512 • The Journal of Neuroscience, January 17, 2018 • 38(3):512–514

Journal Club

Editor’s Note: These short reviews of recent JNeurosci articles, written exclusively by students or postdoctoral fellows, summarize the important findings of the paper and provide additional insight and commentary. If the authors of the highlighted article have written a response to the Journal Club, the response can be found by viewing the Journal Club at www.jneurosci.org. For more information on the format, review process, and purpose of Journal Club articles, please see http://jneurosci.org/content/ preparing-manuscript#journalclub.

Sleep-Stage-Dependent Hippocampal Coordination with Cingulate and Retrosplenial Association Cortices

Andrew S. Alexander1 and Laura E. Shelley2 1Department of Psychological and Sciences, Center for Systems Neuroscience, Boston University, Boston, Massachusetts, 02215, and 2Department of Cognitive Science, University of California, San Diego, La Jolla, California, 92093 Review of Koike et al.

A long line of research has examined the RSC, medial (MEC), and within the region. Koike et al. (2017) per- role of the in learning, epi- the (DG). Accordingly, the formed cross-correlation analyses on pairs sodic spatial and nonspatial memory, and observed pattern of c-Fos expression during of raw LFPs between regions and observed navigation. Although critically important REM reveals a cortico-hippocampal circuit significantly greater synchronization be- in these processes, the hippocampus does with a potentially unique role in REM- tween the RSC and hippocampus during not act in isolation. Effective hippocampal -dependent processing. REM than during waking. Furthermore, mnemonic and spatial processing requires A recent study in The Journal of Neuro- Granger causality analysis indicated that co- cortical input, and the output of hippocampal science sought to identify the neurophysi- herence in the theta frequency range be- network computations must be integrated ological basis for increased c-Fos levels in tween the two regions was driven by the with cortex to guide complex behavior. Much RSC and ACC following REM hypersom- hippocampus rather than RSC. No such re- work on cortical-hippocampal interaction nia and to investigate synchronization lationship between the ACC and either the has focused on synchronization with the between these cortical structures and the hippocampus or RSC was observed. medial prefrontal cortex (mPFC) but hippocampus (Koike et al., 2017). The au- Activation of in several corti- recent data suggest that the anterior cin- thors recorded single units and local field cal and subcortical regions is entrained gulate (ACC) and retrosplenial (RSC) as- potentials (LFPs) simultaneously in the with the hippocampal theta rhythm sociation cortices may play significant ACC, RSC, and hippocampus of rats across across a wide range of behavioral states roles in this coordination process, espe- arousal states in vivo. They found that a ma- and tasks (Berke et al., 2004; Sirota et al., cially during sleep. jority of RSC neurons and nearly half of 2008). At the level of individual neurons, Recent observations indicate that in- ACC units were more active during REM theta-phase coupling is manifested as creased [(REM; than during any other sleep–wake state. The spiking locked to particular phases of the also known as paradoxical sleep (PS)] is increased firing activity during REM forms theta rhythm. At the level of cortical positively correlated with increased ex- electrophysiological support for their ensembles, theta entrainment is often pression of c-Fos, an immediate early gene conclusion that increased c-Fos expres- manifested as phase–amplitude coupling that serves as a biomarker of neural activ- sion in these areas results from REM- between theta phase and increased ampli- ity (Renouard et al., 2015). Critically, in- specific activation. tude within the gamma frequency range creased c-Fos expression following REM In addition to finding increased activity, (40–160 Hz) recorded in the LFP. The ob- is restricted to the ACC, Koike et al. (2017) detected a prominent served interaction between ACC, RSC, theta rhythm in the LFP in all three struc- and hippocampal theta rhythms suggests

