Nature | Vol 587 | 19 November 2020 | 437

Article

Persistent transcriptional programmes are associated with remote memory

Michelle B. Chen1,5, Xian Jiang2,3,5, Stephen R. Quake1,4 ✉ & Thomas C. Südhof2,3 ✉

The role of gene expression during learning and in short-term memories has been studied extensively1–3, but less is known about remote memories, which can persist for a lifetime4. Here we used long-term contextual fear memory as a paradigm to probe the single-cell gene expression landscape that underlies remote memory storage in the medial prefrontal cortex. We found persistent activity-specific transcriptional alterations in diverse populations of neurons that lasted for weeks after fear learning. Out of a vast plasticity-coding space, we identified genes associated with membrane fusion that could have important roles in the maintenance of remote memory. Unexpectedly, astrocytes and microglia also acquired persistent gene expression signatures that were associated with remote memory, suggesting that they actively contribute to memory circuits. The discovery of gene expression programmes associated with remote memory engrams adds an important dimension of activity-dependent cellular states to existing brain taxonomy atlases and sheds light on the elusive mechanisms of remote memory storage.

Long-term memories do not form immediately after learning, but are consolidated over time4. Previous studies have identified important contributions of molecular and cellular processes to learning and mem-ory, such as gene expression changes, cAMP signalling and synaptic plasticity1, and identified a central role for RNA synthesis and protein translation in memory consolidation2. Despite these discoveries, the molecular underpinnings of memory consolidation remain elusive. For instance, while changes in gene expression are found in the first 24 h of learning, it is unclear whether these changes are maintained or whether new changes are acquired to consolidate a long-term memory trace3. Moreover, the dependence on the hippocampus for long-term memory is known to degrade over time, with cortical structures such as the medial prefrontal cortex (mPFC) becoming increasingly impor-tant5. Recently, the development of activity-dependent genetic label-ling tools have allowed identification of sparsely activated neuronal ensembles, enabling access to the molecular mechanisms that underlie experience-dependent connectivity and plasticity6.

Neuron subtypes in remote memory engramsTo identify and study the transcriptional programmes of neurons involved in remote memory, we used TRAP2; Ai14 mice expressing iCre-ERT2 recombinase in an activity-dependent manner along with a iCre-dependent tdTomato (tdT) reporter allele (Extended Data Fig. 1a), enabling us to label memory recall-activated neurons. To genetically label the engrams that are associated with remote mem-ory, we trained mice in a conditioning chamber with three pairs of tone–foot shocks on day 0, and induced fear memory recall (FR) on day 16 by returning mice to the conditioning chamber. Memory recall activates the consolidated memory engrams, thereby labelling the

neuronal ensemble that encodes the consolidated remote memory7, while also inducing memory reconsolidation8. Because the molecular and cellular substrates of consolidation and reconsolidation were indistinguishable in this paradigm, we denote them here collectively as memory consolidation. We used control mice that were not fear conditioned but exposed to the recall context (no fear (NF)), fear conditioned but not subjected to recall (no recall (NR)), and neither fear conditioned nor exposed to the recall condition (homecage (HC)) (Fig. 1a–c). All mice were injected with 4-hydroxytamoxifen (4-OHT) before the FR procedure (or at the equivalent time) to allow activity-dependent production of tdT, thus enabling the distinc-tion between the molecular programmes that are specific to remote memory versus background activation.

Nine days after FR9, single neuronal and non-neuronal cells were col-lected from the mPFC (Extended Data Fig. 1b) via fluorescence-activated cell sorting (FACS) and gating on the tdT signal, followed by plate-based single-cell mRNA sequencing (Fig. 1d). The percentage of TRAPed cells collected via FACS was significantly higher in FR (about 1.5% of all cells) than in other conditions (Extended Data Fig. 2a), further con-firming that the TRAP2 system captured increased neuronal activity during the FR process. We sequenced 3,691 neurons (Snap25 +/tdT + or Snap25 +/tdT − mRNA) and 2,672 non-neuronal cells with high quality and depth (Extended Data Fig. 1c, d). Unbiased transcriptome clustering of cells from all four training conditions allowed the identification of major cell types and confirmed the dominance of neurons among tdT + cells, whereas tdT − cells comprised both neurons and non-neuronal cells (Cldn5 + endothelial, Pdgfra + oligodendrocyte progenitor cells (OPCs), Cx3cr1 + microglia and Aqp4 + astrocytes) (Fig. 1d, e, Extended Data Fig. 1e). Both tdT + and tdT − cells from FR and control groups were represented in all clusters, which suggests that neither the neuronal

https://doi.org/10.1038/s41586-020-2905-5

Received: 26 September 2019

Accepted: 17 August 2020

Published online: 11 November 2020

Check for updates

1Department of Bioengineering, Stanford University, Stanford, CA, USA. 2Department of Molecular and Cellular Physiology, Stanford University, Stanford, CA, USA. 3Howard Hughes Medical Institute, Stanford University School of Medicine, Stanford, CA, USA. 4Chan Zuckerberg Biohub, Stanford, CA, USA. 5These authors contributed equally: Michelle B. Chen, Xian Jiang. ✉e-mail: steve@quake-lab.org; tcs1@stanford.edu

438 | Nature | Vol 587 | 19 November 2020

Article

activation state nor the training paradigm significantly altered funda-mental cell-type identities (Fig. 1e).

Sub-clustering of 3,691 Snap25 + neurons using the top 2,000 highly variable genes revealed seven putative neuron sub-populations—four glutamatergic (C0, C1, C3 and C5) and three GABAergic neuron populations (C2, C4 and C6)—all of which were consistently observed throughout four biological replicates (Fig. 2a–c, Extended Data Fig. 2b). These are molecularly distinct populations (Fig. 2d), with each subtype expressing at least one distinctive marker gene (see Methods; Fig. 2e). All subtypes contained tdT + cells, which indicated activation of all neu-ron subtypes regardless of the training state (Extended Data Fig. 2c). A comparison of key layer-specific marker genes (C0-Dkkl1, C1-Rprm, C2-Calb2, C3-Tesc, C4-Tnfaip8l3, C5-Tshz2 and C6-Lhx6) to existing cortical single-cell expression databases10 confirmed their presence in the mPFC (Extended Data Fig. 2d).

Surprisingly, no significant differences were found in the neuron subtype composition of TRAPed populations between the FR and NF groups (Fig. 2f, Extended Data Fig. 2e), which suggests a lack of training-dependent recruitment of neuron types during consolidation compared to baseline active populations in a NF memory scenario. Both excitatory and inhibitory neuron types were found in active FR popula-tions, with glutamatergic cells comprising about 60–70%. Within the same FR brains, active and inactive populations had roughly similar compositions of neuron subtypes, with the exception of C2-Calb2 and C3-Tesc, suggesting only slight shifts in the recruitment or retirement of neuron subtypes due to activity.

Memory-associated gene expressionTo determine whether remote-memory-associated transcriptional changes occur in recall-activated neurons, we looked for differentially expressed genes (DEGs; log2 fold change (log2FC) > 0.3 and false discov-ery rate (FDR) < 0.01) in TRAPed FR versus NF cells (Fig. 3a). Single-cell resolution enables a comparison of neurons within the same subtype and the identification of genes that are specifically associated with memory consolidation and recall. Of 23,355 genes, 1,292 were found

to be consolidation-dependent. Expression patterns indicated an overall transcriptional activation, with more genes upregulated than downregulated. Interestingly, DEGs were heterogenous across neuron subtypes, which suggests that remote memory consolidation involves subtype-specific transcriptional programmes (Fig. 3b).

We applied a set of strict criteria to identify possible effector genes. First, each DEG had to be differentially expressed in at least three-quarters of biological replicates, enforcing reproducibility. The removal of DEGs that are also differentially expressed between the inactive populations in FR versus NF mice allowed the identification of changes that were specific to active populations (Extended Data Fig. 3a). Next, DEGs must be differentially expressed when FR cells are compared to NR and HC controls, ensuring that DEGs are not just a con-sequence of a fear experience. Last, DEGs had to pass a permutation test with shuffled labels (Extended Data Fig. 3b). These criteria produced a set of 99 ‘remote-memory-associated DEGs’ (Fig. 3c; see Methods). Several genes encoded proteins with regulatory roles, including known regulators of transcription (Hmg20a, Hnrnpk and Zfp706) and transla-tion (Nck2, Alpl1 and Eif2ak1). Interestingly, even among the condensed list of remote-memory-associated DEGs, we found strong enrichments in genes encoding proteins involved in vesicle exocytosis (Vamp2, Gdi2, Rab15, Rab5a, Rab24, Atp6v0c, Syt13, Stx1b and Nsf), transmembrane transport (Slc30a9, Slc25a46, Mfsd14a, Tmem50a, Gpm6a, Mfsd14b and Abcf3), dendritic spine organization (Strip1, Pls3 and Gsk3b) and long-range intracellular transport (Timm29, Atad1, Pak1, Plehkb2, Sarnp, Rtn3, Dmtn, Sar1a and Hid1) (Fig. 3c, Extended Data Fig. 3b, c). More than half of the remote-memory-associated DEGs are associated with neuronal diseases, suggesting links between the functional role of these genes to various memory-affecting neuronal disorders, in addition to the regulation of remote memory.

We further investigated the specificity of our findings by analysing TRAPed neurons that were activated by a salient experience unrelated to fear memory (Extended Data Fig. 4a). Using food deprivation as a sali-ence signal, we identified TRAPed neuronal ensembles that contained the same neuronal subtypes as in TRAPed FR ensembles, but with dif-ferences in the subtype composition ratios (Extended Data Fig. 4b–d).

16 d 25 d

Return +4-OHT

Fear training

HC

scRNA-seq

NF

NR

FR

0 dNovel environment

Foot shock

4-OHT injection

Collect single cells

tdT

Encoding Consolidationa bTRAP

c

NFNRFRHC0 5

ln(cpm + 1)Snap25

tSNE_1

tSN

E_2

e

ACC

d

1 100 10,000Hoescht

log 10

[td

T (A

U)]

Gate1

2

Pos

Neg

3

41.00 0.5 1.5

log2 [tdT (cpm)]

FACS gating

AstrocyteEndothelialMicrogliaNeuronOPC

0 20 40 60 80100

Pos

Neg

Composition (%)

Fear + recall

NF FR0

20

40

60

80

100

Free

zing

(%)

***

Training condition

Fig. 1 | Labelling and collection of single memory engram cells via the TRAP2; Ai14 line. a, The experimental paradigm includes generating remote fear-memory traces via contextual fear conditioning, isolating TRAP+-activated neurons via flow cytometry and plate-based single-cell RNA sequencing (scRNA-seq). b, Representative image of tdT + TRAPed (red) cells in the anterior cingulate cortex (ACC) region of the mPFC 9 days after an injection of 4-OHT (at the time of remote memory recall). Scale bars, 1 mm (0.5 mm for zoomed in images). c, Degree of freezing upon return to the novel context 16 days after fear conditioning (FR) or no conditioning (NF) (n = 4 mice per condition,

***P = 4.89 × 10−9; mean ± s.d.). d, Representative flow gating for tdT + TRAPed cells (about 1.5% of events in FR) in one mouse from the FR condition (left). Post-sequencing analysis of 722 example cells from a representative FR brain shows enrichment of tdT mRNA (pos) in the positive sort gate (orange violin) (middle). Enrichment of neuronal cells in the positive gate and neuronal and non-neuronal cell types in the negative gate in all brains (right). AU, arbitrary units; cpm, counts per million. e, All sorted cells, with neuronal cells identified via the expression of Snap25 mRNA. All training and control conditions are represented in all cell clusters.

Nature | Vol 587 | 19 November 2020 | 439

While we found a total of 143 DEGs (FDR < 0.01) when comparing sali-ence to no salience groups, there was almost no intersection of these DEGs with those found in FR mice (Extended Data Figs. 4e, 5). Thus, while new transcriptional programmes are activated in salience ensem-bles, the nature of these molecular changes is experience-specific and probably modulated by the particular valence of and/or functional requirements arising from the experience.

