This is a sel -a chi ed e sion o an o iginal a icle. This e sion
may di e om he o iginal in pagina ion and ypog aphic de ails.
Au ho (s):
Ti le:
Yea :
Ve sion:
Copy igh :
Righ s:
Righ s u l:
Please ci e he o iginal e sion:
CC BY 4.0
h ps://c ea i ecommons.o g/licenses/by/4.0/
Dis inc Hippocampal Oscilla ion Dynamics in T ace Eyeblink Condi ioning Task o
Re ie al and Consolida ion o Associa ions
© 2024 Kim e al.
Published e sion
Kim, Kayeon; Nokia, Mi iam S.; Pal a, Sa u
Kim, K., Nokia, M. S., & Pal a, S. (2024). Dis inc Hippocampal Oscilla ion Dynamics in T ace
Eyeblink Condi ioning Task o Re ie al and Consolida ion o Associa ions. eNeu o, 11(4),
A icle ENEURO.0030-23.2024. h ps://doi.o g/10.1523/ENEURO.0030-23.2024
2024
Dis inc Hippocampal Oscilla ion
Dynamics in T ace Eyeblink
Condi ioning Task o Re ie al and
Consolida ion o Associa ions
Kayeon Kim,
1,2
Mi iam S. Nokia,
3
and Sa u Pal a
1,4,5
1
Neu oscience Cen e , Helsinki Ins i u e o Li e Sciences, Uni e si y o Helsinki, Helsinki
FI-00014, Finland,
2
Depa men o Neu oscience, Facul y o Heal h and Medical Science,
Uni e si y o Copenhagen, Copenhagen N DK-2200, Denma k,
3
Depa men o Psychology,
Uni e si y o Jy äskylä, Jy äskylä FI-40014, Finland,
4
Cen e o Cogni i e Neu oscience,
School o Psychology and Neu oscience, Uni e si y o Glasgow, Glasgow G12 8QQ, Sco land,
and
5
Di ision o psychology, VISE, Facul y o Educa ion and Psychology, Uni e si y o Oulu,
Oulu, Os obo hnia FI-90014, Finland
Abs ac
T ace eyeblink condi ioning (TEBC) has been widely used o s udy associa i e lea ning in bo h animals
and humans. In his pa adigm, condi ioned esponses (CRs) o condi ioned s imuli (CS) se e as a
measu e o e ie ing lea ned associa ions be ween he CS and he uncondi ioned s imuli (US) wi hin
a ial. Memo y consolida ion, ha is, lea ning o e ime, can be quan i ied as an inc ease in he p opo ion
o CRs ac oss aining sessions. Howe e , how hippocampal oscilla ions di e en ia e be ween success ul
memo y e ie al wi hin a session and consolida ion ac oss TEBC aining sessions emains unknown. To
add ess his ques ion, we eco ded local ield po en ials (LFPs) om he a do sal hippocampus du ing
TEBC and in es iga ed hippocampal oscilla ion dynamics associa ed wi h hese wo unc ions. We
show ha ansien b oadband esponses o he CS we e co ela ed wi h memo y consolida ion, as indexed
by an inc ease in CRs ac oss TEBC sessions. In con as , induced alpha (8–10 Hz) and be a (16–20 Hz)
band esponses we e co ela ed wi h he success ul e ie al o he CS–US associa ion wi hin a session,
as indexed by he di e ence in ials wi h and wi hou CR.
Key wo ds: classical condi ioning; c oss- equency coupling; hippocampus; memo y; phase locking
Signi icance S a emen
T ace eyeblink condi ioning is widely used o s udy he neu al basis o lea ning. How b ain oscilla o y
signa u es o ins an aneous e ie al o associa ions di e om hose e lec ing long- e m memo y
consolida ion is no well unde s ood. We eco ded local ield po en ials om he a hippocampus
du ing condi ioning o dissocia e oscilla ion dynamics associa ed wi h hese unc ions. We show
ha a ansien , ea ly, b oadband esponse is co ela ed wi h memo y consolida ion (inc ease in
condi ioned esponses ac oss aining sessions) whe eas long-la ency sus ained alpha and gamma
oscilla ions a e associa ed wi h he pe o mance wi hin a gi en ial in a session.
In oduc ion
Memo ies a e ep esen ed by he dis ibu ed ac i i y o neu ons o ganized in o neu onal
assemblies in hippocampal–neoco ical ci cui s (Squi e, 1992;Buzsáki and Mose , 2013).
T ace eyeblink condi ioning (TEBC) has been widely used as a model sys em o s udying
he neu al basis o associa i e lea ning and memo y in a wide ange o species om mice
o humans. In TEBC, a condi ioned s imulus (CS, usually audi o y s imulus) is epea edly
pai ed wi h an eyeblink-e oking uncondi ioned s imulus (US, ai pu o elec ic shock o
Recei ed Jan. 24, 2023; e ised Ap il 4,
2024; accep ed Ap il 10, 2024.
The au ho s decla e no compe ing
financial in e es s.
Au ho con ibu ions: M.S.N. and S.P.
designed esea ch; M.S.N. pe o med
esea ch; K.K. and M.S.N. analyzed
da a; K.K., M.S.N., and S.P. w o e he
pape .
