SEARCH

SEARCH BY CITATION

Keywords:

  • flowering;
  • FLC;
  • DNA binding;
  • protein complexes;
  • SOC1;
  • FT

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

The Arabidopsis Flowering Locus C (FLC) protein is a repressor of flowering regulated by genes in the autonomous and vernalization pathways. Previous genetic and transgenic data have suggested that FLC acts by repressing expression of the floral integrator genes SOC1 and FT. We have taken an in vivo approach to determine whether the FLC protein interacts directly with potential DNA targets. Using chromatin immunoprecipitation, we have shown that FLC binds to a region of the first intron of FT that contains a putative CArG box, and have confirmed that FLC binds to a CArG box in the promoter of the SOC1 gene. MADS box proteins are thought to bind their DNA targets as dimers or higher-order multimers. We have shown that FLC is a component of a multimeric protein complex in vivo and that more than one FLC polypeptides can be present in the complex.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

The switch from vegetative to floral growth in plants is controlled by a number of developmental and environmental signals. In Arabidopsis, the minichrome maintenance 1, agamous, deficiens, serum response factor (MADS) box transcription factor Flowering Locus C (FLC) is a repressor of flowering (Michaels and Amasino, 1999; Sheldon et al., 1999); its expression is controlled by loci in the autonomous pathway and by vernalization, the promotion of flowering by an extended cold period (Sheldon et al., 2000). Ecotypes with high levels of FLC expression are generally late-flowering (Shindo et al., 2005) and responsive to vernalization, whereas ecotypes with low FLC expression are early-flowering and have little or no response to vernalization.

MADS box transcription factors are a large family with 107 members in Arabidopsis (Pařenicováet al., 2003); these proteins bind to a loose consensus sequence known as a CArG box (reviewed in Shore and Sharrocks, 1995), which has a typical core motif CC(A/T)6GG. In vitro binding site selection methods showed that the MADS box protein AGAMOUS bound to a consensus CC(A/T)4NNGG motif with the CC and GG dinucleotides conserved (Huang et al., 1993; Shiraishi et al., 1993). Studies utilizing reporter gene fusions have demonstrated that CArG boxes in the promoter of the AP3 gene are involved in the regulation of expression of that gene and are binding sites for the MADS box proteins PI and AP3 (Tilly et al., 1998). The CArG boxes bound by PI and AP3 differ from the consensus CC(A/T)6GG in that the CC and GG dinucleotides are not absolutely conserved.

The genes FT, SOC1 and LFY act as the integrators of floral signaling pathways responding to light, cold, gibberellins and the developmental state of the plant. Expression of the FT and SOC1 genes is repressed in plants with high FLC expression (Lee et al., 2000; Michaels et al., 2005), so both these genes are potentially direct targets of FLC repression. Hepworth et al. (2002) identified a CArG box sequence in the promoter of SOC1. An oligonucleotide containing the core of this CArG box, CCAAAATAAG, is specifically bound by bacterially expressed FLC protein in vitro. A 300 or 1000 bp SOC1 promoter fragment containing this CArG motif directs GUS reporter gene expression at a lower level in 35S::FLC plants than in wild-type Landsberg erecta (Ler) plants which have low endogenous FLC expression. When the CArG box is mutated in otherwise identical constructs, FLC-dependent repression is largely lost. These data show that the CArG box can mediate FLC-dependent repression of SOC1, but this could be indirect, with FLC regulating another MADS box protein that binds to the SOC1 CArG box. There are no data to show that FLC can bind to CArG boxes in genes other than SOC1.

MADS box proteins are thought to bind CArG boxes as complexes. A large body of work, mainly from yeast two-hybrid and in vitro immunoprecipitation experiments (summarized in de Folter et al., 2005), has demonstrated many interactions between MADS box proteins. These data show that MADS box proteins can form hetero- or homodimers (e.g. Riechmann et al., 1996) and may also form tetrameric complexes, such as those proposed to involve combinations of PI, AP1, AP3 and SEP3 (Honma and Goto, 2001). These interactions, demonstrated in vitro, are likely to be of biological significance as they are consistent with mutant phenotypes and expression patterns of the MADS box proteins involved. A yeast two-hybrid study screened all possible dimers between most members of the MADS protein family in Arabidopsis (de Folter et al., 2005). This study identified a large number of potential homodimers and heterodimers among members of the MADS box protein family, but no interactions of the FLC protein with itself or another MADS box protein has been demonstrated.

