Distinct Neural Circuits for the Formation and Retrieval …

Loading...

Taking too long?

Reload document

| Open in new tab

Distinct Neural Circuits for the Formation and
Retrieval of Episodic Memories

Graphical Abstract

Article

Authors
Dheeraj S. Roy, Takashi Kitamura,
Teruhiro Okuyama, …, Yuichi Obata,
Atsushi Yoshiki, Susumu Tonegawa

Correspondence
tonegawa@mit.edu

In Brief
Episodic memories are formed and
retrieved through distinct hippocampal
pathways.

Highlights
d dSub and the circuit, CA1/dSub/EC5, are required for

hippocampal memory retrieval

d The direct CA1/EC5 circuit is essential for hippocampal

memory formation

d The dSub/MB circuit regulates memory-retrieval-induced

stress hormone responses

d The dSub/EC5 circuit contributes to context-dependent

memory updating

Roy et al., 2017, Cell 170, 1000–1012
August 24, 2017 ª 2017 Elsevier Inc.
http://dx.doi.org/10.1016/j.cell.2017.07.013

Article

Distinct Neural Circuits for the Formation
and Retrieval of Episodic Memories

Dheeraj S. Roy,1,5 Takashi Kitamura,1,5 Teruhiro Okuyama,1 Sachie K. Ogawa,1 Chen Sun,1 Yuichi Obata,2
Atsushi Yoshiki,2 and Susumu Tonegawa1,3,4,6,*
1RIKEN-MIT Center for Neural Circuit Genetics at the Picower Institute for Learning and Memory, Department of Biology and Department of
Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, MA 02139, USA
2RIKEN BioResource Center, 3-1-1 Koyadai, Ibaraki 305-0074, Japan
3Howard Hughes Medical Institute, Massachusetts Institute of Technology, Cambridge, MA 02139, USA
4RIKEN Brain Science Institute, 2-1 Hirosawa, Wako-shi, Saitama 351-0198, Japan
5These authors contributed equally
6Lead Contact
*Correspondence: tonegawa@mit.edu
http://dx.doi.org/10.1016/j.cell.2017.07.013

SUMMARY

The formation and retrieval of a memory is thought to
be accomplished by activation and reactivation,
respectively, of the memory-holding cells (engram
cells) by a common set of neural circuits, but this
hypothesis has not been established. The medial
temporal-lobe system is essential for the formation
and retrieval of episodic memory for which individual
hippocampal subfields and entorhinal cortex layers
contribute by carrying out specific functions. One
subfield whose function is poorly known is the subic-
ulum. Here, we show that dorsal subiculum and
the circuit, CA1 to dorsal subiculum to medial ento-
rhinal cortex layer 5, play a crucial role selectively in
the retrieval of episodic memories. Conversely, the
direct CA1 to medial entorhinal cortex layer 5 circuit
is essential specifically for memory formation. Our
data suggest that the subiculum-containing detour
loop is dedicated to meet the requirements associ-
ated with recall such as rapid memory updating
and retrieval-driven instinctive fear responses.

INTRODUCTION

It is generally thought that formation and retrieval of a memory
are accomplished by activation and reactivation of memory-
holding cells (engram cells), respectively, by a largely common
set of neural circuits that convey relevant sensory and pro-
cessed information. However, this hypothesis has not been
well studied. One of the best neural systems to prove this issue
is the medial temporal lobe (MTL), including the hippocampus
(HPC) and entorhinal cortex (EC), which plays crucial roles in
episodic memory (Eichenbaum et al., 2007; Squire, 1992).
Numerous studies have identified specific and crucial roles of
individual HPC subfields and EC layers to the overall mnemonic
function (Deng et al., 2010; Hasselmo and McClelland, 1999;
Hitti and Siegelbaum, 2014; Moser et al., 2014; Nakazawa
function of one HPC
et al., 2004). However, the essential

1000 Cell 170, 1000–1012, August 24, 2017 ª 2017 Elsevier Inc.

subfield, subiculum (Sub), is poorly known. The mammalian
HPC formation is organized primarily as a unidirectional circuit,
where information transferred from the EC’s superficial layers to
the dentate gyrus (DG) is processed successively in CA sub-
fields: CA3, CA2, and CA1. Dorsal CA1 (dCA1) sends its primary
projections directly to medial EC layer 5 (EC5) or indirectly via
dorsal subiculum (dSub) (a detour circuit). One of the interesting
differences between the direct and indirect HPC output path-
ways is that in the latter, dSub projects not only to EC5, but
also to many cortical and subcortical brain regions (Ding,
2013; Kishi et al., 2000).

