This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 19/8/2516    most recent tpc.107.053033v1 Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (1) Citing Articles via Google Scholar Google Scholar Articles by Kiba, T. Articles by Chua, N.-H. Search for Related Content PubMed PubMed Citation Articles by Kiba, T. Articles by Chua, N.-H. Agricola Articles by Kiba, T. Articles by Chua, N.-H.

Targeted Degradation of PSEUDO-RESPONSE REGULATOR5 by an SCF^ZTL Complex Regulates Clock Function and Photomorphogenesis in Arabidopsis thaliana^[W]

^a Laboratory of Plant Molecular Biology, The Rockefeller University, New York, New York 10065-6399
^b Plant Science Center, RIKEN, Institute of Physical and Chemical Research, Tsurumi, Yokohama 230-0045, Japan

² Address correspondence to chua{at}mail.rockefeller.edu.

	ABSTRACT

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Circadian clocks comprise several regulatory feedback loops that control gene transcription. However, recent evidence has shown that posttranslational mechanisms are also required for clock function. In Arabidopsis thaliana, members of the PSEUDO-RESPONSE REGULATOR (PRR) family were proposed to be components of the central oscillator. Using a PRR5-specific antibody, we characterized changes in PRR5 protein levels in relation to its mRNA levels under various circadian conditions. Under long-day conditions, PRR5 mRNA levels are undetectable at dusk but PRR5 protein levels remain maximal. Upon dark transition, however, PRR5 levels decrease rapidly, indicating dark-induced, posttranslational regulation. We demonstrated that the Pseudo-Receiver (PR) domain of PRR5 interacts directly with the F box protein ZEITLUPE (ZTL) in vitro and in vivo. Analyses of mutants and transgenic plants revealed an inverse correlation between PRR5 and ZTL levels, which depends on the PR domain. These results indicate that PRR5 is negatively regulated by ZTL, which likely mediates its ubiquitination and degradation. Phenotypic analyses of prr5 ztl double mutants showed that PRR5 is required for ZTL functions. ZTL contains a Light-Oxygen-Voltage domain, and its activity may be directly regulated by blue light. Consistent with this notion, we found that blue light stabilizes PRR5, although it does not alter ZTL levels. Together, our results show that ZTL targets PRR5 for degradation by 26S proteasomes in the circadian clock and in early photomorphogenesis.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Circadian clocks have evolved to allow organisms to adapt to the rotational movement of the Earth. Upon perceiving environmental stimuli (inputs) such as light and temperature, these internal clocks can generate certain rhythms in the central oscillator, culminating in output mechanisms that allow an organism to adapt to its surroundings (Dodd et al., 2005). In Arabidopsis thaliana, the identification of several clock component genes has led to the first plant clock model, which proposes that CIRCADIAN CLOCK–ASSOCIATED1 (CCA1)/LONG ELONGATED HYPOCOTYL (LHY) (Myb-like transcription factors) and TIMING OF CAB EXPRESSION1 (TOC1) (a pseudo-response regulator) constitute an autoregulatory negative feedback loop. According to this model, TOC1 mRNA (and presumably TOC1 protein) accumulates around dusk and promotes the expression of CCA1/LHY, whose products in turn bind to the TOC1 promoter, repressing its transcription (Green and Tobin, 1999; Alabadí et al., 2001, 2002; Mizoguchi et al., 2002).

In addition to CCA1/LHY1 and TOC1, several other Arabidopsis genes have also been implicated in circadian control, although their sites of action in the clock model are still undefined. TOC1 is a member of the Pseudo-Response Regulator (PRR) family of proteins, which are plant-specific. PRR family members share two domains of unknown function: PR (for Pseudo-Receiver) and CCT (for CONSTANS, CONSTANS-like, TOC1). Interestingly, the PRR genes are transcribed in a sequential manner at different times of the day: PRR9 is transcribed in the morning, then PRR7, followed by PRR5 and PRR3, and finally TOC1 (Matsushika et al., 2000; Strayer et al., 2000). Recent genetic analyses suggested that, in addition to TOC1, the other PRRs also play essential roles within or close to the central oscillator (Somers et al., 1998; Eriksson et al., 2003; Kaczorowski and Quail, 2003; Yamamoto et al., 2003; Farré et al., 2005; Nakamichi et al., 2005a, 2005b). These roles were uncovered through analyses of mutants and transgenic lines in which PRR levels were altered and found to affect the circadian clock of these plants. Furthermore, PRRs are also involved in the control of hypocotyl elongation and flowering time (Makino et al., 2002; Matsushika et al., 2002, 2007b; Sato et al., 2002; Kaczorowski and Quail, 2003; Murakami et al., 2004).

In addition to PRRs, members of the ZEITLUPE (ZTL) family, ZTL (Somers et al., 2000; Jarillo et al., 2001), LOV KELCH PROTEIN2 (LKP2 [Schultz et al., 2001]), and FLAVIN BINDING KELCH F-BOX1 (FKF1 [Nelson et al., 2000]), have also been proposed to regulate clock function. Besides an F box domain, these proteins also possess a Light-Oxygen-Voltage (LOV) motif and six Kelch repeats at the N and C termini, respectively (Somers et al., 2000; Jarillo et al., 2001). The presence of an F box domain suggests that ZTL family members likely operate as components of a Skp1/Cullin/F box (SCF) complex, which presumably has E3 ubiquitin ligase activity. Of the three family members, ZTL is the best characterized with regard to its circadian function (Más et al., 2003b; Han et al., 2004; Somers et al., 2004). Más et al. (2003b) provided evidence that ZTL regulates TOC1 levels in darkness by ubiquitinating the latter and mediating its degradation by 26S proteasomes. The presence of a LOV domain in FKF1 led Imaizumi et al. (2003) to suggest that this protein might be directly regulated by blue light. Interestingly, the LOV domains of the two other family members, ZTL and LKP2, display spectrophotometric characteristics similar to that of the FKF1 LOV domain, suggesting a similar regulation (Imaizumi et al., 2003).

