PerspectivePlant Biology

Flower Arranging in Arabidopsis

See allHide authors and affiliations

Science  02 Jun 2000:
Vol. 288, Issue 5471, pp. 1600-1602
DOI: 10.1126/science.288.5471.1600

One of the key factors contributing to the success of flowering plants is their ability to regulate the timing of flowering so that they can take maximum advantage of the most environmentally favorable conditions. Unraveling the signal transduction pathways that regulate flowering is one of the most exciting areas of plant biology research. Now, Samach et al. (1) on page 1613 of this issue and Blázquez and Weigel (2) in a recent issue of Nature reveal the integrated network of molecular signals that induce flowering in the weed Arabidopsis thaliana under different environmental conditions.

Many plants show a strong seasonality in their flowering, which ensures that their seeds are produced when conditions are optimal for germination and growth. Plants are able to accurately measure day length by integrating signals from photoreceptors and an endogenous circadian clock. In long-day species such as winter wheat (Triticum aestivum), flowering is induced when the duration of light exceeds a certain critical length. In contrast, short-day plants such as soybean (Glycine max) measure the dark period and flower when the night exceeds a certain critical length (3). Many plants also flower in response to cold temperatures (vernalization), mimicking the transition from winter to spring when conditions for seed germination are more favorable (4). Flowering can also be triggered by environmental stresses such as shading by neighboring vegetation. In this case, plant species can best succeed by channeling all of their available energy into reproduction (5).

Arabidopsis is a facultative long-day plant, that is, it flowers vigorously during long days (it also flowers in response to cold). These flowering responses are regulated by the photoperiodic and autonomous signaling pathways (see the figure). However, in the absence of promoting signals, Arabidopsis eventually flowers through a day length-independent pathway that follows an age-dependent developmental program (6). The study of Arabidopsis flowering mutants reveals much about the signaling pathways that regulate each of the flowering responses (6). Embryonic flower mutants have demonstrated that flowering is, in fact, the default state and that the autonomous pathway normally suppresses the transition from the vegetative to the flowering state. Variations in flowering between different Arabidopsis populations have led to the identification of the FLOWERING LOCUS C gene (FLC), a crucial component of the autonomous pathway. The vernalization response in Arabidopsis suppresses the FLC gene, thereby negating the autonomous suppression of flowering (see the figure). Conversely, the photoperiodic pathway and the day length-independent (age-dependent) pathway appear to act independently of the autonomous pathway to promote the transition to flowering.

Flower power.

Integration of the signaling pathways that promote flowering in Arabidopsis. In response to long days, the photoperiodic pathway (blue) induces expression of the CONSTANS gene (CO). The transcription factor encoded by CO induces expression of four target genes. These are AtP5CS2, which promotes stem elongation mediated by proline; ACS10, which is thought to promote flowering mediated by ethylene; and FT and SOC1, two genes previously shown to promote flowering in response to long days. The FCA gene is required for full induction of SOC1 and FT by CO. It forms part of the autonomous pathway (yellow) regulating the floral suppressor FLC. Both the photoperiodic and autonomous pathways converge on the promoters of the FT and SOC1 genes (green). The floral meristem identity gene, LFY, is indirectly regulated by CO and by the phytohormone, gibberellin, which is a component of the day length-independent pathway (red). Through distinct cis elements, these two pathways converge on and interact with the LFY promoter (magenta).

A number of day length-insensitive mutants, including constans (co) and flowering locus t (ft), result in the suppression of flowering normally induced by long days. The co and ft mutants fail to show promotion of flowering during long days, resulting in the assignment of CO and FT to the photoperiodic pathway (see the figure). Finally, mutants such as ga1 (deficient in the phytohormone gibberellin) demonstrate that gibberellins (in the absence of other promoting signals) are required for flowering that is independent of day length (see the figure).