This wo k was suppo ed by unding
om Sig id Juselius Founda ion o S.P.
and Academy o Finland o M.S.N.
(g an numbe s: 275954, 321522).
Co espondence should be add essed
o Sa u Pal a a sa u.pal a@helsinki.fi
o Kim Kayeon a kayeon.kim@sund.
ku.dk.
Copy igh © 2024 Kim e al.
This is an open-access a icle
dis ibu ed unde he e ms o he
C ea i e Commons A ibu ion 4.0
In e na ional license, which pe mi s
un es ic ed use, dis ibu ion and
ep oduc ion in any medium p o ided
ha he o iginal wo k is p ope ly
a ibu ed.
Resea ch A icle: New Resea ch
Cogni ion and Beha io
Ap il 2024, 11(4). DOI: h ps://doi.o g/10.1523/ENEURO.0030-23.2024. 1 o 11
he eyelid). A success ul condi ioned esponse (CR) in esponse o he CS is dependen on lea ning, and upon la e
encoun e wi h he CS, e ie ing om memo y, o an associa ion be ween he CS and he US, ha is, con ingency de ec-
ion (P okasy, 1984;Cla k and Squi e, 1998;Thompson, 2005;Cheng e al., 2008).
Fo ming an associa ion be ween wo e en s sepa a ed in ime is a c i ical b ain unc ion (Pilkiw and Takeha a-Nishiuchi,
2018) ha depends on he hippocampus (Solomon e al., 1986;Moye e al., 1990;Weiss e al., 1996;Tseng e al., 2004).
In iguingly, ecen wo k u ilizing human in ac anial and scalp elec oencephalog am (EEG) sugges ed ha he hippocam-
pus could ep esen bo h senso y and mnemonic in o ma ion and ac as a swi chboa d be ween in e nal and ex e nal ep-
esen a ions (T ede e al., 2021). In line, pa ien s wi h medial empo al lobe (MTL) lesions and memo y impai men also
exhibi ed diminished ma ke s o conscious pe cep ion (U goli es e al., 2018) and showed a enua ed and empo ally
mo e dispe sed esponses o isual s imuli (Rebe e al., 2017), sugges ing ha hippocampal ci cui s also ep esen sen-
so y s imulus p ope ies (K eiman e al., 2002).
A c ucial ole in memo y o ma ion and he ini ial encoding o in o ma ion in o neu al ep esen a ions is played by he a
(θ,4–12 Hz) oscilla ions ( o e iews see Buzsáki, 2002;Colgin, 2013,2016). Hippocampal θoscilla ions a e abundan du -
ing explo a o y beha io in oden s, and hey o ganize bi s o in o ma ion in o cohe en en i ies ( o an example o θand
place cells, see O’Kee e and Recce, 1993). In e es ingly, du ing TEBC, hippocampal θoscilla ions inc ease du ing he
ace pe iod sepa a ing he CS o se and he US onse , bo h in abbi s (Nokia e al., 2009,2015) and in a s (Nokia e
al., 2012). The CS- ela ed θoscilla ions in he hippocampus and connec ed b ain s uc u es a e hough o eflec he
encoding and e ie al o he CS–US con ingency (Wikg en e al., 2010;Nokia and Wikg en, 2013;Su e e al., 2019).
Addi ionally, be a (β;12–30 Hz) and gamma (γ) band (30–40 Hz) oscilla ions a e also p ominen in he hippocampus
(Colgin e al., 2009), wi h γoscilla ions synch onizing he ac i i y in he hippocampal–en o hinal co ical (EC) loop and
ela ing o encoding and e ie al o associa ions (Iga ashi e al., 2014). Whe eas θoscilla ions a e p oposed o allow
chunking o g ouping o in o ma ion, βand γoscilla ions a e p oposed o eflec he in o ma ion pe se. In his con ex ,
c oss- equency phase–ampli ude coupling (PAC) be ween he θphase and β/γoscilla ion ampli udes is hough o enable
he chunking o senso y in o ma ion (Colgin, 2016).
Despi e ex ensi e esea ch on he neu al basis o lea ning using TEBC, he ela ionship be ween mapping s imulus con-
ingencies wi hin a ial (encoding and e ie al o he CS–US associa ion) and lea ning, ha is, en o cing hei associa ions
ac oss aining sessions (memo y consolida ion), emains unclea . While he ole o oscilla ion dynamics in memo y unc-
ions is well es ablished, he di e ences in hese dynamics be ween he o ma ion o mnemonic associa ions be ween
he CS and US wi hin a ial and memo y consolida ion o s imulus con ingencies ac oss aining sessions emains
unexplo ed. Wi h aining on he CS–US con ingency, he p obabili y o a CR o each CS inc eases g adually o e ime.
Howe e , e en wi h ex ensi e aining, some CS p esen a ions ail o elici a CR. In his s udy, we hypo hesized ha
dis inc spa io empo al signa u es would cha ac e ize mapping CS–US con ingencies wi hin sessions (indexed as CR
o he ial) and memo y consolida ion ac oss aining sessions (indexed as an inc easing p opo ion o CRs o e
sessions). To his end, we eco ded local field po en ials (LFPs) om he do sal hippocampus in eely mo ing adul heal hy
Sp ague Dawley male a s du ing classical TEBC and es ima ed oscilla ion dynamics o hippocampal subfields using
s a e-o - he-a analysis app oaches.
