- Split View
-
Views
-
Cite
Cite
Seiya Makino, Akinori Matsushika, Masaya Kojima, Takafumi Yamashino, Takeshi Mizuno, The APRR1/TOC1 Quintet Implicated in Circadian Rhythms of Arabidopsis thaliana: I. Characterization with APRR1-Overexpressing Plants, Plant and Cell Physiology, Volume 43, Issue 1, 15 January 2002, Pages 58–69, https://doi.org/10.1093/pcp/pcf005
- Share Icon Share
Abstract
Several Arabidopsis genes have been proposed to encode potential clock-associated components, including the Myb-related CCA1 and LHY transcription factors and a member of the novel family of pseudo response regulators (APRR1/TOC1). We previously showed that mRNAs of the APRR1/TOC1 family of genes start accumulating after dawn rhythmically and sequentially at approximately 2 h intervals in the order: APRR9→APRR7→APRR5→APRR3 →APRR1/TOC1. Here we constructed APRR1-overexpressing (APRR1-ox) plants, and examined certain circadian profiles for APRRs, CCA1, LHY, GI, CCR2, and CAB2. The free-running circadian rhythms of the APRR1/TOC1 family of genes, including APRR1, were dampened in APRR1-ox plants. In particular, the light-inducible expression of APRR9 was severely repressed in APRR1-ox plants, suggesting that there is a negative APRR1→APRR9 regulation. The free-running robust rhythm of CAB2 was also dampened in APRR1-ox. The circadian profiles of potential clock-associated genes, CCA1, LHY, GI, and CCR2 were all markedly altered in APRR1-ox, each in characteristic fashion. To gain further insight into the molecular function of APRR1, we then identified a novel Myc-related bHLH transcription factor, which physically associated with APRR1. This protein (named PIL1) is similar in its amino acid sequence to PIF3, which has been identified as a phytochrome-interacting transcription factor. These results are discussed in relation to the current idea that APRR1 (TOC1) plays a role within, or close to, the Arabidopsis central oscillator.
(Received July 19, 2001; Accepted October 17, 2001).
Introduction
During the past few years, there has been a tremendous increase in our understanding of the molecular bases of the biological clocks in a wide range of organisms, including Neurospora, Drosophila and Synechococcus (Dunlap 1999). In higher plants, such a clock-dependent circadian rhythm is a very old issue in physiology (Garner and Allard 1920), and a newly emerging paradigm of molecular biology (for reviews see Anderson and Kay 1996, Thomas and Vince-Prue 1997, Kreps and Kay 1997, Koornneef et al. 1998, Piechulla 1999, Staiger and Heintzen 1999, Murtas and Millar 2000, Samach and Coupland 2000, Barak et al. 2000, Carre 2001). A number of circadian-regulated genes have been identified through extensive analyses with DNA microarrays of Arabidopsis (Harmer et al. 2000, Schaffer et al. 2001). In general, a biological clock is presumed to consist of input pathways, central oscillators, and output pathways (Dunlap 1999). A central oscillator is thought to autonomously generate a rhythm through transcription/translation positive/negative feedback loops. In the model higher plant, Arabidopsis thaliana, biological and genetic studies have begun to shed light on the molecular bases of such clocks in higher plants (see the reviews cited above).
With regard to the input pathways responsible for the entrainment of plant circadian clocks, light is obviously an important factor. Members of both the phytochrome (PHY) and cryptochrome (CRY) families of Arabidopsis photoreceptors have been shown to act as part of the input pathways (Somers et al. 1998a). In white light, however, a phyAphyBcry1cry2 quadruple mutant retains robust circadian rhythmicity of leaf movement, suggesting that an, as yet unidentified, photoreceptor(s) can also function in the input pathways (for a review see Barak et al. 2000). Through intensive genetic approaches, several other Arabidopsis genes have also been uncovered as ones encoding additional components of the input pathways. They include the ELF3, GI, ZTL, and FKF1 genes. Mutations in ELF3 (EARLY FLOWERING 3) result in arhythmicity in light but not in the dark (Hicks et al. 1996, Zagotta et al. 1996, Reed et al. 2000). The ELF3 gene product was recently identified and extensively characterized (Hicks et al. 2001, Liu et al. 2001, Covington et al. 2001). Mutations in GI (GIGANTEA) cause altered periods and reduced amplitudes of expression of certain circadian-associated genes, including CCA1 and LHY (Fowler et al. 1999, Park et al. 1999, Huq et al. 2000, and see below for CCA1 and LHY). The ZTL (ZEITLUPE) and FKF1 (flavin-binding, kelch repeat, F-box) genes both encode a protein that contains a common LOV (light, oxygen, voltage) domain, together with an F-box and kelch-repeats (Nelson et al. 2000, Somers et al. 2000). Although all these components are assumed to play roles in controlling input from light signals to the clock, their molecular functions are not yet fully understood.
