Neural Circuit Motifs for Valence Processing

 I am reading a Neuron review by Kay Tye from Salk and MIT, “Neural Circuit Motifs for Valence Processing”. Kay Tye is famous for her approach to neuroscience-- she starts with high-level psychological theories and tries to untangle their low-level neural mechanisms. She begins the review with two old psychological theories on valence processing-- Two-Factor Theory of Emotion (Schachter and Singer, 1962) and Two-Dimensional Theory of Emotion (Lang, 1995). The Two-Factory Theory proposes two stages of valence processing-- one that evaluates arousal and the next that assigns positive or negative valence. In the Two-Dimensional Theory the intensity, arousal and valence, hedonic value span two simultaneous dimensions. Oh my, the experimental support for the Two-Factor Theory is queezy! There was a patient, S.M., with bilateral amygdala damage and when he was exposed to hypoxic suffocation he exhibited autonomic arousal associated with panic! Wow:-D

Kay Tye outlines four models for the amygdalar cellular operations for assigning a valence to a stimulus, i.e. whether a neuron displayed excitatory, inhibitory, or no responses to stimuli of both positive and negative valence. These models are: Labeled Lines model (which encodes a valence in the disparity in the circuits the neurons belong to), Divergent Paths model (based on where the cells project to), Opposing Components model (what the neurons release and where they converge upon) and the Neoromodulatory gain model (how they are differentially modulated by G protein-coupled receptors). Each of these models has an associated algorithmic model. However, it is difficult to conclusively disambiguate the influence of each of these circuit processing motifs because of the lack of anatomical and genetic information overlaid with cellular-resolution recordings. 

Let’s look at these motifs, starting with the Labeled Lines. In this model streams of communication relay rewarding and noxious stimuli to motor systems that drive appropriate behaviors (approach and avoidance). This model confers speed and reliabilty onto the processing elements. The downside is that it does not permit relative weighting of competing motivational drives as there is no comparison through crosstalk. This also precludes interpreting stimuli according to context-- a 10 in black jack might lead to a win or bust you, depending on the cards you hold. It also does not permit learning valences, i.e. reversing innate valences. The neural system in this model is like a divided box with stimuli from different innately aversive or delightful stimuli plugged into different divisions.  

The Divergent Paths motif, in contrast, allows comparisons between different inputs and therefore the “winner-take-all” computation. I quote the paper: “Using an analogy popularized in the prefrontal cortex (Miller and Cohen, 2001), one model for how valence is assigned is to

liken incoming information to a train coming in on a railroad where one set of rails leads it to avoidance and another drives approach behavior. In this ‘‘Divergent Paths’’ model, the railroad forks at certain brain regions such as the amygdala and local computations serve to act as the ‘‘switch operator.’’ This motif is most likely to appear in regions that have similar types of projection neurons, but diverse projections, such as amygdala, hippocampus, thalamus, and cortex.” This is a boon fro the optogenetic-mediated projection specific manipulation technique, where you can force a behavioral phenotype through the optogenetic activation of a “minority population” of projection-defined neurons.

 The Opposing Component motif integrates different signals in downstream effector cells. An example of this motif is the projection of lateral hypothalamus to the the ventral tegmental area. LH contains diverse cell types, including glutamatergic and GABAergic neurons. Both glutamatergic and GABAergic neurons projections to VTA mediate reinforcement. These neurons synapse promiscuously on both GABA and dopamine neurons in VTA. LH→VTA:GABA projections mediate positive and LH→VTA:glutamate negative valence.

The Neuromodulatory Gain motif can work in combination with any of the other circuit motifs. This motif allows for plasticity in the circuit by amplifying the molecular signals in the neuron via G protein-coupled receptors and second messengers.