Effect of aging differs for memory of object identity and object position within a spatial context

There has been considerable focus on investigating age-related memory changes in cognitively healthy older adults, in the absence of neurodegenerative disorders. Previous studies have reported age-related domain-specific changes in older adults, showing increased difficulty encoding and processing object information but minimal to no impairment in processing spatial information compared with younger adults. However, few of these studies have examined age-related changes in the encoding of concurrently presented object and spatial stimuli, specifically the integration of both spatial and nonspatial (object) information. To more closely resemble real-life memory encoding and the integration of both spatial and nonspatial information, the current study developed a new experimental paradigm with novel environments that allowed for the placement of different objects in different positions within the environment. The results show that older adults have decreased performance in recognizing changes of the object position within the spatial context but no significant differences in recognizing changes in the identity of the object within the spatial context compared with younger adults. These findings suggest there may be potential age-related differences in the mechanisms underlying the representations of complex environments and furthermore, the integration of spatial and nonspatial information may be differentially processed relative to independent and isolated representations of object and spatial information.

Advancing age is associated with changes in a number of cognitive domains (Erickson and Barnes 2003; Salthouse 2004; Craik and Bialystok 2006). Particularly age-related changes in episodic memory function have been frequently reported (Grady and Craik 2000; Craik and Bialystok 2006). In addition to a general decline in long-term retention, older adults show a reduced ability to differentiate between highly similar object representations (Yassa et al. 2011; Stark et al. 2013, 2015; Reagh et al. 2016; Berron et al. 2018) and object features (Yeung et al. 2017) compared with young adults. In contrast, spatial representations appear to be relatively spared (Fidalgo et al. 2016; Stark and Stark 2017) with older adults showing performance similar to young adults recognizing subtle changes in a spatial environment (Berron et al. 2018) and changes in the location of an object when presented on a blank screen (Reagh et al. 2016, 2018).The dissociation and integration of object and spatial information has been a key question in memory research. Older adults show significant impairments in memory binding and maintaining associations despite having intact memory for the individual items (Chalfonte and Johnson 1996; Naveh-Benjamin 2000; Old and Naveh-Benjamin 2008). This is also observed in object-location binding with impairments in recalling the specific location of objects (Kessels et al. 2007; Berger-Mandelbaum and Magen 2019; Muffato et al. 2019) as well as recalling the identity of an object within an environment (Schiavetto et al. 2002; Kessels et al. 2007; Mazurek et al. 2015). The binding of object-location information has been hypothesized to involve the medial temporal lobes, including the hippocampus (Postma et al. 2008), although age-related changes in the prefrontal cortex, posterior neocortex and other regions have also been implicated in object location and object identity tasks (Schiavetto et al. 2002; Meulenbroek et al. 2010). In the medial temporal lobes, the integration of object and spatial information is thought to arise from two parallel information processing streams (Eichenbaum 1999; Eichenbaum et al. 1999; Davachi 2006; Ranganath and Ritchey 2012; Knierim et al. 2013). One pathway, commonly referred to as the “what” pathway involves the perirhinal cortex and the lateral entorhinal cortex, and is thought to predominately process information about objects, items and events, while the “where” pathway involving the parahippocampal cortex and the medial entorhinal cortex is thought to process contextual and spatial information. Information from both pathways is projected to the hippocampus, which is then thought to integrate the spatial and nonspatial information into a cohesive “memory space” through a mechanism that is common to both object or episodic and spatial information (Eichenbaum et al. 1999).However, emerging evidence suggests the processing of object and spatial information may be more integrated than previously thought with the lateral entorhinal cortex processing multimodal information, receiving both object and spatial information (Witter et al. 2017; Doan et al. 2019; Nilssen et al. 2019). In rodent studies, the lateral entorhinal cortex has been reported to be involved in the encoding of features from both the object and the environment (Deshmukh and Knierim 2011; Yoganarasimha et al. 2011; Deshmukh et al. 2012; Knierim et al. 2013). Rodent studies using single cell recordings in the lateral entorhinal cortex show that neurons in this region encode object-related information as well as spatial information about the object (e.g., position in relationship to the environment). These studies also show cells in the lateral entorhinal cortex that track the position of an object in the environment and do not fire when that object is no longer present (Deshmukh and Knierim 2011). A different subset of cells (“object trace cells”) have been reported to fire in previously experienced positions of an object within an environment (Deshmukh and Knierim 2011; Tsao et al. 2013). Lateral entorhinal cortex lesioned rodents show no impairment in performing an object-recognition task, but are impaired at recognizing spatial changes and object changes within a set of objects in an environment including position changes of the objects (Van Cauter et al. 2013) and previously learned object-place and object-context associations (Wilson et al. 2013a; Chao et al. 2016). Together, these findings suggest that, beyond encoding information about objects, the lateral entorhinal cortex encodes contextual information and may be binding nonspatial and spatial information, specifically encoding information about objects and certain spatial properties, including information about the object''s position within environment.Few studies have examined the role of the lateral entorhinal cortex in encoding object identity, object position or changes to the spatial context in humans. Reagh and Yassa (2014) report that subtle perceptual differences between similar objects (e.g., two slightly different apples) elicits activity observed with functional magnetic resonance imaging (fMRI) in both the lateral entorhinal cortex and perirhinal cortex, while changes to an object''s position on a blank screen elicited activity in both the parahippocampal cortex and medial entorhinal cortex. Subsequent studies in older adults show impaired performance recalling the identity of object (Reagh and Yassa 2014; Stark and Stark 2017; Yeung et al. 2017; Berron et al. 2018; Reagh et al. 2018) but similar performance recalling position of the object on a screen compared with young adults (Reagh et al. 2016, 2018). However, these studies examined memory for object identity and object position on a blank screen, devoid of any spatial or contextual information. Given the findings from rodent studies, it appears that object identity and object position information are represented in relationship to the spatial environment in which they occur.To examine the integration of nonspatial and spatial information, novel stimuli were developed to mimic real-life environments where objects could occur within the environment, more closely resembling animal studies in which rodents experience objects within an environment. A series of scenes were designed to have the same perspective, spatial dimensions and outdoor scenery (Fig. 1A). Scenes were classified into five general categories: living room, dining room, kitchen, bedroom, and office rooms to allow for the placement of categorically congruent furniture (Fig. 1B). Critically, within each scene, two to five different positions were defined that an object could logically occupy (Fig. 1C). The scene stimuli were first validated using mnemonic ratings and subsequently used in a novel object-in-context task to assess memory for object identity and object position in context and examine age-related changes in performance on this task in cognitively normal older adults compared with young adults.Open in a separate windowFigure 1.Task stimuli. (A) Scenes were designed to have identical dimensions and a similar perspective. (B) Scenes were designed to have two to five different positions where an object could be reasonably placed. Across all scenes, the same general positions were available for object placement. (C) Example of a scene with an object as seen by the participant. No object was present in the scenes for the stimulus validation study.     

A study on episodic memory reconsolidation that tells us more about consolidation

Awake quiescence immediately after encoding is conducive to episodic memory consolidation. Retrieval can render episodic memories labile again, but reconsolidation can modify and restrengthen them. It remained unknown whether awake quiescence after retrieval supports episodic memory reconsolidation. We sought to examine this question via an object-location memory paradigm. We failed to probe the effect of quiescence on reconsolidation, but we did observe an unforeseen “delayed” effect of quiescence on consolidation. Our findings reveal that the beneficial effect of quiescence on episodic memory consolidation is not restricted to immediately following encoding but can be achieved at a delayed stage and even following a period of task engagement.

For labile new memories to be remembered they must be consolidated, that is, strengthened and stabilized, over time (Dudai 2004; Wixted 2004). Extensive research demonstrates that postencoding sleep and awake quiescence (quiet rest) are conducive to human episodic memory consolidation (Clemens et al. 2005; Ferrara et al. 2008; Lahl et al. 2008; Wamsley et al. 2010; Dewar et al. 2012, 2014; Gaskell et al. 2014; Mercer 2015; Brokaw et al. 2016; Craig and Dewar 2018; Craig et al. 2019; Sacripante et al. 2019). Sleep and wakeful rest are hypothesized to support consolidation by providing a state of reduced sensory input and task engagement (Hasselmo 1999; Wixted 2004; Mednick et al. 2011; Dewar et al. 2012, 2014; Craig and Dewar 2018; Craig et al. 2019).Memories are not destined to remain fixed and unmodifiable once consolidated. An increasing body of evidence suggests that the cueing or retrieval of consolidated memories, for example, via internally or externally generated exposure to learned materials (Wichert et al. 2011; Agren 2014), can return such memories into a labile state (Dudai and Eisenberg 2004; McKenzie and Eichenbaum 2011; Schwabe et al. 2014). These relabilized memories are thought to require a process of “reconsolidation” (Rodriguez et al. 1993; Przybyslawski and Sara 1997) in order to restabilize and persist. Evidence for reconsolidation in humans comes from research demonstrating that memories can be disrupted and/or updated by the application of treatments shortly following their retrieval. This includes (1) the disruption and modification of episodic memories via the presentation of similar/related materials (Loftus 2005; Hupbach et al. 2007), (2) a reduction in the number of traumatic memory intrusions following postretrieval engagement in visuospatial tasks (Deeprose et al. 2012; James et al. 2015; Hagenaars et al. 2017), and (3) the extinction of fear memories and increased forgetting of emotional memories via the administration of pharmaceuticals and electrophysiological stimulation (Cahill et al. 1994; Debiec and Ledoux 2004; Kroes et al. 2014). In all cases, the effect of postretrieval treatment is seen only for memories that are retrieved (e.g., via a memory test or cue) prior to the treatment. Memories that were not retrieved (i.e., controls) were unaffected, indicating that the treatment hampered reconsolidation specifically.Episodic memory reconsolidation can also be enhanced. Of particular interest to our study is the finding that reconsolidation, like consolidation, benefits from sleep. If recently or remotely encoded episodic memories are reactivated before an extended period of sleep, the reactivated memories are protected against further forgetting (for reviews, see Stickgold and Walker 2007; Dudai 2012; Spiers and Bendor 2014). Moreover, even naps benefit reconsolidation. It has been shown that a short 40-min nap period facilitated the reconsolidation of remote memories, whereas recently encoded memories did not benefit at all (Klinzing et al. 2016). It is possible that, as is the case for consolidation, sleep is conducive to reconsolidation because it provides a state of reduced cognitive engagement and sensory input (Hasselmo 1999; Wixted 2004; Mednick et al. 2011; Dewar et al. 2012). If so, awake quiescence, which has this in common with sleep, should also be conducive to reconsolidation.In the study reported here, we examined whether a period of awake quiescence following retrieval protects episodic memories from further forgetting by facilitating their reconsolidation. To this end, we combined a computerized item-location memory test with a modified version of our established consolidation paradigm, which compares between-subject effects of postencoding rest versus task engagement on recently encoded memories (e.g., Dewar et al. 2012). Item-location memory tests have previously been sensitive to the effects of postretrieval (reconsolidation) manipulations, including sleep (e.g., Klinzing et al. 2016) hence, subtle differences in memory scores can be recorded. Figure 1 provides an overview of the procedure. Sixty young adults (mean age = 21.47 yr, SD = 1.79 yr; 26M:34F) were sequentially presented photos of 60 everyday items in unique locations on a 22-in touchscreen computer monitor. Immediately after each object''s presentation, participants were asked to recall the object''s location via a touchscreen response as part of a “short-term spatial memory test” (encoding). They were unaware that they would perform subsequent long-term memory tests pertaining to the locations of these items. Location memory for half (N = 30) of the encoded items (“retrieved items”) was probed in a first delayed recall test that occurred after a 10-min delay filled with a visual spot-the-difference game (Dewar et al. 2012). There was no overlap between encoded items and spot-the-difference game stimuli. The first delayed recall test was followed by one of two postretrieval 10-min delay periods: (1) awake quiescence (quiet resting under strict conditions of minimal sensory input in a dimly lit room; N = 30 participants), or (2) a further filler task (a spot-the-difference game comprising different stimuli to earlier; N = 30 participants) (Dewar et al. 2012). Memory for the location of all 60 encoded items was then probed via a second delayed recall test. This test included the 30 items that were probed during the first delayed recall test (retrieved items) and the 30 control items that were not probed during the first delayed recall test (nonretrieved items). Our key measure was the distance of error (in centimeters) between the initially presented locations of the everyday objects during encoding, and the recalled location during the first and second delayed recall tests. We also recorded the time that participants took to respond as well as confidence ratings during encoding and the first and second delayed recall tests. Analyzes were performed using SPSS 19 using ANOVAs, RM ANOVAs. and follow up t-tests that were used to investigate possible between-group differences. See the Supplemental Materials for detailed methods.Open in a separate windowFigure 1.Experimental paradigm. Participants were presented 60 photos of unique everyday items from the Mnemonic Similarity Task (e.g., Stark et al. 2013) in unique locations on a computer screen. They then experienced 10 min of a postencoding delay task (a spot-the-difference game). Memory for the location of half of the encoded items (N = 30) was then probed via a first delayed recall test (via touchscreen response), before participants completed one of two postretrieval delay periods, each of which was 10 min in duration: awake quiescence (N = 30) (A), or an engaging perceptual task (a further spot-the-difference game comprising different stimuli to the earlier one) (N = 30) (B). In the subsequent second delayed recall test, participants’ memory for the location of all the retrieved and nonretrieved items was probed (N = 60) (via touchscreen response). The experimental procedure took place in a single session.Given that our manipulation occurred during the postretrieval delay, we expected groups to be matched in their encoding, postencoding delay task, and first delayed memory test. Crucially, if postretrieval awake quiescence is conducive to episodic memory reconsolidation, then those who experience awake quiescence after the first delayed recall test should demonstrate less forgetting of “retrieved” items (smaller distances of error scores) in the second delayed recall test relative to those who experience the filler task after the first delayed recall test. Given that “control” items were not retrieved during the first delayed recall test immediately prior to our experimental manipulation (postretrieval quiescence vs. task), we predicted that memory (distance of error scores) for control items should be matched between groups in the second delayed recall test.Data from two participants (perceptual task group) were lost due to technical issues. Thus, the following analyzes report data from 58 participants (quiescence group: N = 30, perceptual task group: N = 28). The two groups were well matched in their backgrounds and performance of the postencoding delay task (spot-the-difference game) (see the Supplemental Material). Moreover, as expected, the two groups were well matched in all memory measures occurring prior to our experimental manipulation. Specifically, we found no significant main effect of group in the mean distance of error (in centimeters) during (1) encoding of the items (quiescence: mean = 1.84 cm, SD = 0.57; perceptual task: mean = 2.00 cm, SD = 0.47; F_(1,56) = 1.303, P = .259, η²_ρ = 0.023), (2) the first delayed recall test of the 30 “retrieved” items (quiescence: mean = 13.36 cm, SD = 3.37; perceptual task: mean = 14.17 cm, SD = 3.67; F_(1,56) = = 0.775, P = 0.383, η²_ρ = 0.014), or (3) the response times and confidence ratings in the first delayed recall test (see the Supplemental Material).The second delayed recall test was completed immediately following our experimental manipulation, where participants experienced 10 min of either (1) awake quiescence (N = 30 participants), or (2) ongoing sensory input and cognitive engagement via an unrelated perceptual task (a different spot-the-difference game; Dewar et al. 2012) (N = 30 participants). This delayed recall test included the 30 items that were probed during the first delayed recall test (retrieved items) and the 30 control items that were not probed during the first delayed recall test (nonretrieved items). A RM ANOVA comprised of within-subject factor item type (retrieved vs. nonretrieved) and between-subject factor group (awake quiescence vs. perceptual task) revealed no significant main effect of group in distance of error scores (in centimeters; F_(1,56) = 0.686, P = 0.411, η²_ρ = 0.012). We did, however, find a significant main effect of item type (retrieved vs. nonretrieved; F_(1,56) = 14.138, P < 0.001, η²_ρ = 0.202) because, overall, the distance of error (in centimeters) was larger for nonretrieved (mean = 14.56 cm, SD = 3.49 cm) than retrieved (mean = 13.36 cm, SD = 3.38 cm) items. There was a significant interaction between group (awake quiescence vs. perceptual task) and item type (retrieved vs. nonretrieved; F_(1,56) = 4.561, P = 0.037, η²_ρ = 0.075).Paired t-tests revealed that the above interaction emerged because, in the perceptual task group, the distance of error was significantly larger for nonretrieved (mean = 15.27, SD = 3.32 cm) than retrieved (mean = 13.37 cm, SD = 3.64 cm) items (t₍₂₇₎ = −5.05, P < 0.001; see Fig. 2). This significant finding survived Bonferroni correction for multiple comparisons (α level of 0.050/four comparisons = corrected α level of 0.013). This was not the case in the quiescence group, where no significant item type difference was observed (retrieved: mean = 13.36 cm, SD = 3.49 cm; nonretrieved: mean = 13.89 cm, SD = 3.26 cm; t₍₂₉₎ = −1.018, P = 0.317). Independent t-tests revealed no effect of delay group in the distance of error for retrieved items (quiescence: mean = 13.36 cm, SD = 3.49 cm; task: mean = 13.37, SD = 3.33; t₍₅₆₎ = −0.008, P = 0.994) or nonretrieved items (quiescence: mean = 13.88 cm, SD = 3.26 cm; task: mean = 15.27 cm, SD = 3.64 cm; t₍₅₆₎ = −1.530, P = 0.132). The groups were matched in their second delayed recall test response times and confidence ratings (see the Supplemental Material).Open in a separate windowFigure 2.Memory performance in the second delayed recall test. Mean distance of error (in centimeters) in the second delayed recall test for items that were and were not “retrieved,” that is, probed, during the first delayed recall test. The data for both the quiescence and task groups are shown. Distance of error refers to the deviation in location from the initially presented location of each item. Error bars show the standard error of the mean. Full lines show between-group comparisons (quiescence vs. task) of distance of error (in centimeters) scores, and dashed lines show within-subject comparisons (retrieved vs. nonretrieved).Our data reveal comparable delayed recall performance for recently retrieved item locations in participants who rested quietly after retrieval and those who played a spot-the-difference game after retrieval (see the “retrieved” data in Fig. 2). Although these findings could suggest that awake quiescence had no effect on the reconsolidation of item-location memories, methodological shortfalls (discussed below) render this conclusion unlikely. However, we did find an unforeseen beneficial effect of quiescence for nonretrieved versus retrieved photo locations: Those who experienced the spot-the-difference game in the second delay phase demonstrated significant forgetting of nonretrieved items relative to retrieved items, whereas those who experienced awake quiescence in the second delay phase did not demonstrate such forgetting (see the “nonretrieved” data in Fig. 2). We discuss these two key findings in turn.Why did we not observe any effects of postretrieval awake quiescence on reconsolidation in our study? One possibility is that, unlike sleep, awake quiescence simply does not benefit reconsolidation. A much more likely possibility is that methodological shortfalls mean that our study did not in fact probe reconsolidation. This may be true for a few reasons. First, sometimes retrieval is not sufficient to reactivate memories and reintroduce them into a labile state (e.g., Cammarota et al. 2004; Forcato et al. 2009). In fact, the protective effect of retrieval against forgetting is well documented, even over a delay of a few minutes (Karpicke and Roediger 2008; Roediger and Butler 2011), possibly because it provides a “fast route to consolidation” of new memories (Antony et al. 2017). Notwithstanding the power of retrieval to boost memory, the research discussed earlier shows that memories reactivated via retrieval can (still) benefit from postretrieval sleep. Do these studies differ from our own with regards to memory reactivation? Possibly, reactivation in our study occurred after a short delay (10 min). It is possible that retrieval was so effective at protecting against forgetting in our study because it occurred shortly after encoding and during the initial consolidation of memory traces. Indeed, studies investigating the role of different treatments (e.g., sleep, pharmaceuticals) in reconsolidation typically require participants to retrieve previously encoded memories after a much longer delay period (several hours to days). After lengthy delays such as those, encoded memories would be expected to have completed initial consolidation, or at least be much further along in the consolidation process than in our study. The 10-min delay in our study might simply have been too short for memories to consolidate sufficiently to be relabilized again. As a result, retrieval might have simply strengthened memories against forgetting and did not sufficiently return them to a labile state. These methodological shortfalls mean that our study is unlikely to have probed reconsolidation effectively. Therefore, we cannot conclude whether postretrieval awake quiescence benefits the reconsolidation of episodic memories.How can we explain the unforeseen finding that “postretrieval” awake quiescence protected against subsequent forgetting for nonretrieved (control) items, which were not probed in the first delayed recall test? Our data reveal an effect of delayed rest that was on par with the overall beneficial effect of retrieval-mediated memory reactivation (i.e., the well-established effect of retrieval practice; see above). The findings suggest that in the task group, the lack of reactivation for nonretrieved items in the first delayed recall test negatively affected these memories. However, in the quiescence group, this lack of reactivation was offset by delayed rest, which seemingly had a similarly beneficial effect on nonretrieved memories. It is unlikely that this benefit of delayed rest can be explained by intentional rehearsal. The occurrence of quiet rest after a 10-min filled delay, as opposed to immediately after encoding, makes it unlikely that participants rehearsed the nonretrieved items during the rest period because they would require explicit knowledge that they had been tested on only half of the encoded items in the first delayed recall test and spend their time rehearsing what they could recall from the subset of items that were not tested. While this does not rule out rehearsal completely, our findings reinforce recent data (e.g., Dewar et al. 2014) showing that rest-related enhancements in memory cannot simply be explained by intentional rehearsal. A consolidation account can provide a more likely explanation.Previous research on the beneficial effect of awake quiescence on memory has focused on immediate rest (i.e., rest periods occurring immediately following encoding) (Dewar et al. 2007, 2012, 2014; Craig et al. 2015, 2016; Mercer 2015; Brokaw et al. 2016; Craig and Dewar 2018; Humiston and Wamsley 2018; Sacripante et al. 2019). The rationale for this focus has been that new memories are most labile upon their initial formation (Mednick et al. 2011; Wixted and Cai 2013), and thus, supportive or disruptive interventions applied during this time should be most effective. The findings of the current study suggest that the effect of awake quiescence is sufficiently powerful to influence new memories even 10–20 min after encoding, that is, when memories have—presumably—benefited from some initial consolidation and are no longer highly labile. Moreover, they indicate that awake quiescence can support the consolidation of new memories even if the awake quiescence occurs after a period of task engagement rather than immediately after encoding. The latter hypothesis resonates with the view that consolidation is an opportunistic process (Mednick et al. 2011) and raises interesting questions about how late the onset of rest would need to be for it to no longer facilitate the consolidation of new memories. This remains to be established, but we predict that the benefit of quiescence should diminish in accordance with increasing delay in the onset of the rest period as memories are increasingly stabilized through consolidation that occurs independently of that during intentional rest. Investigation of this question may provide insights into the initial timeline of episodic memory consolidation, which is characterized poorly. Thus, more work is required to establish whether a relationship exists between the precision of item location memory and the time lag between initial encoding and postencoding quiescence.In summary, methodological shortfalls mean that we failed to probe the possible effect of awake quiescence on episodic memory reconsolidation. Nonetheless, we unexpectedly revealed that awake quiescence following encoding reduces forgetting, even when experienced at a delayed point in time and following a period of task engagement. This finding indicates that awake quiescence need not commence immediately following encoding to benefit memory consolidation, at least for object location memories. Further investigation and independent replication of this effect are required to better understand the underlying mechanisms and implications of our unexpected finding.     