With regard to the Arabidopsis circadian oscillator(s), several genes have proposed to encode potential clock-associated components, although none of them have yet been fully established as a bona fide clock component. Best characterized are the Myb-related CIRCADIAN CLOCK ASSOCIATED 1 (CCA1) and LATE ELONGATED HYPOCOTYL (LHY) genes, each of which encodes a putative transcription factor similar to each other (Wang et al. 1997, Green and Tobin 1999, Schaffer et al. 1998). Both CCA1 and LHY show robust and free-running circadian oscillations at both the transcript and protein levels. Overexpression of CCA1 in plants stops overt rhythmicity, including the expression of genes that are under control of the circadian rhythm (Wang and Tobin 1998). The examples of such circadian-controlled genes are those encoding light-harvesting chlorophyll-a/b-binding proteins (e.g. CAB2) and a catalase (CAT3) (for a review see Piechulla 1999). Another gene, TIMING OF CAB 1 (TOC1), is also intriguing, because a semi-dominant (toc1-1) mutant exhibits shortened periods of several rhythmic markers, including the CAB2 gene, and also of the leaf movement and stomatal conductance rhythms (Somers et al. 1998b). Since this single mutation interferes with a wide range of clock-controlled output processes, the TOC1 protein was also proposed to function close to, or as part of, the oscillator itself. The TOC1 gene was recently isolated and found to encode a nuclear protein with a motif similar to that found in the CONSTANS family of transcriptional activators (Robert et al. 1998, Strayer et al. 2000). Besides these candidate components of oscillators, a pair of putative RNA-binding proteins (CCR1/CCR2) was characterized as sub- (or slave) oscillator components (Kreps and Simon 1997, Heintzen et al. 1997).
Recent reverse-genomic studies have provided further insights into such Arabidopsis circadian clocks. Thanks to completion of the entire Arabidopsis genome sequence (The Arabidopsis Genome Initiative 2000), we previously characterized a novel family of genes, encoding Arabidopsis pseudo response regulators (designated as APRR1, APRR3, APRR5, APRR7, and APRR9) (Makino et al. 2000, Matsushika et al. 2000, Makino et al. 2001). Each of these APRR products contains a domain similar to the receiver of response regulators that are implicated in widespread signal transduction systems, generally referred to as “histidine (His)-to-aspartate (Asp) phosphorelay” (Imamura et al. 1999, Makino et al. 2000). However, they are atypical in that the conserved aspartate residue that normally undergoes phosphorylation in other authentic response regulators is replaced with a glutamate residue. Importantly, APRR1 turned out to be identical to TOC1 (Matsushika et al. 2000, Strayer et al. 2000). More importantly, not only APRR1/TOC1 but also other APRRs are all subjected to circadian rhythms at the transcription level. Furthermore, each transcript of APRRs starts accumulating after dawn rhythmically and sequentially at approximately 2 h intervals in the order: APRR9→APRR7→APRR5→APRR3→APRR1/TOC1 (Matsushika et al. 2000). The wave of APRR9-mRNA is first boosted immediately after dawn. Consistent with this last fact, the expression of APRR9 is rapidly and transiently induced when dark-grown etiolated seedlings are exposed to a pulse of red light (or white light) (Makino et al. 2001). Based on these findings, we previously speculated that such sequential and rhythmic events, termed ‘circadian waves of the APRR1/TOC1 quintet’, might be a basis of the biological clock of Arabidopsis.
To gain further insight into such circadian-related properties of APRRs, here we constructed APRR1-overexpressing (APRR1-ox) transgenic plants, and then we characterized them to some extent, with special reference to certain circadian-related gene expression in Arabidopsis. Furthermore, we identified a novel protein that physically associates with APRR1. This protein is a putative transcription factor that resembles PIF3 (PHYTOCHROME INTERACTING FACTOR 3) (Ni et al. 1998, Martinez-Garcia et al. 2000). This novel protein was designated PIL1 (PIF3-LIKE 1), and its possible function was discussed in relation to other findings of this study.
Results
Constructing plants carrying a 35S::APPR1 transgene
According to a conventional Agrobacterium-mediated DNA delivery method, we established independent homozygous transgenic lines (T3 seeds) carrying a 35S::APRR1 transgene, in which the coding region of APRR1-cDNA was placed under the control of the Cauliflower Mosaic Virus (CaMV) 35S promoter so as to be constitutively expressed in plants (tentatively referred to as APRR1-overexpressing plants, APRR1-ox). Two representatives of these putative APRR1-ox transgenic lines (L3 and L5) were characterized. Fig. 1A shows typical seedlings. As compared with wild-type plants (Columbia, Col.), APRR1-ox seedlings exhibited alterations in their petiole lengths. They showed significantly shorter petioles than wild-type plants (upper panel, plants were grown for 16 d under 8 h light/16 h dark). Essentially the same phenotypes were seen for APRR1-ox plants grown under 16 h light/8 h dark (lower panels, plants were grown for 22 d). These morphological alterations appear to be due to developmental regulation rather than a general growth defect, because both the numbers and sizes of leaves, except for petiole lengths, were not significantly different between these transgenic and wild-type plants. The numbers and sizes of roots were also indistinguishable between them (see upper panel). For other transgenic lines, however, the severity of phenotypes varied from one line to another (data not shown). Therefore, clarification of phenotypes of APRR1-ox plants must await more statistic analyses, particularly with regard to circadian-regulated biological events, such as stomatal opening and flowering. Meanwhile, we wanted to address more specific issues for these putative APRR1-ox plants, with special reference to certain circadian-regulated genes, including the APRR1/TOC1 family of genes themselves.
