Synthesis of neurotransmitters serotonin and dopamine, conversion of Phe to Tyr, and breakdown of ether lipids in all animals involve enzymes that require the cofactor 5,6,7,8-tetrahydrobiopterin (BH4) to function. Worms that can't synthesize BH4 are serotonin- and dopamine-deficient, and have leaky, fragile cuticles caused by aberrant lipid metabolism in the hypodermis. The defective cuticle in BH4-deficient worms also alters their susceptibility to bacterial pathogens (Loer et al., 2015, Genetics 200: 237). The first and rate-limiting step in BH4 synthesis is performed by GTP Cyclohydrolase I (GTPCH1, encoded in C. elegans by
cat-4). This first enzyme is regulated by 'end product feedback inhibition' whereby BH4 inhibits GTPCH1 activity. BH4 does not bind to GTPCH1 alone, but requires an additional small protein called 'GTPCH1 feedback regulatory protein' (GFRP) to inhibit GTPCH1. Crystal structures of rat GTPCH1 and GFRP with the BH4 analog BH2 show that the biopterin binds at the interface between the proteins in a large complex (Maita et al., 2004, J Biol Chem 279: 51534). Phe, which stimulates GTPCH1, also binds at this interface, mainly interacting with GFRP. C. elegans encodes an ortholog of GFRP (gene
gfrp-1); sequence analysis and structural predictions suggest worm GFRP will function like the mammalian protein, binding both BH4 and Phe (D. Sykora & C. Loer, unpublished). In mammals, GTPCH1 and GFRP can also bind the selective inhibitor 2,4-Diamino-6-hydroxypyrimidine (DAHP), which is structurally similar to both GTP (substrate) and BH4 (pathway end product). At low concentrations, DAHP apparently inhibits GTPCH1 like BH4 by binding with GFRP; this inhibition is GFRP-dependent. At higher concentrations, DAHP acts as a competitive inhibitor, binding at the GTPCH1 active site like GTP; this inhibition is GFRP-independent (Xie et al., 1998, J Biol Chem 273: 21091). Although DAHP has no obvious effect on wildtype C. elegans, serotonin levels (a proxy for BH4 levels) are reduced by DAHP in worms with a reduction-of-function mutation [
cat-4(
e3015)] that already have less BH4. Preliminary evidence using a
gfrp-1 missense mutant suggests this effect is GFRP-dependent. We are currently testing GFRP function using a CRISPR-generated deletion allele of
gfrp-1.