Proteasomal regulation of circadian proteins has been described in several eukaryotes. In Drosophila, the F box proteins Slimb (Slmb) and JETLAG (JET) were shown to regulate levels of the circadian proteins PERIOD (PER) and TIMELESS (TIM) (Grima et al., 2002; Ko et al., 2002; Koh et al., 2006). Similarly, F BOX/WD-40 REPEAT–CONTAINING PROTEIN1 (FWD-1), the Neurospora homolog of Slmb, targets the circadian protein FREQUENCY (FRQ) for degradation by the ubiquitin-proteasome pathway (He et al., 2003; He and Liu, 2005). Another Slmb homolog, ß-TRCP, was shown to regulate the PER homolog in human cells (Shirogane et al., 2005). Recently, several groups reported that circadian-associated proteins from Arabidopsis, such as GIGANTEA (GI [David et al., 2006]), LHY (Song and Carré, 2005), and CBK4 (Perales et al., 2006), are also regulated by proteasomes. Moreover, several enzymes involved in posttranslational modifications, such as the regulatory subunit of casein kinase II (CKB3 [Sugano et al., 1998]), poly(ADP-ribo)glycohydrase (TEJ [Panda et al., 2002]), and O-linked N-acetylglucosamine (SPINDLY [Tseng et al., 2004]), have been identified as circadian clock-associated factors. All of these results highlight the importance of posttranslational regulation in the clockwork of many different eukaryotes and indicate that analyses at the protein level are necessary for a better understanding of the plant clock.

Here, we investigated the regulation of PRR5, a component of the Arabidopsis central oscillator. PRR5 is one of the less studied members of the PRR gene family, and all data collected to date on this gene pertain only to mRNA variation and the analysis of mutant and overexpression phenotypes (Matsushika et al., 2000; Sato et al., 2002; Eriksson et al., 2003; Michael et al., 2003). Our aim was to investigate changes in, and the regulation of, PRR5 protein levels in a running clock. Here, we show that PRR5 protein is less stable in the dark and that its instability is mediated by the F box protein ZTL, which binds directly to the PR domain of PRR5. We also demonstrated that ZTL-mediated degradation of PRR5 is partially dependent on light quality, suggesting that ZTL activity might be regulated by conformational changes mediated by its LOV domain. Together, our results show that the stability of PRR5 is negatively regulated by ZTL and that this regulation is required for circadian clock function and early photomorphogenesis.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Circadian Rhythms of PRR5
To follow changes in the PRR5 protein in various diurnal and circadian conditions, we prepared an anti-PRR5 antibody and confirmed its specificity using extracts from the wild type, the prr5-1 mutant, and complementation lines expressing PRR5-FLAG (ProPRR5:PRR5-FLAG). Figure 1A shows that the antibody specifically detected PRR5 from the endogenous gene (70 kD) and PRR5-FLAG from the transgene (73 kD), although a faint cross-reacting band was found in all samples.

View larger version (76K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 1. PRR5 Protein and mRNA Levels in Various Circadian Conditions. (A) Characterization of the PRR5 antibody. Total protein was extracted from the wild type (Columbia [Col.]), prr5-1, two independent complemented lines (ProPRR5:PRR5-FLAG, line 1 [L1] and line 2 [L2]), and vector control lines (Empty vector, L1 and L2), entrained in 12-h-light/12-h-dark (12L/12D) cycles, and harvested at ZT9 and ZT21. Note that the 75-kD cross-reacting band (asterisk) is present in all samples. (B) to (D) PRR5 protein (first panel) and mRNA (third panel) levels were determined in wild-type plants grown under 12L/12D for 20 d and then transferred to 12L/12D (B), constant light (LL; [C]), and constant dark (DD; [D]). Quantification of these results is shown at bottom. All quantifications were standardized with the loading control values and calculated assuming that the peak signal in 12L/12D (ZT9 for mRNA and ZT11 for protein) corresponds to 100% of signal. (E) and (F) PRR5 protein (first panel) and mRNA (third panel) levels in 20-d-old wild-type plants grown under 16-h-light/8-h-dark (LD; [E]) and 8-h-light/16-h-dark (SD; [F]) cycles. Quantification of these results is shown at bottom. All quantifications were standardized with the loading control values and calculated assuming that the peak signal in LD (ZT9 for mRNA and ZT11 for protein) corresponds to 100% of signal. Tubulin protein and Ubiquitin10 (UBQ10) mRNA levels were used as protein and RNA loading controls, respectively. Each experiment was performed twice with similar results.

To further characterize the PRR5 circadian pattern, we analyzed wild-type plants grown under 16-h-light/8-h-dark (long-day [LD]) and 8-h-light/16-h-dark (short-day [SD]) conditions. Figures 1E and 1F show that PRR5 mRNA peaked earlier in SD than in LD, although the wave pattern was similar, confirming previous findings (Matsushika et al., 2000). However, the PRR5 protein cycled differently in SD than in LD. In LD, PRR5 was detected for a longer period after the mRNA levels became undetectable (Figure 1E). In SD, the peak of PRR5 was detected within the dark period and the overall PRR5 protein level was reduced (Figure 1F). Under both conditions, PRR5 levels decreased rapidly in the dark, after PRR5 mRNA fell to undetectable levels (Figures 1E and 1F). These results suggest that posttranscriptional regulation of PRR5 leads to a reduction in protein levels preferentially in darkness.

PRR5 Is Degraded by 26S Proteasomes
To gain insight into this possible posttranscriptional regulation of PRR5, we generated transgenic plants expressing FLAG-tagged PRR5 from a constitutive cauliflower mosaic virus (CaMV) 35S promoter (35S-PRR5). RNA gel blot experiments confirmed the constitutive expression of the transgenic PRR5 mRNA in 12L/12D. By contrast, the PRR5 protein displayed a wave pattern (Figure 2A ) confirming the existence of a posttranscriptional regulatory mechanism in light/dark cycles. However, the oscillation pattern of FLAG-tagged PRR5 was different from that of endogenous PRR5.

