X-Linked Imprinted Genes and Sex-Ratio Bias in Autistic Spectrum Disorders

Background:   Two prevailing theories attempt to explain the strong male bias in the occurrence of Autistic Spectrum Disorders (ASD). Baron-Cohen has proposed that the male prevalence can be ascribed to fetal exposure to testosterone, which may masculinize developing neural circuitry exacerbating an underlying predisposition to an “extreme male brain”. Contrarily, Skuse has hypothesized that the male prevalence to ASD can be attributed to hemizygosity of the X chromosome in boys, exposing them to perturbations in gene dosage brought about by X chromosome epigenetic phenomena. Despite their prominent treatment in the literature over the past decade, neither theory has gained much traction due to a lack of general and compelling physiological data supporting the former and genetic or epigenetic data supporting the latter.

Objectives:   We hypothesize that the male bias in ASD may be attributable to the influence of parent-of-origin expression (i.e. imprinting) of X-linked genes. Since males receive only the maternal X chromosome, the true effect is one of “grandparent-of-origin”. In other words, the single X in males is a mosaic of alleles passed from either the maternal grandmother or maternal grandfather. Epigenetic signatures that are disrupted in the grandparents, or are improperly established or reset in the mother may be passed to her sons. While transgenerational effects of parental or grandparental age in ASD susceptibility have been investigated, a genome-wide or candidate gene approach has not been attempted.

Methods:   In a candidate gene approach to identify X-linked imprinted genes we are employing whole transcriptome analysis via microarray and RNA-seq on brain RNA from X monosomic mouse models. It is clear that current GWAS have underutilized data available for the X chromosome. In a genome-wide association approach we are testing for statistical correlation between maternal grandparental X chromosome SNP profile and incidence of ASD in males utilizing novel analytical tools.

Results:   We have identified an X-linked gene that exhibits parentally biased expression: Transketolase-like 1 (Tktl1). Tktl1 exhibits region-specific transcriptional repression of the paternal allele in neonatal mouse brain. TKTL1 maps to the syntenic region (Xq28) of the human X and we have recently confirmed imprinted expression of this gene in human fetal brain tissue. Our genome-wide approach to uncover transgenerational X-linked effects is ongoing.

Conclusions:   TKTL1 encodes a transketolase enzyme that constitutes a rate-limiting step in the bifurcated pentose phosphate pathway (PPP) of glycolysis. The cyclic portion of the PPP is the primary pathway for the production of NADPH, which is necessary for the maintenance of redox potential in cells, protecting them from oxidative damage. Numerous studies in the last several years have established a link between oxidative stress and autism. Our work supports the hypothesis that the male bias in ASD occurrence may be attributable to X chromosome epimutation.