Defects in the pyruvate dehydrogenase (PDH) complex result in severe neurological dysfunction, congenital lactic acidosis, growth retardation, and early death. Current treatments for PDH deficiency are administered postnatally and are generally unsuccessful. Because many patients with this disease are born with irreversible defects, a model system for the development of effective pre-and postnatal therapies would be of great value. In a behavioral genetic screen aimed to identify zebrafish with visual function defects, we previously isolated two alleles of the recessive lethal mutant no optokinetic response a (noa). Here we report that noa is deficient for dihydrolipoamide S-acetyltransferase (Dlat), the PDH E2 subunit, and exhibits phenotypes similar to human patients with PDH deficiency. To rescue the deficiency, we added ketogenic substrates to the water in which the embryos develop. This treatment successfully restored vision, promoted feeding behavior, reduced lactic acidosis, and increased survival. Our study demonstrates an approach for establishing effective therapies for PDH deficiency and other congenital diseases that affect early embryonic development.H uman diseases that affect tissues with high-energy requirements are often caused by defects in mitochondrial function (1). Pyruvate dehydrogenase (PDH) deficiency is a rare metabolic disorder that severely affects the central nervous system due to the high-energy demand of neurons and often results in developmental delay, feeding difficulties, lethargy, ataxia, blindness, and early death (2-4). PDH is a nuclear-encoded mitochondrial matrix multienzyme complex that provides the primary link between glycolysis and the tricarboxylic acid (TCA) cycle by catalyzing the irreversible conversion of pyruvate into acetyl-CoA (5). Defects in PDH result in decreased acetyl-CoA synthesis and the accumulation of pyruvate and lactate, causing reduced energy production and metabolic acidosis, respectively.Current treatments for PDH deficiency, which include the activation of residual PDH with dichloroacetate (6), the administration of cofactors (lipoic acid and thiamine) (7,8), and the provision of ketogenic diets (9, 10), have yielded only limited or variable success (see Fig. 1). The technical challenges of working with small numbers of human patients has made it difficult to include all of the necessary controls and to systematically evaluate therapeutic agents and diets with varying proportions of calories from fat, carbohydrate, and protein (9, 11). Furthermore, preexisting neurological disorders compound the difficulty of evaluating treatments. A therapy that bypasses the entire complex would potentially be of greatest benefit, because it could be used to treat patients with defects in any of the PDH subunits and varying degrees of deficiency.An animal model for PDH deficiency would be of great value to improve current treatments and to allow the development of novel therapies. This model should exhibit phenotypic characteristics similar to those seen in human ...