OBJETIVOS GENERALES

The core oscillator is the central part of circadian clock which is responsible to generate endogenous rhythms, entrain the rhythms to external signals (via input pathways), and produce entrained rhythms to control wide range of biological processes (via output pathway). The first consistent model described in Arabidopsis for the core oscillator is a single transcriptional-translational negative feedback loop comprise of TIMING OF CAB EXPRESSION 1 (TOC1) and partially redundant genes CIRCADIAN CLOCK-ASSOCIATED 1 (CCA1) and ELONGATED HYPOCOTYL (LHY) in which they reciprocally regulate each other (Alabadi et al., 2001). This observation was based on the opposite expression phase of TOC1 and CCA1/LHY around dusk and dawn respectively. The model proposed that in the late evening, TOC1 activates the transcription of CCA1/LHY which results in peaking of their mRNA and protein level around dawn. The increased CCA1/LHY proteins bind directly to an „evening element‟ (EE, AAAATATCT) in the promoter of TOC1 and

represses the transcription of TOC1. This in turn leads to decreasing of CCA1/LHY protein level as TOC1 acts as a positive activator of CCA1/LHY. Evidence supporting this model includes the fact that the expression of TOC1 suppressed in constitutively expressed CCA1 or LHY plants (Alabadi et al., 2002) whereas lhy cca1 double

mutants have increased TOC1 mRNA level (Mizoguchi et al., 2002). Reciprocally, the loss of function of TOC1 in toc1 mutant reduces the level of CCA1 and LHY and shorter period length in transcription rhythms (Más et al., 2003b).

However, this single loop model was inadequate to account for some aspect of the core oscillator such as toc1-2 but not toc1-1 reduces LHY/CCA1 expression

(Alabadi et al., 2001), cca1lhy double mutant is not completely arrhythmic

(Mizoguchi et al., 2002), and overexpression mutants of TOC1 reduce CCA1/LHY expression (Más et al., 2003b). Additionally, there are more genes discovered to affect rhythmicity of clock genes and are not yet fitted into the single loop model (Locke et al., 2005).

In order to explain these experimental data, two recent mathematical

modelling studies (Locke et al., 2006; Zeilinger et al., 2006) have extended the „single loop model‟ into „three-loop model‟ (Figure 1.3). The three-loop model consists of the

existing TOC1-CCA1/LHY loop as a core loop and two loops occur in the morning and evening respectively. In the core loop, a hypothetical component „X’ was also

proposed to act in between TOC1 and CCA1/LHY for CCA1/LHY induction as no evidence has shown direct binding of TOC1 to CCA1/LHY. A recent identified TCP (TB1, CYC, PCFs) transcription factor, CHE (CCA1 HIKING EXPEDITION) was identified as a component for this core loop (Pruneda-Paz et al., 2009). CHE

physically interacts with TOC1 and recruits TOC1 to TCP binding site in the promoter of CCA1 and activates CCA1 transcription. CCA1 then negatively feeds back on CHE expression.

Chapter 1 Genetic control of flowering time

PRR9 PRR7

LHY/

CCA1 CHE _GI/LUX?Y

GI ZTL PRR3 X TOC1 ELF4

Figure 1.2 Core oscillator in Arabidopsis consists of the proposed three-interlocking loops. The three negative feedback loops are represented by three different colours. The white and gray background represent day and night respectively. In the central loop (red), LHY and CCA1 bind to TOC1 promoter in the morning to repress TOC1 expression. LHY and CCA1 expression decreased throughout the day which leads to increased TOC1 expression. Elevated TOC1 level activates LHY and CCA1 expression via an unknown factor „X‟. Another recently identified component in the central loop is CHE. CHE physically interacts with TOC1 and activate CCA1 expression. The morning loop (green) includes activation of PRR9/PRR7 expression by LHY/CCA1 while PRR9/PRR7 negatively feedback on LHY/CCA1 expression. In the evening loop (purple), a hypothetical element „Y‟ (whose activity can be explained by GI or LUX) will promote TOC1 expression in the evening whereas TOC1 will negatively feedback on „Y‟ expression. Additionally, ZTL was suggested to inhibit TOC1 level by protein degradation. This proteasome-mediated degradation is suspected to be stabilised by GI and disrupted by PRR3. Besides this three interlocking loop, ELF4 is hypothesised to inhibit the morning and the evening loops. (Modified from Montaigu et. al. 2010).

