How do some plants know to flower come the long days of summer? Although plant biologists have discovered some of the components of the circadian system that controls this so-called photoperiodic response, they have struggled to understand how both temporal and light information are integrated by plants to control flowering time. Now Marcelo Yanovsky and Steve Kay identify the Arabidopsis transcriptional activator CONSTANS (CO) as the point at which this integration takes place. They show, for the first time, that light has a direct effect on the ability of Arabidopsis to sense and measure a day's length, in addition to its well-known effects on establishing circadian rhythms.

CO is a key component of the photoperiodic response. Its expression is under circadian control — during short days, its daytime expression levels are low and rise only after sunset, whereas during long days, CO mRNA levels start accumulating towards dusk. A direct target of CO is FLOWERING LOCUS T (FT). Its expression also peaks at dusk during long days, at the time when the rise in CO daytime expression coincides with an illuminated part of the day. Could CO function therefore be light dependent?

To investigate this, the authors began by analysing the importance of the circadian control of CO for daylength discrimination in the Arabidopsis mutant toc1 . This mutant flowers early in short days, and was used because its photoperiodic defect is due only to a circadian — and not to a light-response — defect. In toc1 plants, overall CO expression levels remain relatively normal; however, the phase of CO expression is significantly advanced under short-day conditions, resulting in the accumulation of high levels of CO at the illuminated end of the day. FT also accumulates at this time under similar conditions in toc1, but not in wild-type, plants, indicating that the earlier shift in CO expression induces FT expression and that the circadian control of CO expression is required for daylength discrimination. Moreover, the high expression of FT during short days is probably the molecular defect that underlies the toc1 phenotype, because the phenotype was abrogated when FT was mutated in toc1 plants.

So, what is the role of light in this response? When the authors assayed FT expression in plants that constitutively express CO, these plants showed high but rhythmic (peaking at dusk) FT expression patterns under long-day conditions. This pattern, however, depended on exposure to light, indicating that CO regulation of FT is light dependent. Because FT levels are greatly reduced in plants that are mutant for the photoreceptor CRY2, and because these plants have apparently normal CO expression patterns, the authors considered it to be a strong candidate mediator of CO's light-dependent regulation of FT. In fact, they found that both CRY2 and the photoreceptor PHYA are required for light-induced upregulation of FT expression and that this response requires functional CO.

These findings strongly indicate that daylength can regulate flowering time through the coincidence of light — as detected by CRY2 and PHYA — with a particular circadian phase, as manifested by high levels of CO expression. Together these events cause a rise in FT expression, which triggers flowering. The pathways that control this flowering feat are, however, in need of further illumination.