Plant and Cell Physiology Advance Access originally published online on May 31, 2007
Plant and Cell Physiology 2007 48(7):925-937; doi:10.1093/pcp/pcm067
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Rapid Paper |
Genetic Linkages of the Circadian Clock-Associated Genes, TOC1, CCA1 and LHY, in the Photoperiodic Control of Flowering Time in Arabidopsis thaliana
1Laboratory of Molecular Microbiology, School of Agriculture, Nagoya University, Chikusa-ku, Nagoya, 464-8601 Japan
2Division of Biological Science, Graduate School of Science, Nagoya University, Chikusa-ku, Nagoya, 464-8602 Japan
3Institute of Biological Sciences, University of Tsukuba, Tsukuba, Ibaraki, 305-8572 Japan
*Corresponding author: E-mail, i052001d{at}mbox.nagoya-u.ac.jp; Fax, +81-52-789-4091.
 |
Abstract |
|---|
 Top Abstract IntroductionResultsDiscussionMaterials and MethodsAcknowledgmentsReferences |
|---|
In Arabidopsis thaliana, the flowering time is regulated through the circadian clock that measures day-length and modulates the photoperiodic CO–FT output pathway in accordance with the external coincidence model. Nevertheless, the genetic linkages between the major clock-associated TOC1, CCA1 and LHY genes and the canonical CO–FT flowering pathway are less clear. By employing a set of mutants including an extremely early flowering toc1 cca1 lhy triple mutant, here we showed that CCA1 and LHY act redundantly as negative regulators of the photoperiodic flowering pathway. The partly redundant CCA1/LHY functions are largely, but not absolutely, dependent on the upstream TOC1 gene that serves as an activator. The results of examination with reference to the expression profiles of CO and FT in the mutants indicated that this clock circuitry is indeed linked to the CO–FT output pathway, if not exclusively. For this linkage, the phase control of certain flowering-associated genes, GI, CDF1 and FKF1, appears to be crucial. Furthermore, the genetic linkage between TOC1 and CCA1/LHY is compatible with the negative and positive feedback loop, which is currently believed to be a core of the circadian clock. The results of this study suggested that the circadian clock might open an exit for a photoperiodic output pathway during the daytime. In the context of the current clock model, these results will be discussed in connection with the previous finding that the same clock might open an exit for the early photomorphogenic output pathway during the night-time.
Keywords: Arabidopsis thaliana - Circadian rhythm - Clock - CO-FT - Photoperiodic flowering
Abbreviations: CCA1, CIRCADIAN CLOCK-ASSOCIATED 1; CO, CONSTANS; FT, flowering locus T; LD, long day; LHY, LATE ELONGATED HYPOCOTYL; PRR, PSEUDO-RESPONSE REGULATOR; SD, short day; TOC1, TIMING OF CAB EXPRESSION 1.
|
Introduction |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsAcknowledgmentsReferences |
|---|
In the model higher plant Arabidopsis thaliana, many circadian clock-associated genes have been identified through intensive genetic studies (for recent reviews, see Gardner et al. 2006
, McClung 2006, and references therein). Among those, the best candidates of clock components are CCA1 (CIRCADIAN CLOCK-ASSOCIATED 1) and its partially redundant homolog LHY (LATE ELONGATED HYPOCOTYL) (Schaffer et al. 1998, Wang and Tobin 1998, Mizoguchi et al. 2002). Five members of a small family of PSEUDO-RESPONSE REGULATOR (PRR9, PRR7, PRR5, PRR3 and PRR1) genes are also believed to be another type of clock component (Makino et al. 2000, Matsushika et al. 2000, Eriksson et al. 2003, Farre et al. 2005, Nakamichi et al. 2005a, Nakamichi et al. 2005b, Salome and McClung 2005). The first discovered representative of the PRR genes is TOC1 (TIMING OF CAB EXPRESSION 1), which is identical to PRR1 (Makino et al. 2000, Strayer et al. 2000). It was originally proposed that these two types of clock-associated genes (CCA1/LHY and TOC1) form a negative and positive transcriptional feedback loop that generates fundamental circadian rhythms (Alabadi et al. 2001, Alabadi et al. 2002). Other clock-associated components were also identified, for instance ELF4 (EARLY FLOWERIGN 4), GI (GIGANTEA), LUX/PCL1 (LUX ARRHYTHMO/PHYTOCLOCK 1) and ZTL (ZEITLUPE) (Fowler et al. 1999, Doyle et al. 2002, Somers et al. 2004, Hazen et al. 2005, Kikis et al. 2005, Onai and Ishiura 2005). These clock-associated components must coordinately play their distinct and complementary roles to generate stable, robust and sustainable rhythms. Taking these putative clock components into consideration, several versions of interlocking multiloop clock models were then envisaged to explain their complex roles for the clock function (for recent reviews, see Mizuno 2005, Gardner et al. 2006, McClung 2006, and references therein). Based on a mathematical simulation, two independent groups have recently built a consistent multiloop clock model (Locke et al. 2006, Zeilinger et al. 2006).