In this paper, we show a direct interaction of FLC with the SOC1 and FT genes in vivo, and that FLC is present in higher-order complexes that can contain more than one FLC protein.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

We used chromatin immunoprecipitation (ChIP) assays to determine DNA binding sites for the FLC protein in vivo. ChIP has previously identified two target genes of the AGL15 MADS box protein in Arabidopsis that contain putative CArG boxes in their promoters (Wang et al., 2002). In ChIP assays, cross-linked chromatin is extracted and sheared (usually by sonication) to short fragments and the protein of interest is immunoprecipitated along with cross-linked DNA. DNA fragments that are associated with the immunoprecipitated protein are enriched in the immunoprecipitate.

A construct was made to constitutively express the FLC protein with a C-terminal 3xFLAG epitope tag driven by the CaMV 35S promoter (35S::FLC-FLAG; Figure 1). The 3xFLAG tag is relatively small (22 amino acids) and is recognized by a commercially available monoclonal antibody (Brizzard et al., 1994) that has been successfully used in immunoprecipitation experiments (e.g. Lee et al., 2004). The 35S::FLC-FLAG construct was introduced into the L.er ecotype of Arabidopsis; this ecotype has low FLC expression because of inactive alleles of both FRI and FLC (Koornneef et al., 1994; Lee et al., 1994). Of 10 T1 plants, four were significantly late-flowering in the T1 generation. These four lines were also late-flowering in the T2 generation (Figure 1c); this agrees with previous studies where FLC has been expressed in the Ler background under the control of the 35S promoter (Hepworth et al., 2002; Sheldon et al., 1999). We also observed ectopic inflorescences inside some carpels in lines 6 and 9, similar to those reported by Hepworth et al. (2002); no other morphological changes were observed. Line 9 was used in subsequent ChIP experiments and is referred to as 35S::FLAG-FLC. Western blots using the anti-FLAG M2 antibody in the 35S::FLC-FLAG line identified a protein migrating at approximately 33 kDa, consistent with a predicted mass for the FLC-FLAG protein of 26 kDa (Figure 2a). A protein of the same molecular weight was also detected by FLC antiserum (Figure 2b); this protein is present at high abundance compared with the endogenous FLC in the C24 ecotype. We conclude that the 35S::FLC-FLAG line over-expresses the predicted fusion protein, and that this protein is able to act to repress flowering in the same manner as FLC.

image

Figure 1. Constitutive expression of FLC-FLAG protein. (a) Diagram of the 35S::FLC-FLAG construct showing location of GatewayTMattB1 and attB2 sites. The construct is in a pBI121 backbone. 35S, CaMV 35S promoter; Tnos, nopaline synthase terminator; FLAG, 3xFLAG. (b) Additional amino acid residues added to the C-terminus of the FLC protein by the attB2 and 3xFLAG sequences, 3xFLAG is underlined. (c) Leaf number at flowering of wild-type Landsberg erecta and T2 plants from four 35S::FLC-FLAG lines; error bars are the standard deviation.

Download figure to PowerPoint

image

Figure 2. Western blot analysis of (a) total protein extracted from Landsberg erecta and the 35S::FLC-FLAG line probed with anti-FLAG M2 antibody followed by antimouse-HRP, (b) total protein extracts from C24 and 35S::FLC-FLAG probed with FLC antiserum, (c) protein immunoprecipitated by the anti-FLAG chromatin immunoprecipitation method probed with anti-FLAG M2-HRP, and (d) total protein extracts from C24 and flc20; duplicate blots probed with FLC antiserum and affinity-purified FLC antibodies as indicated. Numbers indicate molecular marker mass in kDa.

Download figure to PowerPoint

FLC interacts directly with the CArG box of SOC1

Chromatin immunoprecipitation using anti-FLAG M2 antibody on the 35S::FLC-FLAG line immunoprecipitated the FLC-FLAG protein (Figure 2c). The DNA immunoprecipitated with the FLC-FLAG protein was analyzed by real-time PCR using primers that anneal either side of the SOC1 CArG box (Figure 3a). This demonstrated that the SOC1 CArG box region was 14-fold enriched compared with control primers from a region of the unlinked AMP1 gene (Helliwell et al., 2001) in the immunoprecipitate from the 35S::FLC-FLAG line. The SOC1 CArG box was present at the same concentration as the AMP1 region in the immunoprecipitate from non-transformed Ler (Figure 3a).

image

Figure 3. Chromatin immunoprecipitation analysis of FLC binding to the SOC1 CArG box. (a) Fold enrichment over background (measured using primers against a region of the AMP1 gene) of the SOC1 CArG box in ChIPs with anti-FLAG M2 (Ler and 35S::FLC-FLAG) or anti-FLC. (b) Fold enrichment of the endogenous SOC1 CArG box and the CArG box in the 300 bp CArG::GUS and ΔCArG::GUS constructs in ChIPs with anti-FLC. Specific primer pairs discriminate between the endogenous and transgene CArG boxes. ChIP assays were carried out in triplicate; error bars are the standard error of the mean.