Medial prefrontal cortex and hippocampal activity differentially contribute to ordinal and temporal context retrieval during sequence memory

Remembering sequences of events defines episodic memory, but retrieval can be driven by both ordinality and temporal contexts. Whether these modes of retrieval operate at the same time or not remains unclear. Theoretically, medial prefrontal cortex (mPFC) confers ordinality, while the hippocampus (HC) associates events in gradually changing temporal contexts. Here, we looked for evidence of each with BOLD fMRI in a sequence task that taxes both retrieval modes. To test ordinal modes, items were transferred between sequences but retained their position (e.g., AB3). Ordinal modes activated mPFC, but not HC. To test temporal contexts, we examined items that skipped ahead across lag distances (e.g., ABD). HC, but not mPFC, tracked temporal contexts. There was a mPFC and HC by retrieval mode interaction. These current results suggest that the mPFC and HC are concurrently engaged in different retrieval modes in support of remembering when an event occurred.

Memory for sequences of events is a fundamental component of episodic memory (Tulving 1984, 2002; Allen and Fortin 2013; Howard and Eichenbaum 2013; Eichenbaum 2017). While different experiences share overlapping elements, the sequence of events is unique. Remembering the order of events allows us to disambiguate episodes with similar content and make detailed predictions supporting decision-making.At least two complementary memory processes contribute to the retrieval of events in the correct sequence: ordinal (Orlov et al. 2002) and temporal context (Howard and Kahana 2002) retrieval modes. Whether these disparate retrieval modes operate coincidently or not remains an open question with consequences for understanding basic mechanisms of how we remember the events that unfold throughout our day. According to an ordinal retrieval mode, items are remembered by their position within an event sequence (DuBrow and Davachi 2013; Allen et al. 2014; Long and Kahana 2019), providing sequential memory through well-established semantic or abstracted relationships (first, second, third, etc.). While for a temporal context retrieval mode, events are remembered through a gradually changing temporal context within which specific items have been associated. According to temporal contexts, when an element of a sequence is presented or retrieved (e.g., “C” in ABCDEF), items that are more proximal in the sequence (e.g., the “D” in the sequence) have a higher retrieval rate compared with items that are further away (e.g., the “F” in the sequence). These temporal contexts result from item associations that are dependent on time varying neural activity (e.g., Eichenbaum 2014), and contribute to sequence memory through the reactivation of neighboring items during retrieval (DuBrow and Davachi 2013; Long and Kahana 2019).The medial prefrontal cortex (mPFC) and hippocampus (HC) are thought to contribute to sequence memory through ordinal representations and temporal contexts, respectively (Agster et al. 2002; Fortin et al. 2002; Kesner et al. 2002; DeVito and Eichenbaum 2011; Allen et al. 2016; Jenkins and Ranganath 2016). In rodents, mPFC disruptions impair sequence memory (DeVito and Eichenbaum 2011; Jayachandran et al. 2019), mPFC “time cells” are evident (Tiganj et al. 2017), and positions within a sequence can be the main determinant of differential activity in mPFC neurons during spatial sequences (Euston and McNaughton 2006). In humans, mPFC activation is sensitive to temporal order memory (Preston and Eichenbaum 2013), and codes for information about temporal positions within image sequences regardless of the image itself (Hsieh and Ranganath 2015). HC activations are also generally associated with temporal order memory (Kumaran and Maguire 2006; Ekstrom and Bookheimer 2007; Lehn et al. 2009; Ross et al. 2009; Jenkins and Ranganath 2010; Tubridy and Davachi 2011; Kalm et al. 2013; Hsieh et al. 2014; Goyal et al. 2018). Prior evidence further shows that the medial temporal lobe, specifically the HC formation, plays a critical role in the use of a TCM retrieval mode in the brain (Manns et al. 2007; Hsieh et al. 2014; Bladon et al. 2019). The HC binds events within temporal contexts (Eichenbaum et al. 2007; DuBrow and Davachi 2013; Bladon et al. 2019) through a gradually changing neural context (Manns et al. 2007; Mankin et al. 2012). Similarly, medial temporal lobe neuronal and BOLD activations in humans have demonstrated evidence for gradually evolving temporal contexts (Howard et al. 2012; Kalm et al. 2013; Kragel et al. 2015).Here we tested the contributions of the mPFC and HC during a visual sequence memory task that provides behavioral evidence of both ordinal and temporal context retrieval modes (see Fig. 1A; task modified from Allen et al. 2014). Briefly, participants first memorized six visual sequences (six images each) in a single passive viewing phase, and then were instructed to make judgments as to whether individual items were subsequently presented in sequence (InSeq) or out of sequence (OutSeq) over 240 self-paced presentations of each of the six items from each sequence. In the task, the two retrieval modes are parsed using probe trials that place conflicting demands on ordinal (Orlov et al. 2000; Allen et al. 2014, 2015) and temporal context modes (Jayachandran et al. 2019). We first evaluated ordinal retrieval modes using items that were transferred from one sequence to another while retaining their ordinal position (Ordinal Transfers) (Fig. 1B). Evidence for an ordinal-based retrieval mode occurs when these probes are identified as in sequence, because they occur in the same ordinal position as their original sequence. mPFC activations (but not HC) was strongest for these ordinal retrievals. Second, we evaluated a temporal context retrieval mode using items that skipped ahead (Skips) (Fig. 1B) with shorter lag distances (ABCFEF) compared with larger lag distances (AFCDEF). Skips should be most difficult to detect on the shortest lag distances because proximal items in a sequence are more likely to be retrieved (Howard and Kahana 2002; Kragel et al. 2015) and thus judged as InSeq. HC activations (but not mPFC) tracked with lag distance, providing evidence the HC is more reflective of a temporal context-based retrieval mode. Importantly, a significant interaction was observed such that mPFC and HC differentially activated for ordinal and temporal context retrievals. Altogether, our data show that sequence memory involves both retrieval modes. In line with these results, we suggest that understanding episodic memory requires more insight into the neurobiology of ordinal processing, in addition to the more often studied temporal contexts, in the mPFC and HC system.Open in a separate windowFigure 1.Sequence memory task and overall performance levels. Participants were tested on a sequence memory task that differentially burdens different retrieval modes using different out of sequence probe trial types. (A) An example sequence set that included six sequences. Two sequences were low memory demand sequences and four were high memory demand sequences. (B) There were three out of sequence probe trial types: items that were repeated in the sequence (Repeats), items that were presented too early in the sequence (Skips), and items that transferred from one sequence to another, while remaining in their ordinal position (Ordinal Transfers). Repeats and Skips occurred throughout the whole task, whereas Ordinal Transfers occurred during the second half only. (C) Accuracy throughout the task (error bars = ±1SD). Participants performed best on Repeats, then Skips, and poorest on Ordinal Transfers. (D,E) Distributions of response times for all InSeq trials (D, gray bars) and for all OutSeq trials (E, gray bars) for all participants with a fitted two-term Gaussian curve (black line). (F) A bimodal Gaussian curve fit better than a unimodal curve for InSeq and OutSeq trials. A trimodal curve did not improve the fit and increased the root mean squared error (not shown), suggesting distinct decisions decision-making between two decisions. It was rare to observe responses outside of the two distributions.     

mTOR inhibition impairs extinction memory reconsolidation

Fear-motivated avoidance extinction memory is prone to hippocampal brain-derived neurotrophic factor (BDNF)-dependent reconsolidation upon recall. Here, we show that extinction memory recall activates mammalian target of rapamycin (mTOR) in dorsal CA1, and that post-recall inhibition of this kinase hinders avoidance extinction memory persistence and recovers the learned aversive response. Importantly, coadministration of recombinant BDNF impedes the behavioral effect of hippocampal mTOR inhibition. Our results demonstrate that mTOR signaling is necessary for fear-motivated avoidance extinction memory reconsolidation and suggests that BDNF acts downstream mTOR in a protein synthesis-independent manner to maintain the reactivated extinction memory trace.

Repeated or prolonged nonreinforced recall may induce extinction of consolidated memories, a form of learning involving the formation of a new association that inhibits the expression of the original one (Bouton 2004). On the contrary, brief re-exposure to retrieval cues may destabilize consolidated memories, which must then be reconsolidated to persist (Przybyslawski and Sara 1997; Nader et al. 2000). Psychotherapy based on extinction enhancement or reconsolidation disruption might reduce the intrusive recollection of aversive events and help in the treatment of post-traumatic stress disorder (PTSD), a prevalent mental health condition characterized by the persistent avoidance of places, people, and objects resembling traumatic experiences (Ressler et al. 2004; Schwabe et al. 2014; Dunbar and Taylor 2017; Bryant 2019). Therefore, considerable effort has been lately dedicated to analyze the properties and potential interactions of fear memory extinction and reconsolidation. In this regard, it has been reported that these processes are mutually exclusive (Merlo et al. 2014), and that extinction training during the reconsolidation time window enhances extinction learning and prevents the recovery of fear (Monfils et al. 2009). Moreover, we have previously shown that recall renders fear-motivated avoidance extinction memory susceptible to amnesia, indicating that this memory type is prone to reconsolidation when active and suggesting that targeting extinction memory reconsolidation can be a feasible treatment strategy for PTSD (Rossato et al. 2010; Rosas-Vidal et al. 2015). However, the neurochemical basis of extinction memory reconsolidation has seldom been analyzed.Mammalian target of rapamycin (mTOR) is a 289-kDa phospho-inositide 3-kinase (PI3K)-related serine-threonine protein kinase that functions as a key element of mTOR complex 1 (mTORC1) and mTOR complex 2 (mTORC2) signaling modules to regulate protein synthesis through the phosphorylation of eukaryotic initiation factor 4E (eIF4E)-binding protein 1 (4E-BP1) and p70 ribosomal S6 kinase (p70S6K) (Hay and Sonenberg 2004). A well-known mediator of cell growth and proliferation (Hall 2008; Ryskalin et al. 2017), mTOR involvement in synaptic plasticity was first suggested by studies showing that rapamycin (RAPA), a macrolide that selectively inhibits mTORC1 signaling by interacting with the chaperone FKBP12 and binding to mTOR FKBP12–RAPA-binding domain, impairs long-term facilitation in Aplysia as well as long-term potentiation (LTP) in the rat hippocampus (Casadio et al. 1999; Tang et al. 2002). Interestingly, avoidance memory consolidation and recall need mTOR signaling in the dorsal hippocampus (Bekinschtein et al. 2007; Pereyra et al. 2018), as it also happens with the reconsolidation and extinction of several other memory types (Myskiw et al. 2008; Gafford et al. 2011; Zubedat and Akirav 2017; Jarome et al. 2018; Lee et al. 2018; Yang et al. 2019). Here, we examined whether reconsolidation of fear-motivated avoidance extinction memory requires mTOR activity in the CA1 region of the dorsal hippocampus. To do that, we used 3-mo-old, 300- to 350-g, male Wistar rats (n = 320), housed in groups of five with free access to water and food in a holding room at 22°C–23°C on a normal light cycle (12 h light:12 h dark; lights on at 6.00 a.m.). Animals were implanted with 22-gauge guides aimed at the CA1 region of the dorsal hippocampus (Supplemental Fig. S1, stereotaxic coordinates in millimeters: anteroposterior, −4.2; laterolateral, ±3.0; dorsoventral, −3.0), as previously described (Radiske et al. 2015), and allowed to recover from surgery for 10 d before being handled by the experimenter once per day for 2 d. One day later, the animals were trained in a one-trial step-down inhibitory avoidance (SDIA) task, an aversive learning paradigm in which stepping down from a platform is paired with a mild footshock. Briefly, the SDIA training box (50 × 25 × 25 cm) was made of Plexiglas and fitted with a grid floor through which scrambled electric shocks could be delivered to the rat''s feet. Over the left end of the grid floor there was a 5-cm-high, 8-cm-wide, 25-cm-long wooden platform. For training, the animals were individually placed on the platform facing the left rear corner of the training box and, when they stepped down and placed their four paws on the grid, received a 2-sec, 0.4-mA scrambled footshock, whereupon they were immediately withdrawn from the training box. This training protocol induces a long-lasting, hippocampus-dependent, fear-motivated avoidance memory expressed as an increase in step-down latency at test (Bernabeu et al. 1995; Paratcha et al. 2000; Katche et al. 2013). However, repeated testing in the absence of the footshock causes clear-cut extinction (Cammarota et al. 2005; Rossato et al. 2006; Bonini et al. 2011). Therefore, to extinguish the learned avoidance response, we submitted SDIA trained rats to one daily unreinforced test session for five consecutive days. To that end, we put the animals back on the training box platform until they stepped down to the grid. No footshock was given, and the animals were allowed to freely explore the training apparatus for 30 sec after stepping down. During this time, the animals stepped up onto the platform and down again several times. This procedure induces an SDIA extinction memory immune to spontaneous recovery, reinstatement and renewal that lasts for at least 14 d and requires NMDA receptor activation as well as protein synthesis and gene expression in dorsal CA1 to consolidate (Cammarota et al. 2003; Rossato et al. 2010; Radiske et al. 2015). One day after the last extinction session, extinction memory was reactivated by placing the animals on the training box platform until they stepped down from it. Five minutes or 6 h later, the animals received bilateral intradorsal CA1 infusions (1 µL/side) of vehicle (VEH; 5% DMSO in saline), RAPA (0.02 µg/side) or the selective ATP-competitive inhibitor of mTOR, TORIN2 (TORIN; 0.20 µg/side). RAPA and TORIN were dissolved in DMSO and diluted to working concentration in sterile saline (<5% DMSO). The doses used were determined based on pilot experiments and previous studies showing the behavioral and biochemical effects of each compound (Bekinschtein et al. 2007; Revest et al. 2014; Renard et al. 2016; Lee et al. 2018). Retention was evaluated at different times after extinction memory reactivation by placing the animals on the training box platform and measuring their latency to step down. Because of the 300-sec ceiling imposed on test latency, step-down data were expressed as median ± IQR and analyzed using the Kruskal–Wallis test followed by Dunn''s post hoc comparisons. We found that animals that received VEH recalled SDIA extinction memory normally regardless of the time elapsed between reactivation and test sessions. Conversely, RAPA and TORIN given 5 min, but not 6 h, after SDIA extinction memory reactivation impaired retention of extinction and induced reappearance of the SDIA response 1 d and 7 d later (Fig. 1A, 1 d after RA: H = 24.42, P < 0.001; P < 0.001 for VEH vs. RAPA, P < 0.001 for VEH vs. TORIN; 7 d after RA: H = 26.85, P < 0.001; P < 0.001 for VEH vs. RAPA, P < 0.001 for VEH vs. TORIN in Dunn''s multiple comparisons after Kruskal–Wallis test; Fig. 1B, 1 d after RA: H = 4.510, P = 0.1049; 7 d after RA: H = 4.606, P = 0.0999 in Kruskal–Wallis test). Neither RAPA nor TORIN affected SDIA extinction memory when administered 24 h after the last extinction session in the absence of extinction memory reactivation (Fig. 1C, 1 d after infusion: H = 2.141, P = 0.3428; 7 d after infusion: H = 4.086, P = 0.1296 in Kruskal–Wallis test) or when given 5 min post-reactivation but retention was evaluated 3 h thereafter (Fig. 1D, H = 1.654, P = 0.4375 in Kruskal–Wallis test). Moreover, RAPA and TORIN had no effect on extinction memory retention if injected in dorsal CA1 5 min after an extinction pseudoreactivation session carried out in an avoidance training box rendered nonaversive for SDIA-trained animals (Fig. 1E, After RA: H = 13.86, P = 0.001; P < 0.01 for VEH vs. RAPA, P < 0.01 for VEH vs. TORIN; After _PseudoRA: H = 0.7503, P = 0.6872 in Dunn''s multiple comparisons after Kruskal–Wallis test; Supplemental Fig. S2). mTOR activity is regulated by phosphorylation at different sites (Watanabe et al. 2011). Phosphorylation at Ser2448 is mediated by p70S6K, occurs mainly to mTOR associated with mTORC1 (Chiang and Abraham 2005; Holz and Blenis 2005; Akcakanat et al. 2007), enables mTOR binding to regulatory-associated protein of mTOR (RAPTOR), and correlates with mTORC1 activation (Rosner et al. 2010). On the contrary, Ser2481 is an autophosphorylation site insensitive to acute rapamycin treatment that is phosphorylated only when mTOR makes part of mTORC2 complexes (Peterson et al. 2000; Copp et al. 2009). To analyze mTOR phosphorylation levels, we performed immunoblotting on total homogenates from the CA1 region of the dorsal hippocampus. Samples were not pooled. Equal amounts of proteins (15 µg) were fractionated by SDS-PAGE and transferred to PVDF membranes. Blots were blocked for 1 h, incubated overnight at 4°C with anti-pSer2448 mTOR (1:10,000; RRID:AB_330970), anti-pSer2481 mTOR (1:10,000; RRID:AB_2262884), or anti-mTOR (1:10,000; RRID:AB_330978), and then incubated for 2 h at room temperature with HRP-coupled anti-IgG secondary antibody. Immunoreactivity was detected using the Amersham ECL Prime Western Blotting Detection Reagent and the Amersham Imager 600 system. Densitometric analyses were performed using the ImageQuant TL 8.1 analysis software (GE Healthcare). We found that pSer2448 mTOR levels peaked 5 min after SDIA extinction memory reactivation and returned to control values within 30 min (Fig. 2, F_(5,20) = 2.805, P = 0.0446; P < 0.05 for 5 min vs. No RA in Dunnett''s multiple comparison test after repeated measures ANOVA). No changes in pSer2481 mTOR or total mTOR levels were found up to 6 h post-reactivation (Fig. 2, pSer2481 mTOR: F_(5,20) = 1.241, P = 0.3274; mTOR: F_(5,20) = 1.208, P = 0.3411 in repeated measures ANOVA; Supplemental Fig. S3). mTORC1 activation stimulates brain-derived neurotrophic factor (BDNF) production in hippocampal neurons (Jeon et al. 2015), which in turn may induce mTOR-dependent activation of dendritic mRNA translation (Takei et al. 2004). Previously, we reported that hippocampal BDNF maintains fear-motivated avoidance extinction memory after recall (Radiske et al. 2015). In agreement with this finding, coinfusion of recombinant BDNF (0.25 µg/side) after SDIA extinction memory reactivation impeded the recovery of the avoidance response provoked by RAPA (Fig. 3, 1 d after RA: H = 27.52, P < 0.001; P < 0.001 for VEH vs. RAPA, P < 0.001 for BDNF vs. RAPA, P < 0.05 for RAPA vs. RAPA + BDNF; 7 d after RA: H = 26.76, P < 0.001; P < 0.001 for VEH vs. RAPA, P < 0.001 for BDNF vs. RAPA, P < 0.01 for RAPA vs. RAPA + BDNF in Dunn''s multiple comparisons after Kruskal–Wallis test).Open in a separate windowFigure 1.mTOR is required for fear-motivated avoidance extinction memory reconsolidation. (A) Animals were trained in SDIA (TR; 0.4 mA/2 sec) and beginning 24 h later submitted to one daily extinction session for five consecutive days (EXT). Twenty-four hours after the last session, extinction memory was reactivated (RA) and, 5 min thereafter, the animals received bilateral intradorsal CA1 infusions of vehicle (VEH; 5% DMSO in saline), rapamycin (RAPA; 0.02 µg/side) or TORIN (0.20 µg/side). Retention was assessed 1 and 7 d later (Test). (B) Animals were treated as in A except that they received intra-CA1 infusions of VEH, RAPA, or TORIN 6 h after RA. (C) Animals were treated as in A, except that they received VEH, RAPA, or TORIN in dorsal CA1 24 h after the last extinction session in the absence of RA (No RA). (D) Animals were treated as described in A, except that VEH, RAPA, or TORIN were given 5 min after RA and retention was assessed 3 h later. (E) Animals were treated as in A, except that a subgroup of animals received VEH, RAPA, or TORIN 5 min after an extinction pseudoreactivation session in an avoidance training box rendered nonaversive for SDIA-trained animals. The nonaversive box was similar in dimensions to the SDIA training box, but it was made of dark gray wood and had a Plexiglas platform. (_PRA) Pseudoreactivation session. Data are expressed as median ± IQR. (**) P < 0.01, (***) P < 0.001 versus VEH in Dunn''s multiple comparisons after Kruskal–Wallis test.Open in a separate windowFigure 2.Reactivation of fear-motivated avoidance extinction memory increases mTOR phosphorylation at Ser2448, but not at Ser2481, in the CA1 region of the dorsal hippocampus. Animals were trained in SDIA (0.4 mA/2 s) and beginning 24 h later submitted to one daily extinction session for 5 consecutive days. Twenty-four hours after the last session, extinction memory was reactivated (RA) and the animals killed by decapitation at different post-reactivation times (5–360 min). The CA1 region of the dorsal hippocampus was dissected out, homogenized, and used to determine of pS2448 mTOR, pS2481 mTOR, or mTOR levels by immunoblotting. (N) Naïve animals, (No RA) animals trained in SDIA that were submitted to five daily extinction sessions and killed 24 h after the last extinction session. Data are expressed as mean ± SEM. (*) P < 0.05 versus No RA in Dunnett''s multiple comparison test after repeated measures ANOVA.Open in a separate windowFigure 3.Coinfusion of recombinant BDNF reverses the effect of RAPA on fear-motivated avoidance extinction memory reconsolidation. Animals were trained in SDIA (TR; 0.4 mA/2 sec) and beginning 24 h later were submitted to one daily extinction session for five consecutive days (EXT). Twenty-four hours after the last session, extinction memory was reactivated (RA) and 5 min later the animals received bilateral intradorsal CA1 infusions of vehicle (VEH; 5% DMSO in saline), rapamycin (RAPA; 0.02 µg/side), BDNF (0.25 µg/µL), or RAPA plus BDNF (RAPA + BDNF). Retention was assessed 1 and 7 d later (Test). Data expressed as median ± IQR. (***) P < 0.001 versus VEH in Dunn''s multiple comparisons after Kruskal–Wallis test.Our results show that dorsal CA1 mTOR inhibition during a short post-recall time window persistently impairs retention of SDIA extinction memory and causes avoidance reappearance. This effect took time to develop, was time-dependent, concomitant with SDIA extinction memory reactivation, and occurred after the administration of mTOR inhibitors with different mechanisms of action, suggesting that it was not spontaneous or caused by nonspecific pharmacological interactions but due to bona fide impairment of an active mTOR-dependent reconsolidation process. This conclusion is further supported by findings showing that SDIA extinction memory reactivation rapidly and transiently increased mTOR phosphorylation at Ser2448, a post-translational modification customarily used as a proxy for mTOR activation (Reynolds et al. 2002; Guertin and Sabatini 2007; Rivas et al. 2009; Guo et al. 2017; Dong et al. 2018; Rosa et al. 2019). Most findings indicate that BDNF modulates protein synthesis through mTOR (Takei et al. 2001, 2004). In fact, BDNF controls hippocampal synaptic mRNA translation by regulating mTORC activation state (Briz et al. 2013; Leal et al. 2014), which seems to be necessary for SDIA memory consolidation (Slipczuk et al. 2009). However, in agreement with previous findings that BDNF is sufficient to restabilize a reactivated extinction memory trace, even when hippocampal protein synthesis and gene expression are inhibited (Radiske et al. 2015), our results show that mTOR acts upstream BDNF during the reconsolidation of extinction, and suggest not only that BDNF is a key protein synthesis product for this process but also that its actions are not mediated by mTOR-dependent mRNA translation. Indeed, mTOR signaling controls BDNF activity-dependent dendritic translation (Baj et al. 2016), and several protein synthesis-dependent plastic mechanisms, including late-LTP and memory consolidation, are rescued by BDNF when protein synthesis is impaired (Pang and Lu 2004; Moguel-González et al. 2008; Martínez-Moreno et al. 2011; Ozawa et al. 2014). Exogenous BDNF becomes quickly available for activity-dependent secretion, rapidly replacing the endogenous biosynthetic pathway after its administration (Santi et al. 2006). Thus, the rapid modulation of hippocampal high-frequency transmission produced by this neurotrophin is unaffected by protein synthesis inhibitors (Gottschalk et al. 1999; Tartaglia et al. 2001) and BDNF administration may induce the lasting structural reorganization and potentiation of hippocampal synapses in an mRNA synthesis and protein translation-independent manner (Martínez-Moreno et al. 2020), perhaps through a mechanism involving PKMζ activity regulation (Mei et al. 2011). In fact, hippocampal PKMζ acts downstream BDNF to control AMPAR synaptic insertion through a protein synthesis-independent mechanism during declarative memory reconsolidation (Rossato et al. 2019).In conclusion, our results confirm that extinction does not erase the SDIA response but generates an inhibitory memory that coexists with it and controls its expression. The data also corroborate that avoidance extinction memory enters a labile state when reactivated by recall and needs to be reconsolidated through a mechanism involving hippocampal mTOR/BDNF signaling activation to maintain its dominance over the aversive trace. Finally, though not less important, our findings emphasize the necessity of understanding the dynamics of memory competition in order to develop better therapeutic strategies for PTSD treatment.     

