The morphology of a neuronal dendrite determines what signals it receives and how those signals are integrated. For example, the positions and shapes of dendritic spines determine synaptic partners and synapse strength. In C. elegans, many sensory neurons extend dendrites that terminate at the nose in a sensory cilium ensheathed by glia, and we previously identified genes required for their morphogenesis. In contrast, the oxygen-sensing neuron URX extends a dendrite to the nose that lacks a cilium, is not ensheathed by glia, and does not require any of the genes we previously identified. Therefore, to identify novel regulators of dendrite morphogenesis, we performed an unbiased visual genetic screen for mutants with defects in URX dendrite morphogenesis. We isolated 17 recessive mutants exhibiting four classes of phenotype: (I) short dendrites; (II) ectopic branching; (III) disorganized dendrites, and (IV) dendrite overgrowth, in which the URX dendrite extends up to ~150% of its normal length. We identified two genes that cause this overgrowth phenotype: C05D10.2/mapk-15, a previously uncharacterized gene encoding a kinase with homology to mammalian MAPK15/ERK8, and
sma-1, a homolog of beta-H spectrin. Both
mapk-15 and
sma-1 phenotypes appear during L4 stage and become increasingly pronounced as the animal ages. We determined via mosaic and rescue experiments that MAPK-15 acts cell-autonomously in URX neurons and requires its kinase activity to function. The URX dendrite ending is enriched in signaling molecules including the receptor-type guanylyl cyclase GCY-35, and we found that MAPK-15 also localizes to the dendrite ending in wild-type animals. Further, in
mapk-15 mutants, GCY-35 becomes localized throughout the region of dendrite overgrowth, suggesting the additional dendrite length reflects expansion of this sensory compartment. Surprisingly, overexpression of wild-type, but not inactive, GCY-35 suppresses the
mapk-15 phenotype, suggesting that the size of this sensory compartment may be regulated by sensory signaling in a manner epistatic to
mapk-15. Together, these results suggest that
mapk-15 may mediate activity-dependent size regulation of a dendritic sensory compartment, possibly analogous to regulation of dendritic spine size and shape.