Relationship between genetic variation in the glutaminase gene GLS1 and brain glutamine/glutamate ratio measured in vivo

Abstract

Background: Abnormalities in glutamatergic neurotransmission are implicated in several psychiatric disorders, but in vivo neurochemical studies of the glutamate (Glu) system have been hampered by a lack of adequate probes. By contrast, glutamine (Gln) and Glu can be quantified separately in proton magnetic resonance spectroscopy studies in vivo. Accumulating evidence suggests that the Gln/Glu ratio is a putative index of glutamatergic neurotransmission but interpretation of changes in the Gln/Glu ratio depends on the conditions of the system, including ammonia levels.

Methods: Here, we explored whether variation in GLS1 (the gene encoding the brain isoform of glutaminase, which catalyzes Gln-to-Glu conversion) is associated with Gln/Glu measured in vivo in two brain regions (anterior cingulate cortex, parieto-occipital cortex).

Results: A specific haplotype of four single nucleotide polymorphisms within GLS1 was significantly associated with Gln/Glu in the parieto-occipital cortex in an magnetic resonance spectroscopy-genetics dataset optimized for Gln/Glu detection (n = 42). This finding was replicated in a second magnetic resonance spectroscopy dataset that was optimized for γ-aminobutyric acid detection where Gln and Glu measurements could still be extracted (n = 40).

Conclusions: These findings suggest that genetic variation in a key component of glutamatergic machinery is associated with a putative in vivo index of glutamatergic neurotransmission. Thus, GLS1 genotype might provide insight into normal brain function and into the pathophysiology of many psychiatric conditions where glutamatergic neurotransmission has been implicated. It might also serve as a biomarker for predicting response to existing and novel therapeutic interventions in psychiatry that target glutamatergic neurotransmission.

Figure 1

Outline of GLS1 on human chromosome 2. The very top panel shows the relevant stretch of DNA with the location of GLS1 depicted. The locations of SNPs included in the analysis are marked with “rs” identifiers. The blue bars beneath SNPs represent the percent of sample genotyped at that SNP in Study 1. The next row lists the minor allele frequency of each SNP. The haplotype block structure is shown next, with the most significant haplotype from Study 1 shaded in yellow. The bottom panel shows the linkage disequilibrium (LD) structure of the SNPs within GLS1 with red squares representing the highest degrees of LD between SNPs.

Figure 2

Gln/Glu ratios measured in the POC for individuals carrying 0, 1, or 2 copies of the target haplotype starting with rs13000464 and ending with rs12185688 that was found to be significant in both Study 1 and Study 2. Values are from Study 1 (left) and Study 2 (right) presented as mean ± SE. Note that mean Gln/Glu values are different across the two studies because of differences in data acquisition and analysis but the same relative pattern is observed.