Using functional magnetic resonance imaging of human sub-
jects, several studies have suggested that the DG and CA
subfields are selectively activated during episodic memory
formation, whereas subiculum (Sub) is active during the recollec-
tion of an episode (Eldridge et al., 2005; Gabrieli et al., 1997). In
rodents, ibotenic acid lesions of the CA1 subfield or Sub caused
impairments in the acquisition of place navigation (Morris et al.,
1990). However, since human imaging studies provide only
correlative, rather than causal, evidence and rodent lesions are
not well targeted to a specific hippocampal subregion—espe-
cially given the close proximity of CA1 and dSub—it has not
been possible to identify the essential function of Sub in episodic
memory. Furthermore, previous studies did not address the
potential purpose of
the parallel diverging and converging
dCA1 to medial EC5 and dCA1 to dSub to medial EC5 circuits
in memory formation versus retrieval.

In the present study, we addressed these issues by creating a
mouse line expressing Cre recombinase specifically in dSub
neurons. Combined with circuit tracing and optogenetic manip-
ulations during behavioral paradigms, we found differential roles
of dSub projections in hippocampal memory retrieval and
retrieval-induced stress hormone responses. We demonstrate
that dSub and the circuit, CA1/dSub/EC5, are selectively
required for memory retrieval, while the dSub to mammillary
bodies (MB) circuit regulates stress hormones following memory
retrieval. In contrast, the direct CA1/EC5 circuit is essential for
hippocampal memory formation, but not recall. Our study
reveals a functional double-dissociation between parallel hippo-
campal output circuits that are important for memory formation
versus memory retrieval.

Figure 1. Genetic Targeting of dSub Neu-
rons Using FN1-Cre Mice
(A) FN1-Cre mice were injected with a Cre-
dependent virus expressing eYFP into dSub.
(B) Cre+ dSub neurons (eYFP, green) do not
overlap with dCA1 excitatory neurons (labeled
with WFS1, red). Sagittal
image (left), higher-
magnification image of boxed region (right).
Dashed white line denotes CA1/dSub border
(right).
(C and D) Cre+ dSub neurons (eYFP, green), in
sagittal sections, express the excitatory neuronal
marker CaMKII (red; C). Over 85% of all CaMKII+
neurons in the dSub region also expressed eYFP
(n = 3 mice). Images are taken with a 203 objec-
tive. Cre+ dSub neurons do not express the
inhibitory marker GAD67 (red; D). White arrows
indicate GAD67+ cell bodies (D). Images are taken
with a 403 objective. See also Figure S1A. DAPI
staining in blue.
(E–K) Medial to lateral (ML, in millimeters relative to
Bregma) sagittal sections show that eYFP signal is
restricted to dSub neurons. DAPI staining (blue).
No eYFP signal was observed in ventral subiculum
(vSub) or medial entorhinal cortex (MEC). Dashed
white line denotes perirhinal cortex/MEC border
(J and K).

Cre expression in this mouse line is
highly restricted to dSub and the dorsal
in the brain
tegmental nucleus (DTg)
stem (Figure S1L). Thus, FN1-Cre mice
allows for the genetic manipulation of
dSub excitatory neurons with unprece-
dented specificity.