Received Sept. 26, 2017; revised Nov. 13, 2017; accepted Nov. 20, 2017. tures during REM sleep. A similar rhythm that these cortical regions are similarly The authors declare no competing financial interests. was present in RSC and hippocampus, but coordinated within theta frequency tem- Correspondence should be addressed to Dr. Andrew S. Alexander, De- substantially diminished in ACC, during the poral windows and that interaction is partment of Psychological and Brain Sciences, Center for Systems Neuro- awake state. The theta rhythm is a 6–12 Hz increased specifically during REM. Con- science, Boston University, 610 Commonwealth Avenue, Boston, MA, 02215. E-mail: [email protected]. oscillation measured in the LFP of the hip- sistent with this hypothesis, Koike et al. DOI:10.1523/JNEUROSCI.2781-17.2017 pocampus (among numerous other areas) (2017) found evidence of single-unit and Copyright © 2018 the authors 0270-6474/18/380512-03$15.00/0 that entrains the spiking activity of neurons ensemble coupling to theta-frequency os- Alexander and Shelley • Journal Club J. Neurosci., January 17, 2018 • 38(3):512–514 • 513 cillatory activity. Neurons in both ACC ing, whereas sleep-dependent memory im- processing state, wherein populations of and RSC exhibited theta phase-coupling provement is observed across species in a neurons that were coactive during prior to the theta rhythm recorded in the hip- variety of tasks. Increased hippocampal events strengthen their connections via pocampus during awake states, and RSC communication with cortical areas might spike timing-dependent plasticity mecha- neurons exhibited significantly in- reflect the consolidation of episodic memo- nisms. Indeed, recent work has shown creased theta phase modulation during ries across cortical populations. increased activation of ACC units and in- REM. The authors additionally observed There is some indication that different creased ACC-hippocampus gamma-cohe- phase–amplitude coupling between hip- sleep stages play different roles in memory rence locked to SWRs during waking and pocampal theta phase and RSC gamma- consolidation. Perhaps most relevant to SWS (Remondes and Wilson, 2015; Wang frequency power during REM, but not the current work, Boyce et al. (2016) re- and Ikemoto, 2016). during awake states. Like the preferred cently demonstrated a link between the The lack of theta–gamma phase–am- phase of single-unit theta entrainment, occurrence of REM theta oscillations and plitude coupling between hippocampus gamma oscillations in RSC occurred just consolidation of context-dependent me- and ACC is somewhat at odds with the after the peak of the . mories. Accordingly, hippocampus-RSC presence of ACC units whose firing activ- Phasic entrainment of neurons with synchronization during REM might be im- ity is locked to hippocampal theta, and the theta rhythm has been proposed to portant for the integration of unique sen- suggests that only a small percentage of underlie different cortico-hippocampal sory or behavioral contextual information ACC units exhibit this tendency. This ob- processing states (Hyman et al., 2005). with episodic memories before cortical con- servation could be explained by relatively For instance, activation of neurons during solidation. RSC is reciprocally connected to sparse direct projections from the hip- the peak or trough of the theta wave has multiple cortical sensory processing re- pocampus to ACC compared with hippo- been shown to yield long-term potentia- gions, and neurons within the region campus-RSC interconnectivity (Wyss tion or long-term depression, respectively exhibit activation patterns that can be sen- and Van Groen, 1992). On the other (Huerta and Lisman, 1995). Accordingly, sitive to sensory context and spatial loca- hand, the absence of theta–gamma cou- hippocampal theta modulation of RSC tion (Wyss and Van Groen, 1992; Jacob et pling in ACC during all arousal states and ensembles during REM might indicate a al., 2017). In support of this hypothesis, in RSC during waking could be explained mechanism by which the output of hip- optogenetic activation of RSC ensembles by diminished active navigation during pocampal computations is integrated with initially active during contextual fear