Ma e ials and Me hods
Animals. Eigh adul heal hy male Sp ague Dawley a s (Ha lan Labo a o ies/En igo, weighing ∼300 g, ∼10 weeks)
we e used as subjec s. Food and wa e we e a ailable ad libi um, wi h oom empe a u e and humidi y main ained a
21 ± 2°C and 50 ± 10%, espec i ely. The a s we e kep unde a 12 h ligh /da k cycle. All p ocedu es and expe imen s
we e conduc ed du ing he ligh cycle. The s udy was conduc ed in acco dance wi h Di ec i e 2010/63/EU o he
Eu opean Pa liamen and o he Council on he ca e and use o animals o esea ch. The expe imen s we e app o ed
by he Animal Expe imen Boa d o he Regional S a e Adminis a i e Agency o Sou he n Finland. The ARRIVE guidelines
(h p://a i eguidelines/o g/a i e-guidelines) we e ollowed.
TEBC and analysis o he eyeblink esponse. Da a we e collec ed du ing a TEBC ask (Fig. 1A) using LabVIEW (Na ional
Ins umen s). The animals we e condi ioned using whi e noise (75 dB, 200 ms) as a CS and a 100 Hz bu s o 0.5 ms bipola
pulses o pe io bi al shocks (100 ms) as a US. The ampli ude o he US was adjus ed indi idually o each animal o elici a blink
esponse, ha is, he uncondi ioned esponse in 100% o he ials. Each ial s a ed wi h he 200 ms CS p esen a ion,
ollowed by a 500 ms s imulus- ee ace pe iod, and hen a 100 ms shock US. Each animal pe o med eigh
sessions, each consis ing o 60 ials. O e ime, he animals s a ed o blink in esponse o he CS; ha is, hey acqui ed
he CR.
Eyeblinks we e de ec ed om he EMG signals o fline o de e mine he pe cen age o CRs (he e defined as he hi a e,
HR), as in Nokia e al. (2017). In b ie , o each ial, he esponse was defined as a CR i he signal exceeded a h eshold o
mean + 3 s anda d de ia ions (SD) wi hin he las 200 ms o he ace pe iod (Fig. 1A). Tha is, he eyelid s a ed o close
immedia ely be o e he onse o US. HR was defined as he p opo ion o ials wi h a CR (Fig. 1D). We hen used epea ed
measu es ANOVA ( m ANOVA) o es ima e he change in HR ac oss he eigh sessions as well as ac oss he animals.
Reac ion imes (RTs) we e compu ed o he ials wi h CRs.
Resea ch A icle: New Resea ch 2 o 11
Ap il 2024, 11(4). DOI: h ps://doi.o g/10.1523/ENEURO.0030-23.2024. 2 o 11
Su ge y. Ra s we e anes he ized wi h an in ape i oneal injec ion o pen oba bi al (60 mg/kg) and ea ed o pain wi h
ca p o en (5 mg/kg, s.c.) and bup eno phine (0.03 mg/kg, s.c.). Using a s e eo ac ic ame, we implan ed wo ou -wi e
elec odes (Fo m a -Insula ed Nich ome, ba e diame e 50 µm, no. 762000, A-M Sys ems) ch onically o eco d LFPs
om he do sal hippocampus. The wi es we e glued oge he wi h a ip sepa a ion o 200–250 µm. The bundles
we e implan ed wi h he lowes elec ode ip a he den a e gy us (DG; 3.6–4.5 mm pos e io , 1.5–2.2 mm la e al, and
3.6–4.0 mm below b egma; Fig. 1C). Skull c ews se ed as he e e ence (11 mm pos e io and 2 mm la e al o
b egma) and g ound (4 mm an e io and 2 mm la e al o b egma). To s imula e he eyelid and eco d elec omyog aphy
(EMG) du ing TEBC, wo bipola elec odes made o s ainless s eel wi e insula ed wi h Teflon (ba e diame e 127 µm)
we e implan ed h ough he uppe igh eyelid. Finally, he en i e cons uc ion was secu ed in place using den al
ac ylic cemen . A e he su ge y, each a was allowed o eco e o a leas 1 week and was medica ed o pain
wi h bup eno phine.
Reco dings. A low-noise wi ed p eamplifie (10×) was di ec ly a ached o he elec ode connec o in he a ’s head. The
LFP signals we e fil e ed om 1 o 5,000 Hz, amplified 50×, digi ized a 20 kHz, and hen low-pass fil e ed a 500 Hz.
Finally, all signals we e s o ed a a sampling a e o 2 kHz (USB-ME-64, Mul i Channel Sys ems).
His ology. Ra s we e killed by exposu e o a ising concen a ion o CO
2
and hen decapi a ed. The loca ions o he elec-
ode ips in he b ain we e ma ked by passing a DC anodal cu en (200 mA, 5 s) h ough hem. The b ain was hen
emo ed, fixed in a 4% pa a o maldehyde solu ion, and co onally sec ioned wi h a ib a ome (Leica VT1000). The slices
we e s ained wi h P ussian blue and c esyl iole . The elec ode ip loca ions we e de e mined wi h he help o a con en-
ional ligh mic oscope and a b ain a las (Paxinos and Wa son, 1998).