As summarized above, we have learnt much about the molecular mechanism underlying the plant circadian clock. Such circadian rhythms in higher plants are very relevant to a wide range of biological processes, including movement of organs such as leaves and petals, stomatal opening, regulation of light responses during early photomorphogenesis, and also photoperiodic control of flowering time (for reviews, see McClung 2000, Imaizumi and Kay 2006, Nozue and Maloof 2006). Nevertheless, little is known about how the plant circadian clock plays roles in a wide variety of downstream clock-controlled biological processes. In this context, we have previously addressed the issue as to the genetic linkage between the circadian-associated CCA1/LHY and TOC1 genes and the phyB-dependent light signal transduction pathway that is involved in early photomorphogenesis (Ito et al. 2007). In this study, we address in particular another issue as to the close linkage between the circadian-associated CCA1/LHY and TOC1 genes and the output pathway of photoperiodic flowering time.
In higher plants, to ensure reproductive success, flowering is controlled so as to occur in the optimal environmental conditions for seed production. To this end, plants have evolved several distinct mechanisms that properly control flowering time. In Arabidopsis thaliana, the results of extensive genetic studies revealed the existence of at least four signaling pathways that coordinately promote flowering, namely photoperiodic, autonomous, vernalization and gibberellin pathways (for reviews, see Corbesier and Coupland 2006, Imaizumi and Kay 2006, and references therein). These pathways all converge on the flowering-integrator genes, namely FT (FLOWERING LOCUS) and/or SOC1 (SUPPRESSOR OF OVEREXPRESSION OF CO) (Kobayashi et al. 1999, Samach et al. 2000). In this respect, A. thaliana is classified as a facultative long-day (LD) plant, which is induced to flower by exposure to appropriately longer day-length. The circadian clock can measure the day-length, and determines the time to flower (Yanovsky and Kay 2002, Yanovsky and Kay 2003). During the last decade, we have learnt much about the molecular mechanism underlying the photoperiodic control of flowering time in A. thaliana.
The key player in the photoperiodic induction of flowering is the flowering-time gene CO (CONSTANS) (Putterill et al. 1995). This gene encodes a nuclear-localized zinc finger-containing protein, which was recently suggested to serve as a component of a DNA-binding transcription factor (generally called the HAP complex) (Wenkel et al. 2006). The expression of CO is under the control of the circadian clock so as to show a biphasic diurnal expression profile with peaks both at late daytime and night-time under LD conditions (Suarez-Lopez et al. 2001). In LD conditions, the CO protein is stabilized in the external light conditions, so that the CO protein can actively promote the transcription of FT in leaf phloem (Valverde et al 2004). The FT gene product is believed to move to and act in the shoot apical meristem to induce the floral meristem (Abe et al. 2005, Huang et al. 2005, Wigge et al. 2005). Recently, this critical event on FT was indeed demonstrated at the molecular level (Corbesier et al. 2007, Tamaki et al. 2007). Under short-day (SD) conditions, however, the daytime peak of CO tends to disappear (Suarez-Lopez et al. 2001) and, consequently, no transcription of FT occurs. As a result, the coincidently activated CO–FT pathway depending on the LD photoperiodicity provides predominantly a signal to promote flowering.
To link the circadian clock and the downstream CO–FT pathway, two more flowering time genes have been uncovered, namely GI and CDF1 (CYCLING DOF FACTOR 1) (Fowler et al. 1999, Imaizumi et al. 2005). The GI gene encodes a large nuclear protein which acts as an essential clock-associated component, whereas the clock-controlled CDF1 gene encodes a Dof-domain-containing DNA-binding protein. The GI gene product acts as the transcriptional activator for CO (most probably indirectly), whereas the CDF1 gene product serves as a transcriptional repressor through directly binding to the CO promoter (for a review, see Imaizumi and Kay 2006). In this connection, two more flowering time genes must also be taken into consideration, namely RFI2 (RED AND FAR-RED INSENSITIVE 2) and FKF1 (FLAVIN-BINDING, KELCH REPEAT, F-BOX1) (Imaizumi et al. 2003, Chen and Ni 2006). These two genes are also under the control of the circadian clock, and they are somehow involved in the regulation of the CO–FT pathway. In particular, the FKF1 protein was assumed to be a blue light receptor, and was proposed to promote the degradation of the CDF1 protein by functioning as a component of E3 ligase (Imaizumi et al. 2005). Nevertheless, these clock-controlled and/or light-regulated components (GI, FKF1, CDF1 and RFI2) are believed to play an important role in the photoperiodic flowering pathway at positions upstream of the CO–FT pathway.
A remaining major question is how the clock-associated components (CCA1/LHY and TOC1) modulate the downstream CO–FT pathway to determine the flowering time correctly. In this respect, a close link between the CCA1/LHY, GI, CO and FT genes has been examined previously (Mizoguchi et al. 2005). A cca1/lhy double loss-of-function mutant shows a phenotype of extremely early flowering in SD, whereas a gi loss-of-function mutant displays a phenotype of extremely late flowering in LD. The genetic results supported the view that this gi mutant is epistatic to the cca1/lhy double mutant, albeit not completely. It was thus proposed that CCA1/LHY modulated the CO–FT pathway by negatively regulating the transcription of GI, albeit not exclusively (Mizoguchi et al. 2005). On the other hand, it was reported that an early flowering toc1 mutant (i.e. toc1-1) also exhibits an altered nature in the coincident expression profile of CO in SD (Yanovsky and Kay 2003).