con- waking states, given that most sessions cortical circuits. The presence of theta– ditioning is sufficient to reinstate learned were dominated by sleep with relatively gamma coupling during REM but not behavior (Cowansage et al., 2014), sug- short bouts of waking interspersed. Neu- waking is interesting considering that in- gesting that a contextual memory can be rons in both the ACC and RSC are sensi- dividual RSC neurons exhibit theta en- consolidated for future recall within RSC. tive to action execution, and RSC neurons trainment regardless of arousal state. This The limited interaction between hip- exhibit spatially modulated firing only finding implies that larger populations of pocampus and ACC reported by Koike et recognizable during active navigation (Al- RSC neurons become coupled to the hip- al. (2017) is surprising given the pattern of exander and Nitz, 2017). Accordingly, a pocampal theta rhythm during REM, c-Fos expression levels. Although the mag- lack of significant movement during which could produce greater gamma nitude of ACC-hippocampal theta coher- awake states may decrease firing activity frequency power at the peak of the theta ence reported by Koike et al. (2017).is in both regions which could possibly ex- wave. consistent with previous reports describing plain the significantly greater mean fir- Given these observations, one might both theta coherence generally and task- ing rate observed in REM. Furthermore, ask what neural processes are supported dependent theta frequency coherence shifts, the power and frequency of the hippocam- via increased hippocampal communica- the authors find little other evidence of pal theta rhythm is positively correlated tion with these structures, and especially ACC-hippocampal interaction across arousal with running speed. In combination, de- RSC, during REM. Multiple lines of states (Young and McNaughton, 2009; Re- creased firing activity and potentially weak evidence indicate that sleep-related syn- mondes and Wilson, 2013). However, it is theta rhythms may explain the reduction in chronization between cortex and the hip- possible that ACC communicates with hip- theta–gamma phase–amplitude coupling pocampus might play a special role for pocampus during more discrete time win- during waking states. memory processing. At the most basic level, dows of interaction such as hippocampal Along these lines and as Koike et al. the construction of an episodic memory re- sharp-wave-ripple events (SWRs). SWRs (2017) point out, it would be useful to ex- quires sequentially linking spatially and are short duration (100–200 ms), high- amine interaction within this circuitry temporally discrete events in time, a pro- frequency (150–250 Hz), hippocampal LFP following more complex behavior includ- cess that the hippocampal architecture events that occur during slow-wave sleep ing different spatial and nonspatial learn- and subsequent neurophysiological dy- (SWS) or immobile waking. SWRs are coin- ing paradigms. Engagement of RSC and namics seem especially suited for. It is well cident with the simultaneous activation of ACC neurons is correlated with learning known that initial episodic processing re- large populations of hippocampus neurons and memory, choice, action, various spa- quires the hippocampus, but following in addition to cortical ensembles (Siapas tial variables, and much more (Cowen et encoding, episodic memory becomes and Wilson, 1998; Csicsvari et al., 1999). al., 2012; Jacob et al., 2017, Alexander and hippocampus independent and likely During SWS and SWR-related activity Nitz, 2017). Although Koike et al. (2017) consolidated into distributed cortical bursts, neurons often exhibit sequential ac- do an excellent job characterizing the populations (Frankland and Bontempi, tivation that mimics activity patterning neurophysiological dynamics within this 2005). It is thought that sleep is important during previous experience (Wilson and circuit, it is possible that the profile of in- for this consolidation process, and several McNaughton, 1994;Rothschild et al., 2017). teraction between these association corti- studies have reported that sleep disruption Thus, cortical and hippocampal activation ces and the hippocampus would shift if produces