learn. Mem. 2007, 14: Access the most recent version at doi:10.1101/lm.609807 This article cites 40 articles, 8 of which can be accessed free at: References http://learnmem.cshlp.org/content/14/8/548.full.html#ref-list-1 Freely available online through the Learning & Memory Open Access option. License Receive free email alerts when new articles cite this article - sign up in the box at the Email Alerting top right corner of the article or click here. Service Copyright © 2007, Cold Spring Harbor Laboratory Press http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Learning & Memory Unpaywall

Hippocampal differentiation without recognition: An fMRI analysis of the contextual cueing task

Learning & MemoryAug 9, 2007

Loading next page...
 
/lp/unpaywall/hippocampal-differentiation-without-recognition-an-fmri-analysis-of-rNgM0IHAOR

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Unpaywall
ISSN
1072-0502
DOI
10.1101/lm.609807
Publisher site
See Article on Publisher Site

Abstract

Downloaded from learnmem.cshlp.org on November 4, 2021 - Published by Cold Spring Harbor Laboratory Press Brief Communication Hippocampal differentiation without recognition: An fMRI analysis of the contextual cueing task 1,3 1 2 2 Anthony J. Greene, William L. Gross, Catherine L. Elsinger, and Stephen M. Rao 1 2 Department of Psychology, University of Wisconsin, Milwaukee, Wisconsin, 53211, USA; Department of Neurology, Medical College of Wisconsin, Milwaukee, Wisconsin 53226, USA A central role of the hippocampus is to consolidate conscious forms of learning and memory, while performance on implicit tasks appears to depend upon other structures. Recently, considerable debate has emerged about whether hippocampal-dependent tasks necessarily entail task awareness. In the contextual cueing task, repetition facilitation is implicit, but impaired in patients with amnesia. Whether the hippocampus alone or other MTL structures are required is unclear. Event-related functional magnetic resonance imaging revealed hippocampal activity that differentiates novel from repeated arrays. This pattern of results was observed without recognition of the repeating arrays. This finding provides support for the claim that the hippocampus is involved in processes outside the domain of conscious learning and memory. array of distractors provides a unique context that determines the Hippocampal differentiation without recognition: target (rotated “T”) location (see Fig. 1). Normal memory partici- An fMRI analysis of the contextual cueing task pants show repetition facilitation even though they do not rec- It has long been understood that damage to the hippocampus ognize repeated arrays; amnesic patients with MTL damage do and medial temporal lobe structures (MTL) in humans dimin- not show repetition facilitation (Chun and Phelps 1999). Both ishes or eliminates the capacity to form new long-term episodic normal memory participants and amnesic patients show a prac- (autobiographical) and semantic (knowledge) memory (Milner tice-related reduction in reaction time (RT), but only normal 1972; Takashima et al. 2006). Together, episodic and semantic memory patients show additional RT facilitation for repeated ar- memory are termed declarative memory (Cohen et al. 1985) be- rays. Normal memory participants show near chance perfor- cause they both require conscious or deliberative access. Con- mance on a subsequent recognition task, indicating that con- versely, numerous studies demonstrate implicit forms of learning scious memory processes cannot account for the repetition fa- and memory that are not substantially affected by damage to the cilitation effects. This finding demonstrates repetition hippocampus or MTL (Keane et al. 1995; Stark and Squire 2000). facilitation, which is impaired in amnesia but is not attributable Implicit forms of learning and memory are demonstrated by ex- to declarative memory. perience-dependent changes in task performance and do not re- One problem with the interpretation of this finding is am- quire conscious recollection. Examples of implicit forms of learn- biguity over whether the hippocampus proper or other temporal ing and memory