Acyl-Carnitine Abnormalities In Autistic Children Parallel Abnormalities In A Rodent Model of Autism

Background: Mitochondrial dysfunction has been suggested to explain the complex medical and physiological abnormalities found in some children with autism. However, only 23% of children with mitochondrial disease and autism have a known mitochondria DNA abnormality to explain mitochondrial dysfunction. Some have suggested that the systemic abnormalities seen in autism may arise from environmental triggers in genetically sensitive subpopulations. Mitochondrial are central to this theme as polymorphisms in mitochondrial genes can result in susceptibility to many diseases. Interestingly, mitochondrial dysfunction can be triggered by enteric short chain fatty acids such as propionic acid (PPA) that can be produced as by-product by opportunistic enteric bacteria that have been implicated in autism (i.e. Clostridia,  Desulfovibrio and Bacterioridetes). We have developed an animal model of autism in which intraventricular infusions of PPA produces reversible bouts of autistic-type behaviors. This animal model also demonstrates several characteristics that have been reported in autism such as redox, mitochondrial and acyl-carnitine abnormalities.  

Objectives: To determine whether biochemical abnormalities found in our animal model of autism are also in a subset of children with autism, specifically, we sought to determine whether the pattern of acyl-carnitine elevations, redox abnormalities and mitochondrial dysfunction found in our rodent model could be found in at least a subset of children with autism.

Methods: Fasting acyl-carnitine panel was measured in 213 patients with autism. A workup for secondary causes of fatty-acid oxidation and mitochondrial disorders was recommended for patients with consistent (two or more occasions) elevations in three or more acyl-carnitine species. Mitochondrial and/or nuclear DNA gene abnormalities are examined in a subset as was muscle and/or skin biopsy with functional fatty-acid oxidation pathway and electron transport chain (ETC) testing. Markers of redox metabolism were also examined in a subset. 

Results: Overall, 17% of children with autism had consistent elevations in multiple acyl-carnitines. Statistically significant elevations were found in short (C4OH) and long chain (C14, C16:1), but not medium chain, acyl-carnitines. Examination of the ETC in muscle and fibroblasts demonstrated great variability across individual complex function with particular deficits in the interaction of complex III with complex I or II. In fibroblasts, on average, revealed a relative deficiency in complex II/III was found. Examination of the fatty-acid oxidation pathway revealed no abnormalities except for those secondary to ETC abnormalities. Abnormalities in mitochondrial genes responsible for cytochrome b, an important component of complex III, were also identified in two patients but the majority of patients did not have any genetic abnormalities to explain the metabolic abnormalities. Redox abnormalities were also found in these children. 

Conclusions:  We identified a subset of children with autism with a pattern of acyl-carnitine abnormalities that is similar to our rodent models of autism. Like the rodent model, these children also have abnormalities in ETC function and redox abnormalities.  Few patients demonstrated genetic defects to explain the mitochondrial abnormalities, leaving open the possibility that environmental factors could be resulting in mitochondrial dysfunction, similar to the rodent model.