Input-Output Organization of dSub
Neurons

RESULTS

Generation of FN1-Cre Mice

We took advantage of the finding that fibronectin-1 (FN1) gene
expression is restricted to dSub neurons (Lein et al., 2004) and
created a transgenic mouse line (FN1-Cre) that expresses Cre re-
combinase under the FN1 promoter (Figure 1A and see STAR
Methods). When infected with a Cre-dependent adeno-associ-
ated virus containing an eYFP gene, eYFP expression was highly
restricted to dSub neurons and was completely absent in neigh-
boring dCA1 excitatory neurons identified by WFS1 (Figure 1B).
The expression of eYFP was restricted to CaMKII+ excitatory
neurons in both the deep and superficial layers of dSub (Figures
1C, 1D, and see Figure S1A). This eYFP expression accounted for
over 85% of all excitatory neurons in this brain region and was
dSub specific along the entire medial-lateral axis (Figures
1E–1K). Further, Cre expression was absent in ventral subiculum
(vSub) and medial entorhinal cortex (MEC) in this mouse line (Fig-
ures S1B–S1K). Using in situ hybridization, we confirmed that

We next examined the inputs to dSub
excitatory neurons, as well as their anter-
ograde brain-wide projection pattern. Monosynaptic retrograde
tracing experiments using a Cre-dependent helper virus com-
bined with rabies virus (RV) expressing mCherry (Wickersham
et al., 2007) labeled 78% of dSub cells relative to all cells (i.e.,
DAPI+ cells). The results confirmed that dCA1 provides the major
input to dSub excitatory neurons (Figures 2A–2C) (Ding, 2013;
Kishi et al., 2000). Other brain areas that provide inputs to dSub
include parasubiculum (PaS), retrosplenial agranular cortex
(RSA), superficial layers of EC (MEC II/III), nucleus of the diagonal
band (NDB), nucleus accumbens shell (Acb Sh), and several
thalamic nuclei (Thal Nucl) (Figure 2D, and see Figure S1M).

A Cre-dependent channelrhodopsin-2 (ChR2)-eYFP virus
combined with light sheet microscopy of CLARITY (Chung
et al., 2013, see STAR Methods)-processed brain samples re-
vealed that major efferents of dSub neurons were directed to
RSA, mammillary bodies (MB), medial EC5, and postrhinal
cortex (Pos) (Figures 2E and 2F). No projections from dSub
were observed in the superficial layers (II/III) of MEC (Figure 2G).
These dSub neurons converged on both medial and lateral

Cell 170, 1000–1012, August 24, 2017 1001

(Ipsi) and contralateral

Figure 2. Input-Output Organization of dSub Excitatory Neurons
(A) Monosynaptic retrograde tracing of dSub inputs used a Cre-dependent
helper virus (tagged with eGFP) combined with a rabies virus (RV, mCherry)
injected into dSub of FN1-Cre mice. Avian leukosis and sarcoma virus sub-
group A receptor (TVA) and rabies glycoprotein (G).
(B and C) Representative ipsilateral sections confirmed efficient overlap of
helper and RV-infected dSub neurons. Sagittal image (left; B), higher-magni-
fication images of boxed region (right; B). Quantification revealed that 78% of
dSub cells, relative to DAPI+ neurons, were RV positive (n = 4 mice), which is
the starting population for retrograde tracing. Dashed white lines denote dSub
Cre+ neuron target region. Both ipsilateral and contralateral sagittal sections
revealed that dorsal CA1 provides the major input to dSub Cre+ neurons (C).
(D) Inputs to dSub Cre+ neurons were quantified based on percentage of
neurons in the target brain region relative to DAPI+ neurons (n = 4 mice).
Ipsilateral
(Contra) counts. Parasubiculum (PaS),
retrosplenial agranular cortex (RSA), MEC layers II/III (MEC II/III), nucleus of the
diagonal band (NDB), nucleus accumbens shell (Acb Sh), and thalamic nuclei
(Thal Nucl).
in dSub
(E) FN1-Cre mice expressing ChR2-eYFP (Cre-dependent virus)
neurons were used for CLARITY followed by light sheet microscopy (top).
2.5 mm optical section in sagittal view shows projections to RSA and
mammillary bodies (MB, bottom).
(F) Whole-brain, stitched z stack (horizontal view) shows all major projections
from dSub Cre+ neurons, including RSA, MB, EC5, and postrhinal cortex (Pos).
(G and H) Standard sagittal brain sections of FN1-Cre mice expressing ChR2-
eYFP (Cre-dependent virus) in dSub neurons showing dSub projections to
EC5 and Pos (G), as well as medial and lateral MB (H).
(I–M) Representative standard sagittal brain sections showing dSub neuronal
populations projecting to MB (red, CTB555; I) or EC5 (green, CTB488; J). The
respective CTB was injected into MB or EC5. Overlap image (K). Quantifica-
tion, including weakly labeled CTB+ neurons, revealed that 81% of dSub cells
were double-positive (n = 4 mice).
Scale bar in (I) and (J) apply to (K). Dashed white line denotes CA1/dSub
border. Higher-magnification images of boxed regions indicated in Figure 2K
(L and M).