include perceptual priming or facilitation (rep- lobe structures are mediating repetition facilitation in normal etition leads to greater accuracy and shorter response latency), memory participants. The extent of hippocampal damage was procedural or skill learning, and simple forms of classical condi- assessed in the patient group, but not the extent of concomitant tioning. The most common interpretation of this evidence is that damage to other temporal lobe regions (Chun and Phelps 1999), distinct systems mediate declarative and implicit forms of learn- so the observed impairment could be due to the loss of extra- ing and memory (Squire and Zola 1996; Cohen et al. 1997). Ac- hippocampal structures. More recently the contextual cueing cordingly, the hippocampus would be critically involved in task produced differential outcomes for hippocampus-only am- learning and memory if, and only if, conscious awareness of the nesics (CA fields and dentate gyrus only) compared with tempo- contingencies occurs (Clark and Squire 1998; Reed and Squire ral-lobe amnesics (broad damage to include most of the MTL, 1999; Manns and Squire 2001; Smith et al. 2006). virtually all of the hippocampus, as well as other temporal-lobe While there is no meaningful dispute that the hippocampus regions). The temporal-lobe amnesics showed no repetition fa- is required for conscious learning and memory formation, it may cilitation, while the hippocampus-only amnesics did show near serve a broader function. Several studies have been recently pub- normal repetition facilitation (Manns and Squire 2001). This sug- lished suggesting that the hippocampus is also implicated in cer- gests that the ability to implicitly apprehend specific cue-context tain implicit tasks (Chun and Phelps 1999; Ryan et al. 2000; relations may depend on extra-hippocampal structures and not Greene et al. 2006). However, for some of these tasks there is the hippocampus proper. Thus, the existing evidence does not controversy about whether they are indeed implicit tasks (Smith rule out the hypothesis that the hippocampus is required only for and Squire 2005; Smith et al. 2006; Greene 2007), while for oth- declarative tasks. Problematically, while the temporal-lobe am- ers there is controversy as to whether the task depends upon the nesics had nearly complete hippocampal loss, the hippocampus- hippocampus (Manns and Squire 2001). only amnesics averaged only ∼32% loss of hippocampal tissue In the contextual cueing task (Chun and Jiang 1998), con- with none greater than a 50% loss (Manns and Squire 2001). text-dependent target search is impaired in MTL amnesics but Among the hippocampus-only amnesics, if the damaged tissue does not depend upon recognition (Chun and Phelps 1999). An included critical pathways, the remaining hippocampal volume could be functionally ineffective (e.g., Gold and Squire 2005). On Corresponding author. the other hand, hippocampal damage can spare task-critical E-mail ag@uwm.edu; fax (414) 229-5619. pathways (e.g., Mayanagi et al. 2001), in which case the remain- Article is online at http://www.learnmem.org/cgi/doi/10.1101/lm.609807. ing hippocampal tissue could have mediated performance in Freely available online through the Learning & Memory Open Access option. 14:548–553 © 2007 Cold Spring Harbor Laboratory Press 548 Learning & Memory ISSN 1072-0502/07; www.learnmem.org Downloaded from learnmem.cshlp.org on November 4, 2021 - Published by Cold Spring Harbor Laboratory Press Hippocampal differentiation without awareness interscan interval (TR) of 2 sec. The scanning apparatus and methods for both functional and anatomical data acquisition were identical to those we have published elsewhere (Greene et al. 2006). Functional images were generated with AFNI software (Cox 1996). Each image time series was time-shifted and then spatially registered to reduce the effects of head motion. To pro- vide adequate trials for deconvolution, behavioral blocks (1–20) were imaged in sequences of four at a time, yielding five fMRI runs. The deconvolution analysis included regressors for array type (novel vs. repeated), run (1–5) and RT (continuous) to ex- tract hemodynamic responses