The question then arose as to how these clock-associated TOC1 and CCA1/LHY genes are coordinately linked to the CO–FT pathway. Here, this issue will be addressed by employing a set of single and double mutants, also including an extremely early flowering toc1 cca1 lhy triple mutant. Based on the results of this study, a genetic model will be proposed to explain the genetic linkage between the central clock circuitry and the downstream CO–FT photoperiodic flowering pathway.
|
Results |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsAcknowledgmentsReferences |
|---|
A set of mutants of clock-associated genes
The Arabidopsis mutant alleles used in this study are listed in Table 1. Both the cca1-1 and lhy11 alleles represent loss-of-function mutants (Mizoguchi et al. 2005
), whereas the toc1-2 allele appears to be hypomorphic (Mas et al. 2003). We established a set of homozygous mutants encompassing these defective clock-associated genes in every possible combination, including a toc1-2 cca1-1 lhy11 triple mutant (Ito et al. 2007, and see Fig. 1). This triple mutant will be designated hereafter as 
tcl for clarity of this text. In addition, we employed three double mutants (tc, tl and cl) and individual single mutants (t, c and l), together with the parental wild type (Col). These mutant seedlings germinated well and grew normally on MS agar plates (as well as on soil, see also Fig. 4). As reported (Ito et al. 2007), these mutants displayed a hallmark phenotype regarding the sensitivity to light during early photomorphogenesis. This characteristic was particularly evident when grown in red light. However, the phenotype was evident even when these mutants were grown in white light (10 h light/14 h dark cycle), as judged by their hypocotyl lengths (Fig. 1, upper panel). The results showed that cl is hypersensitive to the light conditions, whereas t is hyposensitive. In this connection, tc, tl and tcl showed the phenotype of long hypocotyls, which was quite similar to that of t. These results indicated that the t mutation is epistatic to both the c and l mutations, suggesting that CCA1/LHY act in a negative manner upstream of TOC1. Taken together, the linear genetic linkage between TOC1 and CCA1/LHY was reasonably deduced to explain the roles of these genes in early photomorphogenesis (Fig. 1; Ito et al. 2007). By employing these mutants, here we focus our attention on another hallmark and common phenotype of these mutants, namely the phenotype as to the photoperiodic control of flowering times.
|
|
|
The set of mutants of clock-associated genes display altered expression profiles of circadian-controlled genes
Before addressing the main issue of this study, it may be of interest (or necessary) to characterize briefly these mutants with regard to the clock function per se. This was done through examining these mutant plants with reference to free-running oscillation profiles of certain clock-controlled genes, including CCA1 and LHY themselves (Fig. 2). The hallmark clock-controlled CAB2 and CCR2 output genes were also examined (Fig. 3).
|
|
When these clock-controlled genes were examined for the plants grown in continuous light,
t, c and l each showed a common phenotype of short period, as has been documented previously (Fig. 2; also see Somers et al. 1998, Mizoguchi et al. 2005, Locke et al. 2006). In this study, we examined for the first time the phenotypes of tc and tl with reference to the circadian rhythm. The results showed that these double mutants exhibited a phenotype of even shorter period compared with those of each single mutant. The free-running rhythms in these double mutants were not robust and were low amplitude (or rapidly dampened after a few cycles) (Fig. 2). These results are consistent with the current view that TOC1 and CCA1/LHY are coordinately involved in the circadian clock function.
This view was further strengthened by the results for tcl (Fig. 3). In this experiment, the tcl triple mutant plants were examined for the rhythmic profiles of CAB2 and CCR2 under the growth conditions of a light/dark cycle, followed by release into continuous light. Both cl and tcl showed a phenotype of extremely short period in continuous light (or seemingly arrhythmic under our experimental conditions of Northern hybridization analyses). Under the light/dark conditions, the phases of expression profiles of both CAB2 and CCR2 were markedly advanced. In particular, the morning gene (CAB2, Fig. 3A) showed its peak at the end of the dark cycle (or night) in both cl and tcl, while the evening gene (CCR2, Fig. 3B) showed a robust peak in the early morning. Such anomalous phasing in the mutants was commonly observed for other circadian-controlled genes tested (such as GI, FKF1 and PRR5) (data not shown, and see Fig. 6 later). These results for cl were in a good agreement with those reported previously (Mizoguchi et al. 2005, Locke et al. 2006). In this study, we showed consistent results for tcl. It was rather difficult to see the difference between cl and tcl under our experimental conditions. Hence, this point remains to be examined by adopting finer analytical methods. Nevertheless, the results from tc, tl and tcl (Figs. 2, 3) strongly supported the current idea that TOC1 and CCA1/LHY play roles very close to the central oscillator. In particular, the tcl triple mutant is severely defective in the circadian clock function under both the light/dark cycle and continuous light conditions. However, the clock function was not completely perturbed in the plants carrying the toc1-2, cca1-1 and lhy11 triple lesions (note that toc1-2 is hypomorphic, perhaps not completely null). Nevertheless, these mutants with the unified Col background will be useful to clarify the molecular mechanism underlying the circadian clock in A. thaliana (this particular issue will be addressed further elsewhere).