‘Sleep-dependent’ memory consolidation? Brief periods of post-training rest and sleep provide an equivalent benefit for both declarative and procedural memory

Sleep following learning facilitates the consolidation of memories. This effect has often been attributed to sleep-specific factors, such as the presence of sleep spindles or slow waves in the electroencephalogram (EEG). However, recent studies suggest that simply resting quietly while awake could confer a similar memory benefit. In the current study, we examined the effects of sleep, quiet rest, and active wakefulness on the consolidation of declarative and procedural memory. We hypothesized that sleep and eyes-closed quiet rest would both benefit memory compared with a period of active wakefulness. After completing a declarative and a procedural memory task, participants began a 30-min retention period with PSG (polysomnographic) monitoring, in which they either slept (n = 24), quietly rested with their eyes closed (n = 22), or completed a distractor task (n = 29). Following the retention period, participants were again tested on their memory for the two learning tasks. As hypothesized, sleep and quiet rest both led to better performance on the declarative and procedural memory tasks than did the distractor task. Moreover, the performance advantages conferred by rest were indistinguishable from those of sleep. These data suggest that neurobiology specific to sleep might not be necessary to induce the consolidation of memory, at least across very short retention intervals. Instead, offline memory consolidation may function opportunistically, occurring during either sleep or stimulus-free rest, provided a favorable neurobiological milieu and sufficient reduction of new encoding.

Of the myriad new experiences we encode each day, only a fraction are remembered over the long-term. The formation of long-term memory is crucial for optimal functioning in our everyday lives and for building knowledge across days, weeks, and years. Such enduring memories require not only the effective encoding of new information, but also a set of postencoding processes, termed “consolidation,” that function to stabilize and transform new memory traces over time (McGaugh 2000; Frankland and Bontempi 2005; Genzel and Wixted 2017).Consolidation of memory is better supported by some states of consciousness than others. For example, sleep has long been known to optimize memory consolidation, purportedly due to specific neurobiology that actively promotes the consolidation process (Diekelmann and Born 2010). Numerous studies have demonstrated that sleep facilitates the consolidation of both declarative and procedural memories. Slow oscillations (Huber et al. 2004; Marshall et al. 2006) and slow wave sleep (SWS) (Alger et al. 2012; Diekelmann et al. 2012) are thought to especially benefit hippocampus-dependent, declarative memory. Meanwhile, various forms of implicit and procedural memory have been linked to rapid eye movement (REM) sleep (Plihal and Born 1997; Mednick et al. 2009) or non-REM stage 2 (N2) sleep (Walker et al. 2002; Tucker and Fishbein 2009).A potential mechanism of offline memory consolidation during sleep is memory “reactivation,” in which patterns of neural activity in the hippocampus and cortex associated with awake experience are reiterated after learning. For example, when rats sleep after being trained on a spatial learning task, hippocampal “place cells” fire again in the same order as when the animals were being trained on the task during wake (Lee and Wilson 2002; Ji and Wilson 2007). Such neural reactivation not only occurs in the hippocampus, but also concurrently in a variety of cortical areas (Ji and Wilson 2007; Peyrache et al. 2009; Kaefer et al. 2020). The recent advent of optogenetics has allowed experimental investigation of memory reactivation in animal models. Experimentally disrupting the hippocampal ripple oscillations during which reactivation occurs impairs memory (Girardeau et al. 2009; Ego-Stengel and Wilson 2010). Conversely, selectively reactivating neural ensembles related to a particular memory appears to induce consolidation, particularly when this manipulation is applied during sleep or light amnesia (de Sousa et al. 2019).But is sleep the only brain state that facilitates memory consolidation in this way? It has been argued that sleep-specific neurobiology, including sleep slow waves (Alger et al. 2012), sleep spindles (Wamsley et al. 2012; Mednick et al. 2013; Laventure et al. 2016), and/or REM sleep (Karni et al. 1994; Stickgold et al. 2000; McDevitt et al. 2015; Boyce et al. 2016), is required for offline memory reactivation and consolidation to occur, or at least to occur optimally. However, a growing body of literature indicates that stimulus-free waking rest can similarly facilitate consolidation (Wamsley 2019). In two influential experiments, Dewar et al. (2012) demonstrated that compared with participants who completed a nonverbal distractor task, those who rested quietly with their eyes closed in a darkened room for 10 min after learning showed better memory for short stories encoded prior to the retention period. The rest group significantly outperformed the wake group after 15 min, 30 min, and 7 d (experiment 1), even in the absence of retrieval practice during the 7-d period (experiment 2).This effect of post-training rest on memory retention has been reported in an increasing number of papers across the last decade (Gottselig et al. 2004; Mercer 2015; Martini et al. 2018; Wamsley 2019; Martini and Sachse 2020). Brokaw et al. (2016) replicated the behavioral observations of Dewar et al. (2012) and showed that this memory benefit was associated with EEG slow oscillation activity, which is thought to facilitate hippocampal–cortical communication and concomitant memory consolidation during sleep (Marshall et al. 2006; Mölle and Born 2011). A recent study by Sattari et al. (2019) also linked improved memory performance to waking EEG slow oscillations, suggesting that the memory-enhancing effects of rest and sleep may share a common mechanism.Of course, it has been known for decades that at least some consolidation must occur during wakefulness. Local, cellular level consolidation begins to stabilize memory immediately following encoding (Bailey and Kandel 2008; Redondo and Morris 2011), enabling us to recall the events of the previous hours in the absence of intervening sleep. The novel suggestion of these more recent studies is that consolidation does not occur equivalently during all types of wakefulness (Dewar et al. 2012; Brokaw et al. 2016). Instead, stimulus-free rest periods appear to have features that are especially suited to facilitate memory.Reduced sensory processing during eyes-closed rest may be one factor accounting for the memory facilitation effect. However, even internally generated stimuli can also function to block consolidation, as demonstrated by the fact that mental tasks such as retrieval of autobiographical memory and focused meditation are also associated with a reduction in rest''s memory benefit (Craig et al. 2014; Collins and Wamsley 2020). Similarly, we reported in two previous studies that individuals with a high propensity for daydreaming show less memory benefit following a period of rest, presumably because intense internally generated mental activity inhibits consolidation (Humiston et al. 2019; Wamsley and Summer 2020).Together, these observations suggest that consolidation occurs during wakefulness when sensory processing is reduced, when low-frequency EEG oscillations are increased, and when internally generated cognition is at a minimum. Therefore, consolidation may not depend on neural mechanisms specific to sleep, instead opportunistically occurring across multiple states of consciousness whenever the correct conditions are met (Mednick et al. 2011). According to the opportunistic theory of memory consolidation, the processes of encoding and consolidation are mutually exclusive: During any brain state in which we are not currently encoding new information, existing memories consolidate as the neural milieu becomes favorable (Mednick et al. 2011; Wamsley 2019). Unoccupied quiet rest, like sleep, is a state in which the encoding of new stimuli is reduced. In addition, quiet rest and sleep also share a number of neurobiological features that are thought to actively promote memory consolidation, including overall slower EEG in comparison with active wakefulness, increased activation of default-mode network brain structures (Buckner and Vincent 2007), and decreased levels of acetylcholine in the brain (Hasselmo and McGaughy 2004). Additionally, the cellular-level offline reactivation of memory occurs not only during sleep, but also during quiet rest (Foster and Wilson 2006; Karlsson and Frank 2009; Carr et al. 2011; Staresina et al. 2013). At the same time, it must be noted that eyes-closed rest does not replicate all aspects of sleep neurobiology proposed to facilitate memory. For example, sleep spindle oscillations and sleep-specific neurohormonal changes are not present during eyes-closed rest.While sleep and rest both benefit memory in comparison with active wakefulness, it is not known whether they do so equivalently. With few studies directly comparing the size of rest''s memory benefit with that of sleep, it remains possible that sleep provides some benefit above and beyond that conferred by waking rest. The limited number of prior studies in this area have shown mixed results. As opposed to the type of truly task-free condition used in the waking rest studies reviewed above, most experiments comparing active wakefulness, rest, and sleep have used a “rest” condition in which participants are asked to complete an undemanding activity such as listening to music or books on tape. This approach sacrifices complete sensory restriction in return for allowing rest condition participants to maintain wakefulness for longer periods of time. For example, Mednick and colleagues have used quiet rest conditions in which participants listen to music or audiobooks, finding that this form of quiet rest facilitates memory equivalently to sleep for some forms of learning (Mednick et al. 2009; Sattari et al. 2019), but not for others (Mednick et al. 2002; McDevitt et al. 2014). Keeping rest participants awake via verbal instructions to alternately open and close their eyes, Simor et al. (2018) found no effect of post-training rest or sleep on performance of a serial reaction time task, relative to an active wake control. While these observations might indicate that only selected forms of memory can be consolidated during resting wakefulness, it is possible that the encoding of meaningful auditory stimuli during these rest conditions prevents optimal consolidation.Only a few prior studies have compared the memory effects of sleep with those of an equivalent duration of eyes-closed, entirely task-free rest. For example, Gottselig et al. (2004) successfully compared the effects of sleep, quiet rest, and active wakefulness on memory consolidation using a carefully controlled rest condition without any stimuli presented to the participants. Using a statistical auditory sequence learning task, they reported that both sleep and rest equivalently facilitated retention. In contrast, using a declarative memory task, Piosczyk et al. (2013) found that neither quiet rest nor sleep improved memory more than active wakefulness. A recent study from our own laboratory directly compared sleep, rest, and active wakefulness using declarative and procedural tasks commonly used in studies of sleep and memory (Tucker et al. 2020). However, high rates of attrition (due to rest participants inadvertently falling asleep and sleep participants failing to obtain sleep) prevented a robust test of our hypothesis that sleep and quiet rest would equivalently benefit memory, relative to active wakefulness.The goal of the current study was to directly compare the memory benefit of a brief nap (<30 min) with that of an equivalent duration of task-free quiet rest. We hypothesized that a brief period of sleep and quiet rest would have an equivalent effect on the consolidation of both declarative and procedural memories, significantly boosting memory retention compared with an equivalent duration of active wakefulness. Following our previous work, we expected that memory retention across sleep and quiet rest would be associated with slow oscillation EEG power in the <1-Hz range, but that participants with high trait daydreaming propensity would show less improvement across quiet rest.     

Striatal dopamine D1 receptors control motivation to respond,but not interval timing,during the timing task

Dopamine plays a critical role in behavioral tasks requiring interval timing (time perception in a seconds-to-minutes range). Although some studies demonstrate the role of dopamine receptors as a controller of the speed of the internal clock, other studies demonstrate their role as a controller of motivation. Both D1 dopamine receptors (D1DRs) and D2 dopamine receptors (D2DRs) within the dorsal striatum may play a role in interval timing because the dorsal striatum contains rich D1DRs and D2DRs. However, relative to D2DRs, the precise role of D1DRs within the dorsal striatum in interval timing is unclear. To address this issue, rats were trained on the peak-interval 20-sec procedure, and D1DR antagonist SCH23390 was infused into the bilateral dorsocentral striatum before behavioral sessions. Our results showed that the D1DR blockade drastically reduced the maximum response rate and increased the time to start responses with no effects on the time to terminate responses. These findings suggest that the D1DRs within the dorsal striatum are required for motivation to respond, but not for modulation of the internal clock speed.

Animals, including humans, can regulate their behavior according to temporal information. This suggests animals can perceive the passage of time (i.e., time perception). Without accurate time perception, we cannot speak, appreciate and play music, drive cars, and play sports. The current paper focussed on time perception in a seconds-to-minutes range; i.e., interval timing (Buhusi and Meck 2005). Interval timing is essential for an optimal foraging (Kacelnik and Bateson 1996) and associative learning (Gallistel and Gibbon 2000).In rodents, interval timing is frequently examined using the peak-interval (PI) procedure (Catania 1970; Roberts 1981). This task consists of randomly ordered two types of trials; i.e., fixed-interval (FI) and probe trials, separated by intertrial intervals (ITIs). During each FI trial, a reward is delivered for the first response made after a criterion time (e.g., 20 sec) has elapsed from the start of the trial, but not for responses made before the criterion time. Probe trials usually last for three times more than the criterion time in FI trials (e.g., 60 sec), and the reward is omitted. During an individual probe trial, rats typically start responding before the criterion time and stop it after the criterion time. The average rate of responses throughout the multiple probe trials in a session increases as the criterion time approaches, reaches a peak around the criterion time, and then decreases. The location of the peak of the response rate function (peak time), the spread of the function (peak spread), and the height of the peak (peak rate) are indices for timing accuracy, timing precision, and motivation to respond, respectively.Many studies have demonstrated that systemic injections of drugs affecting dopamine (DA) receptors impact interval timing in various ways. Some studies suggest that DA modulates the speed of the internal clock (Meck 1996). Systemic injection of DA receptor agonist decreases peak times, while that of DA receptor antagonist increases peak times. Importantly, the degree to which responding is altered is proportional to the duration being timed (Maricq et al. 1981; Maricq and Church 1983; Meck 1983, 1996; Matell et al. 2006). Therefore, it is necessary to use multiple target durations of a PI procedure (e.g., 20 and 40 sec) while examining the occurrence of any changes in clock speed mechanisms adequately. A recent study showed that transient activation or inhibition of DA neurons in the substantia nigra pars compacta (SNc) was sufficient to induce overestimation or underestimation of time in a temporal discrimination task, respectively (Soares et al. 2016). These results suggest that DA neurons in the SNc and DA receptors modulate the speed of the internal clock (“the dopamine-clock hypothesis”). Other studies, however, suggest that DA is involved in motivation during the interval timing task (Balcı 2014). Systemic injection of dopamine D1 receptors (D1DRs) antagonist SCH23390 reduced the peak rate (Drew et al. 2003), whereas systemic injection of DA agonist D-amphetamine increased the peak rate but reduced peak time and start time of responding without affecting stop time of responding (Taylor et al. 2007). The results of these studies are similar to those of studies that investigated the effect of manipulation of motivation on interval timing (Balcı 2014). Manipulations that decrease motivation (e.g., decreasing reward magnitude, prefeeding, and reward devaluation) reduced the peak rate, and sometimes delayed the start time (Roberts 1981; Ludvig et al. 2007; Galtress and Kirkpatrick 2009; Delamater et al. 2014, 2018). In contrast, opposite manipulations on motivation (e.g., increasing reward magnitude) caused earlier initiation of responding and a higher peak rate (Galtress and Kirkpatrick 2009). These findings support the idea that DA contributes to modulating motivation (Ikemoto et al. 2015). Therefore, DA might modulate motivation also during the timing task (“the dopamine-motivation hypothesis”).DA receptors within the dorsal striatum (dSTR) can be responsible for the effect of systemically injected DA agents on interval timing. Lesioning of the dSTR and damaged DA neurons in the SNc projecting to the dSTR caused the severe impairment of interval timing (Meck 2006b; Gouvêa et al. 2015; Mello et al. 2015), and optogenetic manipulation of DA neurons in the SNc affected the performance of the temporal discrimination task (Soares et al. 2016). These results strengthen the notion that the DA pathway from the SNc to the dSTR and the modulation of DA receptors in the dSTR are important for interval timing. Therefore, systemically injected DA agents might affect interval timing task through the DA receptors in the dSTR. Consistently, the affinity of DA agents for D2 dopamine receptors (D2DRs), but not D1DRs, correlated with the degree to produce the rightward shift of the temporal bisection function (Meck 1986), and systemic injection of D2DR blocker, but not D1DR blocker, affected interval timing in the PI procedure (Drew et al. 2003). These findings suggest that the change in the activity of D2DRs is implicated in the effect of DA on interval timing.Some studies, however, suggest that also the D1DRs within the dSTR might also play an important role in interval timing. A study showed that stimulation of the D1DR by systemic injection changed interval timing behavior (Cheung et al. 2007). D1DRs, as well as D2DRs, are highly expressed in the medium spiny neurons in the dSTR (Gerfen and Surmeier 2011). Thus, it is the possible that D1DRs in the dSTR also play an important role in interval timing. Notably, a recent study reported that D1DR blockade in the dSTR increased the time of stop responding (De Corte et al. 2019), and the findings cannot be explained by either “dopamine-clock hypothesis” nor “dopamine-motivation hypothesis.” Therefore, further studies are needed to clarify the role of D1DRs within the dSTR in interval timing.The present study aimed to examine further the role of D1DRs within the dorsal striatum in interval timing using the PI procedure. After a limited amount of training (Cheng et al. 2007), rats were infused with D1DR antagonist SCH23390 into the bilateral dSTR. Our findings support the “dopamine-motivation hypothesis” of D1DRs in the dSTR.     