Signal p ep ocessing, e- e e encing, and fil e ing. T ials wi h la ge LFP fluc ua ions due o mo emen o de ice- ela ed
a i ac s we e excluded i hey exceeded 1,500 µV cu o , esul ing in a ejec ion a e o 0.6%. To emo e he
s imulus-induced spike a i ac , aw signals be ween −2.5 and 1.5 ms om CS onse and o se we e in e pola ed wi h
an alpha blend ac ion o 0.45. Cu en sou ce densi y (CSD) p ofiles we e hen calcula ed o e- e e encing using he
Laplacian e- e e encing me hod wi h adjacen channels (Mi zdo , 1985). The signal was hen fil e ed be ween 3 and
480 Hz using a fini e impulse esponse (FIR) fil e wi h a loga i hmically scaled inc emen o equency and u ilizing a com-
bina ion o high-pass and low-pass fil e pai s (high-pass: 0.6 s op band and low-pass: 1.4 s op band, 60 dB a enua ion).
The fil e ed signals we e u he Hilbe ans o med o ob ain phase and ampli ude ime se ies (Fig. 1E).
Da a analysis o oscilla ion dynamics. We compu ed local oscilla ion dynamics using measu es o oscilla ion ampli udes
and in e ial cohe ence (ITC) wi h a phase-locking ac o (J. M. Pal a e al., 2005) sepa a ely o he hilus and fissu e. The
ampli ude and ITC ime se ies we e a e aged ac oss ials o each condi ion om −600 o 600 ms om CS onse o each
fil e ed equency. The a e age baseline alues om −600 o −100 o CS onse we e hen sub ac ed om pos s imulus
alues. Phase ime se ies da a we e u he used o compu e in e a eal synch oniza ion be ween he fissu e and hilus using
he phase-locking alue (PLV; Fig. 1F). PLV was no malized by di iding he alue by he a e aged PLV alue ob ained by
shu fling ials wi hin a condi ion and measu emen si e 100 imes o con ol o he con ibu ions o s imulus-d i en a i-
ficial synch oniza ion (Hi onen e al., 2018).
To es ima e coupling ac oss he equencies, PAC was compu ed wi h PLV be ween he phase o he slow oscilla ion
(low equency, LF) and he phase o he ampli ude en elope o he high- equency (HF) oscilla ion o he m:n a ios
be ween2and9(Fig. 1G;J. M. Pal a e al., 2005;Siebenhühne e al., 2016). Phase ans e en opy (phase TE; Lobie e
al., 2014) was compu ed o iden i y di ec ionali y in he na ow-band oscilla o y signals. Phase TE was de i ed om he ins an-
aneous phase ime se ies o he signal X
hilus
( )andY
fissu e
( ) exp essed as θ
hilus
( )andθ
fissu e
( )(Fig. 1E). To c ea e a bias- ee
measu e, we compu ed he di e en ial TE (dTE) de i ed by pTE
hil→fis
−pTE
fis→hil
and used his measu e o u he analysis
(Fig. 1H).
S a is ical analysis. To a oid di e en signal- o-noise a ios (SNRs) om influencing he esul s, he numbe o ials
be ween condi ions (CR s no-CR; highes HR s lowes HR) was equalized be o e s a is ical analyses by andomly selec -
ing ials o ma ch he minimum numbe o ials wi hin a session. S a is ically significan changes in oscilla ion ampli udes,
ITC, and in e a eal synch ony we e achie ed by de i ing null dis ibu ions (n= 20,000) using andom flips wi h a p obabili y
o 0.5 (Mon e Ca lo p< 0.025, wo- ailed). To co ec o mul iple compa isons, we used he Benjamini–Hochbe g p oce-
du e o each analysis ime window.
To assess he significan di e ences be ween CR and no-CR ials, we fi s es ima ed he di e ences in neu onal dynamics
wi hin each animal. We hen used a pai ed es o iden i y he - h eshold (p< 0.025, wo- ailed) o each ime and equency
and ob ained he -sum obse ed alue based on empo al adjacency. The maximum -sum obse ed was es ed agains he
-sum null dis ibu ion om 1,000 su oga es (Mon e Ca lo p< 0.05) using clus e -based pe mu a ion s a is ics, which accoun
o mul iple compa isons in s a is ical analysis (Ma is and Oos en eld, 2007). The same clus e ing pe mu a ion s a is ics we e
pe o med o compa e he sessions wi h he lowes and highes HR. The null dis ibu ions we e de i ed by andomizing ials
o compa ison be ween CR and no-CR ials and by andomizing he sessions o compa ing lea ning.
Resea ch A icle: New Resea ch 3 o 11
Ap il 2024, 11(4). DOI: h ps://doi.o g/10.1523/ENEURO.0030-23.2024. 3 o 11
Indi idual-le el s a is ical analysis o oscilla ion ampli ude di e ences was compu ed using a clus e -based pe mu a-
ion me hod by andomizing he CR and no-CR ials and sessions o he lowes %HR and highes %HR, espec i ely, o
ob ain su oga e mean alues. Indi idual-le el s a is ical analysis o ITC alues was ob ained by compa ing each ITC
alue wi h a su oga e null dis ibu ion [Mon e Ca lo p< 0.05, alse disco e y a e (FDR) co ec ed].