Hierarchical clustering of TRAPed FR neurons by the expression levels for each remote-memory-associated DEG allowed distinct populations of ‘highly activated’ and ‘lowly activated’ cells to emerge, suggesting that different transcriptional modules are concertedly regulated during memory consolidation in each neuronal subtype (Fig. 3d). To determine the subtype specificity of these modules, we found that the fraction of cells activated with the subtype-specific DEGs was generally highest in the corresponding subtype when compared to the activation levels in other subtypes or in the inactive populations (Fig. 3e, Extended Data Fig. 6a). Together, this could indicate the presence of subtype-specific common regulatory elements.

To address this possibility, we analysed our DEGs using hypergeomet-ric optimization of motif enrichment (HOMER) to search for common regulatory motifs in an unbiased manner (see Methods; Extended Data Fig. 6b). We found 12 putative de novo and two known motifs enriched within our target DEG set (P < 0.01). While we did not find significant enrichment of motifs within subtype-specific DEGs, the Hif1b binding motif was found in >40% of total DEGs, including the synaptic transmis-sion and plasticity-related genes Rab5a, Rab24, Vamp2, Gdi2, Gpm6a, Strip1, Ptp4a1, Trim32, Mfsd14a, Mfsd14b and Slc25a46. Interestingly, these findings agree with recent studies indicating a potential dual role for HIF1 transcription factors during hippocampal-dependent spatial learning and consolidation under normoxic conditions11. Interestingly, motifs associated with Creb, Nfkb, Cbp and C/ebp—canonical transcrip-tional regulators of neuronal activity and plasticity12,13—were absent near the transcription start site (−400 to +100 bp).

Vesicle exocytosis signatures in memoryTo further elucidate the significance of these remote-memory-dependent transcriptional programmes, we used STRING to look for known and predicted protein–protein interactions. K-means clustering of the gene nodes revealed a significantly connected network (P = 1.75 × 10−6) that was centred around a large cluster of genes related to vesicle-mediated transport, exocytosis and neurotransmitter secretion, all of which were highly connected (confidence = 0.4; Extended Data Fig. 6c). Remark-ably, 20 out of 99 remote-memory-associated DEGs fell within these functional categories, including Stx1b, Syt13, Vamp2, the SNAP receptor (SNARE) ATPase (Nsf) and the GTPase Rab5a, all of which are function-ally linked to the SNARE complex and to vesicle exocytosis (Fig. 4a). Interestingly, the two most highly and ubiquitously upregulated genes across subtypes were Serinc1 and Serinc3, which are thought to be serine incorporators14. Notably, phosphatidylserine phospholipids are calcium-dependent binding partners for synaptotagmins15, sug-gesting that Serinc1 and Serinc3 may have important roles in enhanc-ing phosphatidylserine levels and vesicle membrane fusion during memory consolidation. Finally, in situ hybridization confirmed the endogenous proportions of neuronal subtypes in TRAPed populations (Extended Data Fig. 7a, b), as well as the upregulated expression of key remote-memory-associated DEGs, including Serinc3, Syt13, Vamp2 and Stx1b in respective neuronal subtypes (Fig. 4b, c, Extended Data Fig. 7c).

Non-neuronal gene expression changesRemarkably, we discovered that non-neuronal cells also exhibited tran-scriptional changes associated with remote memory consolidation (FR compared to NF mice; Fig. 5a, b, Extended Data Fig. 8a, b). These signa-tures were distinct from those of neurons, indicating that non-neuronal programmes may support maintenance of the remote fear-memory trace. Surprisingly, >95% of these DEGs were upregulated, which sug-gests an overall transcriptional activation during consolidation. Not only was this response detectable weeks after the initial learning but it was observed even without enrichment of the non-neuronal cells directly associated with the TRAPed engram cells (that is, the TRAP method is neuron-specific).

Astrocytes and microglia showed the greatest number of tran-scriptional changes, with 181 and 308 genes perturbed, respectively (log2FC > 1 and FDR < 0.01) (Fig. 5c). Most of these DEGs represent largely diverging pathways (Fig. 5d). In particular, upregulated astro-cytic genes were enriched in lipid, cholesterol and steroid metabolic functions (Gja1, Hmgcr, Dhcr7, Insig1, Acsl3, Idi1, Acsbg1, 10Asah1 and Hacd3) as well as glucose transport (Abcc5, Slc39a1, Slc6a1, Slc27a1, Slco1c1, Gnb1 and Ttyh1). Enhanced metabolic support from astrocytes may be required during memory consolidation since astrocyte–neu-ron metabolic coupling is elevated during neuronal activity16. Moreo-ver, 95 out of 181 astrocyte DEGs were reproduced when comparing

0

1

2

3

4

5

6

tSNE_1

tSN

E_2

–1

0

1

Glut–GABAsignature

tSNE_1

tSN

E_2

FR: inactiveFR: TRAPNF: TRAP

a b

c

d

Dkkl1

Rprm

Calb2

Tesc

Tnfaip8l3

Tshz2

Lhx6

010

010

010

010

010

010

010

log 2c

pm

C0 C1 C2 C3 C4 C5 C6

Glut

Glut

Glut

Glut

GABA

GABA

GABA

C0 C1 C2 C3 C4 C5 C60

0.1

0.2

0.3

0.4

0.5

Rat

io o

f neu

ron

sub

typ

e in

the

TRA

Ped

or

inac

tive

pop

ulat

ion

*P = 0.02*P = 0.01

Neuron subtype

Fig. 2 | Molecular identification of active neurons during remote memory consolidation. a, Dimensional reduction of all Snap25 + neurons (n = 3,691 cells) reveals seven distinct neuronal subtypes (C0–C6). b, Neuronal subtypes fall into two distinct categories: excitatory (glutamatergic) and inhibitory (GABAergic). The Glut–GABA signature is calculated based on the difference of the scaled expression level of Gad1 and Slc17a7. c, Expression levels of the top marker genes for each neuron subtype (C0–C6). d, Differences in the composition of neuron subtypes of TRAPed populations in FR and NF conditions, as well as inactive populations in FR mice (1 mouse per condition per replicate, n = 5 replicates, two-sided t-test) in the composition of TRAPed populations between FR and NF conditions are found. FR TRAPed populations are composed of significantly more C2 (GABAergic) neurons and less C3 (glutamatergic) neurons than the inactive population (error bars indicate s.e.m.).

440 | Nature | Vol 587 | 19 November 2020

Article

FR to NR mice, suggesting that a large portion of DEGs is specific to the recall experience itself and not merely a remnant of the fear experience.

By contrast, DEGs from microglial cells were enriched in innate immunity (Il6r, Stat6, Csf3r, Il1a, Irf5, Cd86, Tnfrsf1b, Ywhaz, Litaf, Ptgs1, Gdi2 and Rnf13) and cytoskeletal reorganization/focal adhe-sion maintenance pathways (Cdc42, Rhoa, Rhoh, Prkcd, Vasp, Arf6,

Vav1 and Actr2), suggesting that upregulation of specific inflammatory molecules and enhancement of cell migration may be involved in the maintenance of memory. While less is known regarding the immu-nomodulatory roles of microglia in memory and learning, previous studies have shown that low levels of inflammatory cytokines (such as IL-1, IL-6 and tumour necrosis factor) can regulate neuronal circuit remodelling and long-term potentiation17.

log2FC(DE in 3/4 rep)

–4

–2

NS

2

4

No

Yes

Hid1Kctd10Pdha1PigqSkiv2l2Abcf3Miga2MmdSar1aEif2ak1Acsf3Cdc42se2Fam134aVamp2Gsk3bHnrnph2Pja2SdhaGdi2Rab15Fam131aGfra2Pip4k2cNcdnUsp5Nell2Tmem151aDmtnRtn3Pcsk2PfkmTrim32Mfsd14bRab5aTdgEmc1Gpm6aElmod1CycsSarnpRab24Erp29GhitmZfp706Sult4a1AppCckEmc4Psmb6Atp6v0bHint1Guk1Rtn1ClppCrip2Mpc1Atp6v0cAtp5g3Tmem50aPlekhb2Syt13Garnl3Dpysl4Aplp1HnrnpkNsfMfsd14aMir124-2hgNck2Stx1bPak1Slc25a46Itfg1Lmbrd1Tmx1DnerAtad1Ankrd45Timm29Vopp1Pls3Hmg20aCtbp1Strip1Cdv3Inpp5fPrkar1bSlc30a9Alg2Trim35Hacd3Serinc1Serinc3Ptp4a1

Neu

rona

l dis

ease

Met

abol

ic p

roce

ss

Mem

bran

e tr

ansp

ort

Pro

tein

tran

spor

t

Den

driti

c sp

ine

dens

ity

Vesi

cle

exoc

ytos

is

Nuc

leic

aci

d bi

ndin

g

Tran

slat

ion

regl

uatio

n

Spl

icin

g re

gula

tion

Tran

scrip

tion

regu

latio

n

Tran

scrip

tion

fact

or

Sec

rete

dM

embr

ane

C0

C1

C3

C2

C4

EXH INH

Remote-memory-associated DEGsTRAPed FR vs NF

d

TRAPed cells in C0

High MediumLow

c

0

10

20

30

40

log2FC

Num

ber

of D

EG

s

Neuronsubtype

C0C1C2C3C4C5C6

485 158 202 10 173 69 94

308 129 9 101 51 49

702 10 202 67 111

30 9 7 5

433 63 102

118 43

179

C0

C1

C3

C5

C2

C4

C6

C0

C1

C3

C5

C2

C4

C6b

a

0

0.2

0.4

0.6

C0 C0 C1 C2 C3 C4

Frac

tion

of c

ells

activ

ated

with

C0

DE

Gs Activation criteria

(DEGs passed threshold)25%50%75%

C0

e

NF FR

Threshold = 90th percentile of NF distribution

TRAP Inactive

6 120log2cpm

Percentile of expression

0 100

5 6

0 10,000

–log10(number of reads)

Number of genes expressed

TRAPed cells in C0Serinc1Rtn3Gpm6aNsfTrim32Usp5Hacd3SdhaVamp2Rab5aPja2Serinc3NcdnMfsd14bDmtnPfkmRab15Hnrnph2Gdi2Fam131aTmem151aNell2Gfra2Emc1Pip4k2cTdgPcsk2Ptp4a1Gsk3b

DE

Gs

in C

0

DEGs in C0: TRAPed FR vs NF

FDR < 0.01, |log2FC| > 0.3F|

Pcna-ps2Ubc

Hsp90aa1Vamp2

Serinc3

Pja2Gsk3b

Serinc1Hnrnph2Pcsk2

CckGpm6a Hacd3

Trim32Ptp4a1Nell2Gdi2

Rab5aSdhaMfsd14b

Rtn3Nsf

TdgFam131aNcdn

PfkmGfra2

Rab15Usp5

Pip4k2cDmtn

Emc4

Tmem151aEmc10

5

10

15

–2 –1 0

–2 0 2

1 2log2FC

–log

10(F

DR

)

Number of intersecting DEGs

Cellua

r

com

ponen

t

Mole

cular

func

tion

Mala

card

s

GO

Fig. 3 | Transcriptional programmes activated by consolidation of remote memories are distinct across neuron subtypes. a, DEGs in FR versus NF in the C0 neuron subtype (n = 126 cells (NF), n = 289 cells (FR)) with all replicated pooled. Differential expression is defined by FDR < 0.01 and abs(log2FC) > 0.3 (red points) via a two-sided Mann–Whitney test. Remote-memory-associated DEGs (which remain differentially expressed in three-quarters or more of replicates and when FR is compared to the NR and HC groups (see Extended Data Fig. 3)) are labelled in black. b, The number of DEGs per neuron subtype (left) and the number of shared DEGs between each neuron subtype (using pooled cells) (right). c, log2FC of remote-memory-associated DEGs (FR versus

NF) per neuron subtype. Each gene is further annotated with potential biological functions. DE, differentially expressed; EXH, excitatory; GO, Gene Ontology; INH, inhibitory; NS, not significant. Skiv2l2 is also known as Mtrex, and Fam143a is also known as Retreq2. d, The percentile in which a TRAPed neuron (from C0) lies in the distribution of expression of a C0 DEG for all TRAPed C0 cells. The box plots (median ± s.d.) show the log2cpm distribution for each C0 DEG. Hierarchical clustering reveals one common C0 transcriptional programme that is concertedly upregulated. e, The fraction of cells in each neuron subtype that is activated with the transcriptional programme (DEGs) from C0.

