Organisms use specific proteins to sense light: rhodopsin, phytochrome, xanthopsin, cryptochrome, phototropin and BLUF sensors. C. elegans uses a novel protein: the worm's avoidance of shortwave light requires LITE-1, a Drosophila gustatory GPCR homolog (Edwards..Miller 2008). We found that light inhibits pharyngeal pumping. This inhibition is partly independent of
lite-1, indicating that a second light-sensing mechanism functions in the worm. Through laser ablation of pharyngeal neurons we found that loss of the I2 neuron increases the time it takes for light to inhibit pumping. Furthermore, light rapidly increases I2 calcium. I2 calcium increases even in
unc-13 mutants defective in synaptic signaling, suggesting that I2 might directly sense light. Two I2-expressed genes are required for both rapid inhibition of pumping and I2 calcium increase:
gur-3, a
lite-1 paralog, and
prdx-2, a peroxiredoxin. Since peroxiredoxins are antioxidants, we tested the effect of hydrogen peroxide (H2O2) on pumping. H2O2 odor elicits both avoidance and pumping inhibition. H2O2 odor and light act through the same mechanism to inhibit pumping, as both depend on
gur-3 and
prdx-2.
lite-1 is required for H2O2 avoidance, and the
lite-1 gur-3 double mutant is completely defective in the pumping response to H2O2, just like with light. One possibility is that light produces H2O2. H2O2 can be detected by the generation of Prussian blue (Saito..Yoshida 2007). We found that shortwave light also generates Prussian blue in water, likely through the production of H2O2 as addition of catalase reduces pigment formation. Downstream of
gur-3 and
prdx-2,
eat-4, a vesicular glutamate transporter, functions in I2 to inhibit pumping, suggesting that I2 releases glutamate in response to light/H2O2. Two glutamate receptors,
avr-15 and
glc-2, are required for the rapid inhibition of pumping. These receptors are expressed in pharyngeal muscle and function there to inhibit pumping in response to light/H2O2. Overall, these results describe the molecular and cellular components of a light-sensing circuit. Sensing light via an H2O2 intermediary suggests that light-sensing in the worm might have evolved from a pre-existing H2O2 avoidance mechanism.