Plants control their flowering time in order to ensure that they reproduce under favourable conditions. The components involved in this complex process have been identified using a molecular genetic approach in Arabidopsis and classified into genetically separable pathways. The autonomous pathway controls the level of mRNA encoding a floral repressor, FLC, and comprises three RNA-binding proteins, FCA, FPA and FLK. FCA interacts with the 3′-end RNA-processing factor FY to autoregulate its own expression post-transcriptionally and to control FLC. Other components of the autonomous pathway, FVE and FLD, regulate FLC epigenetically. This combination of epigenetic and post-transcriptional control gives precision to the control of FLC expression and flowering time.
- alternative polyadenylation
- floral transition
- gene expression
Flowering time control
The floral transition is complex, involving the integration of responses to environmental cues with an endogenous programme of development. These responses are mediated by genetically separable pathways, which ensure that flowering occurs at a time that will maximize reproductive success. In Arabidopsis, the promotive functions of photoperiod and hormone-responsive pathways are antagonized by the transcription factor FLC, which is a potent repressor of flowering. FLC mRNA expression is regulated by different flowering time pathways, the most important in rapid cycling Arabidopsis accessions being the autonomous pathway  (Figure 1).
An important feature of Arabidopsis flowering is the quantitative nature of the responses involved. The activities of genetically separable promotive and repressive pathways are integrated at common targets known as floral pathway integrators . FLC acts as a ‘rheostat’, determining the capacity of the integrators to respond to promotive cues. Hence, components of the autonomous pathway control the level at which this ‘rheostat’ acts.
FCA and FPA are plant-specific RNA-binding proteins that promote flowering through the autonomous pathway
Despite the large number of Arabidopsis genes coding for novel RRM (RNA recognition motif)-containing proteins , mutations have been identified in only two of them, FPA and FCA, in the course of forward genetic screenings [3,4]. Both mutants flower late, indicating that the function of the wild-type alleles is to promote flowering. Both FCA and FPA function in the autonomous pathway to control FLC mRNA accumulation and flowering time.
Post-transcriptional regulation of FCA expression affects flowering time
FCA pre-mRNA is alternatively processed, with transcripts arising as the result of alternative splicing and alternative polyadenylation . FCA intron 3 is a particularly important site of regulation, as transcripts can be detected in which this intron is retained (transcript α) or excised (transcript γ) or in which premature cleavage and polyadenylation have taken place at a promoter-proximal poly(A)+ (polyadenylated) site (transcript β). The consequence of alternative processing is to alter the transcript coding length. Thus truncated and inactive FCA proteins are produced from transcripts α and β, which neither promote nor inhibit flowering time. Only the fully spliced γ transcript is active in controlling flowering time .
The alternative processing of FCA pre-mRNA is regulated and the formation of these multiple transcripts is explained by autoregulation . At a certain level of expression, FCA ultimately promotes premature polyadenylation within intron 3 of its own pre-mRNA, producing β transcript at the expense of α and γ. This negative feedback is under developmental control since it limits the temporal and spatial patterns of elevated levels of active FCA expression [5,6]. In turn, this has a functional consequence for flowering time control, since the expression of FCA from transgenes lacking cis-elements required for negative feedback results in precocious Arabidopsis flowering .
FCA interacts with FY
FCA physically interacts with another component of the autonomous pathway, FY, through the WW domain (protein–protein interaction domain containing two conserved tryptophan residues) of FCA. FY encodes a protein that is highly conserved in eukaryotes and resembles the Saccharomyces cerevisiae 3′-end RNA-processing factor, Pfs2p . Consistent with a role for FY in 3′-end RNA processing, fy mutants exhibit defects in 3′-end formation. Specifically, fy mutants exhibit less promoter-proximal 3′-end formation and, correspondingly, more promoter-distal 3′-end formation in FCA pre-mRNA .
FCA has at least two functions: first, it promotes flowering and, secondly, it negatively regulates its own expression. Both these functions require an intact WW domain. Consistent with the fact that FY is the protein partner for this domain, FY is genetically required for FCA to carry out both these functions . Since the negative feedback function of FCA ultimately involves 3′-end formation and FY resembles a 3′-end RNA-processing factor, it is possible that FCA and FY perform this autoregulation directly. FCA does not interact with FY to inhibit the activity of the 3′-end RNA-processing complex, but instead, to promote actively 3′-end formation at the promoter-proximal site within FCA pre-mRNA . This is consistent with the idea that FCA acts as an auxiliary factor for the 3′-end-processing complex by specifying the site of 3′-end formation. However, further experimentation is needed to determine the directness of the RNA interactions involved and the interplay between 3′-end formation and splicing.
Since FCA requires FY for both its autoregulation and flowering roles, it is probable that these functions are executed in a mechanistically similar manner. Shifts in the relative abundance of FLC transcripts polyadenylated at different sites between fca-1 and wild-type plants (in the same manner as observed for FCA α, β and γ transcripts) have not been detected. This may be because alternatively polyadenylated FLC transcripts are not stable or because FLC is not a direct target of FCA.
FCA and FPA regulate FLC differently: an emerging paradigm of plant gene expression control?
fca/fpa double mutants flower later than single fca or fpa mutants, indicating that these RNA-binding proteins perform non-overlapping roles in controlling flowering time . Another recently identified component of the autonomous pathway, FLK, encodes a KH-type RNA-binding domain protein, reinforcing the importance of RNA-mediated regulation within this pathway . However, not all autonomous pathway components appear to interact with RNA. The first identified component of the autonomous pathway was LD, a homeodomain-containing protein . FVE and FLD resemble proteins associated with chromatin remodelling [12,13]. Consistent with this, both the mutants fve and fld display changes in the acetylation of histones associated with the FLC locus, whereas other autonomous pathway mutants do not.
None of the autonomous pathway components appear to regulate each other. The emerging picture seems to be one of distinct regulatory mechanisms recruited to control FLC expression. This combination of post-transcriptional and epigenetic regulation may deliver precision to the control of the FLC ‘rheostat’. There are parallels with the manner in which FLC and the floral homoeotic gene AGAMOUS are regulated, which indicates that this combination of epigenetic and post-transcriptional control may not be unique to the autonomous pathway . Forward genetic screenings based on flowering time and floral organ phenotyping have uncovered a role for RNA processing in the regulation of FLC and AGAMOUS; the challenge for the future is to employ molecular approaches to determine how direct this regulation is and to elucidate the mechanisms involved.
Post-Transcriptional Regulation of Plant Gene Expression: Focused Meeting held at the University of East Anglia, Norwich, U.K., 15–17 April 2004. Edited by A.J. Michael (Institute of Food Research, Norwich) and John W.S. Brown (Scottish Crop Research Institute, Dundee, U.K.). Sponsored by Ambion, Biotechnology and Biological Sciences Research Council, Daiwa Foundation, The Gatsby Charitable Foundation, The Institute of Food Research, The John Innes Centre, The Scottish Crop Research Institute and VWR International.
- © 2004 The Biochemical Society