Nature | Vol 587 | 19 November 2020 | 441

In addition to neuron–neuron coupling, communication pro-grammes between neurons and non-neuronal cells may support the memory trace over long periods. We looked for the expression

of receptors or ligands in non-neuronal cells whose known binding partner18 is perturbed in TRAPed FR neurons (Extended Fig. 8c–e). We focused on genes that were differentially expressed in both the ligand-bound and the receptor-bound cell type (Extended Data Fig. 8c). In FR mice, we found upregulation of neuronal neuroligin-1 and neuroligin-3 (encoded by Nlgn1 and Nlgn3, respectively) and its binding partner neurexin-1 (encoded by Nrxn1) on astrocytes, com-plexes that may enhance neuron–glia adhesions and modulate synap-tic function19. Thus, the concerted upregulation of these binding pairs in FR mice strongly suggests a role for astrocyte–neurexin–neuroligin

Rep

licat

e m1m2m3m4

log 10

(FD

R) 0

246

a

0

10

20

30 Syt13

0

5

10

15

0

5

10

15

20

FR NF

Inte

grat

ed in

tens

ity

Serinc3

***P = 2 × 10–8

***P = 1 × 10–10

**P = 0.003

Vamp2

0369

02.55.07.5

10.012.5

02.55.07.5

10.0

0

2.5

5.0

7.5

10.0

02.55.07.5

10.012.5

0

3

6

9

0369

12

1

2

3

00.51.01.52.0

0

1

0123

01234

0 1 3 2 4Glut GABA

Ser

inc3

Ser

inc1

Rab

5a

FR NF

log2cpmavg

log2FC

Stx

1bVa

mp

2S

yt13

Nsf

00.51.01.52.0

01234

c

0

1

2

3

4 ***P = 4 × 10–7

Stx1b

b

FR NF

FR NF

FR NF

Inte

grat

ed in

tens

ityIn

tegr

ated

inte

nsity

Inte

grat

ed in

tens

ity

Ser

inc3

tdT

Dkk

l1D

AP

IM

erge

NF FR

FR/N

F

FR/N

F (In

)

FR/N

R

Fig. 4 | Remote memory consolidation is associated with specific markers for vesicle exocytosis. a, A subset of remote-memory-associated DEGs that are known to regulate vesicle membrane fusion and exocytosis at the presynaptic terminal. The violin plots are overlayed with a drumstick plot indicating the average expression per mouse. The red points represent the median. The bubble plots depict the log2FC and the degree of significance (FDR) for each replicate and are coloured by such. In addition to the FR/NF log2FC (first column), DEGs were also confirmed to be differentially expressed when compared to NR (second column), and their activation is specific only to the active (inactive (In)) (last column). The red dots in coloured panels indicate which neuronal subtype these particular genes are upregulated in. b, Representative in situ images of Serinc3 expression (purple) in Dkkl1+/tdT+/DAPI+ cells in the mPFC. Scale bars, 100 μm. c, In situ validation of key genes involved in vesicle exocytosis in various neuron subtypes, including Serinc3 (in the Dkkl1+ subtype, n = 268 (FR), n = 62 (NF) cells), Stx1b (in the Rprm+ subtype, n = 342 (FR), n = 144 (NF) cells), Syt13 (in the Tnfai8lp+ subtype, n = 244 (FR), n = 44 (NF) cells) and Vamp2 (in the Tesc+ subtype, n = 326 (FR), n = 292 (NF) cells). Each point represents the normalized integrated intensity of the probe per cell analysed.

Oligo

OPC

EC

Astrocyte

Microglia

a b

tSNE_1

tSN

E_2

Oligo

EC

OPC

Astrocyte

Microglia

NS24

Mean log2FC (FR/NF)

0 100 200 300Number of DEGs

(FR/NF)

Rnf13

Lag3V

av1Tm

em86a

Gd

i2C

sf3rIrf5P

tgs1LitafR

nf130

Tnfrsf1b

Stat6

Il1aY

whaz

Il6ra

Cd

86

Bhlhe40

Rhoa

Gsk3b

Hnrnp

kN

ptn

Lrrc8aA

rglu1A

sah1B

mp

2kS

pry2

Slc7a10

Insig1S

lc27a1D

hcr7S

lco1c1H

mgcr

Vcam

1

Gja1

Cholesterol biosynthetic processLipid metabolic process

Glycosaminoglycan metabolismNervous system development

Metal ion transportRegulation of alternative mRNA splicing

Cell adhesionPositive regulation of GTPase activity

Response to cytokineIntracellular signal transduction

Response to lipopolysaccharideActin cytoskeleton organization

Regulation of endocytosis

AstrocyteMicroglia

Score

Pathway enrichment in FR-associated glial cells

Microglia d

Bmp2k

Gm10800Gdi2

Rnf13Rap1b

Tcirg1Vav1

Slc39a1P2ry13 Cmtm6

Lag3Csf3r

Ptgs1

Hnrnpf

Ppp2ca

Asah1Efhd2

Fgd2

Eif4a1

Unc93b1Daglb

Alkbh5 Arglu1Tagap

Arrb2

Dpysl2

0

2.5

5.0

7.5

10.0

12.5

–4 –2 0 2 4log2FC

–log

10(F

DR

)

0 10 20 30

–2–4

cArglu1

Spry2

Gja1 Lrrc8aZnrf3

Gm12940Dpysl2

Sfxn5 Bhlhe41Lgr4

Insig1Slitrk2Slc39a1

LfngPdcd4Msmo1

Zscan26Nfia

Gnb1

Asah1Notch2

Nptn

Mfge8

Dhrs3Msi1

Sema4b

Bmp2k

Paqr7Gria2Gjb6

0

5

10

15

–2 0 2 4log2FC

–log

10(F

DR

)

Astrocyte

Fig. 5 | Transcriptomic changes in non-neuronal cells associated with remote memory consolidation. a, Sub-clustering of non-neuronal cells reveals five non-neuronal cell types (astrocyte, endothelial (EC), microglia, OPC and oligodendrocyte (oligo)) that were collected in an unbiased manner through tdT-negative flow cytometry gates. b, DEGs in non-neuronal cells (FR versus NF) (left). DEGs are defined as log2FC > 1 and FDR < 0.01. The number of DEGs that satisfy these criteria in each non-neuronal cell type is also shown (right). The top DEGs (FR versus NF) for glial cells (astrocytes and microglia)

that are also differentially expressed in FR versus NR are labelled. c, DEGs (determined by two-sided Mann–Whitney test) that are upregulated and downregulated in astrocytes (left) and microglia (right) in FR versus NF mice. DEGs (FDR < 0.01, log2FC > 1) are indicated by red dots, and the top DEGs are labelled in black. d, Pathway analysis of the DEGs (FR versus NF) in microglia and astrocytes. The score is defined as the –log2(P value) using the GeneAnalytics software.

442 | Nature | Vol 587 | 19 November 2020

Articleinteractions in the maintenance of synaptic strength during fear memory storage.

DiscussionWhile high-resolution gene expression atlases of the brain have provided invaluable information about cellular taxonomy10,20, char-acterization of activity-dependent states within these cell types is necessary to understand how experience modulates gene expres-sion, synaptic plasticity and neuronal circuitry. The ability to form and maintain unique synaptic connections that encode a particular memory out of a vast pool of other experiences requires a complex coding space. By using a combination of activity-dependent label-ling of neurons and single-cell transcriptomics, we discovered that (1) all mPFC neuron types can be activated during consolidation of remote memory via heterogenous transcriptional programmes; (2) enhanced membrane fusion and vesicle exocytosis may be a critical mode of synaptic strengthening during memory consolidation; (3) a specific set of exocytosis-related genes out of a vast coding space may be involved in allowing highly unique, experience-specific con-nections to be made; (4) these particular transcriptional programmes are detectable at remote time points and thus are probably involved in maintaining the memory trace weeks after learning; and (5) consoli-dation of remote memory also induces a persistent transcriptional programme in astrocytes and microglia.

Deciphering the temporal evolution of engram populations and their associated gene programmes through the various stages of initial learning, recent memory and remote memory is crucial for understanding the basis of conversion of short-term memo-ries to long-term memories. We found that the majority of gene programmes affected in activated neurons during early stages of learning21–23 and recent memory24 do not intersect with our remote-memory-associated DEGs (Extended Data Fig. 9a), nor with genes enriched in TRAPed FR populations over inactive ones (Extended Data Fig. 9b). This suggests that remote memory could be governed by temporally unique transcriptional programmes. However, future experiments using unified technologies to decon-volve the neuronal compositions of recent and remote engrams and identify the immediate transcriptional changes in recent memory will be of great importance. Relevant to this point, TRAPed neu-rons in NR mice also exhibited continuous transcriptional changes at moderate levels (when compared to NF mice) (Extended Data Fig. 10). However, these DEGs are largely non-intersecting with remote-memory-associated DEGs, which suggests that the experi-ence of fear itself can induce long-lasting changes in gene expression programmes and that the process of recall induces new transcrip-tional programmes in a different set of neurons. The current data therefore provide the first step towards deciphering the transcrip-tional coding landscape that is specifically associated with remote memory consolidation.

Online contentAny methods, additional references, Nature Research reporting sum-maries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author con-tributions and competing interests; and statements of data and code availability are available at https://doi.org/10.1038/s41586-020-2905-5.

1. Kandel, E. R. The molecular biology of memory storage: a dialogue between genes and synapses. Science 294, 1030–1038 (2001).

2. Flexner, L. B. & Flexner, J. B. Effect of acetoxycycloheximide and of an acetoxycycloheximide–puromycin mixture on cerebral protein synthesis and memory in mice. Proc. Natl Acad. Sci. USA 55, 369–374 (1966).

3. Alberini, C. M. & Kandel, E. R. The regulation of transcription in memory consolidation. Cold Spring Harb. Perspect. Biol. 7, a021741 (2014).

4. Squire, L. R. Mechanisms of memory. Science 232, 1612–1619 (1986).5. Kitamura, T. et al. Engrams and circuits crucial for systems consolidation of a memory.

Science 356, 73–78 (2017).6. DeNardo, L. & Luo, L. Genetic strategies to access activated neurons. Curr. Opin.

Neurobiol. 45, 121–129 (2017).7. DeNardo, L. A. et al. Temporal evolution of cortical ensembles promoting remote

memory retrieval. Nat. Neurosci. 22, 460–469 (2019).8. McKenzie, S. & Eichenbaum, H. Consolidation and reconsolidation: two lives of

memories? Neuron 71, 224–233 (2011).9. Guenthner, C. J., Miyamichi, K., Yang, H. H., Heller, H. C. & Luo, L. Permanent genetic

access to transiently active neurons via TRAP: targeted recombination in active populations. Neuron 78, 773–784 (2013).

10. Zeisel, A. et al. Cell types in the mouse cortex and hippocampus revealed by single-cell RNA-seq. Science 347, 1138–1142 (2015).

11. O’Sullivan, C. N., Sheridan, G. & Murphy, K. in Transcription Factors CREB and NF-κB: Involvement in Synaptic Plasticity and Memory Formation 43–65 (Bentham Science, 2012).

12. Suzuki, A. et al. Upregulation of CREB-mediated transcription enhances both short- and long-term memory. J. Neurosci. 31, 8786–8802 (2011).

13. Kida, S. et al. CREB required for the stability of new and reactivated fear memories. Nat. Neurosci. 5, 348–355 (2002).

14. Inuzuka, M., Hayakawa, M. & Ingi, T. Serinc, an activity-regulated protein family, incorporates serine into membrane lipid synthesis. J. Biol. Chem. 280, 35776–35783 (2005).

15. Zhang, X., Rizo, J. & Südhof, T. C. Mechanism of phospholipid binding by the C2A-domain of synaptotagmin I. Biochemistry 37, 12395–12403 (1998).

16. Bélanger, M., Allaman, I. & Magistretti, P. J. Brain energy metabolism: focus on astrocyte–neuron metabolic cooperation. Cell Metab. 14, 724–738 (2011).