To di ec ly s udy he main e ec s and he in e ac ion e ec o success ul memo y e ie al wi hin a session (CR
s no-CR) and consolida ion ac oss TEBC aining sessions, we conduc ed wo-way m ANOVA on bo h he oscilla ion
ampli udes and on he ITC wi hin he ime- equency egions o in e es (p< 0.025). The s a is ical analysis o 1:1 in e a eal
synch oniza ion be ween condi ions was pe o med using a wo-sample es o he ime- equency egions o in e es .
The p esence o significan PAC in he pos s imulus pe iods was assessed pos hoc using a one-sample es ac oss all
subjec s and sessions o each pai o low and high equencies, analysis ime windows, and lamina pai (FDR co ec ed).
Fo s a is ical analysis o phase TE, su oga e da a we e ob ained by shu fling ials 1,000 imes and compa ing he su -
oga e dis ibu ions wi h he empi ical da a o es he s a is ical significance agains a null hypo hesis o no in o ma ion
ans e be ween he hilus and fissu e (Mon e Ca lo p< 0.05, FDR co ec ed).
The code o phase TE is eely a ailable a he Pal a Gi Hub eposi o y (h ps://gi hub.com/pal alab/pal alab).
Resul s
Beha io al da a
Associa i e lea ning was assessed using a TEBC ask (Fig. 1A, see Ma e ials and Me hods o mo e de ails). HR ( he
pe cen age o CRs) significan ly inc eased as a unc ion o he session (Pea son’s = 0.49; p= 4.51 × 10
−5
;Fig. 1B), bu
Figu e 1. O e iew o he app oach. A, Task schema ics and an example o EMG ace. In he TEBC ask, a one was used as CS ollowed by he US
(pe io bi al shock). Eyeblinks we e eco ded wi h EMG. The g ay a ea indica es he ime window when eyeblinks we e coun ed as a CR. B, HR o he
CRs as a unc ion o session wi h mean and SEM. The ci cles ep esen da a om indi idual a s (n= 8). C, A g aphical illus a ion o eco ding elec ode
placemen in he a hippocampus a ge ing fissu e ( , uppe ) and hilus (h, lowe ). D,E, An example o a b oadband aw signals eco ded om fissu e (blue)
and hilus (o ange) o a single ial and o hei na ow-band fil e ing and ans o ma ion o complex alues o ob ain ampli ude and phase ime se ies.
F, In e a eal synch oniza ion be ween fissu e and hilus was es ima ed by compu ing PLV be ween signals and compa ing he empi ical alues o a
su oga e dis ibu ion. G, The n:m PAC was compu ed be ween he phase o a lowe equency (n) and ampli ude o he highe equency (m). The aces
show he θphase in hilus and he γampli ude en elope in fissu e (black). H, Di e en ial phase ans e en opy (phase dTE) was es ima ed om he
phase-lag in o ma ion om each egion using an analysis window o 0.5 s du ing he pos -CS pe iod. (See Ma e ials and Me hods o mo e de ails.)
Resea ch A icle: New Resea ch 4 o 11
Ap il 2024, 11(4). DOI: h ps://doi.o g/10.1523/ENEURO.0030-23.2024. 4 o 11

sessions wi h he bes pe o mance a ied ac oss animals. The mean RT o CRs a e aged ac oss all sessions and animals
was 558.72 ms wi h an SEM o 5.7 ms. To con ol he con ounding e ec s o indi idual di e ences, we also conduc ed m
ANOVA, confi ming a significan change in HR ac oss sessions (F
(7,49)
= 5.72; p= 7.17 × 10
−5
) and indica ing ha he a s
lea ned o an icipa e he shock US and shield he eye wi h a CR as aining p og essed.
Figu e 2. Spa io empo al dynamics o hippocampal oscilla ions du ing TEBC. A, A g oup-a e aged b oadband e oked esponse o he CS, a e aged
ac oss all ials, sessions, and animals om fissu e (le ) and hilus ( igh ). Shade indica es SEM. Dashed lines indica e CS onse and mean (±SEM) and
la ency o CR (558.7 ± 5.7 ms), espec i ely. B, Baseline-co ec ed ime- equency ep esen a ion (TFR) o oscilla ion ampli udes in esponse o he CS
sepa a ely o fissu e (le ) and hilus ( igh ). Oscilla ion ampli udes we e a e aged ac oss ials, sessions, and hen ac oss all a s (n= 8). The ho izon al
ba below he panel indica es analysis pe iods ela i e o CS onse ; 0–0.1 s (black), 0.1–0.3 s (g ay), 0.3–0.5 s (ligh g ay). C, Oscilla ion ampli udes a e -
aged ac oss h ee ime windows. Ho izon al colo ed ba s deno e s a is ical significance a p< 0.025 (see Ma e ials and Me hods o de ails). D,E, In e ial
cohe ence (ITC) as es ima ed wi h PLF as in Band C. Ho izon al ba s deno e s a is ical significance a p< 0.025.
Resea ch A icle: New Resea ch 5 o 11
Ap il 2024, 11(4). DOI: h ps://doi.o g/10.1523/ENEURO.0030-23.2024. 5 o 11
Spa io empo al dynamics o oscilla ions in esponse o he CS
Figu e 2Ashows g oup-a e aged CSD e- e e enced LFP aces om he hippocampal subfields (fissu e, hilus).
