Perturbed Cholesterol Metabolism in Autism Spectrum Disorders: Biochemical and Genetic Analyses

Background: Smith-Lemli-Opitz syndrome (SLOS) is a disease caused by mutations of the  gene encoding 7-dehydrocholesterol reductase (DHCR7), resulting in accumulation of 7-dehydrocholesterol (7DHC) and impaired cholesterol synthesis In an earlier study, we reported that as many as 85% of children with SLOS met criteria for an autism spectrum disorder (ASD). In a recently reported study, sterols were measured in a group of individuals with ASD, and although no sterol profiles were diagnostic of SLOS, low cholesterol levels (<5^thpercentile) were noted in approximately 20% of the study subjects. These studies suggest that cholesterol deficiency or perturbed sterol metabolism is a risk factor for autism.

Objectives: The objective of the study was to determine if children with ASD have perturbed cholesterol metabolism or sequence variations in the DHCR7 gene. We hypothesized that children with ASD would have lower plasma (pl-) cholesterol and 24S (24S hydroxycholesterol, a measure of brain cholesterol turnover) concentrations, elevated pl-7DHC, and a higher prevalence of the DHCR7sequence variants than children who came to the autism clinic for a evaluation for autism for which an ASD diagnosis was ruled out.

Methods: Case /control study consisting of 69 children (42 ASD, 27 non-ASD or controls)  referred to the OHSU Autism Clinic for behavioral evaluation. Plasma cholesterol and 7DHC concentrations were measured by gas chromatography and 24-S-hydroxycholesterol (24S) concentration was measured by liquid chromatography-mass spectrometry. Sequence variants and pathogenic mutations of the DHCR7gene were identified By Sanger sequencing

Results: The two groups were matched for gender, race and body-mass index (BMI).  The median age was lower in the ASD group (30.7 months) than in controls (41.7 month, p<0.03). Plasma cholesterol and 7DHC concentrations were similar in both groups, with low cholesterol levels (<5^th percentile) noted in only 2.4% of the subjects with ASD (3.8% in non-ASD subjects). In contrast, 24S concentration was higher in subjects with ASD (104.1±5.8 ng/mL) than in controls (78.6 ±5.2 ng/mL; p<0.004). This group difference remained statistically significant after adjustment for age (p<0.015). Twenty-six participants (37.7% of the study population) were found to have either a single nucleotide variant or a mutation. The percentage of participants with either a mutation or a variant was significantly greater in ASD (21 subjects, 50%) than in controls (5 subjects, 19%, p=0.01). Interestingly, only children with an ASD were found to have mutations. However, because there were only 3 mutations in the ASD group (7.1%), the difference with controls (no mutation) was not statistically significant.

Conclusions: The study does not confirm the previously reported high prevalence of low plasma cholesterol levels in ASD. However, brain cholesterol turnover may be higher in children with ASD, and DHCR7 sequence variants are more prevalent in children with ASD than controls, suggesting that perturbations in cholesterol synthesis may predispose to ASD.