Odor- and context dependent modulation of mitral cell activity in behaving rats

1
Odor- and context dependent modulation of mitral
cell activity in behaving rats

• Leslie M. Kay Gilles Laurent

Presented by Alexa Hamilton
2
The Olfactory Bulb

• Connected to everything else in the brain
  (according to Kay)
• Signals traveling to the olfactory bulb do not go
  through the thalamus
• Prone to disconnection in traumatic head injuries
  (sieve bone acts as guillotine)
• In rats, the olfactory bulb is very large,
  relatively much larger than in humans.
• OB contains mitral cells that communicate via
  action potentials.

3
The Olfactory Bulb
4
Question

• When assessing the function of neurons, does it
  make a difference whether the animal is
  anesthetized?

5
Alternatives

• Yes
• No

6
Logic

• Look for a difference in the firing of the mitral
  cells between two conditions with different
  contexts.
• If mitral cell firing is influenced by behavior
  and/or context, we conclude that yes, there is a
  difference in behaving vs. anesthetized animals.
  Otherwise no.
• 2 conditions
• Odor identification
• Odor-driven behavior

7
Methods

• Rats are trained to enter a cage and drink from a
  tube.
• Electrodes are implanted in the rats olfactory
  bulbs.
• Trained rats are put through a set of trials,
  varying odors and liquid.
• S odorless trials, 10 sucrose solution
• P1 (odor identification task) two odors ,10
  sucrose solution
• P2 (odor-driven behavior task) two odors, each
  consistently paired with either quinine
  (aversive) or 30 sucrose solution
• P3 (odor identification task) two odors, 10
  sucrose solution
• P4 To ensure rats could not smell the solution,
  after all trials, animals were tested by randomly
  pairing quinine or sucrose solution with
  different odors.

8
Methods - Behavioral

• Trial structure (every trial, regardless of
  P-phase)
• Light goes on ? Door opens ? Rat sniffs ? Rat
  chooses to drink the liquid or not.

9
Methods - Physiological
10
Results
11
Results
S
N
12
Results
13
Results-summary

• During odor identification task (P1 P3), 11 of
  cells were found to have odor-selective
  differences.
• During odor-driven behavior (P2), 94 of cells
  were found to have odor-selective differences.
• It was also found that in P4, the rats performed
  at chance level.

14
Interpretations

• Yes, there is a difference between behaving
  animals and anesthetized animals
• Behaving animals are capable of fast sniffing and
  drinking.
• This suggests that the function of the mitral
  cells (what the spike train reflects/encodes) is
  dynamically dependent on reward history and
  behavioral context/task.
• There is a remarkable flexibility in the ability
  of mitral cells to change what they encode.
• This might reflect the fundamental nature of the
  OB, being hooked up to all kinds of other systems
  (emotional, sensory, behavioral, etc.), flexibly
  changing the task its engaged in/the
  computations it performs.

15
Problems

• Overall, an extremely solid paper
• Performed the necessary controls
• Really shows that the importance of using
  behaving animals in order to make conclusions
  about behavior.
• Does the psychophysics and physiology in the same
  organism.
• Does histology.

16
Actual Problems

• Overall, a very low number of neurons (52 single
  units). Typical minimum standards are around a
  100.
• Reporting most of the results in terms of
  INDIVIDUAL neurons, not populations.
• There is a remarkable, unexplained variance in
  the response between and within mitral cells that
  remain poorly understood and might reflect
  anything, but this possibility is not discussed.
• Is it possible that the crucial difference
  between mitral cell firing in response to
  aversive vs. non-aversive odor in the P2
  condition is due to motor activity (drinking vs.
  non-drinking)? (The authors deny that, but fail
  to provide evidence for this).