Photoperiodic flowering occurs under internal and external coincidence

Abstract

Determining the proper time to flower is important to ensure the reproductive success of plants. The model plant Arabidopsis is able to measure day-length and promotes flowering in long day (LD) conditions. One of the most prominent mechanisms in photoperiodic flowering is the clock-regulated gene expression of CONSTANS (CO) and the stabilization and activation of CO protein by light (regarded as external coincidence). We recently demonstrated that timing of the blue-light dependent formation of FLAVIN-BINDING, KELCH REPEAT, F-BOX 1 (FKF1) and GIGANTEA (GI) protein complex is crucial for regulating the timing of CO gene expression. The expression of FKF1 and GI is clock regulated, and their expression patterns have the same phase in LD (regarded as internal coincidence) but not in short day (SD) conditions, where floral induction is greatly delayed. Hence, timing of the FKF1-GI complex formation is regulated by the coincidence of both external and internal cues. Here, we propose a molecular mechanism for CO regulation by FKF1-GI complex formation.

Introduction

To maximize reproductive success, plants coordinate their transition from a vegetative to a reproductive state with the most favorable season. Photoperiodic flowering, which refers to a plant's ability to promote flowering by measuring the changes in day-length, is regulated by the internal circadian clock and external signals.^1,2 To explain the mechanisms underlying photoperiodism, these are two of the models that have been proposed: the external coincidence model⁴ and the internal coincidence model.⁵ In the external coincidence model, photoperiodism is controlled by circadian clock-regulated expression of a key component and the effect of light on the activity of this component. Only under inductive conditions are sufficient amounts of the key component exposed to light, thereby inducing a photoperiodic response. In the internal coincidence model, induction of a photoperiodic response occurs only when two (or more) regulators, which have differently entrained expression rhythms depending on day length, show the same phase.

Arabidopsis is a facultative long-day plant in which long days (LD) promote and short days (SD) delay flowering. The well-characterized molecular mechanism that mediates photoperiodic flowering is the regulation of FLOWERING LOCUS T (FT) gene expression by CO,^1,3,6,7 which can be explained by the external coincidence model. CO expression is regulated by the circadian clock, and CO protein stability and activity are regulated by light. FT gene expression is induced by CO largely under LD conditions, when the peak expression of CO coincides with external light signals, and CO protein is stabilized and activated by light.⁸ Several regulators that mediate clock-regulated CO gene expression have been identified. These regulators include GI, FKF1 and CYCLING DOF FACTOR 1 (CDF1).^9–11 Previous analyses showed that GI functions essentially in the induction of CO expression,^8,12 and FKF1 is predicted to mediate the peak expression of CO in the late afternoon.¹⁰ FKF1 degrades the CO transcriptional repressor CDF1 in the late afternoon.¹¹ However, the functional relationships among these factors in CO gene regulation were not well understood.

Regulatory Mechanism of CO Gene Expression by FKF1, GI and CDF1

We recently elucidated the molecular mechanism by which FKF1, GI and CDF1 regulate CO expression.¹³ First, we found that GI and FKF1 form a complex in a blue-light dependent manner in vivo. In this interaction, FKF1 functions as a direct blue-light receptor modulating the interaction with GI. Second, timing of the FKF1-GI complex formation is correlated with the CO expression pattern. Third, GI is required for FKF1 to degrade CDF1. Fourth, the chromatin immunoprecipitation (ChIP) experiments showed that all GI, FKF1 and CDF1 proteins are localized near each other on the CO chromatin region. CDF1 is a repressor of CO induction, whose expression peaks in the morning.¹¹ Therefore, we proposed the following molecular regulation of CO expression: In the morning, CDF1 localizes on the CO chromatin to repress CO expression. Later in the afternoon, when GI and FKF1 are expressed and are recruited to the CO chromatin, the GI-FKF1 complex associates with CDF1 in order to degrade it and in turn activates CO expression.

In addition to these functional relationships, the FKF1-GI complex seems to be the prominent regulator in the activation of CO gene expression because overexpression of FKF1 and GI severely disrupted the diurnal expression pattern of CO in both LD and SD conditions.¹³ However, we do not know if GI and FKF1 mediate CO gene expression directly. GI does not have any identified functional domains besides several putative transmembrane regions.⁹ FKF1 is predicted to play a role in protein degradation through the proteasome.^11,14 Hence, we anticipate that other factors might associate with the FKF1-GI complex in order to activate CO expression.

Two Layers of Coincidence Set Timing of the FKF1-GI Complex Formation

Our findings suggest that timing of the FKF1-GI complex formation determines the temporal information for CO expression.¹³ Here, two layers of mechanistic control seem to be important (Fig. 1). One is the external coincidence of FKF1 expression and light. Since FKF1 functions as a blue-light photoreceptor and modulates the interaction with GI, FKF1-GI complex formation is triggered by blue light.¹³ A larger quantity of the complex is formed under LD conditions (16 h of light and 8 h of darkness) than under SD conditions (8 h of light and 16 h of darkness), due to the coincidence between the clock-regulated expression of FKF1 and light in LD.¹³ This likely leads to a sufficient amount of CO induction during the light period in LD but not in SD. The second mechanism is the internal coincidence of FKF1 and GI expression. Expression patterns of FKF1 and GI are highly synchronized with maximum exposure of FKF1 protein under light in LD.¹³ These mechanisms ensure the largest amount of FKF1-GI complex formation in LD conditions. In SD conditions, GI expression peaks around 7 h after light onset, which is about 3 h prior to peak FKF1 expression. Therefore, GI and FKF1 expression patterns show less overlap.¹³ In addition, FKF1 expression occurs mainly in the dark, resulting in minimal formation of the FKF1-GI complex in SD. As we observed the complex in the beginning of the dark period under SD conditions, we presume that, once the complex is formed, it is stable for several hours without light. The difference between FKF1-GI complex formation in LD and in SD determines the timing of daytime CO gene expression in different photoperiods, which is the key step in discriminating day-length.

Figure 1.

FKF1-GI complex formation is regulated by both external and internal coincidence. (A) External coincidence mechanism. Coincidence of the clock-regulated FKF1 expression and light induces a larger amount of FKF1-GI complex formation, which triggers flowering promptly in LD conditions. (B) Internal coincidence mechanism. Clock-regulated expression of FKF1 and GI coincide in LD to maximize FKF1-GI complex formation. In SD, FKF1 and GI have different phases of expression with less overlap under light, thereby minimizing FKF1-GI complex formation.

Conclusions

We elucidated the molecular regulatory mechanism of CO gene expression by FKF1, GI and CDF1. In this mechanism, timing of the FKF1-GI complex formation is regulated by both external and internal coincidence. Considering the well-established external coincidence mechanism of FT regulation by CO, photoperiodic flowering seems to be regulated precisely by a combination of circadian clock and external light signal. To further understand the mechanisms of photoperiodic flowering, we need to analyze how the circadian clock regulates expression of FKF1, GI and CDF1, and how the FKF1-GI complex mediates CO activation in addition to degrading CDF1.

Addendum to: Sawa M, Nusinow DA, Kay SA, Imaizumi T. FKF1 and GIGANTEA complex formation is required for day-length measurement in Arabidopsis. Science. 2007;318:261–265. doi: 10.1126/science.1146994.