for the 14-sec period post-stimulus onset. RT was included as a regressor in the analyses to identify Figure 1. Typical stimulus array. The task was to locate the rotated “T” from among the rotated L distractors. The position and color of the the extent to which changes in hippocampal hemodynamic re- distractors serves as a context that determines the location of the target. sponse could be attributed to differences in dwell time for novel and repeated items. Any hemodynamic effect is evident either as these patients. Whether contextual cueing is a hippocampal- a main effect of RT, a main effect of stimulus type (orthogonal to dependent task is, therefore, unresolved. RT), or as an interaction of RT by stimulus type. Area under the One approach that may clarify ambiguous neuropsychologi- curve (AUC) was calculated by summing the hemodynamic re- cal findings is to use functional imaging to elucidate the role of sponses at all timepoints. the hippocampus during performance of the contextual cueing Individual images were transformed into standard stereo- task in memory-normal participants. If repetition facilitation on taxic space (Talairach and Tournoux 1988) and blurred using a the contextual cueing task depends upon extra-hippocampal 4-mm FWHM Gaussian filter. GLM analyses were conducted for temporal lobe structures, we should observe distinct patterns of (1) the main effect of accuracy, (2) the main effect of RT, and (3) hemodynamic activity associated with novel and repeated items the interaction of accuracy and RT. The region of interest (ROI) within temporal-lobe regions but not within the hippocampus was anatomically defined as the hippocampus (Binder et al. proper. If repetition facilitation on the contextual cueing task 2005). The cluster threshold was 50 µL within the ROI and 200 requires the hippocampus, we should observe distinct patterns of µL for whole-brain analyses. Minimum cluster thresholds were hemodynamic activity for novel and repeated items within the established using Monte Carlo simulations with voxelwise hippocampus proper, which may or may not include extra- P < 0.005 and groupwise  = 0.05 (Cox 1996). hippocampal temporal-lobe regions. Our behavioral results replicate those found elsewhere for We conducted an event-related fMRI study using the con- normal-memory participants (Chun and Phelps 1999; Manns textual cueing task with 26 participants (Sex: 19 f, 7 m; Age: and Squire 2001). Participants’ accuracy for determining the di- range = 18–38, mean = 21.4, SD = 4.2). The behavioral methods rection of the targets was nearly perfect (mean = 97.51% correct). followed those published elsewhere (Chun and Jiang 1998; Chun Figure 2 shows the RT data. The main effect of array type (novel and Phelps 1999) for normal memory participants, with the ex- versus repeated) was significant (F = 12.19, P < 0.01), indicat- (1,16) ceptions that we used visual instead of auditory feedback and ing repetition facilitation (target location for repeated arrays was incorporated a random interstimulus interval (ISI). In each of 20 faster than for novel arrays). The main effect of block (1–20) was blocks, 12 repeated and 12 novel arrays were presented. Arrays significant (F = 7.43, P < 0.01), indicating a practice effect. (19,304) The interaction of array type and block did not reach significance consisted of one target (“T” intersection rotated 90° or 270°) and (F = 1.24, n.s.), indicating that facilitation did not signifi- 11 distractors (“L” intersection rotated either 0°,90°, 180°,or (19,304) cantly increase as the number of exposures to repeated items 270°). Prior to the scanning session, verbal instructions were increased. Importantly, the observed repetition facilitation did given, and participants had one block of training as practice. not depend upon recognition. On the recognition test, we found Items presented during practice were not repeated during the that participants recognized the arrays at chance (binomial experimental scan session. The task was to locate the target and indicate discovery using the right or left key of a button box to indicate the direction of the tail of the “T” target. Each trial consisted