|
Genetic linkages between TOC1 and CCA1/LHY in the mechanism underlying the photoperiodic control of flowering time
We then addressed the main issue of this study, i.e. to clarify possible genetic linkages between TOC1 and CCA1/LHY in the mechanism underlying the photoperiodic control of flowering time. By employing the set of mutants, the flowering phenotypes were examined in both LD and SD by scoring the number of leaves upon the onset of bolting. Each double mutant flowered much earlier than did the wild type in SD (Fig. 4A), although all of the mutants developed the visible primary inflorescence in LD as early as the wild type (data not shown). A quantitative examination (Fig. 4B, upper panel) showed that (i) each single mutant exhibited an early flowering phenotype in SD, as reported previously (Mas et al. 2003
, Mizoguchi et al. 2005); and, more importantly, (ii) both the tc and tl double mutants flowered significantly earlier than did each single mutant. In other words, the toc1 allele acts, to some extent, in a manner additive to both the cca1 and lhy alleles. The cl mutant showed a phenotype of extremely early flowering in a manner almost independent of the photoperiodicity (LD or SD). Notably, the tcl triple mutant showed essentially the same degree of early flowering phenotype as did cl. In this sense, the cl mutation is epistatic to t. Taken together, it was suggested that CCA1/LHY act redundantly as negative regulators of the photoperiodic flowering pathway, and the partly redundant CCA1/LHY functions appeared to be largely, but not absolutely, dependent on the upstream TOC1 gene that serves as an activator. To confirm these results, another independent experiment was carried out in SD, with reproducible results (Fig. 4B, lower panel). These results will provide us with new insight into the genetic linkage between TOC1 and CCA1/LHY in the mechanism underlying the photoperiodic control of flowering time, as will be discussed later (see Fig. 10).
|
TOC1 and CCA1/LHY coordinately modulate the CO–FT photoperiodic flowering pathway
Several lines of evidence from the present study and previous studies suggested that the clock-associated TOC1 and CCA1/LHY genes serve as modulators upstream of the canonical CO–FT photoperiodic flowering pathway (Yanovsky and Kay 2002
, Mizoguchi et al. 2005). Hence, we directly characterized the expression profiles of CO and FT in the mutants. The wild type and mutant (t, c and tc) plants were grown in SD, and RNA samples were prepared at intervals (3 h). The levels of transcript were measured by means of real-time PCR with primers specific for either FT or CO. To obtain firm results, the experiments were repeated, with highly reproducible results (Fig. 5, left hand side, upper and middle panels, Exp.-1 and Exp.-2). In good agreement with the early flowering phenotype of these mutants, the FT gene was highly expressed even in SD in these mutants, while the wild type expressed a very low level of FT under the same growth conditions tested. To extend these analyses, another set of mutants (t, l and tl) were also examined, with similar results (Fig. 5, left hand side, lower panel).
|
Next, the diurnal expression profile of CO was examined in these mutants (Fig. 5, right hand side, upper and middle panels, Exp.-1 and Exp.-2). In the wild type, robust CO expression was observed only during night-time in SD, as documented previously (Imaizumi and Kay 2006
). However, a significant level of CO mRNA was detected during late daytime, evident in tc. A subtle but significant level of CO was also reproducibly detected in daytime in t. Again, the same experiments were conducted by employing another set of mutants (t, l and tl), giving similar characteristic results with regard to the expression profiles of CO (Fig. 5, right hand side, lower panel). These results were consistent with the external coincident model, in which a high level of CO expression during daytime in a certain photoperiodic condition coincidentally results in an enhanced expression of FT, the protein product of which in turn acts as a crucial promoter for flowering (see Introduction). The results of this study supported the view that the clock-associated TOC1 and CCA1/LHY genes coordinately control the flowering time through the canonical CO–FT photoperiodic pathway.
Molecular linkages of the clock-associated functions of TOC1 and CCA1/LHY with the putative downstream flowering-associated genes, GI, CDF1 and FKF1
According to the current molecular view with regard to the photoperiodic control of flowering (see Introduction), it was assumed that the clock-controlled GI, CDF1 and FKF1 act as mediators between the TOC1 and CCA1/LHY clock function and the CO–FT flowering pathway (for a review, see Imaizumi and Kay 2006). In this connection, Mizoguchi et al. (2005) have already addressed this issue with a cl double mutant. Here, we further employed the tc double mutant to address the issue in terms of expression of GI, CDF1 and FKF1. Both the wild-type and tc plants were grown in SD, and RNA samples were prepared at intervals, as mentioned above (see Fig. 5). The levels of transcript were measured by means of real-time PCR with primers specific for GI, CDF1 and FKF1 (Fig. 6). As reported previously (Imaizumi and Kay 2006), these genes were expressed with a diurnal rhythm, with a robust peak in the morning (CDF1) or evening (GI and FKF1) in the wild type. In all instances in tc, however, the peaks appeared with a considerably advanced (or precocious) timing (about 3 h). This is consistent with the previous results shown in Figs. 2 and 3.