View larger version (36K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 2. Posttranslational Regulation of PRR5. (A) Diurnal patterns of PRR5 protein (first panel) and mRNA (third panel) in transgenic plants expressing FLAG-tagged PRR5 (FLAG-PRR5) from a CaMV 35S promoter (35S-PRR5). Plants were grown under 12L/12D cycles for 20 d and harvested at 4-h intervals starting at ZT0. White and black bars indicate subjective day and night, respectively. The bottom panel shows the quantification of these results. All quantifications were standardized with the loading control values and calculated assuming that the highest signal corresponds to 100% of signal. The stained 50-kD band (second panel) and Ubiquitin10 (UBQ10) mRNA levels (fourth panel) were used as protein and RNA loading controls, respectively. (B) and (C) PRR5 is stabilized upon MG132 treatment. 35S-PRR5 plants grown in constant light were incubated in MS medium in the absence (B) or presence (C) of 100 µM MG132 for 9 h (MG132 pretreatment) prior to cycloheximide (CHX) treatment. Plants were washed several times in MS medium to remove MG132 and then incubated with 100 µM cycloheximide in the presence (+) or absence (–) of MG132. Samples were harvested at 4-h intervals. FLAG-PRR5 protein was detected with anti-FLAG antibody and mRNA was detected with the PRR5-specific probe. After immunodetection, the same membranes were stained to visualize protein bands and the 50-kD band was used as a loading control. Each experiment was performed twice with similar results. Arrows indicate a protein band corresponding to the full-length FLAG-PRR5.

The F Box Protein ZTL Negatively Regulates PRR5
Several circadian proteins from a wide range of organisms (ZTL, FWD-1, Slmb, ß-TRCP, and JET) were recently identified as components of SCF complexes and shown to regulate the stability of central oscillators (Ko et al., 2002; He et al., 2003; Han et al., 2004; Shirogane et al., 2005; Koh et al., 2006). In Arabidopsis, TOC1, a component of the central oscillator, is regulated by the F box protein ZTL (Más et al., 2003b). To identify the responsible E3 ligase(s) for PRR5, we determined PRR5 levels at Zeitgeber time 11 (ZT11; PRR5 peak) and ZT23 (PRR5 trough) in several F box protein mutants belonging to the ZTL family: fkf1, lkp2, and ztl-4. We found that PRR5 was still detected at ZT21 in ztl-4 but not in fkf1, lkp2, and the wild type (see Supplemental Figure 1A online), suggesting that ZTL negatively regulates, directly or indirectly, the PRR5 protein. To investigate this further, we analyzed the PRR5 oscillation pattern in wild-type, ztl-3, and ZTL-overexpressor (ZTL-ox) plants grown under 12L/12D cycles. Figure 3A shows higher PRR5 levels in the morning and late night in ztl-3, when no PRR5 was detectable in the wild type. By contrast, PRR5 protein levels were extremely reduced at all time points analyzed in ZTL-ox (Figure 3A). We also detected differences in PRR5 mRNA patterns in ztl-3 mutants and ZTL-ox plants. However, these changes were less pronounced than the differences found at the protein level (Figure 3B).

View larger version (64K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 3. PRR5 Protein and mRNA Patterns in ztl and fkf1 Mutants and ZTL-Ox. (A) and (B) PRR5 protein (A) and transcript (B) levels in wild-type (Columbia [Col.]), ztl-3, and ZTL-ox were analyzed in 20-d-old plants entrained in 12L/12D. Samples were collected at 2-h intervals starting from ZT1. (C) and (D) Detection of PRR5 protein (C) and mRNA (D) levels in etiolated seedlings of wild-type, ztl-3, fkf1, and ZTL-ox plants. Five-day-old dark-grown etiolated seedlings were transferred to light, and samples were collected at the indicated time points. PRR5 protein was detected with a specific antibody. A major band at 50 kD was used as a loading control (Control). Ubiquitin10 (UBQ10) mRNA levels were used as loading controls in (B). PRR5 and -tubulin (-tub) mRNAs were detected by semiquantitative RT-PCR–aided DNA gel blot hybridization in (D). White and black bars represent light and dark periods, respectively. Each experiment was performed twice with similar results.

To test this possibility, we measured PRR5 half-life under white light or in darkness in etiolated seedlings of wild-type, ztl-3, ztl-4, fkf1, and TOC1 RNAi plants. In wild-type seedlings grown in the light, the PRR5 half-life was 6 h, suggesting some level of degradation. However, in dark-treated seedlings, the PRR5 half-life decreased to 3.6 h, confirming our previous results that PRR5 degradation was enhanced in darkness (Figures 4A and 4B ). Similar results were obtained in fkf1 and TOC1 RNAi seedlings (Figures 4A and 4B; see Supplemental Figures 1D and 1E online). By contrast, in ztl-3 or ztl-4 mutants, the PRR5 half-life was extended to 9 to 13 h independent of light or dark conditions (Figures 4A and 4B; see Supplemental Figures 1D and 1E online). These data show that ZTL negatively regulates PRR5 levels, as its depletion results in increased PRR5 stability. Because PRR5 continued to be degraded in ztl mutants, although at a slower rate, factors other than ZTL are also involved in the negative regulation of PRR5.

View larger version (33K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 4. PRR5 Protein Is Stabilized in the ztl Mutant. (A) Five-day-old etiolated wild-type (Columbia [Col.]), ztl-3, and fkf1 seedlings were exposed to white light for 9 h and then treated with 100 µM cycloheximide (CHX) and incubated under light (L) or dark (D) conditions. Samples were collected at the designated time points, and PRR5 levels were detected. A major band at 50 kD was used as a loading control (Control). Each experiment was performed three times with similar results. (B) Quantification of the results obtained in (A) after normalizing with loading controls and using time 0 as 100% of signal. Error bars represent SD of three independent experiments.

View larger version (53K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 5. PRR5 and ZTL Interact in Vitro and in Vivo. (A) Schematic representation of PRR5 deletion derivatives fused to maltose binding protein (MBP) and 3xHA (for hemagglutinin) tags (H). The regions fused to MBP and 3xHA are depicted as solid lines. The deleted regions are shown as dashed lines. (B) In vitro pull down of ZTL with PRR5. Top panel, input of GST-ZTL-myc detected with anti-c-myc antibody. Middle panel, input of MBP fusions PRR5 full length (MBP-PRR5FL-HA), PRR5 PR domain (MBP-PRR5PR-HA), PRR5 without the PR domain (MBP-PRR5PR-HA), PRR5 without the CCT domain (MBP-PRR5CCT-HA), negative control MBP-AIP2-HA protein, and MBP detected with 3xHA or MBP antibody. Bottom panel, detection of ZTL in amylose resin pulled-down fractions with anti-c-myc antibody. (C) In vivo reciprocal immunoprecipitation (IP) of PRR5 and ZTL. PRR5, PRR7, and PRR9 FLAG-tagged overexpressors (35S-PRR5, 35S-PRR7, and 35S-PRR9, respectively) and wild-type plants grown in constant light were treated for 15 h with 100 µM MG132. Top panel, Input levels of PRR5, PRR7, and PRR9 were confirmed with anti-FLAG. A major band at 50 kD was used as a loading control (Control). Middle panel, in vivo immunoprecipitation using anti-ZTL and detection with anti-FLAG. Bottom panel, in vivo immunoprecipitation with anti-FLAG and detection with anti-ZTL. Each experiment was performed twice with similar results.

