Decoupling of the GABA / Gamma Relationship during Development in ASD – the Impact of an Atypical Developmental Trajectory

Background:

It is believed that neuronal ensemble oscillatory activity, recorded in-vivo using electrophysiological techniques such as electroencephalography (EEG) and magnetoencephalography (MEG), reflects the integrity of underlying local circuitry. Furthermore, the precise operation of such local circuitry is maintained by the delicate balance of excitatory and inhibitory influences, mediated by the neurotransmitters, glutamate and GABA.

Objectives:

The objective of this study is to observe the relationship between estimates of GABA levels, derived from edited magnetic resonance spectroscopy (MEGAPRESS-MRS), and gamma-band oscillatory activity elicited from cortex and recorded using magnetoencephalography (MEG). Specifically, the multi modal study is executed in both children (across a broad age range) and adults in order to determine both the developmental trajectory and ultimate asymptote of each measure, as well as their correlation, or coupling.

Methods:

Spectrally-edited MEGAPRESS MRS from voxels in superior temporal gyrus were obtained in children (age 6-14) and young adults (age 18-40) with ASD as well as typically-developing (TD) controls (N=42 ASD, N=32 TD). In the same subjects, gamma-band oscillatory activity elicited by simple auditory stimulation was recorded using magnetoencephalography and localized to auditory cortex using a beamforming algorithm. Dependent variables were GABA/Cr ratios for MRS and gamma-band phase-locking for MEG. Linear mixed modeling and hierarchical regression were performed across the age range as well as separately for children and adults.

Results:

Across the entire age-range, both GABA/Cr ratios and gamma-band phase locking values were significantly reduced in ASD vs. TD. For GABA/Cr: ASD=0.28+/-0.01, TD=0.31+/-0.01, F=7.93, p<0.01. For gamma-band phase locking: ASD=0.058+/-0.003, TD=0.067+/-0.003, F=4.94, p<0.05. Critically, while in TD children there was a strong coupling between GABA and gamma measures (robust regression, p<0.01), no such correlation was observed in children with ASD (p>0.05). Furthermore, in adults, there was a significant difference in gamma-band oscillatory phase locking between TD and ASD (0.069+/-0.04 vs. 0.57+/-0.004, F=5.25, p<0.01).

Conclusions:

Multimodal studies of neurochemistry and electrophysiology suggest a tight coupling of GABA levels and gamma-band phase synchrony during typical development, interpreted as the formation of intact cortical local circuitry. Consequently, in typical adulthood a high level of gamma-band phase synchrony is observed. Conversely, in ASD, such tight coupling is not observed during development; in adulthood gamma-band phase synchrony also appears to be reduced compared with TD controls. We conclude that there is a critical developmental window in childhood for the formation of local neuronal circuitry, which is associated with a tight association between GABA levels and gamma-band activity. This association is disrupted in ASD during development, with consequent reduction in electrophysiological evidence of local circuitry functioning in adulthood.