To extend this intriguing view, we further examined the diurnal expression profiles of GI and FKF1 in the cl double and tcl triple mutants grown in the light/dark cycle. In these experiments, the transcripts were detected at 2 h intervals by Northern blot hybridization analyses in order to obtain more direct data than those obtained by the real-time PCR method (Fig. 7). Consistent with the fact that both GI and FKF1 are typical evening genes, the profiles of both GI and FKF1 in the wild type showed a robust peak in the evening. In cl and tcl, however, the peaks of GI and FKF1 appeared in a greatly advanced position (i.e. in the early morning). In this respect, the cl double mutation appears to be epistatic to the t mutation, suggesting that CCA1/LHY function downstream of TOC1. These characteristic events, observed for the clock-controlled and flowering-associated GI, CDF1 and FKF1 genes, might be relevant to the mechanism by which the circadian clock is linked to the downstream CO–FT flowering pathway, as further addressed later.
|
Molecular linkages of the clock-associated functions of TOC1 and CCA1/LHY with the CO–FT photoperiodic flowering pathway
To confirm the above intriguing results more directly, we then examined the expression profiles of GI, CDF1 and FKF1 in the
cl double and tcl triple mutant, which were grown in SD, under which conditions both the mutants indeed showed an extremely early flowering phenotype (see Fig. 4). In good agreement with the results of Fig. 7, the expression phases of the evening GI and FKF1 genes were markedly advanced, and the peaks appeared in the morning in both cl and tcl (Fig. 8, upper and lower panels). Similarly, the expression phase of the morning CDF1 gene appeared at midnight in cl and tcl (Fig. 8, middle panel). It should be emphasized that such alerted profiles of GI, CDF1 and FKF1 were indistinguishable between cl and tcl, suggesting again that the cl lesions are epstatic to the t lesion at least in this respect.
|
Keeping the above intriguing observations as to the flowering-associated genes in mind, and also by using the same RNA samples prepared in SD, we finally examined the expression profiles of CO and FT in both
cl and tcl (Fig. 9). As expected, no expression of CO in the daytime was seen in Col in SD. In sharp contrast, a marked level of CO expression in the daytime was clearly seen for both cl and tcl, even in SD. Consistently, a large amount of FT was detected in the early flowering mutants grown in SD, while the expression of FT in Col was marginal (Fig. 9). Taken together, it was suggested that the CCA1/LHY–TOC1 circadian clock is closely linked to the downstream CO–FT flowering pathway, particularly through the flowering-associated GI, CDF1 and FKF1 components, as discussed below (see Fig. 10).
|
|
Discussion |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsAcknowledgmentsReferences |
|---|
In this study, we characterized a set of clock-associated mutants, including a toc1 cca1 lhy triple mutant (Fig. 1), with special reference to the photoperiodic control of flowering time. These mutants are defective in the circadian clock function per se (Figs. 2, 3). However, it may be worth mentioning that the
tcl triple mutant has the ability to show rhythmic expression of certain clock-controlled genes in light/dark cycle conditions. In any event, based on the genetic results of this study, the linkage between the clock-associated genes TOC1 and CCA1/LHY and the canonical CO–FT photoperiodic flowering pathway was experimentally deduced. The results were best explained by deducing the linear genetic linkage, as schematically proposed in Fig. 10A. According to this model, CCA1 and LHY act redundantly as negative regulators of the photoperiodic flowering pathway. In this respect, the cl double mutations are epistatic to the t single mutation (Fig. 4), albeit not completely, suggesting that the TOC1 gene acts positively upstream of the partly redundant CCA1/LHY genes. The results of the examination of the expression profiles of CO and FT in the mutants suggested that this clock circuitry is linked to the CO–FT output flowering pathway (Fig. 5). We do not know the precise molecular mechanism by which the clock circuitry is linked to the CO–FT flowering pathway. However, the results suggested that the circadian-controlled GI, FKF1 and CDF1 might serve as such mediators (Fig. 6). At noon in SD, the tc mutation would result in an anomalous up-regulation of GI and FKF1, both of which are known to serve as promoters for flowering (i.e. CO expression), and also the mutation would result in a down-regulation of CDF1, which is known to act as a repressor for CO expression (for these genes, see Imaizumi and Kay 2006). These events were much more striking in the cases of cl and tcl. It was shown that, in both cl and tcl, the diurnal expression profiles of GI and FKF1 were markedly changed in such a way that the evening peaks of GI and FKF1 disappeared, and the precocious and robust peaks appeared in the early morning (Fig. 7). These critical events were confirmed in the plants grown in SD (Fig. 8). These events regarding GI, CDF1 and FKF1 would consistently result in a precocious up-regulation of CO expression during daytime even in SD, as explained above. This view coincides well with the observed tcl phenotype of early flowering in SD (this view was also integrated into the model of Fig. 10A). Indeed, we finally demonstrated that the expression profile of CO was altered in both the cl and tcl mutants, with a markedly high expression level of CO in the late daytime (Fig. 9). Consequently, a large amount of FT was detected in the cl and tcl mutants grown in SD. These striking events are consistent with the extremely early flowering phenotypes of these cl and tcl mutants. It should also be emphasized that the cl double mutations are epistatic to the t single mutation in terms of the phenotypes regarding expression profiles of GI and FKF1 (Figs. 7–9
). This is consistent with the proposed TOC1–CCA1/LHY linear linkage (Fig. 10A). Taken together, we believe that the proposed view is highly reasonable.