regions of MB (Figure 2H). Using a Cre-dependent synaptophy-
sin virus to label dSub axonal terminals, we found that these Cre+
neurons express vesicular glutamate transporters 1 and 2
(Kaneko et al., 2002), reflecting their excitatory nature (Figures
S1N–S1P). Injection of a retrograde tracer, cholera toxin subunit
B (CTB), into the MB revealed a gradient of CTB555 with higher
intensity labeling in the proximal part of dSub (i.e., closer to CA1),
whereas injection into medial EC5 showed a gradient of CTB488
with higher intensity labeling in the distal part of dSub (i.e., away
from CA1) (Figures 2I–2M, and see Figure S1Q) (Witter et al.,
1990). However, neurons in both proximal and distal parts of
dSub were weakly labeled by CTB injected into MB or EC5.
Together, these results indicate that dCA1 serves as the main
input structure to dSub and that the majority of dSub neurons
send projections to multiple downstream target structures.

The dSub/EC5 Circuit Bidirectionally Regulates
Episodic Memory Retrieval

To examine the functional role of dSub neurons and their circuits,
we performed optogenetic inhibition experiments using a Cre-
dependent eArch3.0-eYFP virus. During the contextual fear-
conditioning (CFC) paradigm, we confirmed that green light
inhibition of dSub decreased behavior-induced immediate early
gene cFos-positive neurons (Figures S2A–S2L). Inhibition of
dSub neurons during CFC training had no effect on footshock-
long-term memory formation
induced freezing behavior or

1002 Cell 170, 1000–1012, August 24, 2017

(legend on next page)

Cell 170, 1000–1012, August 24, 2017 1003

(Figure 3A). In contrast, dSub inhibition during CFC recall tests
decreased behavioral performance (Figure 3B).
Inhibition of
dSub neurons had no effect on motor behaviors in an open-field
assay (Figure S2M). Inhibition of dSub terminals in medial EC5,
but not in MB, also revealed a memory-retrieval deficit (Fig-
ure 3C). Since the behavioral effect of dSub inhibition in this
mouse line is based on eArch expression in approximately
85% of excitatory neurons in this brain region, we examined
the effect of a more complete inhibition of dSub neurons. Inhibi-
tion of dSub/EC5 terminals in wild-type (WT) mice using
an EF1a-eArch3.0-eYFP virus revealed a greater memory
retrieval deficit (Figure S3 versus Figure 3C). Further, inhibition
of vSub/EC5 terminals showed normal levels of memory recall
(Figure S3).

Conversely, optogenetic activation of ChR2-eYFP-expressing
tests
dSub projections to medial EC5 during CFC recall
increased recall-induced freezing behavior
in the training
context, but not in a neutral context (Figure 3D, and see Fig-
ure S4A). This result indicates that dSub is involved in hippocam-
pal memory retrieval in a context-specific manner. Activation of
dSub/EC5 in mice that did not receive footshocks during
training lacked freezing behavior during the recall test, support-
ing the specificity of increased memory retrieval in CFC-trained
animals. Our interpretation of
these optogenetic activation
experiments is that, in the training context, natural recall cues re-
activate engram cells in all subfields of the hippocampus, like
DG, CA3, and CA1, but also in dSub. When the activity of
dSub projections to EC5 is further increased by ChR2, this leads
to enhanced freezing due to increased activation of dSub
engram cells. On the other hand, in a neutral context lacking
the specific natural recall cues to reactivate dSub engram cells,
the ChR2 activation without engram labeling is not sufficient to
induce memory recall.
In another hippocampus-dependent
memory paradigm, trace fear-conditioning, dSub/EC5 inhibi-
tion impaired memory recall (Figure 3E, and see Figures S4B

and S4C). In contrast, inhibition of dSub/EC5 had no effect
on the recall of a hippocampus-independent memory formed
during delay fear-conditioning (Figures S4D and S4E). Together,
these experiments indicate that the dSub/EC5 circuit regulates
episodic memory retrieval bidirectionally. We confirmed that the
dSub/EC5 projection is also necessary for the retrieval of a
positive-valence, hippocampus-dependent (Raybuck and Lat-
tal, 2014) memory formed in a conditioned place preference
(CPP) paradigm (Figure 3F, and see Figures S2N, S4F, and S4G).

