Glutamate and GABA in autism spectrum disorder-a translational magnetic resonance spectroscopy study in man and rodent models

Abstract

Autism spectrum disorder (ASD) is a pervasive neurodevelopmental syndrome with a high human and economic burden. The pathophysiology of ASD is largely unclear, thus hampering development of pharmacological treatments for the core symptoms of the disorder. Abnormalities in glutamate and GABA signaling have been hypothesized to underlie ASD symptoms, and may form a therapeutic target, but it is not known whether these abnormalities are recapitulated in humans with ASD, as well as in rodent models of the disorder. We used translational proton magnetic resonance spectroscopy ([1H]MRS) to compare glutamate and GABA levels in adult humans with ASD and in a panel of six diverse rodent ASD models, encompassing genetic and environmental etiologies. [1H]MRS was performed in the striatum and the medial prefrontal cortex, of the humans, mice, and rats in order to allow for direct cross-species comparisons in specific cortical and subcortical brain regions implicated in ASD. In humans with ASD, glutamate concentration was reduced in the striatum and this was correlated with the severity of social symptoms. GABA levels were not altered in either brain region. The reduction in striatal glutamate was recapitulated in mice prenatally exposed to valproate, and in mice and rats carrying Nlgn3 mutations, but not in rodent ASD models with other etiologies. Our findings suggest that glutamate/GABA abnormalities in the corticostriatal circuitry may be a key pathological mechanism in ASD; and may be linked to alterations in the neuroligin-neurexin signaling complex.

Fig. 1. Comparison of glutamate, glutamine and GABA levels observed by [1H]MRS in striatum and mPFC of individuals with ASD and control subjects.

Left panels: striatal levels (in institutional units) of a glutamate, c glutamine, and e GABA. Right panels: mPFC levels (in institutional units) of b glutamate, d glutamine, and f GABA. Data are depicted as mean ± SEM of n = 25–36 per group; *p < 0.05; two-tailed t-test. This difference remains significant with the removal of the putative outlier in the control group, p = 0.037. ASD autism spectrum disorder, i.u. institutional units, mPFC medial prefrontal cortex. For additional data see Supplementary Table 1

Fig. 2. Comparison of glutamate, glutamine and GABA concentrations observed by [1H]MRS in the striatum and mPFC of rodent models of ASD and their corresponding controls.

Left panels: concentrations (in mM) of a glutamate, c glutamine, and e GABA in striatum. Right panels: concentrations (in mM) of b glutamate, d glutamine, and f GABA in mPFC. All the mouse models (colored bars) were on a C57BL/6J background, except for VPA (CD1) and BTBR T+tf/J mice. The rat model (colored bars) was on a Sprague-Dawley background. Wild-type (gray bars) littermates were used as controls, except for VPA and BTBR T+tf/J mice for which age-matched saline-treated CD1 and wild-type C57BL/6J mice served as respective controls. Note the strain-dependent differences in glutamate levels. Data are depicted as mean ± SEM of n = 7–15 per group; *p < 0.05, **p < 0.01, ***p < 0.001; two-tailed t-test. KI knock-in, KO knockout, mPFC medial prefrontal cortex, patDp paternal duplication, VPA valproic acid. For additional data see Supplementary Table 2