These transgenic plants (L3 and L5), together with-wild type plants (Col.), were grown under 16 h light/8 h dark for 20 d, and RNA samples were prepared at both dawn and dusk. They were analyzed by Northern hybridization with a probe specific for the APRR1 coding sequence (Fig. 1B). In wild-type plants expression of APRR1 peaked at dusk, and it was minimal at dawn, as expected (Makino et al. 2000). Under these conditions, two types of APRR1-transcript were detected for APRR1-ox transgenic plants. Judging from their sizes, it was assumed that the longer (or slow migrating) and minor transcript was derived from the endogenous APRR1 gene, and the shorter and major one from the transgene, because the transgene was designed so as to lack the 3′ non-coding sequence. Based on this assumption, the results suggested that the transgenic plants express a high and constitutive level of the transgenic APRR1-transcript, at both dawn and dusk.
We then raised a polyclonal antiserum that is cross-reactive with the C-terminal portion of the APRR1 protein (i.e. its CONSTANS-like domain). The transgenic plants (L3 and L5), together with wild-type plants (Col.), were grown under continuous light for 30 d. Cellular proteins were extracted from whole plants, and then they were analyzed by Western blotting with the anti-APRR1 antiserum (Fig. 1C). A major protein band was detected for one of the transgenic lines (L3), and a considerable amount of the same cross-reactive band was also seen for another (L5). The corresponding band was hardly detected for wild-type plants. Note that the calculated molecular weight of APRR1 is 69,195, which is coincident with the apparent molecular mass (72 kDa), estimated for the immunodetected band on the gel. We thus found that the constructed 35S::APRR1 transgenic plants show the anticipated nature in that they express a larger amount of APRR1 at the levels of both mRNA and protein, as compared with in wild-type plants. These transgenic plants were thus referred to as APRR1-overexpressing (APRR1-ox) plants. In the following studies, the homozygous L3 line of APRR1-ox plants (T3 seeds) was mainly used.
Overexpression and circadian rhythm of APRR1 in APRR1-ox plants
It was first necessary to examine APRR1-ox plants more extensively, particularly, with regard to the circadian rhythm of APRR1 itself. To this end, APRR1-ox plants (L3) were grown for 22 d under 16 h light/8 h dark (LD conditions). From these plants, RNA samples were prepared at appropriate intervals. Northern hybridization analyses with a probe specific for the APRR1 coding sequence were carried out (Fig. 2A). For wild-type plants, rhythmic expression of APRR1 was clearly seen with a peak at evening (see also lower panel). For APRR1-ox plants, however, a higher and constitutive level of the total APRR1-transcript was detected for all RNA samples prepared at any given time. An endogene-specific probe, corresponding to the 3′ non-coding sequence of APRR1, was then used to detect the endogenous APRR1-transcript in APRR1-ox plants (labeled by endo. in Fig. 2A). As far as the light/dark (LD) conditions were concerned, the results showed that the endogenous APRR1-transcript was expressed rhythmically in APRR1-ox plants. However, it should be emphasized that its apparent robustness was considerably less evident, as compared with the case of wild-type plants. In any case, it was concluded that APRR1-ox plants, used in this study, overproduce the total APRR1-transcript constitutively.
Next, free-running expression of the endogenous APRR1 in APRR1-ox was examined after plants were grown under continuous light (LL conditions) (Fig. 2B). Wild-type and APRR1-ox plants (L3) were grown for 20 d under 12 h light/12 h dark, and then they were transferred to LL conditions. RNA samples were prepared from these plants at appropriate intervals, and then Northern hybridization analyses were carried out with a probe specific for the 3′ non-coding sequence of APRR1 (Fig. 2B). For wild-type plants, free-running rhythmic expression of APRR1 was seen with a peak at subjective evening, as reported previously (Makino et al. 2000). For APRR1-ox plants, however, a constitutive level of the endogenous APRR1-transcript was detected for all RNA samples. Note also that essentially the same results were obtained for other RNA samples prepared during the next day (i.e. 48 h to 72 h) (data not shown). These results suggested that overexpression of APRR1 results in loss of rhythmicity of the endogenous APRR1, in a manner that its robust rhythm is abolished.