of an array displayed for 3 sec followed by 1 sec of visual feedback. Accuracy and RT were recorded for each trial and only correct responses were considered for further analysis. Dur- ing each block of trials, 12 fixation screens were randomly inter- mixed with the 24 arrays—12 novel and 12 repeated—which introduced a random ISI necessary for deconvolution analysis of event-related designs. Immediately after scanning, participants were given a recognition test: The 12 repeated arrays were se- quenced randomly with 12 novel arrays, and participants were asked for each array if they recognized the array from the initial trials. The recognition task was not done in the scanner because there were an insufficient number of trials for deconvolution. The RT data for the contextual cueing task were analyzed with a two-way repeated measures ANOVA. The independent variables were (1) Array Type (novel or repeated), and (2) Block (1–20). For the recognition test, accuracy was assessed with a single-sample t-test to determine whether accuracy differed sig- nificantly from chance (0.50). For the fMRI analysis, we grouped the 20 blocks into five Figure 2. Reaction time data for repeated versus novel arrays. imaging runs, each containing 96 trials and 48 fixations, with an Means  SEM computed for each of 20 blocks. www.learnmem.org 549 Learning & Memory Downloaded from learnmem.cshlp.org on November 4, 2021 - Published by Cold Spring Harbor Laboratory Press Hippocampal differentiation without awareness www.learnmem.org 550 Learning & Memory Figure 3. The hippocampus differentiates novel from repeated arrays. (A) Main effect of array type: activity is decreased for repeated items. At left is the Impulse Response Function (IRF; mean percent signal change  SEM) for the largest of the activations (circled in yellow) in the left posterior hippocampus. The IRF shows a decrease in functional activity for repeated items relative to novel items and baseline. Maps of hemodynamic activity are shown on the right. Coordinates (Talairach and Tournoux 1988) and volumes are provided adjacent to each activation. The top left image provides the index (Y-coordinate) for the positions of three coronal slices for four distinct clusters. (B) The interaction of array type by reaction time: only repeated items show activation that is inversely correlated with reaction time. Faster reactions correspond to greater activity. At the left is an interaction plot (the area under the curve for the IRF) for the larger of the two activations. For graphing purposes, the continuous RT variable was split along its median. The functional activity shows that repeated arrays at faster RTs show increased hemodynamic activity compared with novel arrays and repeated arrays at slower RTs, which do not differ significantly from zero. Maps of hemodynamic activity are shown on the right. Downloaded from learnmem.cshlp.org on November 4, 2021 - Published by Cold Spring Harbor Laboratory Press Hippocampal differentiation without awareness chance = 0.50; Mean = 0.48; t = 0.69, n.s.). To further ad- msec; t <1). That recognition took more than twice as long as (15) dress the possibility that explicit recognition could influence tar- target location (see Fig. 2) further underscores that recognition is get-location RT, we assessed decision time on the recognition test unlikely to have contaminated implicit performance. for novel versus repeated arrays and found no significant difference Consistent with our predictions, we found hemodynamic (novel RT = 3026.4  145.1 msec; repeated RT = 3039.7  143.2 activation in the hippocampal ROIs that were differentially sen- sitive to array type. Two principle find- ings are of interest. First, we found a Table 1. Whole brain activations main effect of array type, such that hip- Region Hemisphere Volume (uL) xyz pocampal activation was less for re- peated than for novel arrays (see Fig. Main effect of stimulus type 3A). The observed functional deactiva- Frontal tion for repeated arrays is consistent Middle Frontal Gyrus L 2266 35.1 7.2 46.8 with findings wherein repeated encod- Medial Frontal Gyrus L 1139 5.6 3.5 51.9 ing results in decreased hippocampal