Sleep strengthens integration of spatial memory systems

Spatial memory comprises different representational systems that are sensitive to different environmental cues, like proximal landmarks or local boundaries. Here we examined how sleep affects the formation of a spatial representation integrating landmark-referenced and boundary-referenced representations. To this end, participants (n = 42) were familiarized with an environment featuring both a proximal landmark and a local boundary. After nocturnal periods of sleep or wakefulness and another night of sleep, integration of the two representational systems was tested by testing the participant''s flexibility to switch from landmark-based to boundary-based navigation in the environment, and vice versa. Results indicate a distinctly increased flexibility in relying on either landmarks or boundaries for navigation, when familiarization to the environment was followed by sleep rather than by wakefulness. A second control study (n = 45) did not reveal effects of sleep (vs. wakefulness) on navigation in environments featuring only landmarks or only boundaries. Thus, rather than strengthening isolated representational systems per se, sleep presumably through forming an integrative representation, enhances flexible coordination of representational subsystems.

Wilson and McNaughton (1994) reported that “… information acquired during active behavior is … reexpressed in hippocampal circuits during sleep….” This observation of experience-dependent neural replay activity in the brain during slow-wave sleep (for review, see O''Neill et al. 2010) forms a keystone in our current understanding of how sleep affects memory consolidation in an active system consolidation process that involves the redistribution of hippocampal memory to extrahippocampal regions (McClelland et al. 1995; Diekelmann and Born 2010; Klinzing et al. 2019). According to theory, the emerging extrahippocampal memory representations are essentially schematic, devoid of specific context-information, and lack minute detail (Lewis and Durrant 2011; Payne 2011; Sekeres et al. 2018). Simultaneously, hippocampal replay strengthens hippocampal memory traces in the short-term following Hebbian learning, leading to improved context memory immediately after sleep compared with wakefulness (van der Helm et al. 2011; Weber et al. 2014). In the present study, we sought to test sleep''s role in establishing higher-level memory representations drawing on the example of spatial memory processing.Inspired by the strong role of the hippocampal formation in human spatial memory (Burgess 2008; Hartley et al. 2014) a number of studies examined effects of sleep specifically on spatial memory consolidation (Peigneux et al. 2004; Orban et al. 2006; Ferrara et al. 2008; Rauchs et al. 2008; Wamsley et al. 2010; Nguyen et al. 2013; Noack et al. 2017). In these studies, participants explored a virtual environment during a learning phase before retention periods of sleep and wakefulness and, later on, engaged in specific retrieval tasks that required to reach a predefined goal location in the environment as fast as possible. Results were mixed with, some studies reporting positive effects of sleep on spatial navigation performance (e.g., Peigneux et al. 2004; Wamsley et al. 2010; Nguyen et al. 2013; Noack et al. 2017), whereas in others such sleep effect depended on the length of the retention interval (e.g., Ferrara et al. 2008), or was completely absent (Orban et al. 2006; Rauchs et al. 2008). Interestingly, in the latter studies—despite absent behavioral effects—using a 72-h retention interval between learning and retrieval testing, functional magnetic resonance imaging (fMRI) suggested that sleep favors a shift from activation of hippocampal areas toward preferential activation of striatal areas at retrieval of the relevant spatial representations.Indeed, spatial navigation can rely on two distinct representational systems that involve as key structures hippocampal and striatal circuitry, respectively, and are also linked to different spatial frames of reference (Burgess 2008; Hartley et al. 2014). Doeller et al. (2008) showed in humans that striatal activation is linked to the processing of single proximal landmarks whereas hippocampal activation is related to the processing of spatial boundaries, and that acquisition of representations in both systems may follow different learning rules (Doeller and Burgess 2008). The subject''s reliance on one or the other representation system depends on the specific navigational problem (Maguire et al. 1998; Hartley et al. 2003) as well as familiarity with the environment (Hartley et al. 2003; Iaria et al. 2003; Packard and McGaugh 1996), but both systems can also be activated in parallel and interact. For example, patients with hippocampal atrophy showed impaired memory performance not only for boundary-based but also for landmark-based navigation (Guderian et al. 2015) suggesting the presence of synergistic effects between the representational systems. The activation of the representational systems is presumably coordinated by the medial prefrontal cortex (Ragozzino et al. 1999; Doeller et al. 2008; Rich and Shapiro 2009), that is, a region that is not only involved in the abstraction of schema-like spatial representations (Tse et al. 2011; van Buuren et al. 2014) but, also shows neuronal reactivation during sleep (Euston et al. 2007; Peyrache et al. 2009).In fact, there is first evidence suggesting that sleep supports the formation of abstract representations of space in particular. We found, for example, that sleep benefitted the extraction of semantic structure (regions defined by semantic category of landmarks) in a virtual navigation task (Noack et al. 2017). To date, there is no study, however, to specifically test the interaction between landmark- and boundary-referenced representations of space and their integration during sleep. Here we sought to fill this gap. Drawing on the active systems consolidation concept of sleep (Dudai et al. 2015; Klinzing et al. 2019) and on the existing literature, we followed the hypothesis that, rather than benefiting a specific spatial representation, sleep via neuronal replay primarily supports the formation of an integrative schema-like spatial representation and, thereby, improves flexibility in the use of hippocampus-based and striatum-based representations.To this end, we conducted two experiments, a Main experiment and a Control experiment, using a virtual spatial environment with one proximal landmark and a local boundary (Fig. 1) to preferentially engage striatum and hippocampus-based representational systems, respectively (Doeller et al. 2008). The Main experiment was designed to test the effect of sleep on the integration of landmark-referenced and boundary-referenced representations of space. To this end, participants were first familiarized with an environment featuring both a landmark and a boundary, thereby encoding both hippocampal as well as striatal representations of the environment. In order to test whether sleep enhances the integration of these representations, participants either slept or remained awake on the night after the Familiarization phase. They then learned new objects in impoverished environments featuring the same spatial cues (landmark and boundary) at the same locations but only one at a time. At a final Test session, the integration of the combined environmental layout including landmark and boundary (as presented during Familiarization before sleep) was investigated by the participant''s flexibility to switch from landmark-based to boundary-based navigation in the environment, and vice versa, from boundary-based to landmark-based navigation (Fig. 1). In the Control experiment, we investigated the direct effect of postlearning sleep or wakefulness on the consolidation of spatial memory representation that were either merely boundary-referenced or landmark-referenced, thereby controlling for general effects of sleep on spatial memory performance.Open in a separate windowFigure 1.Task and general procedures. (A) Example views on the three different environments. (Panel i) landmark and boundary present, as used in the Familiarization phase of the Main experiment. Alpine environment (panel ii), and Desert environment (panel iii) as used in the Control experiment. (B) Task procedure: The task featured three different trial types in both experiments. (Panel i) Acquisition trials were presented at the start of Familiarization and Learning phases in both experiments. (Panel ii) Feedback and Test trials started with the presentation of an object on a gray screen. Participants were then placed in the experimental environment containing boundary (thick encirclement), landmarks (traffic cone) or both, and dropped the object at the location where they found it during acquisition. In Feedback trials feedback was given by presenting the object at its correct location. Participants navigated to it to collect it. (C) Design of Main experiment: Environment featured both landmark and boundary cues during Familiarization. The Test session comprised Learning phase and Retrieval phase. Only one spatial cue (landmark or boundary) was present during each trial of the Learning and Retrieval phase (three objects with landmark, three objects with boundary). Object reference switched from Learning to Retrieval phase: Objects presented together with the landmark during learning were presented with boundary during retrieval and vice versa. Note that a specific spatial cue was always at the same relative position when presented during Familiarization, Learning, and Test. (D) Design of Control experiment: Participants were randomly assigned to the Boundary or the Landmark group, whereas all participants performed in Wake and Sleep condition. Each of the two visits (sleep and wake) consisted of two sessions (learning: six Acquisition trials + four blocks and six feedback trials; retrieval: three blocks and six Test trials).To preview our results: Whereas there was no effect of sleep on landmark- and boundary-referenced spatial memory per se in the Control experiment, sleep indeed facilitated the flexible use of different spatial retrieval cues possibly based on a superior integrated spatial memory representation.     

Emotional arousal impairs association memory: roles of prefrontal cortex regions

The brain processes underlying impairing effects of emotional arousal on associative memory were previously attributed to two dissociable routes using high-resolution fMRI of the MTL (Madan et al. 2017). Extrahippocampal MTL regions supporting associative encoding of neutral pairs suggested unitization; conversely, associative encoding of negative pairs involved compensatory hippocampal activity. Here, whole-brain fMRI revealed prefrontal contributions: dmPFC was more involved in hippocampal-dependent negative pair learning and vmPFC in extrahippocampal neutral pair learning. Successful encoding of emotional memory associations may require emotion regulation/conflict resolution (dmPFC), while neutral memory associations may be accomplished by anchoring new information to prior knowledge (vmPFC).

Emotional arousal is well known to enhance memory for individual items (Schümann and Sommer 2018), but can have impairing effects on associative memory, particularly when items cannot be easily unitized and interitem associations have to be formed (Madan et al. 2012; Murray and Kensinger 2013; Bisby and Burgess 2017). The neural substrates of the impairing effect of emotional arousal on associative memory have only begun to be explored (Bisby et al. 2016; Madan et al. 2017). Emotional arousal may disrupt hippocampal functions that are critical to promote binding and thereby lead to reduced associative memory for emotionally arousing items (“disruption hypothesis”) (Bisby et al. 2016). Conversely, encoding of neutral items may engage extrahippocampal medial temporal lobe (MTL) regions, areas we interpreted to promote better incidental unitization of neutral than negative items, leading to a net-decrease in associative memory for negative items (“bypassing hypothesis”) (Madan et al. 2017).Specifically, in our previous high-resolution fMRI study focussing on MTL regions (Madan et al. 2017), extrahippocampal MTL cortex was more active during encoding of neutral than negative picture pairs, showed a subsequent memory effect (SME) for neutral picture pairs, and neutral pair encoding was accompanied by more between-picture saccadic eye movements compared with negative pairs. In line with previous findings of extrahippocampal MTL areas involved in association memory formation of merged or unitized items (Giovanello et al. 2006; Quamme et al. 2007; Diana et al. 2008; Delhaye et al. 2019), we interpreted our fMRI and eye movement findings to suggest better incidental unitization of neutral picture pairs than negative pictures pairs. A behavioral follow-up study confirmed that unitization, that is, imagining the two pictures as one (“interactive imagery”), was indeed rated as higher for neutral than negative pairs, and this advantage in interactive imagery was linked to better associative memory for neutral pairs, lending further support to the bypassing hypothesis (Caplan et al. 2019).What would prevent emotional pairs from being as easily merged as neutral pairs? We observed that during negative pair encoding, each individual picture was fixated longer compared with neutral pictures. These longer fixation durations for negative pictures were related to greater activity during negative than neutral pair encoding in the dorsal amygdala (likely the centromedial group) (Hrybouski et al. 2016), an activation which was functionally coupled to the more ventral amygdala (likely the lateral nucleus) (Hrybouski et al. 2016). This ventral amygdala activation exhibited a subsequent forgetting effect specifically for negative pairs. Given that emotional items attract more attention to themselves and are more likely processed as individual items (Markovic et al. 2014; Mather et al. 2016), we conjectured that this may make pairs of emotional items harder to unitize and to benefit from extrahippocampal unitization-related processes such as interactive imagery. Interestingly, the hippocampus revealed a subsequent memory effect specifically for negative pairs in Madan et al. (2017). We concluded that when sufficiently arousing information precludes extrahippocampal unitization-based encoding, an alternative, compensatory, and effortful relational hippocampus-dependent encoding strategy may be used.Both findings, extrahippocampal MTL encoding for neutral pairs and compensatory hippocampal encoding for negative pairs, raise the question of which cortical areas could be involved in these two dissociable associative encoding processes. Conceivably, successful associative encoding of negative information may require participants to evaluate the emotional content, and regulate emotional arousal/conflict, functions primarily associated with dorso-medial PFC regions (dmPFC; Dixon et al. 2017), the anterior cingulate cortex (ACC) (Botvinick 2007), and posterior areas of the ventro-medial PFC (vmPFC) (Yang et al. 2020). To unitize two pictures through interactive imagery, retrieval of semantic memories and prior knowledge regarding the contents of the two pictures is likely helpful. Semantic memory processes have been attributed to the left inferior frontal gyrus (left IFG) (Binder and Desai 2011). The vmPFC (more anterior portions) could also be involved, owing to its role in relating new information during encoding to prior knowledge, that is, a “unitization-like” process (Gilboa and Marlatte 2017; Sommer 2017). Motivated by our previous discovery and interpretation of the two distinct encoding processes in the MTL (Madan et al. 2017), the potential contribution of these cortical areas in neutral and negative association memory was explored here by using a whole-brain scan and overcoming the limitations of our previous high-resolution fMRI sequence focused only on the MTL in Madan et al. (2017).In the current study, we therefore acquired standard-resolution whole-brain fMRI (3 Tesla Siemens Trio scanner, 3-mm thickness, TR 2.21 sec, TE 30 msec) of 22 male participants during exactly the same task as in Madan et al. (2017). Only male participants were recruited because of known sex-dependent lateralization of amygdala activity (Cahill et al. 2004; Mackiewicz et al. 2006), limiting the conclusions of the current study to males. Eye movements were assessed as a proxy for overt attention (EyeLink 1000, SR Research, 17 participants with usable eye-tracking data). In each of three encoding-retrieval cycles, 25 neutral and 25 negative picture pairs were intentionally encoded. Pictures (e.g., objects, scenes, humans, and animals) were selected from the International Affective Picture System (Lang et al. 2008) and the internet, and confirmed to have different valence and arousal through independent raters (details in Madan et al. 2017). Each encoding round was followed by a two-step memory test for each pair: In a judgement of memory (JoM) task one picture served as retrieval cue and volunteers indicated their memory (yes/no) for the other picture of the original pair. Then the same picture was centrally presented again, surrounded by five same-valence pictures (one correct target, four lures) in a five-alternative forced-choice associative recognition test. Participants chose the target picture from the array with an MR-compatible joystick.As in our previous studies, associative recognition was less accurate for negative (NN) (M = 0.47) than neutral (nn) pairs (M = 0.51; t(22) = 2.49, P = 0.02). Subjective memory confidence (JoM) for neutral pairs (M = 0.41) was not significantly different from confidence for negative pairs (M = 0.43; t(22) = 1.19, P = 0.25) (Fig. 1A; Madan et al. 2017; Caplan et al. 2019).Open in a separate windowFigure 1.Behavioral and eye tracking results. (A) Accuracy in the associative recognition task (5-AFC) for all negative (NN) and neutral (nn) pairs. Chance level in the 5-AFC associative recognition task was 1/5 = 0.20. (B) Mean number of saccades between the two pictures of a pair for remembered (Hit) and forgotten (Miss) negative (NN) and neutral (nn) pairs. (C) Mean number of saccades within pictures. Error bars are 95% confidence intervals around the mean, corrected for interindividual differences (Loftus and Mason 1994).Average fixation duration (a proxy for the depth of processing of individual pictures) was longer for negative than neutral pictures (F_(1,16) = 9.59, P = 0.007), with no main effect of memory (F_(1,16) = 0.11, P = 0.75), nor emotion–memory interaction (F_(1,16) = 1.27, P = 0.28). The number of fixations was also higher for negative than neutral pictures (F_(1,16) = 5.56, P = 0.03), again with no main effect of memory (F₍₁₎ = 1.56, P = 0.23) or interaction (F_(1,16) = 0.26, P = 0.61). The number of saccades within each picture (i.e., visual exploration within but not across items, reflecting intraitem processing) was higher for negative than neutral pairs (Fig. 1B; F_(1,16) = 33.38, P < 0.001), with no main effect of memory (F_(1,16) = 0.02, P = 0.89) nor interaction (F_(1,16) = 0.15, P = 0.71). However, the number of saccades between the two pictures of a pair, which may support associative processing, was substantially lower for negative than neutral pairs (Fig. 1C; F_(1,16) = 7.67, P = 0.01). Importantly, there were more between-picture saccades for pairs that were later remembered than forgotten, that is, a subsequent memory effect based on between-picture saccades (F_(1,16) = 8.43, P = 0.01). This effect did not further interact with emotion (F_(1,16) = 2.64, P = 0.12). Thus, association memory success was driven by interitem saccades, and these were reduced in negative trials. Participants spent more attention to individual negative than neutral pictures (fixation duration and number of within-picture saccades), but this was unrelated to association memory success.The fMRI data were preprocessed (slice timing corrected, realigned and unwarped, normalized using DARTEL and smoothed, FWHM = 8 mm) and analyzed using SPM12. First-level models were created with four regressors that modeled the onsets of the 2 (negative and neutral) × 2 (subsequent hits and misses) conditions of interest. Results of all fMRI analyses were considered significant at P < 0.05, family-wise error (FWE) corrected for multiple comparisons across the entire scan volume or within the a priori anatomical regions of interest (ROIs). ROIs for the hippocampus, amygdala and extrahippocampal MTL were reused from our previous study (Madan et al. 2017). The prefrontal ROIs, that is, dmPFC, ACC, vmPFC and left inferior frontal gyrus ROIs, were manually traced on the mean T1 image using ITK-SNAP 3.6.0 (Yushkevich et al. 2006) following schematic drawings based on meta-analyses (Binder and Desai 2011; Dixon et al. 2017; Gilboa and Marlatte 2017).The second-level analyses based on the resulting individual β images and subject as a random factor replicated a well-established network of brain areas involved in negative emotion processing (Spalek et al. 2015): greater activity during processing negative than neutral picture pairs in the amygdala, insula, right inferior frontal gyrus, mid, and anterior cingulate cortex as well as visual areas (Fig. 2A). As in our previous study, we correlated the difference in left amygdala activity with the difference in eye movements for negative minus neutral trials, showing a significant correlation with the number of within-picture saccades (r = 0.50, P = 0.018). Thus, higher left amygdala activity was associated with increased visual search within negative pictures. We conducted a psychophysiological interaction analysis (PPI) using this amygdala region as seed and contrasted functional coupling during successful versus unsuccessful negative with successful versus unsuccessful neutral pair encoding (i.e., the interaction of valence and subsequent memory success). This PPI revealed stronger coupling during successful encoding of negative compared with neutral pairs with a (nonsignificant) cluster in the dmPFC (Z = 3.01, [−12, 38, 26]). Simple effects showed that the amygdala was more strongly coupled with the dmPFC during successful than unsuccessful negative pair encoding (Z = 3.63, [−2, 16, 42]).Open in a separate windowFigure 2.Main effects of emotion—fMRI results. (A) Greater activity during negative than neutral pair processing irrespective of subsequent memory success. (B) Greater activity during neutral than negative pairs processing. t-maps thresholded at P < 0.001 uncorrected for visualization purposes. t-value color-coded.Neutral-pair processing was associated with greater activity than negative-pair processing in the bilateral extrahippocampal MTL cortex, ventral precuneus (vPC), retrosplenial cortex (RSC), middle occipital gyrus, and putamen (Fig. 2B). In addition, we observed a general SME irrespective of valence in the left hippocampus ([−28, −16, −24], Z = 3.49, P = 0.04).An interaction between pair valence and SME with greater neutral than negative SME was observed in vmPFC (Fig. 3A), together with a (nonsignificant) cluster in right MTL cortex ([26, −24, −28], Z = 3.16, P = 0.11). We conducted a PPI using this vmPFC region as seed and contrasted functional coupling during successful vs. unsuccessful neutral with successful vs. unsuccessful negative pair encoding. This PPI revealed stronger coupling during successful encoding of neutral compared with negative pairs in a cluster at the border of the extrahippocampal MTL cortex reaching into the hippocampus ([−20, −18, −26], Z = 4.61, Fig. 3B).Open in a separate windowFigure 3.SME × Emotion interactions and PPIs. (A) Activity in the vmPFC revealed a SME only for neutral but not negative pairs. (B) This region was stronger coupled during neutral than negative pair encoding with a cluster in the border of left MTL cortex/hippocampus. (C,D) Activity in the right hippocampus and dmPFC revealed a SME only for negative pairs. (E) The dmPFC was stronger coupled during negative than neutral pairs encoding with the bilateral hippocampus. t-maps thresholded at P < 0.001 uncorrected for visualization purposes. Error bars are 95% confidence intervals around the mean, corrected for interindividual differences (Loftus and Mason 1994).Conversely, an interaction between pair valence and SME showing a greater negative than neutral SME was observed in the right hippocampal region (Fig. 3C), replicating our previous finding of compensatory hippocampal encoding, and in the insula (Z = 3.7, [38, 2, 8]). Within prefrontal cortex, the dorsal medial prefrontal cortex (dmPFC, Z = 4.14) (Fig. 3D), also showed this effect. Neutral pairs showed a subsequent forgetting effect, that is, greater activity during unsuccessful encoding of neutral pairs, in these regions (Fig. 3 C,D).Similar to the PPI with the vmPFC seed, we conducted a PPI with the dmPFC cluster as seed. This PPI revealed the bilateral hippocampus to be more strongly coupled with the dmPFC during successful negative than neutral pair encoding (Z = 3.98, [−24, −10, −18], Z = 4.71, [30, −14, −29]) (Fig. 3E). The correlational analyses of activity in the dmPFC and vmPFC (valence × encoding success interactions) with the corresponding eye-tracking measures were nonsignificant, possibly due to low reliability of difference measures (Schümann et al. 2020).The current findings, first, replicated the impairing effects of emotional arousal on association memory previously observed in six experiments across four studies (Madan et al. 2012, 2017; Caplan et al. 2019). We built on these previous findings here by identifying cortical, especially prefrontal areas involved in the associative memory advantage for neutral pairs and those involved in the compensatory mechanism for learning negative pairs. In particular, vmPFC activity more strongly supported successful encoding of neutral than negative pairs and during this process, showed stronger coupling with a cluster at the border between MTL cortex and hippocampus. Conversely, the dmPFC was more engaged and more strongly coupled with the hippocampus during successful negative than neutral pair encoding.We observed more and longer fixations, as well as more within-picture saccades for individual negative pictures compared with neutral pictures, resembling previously reported eye movement findings (Bradley et al. 2011; Dietz et al. 2011). We had previously shown that increased attention (fixation duration) to individual negative pictures is linked to centromedial amygdala activity (not measurable here due to the whole-brain scan resolution), and functionally coupled with a negative pair-specific subsequent forgetting effect in the lateral amygdala (Madan et al. 2017). These findings together suggest that increased attention attracted by individual negative pictures does not support associative memory, or may even be detrimental (cf., Hockley and Cristi 1996).The dmPFC contributed more to negative than neutral association memory and was functionally coupled to the hippocampus, which complements our interpretation of possibly compensatory activity in the hippocampus during negative pair encoding (Madan et al. 2017). The amygdala on the other hand was stronger coupled with the dmPFC during successful encoding of negative pairs which might reflect the detection of aversive stimuli by the amygdala. The dmPFC not only plays a role in emotion regulation (Wager et al. 2008; Ochsner et al. 2012; Kohn et al. 2014; Dixon et al. 2017): It is the central node in the cognitive control network. In particular, the dmPFC regulates conflicts between goals and distracting stimuli by boosting attention toward the relevant task (Weissman 2004; Grinband et al. 2011; Ebitz and Platt 2015; Iannaccone et al. 2015). Consistent with this role in the current task, the dmPFC was functionally more strongly coupled with the bilateral hippocampus during successful negative compared with neutral pair learning. The involvement of the dmPFC during successful negative (but unsuccessful neutral) (discussed below) pair encoding may suggest that it resolves conflicts between the prepotent attention to the individual negative pictures and the current task goals, that is, their intentional associative encoding. One way to do so might involve the dmPFC''s role to regulate the negative emotions elicited by the pictures in order to focus on the associative memory task.Neutral pairs elicited more between-picture saccades than negative pairs, as in (Madan et al. 2017). The vmPFC was more strongly involved in successful associative encoding of neutral than negative pairs and more strongly coupled with the extrahippocampal MTL cortex bordering the hippocampus during successful neutral compared with negative pair encoding. Anterior vmPFC regions and their coupling with the MTL have been implicated in retrieval of consolidated memories and in anchoring new information to prior knowledge (Nieuwenhuis and Takashima 2011; van Kesteren et al. 2013; Schlichting and Preston 2015; Gilboa and Marlatte 2017; Sommer 2017; Brod and Shing 2018; Sekeres et al. 2018). We previously observed that interactive imagery (forming one instead of two images to memorize) was higher for neutral than negative pairs (Caplan et al. 2019), perhaps reflected by the increased between-picture saccades in the current study. Assuming that the anterior vmPFC subserves retrieval of prior knowledge, its engagement during successful neutral pair encoding may have supported such incidental unitization processes here as well. Negative pictures are inherently semantically more related (Barnacle et al. 2016), which implies that they may share even more prior knowledge than neutral pictures. However, the retrieval of this prior knowledge may be inhibited by the attraction of attention to individual negative pictures, not their arbitrary pairing as in the current task. Incidental unitization can occur through rather subtle manipulations (Giovanello et al. 2006; Diana et al. 2008; Bader et al. 2010; Ford et al. 2010; Li et al. 2019) or even entirely without any instruction; for example, when the items’ combination is itself meaningful or familiar (Ahmad and Hockley 2014). We suggest that similar incidental unitization processes may have occurred here as well. Memory for unitized associations is independent of hippocampal memory processes and can be based solely on the extrahippocampal MTL (Quamme et al. 2007; Haskins et al. 2008; Staresina and Davachi 2010). Our previous high-resolution fMRI study supported such a bypassing hypothesis, that is, extrahippocampal MTL cortex involvement in the successful associative encoding of neutral but not negative pairs (Madan et al. 2017). Here, this interaction did not reach significance in the MTL cortex, but the P-value of 0.11 can be considered suggestive based on our strong a priori-hypothesis. Notably, in our previous study using a scanning resolution of 1 mm³ the cluster included only 17 voxels, which would correspond to less than one voxel here. Therefore, we assume the lower sensitivity here was due to the lower spatial resolution.Unexpectedly, we observed greater activity during unsuccessful encoding of neutral pairs in the same regions that promoted successful encoding of negative pairs, that is, the dmPFC and hippocampal region. Hockley et al. (2016) previously observed that incidental but not intentional encoding of associations (for word pairs) improved for items with stronger pre-experimental associations. Perhaps using an effortful (dmPFC/hippocampal) learning strategy for neutral pairs, that is, pairs that are already more likely incidentally linked or linkable (e.g., through interactive imagery) may not have helped encoding. The forgotten neutral pairs underlying the SFE in these regions may then have been simply the hardest-to-learn neutral pairs; that is, pairs where both encoding strategies failed. Evidently, future studies should test such speculations directly.Our interpretation of the dmPFC and vmPFC as signifying in emotion regulation and unitization in this task was based on previous studies. Because we did not manipulate unitization and/or emotional regulation, these processes remain hypothetical. However, within this framework, we addressed two hypotheses regarding interactions between hippocampal/extrahippocampal MTL regions and prefrontal cortex during association memory formation. The disruption hypothesis proposes that the hippocampus is equally responsible for encoding of negative and neutral association memory but that for negative memories, hippocampal activity is inhibited by the amygdala via the prefrontal cortex (Murray and Kensinger 2013; Bisby et al. 2016). The vmPFC has known involvement in negative emotion processing (Yang et al. 2020), and the observed activity pattern in the vmPFC could appear to disrupt hippocampal association memory processes for negative pairs. However, according to the bypassing hypothesis (Madan et al. 2017), successful encoding of negative (compared with neutral) pairs requires the hippocampus since fewer extrahippocampal contributions are available. Supporting the bypassing hypothesis, we observed that the vmPFC was negatively functionally coupled with extrahippocampal MTL cortex (bordering the hippocampus), suggesting that the vmPFC decreased extrahippocampal contributions to association memory for negative pairs. The bypassing hypothesis is also supported by our finding that the hippocampus was not less but more involved in negative compared with neutral pair encoding, that is, we observed no evidence for a prefrontally (e.g., vmPFC)-mediated disruption of hippocampal activity by emotion.In conclusion, the two critical prefrontal cortex regions linked to MTL memory processes in the current study were the dmPFC, involved in successful hippocampal-dependent negative pair learning and the vmPFC, supporting successful neutral pair learning that relied on extrahippocampal MTL involvement.     

