principles of memory organization in the monkey medial temporal lobe wendy a. suzuki center for...

Principles of Memory Organization in the Monkey Medial Temporal Lobe

Wendy A. Suzuki Center for Neural Science

New York University

24/25/32Pi

Ia

Id

Ig

Frontal Lobe Temporal Lobe InsularCortex

CingulateGyrus

ParietalLobe

V4

23/29/30

7a/LIP

STSd STG

TE

STSv

TEO

11/1246

13

DG

SubHPC

30-60%18-29%5-17%<5%

Entorhinal

ParahippocampalPerirhinal

Suzuki and Amaral, 1994

Take Home Message:• All structures of the MTL appear to

provide a similar complement of associative learning-related signals.

24/25/32Pi

Ia

Id

Ig

Frontal Lobe Temporal Lobe InsularCortex

CingulateGyrus

ParietalLobe

V4

23/29/30

7a/LIP

STSd STG

TE

STSv

TEO

11/1246

13

DG

SubHPC

30-60%18-29%5-17%<5%

Entorhinal

ParahippocampalPerirhinal

Suzuki and Amaral, 1994

Take Home Message:• All structures of the MTL appear to provide a

similar complement of associative learning-related signals.

• There appears to be a conditionally dynamic relationship between different MTL structures such that depending on task demands different MTL structure can either contribute in similar/cooperative ways or in more clearly distinct way to mnemonic functions.

Location-Scene Association TaskGaffan (1994), Murray and Wise (1996), Murray et. al., (1998)

+ ++New Scenes

(n = 2-4)

Fixation/Baseline(300 ms)

Scene/TargetPresentation

(500 ms)

Delay(700 ms)

Eye MovementResponse

Pro

babi

lity

Cor

rect

0 10 20 30 40 500

0.2

0.4

0.6

0.8

1

Behavioral Learning Curve and Delay Activity Over Time to Scene 653 R

esponse (Spikes/S

ec)

-10

0

10

20

30

40

50

Baseline Activity

Trials C110600.4 Wirth et al., 2003

Pro

babi

lity

Cor

rect

Behavioral Learning Curve and Delay Activity Over Time to Scene 386 R

esponse (Spikes/S

ec)

Trials

0

0.2

0.4

0.6

0.8

1

1 4 7 10 13 16 19 22 25 28 31 34 37

0

2

4

6

8

10

12

14

16

Wirth et al., 2003

2 Categories of Changing Cells

A. Sustained Changing cells: 56% of all changing cells

B. Baseline Sustained Changing cells: 44% of all changing cells

What about the Perirhinal cortex?

Yanike et al., 2008, In press

Selective

Hippocampus(n = 2 monkeys)

89/145 (61 %)

Changing Cells

25/89 (28%)

Sustained

Baseline Sustained

PR Cortex(n = 2 monkeys)

83/128 (65 %)

26/83 (31%)

8/26 (31%)

18/26 (69%)

14/25 (56%) 11/25 (44%)

Yanike et al., 2008, In press

Early

Late Early

Late+

+ +

+

Object 1 Object 2

Place 1

Place 2

Object-Place Association Task

Wirth et al, 2009

Fixation500ms

Cue500ms

Delay700ms

Early500ms

Late500ms

Reward ITI2000ms

++ ++ ++ ++ ++

Bar Release Period

Fixation500ms

Cue500ms

Delay700ms

Early500ms

Late500ms

Reward ITI2000ms

++ ++ ++ ++ ++

Bar Release Period

Fixation500ms

Cue500ms

Delay700ms

Early500ms

++ ++ ++ ++Reward

Release

Fixation500ms

Cue500ms

Delay700ms

Early500ms

Late500ms

++ ++ ++ ++ ++ Release No Reward

ITI

ITI

Correct Trial

Incorrect Trial

Wirth et al, 2009

50% are outcome-selective

Do hippocampal cells differentiate between correct and error trials?

Time from Release (ms)

-500 0 500 1000 1500

Nor

mal

ized

Firi

ng R

ate

(Spi

kes/

Sec

)

-4

-2

0

2

4

6

8

10

12

Time from Release (ms)

-500 0 500 1000 1500

Nor

mal

ized

Firi

ng r

ate

(Spi

kes/

Sec

)-4

-2

0

2

4

6

8

10

12 CorrectError

CorrectError

N=16 N=18

A B

Correct Up CellsN = 30

Error Up CellsN = 38

Wirth et al, 2009

A

Time from Release or Random Drop (ms)-500 0 500 1000 1500 2000

Spi

kes/

Sec

0

5

10

15

20

25

Correct

Error

Random reward

B

Lat

enc

ies

(ms)

0

200

400

600

800

Standard Delayedwith Sound

DelayedNo Sound

*

Wirth et al, 2009

++ + ++

Fixation500ms

Cue500ms

Delay700ms

Reward ITI2000ms

Fixation500ms

Cue500ms

Delay700ms

Early500ms

Late500ms

Reward ITI2000ms

++ ++ ++ ++ ++

Bar Release Period

Standard Trials

Fixation Only Trials

Time from Release (ms)

-500 0 500 1000 1500

Nor

mal

ized

Firi

ng R

ate

(Spi

kes/

Sec

)