ac- Medial Frontal Gyrus R 570 8.9 9 53.2 tivity (e.g., Zeineh et al. 2003). A second Inferior Frontal Gyrus L 492 43.5 7.6 20.5 Insula L 443 42.2 28.4 18.7 area of activation showed a significant Precentral Gyrus R 369 19.9 22.6 59.9 RT by array type interaction (see Fig. 3B), Insula L 344 33.3 19.5 16 such that for repeated arrays, hippocam- Superior Frontal Gyrus L 242 15.7 26.1 42.3 pal activation increased as RT decreased; Cingulate that is, hippocampal activity was ob- Cingulate Gyrus L 222 6.1 16.9 30.1 Parietal served corresponding to repetition facili- Postcentral Gyrus L 942 34 28.8 46.5 tation. This activation is comparable to Temporal findings from the explicit memory lit- Superior Temporal Gyrus L 244 48.9 25.7 2.1 erature demonstrating greater hippo- Middle Temporal Gyrus L 229 44.6 68.1 13.7 campal activity during successful re- Middle Temporal Gyrus L 219 39.9 68 25.5 trieval (Davachi and Wagner 2002; Parahippocampal Gyrus L 455 21.7 40.9 4.4 Parahippocampal Gyrus R 348 16.5 43.3 4.2 Strange et al. 2005; Greene et al. 2006). Occipital Thus, the hippocampus differentiates re- Middle Occipital Gyrus L 284 42.6 69.4 4.9 peated from novel arrays in the absence Subcortical of conscious recognition. While a com- Culmen L 615 8.7 44.1 4.7 parable division of labor between encod- Caudate L 337 9 6.3 6.8 Thalamus R 285 10.6 18.5 0.1 ing and retrieval is frequently observed Lentiform Nucleus L 226 29 19.8 0.4 in explicit hippocampal tasks, the re- spective hippocampal subregions in- Interaction of RT by stimulus type volved vary considerably by task and Frontal there is little consensus about how to in- Insula L 447 35.7 15.5 15.9 terpret these differences (Gabrieli et al. Medial Frontal Gyrus L 438 6.5 3.9 52.3 1997; Dolan and Fletcher 1999; Greicius Insula L 303 42.8 28.4 18.7 et al. 2003; Eldridge et al. 2005). How- Precentral Gyrus L 289 37.1 9.5 54.6 Medial Frontal Gyrus R 269 8.7 10.1 52.9 ever, in this task, involvement of poste- Superior Frontal Gyrus L 231 15.3 25.3 42.2 rior hippocampal regions for both acti- Middle Frontal Gyrus L 219 29.7 12.8 45.4 vations may be due to the spatial nature Temporal of the task (e.g., Maguire et al. 2003; Parahippocampal Gyrus L 230 23.8 39.1 5.7 Goel et al. 2004). Anatomical connec- Subcortical tions between the posterior hippocam- Culmen L 372 9.4 44.8 6.7 pus, posterior parahippocampal cortices, Main effect of RT and parietal systems (Munoz and Insau- Frontal sti 2005) may constitute a network for Medial Frontal Gyrus R 10349 22.7 0.3 45.5 the acquisition and expression of spatial Medial Frontal Gyrus L 7975 22.3 1.9 43.6 relational learning (Manns and Eichen- Insula R 2969 37.3 13.5 1.5 baum 2006; van Asselen et al. 2006). Insula L 1273 32.5 13.3 4.6 Importantly, the hippocampal acti- Middle Frontal Gyrus R 1204 30.8 34.5 24.7 vations are not attributable to differen- Parietal Inferior Parietal Lobule L 21159 24.5 59.7 38.4 tial dwell time for novel and repeated Inferior Parietal Lobule R 17415 21.9 58.7 43.1 arrays: (1) there was no main effect of RT Occipital within the hippocampus; (2) the main Middle Occipital Gyrus L 9019 35.1 62.8 10.2 effect of array type was orthogonal to Cuneus L 1355 6.5 69 6.3 RT; and (3) in the interaction, only re- Cuneus L 694 1.9 89.7 7.1 Middle Occipital Gyrus R 478 38.5 68.5 7.8 peated arrays showed greater activation Middle Occipital Gyrus R 273 31.3 81.6 16.9 and only at faster RTs. Middle Occipital Gyrus R 248 43.8 55.3 5.7 Whole-brain tables for factors in Subcortical the general linear model (GLM) are Culmen R 696 19.8 55.5 9.7 shown in Table 1. Note that several tem- poral lobe regions to include the MTL Region is defined as center of mass. Coordinates represent distance in millimeters from anterior com- are involved in both the main effect of missure: x right(+)/left(); y anterior(+)/posterior(); z superior(+)/inferior(). Individual voxel prob- ability < 0.005, minimum cluster size > 200 µL. array type and in the interaction of array www.learnmem.org 551 Learning & Memory Downloaded from learnmem.cshlp.org on November 4, 2021 - Published by Cold Spring Harbor Laboratory Press Hippocampal differentiation without awareness type