Nevertheless, verification of the proposed model must await further extensive examinations. Furthermore, the proposed linear linkage model cannot explain some significant details of the results (Fig. 4). For instance, the results showed that the partly redundant CCA1/LHY functions are largely, but not absolutely, dependent on the upstream TOC1 gene that serves as an activator, and the results also showed that the t allele acts in a manner additive to both the c and l alleles, at least to some extent. These details are best explained by assuming that the CCA1/LHY functions are positively regulated by not only TOC1, but also an additional unknown gene(s) (X), as incorporated into the model (Fig. 8A). According to our current knowledge (see Introduction), there are already very good candidates for X. This presumptive activator might be ELF4 or LUX (see Introduction). Similarly to a toc1 mutant, certain loss-of-function mutations in these clock-associated genes result in an early flowering phenotype in SD (Doyle et al. 2002, Hazen et al. 2005). In the current multiloop clock model, furthermore, these genes (ELF4 and LUX) were proposed to function as positive regulators upstream of CCA1/TOC1 in a parallel (or coordinate) manner with TOC1 (for a review, see Gardner et al. 2006, and see also Fig. 8B). To address the relevant issues, further genetic analyses remain to be conducted by employing appropriate elf4 and lux alleles.
With regard to the proposed model, the second set of significant details should be considered, as follows. The results of Figs. 5 and 9 clearly indicated that the TOC1–CCA1/LHY clock regulates the expression of FT through modulating the levels of CO mRNA in the daytime, as discussed above from a qualitative viewpoint. From a strictly quantitative viewpoint, however, the levels of CO mRNA were not perfectly related to the resulting expression levels of FT (see Figs. 5 and 9). This suggests that the TOC1–CCA1/LHY clock might also regulate the expression of FT through an alternative pathway involving an as yet unidentified factor (Y) (see Fig. 10A). This putative pathway could regulate the stability of CO protein (not CO mRNA), or it could directly regulate the transcription of FT in a manner independent of CO.Whatever the case, this additional pathway is consistent with a similar model proposed previously by Mizoguchi et al. (2005). Clarification of the putative pathway must also await further examination, which should include identification of Y. In this context, it is worth mentioning that Y could be GI (Mizoguchi et al. 2005). To address the relevant issues, further epistasis analyses should be conducted by employing an appropriate gi allele.
In our previous study, we proposed the genetic linkage between the clock-associated genes and the red light signal transduction that especially regulates early photomorphogenesis (Ito et al. 2007), as also shown in Fig. 10A (lower part). In this case, it was shown that CCA1/LHY function negatively at a position upstream of TOC1, which in turn positively regulates the light response during early photomorphogenesis (e.g. inhibition of hypocotyl elongation) (see Fig. 1). At a glance, these two linear linkages were not consistent with each other. Furthermore, t and cl display the same phenotypes of early flowering in the flowering pathway (see Fig. 4), while they display the opposite phenotypes to each other (hyposensitive vs. hypersensitive) in the light response pathway (see Fig. 1). However, this problem appears to be superficial because these two linear linkages can be easily unified by loop formation. More importantly, the puzzle is simply solved when these two pathways are superimposed onto the canonical negative and positive feedback loop of CCA1/LHY and TOC1, which is a fundamental genetic linkage proposed for the core clock function per se, as schematically indicated in Fig. 10B. This novel view suggested that the circadian-controlled hallmark output pathways intimately and consistently overlap with the central clock functions per se, which are exerted by the major clock-associated TOC1 and CCA1/LHY components.
The finding of this study provided us with new insight into the linkages between the central clock and the two apparently disparate output pathways; ‘photoperiodic control of flowering time’ and ‘regulation of light signal transduction’. In accordance with this view (Fig. 10B), the clock mechanism can be connected to the downstream output pathways. In particular, we propose that a clock-controlled exit is opened (or gated) for the output pathway of photoperiodic flowering during the daytime (or perhaps the onset of the dark/light transition). In other words, a clock signal might be integrated through the daytime exit into the downstream CO–FT photoperiodic flowering pathway. In this respect, it was previously suggested that the evening GI gene appears to be a connector between CCA1/LHY and CO. (Mizoguchi et al. 2005, see also Fig. 5). Another clock-controlled exit is opened for the light signal transduction during the night-time (or perhaps the onset of the light/dark transition). More specifically, it is tempting to speculate that a clock signal is connected by an unknown mechanism through this night-time exit into the phyB-dependent light signaling pathway that especially regulates the length of hypocotyls, as discussed in the previous study (Ito et al. 2007).