Consis en wi h he fissu e encompassing dend i es o he DG g anule cells ecei ing he main inpu om he EC,
CS-e oked esponses we e obus in he fissu e bu no in he hilus. In addi ion, he e oked esponses we e eflec ed
in bo h oscilla ion ampli udes and ITC, showing ansien b oadband esponses in he fissu e bu also in he hilus
(Fig. 2B–E; Mon e Ca lo p< 0.025, wo- ailed, FDR co ec ed). This ime-locked ac i i y was ollowed by induced alpha
[α,8–12 Hz, no e ha his band is usually defined as he a (θ) in oden li e a u e] and high-gamma (high-γ,50–100 Hz)
band esponses in he fissu e (Fig. 2B,C, g ay, ligh g ay; Mon e Ca lo p< 0.025, wo- ailed, FDR co ec ed). In addi ion,
concu en supp ession o θ(he e 4–6 Hz) and β/γband (23–40 Hz) ampli udes was obse ed bo h in he fissu e and hilus
(Fig. 2B,C, ligh g ay ace; Mon e Ca lo p< 0.025, wo- ailed, FDR co ec ed). Despi e simila spa io empo al oscilla o y
p ofiles in he hilus and fissu e, induced ampli udes we e s onge in he fissu e, which is he p ima y si e o EC inpu s ia
he pe o an pa hway (Scha man, 2016).
Dis inc spa io empo al pa e ns o he e en ion o s imulus con ingencies be ween CS–US and consolida ion
ac oss aining sessions
We nex explo ed whe he dis inc spa io empo al pa e ns associa ed wi h he e ie al o s imulus con ingencies
be ween he CS and he US and wi h memo y consolida ion ac oss he daily aining sessions could be iden ified
(Fig. 3A–D). To examine his, we u ilized he eyeblink–CR as an app oxima ion o he con ingency de ec ion and e ie al
o he CS–US associa ion, compa ing ials wi h CR o hose wi hou CR (=no CR) wi hin he session. A e he ini ial ials
du ing which ep esen a ions o he CS and US a e o med, CR is hough o depend on e ie ing he associa ion be ween
he CS and he US as well as planning and execu ing he subsequen mo o esponse (P okasy, 1984). To assess he
Figu e 3. Dis inc spa io empo al p ofiles o hippocampal oscilla ions. A, Oscilla ion ampli udes ( op panel) and ITC (bo om panel) o he con as be ween
CR and no-CR ials. The colo scale indica es he ampli ude ( op) and ITC (bo om) o each ime- equency elemen . The significan clus e s o ime and
equency (Mon e Ca lo p< 0.05) a e highligh ed wi h anspa ency while nonsignifican obse a ions a e non anspa en . Dashed lines indica e mean RT.
Solid lines indica e CS onse . B, The same as in A, bu o he con as be ween sessions wi h he lowes and he highes HR. C, Indi idual le el s a is ics o
he con as be ween CR and no CR and o D. The lowes e sus highes HR. Colo scale indica es he ac ion o indi iduals ha show significance in each
ime- equency elemen (Mon e Ca lo p< 0.050). E,F, Two-way ANOVA. E, The ep esen a ion o significan ime and equencies deno ing he main e ec s
o CR e sus no-CR (in ed) and Highes %HR e sus Lowes %HR (in blue), along wi h he in e ac ion o he wo ac o s (in pink) on ampli ude (ANOVA p<
0.025). Fissu e (le ), hilus ( igh ). Black lines deno e CS onse F. The same con en ion as in Ebu illus a es he main e ec s o he ANOVA ac o s on ITC.
Resea ch A icle: New Resea ch 6 o 11
Ap il 2024, 11(4). DOI: h ps://doi.o g/10.1523/ENEURO.0030-23.2024. 6 o 11
spa io empo al pa e ns o oscilla ions ela ed o success ul con ingency de ec ion, we compu ed oscilla ion ampli udes
and ITC sepa a ely o ials wi h and wi hou CR o each animal and condi ion. Sus ained ampli udes in αand βbands a
∼300 ms om he CS onse we e s onge o he CR ials han o he no-CR ials in bo h he fissu e and o a lesse ex en
in he hilus (Fig. 3A; Mon e Ca lo p< 0.050). In con as , he e was no no iceable di e ence be ween he CR and no-CR
ials in he ITC (Fig. 3A, bo om). In addi ion o g oup-le el significance, he esul s we e u he ep oduced by pe o ming
s a is ical analysis a he indi idual le el (Fig. 3C; Mon e Ca lo p< 0.050).
We hen es ed whe he we could dissocia e a sepa a e spa io empo al pa e n o p edic ing lea ning ac oss he daily
aining sessions by compa ing he da a om he session wi h he highes and lowes %HR o each animal. T ansien
b oadband ampli udes and ITC we e s onge o he session wi h he highes %HR han o he lowes %HR (Fig. 3B;
Mon e Ca lo p< 0.050). The esul s we e u he ep oduced by pe o ming s a is ical analysis a he indi idual le el
(Fig. 3D; Mon e Ca lo p< 0.050).