Download figure to PowerPoint

To confirm that the endogenous FLC protein can interact directly with the SOC1 CArG box, we affinity-purified FLC antibodies (from antiserum raised against the FLC protein without the conserved MADS domain; Sheldon et al., 2000) and used this to carry out chromatin immunoprecipitation in C24 (an ecotype with high FLC expression) and the loss of function flc20 mutant (a Ds insertion mutant in the C24 ecotype; Helliwell et al., 2002) grown under short-day (SD) conditions (Figure 3a). In C24 plants, the SOC1 CArG box region was enriched 19-fold compared with the AMP1 control region. In the flc20 mutant, a twofold enrichment of the SOC1 CArG box was found. The small enrichment of the SOC1 CArG box in the flc20 mutant may be due to the affinity-purified FLC antibodies interacting with epitopes in MAF (MADS AFFECTING FLOWERING) MADS box proteins which are related to FLC (Ratcliffe et al., 2001, 2003). Consistent with this, we found that the affinity-purified FLC antibodies detect proteins that migrate at a similar position to FLC in the flc20 mutant in Western blots (Figure 2d). This suggests that one or more of the MAF proteins can also interact with the SOC1 promoter.

Long-day photoperiod (LD), vernalization and gibberellins (GA) are all thought to promote flowering by induction of SOC1. FLC and SOC1 mRNA expression was measured in C24 plants grown in SD, LD, SD + GA and SD vernalized conditions. Consistent with previous reports using other ecotypes (Borner et al., 2000; Lee et al., 2000; Moon et al., 2003), there is a small increase in SOC1 mRNA in plants grown under LD and SD + GA compared with SD plants (Figure 4a). There was no significant change in the binding of FLC to the SOC1 CArG box as determined by ChIP using the FLC antibody in the SD + GA-treated plants relative to the SD plants, indicating that GA does not act on SOC1 expression by reducing FLC binding at the CArG box (Figure 3a). There was a slightly lower amount of FLC binding at the SOC1 CArG box in LD-grown plants than SD plants. In vernalized plants, a greatly reduced binding of FLC at the SOC1 CArG box was observed, consistent with the reduced level of FLC expression in these plants (Figure 4a).

image

Figure 4. Characterization of FLC targets. (a) RNA gel blots of FLC and SOC1 mRNA in C24 plants used for ChIP in Figure 3. A 10 μg aliquot of RNA was loaded in each lane. (b) ChIP analysis of FLC binding at the FT promoter in ChIPs from C24 and flc20 using anti-FLC and L.er and 35s::FLC-FLAG using anti-FLAG M2. Input determined by PCR with S-adenosyl methionine synthetase 2 (SAM) primers. (c) RT-PCR analysis of FT, AGL18 and SHP2 expression in Ler and 35S::FLC-FLAG.

Download figure to PowerPoint

To determine whether FLC is interacting with the predicted CArG box in the SOC1 promoter, we analyzed the interaction of FLC with the 300 bp CArG::GUS and 300 bp ΔCArG::GUS constructs in a Ler 35S::FLC background as used by Hepworth et al. (2002). Using specific PCR primers, we distinguished FLC binding to the endogenous SOC1 CArG box from that binding at the CArG or ΔCArG boxes of the GUS constructs. As the ChIP PCR analysis indicates the degree of enrichment over background, it was important to know the number of copies of the CArG transgenes present in the lines being analyzed as an increased copy number would lead to an apparent enrichment compared with the single-copy endogenous AMP1 gene. We used real-time PCR to determine the number of transgene copies, and identified a CArG::GUS line homozygous for a single-copy transgene and a ΔCArG::GUS line hemizygous for a single-copy transgene. In both lines we could detect FLC bound to the endogenous SOC1 CArG region (Figure 3b). Using primers specific for the CArG boxes in the transgene constructs, significant enrichment over background was seen for the CArG::GUS line. The ΔCArG::GUS line showed only a marginal enrichment over background, consistent with FLC binding being disrupted in this line. We conclude that the SOC1 CArG box is an in vivo binding site for the FLC protein.