17. Williamson, L. L., Sholar, P. W., Mistry, R. S., Smith, S. H. & Bilbo, S. D. Microglia and memory: modulation by early-life infection. J. Neurosci. 31, 15511–15521 (2011).

18. Ramilowski, J. A. et al. A draft network of ligand–receptor-mediated multicellular signalling in human. Nat. Commun. 6, 7866 (2015).

19. Südhof, T. C. Synaptic neurexin complexes: a molecular code for the logic of neural circuits. Cell 171, 745–769 (2017).

20. Saunders, A. et al. Molecular diversity and specializations among the cells of the adult mouse brain. Cell 174, 1015–1030.e16 (2018).

21. Lacar, B. et al. Nuclear RNA-seq of single neurons reveals molecular signatures of activation. Nat. Commun. 7, 11022 (2016).

22. Cho, J.-H., Huang, B. S. & Gray, J. M. RNA sequencing from neural ensembles activated during fear conditioning in the mouse temporal association cortex. Sci. Rep. 6, 31753 (2016).

23. Hrvatin, S. et al. Single-cell analysis of experience-dependent transcriptomic states in the mouse visual cortex. Nat. Neurosci. 21, 120–129 (2018).

24. Rao-Ruiz, P. et al. Engram-specific transcriptome profiling of contextual memory consolidation. Nat. Commun. 10, 2232 (2019).

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

© The Author(s), under exclusive licence to Springer Nature Limited 2020

Methods

MiceAll animal experiments were conducted following protocols approved by the Administrative Panel on Laboratory Animal Care at Stanford University. The TRAP2; Ai14 mouse line was kindly gifted by the Luo laboratory at Stanford. TRAP2 mice were heterozygous for the Fos2A-iCreER allele, and homozygous for Ai14, and were bred with Ai14 homozygous mice in the C57BL/6 background. Mice were group-housed (maximum five mice per cage) on a 12 h light–dark cycle (07:00 to 19:00, light) with food and water freely available. Male mice 42–49 days of age were used for all the experiments. Mice were handled daily for 3 days before their first behavioural experiment.

GenotypingThe following primers: GAG GGA CTA CCT CCT GTA CC (forward) and TGC CCA GAG TCA TCC TTG GC (reverse) were used for genotyping of the Fos2A-iCreER allele.

Fear conditioningThe fear conditioning training was performed as previously described25. Briefly, mice were individually placed in the fear conditioning chamber (Coulbourn Instruments) located in the centre of a sound attenuating cubicle, which was cleaned with 10% ethanol to provide a background odour. A ventilation fan provided a background noise at approxi-mately 55 dB. After a 2-min exploration period, three tone–foot shock pairings separated by 1-min intervals were provided. The 85 dB 2-kHz tone lasted for 30 s, and the foot shocks were at 0.75 mA and lasted for 2 s. The foot shocks were co-terminated with the tone. The mice remained in the training chamber for another 60 s before being returned to the home cages. For the context recall, mice were placed back into the original conditioning chamber for 5 min 16 days after the training. 4-OHT injections were performed immediately (within 30 min) before the recall experiments. For the HC and the NR groups, 4-OHT was injected at a similar time when the other two groups were subjected to recall. The behaviour of the mice was recorded and ana-lysed with the FreezeFrame software (v.4; Coulbourn Instruments). Motionless bouts that lasted more than 1 s were considered as freeze. Data were analysed with the tracking software Viewer III (Biobserve).

Food deprivationMice were deprived of food for 16 h, and then 4-OHT was injected to the animals and food was returned to one group (salience group) imme-diately afterwards (within 30 min), while no food was returned to the other group (no salience group) until 10 h later.

tdT florescence examinationMice were deep anaesthetized with tribromoethanol and perfused with PBS followed by fixative (4% paraformaldehyde diluted in PBS). The brains were then removed and postfixed in 4 °C overnight and immersed in 30% sucrose solution for 2 days before being sectioned at a thickness of 50 μm on a cryostat (CM3050 S, Leica Biosystems). Imag-ing was performed with a scanning microscope (BX61VS, Olympus).

Single-cell dissociation and flow cytometrymPFC regions were microdissected from live vibratome sections (300 μm thick) of the prefrontal cortex. Tissue pieces were enzymatically disso-ciated via a papain-based digestion system (LK003150, Worthington). Briefly, tissue chunks were incubated in 1 ml of papain (containing l-cysteine and EDTA), DNase and kynurenic acid for 1 h at 37 °C and 5% CO2. After 10 min of incubation, tissues were triturated briefly with a P1000 wide bore pipette tip and returned. Cells were triturated another four times (around 30 each) with a P200 pipette tip over the rest of the remaining incubation time. At room temperature, cell suspensions were centrifuged at 350g for 10 min, resuspended in 1 ml EBSS with 10% v/v

ovomucoid inhibitor, 4.5% v/v DNase and 0.1% v/v kynurenic acid, and centrifuged again. Supernatant was removed and 1 ml ACSF was added to cells. ACSF was composed of: 1 mM KCl, 7 mM MgCl2, 0.5 mM CaCl2, 1.3 mM NaH2PO4, 110 mM choline chloride, 24 mM NaHCO3, 1.3 mM Na ascorbate, 20 mM glucose and 0.6 mM sodium pyruvate. Cells were passed through a 70-μm cell strainer to remove debris. Hoechst stain was added (1:2,000; H3570, Life Technologies) and incubated in the dark at 4 °C for 10 min. Samples were centrifuged (350g for 8 min at 4 °C) and resuspended in 0.5 ml of ACSF and kept on ice for flow cytometry.

Cells were sorted via the Sony SH800 into 96-well or 384-well plates (Bio-Rad) directly into lysis buffer26 with oligodT, and immediately snap frozen until processing. A positive ‘TRAP’ gate was set for cells that were both Hoechst+ and tdT +. A negative ‘TRAP’ gate was set for all Hoechst+ and tdT − cells in general. No gating on forward or backscatter was used to avoid size biases that might be present in a heterogenous neuronal population. Each plate was kept on the sorter for <25 min to prevent evaporation.

SequencingWhole-cell lysis, first-strand synthesis and cDNA synthesis were per-formed using the Smart-seq-2 protocol as described previously26 in both 96-well and 384-well formats, with some modifications. After cDNA amplification (23 cycles), cDNA concentrations were determined via capillary electrophoresis (96-well format) or the PicoGreen quantita-tion assay (384-well format), and wells were chosen to improve quality and reduce cost of sequencing. Only wells with >0.2 ng μl−1 of cDNA were selected and cDNA concentrations were subsequently normalized to ~0.2 ng μl−1 per sample, using the TPPLabtech Mosquito HTS and Mantis (Formulatrix) robotic platforms. Libraries were prepared, pooled and cleaned using the Illumina Nextera XT kits or in-house Tn5, following the manufacturer’s instructions. Libraries were then sequenced on Nextseq or Novaseq (Illumina) using 2 × 75-bp paired-end reads and 2 × 8-bp index reads with a 200 cycle kit (20012861, Illumina). Samples were sequenced at an average of 1.5 million reads per cell.

RNAscopeThe RNAscope experiment was performed following the manufac-turer’s instructions using the RNAscope multiplex fluorescent reagent kit v2 (323100, ACD). All probes were purchased from existing stocks or custom designed from ACD.

Bioinformatics and data analysisMapping to the genome. Sequences from Nextseq or Novaseq were demultiplexed using bcl2fastq, and reads were aligned to the mm10 genome augmented with ERCC (External RNA Controls Consortium) sequences, using STAR version 2.5.2b. Gene counts were made using HTSEQ version 0.6.1p1. All packages were called and run through a custom Snakemake pipeline. We applied standard algorithms for cell filtration, feature selection and dimensionality reduction. Briefly, genes that appeared in fewer than five cells, samples with fewer than 100 genes and samples with less than 50,000 reads were excluded from the analy-sis. Out of these cells, those with more than 30% of reads as ERCC, and more than 10% mitochondrial or 10% ribosomal were also excluded from analysis. Counts were log-normalized and then scaled where appropriate.

Next, the ‘canonical correlation analysis’ function from the Seurat package27 was used to align raw data from multiple experiments. Only the first ten canonical components were used. After alignment, relevant features were selected by filtering expressed genes to a set of ~2,500 with the highest positive and negative pairwise correlations. Genes were then projected into principal component space using the robust principal component analysis. Single-cell principal component scores and gene loadings for the first 20 principal components were used as inputs into Seurat’s (v2) FindClusters and RunTsne functions to cal-culate 2D tSNE coordinates and search for distinct cell populations.

ArticleBriefly, a shared-nearest-neighbour graph was constructed based on the Euclidean distance metric in the principal component space, and cells were clustered using the Louvain method. Cells and clusters were then visualized using 3D tSNE embedding on the same distance metric. A neuron was characterized as ‘TRAPed’ trapped if it satisfied two conditions: (1) from the tdT + sort gate (tdT protein positive) and (2) tdT mRNA raw count > 0. Neuron subtype marker genes were found by using the FindAllMarkers function in Seurat (min.pct = 0.3, thresh.use = 0.25, min.diff.pct = 0.2). DEG analysis was done by applying the Mann–Whitney U-test on various cell populations. Raw P values were adjusted to an FDR. Permutation tests were then performed on all genes of interest. All graphs and analyses were generated and performed in R. GeneAnalytics and GeneCards packages offered by the gene set enrichment analysis tool were used for GO/KEGG/REACTOME pathway analysis and classification of enriched genes in each subpopulation.

Finding remote-memory-associated DEGs. To reduce our list of DEGs (FR TRAP versus NF TRAP results in 1,291 DEGs, cells from 4 biological replicates pooled, logFC > 0.3, FDR < 0.01) to only the most recall-specific, 4 steps were taken. Analysis was limited to C0–C4 neu-ron subtypes due to insufficient numbers of cells in C5 and C6 across all experimental conditions to make meaningful comparisons. First, DEGs are recalculated by assessing each experiment individually us-ing the whole transcriptome, and only DEGs (via the same criteria as pooled) that intersect in three-quarters of replicates are kept. Three out of four criteria were chosen as a compromise due to the high strict-ness of four out of four, which yielded only a maximum of seven DEGs (for a neuron subtype). All resulting DEGs are found in the initial DEG list (all replicates pooled), indicating that no additional DEGs were found as a result of analysing replicates separately. Second, ‘inactive’ (tdT-negative) populations were also compared (FR inactive versus NF inactive) and any DEGs that intersected with the DEGs left after the first criteria, were removed. This ensures that DEGs are activity-dependent, and not merely an overall upregulation in all cells due to the experience. This routinely removed genes such as Hsp90aa1 and Pcna-ps2. Third, the remaining DEGs had to be differentially expressed when FR TRAP was compared to either NR TRAP or HC TRAP. This ensures that the DEGs are specific to only neuronal ensembles that labelled by memory recall, and not due to forms of baseline activity (HC) or activity that remained from the initial fear learning (NR). Last, the remaining DEGs must pass a permutation test in which the training labels are shuffled and a distribution of log2FC is computed based on these labels. The true observed logFC must be above the 95th percentile of the distribution of the shuffled distribution. After placing these constraints, 99 genes remain from the original list of 1,291.

Assessment of activation score. A TRAPed (or inactive) cell is consid-ered to be ‘activated’ by the remote-memory DEG programme if 25%, 50% or 75% of the subtype-specific DEGs (remote-memory-associated DEGs only) is expressed above the 90th percentile of the distribution of that gene in NF TRAP controls from the same subtype. This calculation is then repeated with DEG programmes that are specific to each neu-ronal subtype. The fraction of cells activated with the subtype-specific signature is calculated as the number of activated cells divided by all cells in the subtype/activity group.

Regulatory motif analysis. Enrichment of known and de  novo motifs was found using HOMER by inputting the list of 99 remote- memory-associated DEGs and using the function findMotifs.pl and

the criteria ‘–start -400 -end 100 -len 8,10 -p 2’. The locations of the motifs in specific DEGs were found using the -find < motif file > option of findMotifs.pl.