The dSub/MB Circuit Regulates Retrieval-Induced
Stress Hormone Responses

During both CFC training and recall, levels of the stress hormone
corticosterone (CORT) increases in the blood (Figure 3G), which
is believed to be important to prepare the animal for a predicted
immediate danger (Kelley et al., 2009). Given our finding that
dSub neurons are required for memory retrieval, but not memory
formation, we investigated whether the dSub/MB circuit is
involved in retrieval-induced stress hormone responses. Opto-
genetic inhibition of dSub/MB projections following CFC recall,
but not following CFC training, prevented the CORT increase
(Figure 3G, and see Figure S2N). This deficit was specific to
dSub/MB terminal inhibition, since dSub/EC5 terminal inhibi-
tion had no effect. In addition, optogenetic activation of ChR2-
expressing dSub/MB projections following CFC recall
increased CORT levels, revealing a bidirectional regulation of
blood stress hormone levels by the dSub/MB circuit following
fear memory retrieval.
Interestingly, we did not observe
increased CORT levels following CPP memory retrieval (Fig-
ure S4H). From our finding that the dSub/EC5 circuit is crucial
for CPP memory retrieval
(Figure 3F), it is clear that dSub
neurons are activated, and therefore, both downstream EC5
and MB circuits would be activated. The lack of increased
CORT levels following CPP memory retrieval suggests that the
dSub/MB circuit is necessary, but not sufficient to induce