FLC can bind other targets in the genome

To see whether FLC could bind to other targets in the Arabidopsis genome, we carried out further PCR analyses using a set of candidate CArG boxes. Putative CArG boxes can be identified in the promoters of many Arabidopsis genes. For example, a search with the sequence CCWWWWWWRG against the 1000 bp upstream of all annotated Arabidopsis genes gives 11 250 hits. A small subset of these putative CArG boxes were selected to test as putative FLC binding sites on the basis of having a sequence similar to the SOC1 CArG box or being in the promoter of a MADS box or flowering time gene. The sequences of these putative CArG boxes and their positions in the promoters are detailed in Table 1. ChIP samples from Ler and the Ler + 35S::FLC-FLAG line using anti-FLAG antibody were analyzed by real-time PCR (Figure 5a). The AMP1 primers showed that the background DNA in the ChIPs was the same for both samples (Figure 5a). Therefore, in real-time PCR reactions with primers amplifying putative CArG box regions, an increased signal in the ChIP from the 35S::FLC-FLAG line compared with Ler indicates an enrichment of that CArG box region. In the 35S::FLC-FLAG lines there was a significant enrichment of CArG box-containing regions from the AGL18 and SHP2 promoters and the first intron of FT (Figure 5a). A smaller enrichment was seen for a region in the promoter of AP1. To determine whether the binding sites identified by 35S::FLC-FLAG were also binding sites for endogenous FLC, the PCR analyses were repeated for the FT intron, AGL18 and SHP2 primers on ChIPs of C24 and flc20 (Figure 5b) using FLC antibodies. Enrichment was seen only for the FT intron CArG region and not for AGL18 or SHP2. The lower degree of enrichment of the FT intron region in ChIP for the endogenous FLC protein may reflect the lower expression level compared with the 35S-driven FLC-FLAG protein.

Table 1.  Putative CArG boxes tested. Both strands of boxes palindromic for CC and GG are shown
GeneAccession numberPutative CArG boxStrandPosition of 1st C from ATG
SOC1At2g45660TTTTCCAAAATAAGTAAATop−949
FT intronAt1g65480TTTTCCTTTTTTGGGGTATop+292
TACCCCAAAAAAGGAAAABottom+301
FT promoterAt1g65480TTTTCCATAATATGGCCGTop−1002
CCGGCCATATTATGGAAABottom−992
COAt5g15840AATGCCTTTTAAGTTTTATop−847
GTAACCATTATAGTAAGTBottom−917
SVPAt2g22540CTTTCCAAATTAGTGCATTop−1303
AGL18At3g57390GCCACCAAATATGGAAATop−290
TTTTCCATATTTGGTGGCBottom−281
AGL24At4g24540TAACCCTAATTAGAAATGTop−1173
ATACCCTTTTAAGTTTTABottom−1157
AAAACCAAAAAGGTAGATTop−1003
TCTACCTTTTTGGTTTTGBottom−995
AGL42At5g62165ATTTCCTATTTTAGCGGATop−1084
TTTGCCTATTTTGGCAAATop−1027
TTACCCTATTTTGGCAAABottom−1018
TGCACCAATTATAGTATTBottom−843
AGL71At5g51870TTCACCATTATTAAGCTTTTop−346
AGL72At5g51860GTTACCAAATTAAAGAAGATop−446
SHP2At2g42830ATTACCAAAAAAGGAAAGTop−755
CTTTCCTTTTTTGGTAATBottom−746
AP1At1g69120AAACCCAAAAATAGTATTBottom−796
image

Figure 5. Chromatin immunoprecipitation analysis of FLC binding at other putative CArG boxes. (a) ChIP with anti-FLAG M2 on Ler and 35S::FLC-FLAG. Real-time PCR carried out using relative quantification on equal inputs. Inputs were shown to be equal using the AMP control primer pair. (b) ChIP using anti-FLC on flc20 and C24. ChIP assays were carried out in triplicate; error bars are the standard error of the mean.

