|n the April 12, 2002 issue of Science (vol. 296:285-289), Gordon G. Simpson and Caroline Dean (Department of Cell and Developmental Biology, John Innes Centre, UK) review mechanisms by which environmental and endogenous factors regulate flowering time in Arabidopsis. A fundamental difference between plants and animals is that plant reproductive development takes place in response to multiple types of environmental cues. Furthermore, most plant development occurs post-embryonically when shoot and root apices continue to generate new cells through meristematic activity. The review focuses upon the molecular pathways that regulate flowering time in Arabidopsis, describes the mechanisms of environmental influence on these pathways that result in differing flowering time phenotypes in response to certain environmental conditions, and discuss mechanisms of flowering time regulation in other species.
In Arabidopsis, the transition from vegetative development to reproductive development is influenced by environmental factors such as exposure to a long period of cold temperatures (vernalization), ambient temperatures, day length, proximity to other plants, and nutrient and water availability. Floral identity genes such as LEAFY (LFY), APETALA1 (AP1), CAULIFLOWER (CAL) and FRUITFULL (FUL) control the transition to flowering. The authors describe the pathway through which vernalization controls flowering time. In this pathway, the FRIGIDA (FRI) gene conditions the plant to overwinter, thereby delaying flowering. The protein encoded by FRI promotes the accumulation of FLOWERING LOCUS C (FLC) transcript. FLC is a strong negative regulator of floral pathway integrator genes responsive to various environmental conditions, and up-regulation of FLC thereby suppresses the transition to flowering. Vernalization causes a reduction in FLC transcript levels, and mRNA levels remain low even after returning to warm temperatures. FLC is probably not the only target of this pathway, as loss-of-function flc mutants are still responsive to vernalization. Vernalization prepares the plant to flower but does not itself elicit flowering, and it is thought that an epigenetic event such as chromatin remodeling may be involved. Vernalization-defective mutants have allowed identification of genes required for this process, such as VRN2. The protein encoded by VRN2 shares homology with the Polycomb-group proteins such as FIS2 and EMF2 of Arabidopsis and Su(z)12 of Drosophila and may help maintain transcriptional repression of FLC.
Light, specifically day length and light quality, may also control flowering time. Day length determines whether a particular variety of a plant species will flower in a particular season of the year. Alterations of the light/dark photoperiod cycle can trigger changes in flowering time. In Arabidopsis, flowering is elicited by long days. Phytochromes A through E and cryptochromes (CRY) 1 and 2 perceive light in Arabidopsis. The duration of day and night is perceived by the circadian clock, or oscillator, which controls many circadian processes in addition to flowering. This is observed in a number of mutants such as lhy, cca1, gi, elf3, toc1, ztland fkf1 that are affected in various circadian processes as well as daylength-dependent flowering. At the molecular level, the link between the central oscillator and flowering time may be the transcription factor CO(CONSTANS). Null co mutants flower late in inductive long days but behave like wild type under short-day conditions. Conversely, over-expression of CO results in early flowering independent of day length. These observations suggest that CO may play a significant role in an output pathway that integrates day length perception and timekeeping mechanisms to promote flowering.
Light quality also affects flowering time, and may serve as a signal that dictates reproductive strategy in crowded conditions. Light that has been reflected from neighboring plants has a reduced red/far-red ratio due to chlorophyll absorbance of red light. Enriched far-red light is thus a signal that other plants are nearby, and exposure to enriched far-red light can trigger flowering under crowded conditions.
Mutants affected in autonomous pathway genes (fca, fy, fpa, ld, fld and fve) flower late in both long-day and short-day conditions, but late flowering can be overcome by vernalization or exposure to far red-enriched light. These genes function to maintain low levels of FLC expression throughout development. The autonomous pathway, vernalization pathway and FRI repression pathway thus all regulate the levels of FLC transcript. The FRI repression pathway is antagonistic to the autonomous and vernalization pathways, and the authors suggest that the interplay between these pathways may have evolved to regulate the FLC levels with precision in response to environmental conditions. It has been suggested that an endogenous input signal to the autonomous pathway regulates age-dependant flowering, requiring plants to reach an adult vegetative stage before shifting to reproductive development. FLC does not, however, appear to repress flowering in an age-dependent manner.
The hormone giberellic acid (GA) promotes flowering. Flowering is promoted in mutants with constitutively active GA signaling and is delayed in GA-sensing or GA biosynthesis mutants. The GA pathway appears to be distinct from other pathways controlling flowering.
The authors illustrate a model depicting a mechanism in which different input signals such as photoperiod, gibberellin, light quality and floral repressor signals affect the floral pathway integrator (FPI) genes FT, AGL20and LFY. The FPI genes in turn up-regulate floral meristem identity genes. The FPI genes are only partially redundant, and are critical controlling factors in flowering time control. Possible interplay between FPI gene activities is still under investigation, and other FPI genes remain to be identified.
The authors illustrate how interplay between flowering time regulation pathways may have been altered during the course of evolution to give rise to two different reproductive strategies characteristic of different ecotypes: winter annual growth habit and rapid cycling growth habit. Winter annuals do not flower until the onset of long spring days because of their requirement for vernalization. Vernalization-responsive late flowering winter ecotypes have a high level of FLC transcripts and protein. The protein encoded by FRI acts in conjunction with VRN2 and autonomous gene products to elevate the levels of the FLC gene product during the winter. The block is removed during the long days of the spring, inducing flowering. On the other hand, rapid cycling plants, which can flower within a single season, have a low level of FLC transcripts and proteins. Several ecotypes with this growth habit have a loss-of-function mutation in FRI, and the autonomous pathway, photoperiod pathway and GA pathway predominate regulation of FLC expression in these lines.
The authors speculate that changes in flowering time pathway predominance may explain regulation of reproductive development in species other than Arabidopsis. A number of components of the various pathways have been found in other plant species, such as rice, which has very different requirements for flowering. In rice, flowering is induced by short-day conditions. The authors suggest that the difference in flowering habit between the two species may be due to different photoperiod pathway output in rice than in Arabidopsis. Not all physiological differences in flowering time are likely to be due to differences in pathway predominance between species, though. In cereal species, the requirement for vernalization is recessive, suggesting that targets other than FLC may be critical for vernalization responses in these species.
The authors conclude that much work remains in flowering time research, particularly to gain a better understanding of endogenous signaling pathways and the genetic and epigenetic control of the flowering process.
Click here for the abstract.