Free-running circadian rhythms of the APRR1/TOC1 family of genes are dampened in APRR1-ox plants
We then examined possible effects of APRR1-ox on expressions of other members of the ARPP1/TOC1 family of genes, namely, APRR9, APRR7, APRR5, and APRR3 (Fig. 3). The conditions used were essentially the same as those described above, except for the photoperiod conditions used (16 h light/8 h dark). After APRR1-ox plants were grown for 22 d, in one line of experiments the photoperiod conditions were maintained (LD conditions), whereas in the other line of experiments plants were transferred to continuous light (LL conditions). From the leaves of these plants, RNA samples were prepared, followed by Northern hybridization analyses, using the probes specific for APRR9, APRR7, APRR5, APRR3, and UBQ10 (loading reference) (Fig. 3A), and then the hybridized bands were quantified (Fig. 3B). It should be first emphasized that the APRR9-transcript was hardly detected in RNA samples prepared at any given time, in either LD or LL conditions. By contrast, rhythmic expressions were observed for APRR7, APRR5, and APRR3 in APRR1-ox plants grown under LD conditions. Under LL conditions, however, such sequential waves of these transcripts were no longer evident. In other words, each expression of ARPP7, APRR5, and APRR3 became arhythmic, or, their rhythmicities were dampened (Fig. 3A and 3B).
In APRR1-ox plants, it was thus suggested that the expression of APRR9 is severely repressed under both LD and LL conditions. Since this observation is particularly crucial, we further examined this event in APRR1-ox plants grown under different conditions (Fig. 3C). Plants were grown for 20 d under conditions of 12 h light/12 h dark (LD conditions), and then plants were transferred to continuous light (LL conditions). RNA samples were prepared from the leaves of these plants at intervals, followed by Northern hybridization analyses. In wild-type plants, rhythmic and free-running expression of APRR9 with a peak at dawn was observed, as expected (Makino et al. 2000). In APRR1-ox plants, however, the APRR9-transcript was hardly detected at any time. This result is consistent with that seen in Fig. 3A. Taken together, for APRR1-ox plants, it was demonstrated that the expression of APRR9 is severely repressed under both LD and LL conditions, whereas the free-running and rhythmic waves of others (APRR7, APRR5, and APRR3) are dampened, particularly, under LL conditions. From these results, one can at least envisage that there are intimate linkages between the expression of APRR1 and those of other members of the APRR1/TOC1 family of genes (APRR9, APRR7, APRR5, and APRR3).
Circadian rhythm of CAB2 is dampened in APRR1-ox plants
A well-known hallmark of clock-regulated events in Arabidopsis is the free-running and rhythmic expression of CAB2, which is presumably the most downstream target of circadian output pathways. This crucial circadian-regulated event was examined in APRR1-ox plants (Fig. 4), after plants were grown under the same conditions as those described above (see Fig. 3A). In APRR1-ox plants grown under 16 h light/8 h dark, expression of CAB2 showed robust rhythmicity with a peak around noon (Fig. 4), as has been reported for wild-type plants (Piechulla 1999). This rhythmic profile was rapidly dampened when APRR1-ox plants were transferred to continuous light (LL) (note that the anticipated 2nd and 3rd peaks disappeared). Thus, the circadian rhythm of the circadian-regulated CAB2 gene was dampened in APRR1-ox plants, suggesting that overexpression of APRR1 results in a defect of the free-running clock function in Arabidopsis.
Circadian rhythms of several clock-associated genes are also altered in APRR1-ox plants
As mentioned in the Introduction, several genes have been proposed to encode such proteins that function within, or close to, the central oscillator(s). They include CCA1/LHY encoding putative clock components, GI encoding a modulator of flowering time, and CCR2 encoding a component of a sub- (or slave) oscillator. It should be noted that it is well known that the accumulations of the respective transcripts themselves are also oscillated under free-running conditions (for a review see Barak et al. 2000). These events are assumed to be crucial for their circadian-associated functions. Thus, it is of interest to examine such oscillated profiles of these circadian-associated genes in APRR1-ox plant. The results are shown in Fig. 5.
In wild-type plants, both CCA1 and LHY showed each typical circadian rhythm with a peak at subjective dawn under LL conditions, as anticipated (Wang et al. 1997, Schaffer et al. 1998). In APRR1-ox plants, however, the amplitude of the peak of CCA1 decreased significantly. Furthermore, the CCA1-transcript peaked at a significantly delayed timing in APRR1-ox plants (Fig. 5A). In other words, in this particular experiment with APRR1-ox plants under LL conditions, the 1st peak of CCA1 appeared about 5 h later than that under LD conditions, and the re-appearance of 2nd peak was delayed further. The estimated period between the 1st and 2nd peaks of CCA1 in APRR1-ox plants was approximately 30 h, considerably longer than anticipated. A similar event was observed for LHY in APRR1-ox plants under LL conditions (Fig. 5B).
Another intriguing event was seen in the case of GI (Fig. 5C). In wild-type plants, the GI-transcript peaked rhythmically around subjective noon under LL conditions, as reported previously (Fowler et al. 1999). In APRR1-ox plants, however, the GI-transcript was hardly detected at any time. It may be noted that this event is similar to that observed for APRR9 in APRR1-ox plants (see Fig. 3). CCR2 was also examined (Fig. 5D). In wild-type plants, the CCR2-transcript peaked rhythmically around subjective evening, as also reported previously (for a review see Staiger and Heintzen 1999). This robust circadian rhythm of CCR2 was no longer evident in APRR1-ox. The rhythm of CCR2 was dampened under LL conditions, and the CCR2-transcript was more or less constitutively detected at a lower level in APRR1-ox plants under these conditions.
