A Time to Grow, a Time to Flower

Flowering is one of the most critical timing decisions for higher plants because it can determine the difference between reproductive success and failure. The transition from vegetative growth to flowering is controlled by a complex network of interconnected pathways influenced by internal developmental cues and external environmental signals that converge on AP1 and LFY, key genes that promote flowering, and their respective positive regulatory genes FT and SOC1. Two key flowering time pathways, the autonomous pathway (default developmental pathway) and vernalization (prolonged exposure to cold), act via effects on FLOWERING LOCUS C (FLC), a strong floral repressor that functions as a negative regulator of FT and SOC1.

Regulation of flowering time has been studied extensively in Arabidopsis thaliana, different accessions of which are classified as winter annuals, which require vernalization to promote flowering, or summer annuals, which are rapid flowering in the absence of vernalization. This trait is determined mainly by the activities of FLC and its positive regulator FRIGIDA (FRI). Most rapid flowering summer annual Arabidopsis accessions, such as Columbia (Col) and Landsberg erecta, are homozygous for a recessive nonfunctional fri allele and/or carry weak flc alleles, whereas winter annuals carry active alleles of both FRI and FLC. When functional FRI and FLC are present (and in the absence of other mutations that inhibit FLC), strong expression of FLC inhibits flowering, until and unless vernalization overrides FRI and suppresses FLC expression, thus allowing increased expression of SOC1, FT, and other genes that promote flowering (reviewed in Amasino, 2004). Summer annuals that lack functional FRI and/or show weak expression of FLC do not produce enough FLC protein to inhibit flowering and flower rapidly in the absence of vernalization. Genes in the autonomous pathway, including FLD, FVE, FY, FCA, FPA, LD, and FLK, act to repress FLC independently of FRI and vernalization. Autonomous pathway mutants in a summer annual parental background (i.e., fri null background) behave as winter annuals due to elevated FLC expression, similar to that of plants having functional FRI alleles (Amasino, 2004).

In this issue of The Plant Cell, ACTIN-RELATED PROTEIN6 (ARP6) is introduced as another regulator of FLC and flowering time in independent reports by Choi et al. (pages 2647–2660) and Deal et al. (pages 2633–2646). Both groups show that ARP6 plays a role in the repression of flowering, at least in part through upregulation of FLC transcription. It is shown that ARP6 controls FLC independently of other known pathways, adding another layer of complexity to this key point in the regulation of flowering time. In addition to its effect on flowering time, ARP6 also appears to play a more fundamental role in regulating plant growth. As shown by Deal et al., ARP6 functions in the growth and development of numerous plant organs by acting to promote cell proliferation and thus control organ size, particularly during rapid growth under long-day conditions. Therefore, ARP6 may represent a pivotal point in the decision between vegetative growth and flowering in Arabidopsis.

Arabidopsis ARP6 was initially described by McKinney et al. (2002) as one of eight ARP sequences in the Arabidopsis genome that were identified based on homology to actin AtACT2 and named ARP2 to ARP9, each corresponding to its most closely related homolog of the same name in yeast and other eukaryotes. The yeast Saccharomyces cerevisiae genome encodes 10 ARPs, named ARP1 to ARP10 based on similarity to actin, with ARP1 being the most similar (Poch and Winsor, 1997). Yeast ARP6 was identified as a component of the ATP-dependent chromatin remodeling complex SWR1 (Krogan et al., 2003; Mizuguchi et al., 2004). The high degree of similarity between AtARP6 and yeast ARP6 suggests that they may have similar functions.

Choi et al. isolated ARP6 by map-based cloning from the mutant suppressor of FRI3 (suf3), obtained from a screen for early flowering mutants in the line Col:FRI^SF2 (a line of Col that carries functional FRI alleles backcrossed from the winter annual accession San Feliu-2). The suf3 mutant flowered significantly earlier than Col:FRI^SF2 under both long- and short-day conditions with or without vernalization treatment. Flowering time characteristics of the suf3 mutant were similar to those of wild-type Col plants (which carry fri null alleles); that is, they behaved as summer annuals and flowered very rapidly (when plants had an average of 11 to 12 rosette leaves) under long-day conditions. By contrast, Col:FRI^SF2 flowered after production of 64 rosette leaves under long-day conditions, and vernalization reduced this number to 13. However, SUF3/ARP6 and FRI appear to act independently of each other, as suf3 fri plants (suf3 mutation introduced into wild-type Col) flowered earlier (with an average of only six rosette leaves) than plants with suf3 or fri alone. If SUF3 and FRI acted through the same pathway, is it expected that the phenotypic effect of one mutation would be epistatic to, or masked by, the other. The suf3:FRI^SF2 mutant plants were found to have a 30 to 60% reduction in the level of FLC transcript, suggesting that SUF3/ARP6 functions to promote FLC transcription. Interestingly, however, the suf3/arp6 mutation introduced into a line that ectopically overexpresses FLC from the cauliflower mosaic virus 35S promoter also led to significantly early flowering compared with 35S:FLC alone, suggesting that SUF3/ARP6 regulates other factors in addition to FLC in the control of flowering time. Choi et al. found that ARP6 is highly expressed in the shoot apex during both vegetative and reproductive stages of growth, and fusion proteins of ARP6 and green or yellow fluorescent protein are localized to the nuclear periphery, consistent with it playing a role in regulating gene expression.