Download figure to PowerPoint

Expression of FT has previously been shown to be inhibited by high FLC expression (Michaels et al., 2005), the repression being mediated through the FT promoter. A putative CArG box is present in the promoter of FT. We were unable to design functional real-time PCR primers to test whether the region containing this putative CArG box is enriched in ChIPs. Instead we used conventional PCR followed by a DNA gel blot (Figure 4b). We did not detect an enrichment of the FT promoter region CArG box in ChIPs from either C24 compared with flc20 using FLC antibodies or in the 35S::FLC-FLAG line compared with Ler using FLAG antibody. This suggests that FLC does not interact directly with the FT promoter. However, measurement of FT expression by RT-PCR in Ler and the 35S::FLC-FLAG line (Figure 4c) showed that FT mRNA is lower in the 35S::FLC-FLAG plants compared with Ler, confirming that over-expression of FLC in this line is repressing expression of FT. AGL18 mRNA abundance is unchanged and SHP2 mRNA is slightly increased in the 35S::FLC-FLAG line (Figure 4c), suggesting that FLC is not acting as an inhibitor of these genes when bound to the respective promoters.

FLC-FLAG interacts with endogenous FLC

A number of Arabidopsis MADS box proteins have been shown to form homodimers in yeast two-hybrid experiments (e.g. de Folter et al., 2005), but there has been no evidence to show that FLC interacts with itself or other MADS box proteins. To test whether more than one FLC polypeptide can be present in the same complex in vivo, we carried out an immunoprecipitation experiment. As Ler expresses the endogenous FLC gene at very low levels, it was not possible to use the 35S::FLC-FLAG line to test for co-immunoprecipitation of endogenous FLC protein with the FLC-FLAG protein. A genomic FLC-FLAG construct (gFLC-FLAG) was constructed consisting of the entire FLC gene from 2026 bp upstream of the translation start fused to 3xFLAG at the stop codon position of FLC. This construct was used to transform the C24 ecotype, which expresses high levels of FLC. Plants transformed with gFLC-FLAG were later-flowering than wild-type C24, indicating that the FLC-FLAG protein was active; a line that flowered with at least 40 leaves compared with about 20 for C24 was analyzed further.

The gFLC-FLAG protein was immunoprecipitated using anti-FLAG M2 agarose (Figure 6a). In a comparison of the immunoprecipitates from C24 and gFLC-FLAG lines, two additional proteins were detected by FLC antiserum (but not anti-FLAG M2) in immunoprecipitates from the gFLC-FLAG line. These proteins were of the same molecular weight as the endogenous FLC proteins detected in the total extract from this line. This result shows that both FLC-FLAG and endogenous FLC polypeptides are present in the same complex.

image

Figure 6. FLC is part of a multiprotein complex. (a) Co-immunoprecipitation of endogenous FLC with FLC-FLAG. C24 expressing gFLC-FLAG and C24 wild-type were immunoprecipitated with anti-FLAG M2 agarose. Proteins were separated by SDS–PAGE and blotted. Blots were probed with anti-FLAG M2-HRP or FLC antiserum and antirabbit-HRP as indicated. Dots indicate the position of FLC polypeptides in the total extract. The FLC-FLAG protein in total extracts migrates slightly faster than in immunoprecipitates (IP). (b) SEC analysis of gFLC:FLAG. Size markers (kDa) are shown. (c) SEC analysis of 35S::FLC-FLAG.

Download figure to PowerPoint

FLC is part of a large protein complex

Size exclusion chromatography (SEC) was used to separate native protein extracts from C24 plants expressing the gFLC-FLAG construct on the basis of molecular weight. The protein fractions from the SEC were analyzed by Western blot to determine the SEC fractions in which FLC-FLAG was present (Figure 6b). The FLC-FLAG protein was only detected in high-molecular-weight fractions with a peak eluting at 600–800 kDa. Western blot analysis using the FLC antiserum on the same SEC fractions showed that the endogenous FLC protein is also present in the same high-molecular-weight fractions. This complex of approximately 800 kDa is larger than would be expected for a complex of two (approximately 50–60 kDa) or four (100–120 kDa) MADS box proteins.