Figure 3. Differential Roles of dSub Projections in Hippocampal Memory Retrieval and Retrieval-Induced Stress Hormone Responses
(A and B) FN1-Cre mice were injected with a Cre-dependent virus expressing eArch3.0-eYFP into dSub. Optogenetic inhibition of dSub neurons during CFC
training had no effect on long-term memory (n = 12 mice per group; A). Inhibition of dSub neurons during CFC recall impaired behavioral performance (n = 12 mice
per group; B). A two-way ANOVA followed by Bonferroni post-hoc tests revealed a behavioral epoch-by-eArch interaction and significant eArch-mediated
attenuation of freezing (A and B: F1,44 = 5.70, p < 0.05, recall). For dSub optogenetic manipulation experiments, injections were targeted to dSub cell bodies, and the extent of virus expression is shown in Figures 1E-1K. (C) Terminal inhibition of dSub projections to EC5 (bottom left), but not MB (bottom right), disrupted CFC memory recall (n = 11 mice per group). A two-way ANOVA followed by Bonferroni post-hoc tests revealed a dSub terminal-by-eArch interaction and significant eArch-mediated attenuation of freezing (F1,40 = 7.63, p < 0.01, dSub/EC5 terminals). (D) FN1-Cre mice were injected with a Cre-dependent virus expressing ChR2-eYFP into dSub. Optogenetic activation of dSub/EC5 terminals during CFC memory recall increased freezing levels (left), which was not observed in a neutral context (middle) or using no-shock mice (right, n = 10 mice per group). (E) Inhibition of dSub/EC5 terminals during trace fear conditioning (TFC) recall decreased tone (Tn)-induced freezing levels (n = 12 mice). A two-way ANOVA followed by Bonferroni post-hoc tests revealed a behavioral epoch-by-eArch interaction and significant eArch-mediated attenuation of freezing (E and Fig- ure S4B: F1,44 = 7.11, p < 0.05, recall). Pre-tone baseline freezing (Pre). Recall-induced freezing levels during individual tone presentations (left), averaged freezing levels during the two light-off tones and the two light-on tones (right). (F) Inhibition of dSub/EC5 terminals during cocaine-induced CPP recall impaired behavioral performance (n = 14 mice per group). Behavioral schedule (top left). Average heatmaps show exploration time during pre-exposure and recall trials (bottom left). Dashed white lines demarcate individual zones in the CPP apparatus. Pre-exposure preference duration (top right) and recall preference duration (bottom right). Saline (S or Sal), cocaine (C or Coc). A two-way ANOVA followed by Bonferroni post-hoc tests revealed a drug group-by-eArch interaction and significant eArch-mediated attenuation of preference duration (F1,52 = 5.16, p < 0.05, cocaine). For CPP training inhibition, see Figure S4F. NS, not significant. (G) Stress hormone: Terminal inhibition of dSub projections to MB, but not EC5, following CFC memory recall tests decreased stress responses as measured by corticosterone (CORT) levels. Optogenetic activation of dSub/MB terminals following CFC memory recall increased CORT levels (n = 10 mice per group). Context (ctx). CORT levels in CPP paradigm are shown in Figure S4H. Unless specified, statistical comparisons are performed using unpaired t tests; *p < 0.05, **p < 0.01, ***p < 0.001. Data are presented as mean ± SEM. 1004 Cell 170, 1000–1012, August 24, 2017 (legend on next page) Cell 170, 1000–1012, August 24, 2017 1005 CORT. These experiments uncovered a neural circuit originating from dSub that regulates stress hormone responses to condi- tioned cues. Heterogeneity of dCA1 Neurons that Project to dSub and EC5 The dCA1 neurons send primary projections directly to medial EC5 or indirectly via dSub (Ding, 2013). We examined whether the same dCA1 neurons send divergent projections to both dSub and EC5 or whether these two circuits involve distinct sub- populations of dCA1 neurons. To test these possibilities, we conducted monosynaptic retrograde tracing by injecting a Cre- dependent helper virus combined with rabies virus expressing mCherry (Wickersham et al., 2007) into dSub of FN1-Cre mice combined with CTB488 injected into medial EC5 (Figures 4A– 4C, and see Figure S1Q). The dCA1 cells revealed a gradient of RV-mCherry with higher intensity labeling in the distal part of dCA1 (i.e., closer to dSub) (Figure 4B) and a gradient of CTB488 with higher intensity labeling in the proximal part of dCA1 (i.e., away from dSub) (Figure 4C). The higher intensity labeling of distal dCA1 by RV-mCherry and proximal dCA1 by CTB488 are consistent with earlier observations (Knierim et al., 2013) and are suggestive of their stronger projections to dSub and EC5, respectively. However, in counting the total number of labeled cells regardless of labeling intensity, we did not find any significant difference in the number of RV-mCherry-positive dCA1 cells nor in the number of CTB488-positive dCA1 cells along the proximal-distal axis (Figure S5J). This suggests that the strength of the projections from dCA1 to dSub and from dCA1 to EC5 are not reflected in the total number of projecting dCA1 cells, but in their labeling intensity. Such differences in pro- jection strength and targets of the distal versus proximal dCA1 cells may provide the basis for their