Temporal proximity to the elicitation of curiosity is key for enhancing memory for incidental information

Curiosity states benefit memory for target information, but also incidental information presented during curiosity states. However, it is not known whether incidental curiosity-enhanced memory depends on when incidental information during curiosity states is encountered. Here, participants incidentally encoded unrelated face images at different time points while they anticipated answers to trivia questions. Across two experiments, we found memory enhancements for unrelated faces presented during high-curiosity compared with low-curiosity states, but only when presented shortly after a trivia question. This suggests processes associated with the elicitation of curiosity—but not sustained anticipation or the satisfaction of curiosity—enhance memory for incidental information.

Curiosity has widely been assumed to benefit memory, with theories postulating that curiosity is a motivational state that stimulates information seeking to reduce uncertainty (Berlyne 1966; Loewenstein 1994; Litman et al. 2005; Kidd and Hayden 2015; Gottlieb and Oudeyer 2018; Gruber and Ranganath 2019). Research that addresses the relationship between curiosity and learning has typically used a trivia paradigm in which participants are tested on memory for answers to trivia questions that elicit different levels of curiosity (e.g., Kang et al. 2009). These studies demonstrate that memory of trivia answers is higher for questions that elicited high levels of curiosity (referred to here as curiosity-enhanced memory) (e.g., Kang et al. 2009; Gruber et al. 2014; McGillivray et al. 2015; Marvin and Shohamy 2016; Fastrich et al. 2018; Wade and Kidd 2019).Evidence also demonstrates that curiosity significantly enhances memory for incidental information. For example, Gruber et al. (2014) presented an incidental face image in the middle of the anticipation period between eliciting curiosity (via the presentation of a trivia question) and satisfying curiosity (via the presentation of an answer to a trivia question). Memory for the incidental face image was higher when participants anticipated answers with high compared with low curiosity. Therefore, states of high curiosity not only improve learning for topics that pique an individual''s curiosity, but a high-curiosity state can also improve memory of information beyond the target of a person''s curiosity (for further replications, see Gruber et al. 2014; Galli et al. 2018; Stare et al. 2018; Fandakova and Gruber 2020). However, it is not clear how memories are enhanced for incidental information during curiosity states.Neuroimaging research has shown that curiosity states increase activity within the dopaminergic circuit (Kang et al. 2009; Gruber et al. 2014; Duan et al. 2020) and thereby benefit hippocampus-dependent memories for curiosity target and incidental information (see the Prediction, Appraisal, Curiosity, and Exploration [PACE] framework for a theoretical framework) (Gruber and Ranganath 2019). Importantly, in one fMRI study (Gruber et al. 2014), we showed that the neural dynamics predicting curiosity-related memory enhancements for incidental images were evident when curiosity was elicited (i.e., during the presentation of a trivia question associated with high curiosity). As activation in the dopaminergic circuit increases phasic dopamine release in the hippocampal memory system (Lisman and Grace 2005; Kang et al. 2009; Rossato et al. 2009; Düzel et al. 2010; Shohamy and Adcock 2010), we theorize that curiosity-enhanced memory for incidental information will be higher when the incidental information is presented in close proximity to when curiosity is elicited. Alternatively, arousal-biased competition theories stipulate that arousal, potentially elicited via trivia stimuli, suppresses competing nontarget mental representations (e.g., such as incidental faces) in favor of goal-relevant stimuli (Mather and Schoeke 2011; Mather et al. 2016). Arousal-based theories would therefore suggest there would be a decrease in memory for incidental items in close proximity to the elicitation of curiosity where biasing of target information would be greater.In order to further disentangle whether early rather than late processes during curiosity states affect memory for incidental information, we performed two behavioral experiments building on previous work using the trivia paradigm. In experiment 1, we used a between-subjects design in which the incidental face image was shown either early or late during the anticipation period (i.e., either subsequently after the presentation of the trivia question or immediately preceding the trivia answer). In experiment 2, we used a within-subjects design and further interrogated the findings of experiment 1 by spanning the presentation of unrelated face images across the whole anticipation period (i.e., at one of four possible time points). This allowed us to investigate whether a linear relationship existed between the magnitude of curiosity-enhanced memory of incidental information and the time point at which it was presented.Across two experiments, participants underwent a three-stage paradigm with (1) a screening phase, (2) a study phase, and (3) a surprise recognition test phase for incidental face images. During the screening phase, we obtained an equal number of low- and high-curiosity questions for which participants did not know the answers (for details, see Fig. 1A; Supplemental Material, section 1.2). In the subsequent study phase (Fig. 1B–D), the selected trivia questions were randomly presented followed by an anticipation period that preceded the presentation of the associated trivia answer. During the anticipation phase, a crosshair was presented that was replaced by an image of an emotionally neutral unrelated face.Open in a separate windowFigure 1.Experimental design. (A) During the Screening phase, trivia questions were randomly selected from a pool of trivia questions (the trivia stimuli are available online at OSF https://osf.io/he6t9). Participants rated their prior knowledge and curiosity for trivia questions on a six-point scale. In experiment 1, the screening phase lasted until the participant selected 56 low-curiosity trials (pressed 1, 2, and 3) and 56 high-curiosity trials (pressed 4, 5, and 6) of which they do not have prior knowledge (112 trials in totals). In experiment 2, this lasted until 64 low-curiosity trials and 64 high-curiosity trials were selected (128 trials in total). (B,C) Study phase experiment 1: Participants encoded trivia questions (4 sec), followed by a 10-sec anticipation period (depicted by the dashed line). Using a between-subjects design, participants were preallocated into early or late conditions. For the early condition an emotionally neutral face (incidental item) is presented after 1 sec (B) and for the late condition an emotionally neutral face is presented after 7 sec (C). Participants rated (yes/no) whether this particular person would be knowledgeable about the trivia topic and could help them figure out the answer. After the anticipation period, the trivia answer was presented (1 sec). The end of a trial is denoted by a white fixation on a gray background (3–5 sec jittered intertrial interval [ITI]). (D) Study phase experiment 2: Participants were presented with trivia questions (4 sec), followed by a 12-sec anticipation period (depicted by the dashed line). After a pseudorandom period of time (2, 4, 6, or 8 sec) an emotionally neutral face (incidental item) was presented in the anticipation period (2 sec), and participants were to rate how pleasant they find the image on a scale from 1 (“not at all pleasant”) to 6 (“extremely pleasant”). The anticipation period then continued until a total of 12 sec had passed since the presentation of the trivia question. The color boxes are for explanatory purposes only and denote the four timing combinations of when the incidental face could be presented. For example, if a fixation period lasted 4 sec before the incidental face image was presented (2 sec) the remaining anticipation period lasted 6 sec (12 sec in total)—highlighted by the blue boxes. After the anticipation period, the trivia answer was presented (2 sec). The end of a trial is denoted by a white fixation on a gray background (2‐‐ to 4-sec ITI). For both experiments, the study phase was divided into four blocks.In experiment 1 (N = 61) (see Supplemental Material, section 1.1), the face image was shown either 1 sec after question offset (early condition) (Fig. 1B) or 7 sec after question offset (late condition) (Fig. 1C). During the presentation of the face, participants had to give a yes/ no response as to whether this particular person would be knowledgeable about the trivia topic and could help them figure out the answer (cf., Gruber et al. 2014). This encoding judgment was used to ensure that faces were likely to be encoded with a similar level of attention across both curiosity conditions. Following the encoding task, a surprise recognition memory test for the faces was administered. The recognition test occurred ∼10 min after the end of the study phase. Participants were tested with a six-way recognition judgment to dissociate between recollection- and familiarity-based recognition of incidental face images (see Supplemental Material, section 1.3).To investigate curiosity-enhanced memory for incidental face images in experiment 1, we used a two-way mixed-effects ANOVA to test if curiosity (two levels: high vs. low) was positively associated with better memory performance and whether the time point of face presentation (two levels between-subjects: early vs. late) interacted with the potential curiosity-enhanced memory for incidental faces (Fig. 2). The results indicated a significant interaction between curiosity and the time point of face presentation (F_(1,59) = 7.86, P = 0.007, partial eta squared = 0.118; Fig. 2). Neither the main effect of curiosity (F_(1,59) = 1.96, P = 0.167, partial eta squared = 0.032) nor the main effect of time point of face presentation (F_(1,59) = 0.46, P = 0.498, partial eta squared = 0.008) reached significance. Follow-up one-tailed t-tests revealed that, in the early presentation group, recollection estimates for faces were significantly higher for high-curiosity compared with low-curiosity trials (t₍₂₉₎ = 2.76, P = 0.005, Cohen''s d = 0.505, mean difference = 3.87, lower = 1.49, upper = ∞). In contrast, for late presentation, we did not find a significant difference in recollection for faces between the high- and low-curiosity condition (t₍₃₀₎ = −1.08, P = 0.854, Cohen''s d = −0.193, mean difference = −1.29, lower = −3.33, upper = ∞) (for further analyses, see Supplemental Material, sections 2.1–2.2).Open in a separate windowFigure 2.Recollection accuracy for incidental faces for experiment 1. Proportion of correctly remembered faces for the early and late face presentation groups split by high and low curiosity conditions. Results reveal a significant interaction between curiosity state and time point of face presentation on subsequent memory for incidental faces. Participants remembered more faces when in a high-curiosity compared with a low-curiosity state if the face was presented early in the anticipation period. Recollection accuracy for incidental items were as follows: high curiosity early (mean = 30.04%, SD = 11.22), low curiosity early (mean = 26.17%, SD = 12.02), high curiosity late (mean = 29.97%, SD = 17.74), and low curiosity late (mean = 31.26%, SD = 16.95), curiosity-enhanced memory (high—low) early (mean = 3.87%, SD = 7.66) and late (mean = −1.29%, SD = 6.68). Standard error is depicted by a vertical black line. (**) P < 0.01. The findings of experiment 1 suggest that high-curiosity compared with low-curiosity states show enhanced memory for incidental information if it is presented in close proximity to the elicitation of curiosity. These effects were specific to recollection, did not generalize across familiarity-based recognition, and could not be explained by performance on the encoding judgments (see Supplemental Material, sections 2.1–2.2).To investigate whether there was a difference in the way the face stimuli were encoded, we investigated whether RTs of the encoding judgment as a potential index of “alertness” during encoding differed between conditions. Therefore, we ran a two-way mixed effects ANOVA on RTs with curiosity (high vs. low) as within-subjects factor and the time point of face presentation (early vs. late) as between-subjects factor (see Supplemental Table S1 for means (SDs)). Neither the main effects of curiosity (F_(1,59) = 3.03, P = 0.087, partial eta squared = 0.049), time point of face presentation (F_(1,59) = 1.09, P = 0.300, partial eta squared = 0.018) nor their interaction (F_(1,59) = 3.76, P = 0.057, partial eta squared = 0.059) reached significance suggesting that “alertness” potentially did not differ between curiosity and timing of face presentation conditions. However, to further interrogate whether RTs, as a potential index of “alertness,” during encoding had any effects on later memory, we included curiosity-related RT differences during encoding (i.e., high-curiosity RTs—low-curiosity RTs) as a covariate in a two-way mixed effects ANCOVA with curiosity (high vs. low) and time point of face presentation (early vs. late) as factors on recollection memory. Consistent with the ANOVA findings, the interaction between curiosity and time point of face presentation remained when the difference in RTs during high vs. low curiosity encoding was controlled for (F_(1,58) = 8.23, P = 0.006, partial eta squared = 0.124). No other main effects or interactions were significant (curiosity: F_(1,58) = 2.27, P = 0.137, partial eta squared = 0.038; time point of face presentation: F_(1,58) = 0.22, P = 0.644, partial eta squared = 0.004; curiosity-related RT difference: F_(1,58) = 0.63, P = 0.432, partial eta squared = 0.011; curiosity * curiosity-related RT difference: F_(1,58) = 0.44, P = 0.509, partial eta squared = 0.008).To determine whether participants gave different ratings on the encoding judgment between curiosity conditions and the time point of face presentation, we ran a two-way mixed effects ANOVA on the proportions rated “helpful” (i.e., “yes” responses) with curiosity (high vs. low) and time point of face presentation (early vs. late) as factors. Helpfulness ratings significantly differed with curiosity states (F_(1,59) = 22.05, P < 0.001, partial eta squared = 0.272) but not with time point of face presentation (F_(1,59) = 0.01, P = 0.915, partial eta squared = 0.000). The interaction between curiosity and time point of face presentation was not significant (F_(1,59) = 0.08, P = 0.775, partial eta squared = 0.001). Potentially surprisingly, faces were rated as significantly more helpful during low-curiosity compared with high-curiosity states (see Supplemental Table S2). Due to this significant difference, the curiosity-related difference in helpfulness ratings (i.e., high–low curiosity) was added as a covariate in a two-way mixed effects ANCOVA on recollection accuracy with curiosity (high vs. low) and timing of face presentation (early vs. late) as factors. Importantly, the interaction between curiosity and time point of face presentation remained when the difference in helpfulness ratings during high vs. low curiosity encoding was controlled for (F_(1,58) = 7.86, P = 0.007, partial eta squared = 0.119). No other main or interaction effects were significant (curiosity: F_(1,58) = 0.84, P = 0.364, partial eta squared = 0.014; timing of face presentation: F_(1,58) = 0.43, P = 0.517, partial eta squared = 0.007; curiosity-related helpfulness rating difference: F_(1,58) = 0.49, P = 0.487, partial eta squared = 0.008; curiosity * curiosity-related helpfulness rating difference: F_(1,58) = 0.27, P = 0.604 partial eta squared = 0.005).In experiment 2 (N = 32) (see Supplemental Material, section 1.1), during the anticipation window an unrelated face replaced the fixation cross after 2, 4, 6, or 8 sec (Fig. 1D) spanning the entire anticipation period. During the presentation of the face, participants had to rate how pleasant they found the face, from 1 (“not at all pleasant”) to 6 (“extremely pleasant”), after which the fixation cross reappeared for the remainder of the anticipation period (either 8, 6, 4, or 2 sec depending on when the face image was presented) (see Fig. 1D). This decision-making judgment was deemed incidental as it was not semantically related to the trivia question. Finally, in experiment 2, we implemented a 1-d delayed surprise memory test, as the experiment served as a pilot experiment for a potential future neuroimaging experiment, and as such, we wished to reduce task demands on individuals on the day of scanning. Participants therefore made a four-point confidence judgment on whether they thought the face was presented during the study phase (see Supplemental Material, section 1.3). The four-point confidence judgment is consistent with previous studies that showed curiosity-enhanced memory for incidental faces in delayed memory tests (Gruber et al. 2014; Stare et al. 2018).Following up the findings of experiment 1, in experiment 2 we used a linear regression model to determine whether curiosity-enhanced memory of incidental information linearly decreased at larger intervals from the elicitation of curiosity. This model indeed revealed a significant effect of time point on curiosity-enhanced memory (F_(1,126) = 9.91, P = 0.002, R² = 0.073; Fig. 3A). Next, to determine at which time point the curiosity-enhanced memory effect was significant, we conducted four follow-up one-tailed t-tests investigating when recognition memory for faces was higher in high-curiosity compared with low-curiosity conditions (i.e., at 2, 4, 6, and 8 sec). This analysis revealed a significant difference at both 2 sec (t₍₃₁₎ = 3.06, P = 0.002, confidence intervals (5.28 and 26.35), mean difference = 15.8), and 4 sec (t₍₃₁₎ = 1.73, P = 0.045, confidence intervals (−0.98 and 11.91), mean difference = 5.5), but no significant difference at 6 and 8 sec (P = 0.380 and P = 0.471, respectively) (Fig. 3B).Open in a separate windowFigure 3.Recognition memory for incidental faces for experiment 2. (A) Curiosity-enhanced memory (high–low) at each presentation time point. (B) Percentage of correctly recognized faces at each presentation time point, split by high- and low-curiosity conditions. Participants recognized significantly more faces when in a high-curiosity trial if the face was presented early in the anticipation window (at both 2 and 4 sec). (**) P < 0.01, (*) P < 0.05. Ninty-five percentage confidence intervals are depicted by the shaded area. Recognition memory for incidental items were as follows: curiosity-enhanced memory (high–low) (2 sec: mean = 15.80%, SD = 29.21; 4 sec: mean = 5.54%, SD = 17.89; 6 sec: mean = 1.33%, SD = 16.26; 8 sec: mean = −0.62%, SD = 20.24). For each time point the recognition memory scores were as followed: 2 sec (high curiosity: mean = 39.58%, SD = 21.37; low curiosity: mean = 23.70%, SD = 29.83), 4 sec (high curiosity: mean = 38.81%, SD = 17.38; low curiosity: mean = 32.92%, SD = 18.28), 6 sec (high curiosity: mean = 38.78%, SD = 19.70; low curiosity: mean = 37.53%, SD = 17.93) and 8 sec (high curiosity: mean = 36.18%, SD = 19.66; low curiosity: mean = 36.76%, SD = 22.01). Extrapolating from this data we found that, consistent with experiment 1, incidental curiosity-enhanced memory is largest when presented in close proximity to the elicitation of curiosity (i.e., 2 sec after the presentation of a trivia question), but its magnitude linearly decreased with increasing time interval from the elicitation of curiosity.To determine whether pleasantness judgment differed across curiosity states or memory performance we ran a 2 (curiosity; high, low) × 2 (memory; hits, misses) repeated measures ANOVA on participants pleasantness ratings. This analysis revealed that pleasantness ratings did not significantly differ across memory (F_(1,31) = 1.43, P = 0.23) or curiosity effects (F_(1,31) = 1.93, P = 0.17). See Supplemental Table S2 for mean pleasantness ratings.Next, to determine whether this effect was due to participants being more “alert” during high curiosity trials we investigated whether response times (RT) to incidental face stimuli differed across time point, condition and memory performance (see Supplemental Table S1 for mean RTs). A 2 (condition; high curiosity, low curiosity) × 2 (memory; hits, misses) × 4 (time point; 2, 4, 6, and 8 sec) repeated-measures ANOVA was conducted revealing no significant difference in RT score across time point (F_(3,48) = 1.58, P = 0.21), condition (F_(1,16) = 0.34, P = 0.57), memory (F_(1,16) = 1.07, P = 0.32), nor the interactions between these.Finally, for consistency with our previous analysis and to determine whether our linear regression effect was due to (1) participants being more “alert” during high-curiosity trials (as inferred by RTs) or (2) participants’ degree of pleasantness rating for the faces (as inferred by encoding judgment), we included two additional independent variables to our previous linear regression model. A multiple regression was therefore run to predict curiosity-enhanced memory performance from time point, curiosity-related RT differences (high–low curiosity) and curiosity-related pleasantness rating (high–low curiosity). These variables significantly predicted curiosity-enhanced memory (F_(3,126) = 4.14, P = 0.008, R² = 0.069). Importantly, only time point added significantly to the model prediction (P = 0.002), whereas curiosity-related RT differences (P = 0.280) and curiosity-related differences in pleasantness rating (P = 0.209) did not. Collectively, this indicates that only the time point of face presentation significantly predicted curiosity-enhanced memory even when controlling for RT and pleasantness ratings.Taken together the current work yields several important contributions to our understanding of how curiosity affects memory for incidental information. First, this work provides further evidence that suggests that high-curiosity states relative to low-curiosity states can improve memory of information beyond the target of a person''s curiosity. Our results suggest that the constraints of this effect are determined by the temporal proximity of incidental information to the elicitation of curiosity. Second, across both experiments we found early curiosity effects on incidental memory were independent of the nature of the incidental encoding judgment (i.e., how knowledgeable or pleasant participants rated the face). In addition, reaction times for encoding judgments as a potential measure of “alertness” did not predict curiosity-enhanced memory across experiments.Recent theories on curiosity (e.g., Gruber and Ranganath 2019; Murayama 2019; Sharot and Sunstein 2020) highlight the importance of dopaminergic brain regions in supporting curiosity and curiosity-related memory enhancements. Although we cannot make strong conclusions about whether our findings are a result of increased release of dopamine, there is reason to believe dopamine may play an important role as our results align with predictions from fMRI findings regarding regions innervated by dopamine including the hippocampus, which suggest that the curiosity-related neural activity is somewhat limited in duration, and potentially fitting with a time-course of clearance of curiosity-triggered dopamine release via reuptake across seconds (Düzel et al. 2010; Shohamy and Adcock 2010; Gruber et al. 2014). As fMRI signals in the dopaminergic midbrain have been shown to positively correlate with dopamine release (Knutson and Gibbs 2007; Schott et al. 2008), this pattern of dopaminergic involvement suggests dopamine might be released during the elicitation of curiosity, which benefits hippocampus-dependent memories for incidental information presented in close succession to this release. Speculatively, this provides a plausible neuromodulatory explanation as to why faces presented at larger intervals from the elicitation of curiosity did not benefit from participants heightened state of curiosity. This finding is also consistent with research in the field of reward, that indicates dopaminergic activity scales with high perceived reward, and this predicts incidental memory (Tobler et al. 2005; Murty and Adcock 2014; Stanek et al. 2019).Perhaps surprisingly, experiment 2 showed that during high-curiosity trials, memory for incidental faces is consistent across all four time points, whereas in low-curiosity trials memory for incidental faces improves at later presentation times (see Fig. 3B; Supplemental Material, section 4.1). Our results are therefore inconsistent with arousal-biased theories (e.g., Mather et al. 2016), which predict that arousal, potentially elicited by high curiosity, would supress attention of irrelevant information (e.g., incidental faces). Instead, our results are consistent with recent literature in the field of reward that showed that early during reward anticipation, memory formation was improved by increased expected reward value (akin to high curiosity) potentially due to a phasic dopamine response, whereas late during reward anticipation, memory formation was enhanced by reward uncertainty reflected by a sustained, ramping of anticipatory dopamine release (Stanek et al. 2019). Notably, the findings by Stanek et al. (2019) might provide a plausible neuromodulatory explanation as to why memory for incidental faces was comparable in high-curiosity trials at all time points as it was supported by potentially both (1) phasic dopamine bursts in the early presentation and (2) sustained anticipatory dopamine ramping in the later presentations. In contrast, low curiosity memory for incidental information would only be facilitated by the later sustained anticipatory dopamine release driven by uncertainty about the correct answer. Our earlier neuroimaging findings showed that individual differences in activation of dopaminergic areas and the hippocampus elicited by curiosity (i.e., during trivia questions) predicted the magnitude of curiosity-enhanced memory for incidental faces (Gruber et al. 2014). Our current findings complement these earlier results suggesting that curiosity-elicited activity in these areas might only enhance memory for incidental faces in close temporal proximity. Despite our findings aligning with previous literature that highlights dopaminergic involvement our interpretations are speculative, therefore the neural investigation of curiosity-triggered modulation of incidental memory require further investigation.Another important finding was that curiosity-enhanced memory for incidental face information was also robustly seen across two different types of incidental encoding judgments and was not influenced by encoding judgment performance. In experiment 1, participants determined “whether the particular person would be knowledgeable about the trivia topic.” Using the same incidental encoding judgment as in previous trivia studies (Gruber et al. 2014; Galli et al. 2018; Stare et al. 2018; Fandakova and Gruber 2020), we replicated and extended the previous findings in showing that incidental curiosity-enhanced memory is specific to the temporal proximity to curiosity elicitation. Although the faces are incidental in the sense that they did not provide any meaningful information relating to the trivia, one could argue that the faces are task-relevant due to the nature of the decision-making judgment. Importantly, curiosity-enhanced memory for incidental faces was still evident with an encoding judgment that is not semantically associated with the trivia stimuli (i.e., rating the pleasantness of the incidental face image in experiment 2). Taken together, these findings suggest high-curiosity states can improve memory of information beyond the target of a person''s curiosity, even when that information is completely incidental to the topics that piqued an individual''s curiosity.In conclusion, our findings provide a better understanding into how curiosity enhances memory for incidental information, suggesting that the elicitation of curiosity is key. However, future studies should examine the role of curiosity on incidental memory for a broader range of stimuli and with different approaches to elicit curiosity. The impact of such work could be applicable to a wide range of areas and would need to be tested in the real world (for example, in news or education). This may be particularly pertinent given the current pandemic in which the need to disseminate rapidly updating policies and educate the public on disease and health, has never been more apparent. As such, understanding when to present critical information that should be remembered is pivotal. Our results indicate it is important that the to-be-enhanced incidental information is presented as early as possible when curiosity is sparked.     