Samach and colleagues investigated the integration of signals from the photoperiodic (day length-dependent) and autonomous pathways. They examined the action of CO in the photoperiodic pathway and observed increases in CO messenger RNA in response to long days (6). The CO gene encodes a GATA-type transcription factor (7), suggesting that an increase in CO protein promotes the switching on of other genes in the flowering pathway. To analyze CO-activated genes, the investigators engineered plants to express a fusion protein composed of CO and the ligand-binding domain of the glucocorticoid receptor. The glucocorticoid receptor fusion protein resides in plant cell cytoplasm and is rendered inactive by chaperone proteins. Treatment with the steroid dexamethasone causes the chaperones to release the fusion protein, which then moves to the nucleus and activates the transcription of genes downstream of CO in the flowering pathway. By combining the fusion protein approach with the application of the protein synthesis inhibitor cycloheximide, Samach and co-workers were able to identify four target genes that are switched on by CO. They are FT and SUPPRESSOR OF OVEREXPRESSION OF CO 1 (SOC1)—already known to be necessary for CO to promote flowering (6)—and AtP5CS2 and ACS10, which are involved in shoot elongation and ethylene biosynthesis, respectively (see the figure).

Within the autonomous pathway, FCA suppresses the expression of the FLC gene. Consequently, fca mutants are constitutively late flowering; that is, flowering is delayed relative to wild-type plants under all conditions (6). By analyzing the effect of the fca mutation on the ability of CO to induce gene expression, the authors were able to deduce that transcription of FT and SOC1 is also under the control of the autonomous pathway. They demonstrated that FCA is required for full induction of FT and SOC1 by CO. These findings highlight the integration of signals from the autonomous and photoperiodic pathways that together directly control expression of FT and SOC1. The balance between regulation by the autonomous and photoperiodic pathways differs for SOC1 and FT. Consistent with this, soc1 mutants flower late under the influence of both long and short days, whereas ft mutants flower late in response to long days only. Vernalization also suppresses FLC, and so, according to this model, SOC1 and FT are common components of pathways regulating flowering in response to different environmental signals.

In a complementary study, Blázquez and Weigel (2) investigate the integration of flowering signals from the photoperiodic and day length-independent pathways. The induction of flowering ultimately leads to the expression of meristem identity genes, consistent with the transition of the meristem from a vegetative state (when it only produces leaves) to a floral state (8). (The apical meristem is the active growing point of the plant producing cells destined to become leaves, stem, and flowers.) The authors generated a reporter construct by fusing the promoter of one meristem identity gene, LEAFY (LFY), to β-glucuronidase (a bacterial gene that produces a colored product when incubated with a substrate). They were able to demonstrate convergence of the photoperiodic and day length-independent pathways (see the figure). The expression of LFY is induced, albeit indirectly, by CO through the photoperiodic pathway (1, 2) and by gibberellins through the day length-independent pathway (2). Until now, it has been unclear whether these pathways converged at LFY or at a point upstream of LFY.

The authors found that gibberellins activated the LFY promoter through cis-acting elements that differed from those inducing the day length flowering response. They generated deletion mutations in the LFY gene promoter of the reporter construct and engineered plants to express the mutant constructs. Disruption of a region of the LFY promoter containing a consensus sequence for binding of MYB transcription factors resulted in abrogation of the gibberellin response, which is essential for promotion of flowering in Arabidopsis during noninductive short days.

The investigators observed normal flowering in response to long days with the mutant constructs. When CO was overexpressed, constitutively activating the photoperiodic pathway, reporter gene expression in the mutant construct was also constitutively activated. However, when FT was overexpressed, activating the photoperiodic pathway, reporter gene expression was silent, indicating that LFY acts downstream of CO but not of FT. Therefore, LFY must be in a signaling pathway that is parallel to that containing FT (see the figure). Similarly, when the same disrupted promoter region was used to drive expression of the LFY gene, the construct was able to rescue flowering in plants deficient in LFY during long days but not short days. This confirms that the disruption constitutes a short-day defect. The LFY promoter thus defines another site where different environmental stimuli that control flowering converge and are integrated.

The integration of diverse stimuli in the control of a response is a feature of many biological systems. The two new studies provide inroads into understanding how several signaling pathways converge to control flowering in higher plants, but there is still much to learn. We do not know the mechanism by which different signaling components integrate the large number of environmental stimuli that regulate flowering. A further question is the way in which these various environmental stimuli are detected. It is still a mystery how signals from photoreceptors and the endogenous circadian clock interact to enable plants to detect day length (9) or how cold temperatures are sensed, triggering the vernalization response (10).


View Abstract

Stay Connected to Science

Navigate This Article