Deal et al. used ARP6 as a starting point and investigated its function by examining gene expression patterns and protein localization of ARP6 fused to green fluorescent protein and analyzing the phenotypes of several arp6 mutants, including T-DNA insertional mutants in wild-type Col and a mutant they created by backcrossing one of the T-DNA mutations into the Col:FRI^SF2 line (i.e., similar to suf3 of Choi et al.). Their results are largely consistent with those of Choi et al., and they made additional observations regarding a potential role for ARP6 in the control of cell proliferation and general plant growth. Similar to Choi et al., they found that arp6 mutants flowered earlier than the wild type and that FLC expression was downregulated in arp6 mutants. They also compared arp6 mutants to an flc-3 null mutant and showed that the arp6 mutants flowered significantly earlier, offering further support to the notion that ARP6 regulates one or more factors in addition to FLC to control flowering time. They therefore examined the transcript levels of a family of five genes called MADS AFFECTING FLOWERING (MAF1 to MAF5), which have been shown to act independently of FLC to repress flowering (Ratcliffe et al., 2001, 2003). They found that transcript levels of MAF4 and MAF5 were reduced approximately twofold in arp6 mutants compared with wild-type plants, suggesting that this might account for at least part of the FLC-independent effect of arp6 mutations on flowering time.

Deal et al. also noted that ARP6 is expressed throughout the plant in all aboveground tissues and most strongly in vascular tissues in both shoot and root. In addition to the effect on flowering time, arp6 mutants showed abnormal leaf development, particularly under long-day conditions. Leaves of arp6 mutants were dramatically smaller than those of wild-type plants, which was found to be correlated with fewer total cells rather than cell size, suggesting that ARP6 functions to promote cell proliferation. Under short-day conditions, arp6 leaves were narrower than, but about the same length as, wild-type leaves and deeply serrated along the margins, an effect also observed by Choi et al. These results suggest that ARP6 has photoperiod-dependent effects on plant growth due to effects on cell proliferation and leaf patterning.

The functional characterization of ARP6 by these two groups opens up several new avenues of investigation into the control of flowering time and plant growth. First, results of both groups suggest that ARP6 affects flowering time by acting to maintain FLC expression and by an FLC-independent mechanism. Deal et al. suggest that FLC-independent effects may involve regulation of MAF4 and MAF5, as transcript levels of both of these FLC-independent floral repressors were decreased in arp6 mutants. Second, homology with yeast ARP6 suggests that AtARP6 may function as a component of an SWR1 chromatin remodeling complex. Yeast SWR1 acts to replace histone H2A with the variant form H2A.Z at specific chromosomal locations, which appears to play a role both in the maintenance of gene expression in regions of euchromatin (Meneghini et al., 2003) and the maintenance of gene silencing in heterochromatin (Dryhurst et al., 2004; Fan et al., 2004). The closest homolog of SWR1 in Arabidopsis is PHOTOPERIOD INDEPENDENT EARLY FLOWERING1 (PIE1), mutants of which, as the name suggests, have an early flowering phenotype similar to that of arp6 mutants (Noh and Amasino, 2003). Interestingly, characteristics of leaves and flowers also were similar between pie1 and arp6 mutants, providing further support for the hypothesis that these genes function in the same pathway or pathways in regulating growth and development.

Deal et al. further hypothesize that ARP6 may act in a common pathway with AINTEGUMENTA (ANT), an AP2-domain transcription factor that has been shown to function in the promotion of cell proliferation during organ development (Mizukami and Fischer, 2000). Another interesting observation is that both ANT and ARP6 may be partially dependent on auxin signaling. Hu et al. (2003) show that the auxin-inducible gene ARGOS functions upstream of ANT and regulates cell proliferation and organ growth through ANT during organogenesis. Both Deal et al. and Choi et al. noted that arp6 mutants showed a loss of apical dominance, indicative of an effect on auxin signaling.

The work of Deal et al. and Choi et al. uncovers a potentially important link between the control of growth and flowering time. In addition, this work should lead to further studies on chromatin remodeling and the presence and function of a putative SWR1 complex in Arabidopsis involving ARP6 and possibly PIE1.