Melanopsin and clock genes: regulation by light and endothelin in the zebrafish ZEM-2S cell line.

It is well known that clocks are present in brain regions other than the suprachiasmatic nucleus and in many peripheral tissues. In the teleost, Danio rerio, peripheral oscillators can be directly synchronized by light. Danio rerio ZEM-2S embryonic cells respond to light with differential growth: cells kept in constant light exhibited a strong inhibition of proliferation, whereas in cells kept in light:dark (LD) cycles (14L:10D and 10L:14D) or in constant darkness (DD), the doubling times were not statistically different. We demonstrated by RT-PCR followed by PCR that ZEM-2S cells express two melanopsins, Opn4x and Opn4m, and the six Cry genes. The presence of the protein OPN4x was demonstrated by immunocytochemistry. The pattern of temporal expression of the genes Opn4x, Per1, Cry1b, and Clock was studied in ZEM-2S cells kept for five days in 12L:12D or DD. In 12L:12D, the clock genes Per 1 and Cry1b exhibited robust circadian expression, while Opn4x and Clock expression seemed to vary in an ultradian pattern. Both Per1 and Cry1b genes had higher expression during the L phase; Clock gene had an increase in expression coincident with the D phase, and during the subjective night. In DD, the temporal variation of Per1 and Cry1b genes was greatly attenuated but not extinguished, and the higher expressions were shifted to the transition times between subjective day and night, demonstrating that Per and Cry1b were synchronized by the LD cycle. Clock and Opn4x kept the ultradian oscillation, but the rhythm was not statistically significant. As endothelins (ET) have been reported to be a potent stimulator of Per genes in rodents, we investigated the effect of endothelin on ZEM-2S cells, which express ETA receptors. Cells were kept in 12D:12L for five days, and then treated with 10(-11) to 10(-8)M ET-1 for 24 h. ET-1 exhibited a biphasic effect on Opn4x expression. At 10(-11)M, the hormone exerted a highly significant stimulation of Opn4x expression during the L phase and introduced a circadian oscillatory pattern. At 10(-10)M, a significant increase was seen at ZT21 and ZT0 (i.e., at the end of the D phase and beginning of the L phase), whereas 10(-9) and 10(-8)M ET-1 inhibited the expression of Opn4x at most ZTs. Clock expression was unaffected by 10(-8)M ET-1; however, in the presence of lower concentrations, the expression was enhanced at some ZTs, strengthening the ultradian oscillation. ET-1 at 10(-11) and 10(-10)M had no effect on Per1 circadian expression; however, 10(-9) and 10(-8)M ET-1 reduced the amplitude of Per1 expression in the beginning of the L phase. ET-1 effects were less evident on Cry 1b. For both genes, the reduction in expression was not sufficient to abolish the circadian oscillatory pattern. Based on these results and data in the literature, a link between ET-1 stimulation of ETA receptors may be established by E4BP4 binding to the promoters and consequent inhibition of gene expression.