impairments in memory encod- during SWRs potentially reflects an offline the animal were engaged in more cogni- 514 • J. Neurosci., January 17, 2018 • 38(3):512–514 Alexander and Shelley • Journal Club tively demanding behaviors during the Huerta PT, Lisman JE (1995) Bidirectional synap- Salin P, Malleret G, Fort P, Luppi PH (2015) awake state. tic plasticity induced by a single burst during The supramammillary nucleus and the claus- cholinergic theta oscillation in CA1 in vitro. trum activate the cortex during REM sleep. Sci 15:1053–1063. CrossRef Medline Adv 1:e1400177. CrossRef Medline References Hyman JM, Zilli EA, Paley AM, Hasselmo ME Rothschild G, Eban E, Frank LM (2017) A corti- Alexander AS, Nitz DA (2017) Spatially periodic (2005) Medial prefrontal cortex cells show cal–hippocampal–cortical loop of information activation patterns of en- dynamic modulation with the hippocampal processing during memory consolidation. Nat code route sub-spaces and distance traveled. theta rhythm dependent on behavior. Hip- Neurosci 20:251–259. CrossRef Medline Curr Biol 27:1551–1560.e4. CrossRef Medline pocampus 15:739–749. CrossRef Medline Siapas AG, Wilson MA (1998) Coordinated in- Berke JD, Okatan M, Skurski J, Eichenbaum HB Jacob PY, Casali G, Spieser L, Page H, Overington D, teractions between hippocampal ripples and (2004) Oscillatory entrainment of striatal Jeffery K (2017) An independent, landmark- cortical spindles during slow-wave sleep. Neu- neurons in freely moving rats. Neuron 43: dominated head direction signal in dysgranular ron 21:1123–1128. CrossRef Medline 883–896. CrossRef Medline retrosplenial cortex. Nat Neurosci 20:173–175. Sirota A, Montgomery S, Fujisawa S, Isomura Y, Boyce R, Glasgow SD, Williams S, Adamantidis A Medline Zugaro M, Buzsa´ki G (2008) Entrainment of (2016) Causal evidence for the role of REM Koike BDV, Farias KS, Billwiller F, Almeida-Filho D, neocortical neurons and gamma oscillations sleep theta rhythm in contextual memory con- Libourel PA, Tiran-Cappello A, Parmentier R, by the hippocampal theta rhythm. Neuron 60: solidation. Science 352:812–816. CrossRef Blanco W, Ribeiro S, Luppi PH, Queiroz CM 683–697. CrossRef Medline Medline (2017) Electrophysiological evidence that the Wang DV, Ikemoto S (2016) Coordinated interac- Cowansage KK, Shuman T, Dillingham BC, retrosplenial cortex displays a strong and spe- tion between hippocampal sharp-wave ripples Chang A, Golshani P, Mayford M (2014) Di- cific activation phased with hippocampal theta rect reactivation of a coherent neocortical during paradoxical (REM) sleep. J Neurosci 37, and anterior cingulate unit activity. J Neurosci memory of context. Neuron 84:432–441. 8003–8013. CrossRef Medline 36:10663–10672. CrossRef Medline CrossRef Medline Remondes M, Wilson MA (2013) Cingulate-hippo- Wilson MA, McNaughton BL (1994) Reactiva- Cowen SL, Davis GA, Nitz DA (2012) Anterior campus coherence and trajectory coding in a tion of hippocampal ensemble memories cingulate neurons in the rat map anticipated sequential choice task. Neuron 80:1277–1289. during sleep. Science 265:676–679. CrossRef effort and reward to their associated action CrossRef Medline Medline sequences. J Neurophysiol 107:2393–2407. Remondes M, Wilson MA (2015) Slow-␥ rhythms Wyss JM, Van Groen T (1992) Connections CrossRef Medline coordinate cingulate cortical responses to between the retrosplenial cortex and the hip- Csicsvari J, Hirase H, Czurko´ A, Mamiya A, Buzsa´ki hippocampal sharp-wave ripples during wake- pocampal formation in the rat: a review. Hip- G (1999) Oscillatory coupling of hippocampal fulness. Cell Rep 13:1327–1335. CrossRef pocampus 2:1–11. CrossRef Medline pyramidal cells and interneurons in the behav- Medline Young CK, McNaughton N (2009) Coupling of ing rat. J Neurosci 19:274–287. Medline Renouard L, Billwiller F, Ogawa K, Cle´ment O, theta oscillations between anterior and poste- Frankland PW, Bontempi B (2005) The organi- Camargo N, Abdelkarim M, Gay N, Scote´- rior midline cortex and with the hippocampus zation of recent and remote memories. Nat Blachon C, Toure´ R, Libourel PA, Ravassard in freely behaving rats. Cereb Cortex 19:24– Rev Neurosci 6:119–130. CrossRef Medline P, Salvert D, Peyron C, Claustrat B, Le´ger L, 40. CrossRef Medline