In summary, the circadian rhythms of certain potential clock components that function within, or close to, the central oscillator were severely disordered in APRR1-ox plants. The amplitudes and periods of the rhythmic profiles of CCA1 and LHY were markedly altered. GI was severely depressed, whereas the rhythmic expression of CCR2 was rapidly dampened.
APRR1 physically associates with a PIF3-like protein
The above results are fully compatible with the idea that APRR1 plays an important role for an as yet unknown mechanism underlying generation of circadian rhythmicity in Arabidopsis. To gain further insight into the molecular function of APRR1, finally, we carried out yeast two-hybrid screenings to search for Arabidopsis proteins that physically interact with the APRR1 protein. The entire APRR1-cDNA sequence was connected to the GAL4 DNA-binding domain of pGBT9, to yield pGBT9-APRR1 (bait vector). On the other hand, an Arabidopsis cDNA-expression library was obtained, in which cDNAs were connected to the GAL4 activation domain in pGAD424 (prey library). Among 107 yeast transformants thus screened, five positive clones were obtained. Each of these was assumed to carry an Arabidopsis cDNA, protein product of which might interact with the APRR1 protein. Analyses involving restriction endonuclease digestion of these cloned DNAs showed that they were classified into three groups [clone-A (two clones), clone-B (two clones), and clone-C (one clone)] (Fig. 6A). Sequencing of each cDNA revealed that they were all derived from the same gene, which has been registered in the GenBank database (gene At2g46970, accession AC004411, chromosome II). In clone-B and clone-C, 5′-proximal regions (or N-terminal portions) had been truncated. In any case, the results of these two-hybrid assays showed that APRR1 might be able to interact with this particular gene product. Sequencing of the entire cDNA of this gene showed that it encodes a protein of 416 amino acids (Fig. 6B). This inferred protein contains a typical Myc-related basic helix-loop-helix (bHLH) motif, which is commonly found in some transcription factors of plants (The Arabidopsis Genome Initiative 2000). Interestingly, the amino acid sequence of this putative bHLH domain is highly similar to that found in the well-characterized Arabidopsis bHLH protein named PIF3 (PHYTOCHROME INTERACTING FACTOR 3) (Fig. 6C). PIF3 has been identified as the protein that specifically associates with red light-activated PHYB (Ni et al. 1998, Martinez-Garcia et al. 2000). Thus, the identified protein that might be able to associate with the APRR1 protein was hereafter designated PIL1 (PIF3-LIKE 1).
To verify the above notion, two-hybrid analyses were further carried out (Fig. 7A), in which PIF3 and APRR9 were also examined as alternative bait and prey, respectively. The results showed that the C-terminal domain of APRR1, lacking a pseudo-receiver domain, has a stronger ability to interact with PIL1. It should be emphasized that APRR1 showed the ability to interact with PIF3 as well. No interaction between APRR9 and PIL1 was observed in this yeast two-hybrid assay. It was thus suggested that APRR1 can physically associate with the putative Myc-related bHLH transcription factors, PIL1, and also PIF3. This was verified by an in vitro protein–protein interaction assay (Fig. 7B). In Schizosaccharomyces pombe, the C-terminal portion of APRR1 (named APRR1-C, see Fig. 6A), fused to a His-tag and Myc-epitope region, was expressed. In Escherichia coli, on the other hand, the C-terminal portion of PIL1 (named PIL1-C, see Fig. 6A, clone-C), fused to glutathione S-transferase (GST), was expressed. With Ni-beads carrying the (His)6-Myc:APRR1-C polypeptide, in vitro pull-down assays were carried out against E. coli cellular proteins containing the GST:PIL1-C polypeptide. The in vitro results supported the view that APRR1 can physically associate with PIL1 (Fig. 6B).
Discussion
Clarification of plant biological clocks is a paradigm of the current Arabidopsis molecular biology (see Introduction). In this respect, the rhythmic waves of the APRR1/TOC1 quintet are intriguing, as addressed above. In this study, certain circadian-regulated genes were extensively examined in Arabidopsis APRR1-ox transgenic plants, mainly, by determining the level of each transcript by means of Northern hybridization analyses. The level of transcript, determined by Northern hybridization, generally does not necessarily give us a direct idea about the level of a given protein that eventually plays its own physiological role. Also, a given circadian profile consists of amplitude, periodic length, and phase angle. However, Northern hybridization mainly used in this study does not allow us to precisely explain the results of this study with special reference to each rhythm component. Nevertheless, the results of this study provided us with much insight into the possible interplay of potential circadian components in Arabidopsis, as addressed below.
