A model of pattern completion based on the CA3 recurrent synapse - - PowerPoint PPT Presentation

▶
a model of pattern completion based on the ca3 recurrent
SMART_READER_LITE
LIVE PREVIEW

A model of pattern completion based on the CA3 recurrent synapse - - PowerPoint PPT Presentation

Dec 29, 2022 19 likes •344 views

A model of pattern completion based on the CA3 recurrent synapse Jose Guzman, Michael Frotscher and Peter Jonas The neural structure for pattern completion Histologie du Systeme Nerveus de lHomme et des Vertebres, Vols. 1 & 2. A.

slide-1
SLIDE 1

A model of pattern completion based on the CA3 recurrent synapse

Jose Guzman, Michael Frotscher and Peter Jonas

slide-2
SLIDE 2

The neural structure for pattern completion

Histologie du Systeme Nerveus de l’Homme et des Vertebres,

  • Vols. 1 & 2. A. Maloine. Paris. 1911. Ramón y Cajal.

CA3-CA3 synapses Guzman, Frotscher and Jonas page 2 of 28

slide-3
SLIDE 3

Synaptic basics of pattern completion

Ramón y Cajal, 1911

12,000 46 3,600 perforant path recurrent collaterals mossy fiber

adapted from Amaral et al., 1990

Auto-associative network

An autoassociative network can perform pattern completion when it associates the input to the principal cells with their own output (Marr, 1971)

1

Storage

2

Activation

3

Recall

CA3-CA3 synapses Guzman, Frotscher and Jonas page 3 of 28

slide-4
SLIDE 4

Synaptic bases of pattern completion

Ramón y Cajal, 1911

12,000 46 3,600 perforant path recurrent collaterals mossy fiber

adapted from Amaral et al., 1990

Auto-associative network

An autoassociative network can perform pattern completion when it associates the input to the principal cells with their own output (Marr, 1971)

1

Storage

2

Activation

3

Recall

1 1 1 CA3-CA3 synapses Guzman, Frotscher and Jonas page 4 of 28

slide-5
SLIDE 5

Synaptic bases of pattern completion

Ramón y Cajal, 1911

12,000 46 3,600 perforant path recurrent collaterals mossy fiber

adapted from Amaral et al., 1990

Auto-associative network

An autoassociative network can perform pattern completion when it associates the input to the principal cells with their own output (Marr, 1971)

1

Storage

2

Activation

3

Recall

1 CA3-CA3 synapses Guzman, Frotscher and Jonas page 5 of 28

slide-6
SLIDE 6

Synaptic bases of pattern completion

Ramón y Cajal, 1911

12,000 46 3,600 perforant path recurrent collaterals mossy fiber

adapted from Amaral et al., 1990

Auto-associative network

An autoassociative network can perform pattern completion when it associates the input to the principal cells with their own output (Marr, 1971)

1

Storage

2

Activation

3

Recall

1 1 1 CA3-CA3 synapses Guzman, Frotscher and Jonas page 6 of 28

slide-7
SLIDE 7

Synaptic bases of pattern completion

1 Macro-connectomics CA3–CA3 connectivity?

CA3-CA3 synapses Guzman, Frotscher and Jonas page 7 of 28

slide-8
SLIDE 8

Synaptic bases of pattern completion

1 Macro-connectomics CA3–CA3 connectivity? 2 Micro-connectomics properties of recurrent synapses

CA3-CA3 synapses Guzman, Frotscher and Jonas page 7 of 28

slide-9
SLIDE 9

Synaptic bases of pattern completion

1 Macro-connectomics CA3–CA3 connectivity? 2 Micro-connectomics properties of recurrent synapses 3 how are memories stored? (i.e. plasticity rules)

CA3-CA3 synapses Guzman, Frotscher and Jonas page 7 of 28

slide-10
SLIDE 10

Macro-connectomics what is the connectivity between CA3 cells?

