Elsevier

NeuroImage

Volume 82, 15 November 2013, Pages 44-52
NeuroImage

Dynamic changes in parietal activation during encoding: Implications for human learning and memory

https://doi.org/10.1016/j.neuroimage.2013.05.113Get rights and content

Highlights

  • Ventral parietal activity is assessed during stimulus encoding and subsequent delay.

  • Activity was negatively correlated with subsequent memory during the stimulus phase.

  • During the following cue, activity was positively correlated with memory formation.

  • Ventral parietal cortex may facilitate beneficial post-encoding processes.

Abstract

The ventral posterior parietal cortex (vPPC) monitors successful memory retrieval, yet its role during learning remains unclear. Indeed, increased vPPC activation during stimulus encoding is often negatively correlated with subsequent memory performance, suggesting that this region is suppressed during learning. Alternatively, the vPPC may engage in learning-related processes immediately after stimulus encoding thus facilitating retrieval at a later time. To investigate this possibility, we assessed vPPC activity during item presentation and immediately following its offset when a cue to remember was presented. We observed a dynamic change in vPPC response such that activity was negatively correlated with subsequent memory during stimulus presentation but positively correlated immediately following the stimulus during the cue phase. Furthermore, regional differences in this effect suggest a degree of functional heterogeneity within the vPPC. These findings demonstrate that the vPPC is engaged during learning and acts to facilitate post-encoding memory processes that establish long-term cortical representations.

Introduction

Advances in our understanding of the neural bases of human memory have implicated a broad cortical network involved in the encoding, retention, and ultimate retrieval of recently learned information. While early investigations focused on the medial temporal lobe (MTL) and prefrontal cortex (PFC; Scoville and Milner, 1957, Shimamura, 1995, Squire and Alvarez, 1995), recent neuroimaging investigations have identified regions in the posterior parietal cortex (PPC) that are particularly active when an item has been successfully retrieved (Cabeza et al., 2012, Konishi et al., 2000, Shimamura, 2011, Wagner et al., 2005). Retrieval-related activations within the PPC are functionally dissociable such that activity related to low confidence, familiarity based responses are clustered within the dorsal PPC (dPPC), and activity related to high confidence, recollective responses are clustered within the ventral PPC (vPPC) regions (Kim and Cabeza, 2009, Wheeler and Buckner, 2004). Furthermore, these retrieval-related activations appear to be dissociable from attention-related activity such that memory tasks generally elicit left-lateralized activity in the angular gyrus (AnG), whereas activations occurring during attention tasks are clustered in right temporo-parietal junction (TPJ) and supra-marginal gyrus (SMG; Hutchinson et al., 2009).

Although neuroimaging findings suggest an important contribution of the vPPC to successful retrieval, its role during memory encoding is unclear. In some studies, the vPPC was negatively correlated with subsequent memory, such that items later remembered elicit greater vPPC deactivation during encoding compared to items later remembered, a phenomenon referred to as a negative subsequent memory effect (SME; see Uncapher and Wagner, 2009). Extant theories of PPC contributions to memory have struggled to explain this apparent conflict between the negative impact during encoding and positive influence during recognition tests, a pattern described as the encoding/retrieval flip (Daselaar et al., 2009). One possibility is that the negative influence of vPPC activity during encoding is related to its role as part of the default mode network (DMN), which also includes regions within the PFC, MTL, and medial PPC. Initially, the DMN was found to be more active during rest periods and inter-trial intervals compared to moments when participants were engaged in task-relevant activity (Buckner et al., 2008, Raichle et al., 2001). Recent findings suggest that this network is suppressed during perceptually-driven/externally attended conditions and engaged during conceptually-driven/internally attended situations (Guerin et al., 2012, Sestieri et al., 2010). Thus, findings of increased vPPC (i.e., DMN) activity at encoding for subsequently forgotten items suggest that participants during such encoding trials may have been sacrificing stimulus-driven encoding for irrelevant conceptually-driven processing (e.g., mind wandering).

