We are interesting in learning how neurons in C. elegans decide to use the neurotransmitter serotonin. Serotonin is synthesized from tryptophan in two steps: Trp is first hydroxylated by an aromatic amino acid hydroxylase (AAAH), then decarboxylated by an L-aromatic amino acid decarboxylase (AAADC, aka dopa decarboxylase/DDC) to yield serotonin. We have previously identified genes encoding both enzymes in C. elegans from ORFs first predicted by the Genome Sequencing Consortium. We have now rescued the serotonin- and dopamine-deficient mutant
bas-1 by injection of a 15.1 kb subclone of the cosmid C05D2 containing two predicted AAADC genes (we are indebted to Fred Wolf, Garriga lab, for this clone). It is noteworthy that the two predicted genes, C05D2.4 and C05D2.3, are separated by only 369 bp, so may be transcribed together. Also, the C05D2.3 predicted protein lacks a region found in all other AAADC's, so may have evolved a novel function. We are interested in determining whether one or both of these ORFs is needed for serotonin synthesis. We are continuing to characterize these and other terminal differentiation genes used by serotonergic neurons. Learning how these genes are controlled is important to understanding how a neuron chooses a particular neurotransmitter during its development. All AAAH enzymes, which include tryptophan hydroxylase (serotonin synthesis), tyrosine hydroxylase (dopamine synthesis) and phenylalanine hydroxylase (tyrosine synthesis/phenylalanine catabolism), require a biopterin cofactor to catalyze their respective hydroxylations. The mutant
cat-4 (
e1141)V is serotonin- and dopamine-deficient (Sulston et al., 1975, J. Comp. Neurol. 163: 215; Desai et al., 1988, Nature 336: 638) and appears to have a leaky cuticle. Sequencing through the region in which
cat-4 maps genetically has identified a likely candidate on cosmid F32G8: a predicted gene with homology to the enzyme GTP cyclohydrolase I (GCH), which catalyzes the first step in biopterin cofactor synthesis. Mutation in the Drosophila Punch locus, which encodes a GCH, causes a phenotype similar to that of the
cat-4 mutant (O'Donnell et al., 1989, Dev Genet 10:273). We are currently injecting the F32G8 cosmid into
cat-4 mutants, subcloning the GCH gene, and constructing reporter fusions. This work is supported by NSF (RUI) Grant IBN9796217 to CML.