Real-Time Imaging of Pavlovian Conditioning in Neurons in the Amygdala

Written by Adam Tozer

December 11, 2018

• Researchers trained mice to fear one type of sound, but expect reward when they heard another
• They performed real time observation of neurons in the basolateral amygdala as animals learned to associate the tones with fear or reward
• They show learning of fear and reward is plastic and can be reversed or remapped
• The research could provide insights into treatments for anxiety and PTSD


The basolateral amygdala is an area of the brain that encodes valence learning, whether something is attractive to us, or aversive to us.
Positive or negative valences induce typical behaviours. In mice, scientists can associate a painful stimulus such as a foot-shock with a tone or an environment. Meaning mice will freeze when they hear the same tone or enter an arena they associate with the foot-shock.
So, the animal’s aversive behaviour becomes conditioned with the stimulus. Similarly, scientists can condition mice to expect reward when they hear a tone or enter an environment they associate with reward. This is akin to Pavlov’s famous experiment of ringing a bell every time he fed his dogs. Soon enough the dogs became conditioned to the sound of the bell and would start to salivate in response to its ringing, even in the absence of food.
Scientists have even shown that this conditioned learning is so strong and pronounced, they can reproduce these behaviours by activating amygdala neurons after the learning trials took place.1


However, how this learning develops in the amygdala circuitry and is represented by individual neurons remained a mystery, until now. Recently, a group from Cold Spring Harbour Laboratory led by Prof. Bo Li published their findings in Nature Communications describing how individual neurons in basolateral amygdala circuits learn to associate fear or reward with a tone, and how this learning can be reversed and remapped.2
Li and his postdoc Xian Zhang conditioned the mice to associate fear with one type of tone by pairing it with a puff of air to the face. And they conditioned the mice to expect reward when another tone was played by giving the mice water.


The scientists used mini microscopes to image the activity of basolateral amygdala neurons expressing the calcium reporter GCamp6 in mice. An increase in fluorescence denotes an increase in the influx of calcium in the neurons, which is a proxy for an increase in the cells’ activity. The scientists observed neurons under their mini microscopes as they played the sounds to the mice. The neurons that underlie learning of the association between the tones and the behaviour were activated in response to the tones.
They initially found that the neurons spiked randomly in response to the tones. However, with repeated trials some of the neurons’ activity became correlated with the playing of the tone. These neurons formed part of the circuit encoding the conditioned response to the stimulus, be it rewarding or aversive.
As the firing patterns became more specific, the animals licked in response to the reward-associated tone – anticipating water, and they blinked in response to the punishment-associated sound – anticipating an air puff.


Most interesting of all, is that after watching the correlation between stimulus and neural activity develop in the basolateral amygdala neurons, the group could then reverse it. They found that switching the meaning of the tones, by changing the pairing of the tones with the air-puff or water-drop was ‘unlearned’ by the neurons in the circuits. Such that they stopped firing in response to the sounds.
The findings of this paper reveal a cool way by which individual neurons can be observed to correlate their activity with a stimulus. The advantage of using imaging techniques over electrophysiological techniques are that many neurons can be imaged at any one time, giving a population view of the neurons in the basolateral amygdala. However, resolution is always a confounding factor in these types of experiments, and the group could only measure changes in firing activity in the cell bodies, not in the dendrites of these cells. It would be interesting to know when and where the strengthening of connections in the circuits arises as the conditioning takes place.
Deep tissue imaging approaches in awake behaving animals are the future of neuroscience research. It will be fascinating to see how this technology advances, as understanding how conditioned behaviors can be reversed could have therapeutic benefits for sufferers of conditions like PTSD and anxiety.
Learn: New lens enables deep tissue imaging of dendrites in the hippocampus 

New Pryer lens from MCI-Neuroscience enables deep tissue imaging in the brain. Video shows dendrites in the hippocampus. Credit: MCI-Neuroscience, YouTube.
1. Felix-Ortiz, A. C., Burgos-Robles, A., Bhagat, N. D., Leppla, C. A., & Tye, K. M. (2016). Bidirectional modulation of anxiety-related and social behaviors by amygdala projections to the medial prefrontal cortex. Neuroscience, 321, 197-209.
2. Zhang, X. and Li, B. (2018). Population coding of valence in the basolateral amygdala. Nature Communications, 9(1).