In A. thaliana, it has long been known that not only toc1, cca1 and lhy, but also certain other circadian clock-associated mutants commonly display hallmark phenotypes with regard to three characteristic biological events: (i) altered rhythmic expressions of circadian-controlled genes; (ii) changes in photoperiodic flowering time; and (iii) altered sensitivities toward red light in elongation of hypocotyls. With regard to the photoperiodic control of flowering time, for instance, an elf4 mutant shows a phenotype of early flowering in SD (Doyle et al. 1999), and a lux mutant also shows a phenotype of early flowering in SD (Hazen et al. 2004). In particular, we recently reported that a prr7 prr5 double mutant shows a phenotype of extremely late flowering in LD in a manner almost independent of the photoperiodicity (Nakamichi et al. 2005a). By characterizing a prr7 prr5 cca1 lhy quadruple mutant, it was further suggested that PRR7 and PRR5 coordinately and positively act on the CO–FT photoperiodic flowering pathway in a manner antagonistic to CCA/LHY (Nakamichi et al. 2007). The molecular bases underlying such complicated clock-associated signaling circuitries have long been the subjects of debate. Indeed, when this study had been completed, Ding et al. (2007) published a paper in which a consistent view with regard to this particular issue was discussed by employing the toc1-21, cca1-11 and lhy-21 alleles in the WS background (note that these are different alleles and background from ours, see Table 1). In this regard, the intriguing view proposed in this and their studies, regarding the linkages between the clock function and the two hallmark output pathways, will provide us with new clues for a better understanding of the molecular mechanisms by which a variety of biological processes are controlled in a manner dependent on the circadian clock in A. thaliana.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsAcknowledgmentsReferences |
|---|
Plant materials and growth conditions
Arabidopsis thaliana accession Columbia (Col) plants were used as the wild type. The cca1-1/lhy-11 (
cl ) mutant (Green and Tobin 1999, Mizoguchi et al. 2002) was introgressed four times into Col from Landsberg erecta (Ler). Three lines homozygous for cl were selected and all showed a similar phenotype (early flowering in SD and short hypocotyls) to that of cca1/lhy mutants (Mizoguchi et al. 2002). One of these lines, named cl (Col background) line B, was used in this work. The toc1-2 (t) mutant was also introgressed six times into Col from C24. Several t homozygous lines in Col were selected. It may be worth mentioning that t mutants in the Col and C24 background both show essentially the same phenotypes (short period in continuous light, early flowering in SD and long hypocotyls under red light). Light intensity was 90–100 and 120–150 µmol m–2 s–1 for LD and SD, respectively.
Construction of multiple mutants
To obtain multiple variants in every possible combination of toc1/cca1/lhy in the unified Col background, the t and cl mutants were crossed (Ito et al. 2007). Multiple variants of toc1/cca1/lhy were isolated from the F3 and F4 generations. We isolated mutants independently for at least two lines and observed the same phenotype (flowering and photomorphogenesis under red light) in each genetic background; c, l, tc, tl and tcl were isolated once, but were confirmed by both genomic PCR for the presence of the T-DNA and by Northern blotting for gene expression.
Preparation of RNA, Northern blotting analysis and real-time PCR
For Northern blotting analysis, seedlings were grown on MS plates containing 1% sucrose under the conditions of a light/dark cycle (12 h light/12 h dark) for 20 d. They were then released into LL (continuous light) and harvested (2 or 3 h intervals) to quantify the mRNA of circadian clock-associated genes and clock-controlled genes. Total RNA was purified by ATA methods, as described previously (Makino et al. 2000). For real-time PCR, seedlings were grown on MS plates containing 1% sucrose under the conditions of LD (16 h light/8 h dark) or SD (10 h light/14 h dark) for 10 d. They were harvested (3 or 3.5 h intervals) to quantify the mRNA of genes involved in flowering regulation. Total RNA was purified by an RNeasy kit (Qiagen, Valencia, CA, USA). To synthesize cDNA, 1 µg of each RNA was reverse transcribed with ReverTea Ace (TOYOBO, Osaka, Japan) and oligo(dT) primer. The synthesized cDNAs were amplified with iQ SYBR Green supermix (Bio-Rad Laboratories, Hercules, CA, USA) and a primer set using a MiniOpticon real-time PCR system (Bio-Rad Laboratories). The primer sets used in this study were, for CO, 5'-CTACAACGACAATGGTTCCATTAAC-3' and 5'-CAGGGTCAGGTTGTTGC-3'; for GI, 5'-GGGTAAATATGCTGCTGGAGA-3' and 5'-CAGTATGACACCAGCTCCATT-3'; for CDF1, 5'-TTTCCCGACGGTTTTAGAGG-3' and 5'-CCATGCTGTTGCATCTTGGA-3'; and for FKF1, 5'-GAAGTCTTCACTGGCTATCG-3' and 5'-GATCAACCAATGGGTGACG-3'. Primer sets for APX3 and FT were described by Hazen et al. (2005) and Endo et al. (2005), respectively. The following standard thermal cycling program was used for all PCRs: 95°C for 120 s, 40 cycles of 95°C for 10 s, and 60°C for 60 s. Data were analyzed using Opticon Monitor version 3.1 software (Bio-Rad Laboratories).
Flowering time measurement
Seeds were imbibed directly on soil. Deionized water was supplied daily. With one exception, water supplemented with Hyponex (Hyponex-Japan, Osaka, Japan) diluted 1/1,000 was used on the first occasion. Plants were grown in soil at 22°C under SD (10 h light/14 h dark) or LD (16 h light/8 h dark). Time to flowering was expressed as the number of rosette leaves at the time of a 1 cm high flower bolt. Light intensity was 90–100 and 120–150 µmol m–2 s–1 for LD and SD, respectively.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsAcknowledgmentsReferences |
|---|
Thanks are due to The Salk Institute Genomic Analysis Laboratory (California, USA) for mutant seeds. This study was supported by Grants-in-Aid for scientific research (to T.M.) from the Ministry of Education, Sports, Culture, Science, and Technology of Japan. N.N. and S.I. were supported by the Japan Society for the Promotion of Science Research Fellowship for Young Scientists.
|
Footnotes |
|---|
4These authors contributed equally to this work.
|
References |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsAcknowledgmentsReferences |
|---|
Abe M, Kobayashi Y, Yamamoto S, Daimon Y, Yamaguchi A, Ikeda Y, Ichinoki H, Notaguchi M, Goto K, Araki T. FD, a bZIP protein mediating signals from the floral pathway integrator FT at the shoot apex. Science (2005) 309:1052–1056.