Reliable induction of sleep spindles with intracranial electrical pulse stimulation

Neocortical sleep spindles have been shown to occur more frequently following a memory task, suggesting that a method to increase spindle activity could improve memory processing. Stimulation of the neocortex can elicit a slow oscillation (SO) and a spindle, but the feasibility of this method to boost SO and spindles over time has not been tested. In rats with implanted neocortical electrodes, stimulation during slow wave sleep significantly increased SO and spindle rates compared to control rest periods before and after the stimulation session. Coordination between hippocampal sharp-wave ripples and spindles also increased. These effects were reproducible across five consecutive days of testing, demonstrating the viability of this method to increase SO and spindles.

Neocortical sleep spindles are prominent oscillations during slow-wave sleep (SWS) (Kandel and Buzsáki 1997; Steriade and Deschenes 1984; Steriade et al. 1993) that play a role in consolidation of both declarative (Gais et al. 2002; Schabus et al. 2004; Schmidt et al. 2006) and nondeclarative memories (Nishida and Walker 2007; Peters et al. 2008; Barakat et al. 2011; Johnson et al. 2012). Spindles are preferred times of memory reactivation in the cortex during sleep, suggesting that reactivation during spindles contributes to memory performance (Peyrache et al. 2009; Ramanathan et al. 2015; Eckert et al. 2020).Based on this evidence, a method to induce spindle oscillations may prove beneficial to memory consolidation. Recent optogenetic experiments have shown that it is possible to induce cortical spindles by activating the thalamic reticular nucleus (Halassa et al. 2011; Kim et al. 2012), and that these induced spindles improve memory performance (Latchoumane et al. 2017). Given the large gap between optogenetic manipulation and clinical practice, a more practical technique for enhancing spindles would be desirable. Electrical stimulation of the neocortex with an implanted electrode can evoke a slow oscillation (SO) and a spindle, but the reliability and longevity of this effect are unknown (Steriade and Deschenes 1984; Contreras and Steriade 1995; Vyazovskiy et al. 2009). Here we show that electrical pulse stimulation reliably evokes spindles for the duration of a 1-h stimulation session, resulting in a significant increase in SO and spindle density.Four adult male Fisher–Brown Norway rats had recording electrodes implanted bilaterally in the motor cortex, hippocampus, and neck. The stimulating electrode was implanted in the deep motor cortex of one hemisphere adjacent to one of the cortical recording electrodes (see Supplemental Fig. S1 for details, as well as the Supplemental Material for greater detail of all methods). After recovering from surgery, the animals were habituated to a quiet, dimly lit room and recording box for 3 d. All procedures and recordings were performed during the animal''s light cycle. The vivarium maintained a 12-h light cycle (7 a.m. on) and the animals were tested in the same order every day such that recordings on different days occurred at approximately the same time (±1 h). On experiment days, rats were taken to the recording room and they rested in a small cage after being connected to the recording system. The recording consisted of three consecutive 1-h sessions: a baseline recording with no stimulation, a recording with repeated single pulse electrical stimulation, and a final recording session with no stimulation. The procedure was repeated on five consecutive days to test the reliability of the method.Recording was done on a Cheetah system (Neuralynx) and all signals were sampled at 2 kHz. Stimulation was delivered by a constant current stimulus isolation unit that was controlled by an Arduino microcontroller. Delivery of stimulation pulses was targeted to SWS using a real-time sleep state detector written in MATLAB. A 3-sec buffer of recording was used to calculate the ratio of δ (1–4 Hz) to θ (5–10 Hz) in the hippocampal LFP as well as the amount of EMG activity. SWS detection occurred when there was a high δ/θ ratio and low EMG activity. For each animal, the threshold value for the SWS detection was determined during a habituation session prior to the start of the stimulation experiment. Once SWS was detected, a single biphasic pulse (150 µS per phase, 500 µA) was delivered every 3 sec for as long as the animal remained in SWS (Fig. 1A). The real-time detection corresponded well with offline sleep structure analysis (Supplemental Fig S3).Open in a separate windowFigure 1.Neocortical stimulation during SWS reliably evokes spindles. (A) Segment of LFP from SWS showing SO and spindle oscillations occurring following stimulation pulses. Colored patches indicate offline detected SO and spindles. (B) Comparison of spontaneous and evoked spindles from each rat. (C) Average waveforms of spontaneous and evoked spindles showing larger amplitude of evoked spindles. Waveforms are aligned to the peak amplitude of the spindle and normalized by the amplitude in the prestim session. Shaded region is SEM of n = 4 rats. (D) Quantification of spindle amplitude across prestim and post-stim recording sessions. Amplitude is normalized to the spindle peak of the prestim session. For the distributions, prestim spindles were used to standardize the amplitude. (*) P < 0.05, t₃ = 10.45; (**) P < 0.01, t₃ = 5.16. Error bars are SEM of n = 4 rats.Spindles, SO, and SWR were detected automatically based on thresholded power signals of filtered LFP recordings (SO 1–4 Hz, spindle 10–20 Hz, SWR 100–250 Hz) (see the Supplemental Material for details). Threshold values for automated detection of LFP features was done while visually inspecting the prestimulation recording from day 1, and then verifying that they were appropriate by checking the prestimulation session from days 2–5. Once set, the same thresholds were used for all recording sessions on all days. Spindles that occurred within 750 msec of a SO (measured from the peak of the hyperpolarization), SWR, or stimulation pulse were considered “coupled” to the SO/SWR or “evoked” by the stimulation pulse. SO within 500 msec of a stimulation pulse were considered “evoked” (Latchoumane et al. 2017). For triple coupling of SO, spindles, and SWR, the SO was taken as time zero and then both a spindle and SWR had to occur within 750 msec of the SO. Although SOs, spindles and SWR appear similar in mice and rats (Mölle et al. 2009; Niethard et al. 2018; Csernai et al. 2019), differences in detection methods and parameters can give rise to different event timings. Furthermore, definitions of “coupling” can vary across studies, with gaps of 2 sec between SO and spindles accepted as coupled (Kam et al. 2019). Because of these issues, we tested several gap durations where we report significant coupling to ensure the coupling is not specific to a particular gap.As previously reported (Vyazovskiy et al. 2009), brief stimulation pulses delivered to the corpus callosum during SWS evoked SO and spindles in the motor cortex (Fig. 1A). Spindle oscillations occurred reliably following the SO throughout the 1-h stimulation session, an effect that occurred ipsilaterally as well as contralaterally to the stimulation electrode.Other than the presence of a stimulation artifact, which was removed with an automated algorithm (Supplemental Fig. S2), stimulation-evoked spindles appeared qualitatively similar to spontaneously occurring spindles (Fig. 1B). Examination of overlaid average waveforms of spindles from the prestim and stimulation sessions revealed that evoked spindles were larger in amplitude (Fig. 1C). Measuring the peak amplitude showed a significant increase in the amplitude of evoked spindles compared to spontaneous spindles in the prestim session (t₃ = 5.16, P < 0.05) (Fig. 1D). This effect did not last beyond the stimulation session, and spindle amplitude in the post-stim session was not significantly different than prestim amplitude (t₃ = 0.47, P = 1). We also compared frequency and duration of evoked and spontaneous spindles. Detected spindles were 400–600 msec in duration on average, which is in the range normally reported for spindles (Gardner et al. 2013; Latchoumane et al. 2017), and stimulation did not alter the duration (Supplemental Fig. S4). Similarly, spindle frequency was not changed by stimulation (Supplemental Fig. S4). Like evoked spindles, the amplitude of evoked SO was significantly larger compared to spontaneous SO from the prestim session (Supplemental Fig. S5). Unlike spindles, the amplitude of spontaneous SO in the post-stim session, although smaller than evoked SO, remained significantly larger than the prestim SO (Supplemental Fig. S5).Importantly, stimulation did not affect the rats’ sleep. They spent most of the time sleeping, and the amount of sleep was similar between the stimulation session and the post-stimulation session (Supplemental Fig. S6A). The amount of sleep in the prestim session was less than the other sessions, likely because the rats were still aroused after being transported to the recording room. However, the proportion of SWS was similar between all epochs, indicating that stimulation did not significantly alter the sleep structure (Supplemental Fig. S6B).SO and spindles occurred reliably following stimulation in both hemispheres. On average, 76% of pulses were followed by a SO and 59% were followed by spindles (Supplemental Fig. S7). To quantify the increase in the number of SO and spindles events over the duration of the 1-h stimulation session, we compared the SO and spindle density (number per minute; averaged across hemispheres) of the prestim and post-stim sessions. SO and spindle density both increased significantly during the stimulation session, SO by an average of 30% (range 18%–45%) (Fig. 2A1), and spindles by an average of 37% (range 23%–54%) (Fig. 2B1; individual hemispheres in Supplemental Fig. S8). Temporal coupling of SO and spindles was previously identified as important for memory (Latchoumane et al. 2017). Using a similar measure, we classified spindles as coupled to a SO if they occurred within 750 msec. Despite the significant increase in both SO and spindles, the SO-spindle coupling was not increased by stimulation (Supplemental Fig. S9). Furthermore, stimulation did not appear to cause any lasting effect on SO or spindle density as both SO and spindle density in the post-stim session were comparable to the prestim session. We also calculated the peristimulus time histogram (PSTH) of SO and spindle occurrences and found a significant increase in SO and spindle rate following stimulation pulses (Fig. 2A2,B2). To test the reproducibility and longevity of the stimulation-induced increase in SO and spindles, we repeated the experiment on five consecutive days and observed similar increases in SO and spindle density on each day (Fig. 2C,D).Open in a separate windowFigure 2.One hour of stimulation reliably increases SO and spindle rate. (A1) Average SO density across prestim and post-stim sessions (t₃ = 5.19). (A2) Peristimulus time histograms (PSTH) of SO relative to stimulation pulses or randomly distributed pulse times (t₃ = 3.86). (B1) Average spindle density across sessions (prestim: t₃ = 8.05; stim-post: t₃ = 4.88). (B2) PSTH of spindle occurrences relative to stimulation pulses (t₃ = 4.71). (C,D) Stimulation-induced increase in SO and spindle rate is consistent across five consecutive days. In all panels, (*) P < 0.05, error bars are SEM of n = 4 rats.We next examined the possible effect of cortical stimulation on hippocampal sharp-wave ripples (SWR). Unlike spindles, stimulation did not affect the overall rate of SWR occurrence (Fig. 3A). However, the timing of SWR was affected by stimulation. SWR were more likely to occur within 300–400 msec following a cortical stimulation pulse, and were less likely to occur at other latencies (ANOVA F_(19,114) = 6.7, P < 0.01) (Fig. 3B). The increased tendency of SWR within this time window led to increased coupling between SWR and spindles (t₃ = 4.72, P < 0.01) (Fig. 3C). To verify that the increased coupling was not due to the definition of coupling (750-msec gap), we tested a range of gap durations (250–2000 msec) and found the coupling to be robust (Supplemental Fig. S10). The increased SWR-spindle coupling was only present during the stimulation session, and there was a decrease in coupling in the post-stim session such that there was no difference between the prestim and post-stim coupling. Finally, we tested for possible triple coupling of SO, spindles, and SWR. Although triple coupling increased in all four rats during the stimulation session, it failed to reach statistical significance (t₃ = 2.10, P = 0.63) (Fig. 3D).Open in a separate windowFigure 3.Effects of cortical stimulation on hippocampal sharp-wave ripples (SWR). (A) The rate of SWR in the prestim and post-stim sessions were similar, indicating that stimulation did not affect the overall rate of SWR. (B) Stimulation altered the timing of SWR. The PSTH of SWR relative to stimulation pulses shows an overall suppression of SWR following the pulse, with a grouping of SWR occurring 300–400 msec after the pulse. (**) P < 0.01, ANOVA F_(19,114) = 6.7. (C) Stimulation caused an increase SWR-spindle coupling. Compared to the prestim session, there was a significant increase in the proportion of spindles that occurred within 750 msec of a SWR during the stim session. (**) P < 0.01, t₃ = 4.72. The increase in SWR-spindle coupling during the stim session did not persist into the post-stim session. (*) P < 0.05, t₃ = 2.49. (D) Triple coupling of SO, spindles, and SWR increased in all four rats during the stimulation session but failed to reach statistical significance (t₃ = 2.10, P = 0.063).In summary, our results show that spindles can be evoked reliably by single-pulse electrical stimulation of the neocortex for the duration of a 1-h stimulation session, resulting in a significant increase in SO and spindle density, and an increase in SWR-spindle coupling. One potential limitation of our study is the within animal design. Because the same parameters were used to detect spindles and SO, the observed increase in spindles and SO could be due to an endogenous increase that occurs over time. While we cannot strictly rule out this possibility, it is worth noting that the increase in spindle and SO density decreased significantly in the post-stim session, so the increase cannot be due to a progressive increase in spindles and SO. Together with the very strong cross-correlation of stimulation pulses and spindles/SO, the evidence favors the view that the increase in spindle/SO density was indeed due to the stimulation. Furthermore, this increase is reproducible across five consecutive days of stimulation, suggesting that this method is viable as a deep-brain stimulation implant for increasing SO and spindles.Previous attempts to boost spindle expression during sleep have used different methods with mixed results. Playing auditory stimuli during sleep is a minimally invasive method that has shown promise in altering SWS oscillations. In the context of a TMR experiment, when the auditory stimulus was previously paired with a learning task, tones presented during sleep increased spindles whose content was related to the memory task (Cairney et al. 2018). Even when auditory tones are used in non-TMR experiment (i.e., the tone was not previously paired with a learning task), tones played during sleep increased spindle power and improved memory performance (Ngo et al. 2013). However, a recent study showed a similar increase in spindle power following tone presentation, yet there was no corresponding memory improvement (Henin et al. 2019). Furthermore, the practical use of auditory stimuli outside of a clinical setting has been questioned because it is less reliable and subject to interference from other sounds (Ngo et al. 2015).Initial studies of transcranial electrical stimulation (TES) used mild oscillating currents during SWS and showed that slow oscillations and spindles were increased, and that memory performance was improved (Marshall et al. 2006). Subsequent studies using similar protocols yielded mixed results, and a recent study showed that currents typically used in TES studies are too weak to affect neural activity (Lafon et al. 2017). Our more invasive method of implanting a stimulating electrode in the cortex and delivering brief pulses of current was first shown to be effective at evoking spindles many years ago in anesthetized cats (Roy et al. 1984; Steriade and Deschenes 1984; Contreras and Steriade 1995, 1996). Since then, an increase in spindle power has been observed following electrical stimulation of the neocortex in sleeping rats (Vyazovskiy et al. 2009), or by transcranial magnetic stimulation in humans (Bergmann et al. 2012; Massimini et al. 2007). Despite these results, it has not been shown that repeated cortical stimulation is capable of increasing spindle rates, and our current results show that stimulation pulses through implanted electrodes can be used to evoke spindles reliably, not only during a 1-h recording session, but across multiple days.A recent optogenetic study showed that thalamic reticular activation increased the coupling of spindles and slow oscillations and improved contextual fear memory (Latchoumane et al. 2017). Interestingly, they did not report a net increase in spindle density, so the improved memory was attributed to the increased coordination between SO, spindles, and SWR. We observed a significant increase in SWR-spindle coupling, as well as a strong trend toward increased triple coupling of SO, spindles, and SWR, although this failed to reach statistical significance. The weaker triple coupling is possibly due to the fact that our stimulation pulses occurred randomly with respect to SO, whereas the reticular stimulation in Latchoumane et al. (2017) was triggered by a SO. If our electrical stimulation was triggered by SO, then it is possible the triple coupling would be increased as well, although this remains to be tested. Another critical next step is to determine if the substantial stimulation-induced increase spindle and SO density, as well as the increased SWR-spindle coupling, is sufficient to improve memory.     

