Visual cortex plasticity

Question 1

how many potential connections are there between neurons?

Answer 1

10^10 (10 billion)

Question 2

what are the rules of laminar connections? (2)

Answer 2

• Laminar connectivity between visual areas
• Genes encoding specific areas

Question 3

how do we refine connections with experience? (2)

Answer 3

• refining the connections with experience => Critical window of development and ocular dominance plasticity in V1
• Adult learning and refinement of RFs in V4 and IT

Question 4

what does V1 produce perceptually?

Answer 4

a line filter or an edge filter which is useful for understanding objects, contours and movement

Question 5

how do we create more sophisticated representations of the visual world that are behaviorally relevant?

Answer 5

Instead of center surround and a little spot now we have a lined filter

Question 6

what happens if one area in our entire visual processing system is blindly connected to another?

Answer 6

we cannot create a sophisticated receptive field ⇒ same as blindly connecting retinal ganglion cells to V1 would result in no orientation selectivity
- Our understanding of how this wiring problem is solved during development is still unknown

Question 7

T/F the complexity of wiring a neural network can be purely genetic?

Answer 7

False
- it cannot it is impossible => There has to be a combination of genes and environmental rules that shape the network
- If something was completely hard wired and specified we would all be similar and we wouldn’t have the flexibility needed for a functioning adult brain

Question 8

what is the laminar structure of the cerebral cortex?

Answer 8

rules of connection in the cerebral cortex => Through the depth of the cortex there is are anatomically distinguishable layers 1-6
- Within a column we find neurons have common orientation selectivity in V1
- There is a lot of interlaminar connectivity that has very specific patterns

Question 9

what are the feedforward vs feedback projections from V1 to V2?

Answer 9

The feedforward is characterized by neurons in the upper layers landing in layer 4 and feedback is characterized by neurons outside of layer 4 in the upper and lower layers (1, 5/6)
- Not just any neuron can connect to any neuron ⇒ they are governed by the laminar specific pattern

Question 10

what are the interpretations for cues that create laminar specific connectivity?

Answer 10

• There is a gene expression transcription factor called OCC1 and it is exclusively expressed in V1 layer 4 => this could be a molecular target that tells lateral geniculate cell axons to go to V1 layer 4
• the second interpretation of that is that geniculate cells are very active with spontaneous activity rates being higher than cortical cells so the layer 4 V1 cells are receiving this background level of activity on the order 10x higher than the neocortex so there are all sorts of metabolic requirements for these synapses with their inputs => Metabolic load of layer 4 V1 cells is much higher than anywhere else in the cortex ⇒ might be a metabolic factor and not a pathfinder

Question 11

what do we see without knockouts in mice for aerial specification?

Answer 11

the C wild type show wild type mice with V1, A1, S1, and frontal motor areas and they tend to be distributed along the same pattern in the brain => There is a transcription factor gradient (similar to ephrins) such that EMX2 is predominant along the posterior pole and then degrades as you go laterally and anteriorly
- In the complete KO V1 is eliminated and auditory is largely gone ⇒ somatosensory and frontal motor are left
- Heterozygous knockouts have shrinkage of V1 and A1 and posterior expansion of the frontal motor and somatosensory regions

Question 12

what happens with overexpression of genes for V1/A1?

Answer 12

If we overexpress this instead we see the opposite occur where there is expansion of V1/A1 and decrease in somatosensory and frontal motor regions
- It also moves toward the anterior pole
- there are genes that play a role in specifying aerial boundaries and is a clue into the genetic restrictions ⇒ neurons in one spot are selectively sampling such as from the retina, cochlea, or skin and there are genetic markers that are helping the axons from those different thalamic regions to reach appropriate targets in the cortex

Question 13

where does mixing of the pathways from the eye eyes first occur?

Answer 13

in the striate V1
- At the lateral geniculate (L and R) there are representations of corresponding visual fields from the left and right visual field sampling from the two eyes
- The eye specific inputs are completely segregated into different layers in the lateral geniculate
- if you record from individual cells in the LGN you end up with monocular cells with no orientation selectivity ⇒ the lateral geniculate cells have an ipsilateral projection to V1 and in V1 layer 4 the eye specific connections are segregated so you get somewhat monocular cells in layer 4 and outside of layer 4 you see binocular cells where the inputs from the two eyes are combined
- This combining follows a pattern that is orthogonal to the structure regarding orientation selectivity as there are alternating bands if you proceed along the surface of the cortex
- It is reminiscent of the gradual change in orientation selectivity but with regard to eye dominance ⇒ alternating patterns of ipsilateral and contralateral dominance that we can project on the surface to see ocular dominance column patterns that show segregation of eye inputs at later 4

Question 14

Ocular dominance columns

Answer 14

axons from the cells in the lateral geniculate nucleus (LGN) serving each eye terminate in separate bands in layer 4
- The ocular dominance show the anatomical segregation of inputs from the 2 eyes into layer 4

Question 15

what does individual variation in the ocular dominance columns look like? Are they wired by genetics alone?

