Methylation hypothesis.

L-Methionine had no behavioral effects in normal humans and failed to increase concentrations of S-adenosylmethionine (methyl donor) in human or rat blood, while increasing rat liver levels more than fivefold. Methionine or S-adenosylmethionine in very high doses had almost no effect on methylation of tritiated levodopa in rodent tissues; various "methyl acceptor" molecules, including nicotinamide, guanidineacetic acid, and estradiol similarly had little effect. In rabbit lung, methionine and S-adenosylmethionine not only failed to increase production of dimethyltryptamine, but actually decreased it, possibly due to end-product inhibition by S-adenosylhomocysteine, which also strongly inhibited methylation of dopa in rat. These results fail to support several predictions of the "methylation hypothesis" concerning the pathophysiology and potential treatment of idiopathic psychotic disorders and leave the consistent clinical worsening effects of methionine in schizophrenia unexplained.