Anxiety and depression: A top-down, bottom-up model of circuit function

Abstract

A functional interplay of bottom-up and top-down processing allows an individual to appropriately respond to the dynamic environment around them. These processing modalities can be represented as attractor states using a dynamical systems model of the brain. The transition probability to move from one attractor state to another is dependent on the stability, depth, neuromodulatory tone, and tonic changes in plasticity. However, how does the relationship between these states change in disease states, such as anxiety or depression? We describe bottom-up and top-down processing from Marr's computational-algorithmic-implementation perspective to understand depressive and anxious disease states. We illustrate examples of bottom-up processing as basolateral amygdala signaling and projections and top-down processing as medial prefrontal cortex internal signaling and projections. Understanding these internal processing dynamics can help us better model the multifaceted elements of anxiety and depression.

Keywords: PFC; amygdala; anxiety; attractor states; bottom-up; depression; top-down.

FIGURE 1

Attractor state dynamics of bottom-up, top-down processing. Bottom-up (orange) and top-down (blue) processing in (A) healthy, (B) anxious, (C) depressed, (D) slow-switching comorbid, and (E) fast-switching comorbid states are represented through attractor state dynamics (upper), where the ball indicates population activity at a given state with arrows showing average range of motion, and through the transition probability between states (lower), where the thicker arrows indicate higher transition probabilities. (F) Disordered transition probabilities result in different phenotype presentations. When the transition probability between TD/BU states is critically low, patients present anxiety, depression, or slow-switching comorbidity. When the transition probability between TD/BU states is critically high, patients present fast-switching comorbidity. (G, upper) Attractor state depth can change from Hebbian (purple) and homeostatic (green) plasticity changes. (G, lower) Plasticity changes are altered on different timescales; whereas Hebbian plasticity deviates from the basal setpoint, homeostatic plasticity resolves plastic deviations and, if needed, re-establishes the basal setpoint. Abbreviations: AMPAR, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic receptor; BU, bottom-up; TD, top-down.

FIGURE 2

Interaction of bottom-up and top-down feedback loops during stimulus processing. (A–C, left) Flow of stimulus and valence processing in (A, left) healthy, (B, left) anxious, and (C, left) depressed states. Larger, thicker arrows indicate increased bias or activation. Dashed, thinner arrows indicate decreased activation. (A–C, right) Valence processing curves. (A, right) In a healthy state, valence processing corresponds appropriately to increases in arousal and positive–negative valuations of valence. The dashed line indicates the threshold of responding to a given stimulus. (B, right) In an anxious state, arousal is increased with a negative valence bias. (C, right) In a depressed state, arousal is decreased with a negative valence bias.

FIGURE 3

The effects of neuromodulation on the depth of attractor states and synaptic plasticity. (A) As neural populations receive higher bouts of 5-HT_1A/B receptor activity and/or increased DA_2–4 receptor activity, the depth of an attractor state decreases. As neural populations experience higher levels of DA_1/5 receptor activity and/or increased levels of 5-HT_2A receptor activity, the depth of an attractor state increases. (B) Variable release of neuromodulation across time after receiving a given stimulus. DA (blue) and NE (yellow) exhibit short encoding of stimulus. 5-HT (pink) imposes a slower, sustained release onto synapses. (C) In healthy mPFC synapses, 5-HT_2A receptors (pink), D1-like receptors (dark blue), and AMPARs (red) interact to encourage NMDAR externalization. These receptors additionally recruit glutamatergic neurons (red). D2-like receptors (blue) internalize surface NMDARs and recruit GABAergic neurons (orange). Under unhealthy conditions in the mPFC, increased pressure from 5-HT_2A and D2-like receptors (in addition to decreased pressure from D1-like receptors) causes increased externalization of NMDARs and glutamatergic neuron recruitment. The consequence of this synaptic change is an excitatory/inhibitory imbalance in the mPFC. Abbreviations: 5-HT, serotonin; AMPAR, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic receptor; DA, dopamine, E, excitation; GABA, gamma-aminobutyric acid; I, inhibition; mPFC, medial prefrontal cortex; NE, norepinephrine; NMDAR, N-methyl-D-aspartate receptor; POST, postsynaptic neuron; PRE, presynaptic neuron.