CA3-CA3 synapses Guzman, Frotscher and Jonas page 8 of 28

slide-11
SLIDE 11

Macro-connectomics: on CA3–CA3 connectivity

On CA3 numbers

∼ 40,000 boutons Witter et al., 2007 Li et al., 1994 Ishizuka et al., 1990 ∼ 330,000 CA3 cells Boss et al., 1987

CA3-CA3 synapses Guzman, Frotscher and Jonas page 9 of 28

slide-12
SLIDE 12

Macro-connectomics: on CA3–CA3 connectivity

On CA3 numbers

∼ 40,000 boutons Witter et al., 2007 Li et al., 1994 Ishizuka et al., 1990 ∼ 330,000 CA3 cells Boss et al., 1987

CA3-CA3 synapses Guzman, Frotscher and Jonas page 10 of 28

slide-13
SLIDE 13

Octuple whole-cell recordings a) Simultaneous recordings b) Recording configurations c) Connectivity matrix to test CA3–CA3 cells

CA3-CA3 synapses Guzman, Frotscher and Jonas page 11 of 28

slide-14
SLIDE 14

Octuple whole-cell recordings a) Simultaneous recordings b) Recording configurations c) Connectivity matrix to test CA3–CA3 cells

CA3-CA3 synapses Guzman, Frotscher and Jonas page 12 of 28

slide-15
SLIDE 15

Motifs of connectivity a) Simultaneous recordings b) Motifs of hyperconnectivity c) Connectivity motifs are representative

CA3-CA3 synapses Guzman, Frotscher and Jonas page 13 of 28

slide-16
SLIDE 16

Macro-connectomics what is the connectivity between CA3 cells?

CA3 neurons use motifs of hyperconnectivity embedded on a sparse network

CA3-CA3 synapses Guzman, Frotscher and Jonas page 14 of 28

slide-17
SLIDE 17

Micro-connectomics Properties of the CA3–CA3 synapses

CA3-CA3 synapses Guzman, Frotscher and Jonas page 15 of 28

slide-18
SLIDE 18

Anatomical identification of synaptic contacts A Biocytine labeling B Identification under light microscopy

performed under 60X or 100X/1.4 oil-immersion objectives

axon and dendrite belong to functionally connected pairs presence of enlargements at the axon/dendrite. appointments located at a single focal plane.

C Nucleated patches

Sather et al., 1992 CA3-CA3 synapses Guzman, Frotscher and Jonas page 16 of 28

slide-19
SLIDE 19
slide-20
SLIDE 20

Micro-connectomics: 1–2 contacts per connection a) Digital reconstruction

100 μm

b) Putative synaptic contacts c) Cable modeling d) Conductance estimation

CA3-CA3 synapses Guzman, Frotscher and Jonas page 18 of 28

slide-21
SLIDE 21

Micro-connectomics: low number of functional sites

CA3-CA3 synapses Guzman, Frotscher and Jonas page 19 of 28

slide-22
SLIDE 22

Micro-connectomics: low number of functional sites

CA3-CA3 synapses Guzman, Frotscher and Jonas page 20 of 28

slide-23
SLIDE 23

Micro-connectomics: low number of functional sites

CA3-CA3 synapses Guzman, Frotscher and Jonas page 21 of 28

slide-24
SLIDE 24

EPSPs are small but add efficiently a) EPSP kinetics b) Temporal summation c) Spatial summation d) Low number of CA3 cells for spiking

CA3-CA3 synapses Guzman, Frotscher and Jonas page 22 of 28

slide-25
SLIDE 25

Micro-connectomics Properties of the CA3–CA3 synapses

CA3 synapses exhibit a low number of sites and summation

CA3-CA3 synapses Guzman, Frotscher and Jonas page 23 of 28

slide-26
SLIDE 26

Towards a quantitative model for pattern completion

CA3-CA3 synapses Guzman, Frotscher and Jonas page 24 of 28

slide-27
SLIDE 27

Towards a quantitative model for pattern completion

CA3-CA3 synapses Guzman, Frotscher and Jonas page 25 of 28

slide-28
SLIDE 28

Towards a quantitative model for pattern completion

CA3-CA3 synapses Guzman, Frotscher and Jonas page 26 of 28

slide-29
SLIDE 29

Summary

1 Macro-connectomics combination of sparse and dense connectivity

CA3-CA3 synapses Guzman, Frotscher and Jonas page 27 of 28

slide-30
SLIDE 30

Summary

1 Macro-connectomics combination of sparse and dense connectivity 2 Micro-connectomics small number of synapses, but very prone to summation

CA3-CA3 synapses Guzman, Frotscher and Jonas page 27 of 28

slide-31
SLIDE 31

Summary

1 Macro-connectomics combination of sparse and dense connectivity 2 Micro-connectomics small number of synapses, but very prone to summation

CA3-CA3 synapses Guzman, Frotscher and Jonas page 27 of 28

slide-32
SLIDE 32

Acknowledgments

1

Peter Jonas,

2

Alois Schlögl,

3

Jonas lab

CA3-CA3 synapses Guzman, Frotscher and Jonas page 28 of 28