The role of functionally distinct sub-regions within the vPPC may also help explain differential effects associated with the encoding/retrieval flip. The anterior portion of the vPPC, including the TPJ and SMG, are thought to be a part of the ventral attention network involved in attentional reorienting (Corbetta et al., 2008, Shulman et al., 2007) and task-switching (Otten and Rugg, 2001, Wagner and Davachi, 2001). During perceptual search, activity within the TPJ/SMG is often suppressed, a phenomenon proposed to reflect filtering (down-regulating) of task-irrelevant stimuli (Shulman et al., 2007). Activation of these areas at encoding is thought to reflect inadvertent bottom–up capture by irrelevant stimulus features or shifts of attention away from task-relevant features (Otten and Rugg, 2001, Uncapher et al., 2011). Either of these possibilities would serve to divert resources away from processes such as elaborative encoding, which would contribute to successful memory formation.

To the extent that vPPC activity is involved in internally mediated mnemonic processes, it is possible that this region is suppressed during stimulus presentations, but becomes engaged immediately thereafter. This possibility is consistent with a recent theory of episodic memory which suggests that the vPPC is critical for the cortical binding of relational activity (CoBRA; Shimamura, 2011). According to CoBRA, the vPPC acts as a convergence zone that enables the binding of event features as an encapsulated episodic representation. Initial binding of event features is established by the MTL, which is also involved in post-encoding strengthening of episodic representations (i.e. consolidation) through reactivation or replay of event features (Eichenbaum et al., 1992, Shimamura and Wickens, 2009). The vPPC participates in cortical binding by establishing more direct links between event features during reactivation (e.g., rehearsal, elaborative encoding). We addressed the role of the vPPC in facilitating memory processes during learning by assessing neural correlates of SME immediately following stimulus presentation. In order to differentially emphasize memory processing during this time period, we presented one of two cues just after stimulus presentation which instructed subjects to either remember the item or ignore it. Later, subjects were given a recognition memory test and the SME was assessed for activity during both stimulus presentation and the cue phase.

Section snippets

Participants

Nineteen healthy subjects were included in this study (mean age 22.05 years, range 18 to 34 years; 13 females). Recruitment was conducted via advertisement on the UC Berkeley Department of Psychology website (http://psychology.berkeley.edu/rsvp). All subjects were native English speakers and were right-handed. None of the subjects reported a history of neuropsychiatric disorders or brain injury or having recently taken psychoactive medication. Subjects were paid for their participation and gave

Behavioral results

Subsequent memory performance for items cued as Remember (71.3% correct) was significantly better than for items cued as Ignore (66.3% correct), [t(17) = 2.35, p = 0.03]. Memory for both conditions was well above chance performance (50%). Response latencies for the People/No People decision were similar between conditions (Remember = 1415.83, Ignore = 1410.61 ms), [t(17) = 0.220, p = 0.83].

fMRI results

We first performed a standard subsequent memory effect (SME) analysis in which item encoding (i.e., stimulus phase)

Discussion

Consistent with previous findings, vPPC activity during stimulus encoding was negatively correlated with subsequent memory (Buckner et al., 2008, Daselaar et al., 2004, Otten and Rugg, 2001). Yet immediately after stimulus presentation, when subjects were presented a cue to remember the item, positive SMEs were observed in the vPPC. This study is the first to demonstrate a dynamic interplay in which both negative and positive SMEs in the vPPC occur during learning. Thus, these regions are not

Acknowledgments

This research was supported by the NSF Grant BCS-0745835 and NIH Grant NS040813.

Conflict of interest statement

The authors have no conflict of interest to declare.

References (54)

  • H. Kim et al.

    Common and specific brain regions in high- versus low-confidence recognition memory

    Brain Res.

    (2009)
  • S. Konishi et al.

    Neural correlates of episodic retrieval success

    NeuroImage

    (2000)
  • S.M. Nelson et al.

    A parcellation scheme for human left lateral parietal cortex

    Neuron

    (2010)
  • L.J. Otten et al.

    When more means less

    Curr. Biol.

    (2001)
  • Z.A. Rosner et al.

    The generation effect: activating broad neural circuits during memory encoding

    Cortex

    (2013)
  • L.R. Squire et al.

    Retrograde amnesia and memory consolidation: a neurobiological perspective

    Curr. Opin. Neurobiol.

    (1995)
  • B.P. Staresina et al.

    Mind the gap: binding experiences across space and time in the human hippocampus

    Neuron

    (2009)
  • N. Tzourio-Mazoyer et al.

    Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain

    NeuroImage

    (2002)
  • M.R. Uncapher et al.

    Posterior parietal cortex and episodic encoding: insights from fMRI subsequent memory effects and dual-attention theory

    Neurobiol. Learn. Mem.