To confirm the PRR5–ZTL interaction in planta, we generated transgenic lines in the prr5-1 mutant background. We used the PRR5 promoter (ProPRR5) to express full-length PRR5, a deletion mutant without the PR domain, and a mutant with the PRR5 PR domain alone. These transgenic lines were named ProPRR5:PRR5FL, ProPRR5:PRR5PR, and ProPRR5:PRR5PR, respectively. All of these deletion derivatives were shown previously to localize in the nucleus (Matsushika et al., 2007a). Plants were grown under 12L/12D cycles, and protein and mRNA levels were analyzed at ZT9 and ZT21 (Figure 6A ). In both ProPRR5:PRR5FL and ProPRR5:PRR5PR, mRNA and protein accumulated to higher levels only at ZT9, whereas in ProPRR5:PRR5PR, mutant PRR5 protein remained maximal even at ZT21, although the mRNA was undetectable (Figure 6A). We found that PRR5PR was more stable than PRR5FL and PRR5PR (Figure 6B). Furthermore, interaction with ZTL was detected for PRR5FL and PRR5PR but not PRR5PR in in vivo immunoprecipitation experiments (Figure 6C), confirming that ZTL-mediated regulation of PRR5 occurred through binding to its PR domain.

View larger version (42K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 6. ZTL Associates with the PR Domain of PRR5. Transgenic lines expressing PRR5 full-length protein (ProPRR5:PRR5FL) and deletion series, PRR5 deletion of the PR domain (ProPRR5:PRR5PR), and PRR5 PR domain (ProPRR5:PRR5PR), under the control of the PRR5 native promoter (ProPRR5) in the prr5-1 mutant background, were used. Note that all of the proteins were tagged with 3xFLAG on their N termini and with EGFP on their C termini. The major protein band of 50 kD and -tubulin (-tub) mRNA levels were used as protein and RNA loading controls, respectively. Each experiment was performed twice with similar results. (A) Expression of each deletion mutant protein and mRNA in planta. Protein and RNA samples were prepared from transgenic plants entrained in 12L/12D cycles. Samples were collected at the times indicated (ZT9 and ZT21) and analyzed by protein gel blot with anti-FLAG antibody and semiquantitative RT-PCR with specific primer sets. (B) Half-life of each deletion mutant of PRR5. Twenty-day-old plants grown in 12L/12D cycles were incubated with 100 µM cycloheximide (CHX) starting at ZT9 and harvested at 3-h intervals up to 9 h under white light. (C) In vivo immunoprecipitation (IP) of ZTL with PRR5 deletion mutants. Twenty-day-old transgenic plants grown in 12L/12D cycles were incubated with 100 µM MG132 for 9 h before harvesting at ZT15. Top panel, input levels of deletion mutants were confirmed with anti-FLAG. Middle panel, in vivo immunoprecipitation using anti-FLAG and detection with anti-ZTL. Bottom panel, in vivo immunoprecipitation using anti-ZTL and detection with anti-FLAG.

View larger version (49K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 7. PRR5 Protein Stability Increases under Blue Light. (A) Five-day-old wild-type (Columbia) etiolated seedlings were incubated in white light for 9 h, treated with 100 µM cycloheximide (CHX), and transferred to white light, blue light, red light, and dark conditions. Protein samples were analyzed by protein gel blotting using anti-PRR5 antibody. A major protein band at 50 kD was used as a loading control (Control). Each experiment was performed three times with similar results. (B) Quantification of the results obtained in (A) after normalizing with the loading controls and assuming time 0 as 100% of signal. Error bars represent SD of three independent experiments.

First, we analyzed hypocotyl elongation during early photomorphogenesis in various mutants, transgenic lines, and the wild type. Under 1 µmol·m^–2·s^–1 red light, prr5-1 and ZTL-ox displayed longer hypocotyls, whereas ztl mutants and 35S-PRR5 showed shorter hypocotyls (Figure 8A ) (Sato et al., 2002; Yamamoto et al., 2003; Somers et al., 2004). In prr5-1 ztl-3 and prr5-1 ztl-4 double mutants, the red light hypersensitivity of ztl was only partially suppressed by the prr5-1 mutation (Figure 8A), suggesting the existence of other ZTL-targeted factors that act as positive regulators in the red light response. This notion is further supported by the differential red light hyposensitivity phenotype between ZTL-ox and prr5-1. We have shown that PRR5 levels were extremely reduced in ZTL-ox similar to prr5-1 (Figure 3C), although its hyposensitivity phenotype was stronger than that of prr5-1 (Figure 8A). This effect was observed over a wide range of red light fluence rates > 1 µmol·m^–2·s^–1 (Figure 8B; data not shown), but no statistically significant difference in hypocotyl lengths was seen under blue light and in the dark (Figure 8C).

View larger version (21K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 8. Hypocotyl Elongation of the prr5 ztl Double Mutant during Early Photomorphogenesis. Seedlings were grown under the indicated fluence rates of red light ([A] and [B]) and blue light (C). Wild-type (Columbia [Col.]), prr5-1, ztl-3, ztl-4, prr5-1 ztl-3, prr5-1 ztl-4, 35S-PRR5, and ZTL-ox seeds were stratified for 48 h, exposed to white light for 3 h, and then placed under the appropriate light quality and fluence rate for 5 d. A photograph was taken of the seedlings grown under red light with a fluence rate of 1 µmol·m–2·s–1 ([A], top). Hypocotyl lengths were measured and SD was calculated (n = 20 to 25). The asterisks indicate significant differences from the prr5-1 single mutant (P < 0.001, as determined by Student's t test). Bar in (A) = 5 mm.

