Identifying targets of FLC at 27oC by expression profiling





FLOWERING LOCUS C (FLC) is a MADS box transcription factor that plays a well characterised role in repressing the vegetative to floral transition of Arabidopsis thaliana. FLC has also been shown to affect the Arabidopsis circadian clock, with mutant seedlings showing short circadian periods. In a previous study, we identified the temperature-dependent circadian period QTL PerCv5b near the FLC locus on the top arm of Chromosome 5. PerCv5b caused a significant period effect at 27oC but not at 12oC or 22oC. Temperature-dependent circadian period phenotypes and a known polymorphism in the Ler allele made FLC a strong candidate gene for PerCv5b. The period effect of FLC was enhanced by combination with alleles of FRIGIDA (FRI), a gene shown to up-regulate FLC's expression.
We were interested in identifying how FLC affects the circadian clock, so we decided to identify its target genes. Greatest phenotypic difference was observed between fri; flc and FRI; FLC genotype seedlings at 27oC, so expression was compared between these lines (previously described in Michaels and Amasino1999 and 2001) on the Affymetrix ATH1 microarray. Seedlings were grown on media (MS 1.5% Agar containing 3% Sucrose) for 6 days under constant cool white fluorescent light (55-60 micro EINSTEINS) at 22oC then entrained for 4 days under (12h , 12h) light , dark cycles at 22oC. At dawn on the fourth day of entrainment they were transferred to constant light (25-30 micro EINSTEINS) at 27oC. Four samples were taken at 6 hour intervals starting 24h after the transfer to continuous conditions at the times 24h, 30h, 36h and 42h.
Equal amounts of total RNA were pooled from the three time points to produce one sample per genotype. The pooling strategy was employed to reduce the effect of circadian regulation on genes expression. This was particularly important in our case as some interesting genes would likely be regulated by the circadian clock and may only show expression differences at particular phases that could easily be missed if using just one time point.
Date made available2006
PublisherUniversity of Edinburgh

Cite this