Alabadi D, Oyama T, Yanovsky MJ, Harmon FG, Mas P, Kay SA. Reciprocal regulation between TOC1 and LHY/CCA1 within the Arabidopsis circadian clock. Science (2001) 293:880–883.
Alabadi D, Yanovsky MJ, Mas P, Harmer SL, Kay SA. Critical role for CCA1 and LHY in maintaining circadian rhythmicity in Arabidopsis. Curr. Biol. (2002) 12:757–761.[CrossRef][ISI][Medline]
Chen M, Ni M. RFI2, a RING-domain zinc finger protein, negatively regulates CONSTANS expression and photoperiodic flowering. Plant J. (2006) 46:823–833.[CrossRef][ISI][Medline]
Corbesier L, Coupland G. The quest of florigen: a review of recent progress. J. Exp. Bot. (2007) 57:3395–3403.[CrossRef][ISI]
Corbesier L, Vincent C, Jang S, Fornara F, Fan Q, Searle I, Giakountis A, Farrona S, Gissot L, Turnbull C, Coupland G. FT protein movement contributes to long-distance signaling in floral induction of Arabidopsis. Science (2007) 316:1030–1033.
Ding Z, Doyle MR, Amasino RM, Davis SJ. A complex genetic interaction between Arabidopsis thaliana TOC1 and CCA1/LHY in driving the circadian clock and in output regulation. Genetics (2007) (online PMID17483414).
Doyle MR, Davis SJ, Bastow RM, McWatters HG, Kozma-Bognar L, Nagy F, Millar AJ, Amasino RM. The ELF4 gene controls circadian rhythms and flowering time in Arabidopsis thaliana. Nature (2002) 419:74–77.[CrossRef][Medline]
Endo M, Nakamura S, Araki T, Mochizuki N, Nagatani A. Phytochrome B in the mesophyll delays flowering by suppressing FLOWERING LOCUS T expression in Arabidopsis vascular bandles. Plant Cell (2005) 17:1941–1952.
Eriksson ME, Hanano S, Southern M, Hall A, Millar AJ. Response regulator homologues have complementary, light-dependent functions in the Arabidopsis circadian clock. Planta (2003) 218:159–162.[CrossRef][ISI][Medline]
Farre ME, Harmer SL, Harmon FG, Yanovsky MJ, Kay SA. Overlapping and distinct roles of PRR7 and PRR9 in the Arabidopsis circadian clock. Curr. Biol. (2005) 15:47–54.[CrossRef][ISI][Medline]
Fowler S, Lee K, Onouchi H, Samach A, Richardson K, Morris B, Coupland G, Putterill J. GIGANTEA: a circadian clock-controlled gene that regulates photoperiodic flowering in Arabidopsis and encodes a protein with several possible membrane-spanning domains. EMBO J. (1999) 18:4679–4688.[CrossRef][ISI][Medline]
Gardner MJ, Hubbard KE, Hotta CT, Dodd AN, Webb AA. How plants tell the time. Biochem. J. (2006) 397:15–24.[CrossRef][ISI][Medline]
Green RM, Tobin EM. Loss of the circadian clock-associated protein 1 in Arabidopsis results in altered clock-regulated gene expression. Proc. Natl Acad. Sci. USA (1999) 96:4176–4179.
Hazen SP, Schultz TF, Pruneda-Paz JL, Borevitz JO, Ecker JR, Kay SA. LUX ARRHYTHMO encodes a Myb domain protein essential for circadian rhythms. Proc. Natl Acad. Sci. USA (2005) 102:10387–10392.
Huang T, Bohlenius H, Eriksson S, Parcy F, Nilsson O. The mRNA of the Arabidopsis gene FT moves from leaf to shoot apex and induces flowering. Science (2005) 309:1694–1696.
Imaizumi T, Kay SA. Photoperiodic control of flowering: not only by coincidence. Trends Plant Sci. (2006) 11:550–558.[CrossRef][ISI][Medline]
Imaizumi T, Schultz TF, Harmon FG, Ho LA, Kay SA. FKF1 F-box protein mediates cyclic degradation of a repressor of CONSTANS in Arabidopsis. Science (2005) 309:293–297.
Imaizumi T, Tran HG, Swartz TE, Briggs WR, Kay SA. FKF1 is essential for photoperiodic-specific light signalling in Arabidopsis. Nature (2003) 426:302–306.[CrossRef][Medline]
Ito S, Nakamichi N, Nakamura Y, Niwa Y, Kato T, Murakami, M. Kita M, Mizoguchi T, Niinuma K, Yamashino T, Mizuno T. Geneti