In order to guide behavior, humans and animals must flexibly evaluate the motivational significance of stimuli in the environment. We sought to determine if, in different contexts, neurons in the amygdala and orbitofrontal cortex (OFC) indeed rescale their calculation of the motivational significance of stimuli that predict rewards. We used a "contrast revaluation" task in which the reward associated with one stimulus is held constant while other rewards within a particular context (or block of trials) change. This manipulation modulates the relative significance of the reward associated with one stimulus without changing its absolute amount. We recorded the activity of individual neurons in the amygdala and OFC of two monkeys while they performed the contrast revaluation task. On every trial, a monkey viewed one of two conditioned stimuli (CSs; distinct fractal patterns), each predictive of a different reward amount. CSs were novel for every experiment. Unconditioned stimulus (US, liquid reward) delivery followed CS presentation and a brief temporal gap (trace interval). The task consisted of three trial blocks, with switches between blocks occurring without warning. The presentation of CS2 predicted either a small (first and third blocks) or large US (second block). The presentation of CS1 predicted delivery of a medium US in all blocks. Thus CS1 corresponded to the "better" trial type in blocks 1 and 3, but not 2. Anticipatory licking behavior indicated that the monkey adapted its behavior depending upon the relative amount of expected reward. Although the reward amount associated with CS1 remained constant throughout the experiment, anticipatory licking decreased in block 2 and increased in block 3 - the blocks in which CS1 trials had become relatively less (block 2) and more (block 3) valuable. Strikingly, many individual amygdala and OFC neurons also modulated their responses to CS1 depending upon the block. Because this CS predicts the exact same reward in each block, these neurons cannot simply represent the sensory properties of a US associated with a CS. This finding demonstrates that amygdala and OFC neurons are often sensitive to the relative motivational significance of a CS, and not just to the sensory properties of its associated US or to the absolute value of the specific reward. Neurons in both the OFC and amygdala encode the relative value of CS1 but OFC neurons significantly encode relative value earlier than amygdala neurons. Cells in the amygdala and OFC code different properties during different time intervals during the trial and are consistent in valence when they code multiple properties. This implies that neurons are tracking state value: the overall motivational value of an organism's internal and external environment across time and sensory stimuli. Neurons that code relative value during the CS-trace interval and during reinforcement are also consistent in the valence that they code further supporting that these cells track state value. The neurons code with the same sign and strength whether the neuron is representing the relative value of the reward with no sensory input of the reward during CS or trace interval, or actually experiencing the reward during the US interval. Further, amygdala and OFC neural activity was correlated with the animal's behavioral performance, suggesting that these neurons could form the basis for animal's behavioral adaptation during contrast revaluation. These neural representations could also support behavior in other situations requiring flexible and adaptive evaluation of the motivational significance of stimuli.
| Identifer | oai:union.ndltd.org:columbia.edu/oai:academiccommons.columbia.edu:10.7916/D81C23ZP |
| Date | January 2013 |
| Creators | Saez, Rebecca |
| Source Sets | Columbia University |
| Language | English |
| Detected Language | English |
| Type | Theses |
Page generated in 0.0075 seconds