View larger version (96K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 9. Analysis of Circadian-Controlled Genes in the prr5 ztl Double Mutant. Wild-type (Columbia [Col.]), prr5-1 ztl-3, prr5-1, and ztl-3 plants were grown under 12L/12D for 20 d. Plants were released to continuous light conditions, and samples were collected at 2-h intervals. PRR7 (A) and CCA1 (B) mRNAs were detected with each specific probe. Ubiquitin10 (UBQ10) mRNA levels were used as loading controls. White and gray bars indicate subjective days and nights, respectively. Each experiment was performed twice with similar results.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Posttranscriptional Regulation of PRR5
Matsushika et al. (2000) previously showed that the PRR5 mRNA temporal pattern of accumulation is stringently regulated by the circadian clock, but no data were shown for PRR5 protein levels under the same conditions. Using a PRR5-specific antibody, we demonstrated that PRR5 protein and mRNA levels do not always synchronize. As expected, PRR5 wave patterns were different in LL, DD, LD, and SD. Interestingly, constitutive expression of PRR5 still resulted in a cycling of the PRR5 protein, suggesting the existence of posttranscriptional regulatory mechanism(s) in response to light conditions. Our results showed that the posttranscriptional regulation of PRR5 was most evident in darkness, similar to TOC1 (Más et al., 2003b) and GI (David et al., 2006). Moreover, the dark-induced degradation of TOC1 and GI is mediated by 26S proteasomes (Más et al., 2003b; David et al., 2006), and our results indicate a similar regulatory pathway for PRR5. However, although both PRR5 and TOC1 levels are controlled by the proteasomal pathway, the rate of PRR5 depletion in darkness was faster than that of TOC1 (Figures 1 and 3) (Más et al., 2003b), indicating a more stringent regulation of PRR5 by light/dark transition. This mechanism might account for the different PRR5 waveforms detected in SD and LD. Our results suggest that PRR5 might be involved in a daylength-sensing process, which could explain its role in the control of flowering time (Sato et al., 2002; Yamamoto et al., 2003; Nakamichi et al., 2007).

PRR5 Degradation Requires ZTL
ZTL is an F box protein that targets TOC1 for degradation in darkness (Más et al., 2003b). Several lines of evidence presented here support the view that PRR5 is also negatively regulated by ZTL at a posttranslational level: (1) PRR5 protein levels vary in ztl mutants and ZTL-ox plants, inversely correlating with changes in ZTL levels; (2) PRR5 half-life is extended in ztl mutants compared with the wild type; (3) PRR5 and ZTL interact in vitro and in vivo, and the PR domain is critical for this association; and (4) deletion of the PR motif results in PRR5 stabilization similar to ztl mutants. Matsushika et al. (2007a) reported that constitutive expression of the different PRR5 domains produced different phenotypes. The overexpression of the C-terminal region of PRR5 alone could account for the red light hypersensitivity, early flowering, and the circadian phenotype displayed in full-length overexpressors. By contrast, overexpressing the PR domain alone resulted in long-period rhythms phenocopying ztl-21 mutants (Kévei et al., 2006; Matsushika et al., 2007a). Consistent with our results, Matsushika et al. (2007a) proposed that the phenotype could be explained by the formation of nonproductive complexes between the PR domain and ZTL, thereby depleting the latter. Among the five members of the PRR family, PRR5 and PRR9 possess the most closely related PR domains, sharing 70% amino acid similarity. Notwithstanding this high sequence homology, we were unable to detect any PRR9–ZTL interaction, confirming the observations of Yasuhara et al. (2004). Since TOC1 and PRR5 PR domains share only 40% similarity, and we could not identify any other motif specifically conserved between the two PR domains, it seems unlikely that the TOC1–ZTL interaction is mediated by the PR domain, although the interacting region of TOC1 has not been characterized. It is noteworthy that Yasuhara et al. (2004) as well as Más et al. (2003b) reported that the LOV domain of ZTL is responsible for its interaction with both PRR5 and TOC1. The differences in kinetics of the waveforms of PRR5 and TOC1 also suggest that the mechanism of the ZTL(LOV)–PRR5 interaction is probably different from that of the ZTL(LOV)–TOC1 interaction.

PRR5 Is Required for ZTL Regulation of the Clock and Early Photomorphogenesis
Circadian-affected mutants typically display two phenotypes: differential hypocotyl elongation under red light and alterations in free-running rhythms of circadian-associated genes. Mutants deficient in PRR5 and ZTL show opposite phenotypes (Eriksson et al., 2003; Yamamoto et al., 2003; Somers et al., 2004), and our biochemical data revealed that the two proteins interact. Analysis of the hypocotyl elongation and circadian phenotype of prr5 ztl double mutants provides the physiological relevance of this interaction. We observed a partial suppression of ztl red light hypersensitivity and a lengthening of the free-running period by the prr5-1 mutation, confirming that PRR5 is required for the ZTL-dependent phenotypes. The partial effect found is likely due to the existence of multiple ZTL targets, such as TOC1. TOC1 is a positive regulator of both red light signaling during deetiolation and period length of the clock (Más et al., 2003a, 2003b). Accumulation of TOC1, along with other putative targets of ZTL, could lead to an intermediate phenotype in prr5-1 ztl double mutants. Interestingly, toc1 ztl double mutants did not show an intermediate phenotype but rather displayed the toc1 short-period phenotype (Más et al., 2003a, 2003b). This discrepancy could be explained as follows: prr5-1 shows a weaker period phenotype than toc1, and the accumulation of PRR5 does affect period length (Sato et al., 2002; Yamamoto et al., 2003). Moreover, we could not detect PRR5 stable accumulation at a maximum level in ztl mutants, as was described for TOC1 (Más et al., 2003b). Therefore, the contribution of PRR5 in toc1 ztl phenotypes might be weaker than that of TOC1 in prr5 ztl mutants. Further analysis of hypocotyl elongation in toc1 ztl would be important to clarify this issue, as well as the study of prr5 toc1 double and prr5 toc1 ztl triple mutants.

Blue Light Affects PRR5 Stability through ZTL Regulation
LOV domains present in ZTL, LKP2, and FKF1 show spectral properties that are typical of blue light photoreceptors (Imaizumi et al., 2003), indicating the existence of a blue light–dependent regulatory mechanism. This possibility has been suggested for FKF1, although the exact mechanism is not clear (Imaizumi et al., 2003, 2005).