RNAscope image analysis. Images were taken using a Nikon Confocal Microscope (at ×10 or ×20, NA = 0.45) and images were processed in ImageJ to only obtain the mPFC regions. The resulting images were fed into a custom image analysis pipeline on CellProfiler (using a com-bination of the functions IdentifyPrimaryObjects, RelateObjects, FilterObjects, MeasureObjectIntensity, ClassifyObjects and Calcu-lateMath. The custom pipeline can be found in Supplementary Meth-ods). Briefly, images were corrected with control slides (unstained sample and negative control probes) and cells were segmented using the DAPI signal. Those harbouring a signal (above a set threshold level) for both the subtype marker and the tdT probe were retained. The integrated fluorescence intensity of the DEG probe was calculated for each DAPI+/subtype+/tdT + cell. Cells that were not double-positive were not considered. The integrated fluorescent intensity was then normalized to the integrated DAPI signal per cell and results were plotted with custom scripts in R.

Reporting summaryFurther information on research design is available in the Nature Research Reporting Summary linked to this paper.

Data availabilityThe accession number for the single-cell RNA sequencing data reported in this paper is GSE152632.

Code availabilityCustom scripts can be found at https://github.com/mbchen-424/memory-sc-rnaseq.

25. Zhou, M. et al. A central amygdala to zona incerta projection is required for acquisition and remote recall of conditioned fear memory. Nat. Neurosci. 21, 1515–1519 (2018).

26. Picelli, S. et al. Full-length RNA-seq from single cells using Smart-seq2. Nat. Protoc. 9, 171–181 (2014).

27. Butler, A., Hoffman, P., Smibert, P., Papalexi, E. & Satija, R. Integrating single-cell transcriptomic data across different conditions, technologies, and species. Nat. Biotechnol. 36, 411–420 (2018).

Acknowledgements We thank L. Chen, M. Zhou and J. Li for discussion of the experimental design; S. Kolluru and D. Henderson for assistance in library preparation; N. Neff and J. Okamoto for assistance with sequencing; J. Lui for advice on brain dissociation; L. Denardo, J. Lui and L. Luo for the gift and help with the TRAP2 line; and W. Wang, G. Stanley and F. Horns for helpful discussions and computational assistance. S.R.Q. is a Chan Zuckerberg Investigator. This work is supported by a grant from the NIH (MH115999 to T.C.S.)

Author contributions X.J. and T.C.S. designed the animal experiments. M.B.C. and S.R.Q. designed the single-cell RNA sequencing experiments. X.J. performed the animal experiments, brain dissection, and in situ hybridization and imaging. M.B.C. performed the brain dissociation, flow cytometry, single-cell library preparation and sequencing pipelines. M.B.C. performed all single-cell RNA sequencing data and image analysis, with input from X.J., S.R.Q. and T.C.S. M.B.C. wrote the manuscript with substantial contributions from X.J., S.R.Q. and T.C.S. T.C.S. and S.R.Q. oversaw the project.

Competing interests The authors declare no competing interests.

Additional informationSupplementary information is available for this paper at https://doi.org/10.1038/s41586-020-2905-5.Correspondence and requests for materials should be addressed to S.R.Q. or T.C.S.Peer review information Nature thanks Steve Ramirez and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.Reprints and permissions information is available at http://www.nature.com/reprints.

replicatem1m2m3m4

tSNE_1

tSN

E_2

a

0

4000

8000

12000

m1 m2 m3 m4replicate

nGen

es

training

NFNRFRHC 3000

6000

9000

5.0 5.5 6.0 6.5log10(nReads)

nGen

es

5.0

5.5

6.0

6.5

7.0

log1

0(nR

eads

)

m1 m2 m3 m4replicate

c

FC HC0

20

40

60

80

100

% o

f cFo

s+ c

ells

that

are

iCre

+

FC HC0

2

4

6

8%

cel

ls c

Fos+

*p=0.02

d

e

dd

50 μm

cFos iCreDAPI cFos iCre

d

`b

ACC

PrLIL

ACC

PrLIL

ACC

IL

Extended Data Fig. 1 | Fidelity of the TRAP2; Ai14 line and sequencing quality metrics. a, Fidelity with which the tdTomato reporter in TRAP2; Ai14 mice captures endogenous cFos expression during fear memory encoding by in situ hybridization of cFos and iCre in the mPFC directly following fear conditioning. Left, percentage of cFos+ cells in either fear conditioned (FC) or homecage (HC) mice as seen by in situ staining (mean ± s.d.). Percentage of cFos+ cells that are also iCre+(mean ± s.e.m). Right, representative in situ hybridization images of FC mice (mPFC). b, Representative images of regions of

the mPFC analysed in this study (anterior cingulate cortex (ACC), the prelimbic (PrL) and the infralimbic cortex (IL)). Scale bars are 1mm (and 0.5 mm in the insets). c, Violin plots of the number of reads and number of genes per biological replicates (m1–m4), per cell. d, Scatterplot depicting the lack of strong relationship between number of genes detected and number of reads obtained per cell. e, Scaled expression of canonical markers in non-neuronal cells (Cldn5-BECs, Pdg fra-OPCs, Cx3cr1-microglia, Aqp4-astrocytes).

Article

Fear-Recall (FR)

0

100

200

300

400

# ce

lls

0

100

200

300

400

# ce

lls InactiveTRAPed

C0 C1 C2 C3 C4 C5 C6 C0 C1 C2 C3 C4 C5 C6

No Fear (NF)

11

280

10499

12

149

8564

23

157

10368

16

104

4847

6

653726

1392617 10

5027230

100

200

# of

TR

APed

cel

ls

neuron subtypeC0 C1 C2 C3 C4 C5 C6

neuron subtype neuron subtype

a

c

b

Allen mouse brain atlas (2018)

Dkkl1

Tnfaip8l3

Calb2

Rprm Tesc

d

NFNRFRHC

Lhx6Tshz2

e

f

0

200

400

600

neuron subtype

# ce

lls

replicatem1m2m3m4

C0 C1 C2 C3 C4 C5 C6tSNE_1

tSN

E_2

Replicatem1m2m3m4

Neuron subtype composition

TRAP

In

Fear

-Rec

all (

FR)

C0C1C2C3C4C5C6

TRAP

No

Fear

(NF)

TdTom_transgeneCrtac1Rbp4CygbZcchc12Laptm4bGpx3Cox6a2Lypd6bRab3bKcnc2CartptCortLypd6SstLhx6Lgals1Nxph1CrhbpTrp53i11Plcxd2Pid1Uap1Tmem159Wscd1Gsta4Efr3aSatb1Plcxd3Nrsn2Sla2Fezf2Neto2Chrm2Il11ra1Lypd1Etv1Myl4Slc17a6GrpAbi3bpStard5Tshz2Nrip3CryabQpctKitDnerCplx3Rgs12Npas1Cnr1NpyDlx6os1Adarb2Nr2f2RelnSncgTnfaip8l3PrkcgLamp5NeflDok5Cdh13Ak5Wfs1Syt17Sv2bGria3NnatPtk2bOciad2Cacna1eNptxrAtp2b4ItpkaCalb1Mas1Rasl10aIgfbp6TescSerpine2Gpd1Cbln2Pcp4l1Kcnip1Limch1Cxcl14PnocRpp25PenkRgs10Cd63Tac2Gad2Slc6a1SynprAp1s2Htr3aGad1Dlx1PthlhCalb2Igf1CrhVipCyr61Crip2Baiap2NgefPde1aKcnip3Cxcl12IghmRab26Igsf21Hs3st2Pcp4Lmo3Adora1Spon1Ifi27RP23Crym3110035E14RikPpp1r1bNptx1Ramp3Hs3st4Arhgap25RprmCdkn1aArhgap32Camk4Lingo1NrepBdnfLmo4Dusp14NrgnStx1aRtn4rGfra2Foxp1Arpp21Cnih3Rgs4PtnEpha4Dnajc21Cpne9Krt12Tnnc1Dkkl1Slc30a3Nrn1

C0 C1 C2 C3 C4 C5 C6subtype:

Enric

hed

gene

s pe

r neu

rona

l typ

e

Extended Data Fig. 2 | Distribution of cell numbers and neuronal subtypes across various training conditions. a, Representative flow cytometry plot of the amount of tdT + events per training condition. In scatter plot, each point represents one mouse (mean ± s.d.). b, Number of cells from each biological replicate that were annotated as one of 7 defined neuron subtypes (C0–C6). c, Scaled expression of the top marker genes for each neuron subtype.

d, Number of TRAPed and Inactive cells (as defined by non-zero expression of tdT mRNA) collected per neuron subtype, in either fear-recall (FR) or no-fear (NF) mice. e, Representative images of subtype marker genes (C0–C6) in the Allen Brain ISH atlas. f, Neuron subtype composition of FR and NF populations. Number of TRAPed (tdT mRNA+) cells collected in each experimental condition that fall in one of 7 neuronal subtype categories.

0 5 10 15 20

Regulation of TranslationProteining folding

Intracellular Protein TransportPositive Regulation of Receptor Recycling

Nervous System DevelopmentRegulation of Filopodium Assembly

L serine TransportNegative Regulation of Peptidyl serine Phosphorylation

Vesicle mediated TransportCellular Response to Norepinephrine Stimulus

Regulation of Vesicle mediated exocytosis

Score

Vamp2

Ubc

Gsk3b

Ptp4a1

Serinc3

Hacd3

Trim35Alg2

Skiv2l2

Mmd

Pdha1

Cdc42se2Eif2ak1Kctd10

App

Hid1

Abcf3

Pigq

Sult4a1Zfp706

Ghitm

Miga2

Fam134a

Sar1aAcsf3

Rtn1 Guk1

Atp5g3 Crip2Atp6v0c

Mpc1Hsbp1

ClppTmem50a

1

0

1

2

0.25 0.50 0.75 1.00abs(99th percentile of permuted distribution (logFC))

True

obs

erva

tion

(logF

C)

Gpm6aVamp2

Ncdn

Ubc Gsk3b Ptp4a1Serinc1

Serinc3

Hnrnph2Usp5

Nell2

Fam131a

Trim32

Rtn3

Rab5a

Dmtn

Pcsk2 Nsf

Mfsd14b

Hacd3

Sdha

Tmem151aPfkm

Pip4k2c

Pja2

Gdi2

Rab15

Tdg

Gfra2Emc1

Emc4

Cck0.5

0.0

0.5

1.0

1.5

0.2 0.4 0.6abs(99th percentile of permuted distribution (logFC))

True

obs

erva

tion

(logF

C)

Gpm6aUbc

Serinc1

Serinc3

Vcp rs

Pak1

Mfsd14aNck2

Gm5436

Slc25a46

Mir1242hg

Stx1b

Itfg1

Rtn1

Guk10.5

0.0

0.5

1.0

1.5

0.2 0.4 0.6 0.8

abs(99th percentile of permuted distribution (logFC))

True

obs

erva

tion

(logF

C)

Ubc

Serinc1

Hacd3

Prkar1b

Slc30a9

Lmbrd1Ankrd45 Pls3

Ctbp1 Atad1

Strip1

DnerTimm29

Vopp1

Trim35

Tmx1

Cdv3

Hmg20a

Inpp5f

Alg2

Rtn1

Psmb6

1110051M20Rik

Atp6v0bGm8797 Hint1

1

0

1

2

0.3 0.6 0.9abs(99th percentile of permuted distribution (logFC))

True

obs

erva

tion

(logF

C)

Ubc

Ptp4a1

Serinc1 Serinc3

Nsf

Slc25a46

Itfg1

Aplp1 Hnrnpk

Syt13

Dpysl4

Garnl3

Gm13340

Plekhb2

AI413582

Cycs

Elmod1

Rab24

Erp29

Sarnp

1

0

1

0.25 0.50 0.75abs(99th percentile of permuted distribution (logFC))

True

obs

erva

tion

(logF

C)

C0

- D

kkl1

(G

lut)

C1

- R

prm

(G

lut)

