Temperature Induced Flowering in Arabidopsis thaliana

Abstract

The formation of flowers, a tightly controlled process, is modulated by developmental as well as environmental cues. Among the environmental factors that influence flowering, our knowledge of the effects of growth temperature is relatively less. In the July issue of PLoS Genetics, we have shown that higher temperatures can induce flowering by passing the requirement for long photoperiods for floral transition in Arabidopsis thaliana. By exploiting natural variation in combination with mutant analysis and genome wide expression profiling, we have shown that this floral induction has a novel genetic basis and is possibly associated with genome wide alterations in splicing patterns of several transcripts including FLOWERING LOCUS M, which we have shown to be a major effect QTL for thermo-sensitivity. Our study nicely demonstrates the power of a combinatorial approach to understand the molecular genetic basis of complex traits. In this addendum, we propose a testable hypothesis for FLM function in regulation of temperature mediated floral transition as well as the function of FLM in flowering time regulation.

Introduction

When to flower is a crucial developmental decision in the life cycle of a plant that is directly associated with its reproductive success. Subsequently the timing of flowering is regulated both on by the developmental status of the plant as well as external cues.¹ The major environmental factors that modulate flowering include the quantity and quality of light, vernalization (exposure to winter conditions) and growth temperature. Arabidopsis is a facultative long day plant in which flowering is accelerated by long days (16 hrs light). The molecular genetic basis of the photoperiodic induction of flowering has been deciphered using mutants that are late flowering specifically in long days. The nuclear protein CONSTANS plays a major role in integrating the effects of long days.² The expression of CO as well as the activity of the CO protein itself is under light regulation.^3,4 Analysis of naturally occurring flowering time variation resulted in the identification of the vernalization pathway that promotes flowering in response to winter conditions.^5,6 Vernalization is a mechanism that plants appear to have evolved in order to avoid flowering in peak winter. There is extensive natural variation in the requirement as well as response to vernalization and most of this effect is mediated through epistatic interaction between two loci FRIGIDA and FLOWERING LOCUS C.^7–10 In contrast to our understanding of the effects of photoperiod and vernalization, our knowledge of the effects of growth temperature on flowering remains enigmatic, although it is generally known that growth temperature modulates flowering time.

Previous studies have shown that in long days there is a delay in flowering under lower temperatures (16°C LD) and the analysis of flowering time mutants revealed that cry2 photoreceptor mutants show an extensive delay in this response.¹¹ It is also known that the early flowering behavior of another photoreceptor mutant phyB is temperature sensitive and this early flowering effect was abolished at lower temperatures.¹² We have previously shown that there is extensive natural variation in thermal response of accessions in long days, although it was clear that specifically dissecting the thermal effects under these conditions would be difficult.⁹ Subsequently, we set out to find a way to analyze the interaction of flowering and temperature. Since previous studies have indicated that flowering is not induced by a simple heat shock,¹³ we tested plants by growing at different temperatures that will not induce a typical heat shock response. We realized that growing plants at 25°C or 27°C, results in early flowering compared to the normally used 23°C in short days.¹⁴ Armed with this new assay that allows us to specifically address the growth temperature effects on flowering, we addressed the molecular genetic basis for the same.

We took a combinatorial approach of using natural variation in combination with mutant analysis and genomics to decipher the molecular genetic basis of thermal induction of floral transition. A simple analysis of flowering time mutants at 23°C and 27°C in short days revealed that the thermal response in independent of CO, suppressed by FLC and integrated mainly through FT. In order to understand the genetic cascade that is involved in the thermal response, we took advantage of the enormous natural variation and performed QTL mapping with parental strains that differ in their thermal sensitivity. Interestingly, the thermal sensitivity mapped to FLOWERING LOCUS M, a gene that is a FLC paralogue. However, in contrast to the flc mutants, which are more sensitive to thermal cues, flm mutants display less sensitivity indicating temperature acts either through FLM or acts in the same genetic cascade in which FLM participates. Genome wide expression profiling showed temperature specific enrichment in factors associated with alternative splicing and consistent with this, the splicing patterns of several flowering time regulators including FLM are altered during thermally induced flowering.

Further mechanistic work is required to decipher the mechanism through which temperature acts and how FLM might be connected in this response. We would like to propose a simple testable model that could account for the results reported in this paper (Fig 1). According to our model, FLM rather than FLC is the major repressor that suppresses flowering in short days. This is consistent with the fact that flm mutants are much earlier than flc mutants in short days.¹⁵ Normally FT is repressed in short days by the floral repressor FLM. When FLM levels are perturbed this leads to a quantitative alterations in flowering time.^15,16 When plants are grown at 27°C, this leads to a general reprogramming of the physiology, which includes changes in the splicing patterns of several transcripts including FLM. This altered splicing of FLM leads to reduction in the splice form that represses FT, subsequently leading to an increased FT expression that ultimately leads to early flowering. In addition, we would like to suggest that FT expression is regulated in both positive and negative manner such that under long days there is both de-repression as well as activation of FT. The factors that repress FT include FLC, FLM, SVP and MAF2.^15,17–21 Among these the main function of FLM is to prevent flowering in non-inductive conditions such as short days. When plants encounter long days, there is both de-repression as well as activation of FT. Under higher temperatures (27° short days), there is only the de-repression and this de-repression is sufficient to provide an early flowering response. The model we have proposed here is testable and should provide further insights into regulation of flowering time.

Figure 1.

A schematic representation of a model for temperature effects and FLM function in short days. This model proposes a transcriptional repression of FT by FLM in short days, which is relieved through changes in the splicing pattern of FLM at higher temperatures.

This work has strengthens the ground work by Blázquez et al and Halliday et al.^11,12 for the future analysis of thermal response in flowering and demonstrates the power of combining natural variation approach together with mutant analysis.

Addendum to: Potent Induction of Arabidopsis thaliana Flowering by Elevated Growth Temperature. Balasubramanian Sureshkumar, Sureshkumar Sridevi, Lempe Janne, Weigel Detlef. PloS Genetics. 2006;2:e106. doi: 10.1371/journal.pgen.0020106.