by RT. While the present experiment was designed to test (McEchron and Disterhoft 1997). Whereas the present findings the involvement of the hippocampus proper in the contextual tend to argue against declarative memory as the core function of cueing task, most treatments of declarative memory assert that the hippocampus, future research is needed to determine the neither the hippocampus nor surrounding MTL regions are in- relative merit of binding and context-driven models. volved in implicit tasks (Squire and Zola 1996; Squire et al. 2004); on the other hand, processes involved in familiarity may involve Acknowledgments certain MTL regions, but not the hippocampus (for review, see This study was supported in part by a grant from the National Eichenbaum et al. 2007). Institute of Aging (R01 AG022304) to S.R., and the National Cen- The purpose of this experiment was to provide converging ter for Research Resources, National Institutes of Health (M01- evidence that the hippocampus is involved in implicit contex- RR00058) to the Medical College of Wisconsin General Research Center, and the W.M. Keck Foundation. tual learning. We found that the hippocampal hemodynamic response differentiated novel from repeated items despite the fact References that participants could not differentiate the repeated items in a recognition task. Our findings are consistent with neuropsycho- Binder, J.R., Bellgowan, P.S., Hammeke, T.A., Possing, E.T., and Frost, J.A. 2005. A comparison of two fMRI protocols for eliciting logical findings suggesting that the hippocampus plays a poten- hippocampal activation. Epilepsia 46: 1061–1070. tially important role in this implicit task (Chun and Phelps Chun, M.M. and Jiang, Y. 1998. Contextual cueing: Implicit learning 1999). To our knowledge, no study has found task-differential and memory of visual context guides spatial attention. Cognit. hippocampal activation in a perceptual priming task. Psychol. 36: 28–71. Chun, M.M. and Phelps, E.A. 1999. Memory deficits for implicit This finding also helps establish an important and necessary contextual information in amnesic subjects with hippocampal bridge between amnesic studies (which are limited by small damage. Nat. Neurosci. 2: 844–847. samples and substantial variation in focus and extent of damage) Clark, R.E. and Squire, L.R. 1998. Classical conditioning and brain and functional imaging studies (limited to showing that an area systems: The role of awareness. Science 280: 77–81. Cohen, N.J., Eichenbaum, H., Deacedo, B.S., and Corkin, S. 1985. is involved in a particular process, but not necessarily critical for Different memory systems underlying acquisition of procedural and its performance). Where possible, converging evidence between declarative knowledge. Ann. N. Y. Acad. Sci. 444: 54–71. neuropsychological and functional imaging methodologies pro- Cohen, N.J., Poldrack, R.A., and Eichenbaum, H. 1997. Memory for vides stronger evidence than either can provide alone. items and memory for relations in the procedural/declarative memory framework. Memory 5: 131–178. According to the declarative memory model, context- Cox, R.W. 1996. AFNI: Software foranalysis and visualization of dependent tasks, such as contextual cueing, involve the hippo- functional megnetic resonance neuroimages. Comput. Biomed. Res. campus because the contingencies are sufficiently complex that 29: 162–173. they require reorganization that may only take place as a delib- Davachi, L. and Wagner, A.D. 2002. Hippocampal contributions to episodic encoding: Insights from relational and item-based learning. erate act of memory manipulation (Clark and Squire 1998; Reed J. Neurophysiol. 