C2

- C

alb

2 (

GA

BA

)C

3 -

Te

sc (

Glu

t)C

4 -

Tn

faip

8l3

TRAPed FR vs NF Permutation test

40 0 40 80

Protein TransportIntracellular Protein Transport

Ephrin Receptor Signaling PathwayProtein Ubiquitination

Vesicle mediated TransportMembrane Organization

Nervous System DevelopmentUbiquitin dependent Protein Catabolic Process

Rhythmic ProcessSynapse Organization

ER to Golgi Vesicle mediated TransportLong term Synaptic Potentiation

Learning or MemoryNeuron Projection Development

MRNA ProcessingRegulation of Circadian Rhythm

Negative Regulation of Epidermal Growth Factor Receptor Signaling PathwayTricarboxylic Acid Cycle

Regulation of DNA binding Transcription Factor ActivityRegulation of Neuronal Synaptic Plasticity

Regulation of AxonogenesisCellular Process

Regulation of Synaptic PlasticityRNA Splicing

Neuron Projection MaintenanceGolgi to Plasma Membrane Protein Transport

Regulation of Synaptic Vesicle ExocytosisNegative Regulation of Neuron Apoptotic Process

ExocytosisViral Process

Antigen Processing and Presentation Via MHC Class IILearning

Positive Regulation of Synapse AssemblyRegulation of Alternative MRNA Splicing, Via Spliceosome

Protein DephosphorylationNegative Regulation of Protein Autophosphorylation

Regulation of TranslationRegulation of Catalytic Activity

Regulation of SNARE Complex AssemblyRegulation of Axon Extension

ATP Metabolic ProcessNeuron Apoptotic Process

Positive Regulation of GTPase ActivityProtein Stabilization

Phagosome AcidificationFc epsilon Receptor Signaling Pathway

Proteasome mediated Ubiquitin dependent Protein Catabolic ProcessRegulation of Macroautophagy

Protein DeubiquitinationMitochondrial Electron Transport, Cytochrome C to Oxygen

Post translational Protein ModificationATP Hydrolysis Coupled Cation Transmembrane Transport

Positive Regulation of Canonical Wnt Signaling PathwayStimulatory C type Lectin Receptor Signaling Pathway

Microtubule based ProcessProteolysis Involved in Cellular Protein Catabolic Process

Proteasomal Protein Catabolic ProcessTumor Necrosis Factor mediated Signaling Pathway

Regulation of MRNA StabilitySCF dependent Proteasomal Ubiquitin dependent Protein Catabolic Process

Mitochondrial Translational ElongationMitochondrial Translational Termination

Regulation of Mitotic Cell Cycle Phase TransitionElectron Transport Chain

Proteasomal Ubiquitin independent Protein Catabolic ProcessInterleukin 1 mediated Signaling Pathway

Proton Transmembrane TransportAntigen Processing and Presentation

Substantia Nigra DevelopmentCristae Formation

Mitochondrial ATP Synthesis Coupled Proton TransportNIK/NF kappaB Signaling

Regulation of Hematopoietic Stem Cell DifferentiationATP Synthesis Coupled Proton Transport

ATP Biosynthetic ProcessAnaphase promoting Complex dependent Catabolic Process

Negative Regulation of G2/M Transition of Mitotic Cell CycleWnt Signaling Pathway, Planar Cell Polarity Pathway

Regulation of Transcription From RNA Polymerase II Promoter in Response to HypoxiaRegulation of Cellular Amino Acid Metabolic Process

Translational InitiationViral Transcription

Oxidation reduction ProcessNuclear transcribed MRNA Catabolic Process, Nonsense mediated Decay

TranslationMitochondrial Respiratory Chain Complex I Assembly

SRP dependent Cotranslational Protein Targeting to MembraneMitochondrial Electron Transport, NADH to Ubiquinone

40 0 40 80

#DEGs

Score

downreg with FRupreg with FR

Remote memory-associated DEGs (99)

All DEGs (C0-C4) - TRAPed FR over NF

112

8

7

23

292

54

Yme1l1

PreplUap1Ndufs1Lgi1Ppp2r2aStx12Upf1Tmem263

Pdcd4NarsUba2Impact

CpeChchd2Hsp90aa1Cdc123Sra1Pcna

ps2Slc25a46Pak3

C0C1C3C2C4

log2 FC(DEG in 3/4 rep)

0

2

4

-4

-2

FR inactive vs NF inactivea

b

c

Extended Data Fig. 3 | Differential gene expression in distinct neuronal subtypes (FR over NF TRAPed populations). a, DEGs in each neuron subtype when inactive (tdT-) neurons are compared between FR and NF mice. b, Left, volcano plots of DEGs in FR vs NF mice for each neuron subtype (C0 to C4). DEGs found when all replicates are pooled analysed in a combined manner are shown in red. Recall-dependent DEGs (defined as being differentially expressed in at least 3 out of 4 replicates, when analysed replicates are analysed individually) are labelled in black. Right, permutations are performed for every recall-dependent DEG for each neuronal population. Upregulated DEGs lying above the y = x line (red) and downregulated DEGs lying below the y = x line

(blue) are considered to be above the 99th percentile of the permuted distribution. c, GO enrichment analysis of all up- and downregulated DEGs (941 DEGs up, 384 DEGs down, all neuron subtypes combined) when all replicates are pooled. Bars show the enrichment scores (GeneAnalytics) for the GO pathway and dot indicates the number of DEGs involved in that pathway. d, GO enrichment analysis of only the upregulated remote-memory-specific DEGs (from Fig. 3c) from all neuron subtypes combined. Bar indicates enrichment score (Gene Analytics) and number indicates number of recall-dependent DEGs involved in that pathway.

Article

1d 2d

No Salience (NS)

Salience (S) deprive

10 d

Food & 4-OHT

tdT production

0d

4-OHT injection

Active neuronal ensembleCollect single cells

Chow

WT

Homec

age

No Fea

r

Fear-R

ecall

No Sali

ence

Salien

ce0

1

2

3

experience

% o

f eve

nts

that

are

tdT+

in fl

ow

**p=0.002

p=0.013*

50

25

0

25

25 0 25tSNE_1

tSN

E_2

Fear paradigm (FR, NF, NR, HC)Food salience paradigm (S, NS)

50

25

0

25

25 0 25tSNE_1

tSN

E_2

C0-Dkkl1+C1-Rprm+

C2-Calb2+C3-Tesc+

C4-Tnfaip8l3C5-Tshz2C6-Lhx6

SalienceTRAPed

C0C1C2C3C4C5C6

neuronsubtype

FRTRAPed

Inactive

*

*

**

C0 C1 C2 C3 C4 C0 C1 C2 C3 C4Fear-Recall vs No Fear Food Salience vs No Salience

Pcna-ps2

avg log2(FC)

21

012

Gm13340

intersecting DEGs

a

b c

d e

deprive10 hr later

4-OHT

Extended Data Fig. 4 | Analysis of TRAPed ensembles in food salience (S) versus no salience (NS) mice. a, Schematic of experimental paradigm for generating TRAPed neuronal ensembles as a result of food salience (food deprivation followed by food return (salience) or no food return (no salience)). b, Percentage of events in flow cytometry that were tdT +, by experience (mean ± s.d.). c, tSNE of the merge of data from fear-recall experiments and

food-salience experiments coloured by neuron subtype (left) and experimental paradigm (right). d, Subtype composition differences between TRAPed fear-recalled ensembles and TRAPed food salience ensembles, as compared to background Inactive ensembles. e, Heat map of the average log2 fold change of DEGs in each neuron subtype when comparing fear-recall vs no-fear, and salience vs no salience. Only DEGs with FDR <0.01 are shown.

C0 C1

C2 C3

C4

Extended Data Fig. 5 | DEGs when comparing ensembles from food salience (S) to no salience (NS) mice. Volcano plots showing the log2 fold change and adjusted P values (in log10 scale) of genes when comparing food

salience over no salience groups within each neuron subtype (C0 through C4). Top DEGs per neuron subtype are labelled in red. Positive log2 fold change indicates upregulation in food salience (S) group.

Article

Stx1b

Emc1

Atp6v0b

Hnrnph2

Gdi2

Nck2

Rab24

Dpysl4

Pja2

Rab15

Vamp2

Cck Rab24

Pja2

Gsk3b

NsfStx1b

Gdi2

Rtn1

Gsk3b

Pdha1

Pfkm

Atp6v0c

Erp29

Aplp1

Nck2

Rtn3

Sdha

Cycs

Lmbrd1

Eif2ak1

Trim32

Sdha

Gdi2

Cycs

NsfNsf

Hid1

Rtn1

Nel l2

Hnrnpk

Mpc1

Pja2

Ctbp1

Vamp2

Cdc42se2

Rtn1

Pja2

Syt13

Hnrnph2Emc4

Hint1

I t fg1

Gpm6aStx1b

Ghitm

Dner

Rtn1Nsf

Rtn1

Cck

Sar1a

Cycs

Rtn1

Serinc1

Ankrd45

Vamp2

Emc4

Atp6v0c

Hnrnpk

Usp5Rab5a

Pak1

Psmb6Gsk3b

Rab5aRab5aApp

Trim32

Gsk3b

Vamp2

Rab5aRab5a

Trim32

Psmb6

Vamp2

Nsf

Rab24

Syt13

App

Sdha

Pdha1

Gdi2App

Hnrnpk

App

Atp5g3Sdha

Mmd

Abcf3

Tr im35

Clpp

Gdi2

Atp6v0c

Usp5

Dner

Rab24

Gsk3b

Atp6v0c

Ghitm

Pak1

App

Hint1

Aplp1

Serinc1Vamp2

Rab5a

Tmx1

Sdha

Vamp2

Syt13

Serinc1

Aplp1

App

Syt13

Rtn3

Pcsk2

AppApp

Pdha1

Serinc1

Rab5aSar1a

Serinc3

Up or down regulat ion(FR over NF)

Down

Up

Stx1b

Emc1

Atp6v0b

Hnrnph2

Gdi2

Nck2

Rab24

Dpysl4

Pja2

Rab15Cck Rab24

Pja2

Stx1b

Gdi2

Pdha1

Pfkm

Atp6v0c

Erp29

Aplp1

Nck2

Atp6v0b

Rtn3

Sdha

Cycs

Lmbrd1

Eif2ak1

Trim32

Sdha

Gdi2

Cycs

Hid1

Nell

Hnrnpk

Mpc1

Pja2

Ctbp1

Cdc42se2

Pja2

Syt13

Emc4

Hint1

Itfg1

Gpm6a Stx1b

Ghitm

Dner

Cck

Sar1a

Cycs

Rtn1

Ankrd45

Emc4

Atp6v0c

Hnrnpk

Usp5Rab5a

Pak1

Psmb6

Rab5aRab5aApp

Trim32

Rab5aRab5a

Trim32

Psmb6

Nsf

Rab24

Syt13

App

Sdha

Pdha1

Atp5g3

Gdi2App

Hnrnpk

App

Sdha

Mmd

Abcf3

Trim35

Clpp

Gdi2

Atp6v0c

Usp5

Dner

Rab24

Atp6v0c

Ghitm

Pak1

App

Hint1

Aplp1

Rab5a

Tmx1

Sdha

Syt13Aplp1

App

Syt13

Rtn3

Pcsk2

AppApp

Pdha1

Rab5aSar1a

Vamp2

Gsk3b

Serinc1

neuron subtype

C0

C1

C2

C3

C4

neurotransmitter

GABA

GLUT

Serinc3

a

0.1

0.2

0.3

0.4fra

ctio

n in

duct

ion

(C1

DEG

s)

C1 C0 C1 C2 C3 C4 C1

NF FRTRAP Inactive

0.2

0.4

0.6

fract

ion

indu

ctio

n (C

2 D

EGs)

TRAP InactiveC2 C0 C1 C2 C3 C4 C2

NF FR

0.20.40.6

TRAP InactiveC3 C0 C1 C2 C3 C4 C3fra

ctio

n in

duct

ion

(C3

DEG

s)