Nex , we pe o med an m ANOVA o p obe whe he he di e en oscilla o y pa e ns associa ed wi h he CR and
no-CR ials and memo y consolida ion ac oss sessions in he p e ious analysis we e s a is ically significan ly di e en .
This analysis confi med ha sus ained αand βband ampli udes we e s onge o he CR e sus no-CR di e ence
(α,F
(1,127)
= 25.94, p= 1.44 × 10
−6
;β,F
(1,127)
= 13.12, p= 4.40 × 10
−4
) han hose o he memo y consolida ion ac oss
sessions in he fissu e (α,F
(7,127)
= 1.19, p= 0.32; β,F
(7,127)
= 0.88, p= 0.53; Fig. 3E, le panel). Simila ends we e
obse ed in he hilus (α, CR s no-CR, F
(1,127)
= 10.1, p= 0.002; memo y consolida ion, F
(7,127)
= 1.39, p= 0.21; Fig. 3E,
igh panel).
In con as , he ansien γband ampli ude esponse was s onge o memo y consolida ion (fissu e, F
(1,31)
= 8.23,
p= 0.008; hilus, F
(1,31)
= 13.17, p= 0.001; Fig. 3E) han ha o CR s no-CR (fissu e, F
(1,31)
= 4.23 × 10
−4
,p= 0.98; hilus,
F
(1,31)
= 0.87, p= 0.36), whe eas he e we e no significan e ec s in he lowe equencies despi e he s a is ical di e ences
shown in Figu e 3B. Howe e , he ansien αband ITC exhibi ed a s onge associa ion wi h memo y consolida ion ac oss
TEBC sessions in he fissu e (F
(1,31)
= 10.75; p= 0.003) compa ed wi h ha in CR s no-CR ials (F
(1,31)
= 0.24; p= 0.63;
Fig. 3F, le panel). An in e ac ion e ec was obse ed be ween e ie al o he CS–US con ingency and memo y consol-
ida ion ac oss he daily aining sessions in he hilus (F
(1,31)
= 10.85; p= 0.003; Fig. 3F, igh panel). This analysis confi med
ha he e ie al o s imulus con ingencies be ween he CS and he US and memo y consolida ion ac oss he daily aining
sessions a e associa ed wi h dis inc spa io empo al pa e ns o neu onal oscilla ions.
Figu e 4. In e a eal synch oniza ion and di ec ed in e ac ions be ween hilus and fissu e. A, G oup-a e aged in e a eal synch oniza ion be ween hilus and
fissu e es ima ed wi h PLV. Da a a e a e aged ac oss ials, sessions, and hen ac oss all a s (le ). B, Same as in Abu p esen ing PLV o he con as
be ween CR and no-CR ials (le ) and he con as be ween he highes and he lowes HR ( igh ). The anspa ency highligh s he ime and equencies
wi h a significan di e ence. The sca e plo s below show indi idual a da a, wi h ed ho izon al lines indica e he median, (pai ed es , p< 0.05).
C, G oup-a e aged di e en ial phase ans e en opy (dTE) ac oss sessions and a s du ing a 0.5 s pos -CS pe iod. G een ma ks indica e significan
dTE om hilus o fissu e, blue ma k indica es significan dTE om fissu e o hilus. D, Dis ibu ion o mean dTE de i ed om su oga e da a by ial shu fling.
The g een o blue a ows in each subplo indica e he mean dTE o he empi ical da a.
Resea ch A icle: New Resea ch 7 o 11
Ap il 2024, 11(4). DOI: h ps://doi.o g/10.1523/ENEURO.0030-23.2024. 7 o 11
Phase synch oniza ion and di ec ional in e ac ions among he hippocampal subfields
In he hippocampus, in o ma ion flows om he EC o he dend i es o he DG g anule cells and CA1 py amidal cells
lining he fissu e (pe o an pa h) and om he DG o CA3 py amidal cells ia he hilus (mossy fibe s). F om he hilus, in o -
ma ion eeds back o he CA1 py amidal cell dend i es lining he fissu e (Fig. 1C). To s udy he in o ma ion flow be ween
he hippocampal subfields, we compu ed he phase synch oniza ion o oscilla ions be ween he hilus and fissu e. Robus
phase synch ony be ween he hilus and fissu e was ound in he αband and in he high-γband (Fig. 4A, igh ; Mon e Ca lo
p< 0.010, FDR co ec ed) despi e he lack o concu en oscilla ion ampli ude inc eases. In e es ingly, o local oscilla-
ions, sus ained α-βband (9–16 Hz) phase synch oniza ion be ween he hilus and fissu e was significan ly highe o
he highes %HR e sus lowes %HR (Fig. 4B, igh ; p< 0.05), whe eas he ea ly ansien αand βband phase synch o-
niza ion was s onge o CR e sus no-CR ials (Fig. 4B, le ; p< 0.05).
To u he es ablish whe he oscilla o y in e ac ions would be di ec ional as p edic ed by he eed- o wa d in o ma ion
flow om he fissu e o he hilus o eedback p ocesses om he hilus o he fissu e, we es ima ed he di ec ionali y o
coupling using pTE (Lobie e al., 2014), ocusing on he equencies showing significan in e a eal synch ony.