88: 982–990. and Squire 1999; Manns and Squire 2001). By this account, con- Dolan, R.J. and Fletcher, P.F. 1999. Encoding and retrieval in human text-dependent learning is a subset of declarative memory, which medial temporal lobes: An empirical investigation using functional necessarily requires the hippocampus. Conversely, it has been magnetic resonance imaging (fMRI). Hippocampus 9: 25–34. Eichenbaum, H. 2006. Remembering: Functional organization of the argued that episodic and semantic memory are specific instances declarative memory system. Curr. Biol. 16: R643–R645. of context-dependent learning (Ryan et al. 2000; Greene et al. Eichenbaum, H. and Cohen, N.J. 2001. From conditioning to conscious 2006; Manns and Eichenbaum 2006; Greene 2007). Accordingly, recollection: Memory systems of the brain. Oxford University Press, New declarative memory is encoded as associations to context and the York. Eichenbaum, H., Yonelinas, A.P., and Ranganath, C. 2007. The medial necessary context for recognition and recall is provided by re- temporal lobe and recognition memory. Annu. Rev. Neurosci. 30: trieval cues (Nyberg et al. 1996; Rudy et al. 2002). A third hy- 123–152. pothesis is that the hippocampus is involved in binding nonpre- Eldridge, L.L., Engel, S.A., Zeineh, M.M., Bookheimer, S.Y., and disposed associations. That is, some types of association forma- Knowlton, B.J. 2005. A dissociation of encoding and retrieval processes in the human hippocampus. J. Neurosci. 25: 3280–3286. tion require synaptic connections where existing connectivity is Gabrieli, J.D., Brewer, J.B., Desmond, J.E., and Glover, G.H. 1997. inadequate. Some of these instances occur where the contextual Separate neural bases of two fundamental memory processes in the associations require a complex network (Chun and Phelps 1999; human medial temporal lobe. Science 276: 264–266. Reed and Squire 1999; Ryan et al. 2000; Manns and Squire 2001; Goel, V., Makale, M., and Grafman, J. 2004. The hippocampal system mediates logical reasoning about familiar spatial environments. J. Davachi and Wagner 2002; Greene et al. 2006) and some involve Cogn. Neurosci. 16: 654–664. multiple cortical systems that span distant regions of cortex Gold, J.J. and Squire, L.R. 2005. Quantifying medial temporal lobe (McEchron and Disterhoft 1997; Davachi and Wagner 2002; Ei- damage in memory-impaired patients. Hippocampus 15: 79–85. chenbaum 2006). According to this view, the large proportion of Greene, A.J. 2007. Human hippocampal-dependent tasks: Is awareness necessary or sufficient? Hippocampus 17: 429–433. observed hippocampal-dependent tasks that are declarative oc- Greene, A.J., Gross, W.L., Elsinger, C., and Rao, S.M. 2006. An fMRI curs because declarative memory is generally both context de- analysis of the human hippocampus: Inference, context and task pendent and involves numerous cortical regions (Eichenbaum awareness. J. Cogn. Neurosci. 18: 1156–1173. and Cohen 2001). Greicius, M.D., Krasnow, B., Boyett-Anderson, J.M., Eliez, S., Schatzberg, A.F., Reiss, A.L., and Menon, V. 2003. Regional analysis of The present results, taken together with neuropsychological hippocampal activation during memory encoding and retrieval: findings (Chun and Phelps 1999), provide converging evidence fMRI study. Hippocampus 13: 164–174. to suggest hippocampal involvement in a context-dependent Keane, M.M., Gabrieli, J.D., Mapstone, H.C., Johnson, K.A., and Corkin, task that does not rely on declarative memory. One possibility is S. 1995. Double dissociation of memory capacities after bilateral occipital-lobe or medial temporal-lobe lesions. Brain 118: 1129–1148. that declarative memory is a subset of context-dependent learn- Knowlton, B.J. and Squire, L.R. 1996. Artificial grammar learning ing and memory, and accordingly, contextual learning is closer depends on implicit acquisition of both abstract and exemplar-specific to the central role of the hippocampus. Problematically, learning information. J. Exp. Psychol. Learn. Mem. Cogn. 22: 169–181. tasks that one might argue are contextual in nature may be in- Maguire, E.A., Spiers, H.J., Good, C.D., Hartley, T., Frackowiak, R.S., and Burgess, N. 2003. Navigation expertise and the human dependent of the hippocampus (Knowlton and Squire 1996). hippocampus: A structural brain imaging analysis. Hippocampus Other tasks such as trace conditioning are hippocampal depen- 13: 250–259. dent but have no greater claim to context dependency than com- Manns, J.R. and Eichenbaum, H. 2006. Evolution of declarative parable nonhippocampal-dependent conditioning tasks memory. Hippocampus 16: 795–808. www.learnmem.org 552 Learning & Memory Downloaded from learnmem.cshlp.org on November 4, 2021 - Published by Cold Spring Harbor Laboratory Press Hippocampal differentiation without awareness Manns, J.R. and Squire, L.R. 2001. Perceptual learning, awareness, and hippocampus-dependent memory. J. Neurosci. 