The Amygdala: new tricks for an old structure
Clayton Mosher & Peter Rudebeck, Mount Sinai Hospital :: Nature Neuroscience, Vol 18, Number 3, March 2015 :: http://www.nature.com/neuro/
Summary and review of the above paper
INTRODUCTION: This study demonstrates how neurons that are active when the subject chooses whether to have a small reward now or a larger reward in the future are inactive when the choice is made for the subject, implying a physical brain distinction between freewill and its absence.
This paper argues that the amygdala may be involved in choosing more distant but also more valuable goals, in preference to lesser but closer goals. Neurons in the amygdala, a subcortical structure in the temporal lobe, signal long-term goal orientated plans. The amygdala was already known to signal the achievement of rewards, but was not previously thought to be involved in foresight of rewards. The study involved monkeys choosing an immediate juice reward, or deferring the reward in favour of a subsequent larger reward.
Two types of neurons in the amygdala correlated with the future reward choices of the monkeys. The first type of neuron, described as ‘sequence value neurons,’ signalled the monkeys’ willingness to work for a particular quantity of juice. A second type of neuron, described as ‘sequence length neurons,’ signalled the amount of time (taken as the number of trials) that the monkey was prepared to wait for its enlarged reward. When the monkeys did eventually take a juice reward, these neurons recalibrated to reflect forecasts for the next round of the juice trial.
Free and non-free will
The amygdala was already known to respond to stimuli that predicted reward and create the perception that more distant rewards were of less value than nearer-term rewards. However, the neurons referred to in this paper only became active when the subjects first chose to pursue a reward. Thus in an experiment where the monkeys freedom to choose a short or longer-term reward was removed, and the monkey was compelled to follow a particular reward pattern of waiting for rewards, rather in the nature of a compulsory pension contribution, the absence of a self-generated plan meant that the sets of neurons discussed here were no longer active, and no longer tracked the values or time horizons of the monkeys behaviour; this was despite the behaviour of the monkeys being exactly the same as it could have been with self-generated behaviour. The study thus indicates a physical brain distinction between free choice, and exactly the same behaviour, but with the subject not having free choice.
Normally, planning is associated with the prefrontal cortex rather than the amygdala. However, it is suggested that while some parts of the amygdala are confined to automatic reactions, other parts have evolved to work along with the prefrontal areas. The neurons referred to here are located in the basolateral nuclei of the amygdala, an area strongly connected to the prefrontal.
It is thought likely that these amygdala neurons are a node within a brain-wide network that generates plans. The anterior cingulate is a frontal brain region that is important in choosing to pursue an existing goal or move to another goal. It is also suggested that the amygdala could exert a bottom-up influence on the orbitofrontal, another region important in evaluating stimuli. The authors view the amygdala neurons discussed here as being part of a reward planning circuit. Behaviour is seen to suffer whenever such a reward planning circuit is compromised.