Although a number of studies have documented abnormal expression of metabolites in blood and urine samples from individuals with autism, no clear metabolic biomarkers or unifying concepts regarding metabolic dysregulation and brain development have emerged from such studies.

Valerie Hu and her colleagues at George Washington University in Washington, D.C., investigated the feasibility of analyzing frozen human brain tissues from individuals with autism and from controls in order to identify specific metabolite differences in the brains of individuals with autism.

The researchers employed a novel technique called laser ablation electrospray ionization mass spectrometry (LAESI-MS), developed by co-investigator Akos Vertes of the chemistry department at George Washington University. The technique allows unbiased detection of tissue metabolites in the tissue itself without solvent extraction. The researchers coupled this metabolite analysis with whole-transcriptome expression analyses of the same tissues to help identify autism-linked metabolic pathways and neurological functions impacted in the autism brain.

Their analyses showed relatively few metabolite differences between the brain tissue of adults (ages 29 to 39) with autism and that of age-matched controls. However, the LAESI-MS studies of the frontal cortex indicated increases in creatine and choline, consistent with the finding of magnetic resonance imaging studies on living adults. They also showed about twofold increases in pantetheine or a-CEHC (derivatives of vitamin B5 and E, respectively), suggesting possible dysregulation of these biosynthetic pathways.

Although it is unclear how the observed increases in these specific vitamin-associated metabolites relate to brain function in autism, they may provide useful biomarkers for diagnostic screening. They may prove particularly useful if increases in these metabolites can be reliably detected in the blood or urine of individuals with autism and if they can be detected at early ages.

Gene expression profiling of postmortem brain tissues from individuals with autism and controls provided information on various metabolic pathways involved in autism, which prominently includes lipid metabolism. Although the researchers could not detect a one-to-one correspondence between the metabolites identified by mass spectrometry and those implicated by gene expression studies, these data are consistent with the substantial differences in lipid profiles observed by the metabolomic analyses between autism brain tissue and control brain tissue.

The expression analyses also revealed significant differences in alternative splicing — the use of different exons in producing a final protein product — of genes in the brain tissue from people with autism compared with that of the controls. The differences affect folic acid and vitamin D metabolism, in addition to a number of signaling pathways and functions implicated in autism.

These studies showed that although many metabolic and signaling pathways may be altered in autism, the overall metabolite profile in the brains of adults with autism is, with the exception of lipids, remarkably similar to that of controls, perhaps reflecting the importance of metabolic homeostasis in the brain.