Overlapping Emotion Gradients in the Human Temporo-Parietal Cortex

Post by Deborah Joye

What's the science?

Understanding our own emotions and the emotions of others is an important aspect of our social experience. While we know that our emotional experience is reflective of changes in our brain, precisely how our brains encode emotional states remains unclear. Some research has reported that individual brain regions process specific emotional features, while other research reports that multiple brain networks must converge to create our emotional experience. It is also unclear whether emotional processing involves discrete basic emotions like happiness, sadness, and fear, or whether emotions are processed on gradient scales such as positive vs. negative or intense vs. weak emotions. One interesting possibility is that emotional processing is spatially organized in the brain similar to other sensory systems, such as vision. For example, the visual cortex is retinotopically mapped so that areas that are next to each other on the retina are next to one another in the cortex. Similarly, it is possible that one region of the brain contains cells that are specifically tuned to respond to intense feelings whereas other cells in that region only respond to weak feelings. This type of gradient organization is biologically efficient because it means that several gradients can be spatially overlapping, allowing one region of the brain to process multiple features of an emotional state. This week in Nature Communications, Lettieri, Handjaras, and colleagues present evidence that one patch of the human right temporal-parietal cortex contains spatially overlapping gradients that encode features of subjective emotions including polarity (pleasant vs. unpleasant), intensity (strong vs. weak), and complexity (automatic bodily responses like fear vs. mentalized emotions such as a mixture of happiness and sadness).

How did they do it?

The authors recruited 12 native Italian participants to watch an edited version of the movie Forrest Gump while continuously indicating the intensity of their moment-to-moment perceived, subjective emotions. Participants were able to select from 6 basic emotions including happy, sad, afraid, angry, surprised, or disgusted and could also select more than one subjective emotion at a time to indicate more complex emotional states. The authors analyzed these moment-to-moment data and examined how well emotion ratings agreed between subjects. The authors also determined the complexity of emotional responses made possible by the option to respond with multiple perceived emotions. The authors performed a principal component analysis to isolate and dissect variables that explained the most variability in emotional responses.

The authors then compared their emotion time series data with data from an independent, publicly available dataset in which 15 different German-native participants viewed the same segments of Forrest Gump while undergoing fMRI. The authors then used their collected data to determine the extent to which subjective emotional responses correlated with changes in brain blood flow. The authors also took advantage of publicly available emotion-tagging of characters in Forrest Gump to investigate whether the emotions reported by participants reflected their own subjective internal state, or rather, emotions portrayed by characters in the movie.

What did they find?

First, emotional responses from participants were categorized into 15 distinct emotional states, suggesting that 6 basic emotions do not capture the range of human emotional experiences even when considering something as simple as responses to a movie narrative. Interestingly, the authors also found that participant emotion ratings were able to explain brain activity changes from independent participants who watched the same movie, even though they were from a different country and spoke a different language. The association between the emotion ratings and brain activity was localized to the right temporo-parietal junction, which is known to be important for social cognition. Specifically, the authors found that this region contained spatially overlapping emotion gradients that encode the polarity, complexity, and intensity of emotional states. These emotion gradients have specific directionality and were only found in the right (not left) temporo-parietal junction. Since the temporo-parietal junction is known to be involved in social cognition, the authors investigated whether emotional responses from the participants could be explained by emotions portrayed by movie characters; they found that the TPJ topography was better explained by subjective emotion reports than by emotions portrayed by movie characters. Finally, the authors demonstrate that populations of brain cells in the TPJ may be specifically tuned to encode specific gradient features, such as neurons that maximally respond to highly positive or negative events or neurons that selectively respond depending on the intensity of the emotional experience.

What's the impact?

This is the first study to demonstrate the existence of spatially overlapping gradients within the temporo-parietal junction that can encode the polarity, complexity, and intensity of subjective emotional responses. Spatially overlapping gradients allow for a relatively small region of the brain to process many different aspects of an emotional state in parallel. Overall, this emotionotopic mapping of the cortex presents an intriguing model for how our brain may process various features of a complex emotional state and resembles how other sensory regions represent stimuli from the environment.

Lettieri & Handjaras et al., Emotionotopy in the human right temporo-parietal cortex, Nature Communications (2019). Access the original scientific publication here.