In this new model, a morning loop comprised of PSEUDO RESPONSE REGULATOR 7 (PRR7) and PSEUDO RESPONSE REGULATOR 9 (PRR9) is activated by CCA1/LHY and PRR7/PRR9 then negatively feed back to regulate CCA1/LHY expression. PRR7/PRR9 are part of the Pseudo Response Regulator (PRR) gene family which includes TOC1 (PRR1), and all members of the family contain two characteristic domains of C-terminal Pseudo-Receiver (PR) domain and N-terminal Constans, Constans-like, TOC1 (CCT) domain(Matsushika et al., 2000). The

members of this family show sequentially expression peaks every 2 hours after dawn in the order of PRR9, PRR7, PRR5, and PRR3 and until TOC1 is eventually expressed in the evening (Matsushika et al., 2000). Experimental data which supports this

morning loop includes the CCA1 binding directly to CBS (CCA1-binding site) in PRR7 and PRR9 promoters and a later phase shift in CCA1 and LHY expression in prr7prr9 double mutant (Farré et al., 2005). Moreover, PRR7 and PRR9 were found to be essential for temperature sensitivity of circadian clock (Salome and McClung, 2005b). Therefore, PRR7 and PRR9 could be important components that respond to light and temperature input signals.

Separately, a hypothetical component „Y‟ was proposed to induce TOC1

expression in an evening loop and ZEITLUPE (ZTL) was suggested to repress TOC1 expression. The expression of Y is in turn negatively regulated by TOC1. ZTL encodes a blue light photoreceptor which involved in the input pathways. ZTL interacts with TOC1 in vivo and target TOC1 for proteasome-mediated degradation in the dark. This finding is consistent with the observation that ztl mutant has elevated TOC1 protein level (Más et al., 2003a). GIGANTEA (GI) (Fowler et al., 1999; Park et al., 1999) and LUX ARRHYTHMO (LUX) (Hazen et al., 2005; Onai and Ishiura, 2005) are possible candidates for Y component. GI encodes a nuclear protein with several

membrane-spanning domains. GI transcript is regulated by clock and peaks around late afternoon supporting a role for GI in the feedback loop as a positive regulator of TOC1 expression. Furthermore, decreased CCA1/LHY expression is observed in gi mutant that might due to the reduced TOC1 levels in the mutant (Fowler et al., 1999). GI was also found to stabilise ZTL in the blue light to ensure robust rhythm of TOC1 (Kim et al., 2007). Interestingly, PRR3 was found to disrupt TOC1 degradation by ZTL by binding to TOC1 (Para et al., 2007). Besides its role in circadian clock, GI was also found to function in phytochromeA-mediated photomorphogenesis (Oliverio et al., 2007), temperature compensation (Gould et al., 2006), and also interact with SPY in the GA signalling pathway (Tseng et al., 2004). LUX encodes for a MYB

Chapter 1 Genetic control of flowering time

transcription factor which expressed at the same time during the day as TOC1 and mutation in LUX cause arrhythmia in both TOC1 and CCA1/LHY expression rhythms (Onai and Ishiura, 2005).

In addition to this „three-loop model‟, EARLY FLOWERING 4 (ELF4) was

integrated into this model via recent structural and functional analysis (Kolmos et al., 2009). ELF4 encodes a novel protein with no known domains and its transcript is clock regulated with a peak at dusk (Doyle et al., 2002). Furthermore, elf4 mutant has reduced CCA1/LHY expression and elevated TOC1 expression (Kikis et al., 2005). Recently, ELF4 was found to repress PRR7/PRR9 expression in the morning loop and GI/LUX expression in the evening loop (Kolmos et al., 2009). This suggests ELF4 may be part of the core oscillator. Moreover, several genes known to affect

rhythmicity under constant conditions have not yet placed into this model such as FIONA1 (FIO1) (Kim et al., 2008), TEJ (Panda et al., 2002), and PRR5 (Ito et al., 2008).

Besides transcriptional feedback loop, core oscillator is also regulated by other mechanisms. Histone acetylation in TOC1 chromatin regulates the transcription activation or repression of TOC1 (Más, 2008). Post-translational phosphorylation of CCA1 by CK2 regulates stability and activities of CCA1 (Daniel et al., 2004). In addition to TOC1, several other components of core oscillator are also regulated by proteasome-dependent degradation such as PRR3 by WNK1 (Murakami et al., 2002). Therefore, it is highly likely that the „three-loop model‟ is not enough to explain the

operation of core oscillator and will need to be repeatedly revised after exploring relationship of different components using either experimental or modelling approaches.