NF FR

Rtn

1H

sbp1

Vam

p2Sl

c25a

46Ab

cf3

Rab

24N

sfPr

kar1

bPl

ekhb

2Pi

p4k2

cD

pysl

4M

iga2

Kctd

10H

int1

Ghi

tmSl

c30a

9G

uk1

Cdv

3C

ckEm

c4Sd

haAt

p6v0

bC

lpp

Tmem

50a

Mfs

d14b

Lmbr

d1Ap

pAt

p6v0

cU

bcItf

g1C

rip2

Hac

d3Td

gPc

sk2

Gar

nl3

Hnr

nph2

Hm

g20a

Strip

1Sk

iv2l

2An

krd4

5Tm

x1H

nrnp

kPt

p4a1

Tmem

151a

Eif2

ak1

Mpc

1In

pp5f

Serin

c3C

dc42

se2

Sar1

aZf

p706

Hid

1Pd

ha1

Fam

131a

Tim

m29

Ctb

p1Em

c1Ap

lp1

Pigq

Acsf

3G

pm6a

1110

051M

20R

ikSe

rinc1

Trim

35D

mtn

Atp5

g3Vo

pp1

Cyc

sG

di2

Usp

5R

ab5a

Dne

rM

ir1242

hgR

tn3

Mfs

d14a

Elm

od1

Trim

32Pj

a2Su

lt4a1

Rab

15Er

p29

Pls3

ATCG

AG TCAGTC

G TAC

GATC

C TAG

TGCA

AC TGACGT

AC TGATCGTGACGACTGACTCGTAATGCACTGCGAT

GACT

ATCG

ATCG

AC TGCG TAAGCT

ATGC

TGCA

GACTCAGTCTGA

CATGCATGTGACAGCT

ATGCGATCTCAG

CTGACAGTCTGA

AGTC

CTAG

G ACT

ATCGGTAC

ACGT

C TGA

G ATC

ATCG

G ACTTCAG

GTACTAGC

ACGTACGT

CG TAC TGA

AG TCGCAT

AC TGACGTCGAT

AGCT

AGTC

CGAT

CGAT

GATC

AC TGGTAC

C TAG

ATCG

AGCT

AG TCATGC

AGCT

CG TAAC TGAC TGG TCA

AGTC

AGTC

TCGA

ACGT

AGCT

C TAG

ACGT

AG TCAG TCACGTACGTATCG

ATGC

GCTAGTCA

TGAC

AG TCAGTC

AC TGC TAG

G TCA

CG TACGTA

AC TGCG TAAG TCCG TAAGCT

ATGC

CTGA

CG TAACTG

CATG

CG TAAC TGACGT

GCAT

CG TACG TACGTA

C TAG

ACTG

AG TCTCGA

CATG

CG TAC TGA

CGTA

0 10 20 30 40

E2F1

ZNF652

JUN

Ddit3

Sox6

ZBTB12

POL008.1

MYBL1

HIF1a

HIF1b

POL013.1

TFAPC2

RHOXF1

GMEB2

Best match

Upregulated and Downregulated Exemplary DEGs (FR over NF)

% of Target or background

Motif score

0

5

10

15

Motif

0.2

0.4

0.6

TRAP InactiveC4 C0 C1 C2 C3 C4 C4

fract

ion

indu

ctio

n (C

4 D

EGs)

NF FR

c

b

Extended Data Fig. 6 | Neuron subtype-specific activation programs, hypothesized protein–protein interactions and upstream regulatory motifs. a, Fraction of cells in each neuron subtype that are induced with the transcriptional program (that is, DEGs) from a neuron subtype. Overall, the activation program of each TRAPed neuron subtype is found to be more specific to it than the inactive population, or other subtypes. b, Left, de novo regulator motif discovery: analysis was performed using HOMER on the subset of 99 remote-memory-associated DEGs by looking at the sequences -400 to +100 bp from the TSS. 12 de novo and 2 known motifs were found (only motifs with an enrichment P  value <10−2) were kept). Heat map depicts the ‘motif score’ of each DEG for each motif, and genes and motifs were clustered via the ward.D

method. Right, bar graph depicting the percentage of the DEGs (target sequences) that possess a match for the motif within -400 to +100 bp from the TSS, vs the percentage of background sequences. For de novo motifs, the best match gene is listed on the right. HIF1b and HIF1a are matches to known motifs. c, Left, hypothesized protein–protein interactions of a subset of recall-dependent DEGs (TRAPed FR/NF) using the STRING database (https://string-db.org/). Only genes that are connected at a confidence level of 0.4 (medium) are shown. Connections indicate a possible existence of an interaction between two proteins. Genes are coloured by up of downregulation in FR/NF. Right, same network plot, with nodes coloured by the neuron subtype which differentially regulates the DEG.

0.0

0.1

0.2

0.3

0.4

0.5

Frac

tion

of T

RAP

ed c

ells

that

are

Dkk

l1+

RNA-scope scRNA-seq

0.0

0.1

0.2

0.3

0.4

Frac

tion

of T

RAP

ed c

ells

that

are

Tnf

+

RNA-scope scRNA-seq

0.0

0.1

0.2

0.3

0.4

Frac

tion

of T

RAP

ed c

ells

that

are

Tes

c+

RNA-scope scRNA-seq

0.00 0.02 0.04 0.06 0.08

NF

FR

Ratio of Nuclei that are TRAP+

**p=0.006

FRNF

a

bDAPIMerge tdTTnfaip8l3 Syt13

Merge DAPI tdT Vamp2Tesc

NF

FR

FR

NF

DAPIMerge tdTRprm Stx1bNF

FR

c

Extended Data Fig. 7 | In situ validation of tdT levels, neuronal subtype compositions and remote-memory-specific DEGs in the mPFC. a, Ratio of Nuclei that are tdT+ (mRNA level) per training condition. Each data point represents one region of interest. (mean ± s.d.) b, Ratio of TRAPed cells that are positive for a neuronal subtype marker obtained either via the RNA-scope method, or by scRNA-seq (mean ± s.e.m.) (see Fig. 2). TRAPed cells are defined as DAPI+/tdT+ in RNAscope quantification, and as tdT mRNA count >1 in

scRNA-seq (post-QC). No significant differences are found between FR and NF within either RNAscope or scRNA-methods, indicating no major changes in neuronal subtype composition of active populations in different training conditions. c, in situ hybridization of key remote-memory specific DEGs including Stx1b in Rprm+/tdT + cells, Syt13 in Tnfaip8l3+/tdT + cells, Vamp2 in Tesc+/tdT + cells. Scale bars, 100 μm.

Article

AgrnCol4a4Dusp18LplThbs2Lama2Col4a5AppNtng2NrtnCntn2A2mTncEfna5CalrHsp90b1Psen1PsapLrpap1ApoeSerpine2Fgf13MdkRph3aF8Efna2Nlgn3Nlgn1Nlgn2L1camNxph1Ntng1TnfC1qbF9Nxph3PlauGpc3VtnCtgfCol4a1Efna3Col4a3Lama1NgfPlatEfna1Efna4Col4a6Lamb2Serping1TfpiSerpine1Thbs1Pdgfb

Maged1

Atp1a3

Tnfrsf21

Lrrc4c

Lrp1

Ncstn

Epha7

Scn8a

Nrxn1

Gfra2

Cntn1

Cd47

Cd81

Rpsa

ECm

icrog

lia

OPC

olig

oas

trocy

te

C0 C1 C3 C2 C4

EXH INH 0 1

log2FC (FR/NF)

2 2

scaled avg eprs

FDR<0.05

Neuronal RECEPTORSenriched in FR over NF Avg expression of

non-neuronal LIGANDS

Adam15

Adcyap1

App

Calm1

Calm2

Calm3

Cck

Cx3cl1

Efna3

Efna5

Fgf14

Gad1

Gas6

Hsp90b1

Icam5

Ncam1

Nlgn1

Nlgn3

Pkm

Ptdss1

Ptma

Qdpr

Rgmb

Rps19

Sema4b

Sema4f

Sorbs1

Itga5

Itga9

Itgav

Itgb1

Itgb3

Adcyap1r1

Sctr

Vipr1

Vipr2

Cav1

Cd74

Gpc1

Lrp1

Ncstn

NgfrSlc45a3

Tnfrsf21

Abca1

Adcy8Adcyap1r1

Cacna1cCrhr1

Egfr

Fas

Grm3

Grm4

Grm5

Grm7

Hmmr

Htr2c

Insr

Kcnn4

Kcnq1

Kcnq3

Kcnq5

Mylk

Oprm1

Pde1a

Pde1b

Pde1c

Ptpra

Sctr

Trpc3

Trpc5

Vipr1

Abca1

Adcy8

Cacna1c

Egfr

Grm5

Grm7

Insr

Kcnq1

Kcnq3

Kcnq5

Mylk

Pde1a

Pde1b

Pde1c

Trpc5

Abca1

Adcy8

Egfr

Grm5

Grm7

Insr

Kcnq1

Kcnq3

Kcnq5

Mylk

Pde1a

Pde1b

Pde1c

Trpc5

Cckbr

Cx3cr1

Epha1

Epha2

Epha3

Epha4

Epha5

Epha6

Epha7

Ephb1

Ephb6

Epha1

Epha2

Epha3

Epha4

Epha5

Epha6

Epha7

Ephb1

Ephb2

Ephb6

Fgfr1

Fgfr2

Fgfr3

Grm4

Axl

Mertk

Tyro3

Asgr1

Erbb2

Lrp1

Tlr1

Tlr2

Tlr4

Tlr7

Tlr9

Itgb2

Cacna1c

Fgfr1

Fgfr2

Gfra1

Ptpra

Robo1

Nrxn1

Nrxn2

Nrxn3

Nrxn1

Nrxn2

Nrxn3

Cd44

Jmjd6

Scarb1

Vipr1

Dysf

Bmpr1b

Bmpr2

Neo1

C5ar1

Dcbld2

Nrp2

Insr

Itga1

Itgb5

C0 C1 C3 C2 C4

EXH INH

mic

rogl

iaE

CO

PC

olig

oas

trocy

te

Neuronal LIGANDSenriched in FR over NF

Avg expression of non-neuronal RECEPTORS

0 1

log2FC (FR/NF)

2 2

scaled avg eprs

FDR<0.05

d

e

Adam15

Adcyap1

App

Calm1

Cx3cl1

Fgf14

Gas6

Hsp90b1

Icam5

Ncam1

Nlgn1

Nlgn3

Rps19

Sorbs1

ItgavItgb1SctrNcstnTnfrsf21Abca1EgfrGrm3Grm5Kcnq3

Sctr

Cx3cr1Fgfr2Fgfr3MertkTlr2Tlr4Tlr7Itgb2

Fgfr2

Nrxn1C5ar1Itgb5

-log10(FDR)<1.32

4

6

21

012

log2FC (FR / NF)

C0 C1 C3 C2 C4

astro

cyte

micr

oglia

Perturbed neuronal ligands Perturbed glial receptorsc

0

200

400

astro

cyte EC

microg

liaoli

goOPC

# ce

lls

NFFR

Endothelial OPC Oligodendrocytea

b

Extended Data Fig. 8 | DEGs and potential cell–cell interactions in non-neuronal cells during memory consolidation. a, Volcano plots of non-neuronal cell types when comparing cells in FR over NF nice. DEGs (FDR >0.01, log2FC >1) are labelled in red, and exemplary DEGs (high log2FC and log10FDR) are labelled in black. b, Number of non-neuronal cells collected in this study, for each cell type and experimental condition. c, Heat map of a subset of neuronal ligands and glial receptors that are found to be differentially perturbed upon memory consolidation. Only receptors and ligands which were

found to be (differentially) expressed are shown. d, Left, heat map of the log2FC of DEGs (FR over NF) in neurons that are classified as ligands. Middle and right, Sankey plot of known ligand-receptor pairs and heat map of the average scaled expression level of the corresponding receptors in each cell type. e, Left, heat map of the log2FC of DEGs (FR over NF) in neurons that are classified as receptors. Middle and right, Sankey plot of known ligand-receptor pairs and heat map of the average scaled expression level of the corresponding ligands in each cell type.