-4

-2

0

2

4

6

8

10

12

Time from Release (ms)

-500 0 500 1000 1500

Nor

mal

ized

Firi

ng r

ate

(Spi

kes/

Sec

)-4

-2

0

2

4

6

8

10

12 CorrectError

CorrectError

N=16 N=18

A B

Correct Up CellsN = 30

Error Up CellsN = 38

Wirth et al, 2009

++ + ++

Fixation500ms

Cue500ms

Delay700ms

Reward ITI2000ms

Fixation500ms

Cue500ms

Delay700ms

Early500ms

Late500ms

Reward ITI2000ms

++ ++ ++ ++ ++

Bar Release Period

Standard Trials

Fixation Only Trials

Time from Trial End (ms)

-500 0 500 1000 1500

Firi

ng R

ate

(Spi

kes/

Sec

)

6

8

10

12

14

16

18

20

Time from End of Fixation (ms)

-500 0 500 1000 1500F

iring

Rat

e (S

pike

s/S

ec)

0

10

20

30

40Completed FixationBreake Fixation

N=9N=13 Rewarded TrialsOmit

A B

Wirth et al, 2009

Time from Release (ms)

-500 0 500 1000 1500

Nor

mal

ized

Firi

ng R

ate

(Spi

kes/

Sec

)

-4

-2

0

2

4

6

8

10

12

Time from Release (ms)

-500 0 500 1000 1500

Nor

mal

ized

Firi

ng r

ate

(Spi

kes/

Sec

)-4

-2

0

2

4

6

8

10

12 CorrectError

CorrectError

N=16 N=18

A B

Correct Up CellsN = 30

Error Up CellsN = 38

Wirth et al, 2009

Early

Late Early

Late+

+ +

+

Object 1 Object 2

Place 1

Place 2

Object-Place Association Task

Wirth et al, 2009

Fixation500ms

Cue500ms

Delay700ms

Early500ms

Late500ms

Reward ITI2000ms

++ ++ ++ ++ ++

Bar Release Period

Fixation500ms

Cue500ms

Delay700ms

Early500ms

Late500ms

Reward ITI2000ms

++ ++ ++ ++ ++

Bar Release Period

Sessions without learning

BEFORE AFTERS

elec

tivity

Inde

x

0.35

0.40

0.45

0.50

0.55

0.60

0.65CORRECT UPERROR UPCONTROL

Sessions with learning

BEFORE AFTER

Sel

ectiv

ity In

dex

0.35

0.40

0.45

0.50

0.55

0.60

0.65CORRECT UPERROR UPCONTROL

*

Sessions without learning

Differential Selectivity

-0.7 -0.5 -0.3 -0.1 0.1 0.3 0.5 0.7 0.9

Inci

denc

es (

perc

enta

ge o

f ca

ses)

0

10

20

30

40

50

60

A B

C DSessions with learning

Differential Selectivity

-0.7 -0.5 -0.3 -0.1 0.1 0.3 0.5 0.7 0.9

Inci

denc

es (

perc

enta

ge o

f ca

ses)

0

10

20

30

40

Control cellsError up cellsCorrect up cells

Control cellsError up cellsCorrect up cells

Wirth et al, 2009

24/25/32Pi

Ia

Id

Ig

Frontal Lobe Temporal Lobe InsularCortex

CingulateGyrus

ParietalLobe

V4

23/29/30

7a/LIP

STSd STG

TE

STSv

TEO

11/1246

13

DG

SubHPC

30-60%18-29%5-17%<5%

Entorhinal

ParahippocampalPerirhinal

Suzuki and Amaral, 1994

Take Home Message:• All structures of the MTL appear to provide a

similar complement of associative learning-related signals.

• There appears to be a conditionally dynamic relationship between different MTL structures such that depending on task demands different MTL structure can either contribute in similar/cooperative ways or in more clearly distinct way to mnemonic functions.

Temporal Order Task

0.3 s 0.9 s 0.3 s 0.5 s0.4 s

1.0 s > 4.0 s

Encoding Phase

Retrieval Phase

1st Cue 2nd Cue

1st Choice 2nd Choice

SmallReward

LargeReward

0.7-1.5 s

Pre-responseDelay

Inter-phaseDelay

Naya and Suzuki, SFN 2008

24/25/32Pi

Ia

Id

Ig

Frontal Lobe Temporal Lobe InsularCortex

CingulateGyrus

ParietalLobe

V4

23/29/30

7a/LIP

STSd STG

TE

STSv

TEO

11/1246

13

DG

SubHPC

30-60%18-29%5-17%<5%

Entorhinal

ParahippocampalPerirhinal

Suzuki and Amaral, 1994

Findings

• The most stimulus selective response are found in area TE and the perirhinal cortex.

• Perirhinal cortex has the strongest proportion of order and stimulus selective responses

• The hippocampus has the highest proportion of trial outcome selective responses.

Take Home Message:• All structures of the MTL appear to provide a

similar complement of associative learning-related signals.

• There appears to be a conditionally dynamic relationship between different MTL structures such that depending on task demands different MTL structure can either contribute in similar/cooperative ways or in more clearly distinct way to mnemonic functions.

Sylvia WirthMarianna Yanike

Emin Avsar

Eric Hargreaves Yuji Naya