Larval zebrafish display dynamic learning of aversive stimuli in a constant visual surrounding

Balancing exploration and anti-predation are fundamental to the fitness and survival of all animal species from early life stages. How these basic survival instincts drive learning remains poorly understood. Here, using a light/dark preference paradigm with well-controlled luminance history and constant visual surrounding in larval zebrafish, we analyzed intra- and intertrial dynamics for two behavioral components, dark avoidance and center avoidance. We uncover that larval zebrafish display short-term learning of dark avoidance with initial sensitization followed by habituation; they also exhibit long-term learning that is sensitive to trial interval length. We further show that such stereotyped learning patterns is stimulus-specific, as they are not observed for center avoidance. Finally, we demonstrate at individual levels that long-term learning is under homeostatic control. Together, our work has established a novel paradigm to understand learning, uncovered sequential sensitization and habituation, and demonstrated stimulus specificity, individuality, as well as dynamicity in learning.

Learning while being exposed to a stimulus (i.e., nonassociative learning) is of great importance in that it triggers intrinsic constructs for subsequent recognition of that stimulus and provides a foundation for associative learning (e.g., learning about relations between stimuli in Pavlovian conditioning and stimuli responses-outcomes in instrumental conditioning). Nonassociative learning precedes associative learning in the evolutionary sequence and involves a broad range of behavioral phenomena including habituation, sensitization, perceptual learning, priming, and recognition memory (Pereira and van der Kooy 2013; Ioannou and Anastassiou-Hadjicharalambous 2018).As the simplest learning form, habituation is defined as the progressively reduced ability of a stimulus to elicit a behavioral response over time (Glanzman 2009; Rankin et al. 2009; Thompson 2009). Such a response reduction is distinguished from sensory adaptation and motor fatigue and is often considered adaptive in that it helps animals to filter out harmless and irrelevant stimuli (Rankin et al. 2009). Since an early study of EEG arousal in cats (Sharpless and Jasper 1956), the habituation phenomenon has been widely documented in invertebrates such as C. elegans (Rankin and Broster 1992; Rose and Rankin 2001; Giles and Rankin 2009; Ardiel et al. 2016) and Aplysia (Glanzman 2009) as well as in vertebrates such as rodents (Bolivar 2009; Salomons et al. 2010; Arbuckle et al. 2015), zebrafish (Best et al. 2008; Wolman et al. 2011; Roberts et al. 2016; Randlett et al. 2019; Pantoja et al. 2020) and humans (Bornstein et al. 1988; Coppola et al. 2013).Accompanying habituation is a process termed sensitization, which in contrast enhances responses to a stimulus over time (Kalivas and Stewart 1991; McSweeney and Murphy 2009; Robinson and Becker 1986). This counterpart of habituation may also be adaptive if it helps animals avoid potentially risky or costly situations (Blumstein 2016; King et al. 2007). Like habituation, sensitization has also been documented in a phylogenetically diverse set of organisms (Cai et al. 2012; Kirshenbaum et al. 2019; Tran and Gerlai 2014; Watkins et al. 2010), suggesting an evolutionarily conserved biological underpinning for both processes. Furthermore, these simple learning forms are observed in various functional outputs of nervous systems ranging from simple reflexes (Blanch et al. 2014; Pantoja et al. 2020; Pinsker et al. 1970; Randlett et al. 2019) to complex cognitive phenotypes (Bolivar 2009; Leussis and Bolivar 2006; Thompson and Spencer 1966) and may represent deeper neurobiological constructs associated with anxiety-memory interplay (Morgan and LeDoux 1995; Ruehle et al. 2012; Sullivan and Gratton 2002). Therefore, understanding basic building blocks of habituation and sensitization is essential to fully understand complex behaviors.Habituation and sensitization have been reported with short-term (intrasession) and long-term (intersession) mechanisms in a number of systems (Rankin et al. 2009; Thompson 2009). Short-term mechanisms sensitize or habituate a response within a session (Meincke et al. 2004; Leussis and Bolivar 2006; Rahn et al. 2013; Byrne and Hawkins 2015). In contrast, long-term mechanisms retain memory formed in previous session and use it to modify behavioral responses in a subsequent session (Rankin et al. 2009).Although both short- and long-term learning and memory have been demonstrated in young larval zebrafish (Wolman et al. 2011; O''Neale et al. 2014; Roberts et al. 2016; Randlett et al. 2019), so far, most paradigms use unnatural stimuli and are designed without integrating sensitization and habituation in the same paradigm. The latter limitation is crucial as the influential dual-process theory, proposed by Groves and Thompson (1970), suggests that the two learning forms interact to yield final behavioral outcomes and therefore assessment of only one process might be confounded by alteration in the other process (Meincke et al. 2004).In this study, we examined stimulus learning in a large population of larval zebrafish using a light/dark preference paradigm over four trials across 2 d. Light/dark preference as a motivated behavior is observed across the animal kingdom (Serra et al. 1999; Bourin and Hascoët 2003; Gong et al. 2010; Lau et al. 2011). Larval zebrafish display distinct motor behaviors that are sensitive to the intensity of both preadapted and current photic stimuli (Burgess and Granato 2007; Burgess et al. 2010; Facciol et al. 2019). In our paradigm with well-controlled luminance history and constant visual surrounding, larval zebrafish generally display dark avoidance and center avoidance (also known as thigmotaxis) with heritable individual variability and are considered fear- or anxiety-related (Steenbergen et al. 2011; Schnörr et al. 2012; Bai et al. 2016; Wagle et al. 2017; Dahlén et al. 2019). From an ethological perspective, the extent of avoidance is likely a readout of the circuitry that balances anti-predation (i.e., avoid the dark and the center) and free exploration (i.e., approach the dark and the center). As described below, we have uncovered stimulus-specific temporal dynamicity of learning (both short term and long term), as well as individual differences in learning that are under homeostatic control.     

Inhibiting ventral hippocampal NMDA receptors and Arc increases energy intake in male rats

Research into the neural mechanisms that underlie higher-order cognitive control of eating behavior suggests that ventral hippocampal (vHC) neurons, which are critical for emotional memory, also inhibit energy intake. We showed previously that optogenetically inhibiting vHC glutamatergic neurons during the early postprandial period, when the memory of the meal would be undergoing consolidation, caused rats to eat their next meal sooner and to eat more during that next meal when the neurons were no longer inhibited. The present research determined whether manipulations known to interfere with synaptic plasticity and memory when given pretraining would increase energy intake when given prior to ingestion. Specifically, we tested the effects of blocking vHC glutamatergic N-methyl-D-aspartate receptors (NMDARs) and activity-regulated cytoskeleton-associated protein (Arc) on sucrose ingestion. The results showed that male rats consumed a larger sucrose meal on days when they were given vHC infusions of the NMDAR antagonist APV or Arc antisense oligodeoxynucleotides than on days when they were given control infusions. The rats did not accommodate for that increase by delaying the onset of their next sucrose meal (i.e., decreased satiety ratio) or by eating less during the next meal. These data suggest that vHC NMDARs and Arc limit meal size and inhibit meal initiation.

Research into the higher-order cognitive controls of eating behavior has demonstrated that hippocampal neurons, which are critical for learning and memory, also regulate energy intake (Benoit et al. 2010; Parent 2016; Kanoski and Grill 2017). The hippocampus is functionally divided along its longitudinal axis into dorsal (posterior in primates) and ventral (anterior in primates) poles (Moser and Moser 1998; Fanselow and Dong 2010; Strange et al. 2014). Generally, dorsal hippocampal (dHC) neurons are necessary for episodic and spatial memory, whereas ventral hippocampal (vHC) neurons are essential for affective and motivational processes and emotional memory (Fanselow and Dong 2010; Strange et al. 2014). dHC and vHC have different anatomical connections, cellular and circuit properties and patterns of gene expression that likely contribute to the different functions that they serve (Moser and Moser 1998; Thompson et al. 2008; Dong et al. 2009; Barkus et al. 2010; Fanselow and Dong 2010; Bienkowski et al. 2018).vHC neurons, in particular, are poised to integrate energy-related signals with mnemonic processes because they contain receptors for numerous food-related signals (Kanoski and Grill 2017) and project to several brain regions critical for food intake (Namura et al. 1994; Cenquizca and Swanson 2006; Radley and Sawchenko 2011; Hsu et al. 2015b). vHC lesions increase food consumption and body mass (Davidson et al. 2009, 2012, 2013), and activation of vHC receptors for gut hormones affects food intake and food-related memory (Kanoski et al. 2011, 2013; Hsu et al. 2015a, 2017, 2018). Additionally, vHC glutamatergic projections to the bed nucleus of the stria terminalis, lateral septum, and prefrontal cortex inhibit energy intake (Sweeney and Yang 2015; Hsu et al. 2017).It is possible that vHC neurons contribute to the representation of the memory of a meal and inhibit subsequent intake. In support, we have shown that vHC neurons inhibit energy intake during the postprandial period. Specifically, optogenetic inhibition of vHC principle glutamatergic neurons given after the end of a sucrose or chow meal, timed to occur when the memory of the meal would be undergoing consolidation, accelerates the onset of the next meal and increases the amount eaten during the next meal when the neurons are no longer inhibited (Hannapel et al. 2019). Inactivation of these neurons given after a saccharin meal also hastens the initiation of the next saccharin meal and increases the size of that next meal, suggesting that vHC inhibition does not increase intake by disrupting the processing of interoceptive visceral signals (Hannapel et al. 2019).If vHC neurons inhibit intake through a process that involves memory, then well-defined molecular events necessary for vHC synaptic plasticity should play a role in controlling meal timing and meal size because synaptic plasticity at hippocampal excitatory synapses is a critical mechanism underlying memory formation (Bailey et al. 2015; Bartsch and Wulff 2015). Activation of glutamatergic N-methyl-D-aspartate receptors (NMDARs) is required for most forms of hippocampal synaptic plasticity (Malenka and Nicoll 1993; Volianskis et al. 2015). NMDAR-dependent increases in intracellular calcium activate proteins and stimulate mRNA synthesis and protein translation that collectively act to increase glutamate AMPA receptor function in the postsynaptic cell, thereby increasing glutamate signaling and synaptic strength (Shanley et al. 2001; Bevilaqua et al. 2005; Herring and Nicoll 2016). Synaptic plasticity in vHC is NMDAR-dependent and vHC NMDARs are often necessary for vHC-dependent memory (Zhang et al. 2001; Xu et al. 2005; Kent et al. 2007; Czerniawski et al. 2012; Portero-Tresserra et al. 2014; Zhu et al. 2014; Clark et al. 2015; Maggio et al. 2015). Of note, feeding-related hormones such as insulin and leptin enhance NMDAR functionality in hippocampal cultured neurons and slices (Liu et al. 1995; Shanley et al. 2001).Hippocampal synaptic plasticity is also dependent on the activation of the immediate early gene (IEG) activity-regulated cytoskeleton-associated protein (Arc). Arc is considered a master regulator of synaptic plasticity (Bramham et al. 2010; Korb and Finkbeiner 2011; Shepherd and Bear 2011). It is downstream from many molecular signaling pathways and is necessary for virtually every type of synaptic plasticity (Bramham et al. 2008; Korb and Finkbeiner 2011; Shepherd and Bear 2011). Learning experiences produce small but significant increases in Arc that are typically maximal within 15 min of the experience, and unlike other IEGs, Arc expression reflects synaptic plasticity rather than neuronal activity (Fletcher et al. 2006; Guzowski et al. 2006; Carpenter-Hyland et al. 2010). vHC Arc is necessary for memory consolidation because disrupting vHC Arc expression with Arc antisense (anti-Arc) oligodeoxynucleotides (ODN) disrupts vHC-dependent memory (Czerniawski et al. 2011, 2012; Chia and Otto 2013). We have shown that sucrose consumption increases vHC Arc expression during the early postprandial period (Hannapel et al. 2017), suggesting that ingestion activates molecular processes required for synaptic plasticity in vHC.Although it is well established that vHC neurons influence energy regulation, it is unknown whether vHC neurons regulate energy intake through a process that requires NMDARs and Arc. In the present experiments, we tested the prediction that disrupting vHC NMDAR activation and Arc expression would increase meal size and decrease the interval between meals. Specifically, NMDAR antagonists or anti-Arc ODNs were infused into the vHC and subsequent intake of sucrose was assessed.     

Awareness of errors and feedback in human time estimation

Behavioral and electrophysiology studies have shown that humans possess a certain self-awareness of their individual timing ability. However, conflicting reports raise concerns about whether humans can discern the direction of their timing error, calling into question the extent of this timing awareness. To understand the depth of this ability, the impact of nondirectional feedback and reinforcement learning on time perception were examined in a unique temporal reproduction paradigm that involved a mixed set of interval durations and the opportunity to repeat every trial immediately after receiving feedback, essentially allowing a “redo.” Within this task, we tested two groups of participants on versions where nondirectional feedback was provided after every response, or not provided at all. Participants in both groups demonstrated reduced central tendency and exhibited significantly greater accuracy in the redo trial temporal estimates, showcasing metacognitive ability, and an inherent capacity to adjust temporal responses despite the lack of directional information or any feedback at all. Additionally, the feedback group also exhibited an increase in the precision of responses on the redo trials, an effect not observed in the no-feedback group, suggesting that feedback may specifically reduce noise when making a temporal estimate. These findings enhance our understanding of timing self-awareness and can provide insight into what may transpire when this is disrupted.

