Disrupted Glutamate Signaling in Drosophila Generates Locomotor Rhythms in Constant Light.
2020-05-22T14:08:48Z (GMT) by
We have used the Cambridge Protein Trap resource (CPTI) to screen for flies whose locomotor rhythms are rhythmic in constant light (LL) as a means of identifying circadian photoreception genes. From the screen of ∼150 CPTI lines, we obtained seven hits, two of which targeted the glutamate pathway, Got1 (Glutamate oxaloacetate transaminase 1) and Gs2 (Glutamine synthetase 2). We focused on these by employing available mutants and observed that variants of these genes also showed high levels of LL rhythmicity compared with controls. It was also clear that the genetic background was important with a strong interaction observed with the common and naturally occurring timeless (tim) polymorphisms, ls-tim and s-tim. The less circadian photosensitive ls-tim allele generated high levels of LL rhythmicity in combination with Got1 or Gs2, even though ls-tim and s-tim alleles do not, by themselves, generate the LL phenotype. The use of dsRNAi for both genes as well as for Gad (Glutamic acid decarboxylase) and the metabotropic glutamate receptor DmGluRA driven by clock gene promoters also revealed high levels of LL rhythmicity compared to controls. It is clear that the glutamate pathway is heavily implicated in circadian photoreception. TIM levels in Got1 and Gs2 mutants cycled and were more abundant than in controls under LL. Got1 but not Gs2 mutants showed diminished phase shifts to 10 min light pulses. Neurogenetic dissection of the LL rhythmic phenotype using the gal4/gal80 UAS bipartite system suggested that the more dorsal CRY-negative clock neurons, DNs and LNds were responsible for the LL phenotype. Immunocytochemistry using the CPTI YFP tagged insertions for the two genes revealed that the DN1s but not the DN2 and DN3s expressed Got1 and Gs2, but expression was also observed in the lateral neurons, the LNds and s-LNvs. Expression of both genes was also found in neuroglia. However, downregulation of glial Gs2 and Got1 using repo-gal4 did not generate high levels of LL rhythmicity, so it is unlikely that this phenotype is mediated by glial expression. Our results suggest a model whereby the DN1s and possibly CRY-negative LNds use glutamate signaling to supress the pacemaker s-LNvs in LL.