    (2009)
  • D.C. Van Essen

    A population-average, landmark-and surface-based (PALS) atlas of human cerebral cortex

    NeuroImage

    (2005)
  • K.L. Vilberg et al.

    Memory retrieval and the parietal cortex: a review of evidence from a dual-process perspective

    Neuropsychologia

    (2008)
  • A.D. Wagner et al.

    Cognitive neuroscience: forgetting of things past

    Curr. Biol.

    (2001)
  • A.D. Wagner et al.

    Parietal lobe contributions to episodic memory retrieval

    Trends Cogn. Sci.

    (2005)
  • M.E. Wheeler et al.

    Functional–anatomic correlates of remembering and knowing

    NeuroImage

    (2004)
  • A. Ben-Yakov et al.

    Constructing realistic engrams: poststimulus activity of hippocampus and dorsal striatum predicts subsequent episodic memory

    J. Neurosci.

    (2011)
  • B.R. Buchsbaum et al.

    Recency effects in the inferior parietal lobe during verbal recognition memory

    Front. Hum. Neurosci.

    (2011)
  • R.L. Buckner et al.

    The brain's default network: anatomy, function, and relevance to disease

    Ann. N. Y. Acad. Sci.

    (2008)
  • Cited by (21)

    • Post-encoding frontal theta activity predicts incidental memory in the reward context

      2019, Neurobiology of Learning and Memory
      Citation Excerpt :

      Another possibility is that the post-encoding activity is associated with the initial stages of memory consolidation. A recent study, using a different approach in which a post-encoding cue instructed participants to remember or to forget the preceding stimulus, indicated that post-encoding ventral posterior parietal cortex (vPPC) activity was positively correlated with subsequent memory and was involved in the consolidation of cortically binding features of the prior stimulus (Elman, Rosner, Cohn-Sheehy, Cerreta, & Shimamura, 2013). These studies suggest a significant role of immediate post-encoding brain activity in predicting memory performance.

    • Contributions of the ventral parietal cortex to declarative memory

      2018, Handbook of Clinical Neurology
      Citation Excerpt :

      As illustrated by Figure 27.4, the same VPC regions that showed less activity for subsequently remembered than forgotten trials during encoding showed greater activity for remembered than forgotten trials during retrieval (Daselaar et al., 2009); this finding was consistent across several experiments using different stimuli. The E-R flip can be also observed when encoding and retrieval are separated by a few seconds (Elman et al., 2013b). In SM retrieval, VPC typically shows greater activity for stimuli that are more meaningful (e.g., words) than for stimuli that are less meaningful (e.g., pseudowords).

    • Peri-encoding predictors of memory encoding and consolidation

      2015, Neuroscience and Biobehavioral Reviews
      Citation Excerpt :

      In subsequent studies, they demonstrated that retroactively-interfering stimuli attenuated the hippocampal offset response in a manner corresponding to the behavioral interference (Ben-Yakov et al., 2013) and that the magnitude of the offset response is modulated by stimulus novelty (Ben-Yakov et al., 2014). Using a different approach, Elman et al. (2013) measured activity in response to a post-stimulus cue instructing participants to remember or to forget the preceding stimulus. They presented photographs of outdoor scenes and several seconds later a cue appeared indicating whether the photograph should be remembered for a subsequent recognition test.

    • A parietal memory network revealed by multiple MRI methods

      2015, Trends in Cognitive Sciences
      Citation Excerpt :

      Although early neuroimaging studies focused on the medial temporal and frontal lobes, parietal cortex has consistently been linked to human memory processing [10,11]. Numerous studies have associated parietal cortex with retrieval-success effects (see Glossary) [12–15], but regions of parietal cortex also show differential activity for: (i) subsequently forgotten, relative to subsequently remembered, items during memory encoding [16,17]; (ii) intentional retrieval tasks relative to intentional encoding tasks [18]; (iii) items that have been encoded multiple times relative to items that are encoded for the first time [19,20]; (iv) false alarms (new items judged as old) relative to misses (old items judged to be new) [21]; (v) more confidently, relative to less confidently, retrieved items [6]; and (vi) items for which rich, relative to sparse, contextual information is accessible at retrieval [7,22]. Hypotheses have been forwarded to explain these different patterns of data (see [10,23] for reviews), but no single hypothesis can account for all of the patterns described above.

    View all citing articles on Scopus
    View full text