The tendency to reflect on one''s mental state, or metacognition, is an essential human trait that is a central component of consciousness, memory processing, language, and decision-making, and, most importantly, time perception (Fleming and Dolan 2012; Yeung and Summerfield 2012). Accurate time perception is also a hallmark of consciousness and critical for everyday behaviors and cognitive functions ranging from speech, motor control, adaptive behavior, and survival (Meck 2005; Grondin 2010). Self-assessment of one''s own timing ability without any external feedback, known as temporal metacognition, is vital for reliably determining temporal accuracy and variance despite uncertainty (Balcı et al. 2011; Lamotte et al. 2012; Akdoğan and Balcı 2017). Past research shows that humans and rodents can successfully incorporate the endogenous timing uncertainty and trial-by-trial variability associated with time perception tasks into their behavioral responses in a way that maximizes performance, updates temporal representations, and boosts reward (Balcı et al. 2011; Li and Dudman 2013). Human performance is measured via interval timing tasks that instruct participants to estimate temporal intervals of several seconds (Buhusi and Meck 2005). One particular task, temporal reproduction, involves exposure to a specific duration (encoding) and then an opportunity to recreate the interval duration via keypress (reproduction), with or without feedback. Previous research has shown that human subjects can accurately infer the distribution of intervals presented, and use this information to guide reproductions in the phase of measurement uncertainty (Jazayeri and Shadlen 2010; Acerbi et al. 2012). While the evidence already points toward an existing self-awareness of time, evaluating timing aptitude in context of feedback and learning may broaden our understanding of internal metacognitive process and its role in time perception.Detecting and correcting errors is an important component for metacognition and self-awareness of one''s cognitive state (Fleming and Dolan 2012; Yeung and Summerfield 2012). The brain''s performance monitoring system is responsible for assessing and minimizing these errors and facilitating the selection of an appropriate motor program to successfully complete the chosen task or behavior (Ullsperger et al. 2014). Error tracking mechanisms have been reported for temporal, numerosity, and spatial errors, implying that there may be a common metric error-monitoring system that underpins magnitude-based representations (Duyan and Balcı 2019). In particular, how errors related to early or late timing either with or without external feedback are managed is not fully understood. Studies on the impact of feedback on time perception have produced conflicting results, particularly due to the variability in the type of feedback delivery. Experimental timing paradigms offer a broad array of options ranging from no feedback, magnitude and directional feedback, magnitude-based feedback only, or directional-only feedback. Our study is an initial assessment of whether there is self-awareness of directional temporal information and compares the no feedback and magnitude-based conditions, serving as a launching point for further studies.Akdoğan and Balcı (2017) used a temporal reproduction task to assess how subjects reproduced a range of suprasecond intervals; their findings demonstrated that humans are aware of both the magnitude and direction (early/late) of their timing errors despite not receiving any external feedback. Another recent behavioral study used a temporal production task, in which subjects were asked to repeatedly produce a single duration (3 sec) and compared performance during a condition when only the magnitude of the error was given (absolute) against another condition in which both the magnitude and direction (signed) were given (Riemer et al. 2019). Signed feedback delivery yielded more behavioral adjustments in opposition to the direction of the error in subsequent trials, reduced bias in temporal estimates, and produced a more accurate and better calibrated performance when compared with absolute feedback. This study illustrated that directional information was not intrinsically accessible to the subject and that the participant''s internal timing error representation failed to include that error direction. Furthermore, subjects assigned to the absolute feedback group also tended to report an overreproduction of the interval duration when in reality, they were underreproducing (Riemer et al. 2019). A key difference should be noted between these two studies, notably that feedback and retrospective self-judgments respectively were given following an entire block (Riemer et al. 2019) rather than trial by trial as in the Akdoğan and Balcı (2017) experiment. Task context also played a crucial role, as different tasks were used for the two studies; Akdoğan and Balcı (2017) used a temporal reproduction with a mixed set of intervals while Riemer et al. (2019) used a temporal production task with a singly presented interval.In addition to simply supplying knowledge and guidance about the response (Salmoni et al. 1984), feedback has numerous other uses. It reduces response drifts over the experimental trajectory (Salmoni et al. 1984; Riemer et al. 2019) and may be erroneous or correct, but our study concentrates on correct feedback that tends to positively adjust behavioral responses (Salmoni et al. 1984). Its delivery may be absolute—after every trial or on a percentage of trials (relative). Additionally, feedback can be a motivational factor and act as an implicit reward for behavioral learning (Salmoni et al. 1984; Tsukamoto et al. 2006). Posterror feedback can also facilitate the learning of time intervals (Ryan and Robey 2002). Feedback may be processed differently depending on the quality of the learners and is reflected in a well-functioning performance monitoring system (Luft et al. 2013).Numerous timing studies demonstrate that participants are aware of their errors prior to or independent of the administration of feedback (Akdoğan and Balcı 2017; Brocas et al. 2018; Kononowicz et al. 2018). In a recent M/EEG study using a temporal production task with the objective of repeatedly producing the same single interval duration, Kononowicz et al. (2018) asked respondents to first judge their own performance. Afterwards, they were provided with 100% directional feedback in two blocks out of the six and 15% feedback on the remaining blocks. The initial self-assessment of their performance prior to feedback matched the true interval duration and β power operated as an index of the actual duration and the self-evaluative ability to track timing errors (Kononowicz et al. 2018). In yet another single duration temporal production study, Brocas et al. (2018) tasked participants to generate interval durations of 30+ sec repeatedly for 10 trials and introduced a reward scheme to incentivize making accurate temporal estimates. Similarly to Riemer''s paradigm, participants accurately self-evaluated their performance after a block of trials and correctly identified what proportion of trials were above or below the target interval duration (Brocas et al. 2018). Although accurate in their assessments of bias or tendency to overestimate or underestimate, the participants were less successful in prediction beforehand or correction of their responses (Brocas et al. 2018).Self-knowledge about your internal timing behavior also incorporates reinforcement learning, another adaptive behavioral operation that integrates previous behavioral experiences and applies them to future scenarios to improve outcomes and maximize future rewards (Lee et al. 2012). Predicting the value of a set action along with its outcome and determining whether it will be awarded involves precise timing mechanisms (Petter et al. 2018; Mikhael and Gershman 2019). Errors are prone to occur in this process, particularly when there is a mismatch between the expected and actual outcomes, manifesting itself in the form of a reward prediction error (RPE) (Hollerman and Schultz 1998).Notwithstanding errors, the process of learning interval durations transpires fairly quickly and is achieved with only one trial (Simen et al. 2011). To measure the speed of temporal learning, Simen et al. (2011) devised the “beat the clock task,” a paradigm where a green square is displayed on a computer screen for an unknown amount of time. It remains onscreen until the appearance of a red square outline signals that the interval is ending. Participants must respond with a keypress before the stimulus terminates and are rewarded for on-time responses. Larger rewards are delivered the closer a participant makes an on-time response (Simen et al. 2011). Although the task comprises mixed interval durations that transition rapidly, early responding is minimized and on-time responding becomes the norm as participants learn the structure of the task. It is noteworthy that despite the endogenous timing uncertainty stemming from the presentation of rapidly changing durations, performance improvement was not impeded (Simen et al. 2011). In fact, all beat the clock participants improved in response times after only a single trial of beat-the-clock task, quickly reduced their timing errors, and were able to implement a strategy that facilitated learning interval durations (Simen et al. 2011).Using an appropriate feedback technique is pivotal to understanding self-timing awareness. The traditional feedback structure in psychophysics studies is best suited for single interval reproductions because the same interval is successively reproduced; therefore, the majority of the studies described above are single interval experiments. Challenges arise when reproducing a random set of a mixed range of intervals because corrective guidance is given on one interval duration, yet the next interval may be of a different duration (Ryan 2016). This is problematic since there is no opportunity to use the feedback from the previous trial to the new trial, so the feedback is frequently misapplied to an entirely different duration (Ryan 2016). To rectify this issue, we introduced a “redo” trial, which allows the subject to use the original feedback from the first trial in a second trial of the same duration. We hypothesized that the redo trials will be beneficial to subjects for improving performance and in minimizing the Vierordt effect (underestimation of long intervals and overestimation of short intervals) (Ryan 2016). It is for this reason our study also incorporated absolute (nondirectional) feedback: to assess whether subjects possess awareness of the direction of the timing errors. If there is a substantial directional awareness of timing error, then the central tendency would be reduced.     

How does episodic memory develop in adolescence?

Key areas of the episodic memory (EM) network demonstrate changing structure and volume during adolescence. EM is multifaceted and yet studies of EM thus far have largely examined single components, used different methods and have unsurprisingly yielded inconsistent results. The Treasure Hunt task is a single paradigm that allows parallel investigation of memory content, associative structure, and the impact of different retrieval support. Combining the cognitive and neurobiological accounts, we hypothesized that some elements of EM performance may decline in late adolescence owing to considerable restructuring of the hippocampus at this time. Using the Treasure Hunt task, we examined EM performance in 80 participants aged 10–17 yr. Results demonstrated a cubic trajectory with youngest and oldest participants performing worst. This was emphasized in associative memory, which aligns well with existing literature indicating hippocampal restructuring in later adolescence. It is proposed that memory development may follow a nonlinear path as children approach adulthood, but that future work is required to confirm and extend the trends demonstrated in this study.

Episodic memory (EM) describes the ability to encode, store, and retrieve representations of previously experienced episodes and their temporal-spatial context (Tulving 1972). EM development continues well into the third decade of life (Ruggiero et al. 2016); however, its developmental trajectory after the preschool years remains controversial, with some studies suggesting linear improvements (Ofen et al. 2007) and others no improvement (Picard et al. 2012) or a nonlinear pattern (Tulving 1985; Keresztes et al. 2017). While there has been some debate as to the “defining features” of EM (Cheke and Clayton 2013, 2015) most theorists agree that it is not a unitary ability, instead reflecting the combination of a number of contributing features. Given that many of these studies used different methods for testing EM, and that different tests may emphasize different features (Cheke and Clayton 2013, 2015), it is likely that empirical differences reflect the fact that different features of memory may develop differently during later childhood and adolescence (Picard et al. 2012).The importance of understanding the developmental trajectory of EM in adolescence is highlighted in the close association between EM and other cognitive processes. EM is thought to support decision-making, particularly in the incorporation of memories into task- and goal-relevant responses (Murty et al. 2016); thus, immaturity of EM may influence the high levels of risk taking observed in adolescence. Adolescence also represents a period of vulnerability to the development of mental illness (Kessler et al. 2007). Evidence that deficits in EM have been linked to a number of mental health disorders such as depression (Goodwin 1997) and anxiety (Airaksinen et al. 2005) raises the possibility that individual differences in memory development during this period may influence this vulnerability. Finally, adolescence is a demanding time academically: During these school years, large quantities of knowledge must be acquired to be successful in exams, which have long-term impacts on individuals’ academic and professional future. It is therefore important to understand factors that may contribute to individual differences and challenges in learning and memory during this period.Memory development in adolescence has attracted considerable research attention in recent years, with the majority of work conducted on developmental trajectories of brain areas within the memory network. EM relies on a distributed network of brain areas, including the medial temporal and superior parietal lobes and the prefrontal cortex (PFC) (Simons and Spiers 2003). Each area within the network, as well as the network itself, shows protracted maturation across adolescence.     

Short-term memory reactivation of a weak CS–US association promotes long-term memory persistence in conditioned odor aversion

In conditioned odor aversion (COA), the association of a tasteless odorized solution (the conditioned stimulus [CS]) with an intraperitoneal injection of LiCl (the unconditioned stimulus [US[), which produces visceral malaise, results in its future avoidance. The strength of this associative memory is mainly dependent on two parameters, that is, the strength of the US and the interstimuli interval (ISI). In rats, COA has been observed only with ISIs of ≤15 min and LiCl (0.15 M) doses of 2.0% of bodyweight, when tested 48 h after acquisition (long-term memory [LTM]). However, we previously reported a robust aversion in rats trained with ISIs up to 60 min when tested 4 h after acquisition (short-term memory [STM]). Since memories get reactivated during retrieval, in the current study we hypothesized that testing for STM would reactivate this COA trace, strengthening its LTM. For this, we compared the LTM of rats trained with long ISIs or low doses of LiCl initially tested for STM with that of rats tested for LTM only. Interestingly, rats conditioned under parameters sufficient to produce STM, but not LTM, showed a reliable LTM when first tested for STM. These observations suggest that under suboptimal training conditions, such as long ISIs or low US intensities, a CS–US association is established but requires reactivation in the short-term in order to persist in the long-term.

The dynamic and malleable nature of memories is a well-studied phenomenon. Traditionally, for memory formation to occur, a set of processes collectively known as consolidation are thought to be needed in order to stabilize memories, making them susceptible to modification during this period (Dudai et al. 2015). More recently a slightly distinct theory, known as memory integration, was proposed according to which memories are rapidly formed during learning without the need for consolidation, but any relevant information around the event can be integrated modifying them (Gisquet-Verrier and Riccio 2019). Common to both theories though, is that memories alternate between an inactive and an active state and modifications can mostly occur during the active state, which lasts for some time after learning, or during its reactivation due to retrieval (Lee et al. 2017; Albo and Gräff 2018; Gisquet-Verrier and Riccio 2019). Thus, memory malleability is explained either because consolidation can be altered or because additional information can be integrated with the initial memory (Bailey et al. 1996; Dudai 2004; Wixted 2004; Alberini et al. 2006; Lee et al. 2008, 2017; McGaugh and Roozendaal 2009; Roesler and Schröder 2011; Dudai et al. 2015; Nader 2015; Crossley et al. 2019; Gisquet-Verrier and Riccio 2019).In conditioned odor aversion (COA), an odorized tasteless solution (conditioned stimulus, CS) whose ingestion is followed by gastrointestinal malaise (unconditioned stimulus, US) is rejected in future encounters (conditioned response, CR). In most COA studies, a robust aversion has been observed only when the interstimulus interval (ISI) is ∼5 min, and no significant aversion can be seen when the ISI is >15 min (Hankins et al. 1973; Palmerino et al. 1980; Ferry et al. 1995, 1996; Ferry and di Scala 1997; Ferry et al. 2006; Chapuis et al. 2007). This observation has been attributed to a short-lasting memory of the odor that becomes unavailable for its association with the US after ISI >15 min. However, in all these instances the CR was measured 48 h after conditioning (LTM test), leaving up the possibility that CS–US association was formed but somehow did not last till the long-term. In keeping with this possibility, we previously reported a significant aversion during a test performed 4 h after conditioning (i.e., STM test) in rats trained with ISIs up to 60 min, three times longer than previously described (Tovar-Díaz et al. 2011). The LTM, however, was not tested so no further insight was provided regarding its persistence due to STM reactivation.Thus, in the current paper we hypothesized that a STM test would reactivate the initial memory, allowing it to further consolidate/integrate the information and to persist in the LTM. To test this possibility, we trained independent groups of rats with reduced US intensities or prolonged ISIs in a standard two-bottle choice COA paradigm and tested them twice at 4 and 48 h after conditioning. Our findings suggest that COA takes place under milder US and longer ISIs than previously thought and reactivating this memory during the STM test promotes its persistence in the LTM test.     

Depolarization-initiated endogenous cannabinoid release and underlying retrograde neurotransmission in interneurons of amygdala

The depolarization is also important for the short-term synaptic plasticity, known as depolarization-induced suppression of excitation (DSE). The two major types of neurons and their synapses in the lateral nucleus of amygdala (LA) are prone to plasticity. However, DSE in interneurons has not been reported in amygdala in general and in LA in particular. Therefore, we conducted the patch-clamp experiments with LA interneurons. These neurons were identified by lack of adaptation in firing rate of action potentials. In this study, we show for the first time a transient suppression of neurotransmission at synapses both within the local network and between cortical inputs and interneurons of the LA. The retrograde neurotransmission from GABAergic interneurons were comparable with that of glutamatergic pyramidal cells. That is the axonal terminals of cortical inputs do not posses selectivity toward two neuronal subtypes. However, the DSE of both types of neurons involve an increase in intracellular Ca²⁺ and the release of endogenous cannabinoids (eCB) and activation of presynaptic CB1 receptors. The magnitude of DSE was significantly higher in interneurons compared with pyramidal cells, though developed with some latency.

…I made experiments on myself and my assistant, using smaller doses, and not repeating them so often… — Clouston 1870

The biological actions of endogenous cannabinoids (eCB) occur by binding to the CB₁ and CB₂ receptors throughout the whole body (Ameri 1999; Pertwee 2006; Hill et al. 2007; Yoshida et al. 2011). The density of CB₁ receptors in the amygdala is comparably high in mammals (Herkenham et al. 1990).Amygdala similar to hippocampus is important for memory formation and often studied to elucidate plasticity at cellular level using the classical paradigm of Pavlov that continuously serves as a substrate (Pavlov 1927; Bliss and Lomo 1973; Rogan et al. 1997). The amygdala not only receives, but also sends behavior underlying signals into other regions (Racine et al. 1983; Aggleton and Mishkin 1984). While the role of hippocampus is crucial for memory formation, those associated with many different kinds of emotions are mainly modulated by the amygdala (Bucherelli et al. 2006; Fujii et al. 2020). The memory enabling substrate is a long-term potentiation (LTP) of neurotransmission into the postsynaptic neurons (Rogan et al. 1997; Kodirov et al. 2006).The short-term synaptic plasticity in the form of depolarization-induced suppression of either excitation or inhibition (DSE and DSI) has been reported in several regions of the brain (Alger et al. 1996; Kano et al. 2009; Ivanova and Storozhuk 2011). We have discovered DSE in the lateral amygdala (LA), specifically at cortical inputs into the pyramidal neurons (Kodirov et al. 2010).Despite the extensive studies on DSE and DSI, there are only three papers on interneurons that we are aware of. Two of them describe the presence of these phenomena: One study was carried out on parvalbumin immunoreactive interneurons of the stratum radiatum in the hippocampus (Ali 2007) and another on cerebellar stellate and basket cells (Beierlein and Regehr 2006). However, none of the cortical interneurons exhibited DSI despite the presence of a functional and cannabinoid-sensitive inhibitory inputs (Lemtiri-Chlieh and Levine 2007). The retrograde neurotransmission (Llano et al. 1991) takes place via the release of two natural ligands of endogenous cannabinoids anandamide and 2-arachidonoyl-glycerol (Urbanski et al. 2009). These ligands also suppress the evoked excitatory neurotransmission when applied exogenously in vitro (Ameri et al. 1999; Ameri and Simmet 2000; Lemak et al. 2007).Since DSE in interneurons has not been reported in amygdala and we demonstrated the existence of DSE in pyramidal cells of LA (Kodirov et al. 2010), we then studied the same phenomenon in regard to interneurons; the main question was whether or not does depolarization-induced mobilization of eCBs from the two types of postsynaptic LA neurons cause similar retrograde modulation of cortical inputs? Subsequently, DSE between the presynaptic terminals and interneurons was shown, and we found that its properties are similar to those in pyramidal cells of LA. Our study documents the participation of endogenous cannabinoids of interneurons in DSE.     

Contributions of glucocorticoid receptors in cortical astrocytes to memory recall

Dysfunctions in memory recall lead to pathological fear; a hallmark of trauma-related disorders, like posttraumatic stress disorder (PTSD). Both, heightened recall of an association between a cue and trauma, as well as impoverished recall that a previously trauma-related cue is no longer a threat, result in a debilitating fear toward the cue. Glucocorticoid-mediated action via the glucocorticoid receptor (GR) influences memory recall. This literature has primarily focused on GRs expressed in neurons or ignored cell-type specific contributions. To ask how GR action in nonneuronal cells influences memory recall, we combined auditory fear conditioning in mice and the knockout of GRs in astrocytes in the prefrontal cortex (PFC), a brain region implicated in memory recall. We found that knocking out GRs in astrocytes of the PFC disrupted memory recall. Specifically, we found that knocking out GRs in astrocytes in the PFC (AstroGRKO) after fear conditioning resulted in higher levels of freezing to the CS+ tone when compared with controls (AstroGRintact). While we did not find any differences in extinction of fear toward the CS+ between these groups, AstroGRKO female but not male mice showed impaired recall of extinction training. These results suggest that GRs in cortical astrocytes contribute to memory recall. These data demonstrate the need to examine GR action in cortical astrocytes to elucidate the basic neurobiology underlying memory recall and potential mechanisms that underlie female-specific biases in the incidence of PTSD.

Recalling important information about salient environmental cues is an integral part of how we navigate our world. Recalling too much, or too little, information about salient environmental cues is a part of the psychopathology of posttraumatic stress disorder (PTSD) (Milad and Quirk 2012). More specifically, the augmented recall of an association between an environmental cue and a traumatic event results in debilitating fear toward the cue, even in the absence of any threat. In contrast, impoverished recall of information that a cue, previously associated with trauma, is no longer a threat also results in debilitating fear toward the cue after it is no longer dangerous. Therefore, one way to mitigate debilitating fear that characterizes PTSD is to understand the neurobiological mechanisms underlying memory recall.Among many mechanisms, glucocorticoid action via signaling through glucocorticoid receptors (GRs) is an important neurobiological pathway that underlies the recall of salient information. When trauma-associated cues are encountered, the hypothalamic-pituitary-adrenal axis is activated and GR signaling is consequently triggered (McEwen et al. 1988; McEwen 1992; Lupien et al. 2009). Existing literature demonstrates that glucocorticoids and GRs do in fact influence learning, memory, and the recall of learning (Pugh et al. 1997; de Quervain et al. 1998, 2009, 2011, 2017, 2019; Roozendaal 2002, 2003; Conrad et al. 2004; Hui et al. 2004; Donley et al. 2005; Roozendaal and de Quervain 2005; Cai et al. 2006; Soravia et al. 2006; Yang et al. 2006; Roozendaal et al. 2009; Bentz et al. 2010; Blundell et al. 2011; Clay et al. 2011; Nikzad et al. 2011; Roesler 2012; Liao et al. 2013; Wislowska-Stanek et al. 2013; Arp et al. 2016; Reis et al. 2016; Dadkhah et al. 2018; Inoue et al. 2018; Scheimann et al. 2019; Lin et al. 2020). The relationship between glucocorticoids, GRs, learning and memory is complicated and within the literature cited above, one can find examples of GR action being facilitatory as well as inhibitory to learning and memory recall. As expansive as this research is, the influence of GRs on learning, memory, and recall of learning has mostly focused only on GR action in neurons or has ignored cell type specific contributions. While glia are approximately as common as neurons in the nervous system (von Bartheld et al. 2016; von Bartheld 2018), the role of GRs in glial cells on the recall of salient environmental cues has been neglected. More specifically, while astrocytes comprise a significant proportion of the glial cell population (von Bartheld et al. 2016) and express GRs (Vielkind et al. 1990; Bohn et al. 1991), the influence of GRs in astrocytes on memory recall remains largely unappreciated (for one exception, see the Discussion).Our goal in this study was to determine the influence of GRs in astrocytes on memory recall. To do so, we combined the robust and reliable experimental framework of classical fear conditioning in rodents (Santini et al. 2008; Dias et al. 2014; Bukalo et al. 2015; Keiser et al. 2017; Giustino and Maren 2018; Greiner et al. 2019; Gunduz-Cinar et al. 2019; Venkataraman et al. 2019) with molecular genetic manipulations in the prefrontal cortex (PFC), a brain region critical for the recall of memory (Morgan and LeDoux 1995; Quirk et al. 2000; Mueller et al. 2008; Quirk and Mueller 2008; Giustino and Maren 2015; Rozeske et al. 2015; Maren and Holmes 2016). We first trained mice to associate tone presentations with mild footshocks. After this auditory fear conditioning, we used a CRE-loxP strategy to specifically knock out GRs in astrocytes in the PFC (hereafter termed cortical astrocytes) of these trained mice. We then exposed animals to extinction training: 30 presentations of the tone in the absence of any footshocks. Finally, 1 d after the extinction training, we exposed animals to two presentations of the tone. This experimental timeline allowed us to ask how a lack of GRs in cortical astrocytes influences (1) the recall of the previous aversive association of the tone presentation with the footshock, (2) the extinction of fear that would typically occur during extinction training, and (3) the recall of extinction training allowing us to measure the influence of GRs in cortical astrocytes on the recall of extinction training. Broadly, our results demonstrate that knocking out GRs in cortical astrocytes disrupts fear memory recall in both male and female mice, while only disrupting extinction recall in female mice.