Our results with APRR1-ox plants are fully compatible with the current idea that APRR1 (TOC1) plays a crucial role within, or close to, the Arabidopsis central oscillator. The following findings of this study supported this current idea. (1) The free-running and robust circadian rhythms of the APRR1/TOC1 family of genes, including APRR1 itself, are dampened in APRR1-ox plants. (2) A hallmark of clock-regulated circadian rhythms is the free-running rhythmicity of CAB2, a downstream target of the output pathways. This robust rhythmicity of CAB2 is also dampened in APRR1-ox plants. (3) Circadian profiles of some clock-associated genes are also significantly altered in APRR1-ox plants. These genes include CCA1/LHY, GI, and CCR2. The effect on CCA1 and LHY are particularly interesting, because these proteins are the potential components of the central oscillator (for a review, see Barak et al. 2000). Namely, expressions of CCA1 and LHY seem to keep their rhythmicities, but their robustness and period are markedly altered, resulting in a longer period in APRR1-ox plants under LL conditions. In the accompanying paper (Matsushika et al. 2002), in fact, we showed also that constitutive overexpression of CCA1 in plants dampened the free-running rhythmic expression of APRR1 (and other APRRs), and vice versa. These results support the view that there are intimate interactions between APRR1 and certain circadian-associated components, including CCA1 and LHY.
In APRR-ox plants, the APRR9-transcript was hardly detected in RNA samples prepared at any time, irrespective of LD or LL conditions. Otherwise, it was suggested that constitutive expression of APRR1 does not significantly affect the expression profiles of other APRR1/TOC1 family members (APRR7, APRR5, and APRR3), under LD conditions. However, it should be emphasized that their free-running circadian waves are severely dampened under LL conditions in APRR1-ox plants. This suggests that the seemingly sequential events of the APRR9→APRR7→APRR5→APRR3→APRR1/TOC1 waves may not be such a simple unidirectional linear pathway. Rather, they may form interlocking loops more complex than we originally thought (Matsushika et al. 2000), in such a manner that APRR1 and APRR9 are more closely linked, as discussed further below.
One of the most intriguing findings of this study is that constitutive expression of APRR1 results in a severe depression of APRR9. The fact that APRR9 is induced by white light (or red light pulse) previously led us to speculate that the APRR9 protein might act at an interface between the light signal input pathways and the oscillator components (Matsushika et al. 2000). In wild-type plants the APRR9-transcript was rapidly accumulated when dark-grown young seedlings were exposed to light, it peaks within 2 h, and has decreased by 6 h (Makino et al. 2001), whereas the level of the APRR9-transcript was dramatically reduced on a phyB background (Matsushika et al. unpublished result). Such a light-dependent induction of APRR9 was also no longer seen in APRR1-ox (Matsushika et al. unpublished result). These events are indicative of a negative APRR1→APRR9 regulation. It is thus of interest to know how APRR1 and PHYB (or PHYA) regulate the expression of APRR9. In this connection, another intriguing finding of this study is that APRR1 can interact with a Myc-related bHLH factor, named PIL1, which is similar to PIF3. APRR1 appears to bind to PIF3 as well. The PIF3 gene has been identified as the one encoding a phytochrome-interacting transcription factor. In plants with reduced levels of PIF3, in fact, there is a decrease in red-light induction of CCA1 and LHY (Martinez-Garcia et al. 2000). Taking these together, one can assume that PIL1 (and/or PIF3) might also be involved in the phytochrome-dependent induction of APRR9 by light, and that APRR1 negatively affects the expression of APRR9 through its interaction with PIL1 (and/or PIF3). Although these views are speculative at present, they are worth being tested because the findings of this study addressed above are best explained by such a hypothesis.
It is premature to discuss in details about the molecular bases of the Arabidopsis central oscillator(s) solely based on the results of this study. However, we have provided several insights into the importance of the APRR1/TOC1 circadian waves. These views will provide at least a preliminary framework for better understanding the molecular mechanisms underlying the circadian rhythms in Arabidopsis.
Materials and Methods
Plant growth conditions and related materials
Arabidopsis thaliana (Columbia, hereafter designated Col.) was mainly used as wild-type plants. Seeds were imbibed and cold treated at 4°C for 2 d before germination and growth at 22°C. To grow plants in the dark, the imbibed seeds were exposed to white light for 30 min before growth in the dark. Otherwise, plants were grown in a chamber with conditions of light from fluorescent lights (150–200 µmol m–2 s–1) at 22°C on soil, or on agar-plates containing MS salts and 2% sucrose, as described previously (Taniguchi et al. 1998). Light/dark conditions used were either 16 h light/8 h dark (long-day conditions), 12 h light/12 h dark, or 8 h light/16 h dark (short-day conditions), as specifically noted for each experiment in the text.