Our evidence shows that PRR5 is regulated by blue light in a ZTL-dependent manner. Moreover, ZTL levels were constant under all of the light conditions tested, indicating that ZTL is subjected to qualitative modulation rather than quantitative regulation in our experimental conditions. Blue light could induce activation of the LOV domain, leading to conformational changes and disruption of the PRR5/ZTL association. Alternatively, photoactivation of the LOV domain could interfere with the assembly or function of the SCF E3 ligase complex. In any case, our results suggest that PRR5 regulation by ZTL is partially dependent on blue light signaling, which may constitute a fine-tuning mechanism to respond to changes in the environment.

Posttranslational Regulatory Mechanisms Account for the Regulation of Clock Components
The basic clock mechanism is being elucidated in cyanobacteria, fungi, Drosophila, and mammals, in which it was proposed that the central oscillator consists of a transcriptional–translational feedback loop. Recent studies have uncovered several layers of posttranslational regulation, including protein phosphorylation, degradation, complex formation, and subcellular translocation, all of which are crucial for the stability and flexibility of the clock (Daniel et al., 2004; Tseng et al., 2004; Tamanini et al., 2005). In plants, the current clock model consists of mostly transcriptional regulatory loops because of a paucity of experimental data for posttranslational regulatory events. Our detailed analyses of PRR5 revealed that protein and transcript levels do not always synchronize. Furthermore, we have uncovered a step in PRR5 negative regulation mediated by ZTL, which could have not been identified by transcriptional analysis alone. Proteasomal degradation in many organisms is preceded by other posttranslational modifications, such as phosphorylation (Sugano et al., 1998; Ko et al., 2002; He et al., 2003; Shirogane et al., 2005). Whether this is the case for PRR5 remains to be determined. We expect that several levels of posttranslational regulation remain to be uncovered for PRR5 and other clock-related factors. Deciphering these processes will lead to a better understanding and a more precise description of the plant circadian clock.

	METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Plant Materials and Growth Conditions
Unless otherwise stated, Arabidopsis thaliana Columbia ecotype was used as the wild type. Seedlings were grown on Murashige and Skoog (MS) plates supplemented with 1% sucrose under white light (60 to 80 µmol·m^–2·s^–1, in constant light, 16 h of light/8 h of dark [LD], 12 h of light/12 h of dark [12L/12D], or 8 h of light/16 h of dark [SD]) at 22°C for 20 d. To obtain etiolated seedlings, seeds were sown on MS plates without sucrose. After 2 d of stratification, seeds were exposed to white light (60 to 80 µmol·m^–2·s^–1) for 3 h at 22°C to enhance germination and then incubated for 5 d in the dark at 22°C.

The T-DNA insertion alleles prr5-1 (SALK_006280 [Eriksson et al., 2003]), ztl-3 (SALK_035701 [Jarillo et al., 2001; Kim et al., 2003]), ztl-4 (SALK_012440 [Salomé and McClung, 2005]), and lkp2 (SALK_036083) were isolated from the Salk collection (http://signal.salk.edu). These alleles, as well as fkf1 (Nelson et al., 2000), phyA-201 phyB-9 (Devlin et al., 1999), cry1 cry2 (Lascève et al., 1999), phot1-5 phot2-1 (Kinoshita et al., 2001), ZTL-ox H6 (Nelson et al., 2000), ZTL-ox G3 (Nelson et al., 2000), and TOC1 RNAi (Más et al., 2003a), have been described. The prr5-1 ztl-3 and prr5-1 ztl-4 double mutants were generated by standard genetic crosses and PCR analysis using gene-specific and insertion-specific primers as follows: 5'-TACACAGTATTCACCGAGTCCT-3' and 5'-CAAAGCTAAGTGTTTGTGTTGTG-3' for prr5-1, 5'-ATGAAGATATTGTCACGCCTGA-3' and 5'-TTCTCTTTCATACATAGGCTCTC-3' for ztl-3 and ztl-4, and 5'-GCGTGGACCGCTTGCTGCAACT-3' for the insertion.

Generation of Transgenic Plants
PRR5, PRR7, and PRR9 full-length genomic fragments (without stop codon) were amplified with the KOD plus DNA polymerase (EMD Biosciences) and cloned into pLITMUS29 (New England Biolabs), which carries a 3xFLAG tag N-terminal fusion cassette. These constructs were transferred into pBA002-HA (Jang et al., 2005), generating N-terminal 3xFLAG and C-terminal 3xHA fusions under the control of a CaMV 35S promoter. Wild-type (Columbia) plants were vacuum-infiltrated with Agrobacterium tumefaciens (Zhang et al., 2006), and transgenic lines were selected with the appropriate antibiotics. Lines segregating as single insertions were maintained and homozygous plants were selected. These lines were named 35S-PRR5, 35S-PRR7, and 35S-PRR9.

The PRR5 promoter sequence (ProPRR5; 2842 bp upstream of the inferred initiation codon), a 3xFLAG cassette, and a 3' untranslated region sequence (910 bp downstream of the stop codon) were assembled in the 5' to 3' order in the pBA002a binary vector, generating the ProPRR5:3xFLAG:3'-UTR construct, which was named pBA-PF5. The pBA002a was generated from the pBA002 binary vector (Jang et al., 2005) by eliminating the CaMV 35S promoter region. The ProPRR5:PRR5-FLAG construct was prepared by introducing the PRR5 full-length genomic fragment into pBA-PF5 vector. To create ProPRR5:FLAG-PRR5FL-EGFP, ProPRR5:FLAG-PRR5PR-EGFP, and ProPRR5:FLAG-PRR5PR-EGFP constructs, cDNA fragments corresponding to amino acid residues 1 to 558 (PRR5 full length; PRR5FL), 1 to 180 (PR domain; PRR5PR), and 172 to 558 (deletion of the PR domain; PRR5PR) were fused with EGFP coding sequence (Clontech) and introduced to pBA-PF5. These constructs were transferred into Agrobacterium, and positive colonies were confirmed by PCR. The prr5-1 mutants were then vacuum-infiltrated (Zhang et al., 2006) with agrobacterial cells carrying different plasmids. Transgenic lines with a single insertion were maintained, and T3 homozygous plants were identified. These lines were named ProPRR5:PRR5-FLAG, ProPRR5:PRR5FL, ProPRR5:PRR5PR, and ProPRR5:PRR5PR, respectively.

