Contribution of Excitatory and Inhibitory Neuronal Activity to BOLD fMRI

• Published in: Cerebral cortex (New York, N.Y. 1991) Vol. 31; no. 9; pp. 4053 - 4067
• Main Authors: Moon, Hyun Seok, Jiang, Haiyan, Vo, Thanh Tan, Jung, Won Beom, Vazquez, Alberto L, Kim, Seong-Gi
• Format: Journal Article
• Language: English
• Published: United States Oxford University Press 29.07.2021
• Subjects: Animals; Cerebral Cortex - diagnostic imaging; Cerebral Cortex - physiology; Cerebrovascular Circulation - physiology; Excitatory Postsynaptic Potentials - physiology; Female; Humans; Magnetic Resonance Imaging - methods; Male; Mice; Mice, Inbred C57BL; Mice, Transgenic; Models, Neurological; Neural Inhibition - physiology; Neurons - physiology; Neurovascular Coupling; Optogenetics; Original; Oxygen - blood; Somatosensory Cortex - diagnostic imaging; Somatosensory Cortex - physiology; Vesicular Inhibitory Amino Acid Transport Proteins - genetics; neurovascular coupling; optogenetic fMRI; BOLD; excitatory neurons; inhibitory neurons
• ISSN: 1047-3211; 1460-2199
• DOI: 10.1093/cercor/bhab068

Abstract The BOLD fMRI response in the cortex is often assumed to reflect changes in excitatory neural activity. However, the contribution of inhibitory neurons to BOLD fMRI is unclear. Here, the role of inhibitory and excitatory activity was examined using multimodal approaches: electrophysiological recording, 15.2 T fMRI, optical intrinsic signal imaging, and modeling. Inhibitory and excitatory neuronal activity in the somatosensory cortex were selectively modulated by 20-s optogenetic stimulation of VGAT-ChR2 and CaMKII-ChR2 mice, respectively. Somatosensory stimulation and optogenetic stimulation of excitatory neurons induced positive BOLD responses in the somatosensory network, whereas stimulation of inhibitory neurons produced biphasic responses at the stimulation site, initial positive and later negative BOLD signals, and negative BOLD responses at downstream sites. When the stimulation duration was reduced to 5 s, the hemodynamic response of VGAT-ChR2 mice to optogenetic stimulation was only positive. Lastly, modeling performed from neuronal and hemodynamic data shows that the hemodynamic response function (HRF) of excitatory neurons is similar across different conditions, whereas the HRF of inhibitory neurons is highly sensitive to stimulation frequency and peaks earlier than that of excitatory neurons. Our study provides insights into the neurovascular coupling of excitatory and inhibitory neurons and the interpretation of BOLD fMRI signals.