Preparing RNA, and Northern blotting
Total RNA was isolated from appropriate organs (mainly leaves) of Arabidopsis by the phenol-sodium dodecyl sulfate (SDS) method (Taniguchi et al. 1998). For Northern hybridization, RNA was separated in agarose gels (1%) containing 2.2 M formaldehyde, then transferred to Hybond-N+ membranes. The fixed membranes were hybridized with 32P-labeled DNA fragments in 6× standard saline phosphate and EDTA (1× SSPE = 0.18 M NaCl, 10 mM phosphate buffer, 1 mM EDTA, pH 7.4), 5× Denhardt’s solution, and 0.5% SDS containing 10% dextran sulfate and 100 µg ml–1 salmon sperm DNA, at 65°C for 18 h. The membranes were washed twice with 2× SSPE and 0.5% SDS for 15 min at room temperature, twice with 2× SSPE and 0.5% SDS for 30 min at 65°C, and then with 0.2× SSPE and 0.5% SDS for 15 min at 65°C. The washed membranes were exposed and analyzed on a phosphoimage analyzer (BAS-2500II) (FujiXerox, Tokyo, Japan).
PCR amplification
Using appropriate pairs of primers, if necessary, polymerase chain reaction (PCR) was carried out to prepare DNA segments. The standard conditions were primarily 94°C for 1 min, 55°C for 2 min, and 72°C for 3 min, 25 cycles with Thermal Cycler 480 (Takara Shuzo, Kyoto, Japan). A PCR kit was used, according to the supplier’s instructions (Takara Shuzo, Kyoto, Japan).
Probes for Northern blotting
Several double-stranded 32P-labeled DNA probes were used to detect each specific mRNA. The probes used were amplified by PCR with appropriate sets of primers. They were designed appropriately, according to the previous reports, including UBQ10 (Taniguchi et al. 1998), CCA1, GI, CCR2 (Wang and Tobin 1998), LHY (Schaffer et al. 1998), CAB2 (Brusslan and Tobin 1992), APRR1, APRR3, APRR5, APRR7, and APRR9 (Matsushika et al. 2000). They include: APRR9 (5′-CAAGATTCGATAAAAATGGTGTTG and 5′-TCTCTGGCTTCTACTCCAACG), APRR7 (5′-ATGTCATCTCATGACTCAATGGGG and 5′-AGAGATCTTATTTTCATCCGC), APRR5 (5′-ACACAGGACTCGGTGAATACTGTG and 5′-TTCTCGTTGAAGAGATTGCTG), APRR3 (5′-TCACATGATTCTATGGTTCTGGTC and 5′-TTGTGCCCATCGGTCACTTCC), APRR1-coding (5′-TCGAGGCAAGACGAAGTCCCTGTC and 5′-TCTTCTGTCAAGTTTATTTACCC), APRR1-3′-non-coding (5′-AGATACACCAAGAACTGAAAACCGTTG and 5′- CAAGATTCAAGTAAATGATACTGGAAGG), CCA1 (5′-GGCCTAAGCGTAAACCAAACAATCC and 5′-GGGATCTTTCTGTTCCACATGAATG), LHY (5′-TGGACATAGAAATTCCGCCTCCTCG and 5′-GCTTTTGAAATTAGGAGCCAATGGC), GI (5′-GAGTTTTACCACAAGATCAGCGACC and 5′-GATGTCACCGCTGTATCAAAGCATG), CCR2 (5′-GCTCTTGAGACTGCCTTCGCTC and 5′-CTCGTTAACAGTGATGCTACGG), CAB2 (5′-GGGAATTCAAATCCAATGAGTAGAGAGA and 5′-GGGGATCCTCACACGGCCGCTTCCGAGGA). Each 32P-labeled probe was prepared with a random-labeling kit (Takara Shuzo, Kyoto, Japan).
Constructing APRR1-ox transgenic plants
The entire coding sequence of the APRR1 cDNA (Makino et al. 2000) was cloned by PCR-aided cloning into the pSK1 vector (Kojima et al. 1999) at the XbaI site downstream of the CaMV 35S promoter, to yield pSK1-APRR1. This construct was transformed into Agrobacterium tumefaciens strain EHA101, and then Arabidopsis plants (Columbia) was transformed by vacuum infiltration procedures, as described previously (Bechtold et al. 1993). Only transgenic lines segregating the hygromycin resistance as a single locus were used in further analysis, after establishing homozygous T3 seeds.
Yeast two-hybrid screening
A kit for two-hybrid analysis (MATCHMAKER™, Clontech) was obtained through TOYOBO Co. This kit contained all tools essential for two-hybrid screening, including the vectors: pGBT9 providing the GAL4 DNA-binding domain (TRP1 marker), and pGAD424 providing the GAL4 activation domain (LEU2 marker). The kit also included the yeast host strains: HF7c carrying both the GAL1-HIS3 and (GAL4 17-mers)3 CYC1-lacZ reporters, and SFY526 carrying the GAL1-lacZ reporter. Two-hybrid screening was carried out essentially according to the manual accompanying the kit. An Arabidopsis cDNA-expression library was obtained from Clontech, namely, A. thaliana MATCHMAKER cDNA Library. cDNAs were prepared from mRNA of 3-week-old green vegetative tissue of Arabidopsis (Columbia). The following plasmid were constructed from pGBT9 for baits, namely, pGBT9-APRR1, pGBT9-APRR1-R, pGBT9-APRR1-C, and pGBT9-APRR9. The plasmid pGAD424-PIF3 was constructed from pGAD424 for preys. To construct these plasmids, we designed each appropriate set of PCR-primers for PCR-aided cloning, and the nucleotide sequences of the resulting plasmids were confirmed by sequencing. Plasmids, pGAD10-PIL1-A, pGAD10-PIL1-B, pGAD10-PIL1-C were isolated by screening (see Fig. 6).
