Previous work in several laboratories has shown that mutations in
eat-4 cause multiple behavioral defects.
eat-4 mutants are feeding abnormal (Eat) (Avery, Genetics 133: 897), insensitive to light touch on the nose (Not) (Kaplan et al., WBG 12(3):105), foraging abnormal (Fab) (J Kaplan, pers. comm.), and thermotaxis defective (Ttx) (Mori, et al., WBG 12(5):73). However,
eat-4 affects only a subset of the nervous system since
eat-4 mutant animals do not show any obvious defects in many other behaviors, including body movement, egg-laying, and mating.
eat-4 is necessary for the function of some or all glutamatergic neurons.
eat-4 is necessary for the function of a pair of neurons in the pharynx called the M3s (Raizen and Avery, Neuron 12:483). The M3s are motor neurons that appear to use glutamate as their neurotransmitter. The
eat-4 mutant defect is presynaptic, suggesting that the M3 neurons (or other neurons that modulate M3) are defective (Dent et al., pers. comm.).
eat-4 also affects the function of a mechanoreceptor neuron AVM. AVM is probably glutamatergic since a recently characterized putative glutamate receptor subunit (AVR-15, Dent et al., pers. comm.) is necessary for AVM synaptic transmission (Lee, et al., WBG 14(1): 46). Yet another glutamate receptor subunit, GLR-1, has been shown to mediate the nose-touch response (Maricq et al., Nature 378:78 and Hart et al., Nature 378:82). The fact that
eat-4 mutants are Not further supports the notion that
eat-4 is necessary for glutamatergic transmission. We are interested in understanding the molecular basis of
eat-4 function: why does
eat-4 affect only a subset of neurons in the worm, and does it play a specific role in glutamatergic neurotransmission? We have cloned
eat-4 by positional cloning and germ-line transformation rescue. We have shown that a genomic promoter::cDNA fusion construct can rescue
eat-4 mutants. We have also shown that the presence of an RFLP in an
eat-4 mutant is correlated with
eat-4 phenotypes.
eat-4 potentially encodes a polypeptide of 576 amino acids with six to eight putative membrane spanning domains. We found that EAT-4 shares extensive sequence similarity (46% identity) with a brain and neural specific sodium-dependent inorganic phosphate cotransporter discovered in rat (BNPI, Ni et al., PNAS 91: 5607). We have further shown, using reporter constructs whose expression is driven by the
eat-4 promoter, that
eat-4 may be expressed in more than 24 neurons, including the M3 and AVM neurons. Our results suggest that EAT-4, as a possible Na-Pi cotransporter, is required in glutamatergic neurons for their normal function. In addition to M3 and AVM, at least some of the other neurons expressing EAT-4 may also be glutamatergic, eg, neurons mediating the nose-touch response. So far, there has been no known function assigned to the mammalian BNPI. Our results indicate that BNPI-like proteins may have a critical role in neural transmission.