Answer 15

the first clue that these things aren’t completely hard wired via genetics you get from looking at the pattern of ocular dominance across different animals
- There are rich alternating patterns of V1 and monocular cells but the pattern, details, and frequency of banding is different between animals

Question 16

what are the effects of monocular deprivation on the patterning of V1 dominance columns?

Answer 16

deprived eye columns are shrunken, not disappeared ⇒ so not disuse mediated withering away but competitive imbalances
- if you restrict input from one eye during a critical period you find that this normal pattern changes so the open eye (gaps between dark labels) takes over the brain while prior there were equal amounts of light and dark for equal cortical territory devoted to the two eyes
- But with deprivation one eye takes over the brain even in the adult

Question 17

what does monocular deprivation do to the temrinal arborizations of LGN axons in cat V1?

Answer 17

if you look at afferents landing in V1 layer 4 with short term vs long term monocular deprivation you can see in the open eye there is a robust pattern of axonal termination but the closed eye has almost nothing going on ⇒ even after only one week of deprivation during the critical period
- even across long periods of time

Question 18

what can electrophysiology show us about input from the eyes?

Answer 18

recording and noting for each eye the ocular dominance index which describes the ratio when you stimulate one eye and don’t stimulate the other (you get a response) and when you exclusively stimulate the other eye you get a response
- is completely equal activation from the left and right eye with the optimal stimulus from an appropriately oriented bar in an appropriate location

Question 19

for adult cats with normal raising what do the ocular dominance indices show?

Answer 19

when you do this experiment in an adult cat that has been normally raised, there is a broad distribution of ocular dominance indices and the peak is at binocular cells ⇒ lots of cells integrate inputs from the two eyes roughly equally with a slight contralateral vs ipsilateral bias
- We need the binocular cells

Question 20

what do ocular dominance indices show for a (sutured) closed eye during the critical period for a short amount of time?

Answer 20

when we environmentally manipulate so that one eyelid is sutured closed and there are no formed vision activating center surround retinal ganglion cells for the first 2 ½ months of the cats life and we open the suture until adulthood with normal visual input, the pattern of eye dominance in V1 shows the ipsilateral (open eye) is exclusively dominating responses in V1 ⇒ there are no binocular cells and no cells corresponding to the input for the closed eyelid despite the retinal activity patterns were normal for most of the animals life aside from the first 2 ½ months

Question 21

what happens if both eyes are sutured shut for the same critical period time as suturing one eye?

Answer 21

if for the same brief period of time both eyes are sutured and we repeat the experiment, we find a relatively normal pattern of ocular dominance ⇒ this means competition is the key
- The open eye can essentially take over

Question 22

does suturing one eye show disuse or competition?

Answer 22

competition
- disuse does not induce ocular dominance plasticity in binocular deprivation ⇒ competition and local correlations in neural activity (spikes) are the drivers of plasticity

Question 23

how does hebbian firing work with ocular deprivation?

Answer 23

competition may occur because in the normal pattern you have a strengthening of synapses from left eye activity and right eye activity ⇒ there is cooperation and strengthening of things within the eye and competition between the eyes because patterns of input are different
- With only one eye (left eye) and right is restricted then the activity of the right eye is decreased so activity of the cells in the left will have strengthened synapses at the expense of the right eye synapses
- Cooperation of similar inputs and competition between different inputs is the fundamental synaptic theory of how plasticity might work throughout the cortex

Question 24

what does orientation tuning in V4 show us after training?

Answer 24

if you go into V4 which has orientation selective cells and in adult macaques we can train them in an orientation task where they have to discriminate between two fine orientations => if you look at the neurons in V4 after training there is some sharpening of orientation selectivity that indicates the neurons have changed their orientation selectivity so it is consistent with an improved ability to discriminate

Question 25

Perceptual learning

Answer 25

as you train in a perceptual task you improve your performance in that task

Question 26

what do shapes specific response show in the IT?

Answer 26

at a cellular level there is adult plasticity going on that is able to refine response on the basis of training => in the inferotemporal cortex we think this is associated with shape processing and you can find that if you present a certain subset of shapes to an animal and ask them to do detection tasks on the shapes - After several months you present the shapes in combination with shapes they haven’t been exposed to - There is some weak tendency for neurons to fire better to shapes they have been exposed to and trained on than shapes they have not