26: 11304–11312. the hippocampus. Hippocampus 11: 776–782. Squire, L.R. and Zola, S.M. 1996. Structure and function of declarative Mayanagi, Y., Watanabe, E., Nagahori, Y., and Nankai, M. 2001. and nondeclarative memory systems. Proc. Natl. Acad. Sci. Psychiatric and neuropsychological problems in epilepsy surgery: 93: 13515–13522. Analysis of 100 cases that underwent surgery. Epilepsia 42 (Suppl Squire, L.R., Stark, C.E., and Clark, R.E. 2004. The medial temporal lobe. 6): 19–23. Annu. Rev. Neurosci. 27: 279–306. McEchron, M.D. and Disterhoft, J.F. 1997. Sequence of single neuron Stark, C.E. and Squire, L.R. 2000. Recognition memory and familiarity changes in CA1 hippocampus of rabbits during acquisition of trace judgments in severe amnesia: No evidence for a contribution of eyeblink conditioned responses. J. Neurophysiol. 78: 1030–1044. repetition priming. Behav. Neurosci. 114: 459–467. Milner, B. 1972. Disorders of learning and memory after temporal lobe Strange, B.A., Duggins, A., Penny, W., Dolan, R.J., and Friston, K.J. 2005. lesions in man. Clin. Neurosurg. 19: 421–446. Information theory, novelty and hippocampal responses: Munoz, M. and Insausti, R. 2005. Cortical efferents of the entorhinal Unpredicted or unpredictable? Neural Netw. 18: 225–230. cortex and the adjacent parahippocampal region in the monkey Takashima, A., Petersson, K.M., Rutters, F., Tendolkar, I., Jensen, O., (Macaca fascicularis). Eur. J. Neurosci. 22: 1368–1388. Zwarts, M.J., McNaughton, B.L., and Fernandez, G. 2006. Declarative Nyberg, L., McIntosh, A.R., Houle, S., Nilsson, L.G., and Tulving, E. memory consolidation in humans: A prospective functional 1996. Activation of medial temporal structures during episodic magnetic resonance imaging study. Proc. Natl. Acad. Sci. memory retrieval. Nature 380: 715–717. 103: 756–761. Reed, J.M. and Squire, L.R. 1999. Impaired transverse patterning in Talairach, J. and Tournoux, P. 1988. Co-planar stereotaxic atlas of the human amnesia is a special case of impaired memory for two-choice human brain. Thieme, New York. discrimination tasks. Behav. Neurosci. 113: 3–9. van Asselen, M., Kessels, R.P., Neggers, S.F., Kappelle, L.J., Frijns, C.J., Rudy, J.W., Barrientos, R.M., and O’Reilly, R.C. 2002. Hippocampal and Postma, A. 2006. Brain areas involved in spatial working formation supports conditioning to memory of a context. Behav. memory. Neuropsychologia 44: 1185–1194. Neurosci. 116: 530–538. Zeineh, M.M., Engel, S.A., Thompson, P.M., and Bookheimer, S.Y. 2003. Ryan, J.D., Althoff, R.R., Whitlow, S., and Cohen, N.J. 2000. Amnesia is Dynamics of the hippocampus during encoding and retrieval of a deficit in relational memory. Psychol. Sci. 11: 454–461. face-name pairs. Science 299: 577–580. Smith, C.N. and Squire, L.R. 2005. Declarative memory, awareness, and transitive inference. J. Neurosci. 25: 10138–10146. Smith, C.N., Hopkins, R.O., and Squire, L.R. 2006. Experience-dependent eye movements, awareness, and Received April 17, 2007; accepted in revised form June 25, 2007. www.learnmem.org 553 Learning & Memory Downloaded from learnmem.cshlp.org on November 4, 2021 - Published by Cold Spring Harbor Laboratory Press Hippocampal differentiation without recognition: An fMRI analysis of the contextual cueing task Anthony J. Greene, William L. Gross, Catherine L. Elsinger, et al. Learn. Mem. 2007, 14: Access the most recent version at doi:10.1101/lm.609807 This article cites 40 articles, 8 of which can be accessed free at: References http://learnmem.cshlp.org/content/14/8/548.full.html#ref-list-1 Freely available online through the Learning & Memory Open Access option. License Receive free email alerts when new articles cite this article - sign up in the box at the Email Alerting top right corner of the article or click here. Service Copyright © 2007, Cold Spring Harbor Laboratory Press

Journal

Learning & MemoryUnpaywall

Published: Aug 9, 2007

There are no references for this article.