PRR5 and ZTL Antibodies
cDNA fragments encoding amino acid residues 178 to 500 of PRR5 and 1 to 283 of ZTL were cloned into the expression vector pQE30 (Qiagen). Recombinant proteins were expressed in Escherichia coli BL21 (DE3) and purified using a nickel-nitrilotriacetic acid agarose column (Qiagen). Recombinant proteins were used to generate polyclonal anti-PRR5 and anti-ZTL antisera (Cocalico Biological). Each antiserum was subjected to immunoaffinity purification against its cognate antigen immobilized to a nitrocellulose membrane.

MG132 and Cycloheximide Treatments
Seedlings were incubated with 100 µM MG132 (EMD Biosciences) and/or 100 µM cycloheximide (EMD Biosciences) in liquid MS medium for the indicated durations and used for total protein extraction.

RNA Analysis
Total RNA was extracted by the aurintricarboxylic acid method and blotted as described (Taniguchi et al., 1998). Blots were hybridized with gene-specific radiolabeled probes for PRR5, PRR7, GI, CCA1, and UBQ10 genes as described (Makino et al., 2002; Yamamoto et al., 2003). For RT-PCR, total RNA was isolated with the RNeasy plant mini kit (Qiagen), and 1 µg of RNA was used for cDNA synthesis using the SuperScriptIII first-strand synthesis system (Invitrogen). PCR was performed as described (Ishikawa et al., 2006) with the following primer sets: 5'-GACAGATTATCGCTGCTCTTCT-3' and 5'-ACTCTCATTCCATGGAAAGCTAT-3' for PRR5, 5'-GACTACAAAGACCATGACGGTGATTAT-3' and 5'-ATGAACTTCAGGGTCAGCTTGCCGTAG-3' for FLAG-EGFP fusion genes, and 5'-GGACAAGCTGGGATCCAGG-3' and 5'-CGTCTCCACCTTCAGCACC-3' for -tubulin.

Protein Extraction and Blotting
Frozen plant materials were ground to a fine powder and suspended in a 1:1 ratio (w/v) with 2x lithium dodecyl sulfate sample buffer, and the suspension was incubated for 5 min at 95°C. Protein samples were separated by 8% SDS-PAGE, blotted to an Immobilon-P membrane (Millipore), and incubated with the following primary antibodies: affinity-purified anti-PRR5 and anti-ZTL, commercial monoclonal antibodies, anti-c-myc, anti-HA, and anti-MBP (Santa Cruz Biotechnology), anti-FLAG M5, and anti--tubulin (Sigma-Aldrich). Anti-mouse IgG and anti-rabbit IgG conjugated with horseradish peroxidase were used as secondary antibodies, and protein signals were detected using the ECL system (GE Healthcare). After immunoblotting, all membranes were stained with a Commassie Brilliant Blue R 250 solution (0.2% Commassie Brilliant Blue R 250, 45% methanol, and 10% acetic acid) to visualize protein bands. A major band at 50 kD on stained membranes was used as a loading control.

Protein Half-Life Determination
Seedlings were preincubated for 30 min in MS liquid medium with 100 µM cycloheximide under white light (60 to 80 µmol·m^–2·s^–1) to inhibit de novo protein synthesis and then transferred to specific light conditions for the indicated periods of time. Fluence rates for the various light treatments were as follows: white, 20 µmol·m^–2·s^–1; blue, 20 µmol·m^–2·s^–1; red, 20 µmol·m^–2·s^–1; and far-red, 4 µmol·m^–2·s^–1.

In Vitro Pull-Down Assays
cDNA fragments encoding amino acid residues 1 to 558 (PRR5 full length; PRR5FL), 1 to 180 (PR domain only; PRR5PR), 172 to 558 (deletion of the PR domain; PRR5PR), and 1 to 501 (deletion of the CCT domain; PRR5CCT) were cloned into the pMAL-HA expression vector, a derivative of the pMAL-c1 vector (New England Biolabs) modified to express an N-terminal MBP-tagged and a C-terminal 3xHA-tagged protein. Recombinant proteins (MBP-PRR5FL-HA, MBP-PRR5PR-HA, MBP-PRR5PR-HA, and MBP-PRR5CCT-HA) were expressed in E. coli BL21 (DE3) and purified on an amylose column, according to the manufacturer's instructions (New England Biolabs). ZTL coding sequence was cloned into the pGEX-4T-1 vector (GE Healthcare) with N-terminal GST and C-terminal 3xmyc tags. This recombinant protein (GST-ZTL-myc) was expressed in E. coli BL21 (DE3) and purified using a glutathione–Sepharose column.

In vitro pull-down assays were performed as described using MBP, GST, and MBP-AIP2-HA (Zhang et al., 2005) as negative controls. The assays were repeated at least three times with similar results.

In Vivo Immunoprecipitation
Plants grown for 20 d in 12L/12D cycles were used. Wild-type (Columbia), 35S-PRR5, 35S-PRR7, and 35S-PRR9 plants were treated with 100 µM MG132 for 15 h, and ProPRR5:PRR5FL, ProPRR5:PRR5PR, and ProPRR5:PRR5PR plants were treated with 100 µM MG132 for 9 h prior to sampling. Protein was extracted from these samples as described (Más et al., 2003b). In vivo immunoprecipitation experiments were done in triplicate using 500 µg of total protein that was initially incubated with 20 µL of 50:50 protein A or G beads in immunoprecipitation buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.5% Nonidet P-40, 1 mM EDTA, 3 mM DTT, 1x plant protease inhibitor cocktail [Sigma-Aldrich], 100 µM N-acetyl-Leu-Leu-Nle-CHO, 1 µM epoxomicin, and 4 µM clasto-lactacystin ß-lactone [EMD Biosciences]) for 1 h at 4°C to remove nonspecific binding. The mix was centrifuged for 3 min at 1000g, and the supernatant was collected. The precleared supernatant was incubated with 15 µL of anti-ZTL or 5 µL of anti-HA (Santa Cruz Biotechnology) antibodies for 2 h at 4°C. Then, 30 µL of protein A (for anti-ZTL immunoprecipitation) or protein G (for anti-HA immunoprecipitation) was added, and the mixture was incubated for another 2 h at 4°C. After this step, the triplicate samples were pooled and each immunoprecipitation was washed three times with the immunoprecipitation buffer. The immunoprecipitation beads were then incubated with 40 µL of 2x lithium dodecyl sulfate loading dye and denatured for 3 min at 95°C. Half of the denatured sample was separated by 8% SDS-PAGE and detected with the appropriate antibodies. The immunoprecipitation experiments were repeated at least three times with similar results.