C0

C1

C3

C2

C4

Rao

-Rui

z, 2

019

Hrv

atin

, 201

8La

car,

2016

Cho

, 201

6

Serinc3Ptp4a1Serinc1Slc25a46Itfg1Mfsd14aMir1242hgPlekhb2HnrnpkAplp1NsfGarnl3Gfra2NcdnNell2PfkmSyt13Rtn1CckInpp5fPip4k2cNck2Stx1bHnrnph2Pja2SdhaGdi2Fam131aRab15Usp5Tmem151aDmtnRtn3Pcsk2Gpm6aTrim32Mfsd14bRab5aTdgEmc1Psmb6Hint1Emc4Atp6v0bGuk1ClppMpc1Atp6v0cAtp5g3Tmem50aElmod1CycsSarnpRab24Erp29Hid1Kctd10Pdha1PigqSkiv2l2Abcf3Miga2Cdc42se2Sar1aFam134aGsk3bVamp2MmdGhitmZfp706Sult4a1AppAcsf3Eif2ak1Dpysl4Pak1Crip2Lmbrd1Tmx1DnerAtad1Ankrd45Timm29Cdv3Prkar1bHmg20aCtbp1Vopp1Strip1Slc30a9Pls3Alg2Trim35Hacd3

C0

C1

C3

C2

C4

Rao

-Rui

z, 2

019

Hrv

atin

, 201

8La

car,

2016

Cho

, 201

6

log2_FC

21

012

Che

n et

al,

1292

DEG

s, c

ells

poo

led

(FR

vs

NF,

TR

APed

pop

ulat

ions

)

Che

n et

al,

rem

ote

mem

ory-

asso

ciat

ed D

EGs

(FR

vs

NF,

TR

APed

pop

ulat

ions

)

a Tmem151bPsme3NckipsdNsmfPpme1Susd6Atp13a2Pias4Fam131aKlc1Gdi1AppPrdx1Hdac3Utp15NcdnPkmNeflKcnv1Zbtb16MlxEgr4Adora1HmgcrMrpl37Rrp1Nudt3Rell2Efhd2Chst2GnazFkbp89530068E07RikYarsClcn4Itfg1Hspd1Flot1Med25Spock2Exoc7Zfp426Inpp5fKcnip4Gpd1Tdtomato

C0

C1

C3

C2

C4

Rao

-Rui

z, 2

019

Hrv

atin

, 201

8La

car,

2016

Cho

, 201

6

FR T

RAP

vs

FR In

activ

e

b

Extended Data Fig. 9 | Comparison of remote-memory DEGs with previously published datasets of experience-dependent transcriptional activity. a, Left, heat map of the log2FC of all 1,292 DEGs (FDR <0.01, FR over NF, all cells pooled) in this manuscript, and their log2FC values in previously published datasets of experience-dependent DEGs in activated neurons during: recent fear memory retrieval24, associative fear-learning22, post-visual stimulus23, or novel environment exposure21. A value of zero log2FC indicates the gene was not differentially expressed in a dataset. Right, same as left, but

now DEGs are filtered down to the ‘Recall-dependent DEG’ set derived from this manuscript. Only genes differentially expressed in three out of four replicates are remaining. b, Left, log2 fold change heat map of the recall-dependent DEGs between tdT + vs tdT − neurons in FR mice (genes are differentially expressed in >3/4 replicates) undergoing remote fear memory consolidation. Right, the log2FC values of these genes if they are found in previously published datasets of experience-dependent DEGs (see a). A value of zero log2FC indicates the gene was not differentially expressed in that dataset.

Article

Elmod1Erp29

AI413582Psmb6

Tmem50aMpc1Atp6v0bHsbp1Hint1CckEmc4Guk1Atp6v0cRtn1Hsp90aa1CycsCrip2ClppAtp5g3

Rab15Pip4k2cPcsk2Syt13Nell2

Gfra2

Stx1b

Mir1242hgMfsd14b

Gsk3b

Pigq

Acsf3Mmd

Eif2ak1

Kctd10Abcf3Miga2Sult4a1

HnrnpkDmtn

Fam134a

Tmem151aFam131aDpysl4Pak1Skiv2l2Mfsd14aNck2Vamp2Cdv3Ankrd45Strip1Hmg20aNsfTimm29Zfp706AppSdhaPfkmCtbp1Usp5Aplp1Prkar1bLmbrd1Hid1Emc1Pls3TdgVopp1Sar1aAtad1Plekhb2Trim35Itfg1Hacd3Rab5aSlc30a9DnerUbcRtn3GhitmGpm6aGdi2Pja2Tmx1Cdc42se2NcdnHnrnph2Serinc1Ptp4a1Pdha1Pcna ps2Slc25a46Trim32Alg2Serinc3Inpp5f

FR vs NF NR vs NF

Recall-specific DEGs

Fear-relatedDEGs

FR vs NF NR vs NF

C0

C1

C3

C2

C4

C0

C1

C3

C2

C4

C0

C1

C3

C2

C4

C4

log2_FC

21

012

C4

C0

C1

C3

C2

C4

Comparison of Recall-specific and fear-related DEGs

All Recall-specific DEGs

Extended Data Fig. 10 | Comparison of remote-memory-specific DEGs and fear-experience-related DEGs. Left, log2 fold change heatmap of the union of DEGs from comparing FR vs NF (remote-memory specific) and NR vs NF (fear-related). Sustained transcriptional changes from the fear-experience

itself is shown in the yellow highlighted columns. Right, zoomed in view of the portion of the heat map within the green boxes where most fear experience-related DEGs are located.

1

nature research | reporting summ

aryO

ctober 2018

Corresponding author(s): Chen MB, Jiang X, Quake SR, Sudhof TC.

Last updated by author(s): Jul 21, 2020

Reporting SummaryNature Research wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency in reporting. For further information on Nature Research policies, see Authors & Referees and the Editorial Policy Checklist.

StatisticsFor all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section.

n/a Confirmed

The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement

A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly

The statistical test(s) used AND whether they are one- or two-sided Only common tests should be described solely by name; describe more complex techniques in the Methods section.

A description of all covariates tested

A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons

A full description of the statistical parameters including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient) AND variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals)

For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted Give P values as exact values whenever suitable.

For Bayesian analysis, information on the choice of priors and Markov chain Monte Carlo settings

For hierarchical and complex designs, identification of the appropriate level for tests and full reporting of outcomes

Estimates of effect sizes (e.g. Cohen's d, Pearson's r), indicating how they were calculated

Our web collection on statistics for biologists contains articles on many of the points above.

Software and codePolicy information about availability of computer code

Data collection - Sequences from the Nextseq or Novaseq were demultiplexed using bcl2fastq, and reads were aligned to the mm10 genome augmented with ERCC sequences, using STAR version 2.5.2b. Gene counts were made using HTSEQ version 0.6.1p1. All packages were called and run through a custom Snakemake pipeline - FreezeFrame (version 4)by Coulbourn Instruments

Data analysis - R (version 4.0.2 Taking Off Again) - RStudio Version 1.1.456 - Seurat v3 by the Satija Lab (https://satijalab.org/seurat/) - HOMER (version 4.8) (http://homer.ucsd.edu/homer/motif/) - GeneAnalytics (2020) (https://geneanalytics.genecards.org/)

For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors/reviewers. We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Research guidelines for submitting code & software for further information.

DataPolicy information about availability of data

All manuscripts must include a data availability statement. This statement should provide the following information, where applicable: - Accession codes, unique identifiers, or web links for publicly available datasets - A list of figures that have associated raw data - A description of any restrictions on data availability

- The accession number for the single cell RNA-sequencing data reported in this paper is GSE152632.

2

nature research | reporting summ

aryO

ctober 2018- STRING database (https://string-db.org/) - Malacards: The human disease database (https://www.malacards.org/) - RNA-seq data from Rao-Ruiz, 2019 (NCBI GEO Accession GSE129024) - RNA-seq data from Hrvatin, 2017 (GSE102827) -RNA-seq data from Lacar, 2016 (NCBI GEO Accession Number: GSE77067) -RNA-seq data from Cho, 2016 (GSE85128) - Receptor-ligand data from Ramilowski, 2015 (http://fantom.gsc.riken.jp/5/suppl/Ramilowski_et_al_2015/)

Field-specific reportingPlease select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection.

Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences

For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pdf

Life sciences study designAll studies must disclose on these points even when the disclosure is negative.

Sample size Sample sizes (both number of mouse replicates and number of cells sequences) were determined via a combination of factors including: the total expected size of the population of active neurons in the prefrontal cortex (~1.5%), (2) cost of sequencing and mice (3) expected margin of error and confidence levels

Data exclusions No data exclusions

Replication Replication was achieved by performing the behavioral training on 4 sets of mice, on 4 different independent days. Results were consistent in all replicates (4).

Randomization Mice were randomly grouped. Mice were fed and house in the same situation just prior to exposure to various behavioral trainings

Blinding Blinding was not performed during this study as the number of mice per training per biological replicate was too few to effectively allow this

Reporting for specific materials, systems and methodsWe require information from authors about some types of materials, experimental systems and methods used in many studies. Here, indicate whether each material, system or method listed is relevant to your study. If you are not sure if a list item applies to your research, read the appropriate section before selecting a response.

Materials & experimental systemsn/a Involved in the study

Antibodies

Eukaryotic cell lines

Palaeontology

Animals and other organisms

Human research participants

Clinical data

Methodsn/a Involved in the study

ChIP-seq

Flow cytometry

MRI-based neuroimaging

Animals and other organismsPolicy information about studies involving animals; ARRIVE guidelines recommended for reporting animal research

Laboratory animals TRAP2 was generated in a 129Sv/SvJ background. For behavior experiments, they were backcrossed to C57Bl6/J for 3 generations. Mice were housed at room temperature and 40-60% humidity. We have described the usage of C57bl6J mice in the method part of the manuscript. Male mice at age between 6-7 weeks were used in this study.

Wild animals None

Field-collected samples None

Ethics oversight All animal procedures followed animal care guidelines approved by Stanford University's Administrative Panel on Laboratory Animal Care (APLAC)

Note that full information on the approval of the study protocol must also be provided in the manuscript.

3

nature research | reporting summ

aryO

ctober 2018Flow CytometryPlots

Confirm that:

The axis labels state the marker and fluorochrome used (e.g. CD4-FITC).

The axis scales are clearly visible. Include numbers along axes only for bottom left plot of group (a 'group' is an analysis of identical markers).

All plots are contour plots with outliers or pseudocolor plots.

A numerical value for number of cells or percentage (with statistics) is provided.

Methodology

Sample preparation mPFC regions were micro-dissected from live vibratome sections (300 um thick) of the prefrontal cortex. Tissue pieces were enzymatically dissociated via a papain-based digestion system (Worthington, Cat # LK003150). Briefly, tissue chunks were incubated in 1mL of papain (containing L-cysteine and EDTA), DNAse, and kyneurenic acid for 1 hour at 37C and 5% CO2. After 10 min of incubation, tissues were triturated briefly with a P1000 wide bore pipette tip and returned. Cells were triturated another 4 times (~30 each) with a P200 pipette tip over the rest of the remaining incubation time. At room temperature, cell suspensions were centrifuged at 350g for 10 min, resuspended in 1mL of EBSS with 10% v/v ovomucoid inhibitor, 4.5% v/v DNAse and 0.1% v/v kyneurenic acid, and centrifuged again. Supernatant was removed and cells 1mL ACSF was added. ACSF was composed of: 1mM KCl, 7mM Mgcl2, 0.5 mM Cacl2, 1.3 mM NaH2PO4, 110 mM choline chloride, 24mM NaHCO3, 1.3 mM Na Ascorbate, 20mM glucose and 0.6mM sodium pyruvate. Cells were passed through a 70 um cell strainer to remove debris. Hoescht stain was added (1:2000, Life Technologies, Cat #H3570) and incubated in the dark at 4C for 10 min. Samples were centrifuged (350g for 8 min at 4C) and resuspended in 0.5mL of ACSF and kept on ice for flow cytometry.

Instrument Sony SH800

Software Sony SH800

Cell population abundance tdtomato+/ Hoescht+ cells were ~0.5-2% of all total events

Gating strategy There was gating on BSC or FSC as size distribution of neurons/non-neuronal cells of interest is unknown. Only gating for a population that was double positive for Hoescht (above 1000 arbitrary fluorescent units) and tdTomato (above 10e4 arbitrary fluorescent units) was used (see Extended Data 2 for examples).

Tick this box to confirm that a figure exemplifying the gating strategy is provided in the Supplementary Information.