differential roles in behaviors. It has been suggested that proximal and distal dCA1 may play differential roles in memory formation (Nakazawa et al., 2016). Importantly, we observed three neuronal populations distrib- uted throughout the proximal-distal axis of dCA1—namely RV- mCherry-positive dCA1 cells, CTB488-positive dCA1 cells, and double-positive dCA1 cells (Figures 4D–4F)—indicating that dCA1 neurons project collaterally to both dSub and medial EC5 (22%), project to dSub alone (18%), or to medial EC5 alone (23%) (Figure 4G, and see Figures S5A–S5H). A signifi- cant proportion of the remaining dCA1 neurons most likely send primary projections to the deep layers of the lateral EC (LEC5) (Knierim et al., 2013), which we confirmed using CTB retrograde tracing (Figure S5I). Thus, these data demonstrate that, although there are distinct dCA1 subpopulations that proj- ect to either dSub or EC5, a significant proportion of dCA1 neurons projecting to dSub and EC5 are shared between these two circuits. The dCA1/EC5 Circuit Is Crucial for Episodic Memory Encoding Given the selective role of the dSub/EC5 circuit in memory retrieval and our finding that heterogeneous subpopulations of dCA1 neurons project to dSub and medial EC5, respectively, we next investigated the behavioral contributions of the direct dCA1/EC5 circuit. The injection of a Cre-dependent H2B- GFP virus into dCA1 of CA1 pyramidal cell-specific Cre trans- genic mice, TRPC4-Cre (Okuyama et al., 2016), resulted in GFP expression restricted to dCA1 pyramidal cells without any expression in dSub (Figure 4H). Terminal inhibition of CA1 axons at medial EC5 during CFC training impaired long-term memory formation (Figure 4I), whereas the same manipulation during CFC recall had no effect on behavioral performance (Figure 4J). Further, consistent with the role of dSub in CFC recall, terminal inhibition of dCA1/dSub during CFC recall, but not during CFC training, decreased behavioral performance (Figures 4K-4L). Therefore, the direct dCA1/EC5 circuit plays a crucial role in the encoding, but not recall, of CFC long-term memory, Figure 4. Projection from CA1 to EC5 Is Crucial for Encoding, but Not for Retrieval, of Hippocampal Memories (A–C) Retrograde monosynaptic identification of dCA1 neurons projecting to dSub (in FN1-Cre mice) using a Cre-dependent helper virus combined with a rabies virus (RV). The extent of RV-positive dSub cells, which is the starting population for retrograde tracing, is shown in Figure 2B. Simultaneous retrograde monosynaptic identification of dCA1 neurons projecting to EC5 using CTB. DAPI (blue; A), RV-mCherry (red; B), CTB488 (green; C). Representative sagittal sections, dashed white line denotes CA1/CA2 border. (D–F) Higher-magnification images of boxed regions indicated in Figure 4C. (G) Percentage of dCA1 neurons labeled with mCherry (dSub only), CTB488 (EC5 only), or mCherry and CTB double-positive (dSub+EC5, n = 4 mice). Dashed line indicates chance level (6%), calculated from a control experiment (Figures S5A–S5H, and see STAR Methods). One-sample t tests against chance level were performed. (H) Representative sagittal sections of hippocampus from TRPC4-Cre mice showing dCA1 neurons labeled with a Cre-dependent histone H2B-GFP virus (green, bottom) and stained with DAPI (blue, top). (I and J) TRPC4-Cre mice were injected with a Cre-dependent virus expressing eArch3.0-eYFP into dCA1. Terminal inhibition of CA1/EC5 during CFC training impaired long-term memory (n = 10 mice per group; I). Inhibition of CA1/EC5 terminals during CFC recall had no effect on behavioral performance (n = 10 mice per group; J). A two-way ANOVA followed by Bonferroni post-hoc tests revealed a behavioral epoch-by-eArch interaction and significant eArch-mediated attenuation of freezing (I-J: F1,36 = 9.19, p < 0.01, training). (K and L) Terminal inhibition of CA1/dSub during CFC training had no effect on long-term memory (n = 13 mice per group; K). Inhibition of CA1/dSub terminals during CFC recall disrupted behavioral performance (n = 13 mice per group; L). A two-way ANOVA followed by Bonferroni post-hoc tests revealed a behavioral epoch-by-eArch interaction and significant eArch-mediated attenuation of freezing (K-L: F1,48 = 5.16, p < 0.05, recall). (M and N) Memory updating. Experimental schedule (top) for pre-exposure mediated contextual fear conditioning (PECFC) with optogenetic terminal inhibition of CA1/EC5 (using TRPC4-Cre mice; M) and dSub/EC5 (using FN1-Cre mice; N) during the pre-footshock period (left) or footshock period alone (right) on day 2. Freezing levels during recall tests (day 3) to the conditioned context (bottom). eYFP and eArch conditions (n = 12 mice per group). NS, not significant. Immediate shock (Imm. shk). A two-way ANOVA followed by Bonferroni post-hoc tests revealed a behavioral epoch-by-eArch interaction and significant eArch-mediated attenuation of freezing (M: F1,44 = 9.81, p < 0.01, recall in right panel; N: F1,44 = 4.75, p < 0.05, recall in left panel). Unless specified, statistical comparisons are performed using unpaired t tests; *p < 0.05, **p < 0.01. Data are presented as mean ± SEM. 1006 Cell 170, 1000–1012, August 24, 2017