Hypocotyl Length Analyses
Seeds were sown on MS plates without sucrose and stratified for 48 h in the dark. After exposure to white light for 3 h, plates were placed at 22°C under the indicated light quality and fluence rate. Hypocotyl lengths were measured after 5 d of incubation.

Image Quantification
Exposed films of RNA and protein gel blots, as well as stained membranes, were scanned and analyzed using NIH ImageJ software (http://rsb.info.nih.gov/ij/). Changes in transcript and protein levels were normalized using appropriate loading controls (UBQ10 for RNA and tubulin or the 50-kD stained protein band).

Accession Numbers
Sequence data for the genes described in this article can be found in the Arabidopsis Genome Initiative and GenBank/DDBJ/EMBL data libraries under the following accession numbers: TOC1 (At5g61380), PRR5 (At5g24470), PRR7 (At5g02810), PRR9 (At2g46790), FKF1 (At1g68050), and ZTL (At5g57360).

Supplemental Data
The following materials are available in the online version of this article.

Supplemental Figure 1. PRR5 Protein Levels in Various Genetic Backgrounds.

Supplemental Figure 2. Characterization of ZTL-Specific Antibody.

Supplemental Figure 3. Blue Light Stabilization of PRR5 Is Independent of Phytochromes, Cryptochromes, or Phototropins.

	Acknowledgments

 
We thank B. Bartel for fkf1, ZTL-ox H6, and ZTL-ox G3; G.C. Whitelam for phyA-201 phyB-9; C. Lin for cry1 cry2; and T. Kagawa and M. Wada for phot1-5 phot2-1. We are grateful to T. Mizuno, N. Nakamichi, and S. Ito for helpful discussion. We are also thankful to Raquel Martin and Isabel Abreu for critical reading of the manuscript. T.K. was supported in part by a postdoctoral fellowship for research abroad from the Japan Society for the Promotion of Science, and R.H. was supported in part by a postdoctoral fellowship from the Gulbenkian Foundation, Portugal (78682/2006). This work was supported by National Institutes of Health Grant GM-44640 to N.-H.C.

	Footnotes

 
¹ These authors contributed equally to this work.

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantcell.org) is: Nam-Hai Chua (chua{at}mail.rockefeller.edu).

^[W] Online version contains Web-only data.

www.plantcell.org/cgi/doi/10.1105/tpc.107.053033

Received May 21, 2007; Revision received July 5, 2007. accepted July 17, 2007.

	REFERENCES

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Alabadí, D., Oyama, T., Yanovsky, M.J., Harmon, F.G., Más, P., and Kay, S.A. (2001). Reciprocal regulation between TOC1 and LHY/CCA1 within the Arabidopsis circadian clock. Science 293: 880–883.[Abstract/Free Full Text]

Alabadí, D., Yanovsky, M.J., Más, P., Harmer, S.L., and Kay, S.A. (2002). Critical role for CCA1 and LHY in maintaining circadian rhythmicity in Arabidopsis. Curr. Biol. 12: 757–761.[CrossRef][ISI][Medline]

Daniel, X., Sugano, S., and Tobin, E.M. (2004). CK2 phosphorylation of CCA1 is necessary for its circadian oscillator function in Arabidopsis. Proc. Natl. Acad. Sci. USA 101: 3292–3297.[Abstract/Free Full Text]

David, K.M., Armbruster, U., Tama, N., and Putterill, J. (2006). Arabidopsis GIGANTEA protein is post-transcriptionally regulated by light and dark. FEBS Lett. 580: 1193–1197.[CrossRef][ISI][Medline]

Devlin, P.F., Robson, P.R.H., Patel, S.R., Goosey, L., Sharrock, R.A., and Whitelam, G.C. (1999). Phytochrome D acts in the shade-avoidance syndrome in Arabidopsis by controlling elongation growth and flowering time. Plant Physiol. 119: 909–915.[Abstract/Free Full Text]

Dodd, A.N., Salathia, N., Hall, A., Kévei, E., Tóth, R., Nagy, F., Hibberd, J.M., Millar, A.J., and Webb, A.R. (2005). Plant circadian clocks increase photosynthesis, growth, survival, and competitive advantage. Science 309: 630–633.[Abstract/Free Full Text]

Eriksson, M., Hanano, S., Southern, M.M., Hall, A., and Millar, A.J. (2003). Response regulator homologues have complementary, light-dependent functions in the Arabidopsis circadian clock. Planta 218: 159–162.[CrossRef][ISI][Medline]

Farré, E.M., Harmer, S.L., Harmon, F.G., Yanovsky, M.J., and Kay, S.A. (2005). Overlapping and distinct roles of PRR7 and PRR9 in the Arabidopsis circadian clock. Curr. Biol. 15: 47–54.[CrossRef][ISI][Medline]

Green, R.M., and Tobin, E.M. (1999). Loss of the circadian clock-associated protein 1 in Arabidopsis results in altered clock-regulated gene expression. Proc. Natl. Acad. Sci. USA 96: 4176–4179.[Abstract/Free Full Text]

Grima, B., Lamouroux, A., Chélot, E., Papin, C., Limbourg-Bouchon, B., and Rouyer, F. (2002). The F-box protein Slimb controls the levels of clock proteins Period and Timeless. Nature 420: 178–182.[CrossRef][Medline]

Han, L., Mason, M., Risseeuw, E.P., Crosby, W.L., and Somers, D.E. (2004). Formation of an SCF^(ZTL) complex is required for proper regulation of circadian timing. Plant J. 40: 291–301.[CrossRef][ISI][Medline]

He, Q., Cheng, P., Yang, Y., He, Q., Yu, H., and Liu, Y. (2003). FWD1-mediated degradation of FREQUENCY in Neurospora establishes a conserved mechanism for circadian clock regulation. EMBO J. 22: 4421–4430.[CrossRef][ISI][Medline]