GABAA receptor availability relates to emotion-induced BOLD responses in the medial prefrontal cortex: simultaneous fMRI/PET with [11C]flumazenil

Alexander Heinzel, Jörg Mauler, Hans Herzog, Frank Boers, Felix M. Mottaghy, Karl‐Josef Langen, J. Scheins, Christoph Lerche, Bernd Neumaier, Georg Northoff, N. Jon ShahFrontiers in Neuroscience 2023 · DOI

TL;DR

This study uses simultaneous fMRI and PET to show that the availability of GABAA receptors in the medial prefrontal cortex (MPFC) is linked to how our brains respond to emotional stimuli. Specifically, different parts of the MPFC seem to play distinct roles in processing emotions, with one area involved in passively viewing emotions and another in actively appraising them.

What Problem Does This Paper Address?

Understanding how brain activity, measured by fMRI's Blood-Oxygen-Level-Dependent (BOLD) signal, relates to the underlying neurochemistry is a major challenge in neuroscience. While fMRI is great at showing where and when the brain is active, it doesn't tell us why. Gamma-aminobutyric acid (GABA) is the primary inhibitory neurotransmitter in the brain, and its levels have been linked to emotional processing. However, previous studies often used different methods (like proton magnetic resonance spectroscopy, ¹H-MRS) to measure GABA levels and fMRI to measure BOLD activity, making it difficult to directly link these. This paper aims to bridge this gap by investigating the relationship between GABAA receptor availability (using PET) and BOLD responses during emotional processing within the same individuals and at the same time.

Methods

The researchers employed a sophisticated simultaneous fMRI and PET imaging technique using a hybrid 3T MR-BrainPET scanner.

  • PET Imaging: They used [¹¹C]flumazenil, a radiotracer that binds to GABAA receptors. This allowed them to measure the "binding potential" (BP) of these receptors, essentially quantifying their availability in different brain regions. A bolus plus infusion protocol was used to achieve a steady state of the tracer in the brain.
  • fMRI Imaging: Functional MRI (fMRI) was used concurrently to measure the BOLD signal, which reflects changes in blood flow and oxygenation associated with neural activity.
  • Emotional Stimuli and Tasks: Participants viewed images from the International Affective Picture System (IAPS) and performed two types of tasks:
  • Passive Viewing: Simply looking at the emotional (positive or negative) or neutral images.
  • Picture Appraisal: Actively judging the emotional valence (positive/negative) or personal relevance (high/low) of the images.
  • Data Analysis: The core of the analysis involved correlating the BOLD signal changes (specifically, the percent signal change) during these tasks with the [¹¹C]flumazenil binding potential in various brain regions, particularly the medial prefrontal cortex (MPFC). They used statistical parametric mapping (SPM12) for preprocessing and analysis, and defined regions of interest (ROIs) using the Harvard-Oxford Cortical Atlas.

Datasets

The study utilized the International Affective Picture System (IAPS) dataset for emotional stimuli. No new benchmark datasets were introduced.

Key Insights and Ideas

  • Bridging Neurochemistry and Activity: The study successfully links neurochemical availability (GABAA receptors) with functional brain activity (BOLD signal) in the same individuals, providing a more direct understanding of their relationship.
  • MPFC Specialization in Emotion: The findings suggest that different sub-regions within the MPFC are not functionally uniform when it comes to processing emotions.
  • Passive vs. Active Processing: The distinction between passively experiencing an emotion and actively appraising it is reflected in the neurochemical-functional correlations within the MPFC.

Key Results

  • Ventromedial Prefrontal Cortex (vMPFC) and Passive Viewing: A positive correlation was found between [¹¹C]flumazenil binding potential in the vMPFC (specifically the frontal medial cortex, FMC) and the BOLD signal change during passive viewing of emotionally valenced (positive and negative) pictures. No such correlation was observed for neutral pictures.
  • Paracingulate Gyrus and Picture Appraisal: In contrast, the picture appraisal task showed a correlation between [¹¹C]flumazenil binding potential and BOLD signal change in the paracingulate gyrus. This correlation was observed for both emotional and neutral stimuli, and was particularly strong for neutral stimuli.
  • No Correlation in vMPFC for Appraisal: Notably, no significant correlation was found in the vMPFC for the picture appraisal task.
  • Whole-Brain Findings: Beyond the MPFC, other brain regions like the insular cortex, planum temporale, Heschl's gyrus, and planum polare also showed correlations, some positive and some negative, with GABAA receptor availability during passive viewing and appraisal.

Limitations and Future Work

  • No Direct GABA Measurement: The study measured GABAA receptor availability, not GABA levels themselves. While related, they are not the same. The authors speculate that their positive correlations might imply lower GABA levels (as measured by ¹H-MRS) in regions with higher GABAA receptor availability, based on existing literature suggesting a homeostatic balance. Future work could combine fMRI, PET, and ¹H-MRS to directly investigate these relationships.
  • Sample Size: While 15 participants were included in the final analysis, the authors acknowledge that larger sample sizes are often needed for robust fMRI-PET correlation studies due to the low signal-to-noise ratio of fMRI. Multi-center studies are suggested as a way to overcome this.
  • ROI Definition: The analysis relied on predefined ROIs from an atlas. While standard, this approach treats entire regions as uniform, potentially missing finer-grained regional differences.
  • PET Measurement Method: The study used a steady-state distribution volume method for PET analysis, which is simpler but may be less precise than kinetic modeling with arterial blood sampling, especially given the observed larger standard deviations in binding potential values compared to some literature.
  • Future Directions: The authors suggest future studies should aim to measure all three parameters (GABA levels, GABAA receptor availability, and BOLD signal) simultaneously to gain a more complete picture of the neurochemical-functional interplay in emotion processing. Further investigation into the specific roles of different brain regions, like the planum temporale and Heschl's gyrus, in emotional processing is also warranted.

Summary updated

Abstract

Our data deliver first evidence for a relationship between MPFC GABAA receptors and emotion processing in the same region. Moreover, we observed that GABAA receptors appear to play different roles in emotion processing in the vMPFC (passive viewing) and paracingulate gyrus (appraisal).
BibTeX
@inproceedings{heinzel2023gabaa,
  author = {Alexander Heinzel and Jörg Mauler and Hans Herzog and Frank Boers and Felix M. Mottaghy and Karl‐Josef Langen and J. Scheins and Christoph Lerche and Bernd Neumaier and Georg Northoff and N. Jon Shah},
  title = {{GABAA receptor availability relates to emotion-induced BOLD responses in the medial prefrontal cortex: simultaneous fMRI/PET with [11C]flumazenil}},
  booktitle = {Frontiers in Neuroscience},
  year = {2023},
  doi = {10.3389/fnins.2023.1027697},
}
Figure