e., Pdf-Gal4 > UAS-Mef2 flies, was associated with substantial arrhythmicity as previously reported ( Blanchard et al., 2010) ( Table 2). Flies with decreased Fas2 levels in LNvs also manifest constant defasciculation of s-LNv axons (albeit a weaker morphological phenotype than Mef2 overexpression; Figures 3B, 3C, and data not shown), and these flies had a substantially weaker behavioral phenotype than Pdf-Gal4 > UAS-Mef2 flies, namely, only

about 80% rhythmic flies on days 1–4 of DD and 69% rhythmic flies on days 5–8 compared to ∼98% for control strains (p < 0.01 Fisher’s test, Table 2). Similar morphological and behavioral phenotypes (p > 0.5 Fisher’s test, Table 2) were observed with Pdf-GAL4 > UAS-Fas2RNAi/UAS-Mef2RNAi flies. Importantly, overexpression of Fas2 in the Pdf-Gal4 > UAS-Mef2 background not only rescued the check details constant defasciculation of the background strain but also significantly increased the percentage of rhythmic flies (p < 0.01 Fisher’s test, Table 2). There was no significant change in selleck kinase inhibitor rhythmicity due to the addition of an extra UAS element, i.e., PDF-GAL4 > UAS-Mef2/UAS-mCD8GFP is indistinguishable from Pdf-Gal4 > UAS-Mef2 (p > 0.5 Fisher’s test Table 2). These data strongly indicate that PDF neuron defasciculation contributes to the Mef2 overexpression phenotype. How is Mef2 itself regulated? CLK and CYC

ChIP-Chip experiments in our laboratory identified Mef2 as a direct target of CLK and CYC ( Abruzzi et al., 2011), and the Mef2 promoter manifests canonical cycling of CLK/CYC binding with peak levels at ZT14 ( Figure 5A). Indeed, previous expression analysis ( Kula-Eversole et al., 2010) demonstrated that Mef2 transcript levels cycle in l-LNvs with a peak phase consistent with this rhythmic CLK binding ( Figure 5B). As Mef2 transcript levels do not oscillate in whole Drosophila heads (see Figure 5B;

McDonald and Rosbash, 2001), we speculate that Mef2 is regulated by rhythmic CLK binding only in certain cell types (see Discussion). This notion is in agreement with the previously observed decrease of Mef2 staining levels within no PDF neurons in the clk and cyc mutants, Clkar and cyc01, respectively ( Blanchard et al., 2010). To verify that the link between CLK and neuronal plasticity goes through Mef2, we assayed the epistatic relationship between Clk and Mef2. As the loss-of-function Clk mutant ClkJrk leads to loss of s-LNv neurons ( Park et al., 2000; data not shown), we used RNAi to decrease Clk activity levels in PDF cells. The knockdown causes arrhythmic locomotor behavior (F. Guo and M.R., unpublished data) and disrupts rhythmic remodeling of s-LNv projections as expected. In addition to the loss of circadian plasticity, the Clk knockdown causes an overfasciculated phenotype, also characteristic of the Mef2 RNAi knockdown ( Figure 5C).