Cumulative Final Flashcards
Emphasis on post midterm material (35 cards)
In the Odor representational drift paper, what are three ways the authors use to convince us they are recording from the same neurons?
Waveform comparison
- Waveform of a neuron shouldn’t change too much overtime
- dif neurons will have differently shaped AP
- Compare waveform recordings, if they are highly similar, good chances are you’re recording from same neuron
Displacement:
- Neuron probes shouldn’t move that much to make sure that they were recording from the same neurons
Autocorrelation:
- Correlating AP value at time T w AP value at time T-1
- Calculate membrane potential every micro s. and compare it with what it was a micro s. ago
- During base level activity, should be highly autocorrelated, during AP should be less and less correlated
- Point is to see refractory period which should be stable in within same neurons and across time
How did single neuron activity in response to odor change overtime?
- got less similar overtime while percentage of responsive neurons remained the same
I.e same amount of neurons, dif neurons overtime
What technique did the authors use to evaluate changes in population activity in response to odor? What was their conclusion?
Technique: linear classifier. If classifier can identify odors with same accuracy on day 30 as on day 1, then no population change.
Results:
- classifier got worse overtime when using day 0 data
- at day 32, as accurate as with shuffled data (control)
How do we know changes to odor response aren’t due to global changes in piriform activity?
of neurons responding to any given odor is stable
- same # of neurons active in response to odors on any given test day: piriform neuron pop isn’t becoming more or less active
- # of neurons responding to any given odor is stable
- # of odors a single neuron responds to is stable
After single neuron activity and population activity, what did the authors evaluate to determine drift in response to odor? What were the results?
Evaluated neural geometry:
- measures relationship b/w neuron activity by angles b/w firing activity of dif neurons
- if geometry is preserved, pop and single neuron activity doesn’t need to be stable
Results:
- angles differ overtime
What was the rational for the virtual burrow assay experiment in the odor representation paper?
- it was used to measure whether stimuli (i.e. odors) were aversive or appealing
- goal was to associate odor with shock and see if that odor saw less drift
- hypothesis was that a more important stimulus (e.g. one that should be avoided or predicts bad outcome) would lead to less representational drift for that odor
Did the odor associated with shock decreased amount of drift for that odor?
- no, it had the same drift as other non-aversive odors
What was the one condition that reduced representational drift of odors?
Familiarity
- prior familiarity + constant exposure to familiar odor b/w test days was the ONLY condition to reduce shift
- prior familiarity alone did NOT decrease drift
What was the behavioral paradigm in the stable neural representation paper?
- rats implanted w electrodes in MC and Dorsalateral Striatum (DSL)
- trained to press lever in home cage twice in certain time frame to get reward
What was the very first technique they used in the lever-press paper?
Computational Modelling to determine if drifting representation can theoretically have consistent motor output
- Claim supported by computational modelling
How did the authors determine that rat behavior was sterotyped?
velocity for lever presses was stable for expert rat
How did single neuron activity eliciting lever-presses change over time?
- they DIDN’T
- neurons activity over time was extremely well correlated, suggesting stable neural representation
After finding very stable single neuron activity over time in lever-pressing task, what was the next question the authors asked?
- still some drift
- but there is also some slight behavioral changes b/w presses
Questions: Do those changes correlate with each other?
- if they do, then observed drift is not actual drift, just a result of slight movement variations
What caused the observed drift in neural activity driven by lever-pressing task?
- slight changes in motor output
State the potential reasons for drift and their assumed mechanism(s):
Continual learning : regularization to prevent overfitting, stochastic randomization, drop-out
Keep representations separate : orthogonalization
Noisiness : internal error signals driven by predictive coding
What is one explanation for prior and constant exposure to an odor decreasing its rate of drift?
- better predictive coding set-up
- would drive internal error signals from higher order areas when wrong or unexpected representation happen, driving changes in activity of lower representation areas
Why did the authors want to label neurons that express Fos during sleep in Cre-dependent manner even though they had just identified them with immunohistochemistry?
- immunohistochemistry was the first step to see IF and WHERE there were any neurons that were more active during sleep (helped identify pIII regions of midbrain)
- labeling those same neurons with Cre allowed researchers to permanently label them and therefore have time to study them –> find proteins only expressed in those neurons –> with end-goal to manipulate them with Cre-dependent virus
How did researchers permanently label neurons that express FOS during sleep?
By crossing a FOS-CreER mouse line with a generic stop codon-GFP mouse line:
- resulting mice have Cre-ER proteins in all neurons
- Cre being attached to ER cannot go and excise the stop codon before the GFP gene
- when animal is sleeping, give drug that binds to Estrogen receptor and forces chaperone protein to let go of CreER
- CreER in active neurons (i.e. sleep-active neurons) will now be able to excise stop codon before GFP gene
- neurons that expressed Fos during sleep are now permanently labeled
Name the drug that forces chaperone protein to let go of CreER protein
Tamoxifen
Efficiency and specificity. What are they and how did it relate to GFP reporter mice for sleep-active neurons?
Efficiency refers to amount of neurons that can be identified using a technique.
Specificity refers to the accuracy of the technique in identifying the right neurons.
GFP reporter mice were able to identify 15% of the neurons of interest (i.e. the neurons that expressed FOS during sleep) meaning low efficiency
80% of the GFP-marked neurons were the neurons of interest (i.e. the neurons that expressed FOS during sleep) meaning high specificity
Overall, GFP reporter mice are a good method to determine proteins only expressed in neurons that express FOS during sleep.
Using FISH technique, what did research find out about pIII neurons?
70% of cells that express FOS during sleep express CALCA and vice versa but not in neighboring regions
CALCA-Cre mice will allow to use cre-dependent virus to selectively record and manipulate activity of these cells
How did researchers selectively and accurately record from CALCA neurons?
- electrodes are most accurate way of recording neural activity but –> non-selective
To make selective:
- ChR2 delivery to CALCA-Cre mice
- shine line with fiber optic attached to recording probe
- the neurons that have strong response to light are the CALCA neurons with ChR2
How does chemogenetics work?
- key concept is that we induce gene expression of a designer receptor that is activated by its corresponding designer drug ONLY
- no endogenous molecule can activate designer receptor
- induces slight changes in excitability but keeps the naturally occuring neural dynamics: more ecologically valid compared to optogenetics
What did opto activation of cholinergic neurons in pIII do to sleep?
nothing, no effect
