Neuroscience of Learning
Neuroscience of Learning
Entire field of research, behavioural neuroscience, in one lesson
Use the reading as background
Challenges:
Brains are complex
Made of neurons, which are about 10-20 um (micrometers) in diameter
About ⅕ thickness of paper
Neurons wok fast:
Action potentials take about 1-2 ms (1/1000 second)
Billions of neurons (about 80-100 billion in a human)
In complex networks
Need to study large numbers to see interesting effects
Synapse = connection between neurons
Human brain has about 1000 trillion synapses
Basics of Brains
Brains, 2 kinds of cells: neurons and glia (we won’t discuss them)
Focus on learning:
Acquiring new information
Remembering it
Using it to drive behavior
Brains do this:
Information comes in from the world
Through special organs (eyes, ears…)
Sent to the brain via sensory neurons, connect organs to spinal cord
Brain processes the information
Information used to drive behavior
Moving muscles: spinal cord uses motor neurons
Neurons
Designed to receive and pass on information: directional
Info enters at dendrites = soma = axon = synapse
Synapse connects axon terminal of presynaptic cell of dendrite of postsynaptic cell
Action Potentials
What it means for a neuron to “fire”
Action potentials are:
Electrical:
Neuron is charged (membrane potential)
Maintained by movement of ions in and out of the cell
All-or-nothing:
Cell has a threshold; if the potential goes above that, it fires
Action potential starts in the dendrite and moves down
After firing, cell enters refractory period
Cannot fire again until it is over
Synapses
Involves 2 neurons:
Axon terminal of presynaptic neuron
Dendrite of postsynaptic neuron
Neurons don’t touch
Gap = synaptic cleft
Crossed by neurotransmitter
Live in synaptic vesicles
Released by presynaptic neuron
Diffuses across cleft
Binds to receptors on membrane of postsynaptic neuron
Each cell’s action potential is electric; communication between them is chemical
Synapses Process
Action potential arrives, coming down axon of presynaptic neuron
Reaches axon terminal, causes a change in the membrane, leads to some vesicles fusing with membrane, spilling neurotransmitter into synaptic cleft
Neurotransmitter diffuses across the synaptic cleft, and arrives at the postsynaptic neuron’s dendrite
On the membrane of the dendrite are receptors – proteins that bind neurotransmitters
When they detect a neurotransmitter, they change shape, leading to ions moving in or out of the postsynaptic cell, changing its membrane potential
If membrane potential changes enough, exceeds firing threshold, new action potential is formed, starts moving down the axon
Usually one stimulation is not enough to get to the threshold
Neurons have thousands of synapses:
Several simulations at the same time can sum up to reach threshold
Repeat from #1.
Regulation
Real neuronal communication is a lot more complex
Neurons can modulate the communication process:
Over 40 kinds of neurotransmitters, different effects
Excitatory (e.g., glutamate): binding to receptor increases potential of postsynaptic neuron, makes it more likely to fire
Excitatory synapse
Inhibitory (e.g., GABA): binding to receptor decreases potential of postsynaptic neuron, makes it less likely to fire
Inhibitory synapse
Different thresholds
Release more/less neurotransmitter on each firing
Postsynaptic cell can change receptor density and sensitivity
Grow new synapses; kill off old ones
A Simple Behavior
Reflex arc (e.g., knee-jerk reflex):
1. Doctor hits you with hammer
2. Sensory neuron (purple) detects stretching of the tendon; fires
Dendrite in the knee, axon terminal in spinal cord
Very long axon (over a meter)
3. Motor neuron (green)
Dendrites in spinal cord
Axon terminal in knee, flexes quad
Neuromuscular junction
No involvement of ‘brain’
Involuntary, fast, fixed action pattern
Need to also relax the opposing muscle:
Sensory neuron also synapses onto inhibitory interneuron (red)
Inhibits the hamstring from flexing
Neurons Learning
Learning = change in reaction to situation
Brain change: altering synapse strength
Hebbian learning
Synapse strength:
Excitability of postsynaptic neuron
Likelihood that one stimulation will get to threshold
Long-term potentiation (LTP)
LTP
Occurs at glutamate synapses
2 receptor types: AMPA, NMDA
AMPA (normal receptor):
Glutamate binds, ions enter cell, polarization increases, more likely to fire
NMDA:
To open, needs glutamate bound + potential of cell to already be high
“Coincidence detector”
When open: causes lots of long-term changes to cell (genes transcribed):
More receptors at the synapse
More sensitive receptors
Synapse gets stronger
LTP Example
Neuron A fires
Activates AMPA on neuron C, potential increases
No activation of NMDA, no long term changes
Neurons A and B fire at same time:
Activate AMPA on neuron C, a lot
Potential increases a lot
NMDA also opened
Both synapses (A-C, B-C) stronger
Pavlov e.g.:
Neuron A represents buzzer (CS)
Neuron B represents food (US)
Neuron C drives salivating (UR/CR)