Di e en ial TE (dTE) was de i ed by he di e ence be ween pTE
hil→fis
and pTE
fisl→hil
ac oss all sessions and a s. In he
αand βbands, in o ma ion flow was significan om he hilus o he fissu e (Fig. 4C,D, g een ma ks; Mon e Ca lo
p< 0.050), whe eas oscilla ions a high-γ(80Hz)bandshowed heopposi edi ec ion, om hefissu e o he hilus
(Fig. 4C,D,bluema ks;Mon eCa lop< 0.050).
Nes ed α-γoscilla ions associa ed wi h lea ning ac oss sessions
We hen examined he PAC be ween oscilla ions ac oss equencies and lamina pai s. Specifically, we ocused on he
coupling o αphase (9 Hz) wi h highe equencies. Robus α:γPAC cha ac e ized he LFP signal wi hin he hilus and
be ween he hilus and fissu e ac oss all windows (Fig. 5A). PAC did no di e be ween CR and no-CR ials (Fig. 5B); how-
e e , he e was a no able inc ease in α:γPAC du ing sessions wi h he highes %HR compa ed wi h he lowes %HR ses-
sions be ween 100 and 500 ms om CS onse (Fig. 5C; pai ed es ; p< 0.05, co ec ed), showing ha α:γs eng hens as a
unc ion o lea ning.
Discussion
TEBC has been widely used o s udy neu onal mechanisms o lea ning in humans (Cason, 1922), abbi s (Schneide man
e al., 1962), and oden s (Weiss e al., 1996;McEch on and Dis e ho , 1999;Tseng e al., 2004). In his s udy, we used a
classical TEBC ask in a s o in es iga e hippocampal oscilla ion dynamics linked o he e ie al o s imulus con ingen-
cies wi hin a ial (app oxima es as CR) om hose dynamics associa ed wi h memo y consolida ion, ha is, he imp o e-
men in pe o mance ac oss he aining sessions. Ou esul s demons a e ha di e en spa io empo al pa e ns o
hippocampal oscilla ions eflec hese unc ions. T ansien CS ime-locked esponses we e co ela ed wi h memo y con-
solida ion (lea ning ac oss sessions), bu no wi h he wi hin-session e ie al o s imulus con ingencies. In con as ,
induced θ/αand γband oscilla ions a la e ime windows we e co ela ed wi h he success ul e ie al o he CS–US con-
ingency wi hin sessions, bu no wi h memo y consolida ion ac oss he aining sessions. These esul s demons a e ha
lea ning subp ocesses a e associa ed wi h dis inc oscilla ion dynamics.
T ansien ea ly b oadband ac i i y p edic s lea ning ac oss sessions
The ea ly ansien b oadband esponse was simila o ha ound in he senso y co ices in humans o pe cep ion (S.
Pal a e al., 2005;Pal a e al., 2011;Hi onen and Pal a, 2016;Julku e al., 2021) and sho - e m memo y encoding
(Pal a e al., 2011), as well as in he p ima y audi o y co ex in guinea pigs (Voig e al., 2018). The ansien γband ampli ude
esponse, and θ-αITC, p edic ed memo y consolida ion (imp o ed pe o mance) ac oss he aining sessions. This esul
ag ees wi h p e ious findings o θband phase a CS onse p edic ing hippocampal esponses and lea ning in TEBC in ab-
bi s (Seage e al., 2002;Nokia e al., 2015) and in human episodic memo y asks (G i fi hs e al., 2021). The ea ly la ency o
his ansien esponse sugges s ha i migh eflec he encoding o he CS ea u es and eed- o wa d p ocessing o sen-
so y in o ma ion (Lamme and Roel sema, 2000), a he han sho - e m memo y o he CS, e ie ing he CS–US associa-
ion om memo y, o he mo o esponse planning. This suppo s he idea ha imp o ed beha io al pe o mance ac oss
he aining sessions is due o s eng hened s imulus ep esen a ions o CS, which leads o memo y consolida ion o e
ime. This is also suppo ed by he in e ac ion e ec be ween memo y consolida ion and con ingency de ec ion in he hilus.
Wi hin- ial e ie al o s imulus con ingencies is associa ed wi h sus ained oscilla ions
Beyond he encoding o he physical en i onmen , such as spa ial mapping du ing na iga ion o explo a ion (O’Kee e
and Dos o sky, 1971;Mo gan e al., 2011), he hippocampus is also implica ed in ep esen ing and encoding isual in o -
ma ion in p ima es (Lee e al., 2012;Ju as e al., 2013;Zeidman e al., 2015). In his s udy, sus ained αand γband
esponses we e co ela ed wi h success ul e ie al o CS–US con ingencies, bu no wi h lea ning ac oss daily aining
sessions. This indica es ha p ocesses ela ed o e ie ing he CS–US associa ion a e dis inc om hose media ing
imp o emen s in lea ned beha io o e ex ended pe iods (hou s, days). Ou s udy dissocia ing hese wo p ocesses in
a classical TEBC ask p o ides a no el iew o he ole o concu en θand γoscilla ions, whe e he adi ional iew has
Resea ch A icle: New Resea ch 8 o 11
Ap il 2024, 11(4). DOI: h ps://doi.o g/10.1523/ENEURO.0030-23.2024. 8 o 11