Acknowledgments
This study was supported by a Grant-in-Aid for scientific research on a priority area (12142201 to TM) from the Ministry of Education, Science, Sports and Culture of Japan. We thank A. Nagatani (Kyoto University) for the kind gifts (phyA and phyB mutants).
Corresponding author: E-mail: tmizuno@agr.nagoya-u.ac.jp; Fax, +81-52-789-4091.
References
Anderson, S.L. and Kay, S.A. (
Barak, S., Tobin, E.M., Andronis, C., Sugano, S. and Green, R.M. (
Bechtold, N., Ellis, J. and Pelletier, G. (
Brusslan, J.A. and Tobin, E.M. (
Carre, I.A. (
Covington, M.F., Panda, S., Liu, X.L., Strayer, C.A., Wagner, D.R. and Kay, S.A. (
Fowler, S. Lee, K., Onouchi, H., Samach, A., Richardson, K., Morris, B., Coupland, G. and Putterill, J. (
Garner, W.W. and Allard, H.A. (
Green, R.M. and Tobin, E.M. (
Harmer, S.L., Hogenesch, J.B., Straume, M., Chang, H.-S., Han, B., Zhu, T., Wang, X., Kreps, J.A. and Kay, S.A. (
Heintzen, C., Nater, M., Apel, K. and Staiger, D. (
Hicks, K.A., Millar, A.J., Carre, I.A., Somers, D.E., Straume, M., Meeks-Wagner, D.R. and Kay, S.A. (
Hicks, K.A., Albertson, T.M. Wagner, D.R. (
Huq, E., Tepperman, J.M. and Quail, P.H. (
Imamura, A., Hanaki, N., Nakamura, A., Suzuki, T., Taniguchi, M., Kiba, T., Ueguchi, C., Sugiyama, T. and Mizuno, T. (
Kojima, S., Banno, H., Yoshioka, Y., Oka, A., Machida, C. and Machia, Y. (
Koornneef, M., Alonso-Blanco, C., Peeters, A.J.M. and Soppe, W. (
Kreps, J.A. and Kay, S.A. (
Kreps, J.A. and Simon, A.E. (
Liu, X.L., Covington, M.F., Fankhouser, C., Chory, J. and Wagner, D.R. (
Makino, S., Kiba, T., Imamura, A., Hanaki, N., Nakamura, A., Suzuki, T., Taniguchi, M., Ueguchi, C., Sugiyama, T. and Mizuno, T. (
Makino, S., Matsushika, A., Kojima, M., Oda, Y. and Mizuno, T. (
Martinez-Garcia, J.F., Huq, E., and Quail, P.H. (
Matsushika, A., Makino, S., Kojima, M. and Mizuno, T. (
Matsushika, A., Makino, S., Kojima, M., Yamashino, T. and Mizuno, T. (
Millar, A.J., Carre, I.A., Strayer, C.A., Chua, N.H. and Kay, S.A. (
Nelson, D.C. Lasswell, J. Rogg, L.E., Cohen, M.A. and Bartel, B. (
Ni, M., Tepperman, J.M. and Quail, P.H. (
Park, D.H., Somers, E.D., Kim, Y.S., Choy, Y.H., Lim, H.K., Soh, M.S., Kim, H.J., Kay, S.A. and Nam, H.G. (
Piechulla, B. (
Reed, J.W., Nagpal, P., Bastow, R.M., Solomon, K.S., Dowson-Day, M.J., Elumalai, R.P. and Millar, A.J. (
Robert, L.S., Robsons, F., Sharpe, A., Lydiate, D. and Coupland, G. (
Samach, A. and Coupland, G. (
Schaffer, R., Ramsay, N., Samach, A., Corden, S., Putterill, J., Carre, I.A. and Coupland, G. (
Schaffer, R., Landgraf, J., Accerbi, M., Simon, V.V., Larson, M. and Wisman, E. (
Somers, D.E., Devlin, P.F. and Kay, S.A. (
Somers, D.E., Webb, A.A.R., Pearson, M. and Kay, S.A. (
Somers, D.E., Schultz, T.F., Milnamow, M., and Kay, S.A. (
Staiger, D. and Heintzen, C. (
Strayer, C., Oyama, T., Schultz, T.F., Raman, R. Somers, D.E., Mas, P., Panda, S., Kreps, J.A. and Kay, S.A. (
Taniguchi, M., Kiba, T., Sakakibara, H., Ueguchi, C., Mizuno, T. and Sugiyama, T. (
The Arabidopsis Genome Initiative (
Wang, Z.-Y., Kenigsbuch, D., Sun, L., Harel, E., Ong, M.S. and Tobin, E.M. (
Wang, Z.-Y. and Tobin, E.M. (