Over the last fifty years the search for the biochemical basis of schizophrenia
has focused on diVerent subjects at diVerent times. The first specific
biochemical hypothesis was put forward in 1952 by Osmond and Smythies.
This was the transmethylation hypothesis that was based on the chemical
similarity between the psychotomimetic drug mescaline and the neurotransmitter
adrenaline. Later similar hypotheses were advanced based on
the chemical similarity between other psychotomimetic agents, such as
dimethyltryptamine and o-methylbufotenin, and the neurotransmitter
serotonin. It was supposed that a simple genetically induced metabolic
fault might result in the synthesis of some neurotoxic compound, as in
the case of phenylketonuria. However, as no such compounds were
detected in either the normal or the schizophrenic brain in any meaningful
quantity, interest in these hypotheses lapsed. The next candidate hypothesis
was the adrenochrome theory of HoVer, Osmond and Smythies
(1954). Adrenochrome is the oxidative product of adrenaline and is a
psychotomimetic compound (HoVer et al., 1954; Grof et al., 1963). But,
as adrenochrome was not detected in vivo, interest in this theory also
lapsed. However, recently interest has revived as it is now apparent that
adrenochrome may be formed normally in the C2 adrenergic group in
the medulla. Some adrenergic neurons in this nucleus are pigmented
and the pigment is probably neuromelanin (Gai et al., 1993). In the synthesis
of neuromelanin from adrenaline adrenochrome is an essential intermediary
metabolite. Furthermore, catecholamine o-quinones (including
adrenochrome) are, in part, detoxified by 5-conjugation with glutathione
S-transferase. Harada et al. (2001) report that schizophrenics have an increased
frequency of deletion of the gene for this enzyme (p ¼ 0.0075)
and an even higher rate in the subgroup of disorganized schizophrenics
(p ¼ 0.0008). Neurotoxic catecholamine o-quinones may play a normal
role in synapse deletion. There is also now strong evidence to support a
role for such o-quinones derived from dopamine in Parkinson’s disease
(Smythies 2002). So this hypothesis merits further research.