Size exclusion chromatography analysis of Ler expressing the 35S::FLC-FLAG construct showed that the FLC-FLAG protein in these plants is also present in a similar high-molecular-weight complex (Figure 6c). The 35S::FLC-FLAG lines also contained FLC-FLAG protein eluting at 63–100 kDa as expected for dimers or tetramers of MADS box proteins. As FLC is only detected in the high-molecular-weight complex in the gFLC-FLAG lines, it is likely that this is the active form of FLC.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

We have used chromatin immunoprecipitation to demonstrate that the FLC protein binds in vivo to the CArG box in the SOC1 promoter, and that it binds to a region containing a putative CArG box in the first intron of FT. Both these interactions are consistent with FLC acting to repress expression of these two floral integrator genes. The CArG box we have identified in the FT intron does not completely explain the action of FLC on the FT gene, as an FT::GUS promoter fusion that does not contain the first intron of FT showed reduced expression in lines with a late FRI allele or 35S::FLC compared with a Columbia background (Michaels et al., 2005). These results showed that FLC can repress FT expression through the FT promoter region; our ChIP data did not show FLC binding to the FT promoter, suggesting that this repression may be indirect.

Using the 35S::FLC-FLAG line, we also found strong binding to CArG boxes in the SHP2 and AGL18 promoters. The ChIP data suggest that these are strong interactions; however, these interactions were not observed with endogenous FLC protein. These interactions could be the result of an excess of FLC-FLAG protein in the over-expression line, ectopic expression of FLC-FLAG protein by the 35S promoter, or alteration of FLC activity by the addition of the FLAG epitope to the protein. The latter possibility seems least likely as the FLC-FLAG protein still acts to repress flowering and gave the same ecotopic inflorescence phenotype that has been observed for over-expressed FLC (Hepworth et al., 2002). The binding of FLC-FLAG to the AGL18 and SHP2 promoters does not appear to cause any repression of expression of these two genes; a small increase in SHP2 expression was observed in 35S::FLC-FLAG. This suggests that the FLC bound at these promoters does not repress gene expression. This could be because the FLC protein is bound at a promoter position where it cannot affect the transcription mechanism or because an inactive form of FLC or FLC protein complex (such as the lower-molecular-weight complex, Figure 6c) is binding.

There is evidence for sequence conservation amongst the putative CArG boxes to which FLC binds. The boxes in SOC1, FT, AGL18 and SHP2 all contain a CCAAA half site (Table 1), and those in FT, AGL18 and SHP2 contain a GGTTT half site. However, other putative CArG boxes that we tested that were not bound by FLC also contain these sequences, so they may be necessary but not sufficient for FLC binding.

There is reduced binding of FLC at the SOC1 promoter in plants that have been vernalized, as would be expected with a reduction in FLC content in the plant. LD and GA treatment both gave a small increase in SOC1 mRNA expression. In both cases, FLC is still bound to the SOC1 promoter suggesting that some other factor can override FLC repression while it is bound to the SOC1 CArG box. Constans (CO) has been shown to act indirectly on a separate region of the SOC1 promoter to FLC (Hepworth et al., 2002), so binding of a factor to this region could be the mechanism for overcoming FLC repression.

MADS box proteins are thought to bind to CArG boxes as dimers or higher-order multimers (Riechmann et al., 1996). Although previous studies using yeast two-hybrid systems have not identified any MADS protein partners for FLC (de Folter et al., 2005), our data demonstrate that biologically active FLC-FLAG and endogenous FLC proteins interact in vivo. The lack of an FLC–FLC interaction in previous yeast two-hybrid studies may be because a higher-order structure involving another protein(s) or DNA, not present in yeast, is required to stabilize the interaction. Alternatively, so as to dimerize, FLC may require a post-translational modification that cannot be performed by yeast. We have observed two FLC-specific bands in high-resolution Western blots of C24 protein which could be indicative of a modification; two FLC bands were also observed in C24 by Rouse et al. (2002).

FLC-FLAG protein expressed from either its own promoter or the 35S promoter is present in a complex with a molecular weight of around 800 kDa as is the endogenous FLC protein. This 800 kDa complex is the only SEC peak where FLC-FLAG is detected in the late-flowering gFLC-FLAG line, and so this complex probably represents the active form of FLC. A significant amount of FLC protein is co-immunoprecipitated with the FLC-FLAG protein (Figure 6a), suggesting a high abundance of complexes that include at least two FLC proteins. The size of a dimer or tetramer of FLC (approximately 50–100 kDa) clearly does not account for the whole of the complex of approximately 800 kDa complex we have identified. The gFLC-FLAG extract was treated with micrococcal nuclease prior to SEC, excluding the possibility that DNA is part of the complex. It is therefore likely that other proteins besides FLC are in the 800 kDa complex. One possibility is that this complex contains chromatin-remodeling proteins such as histone deacetylases, and that FLC represses gene expression by formation and maintenance of an inactive chromatin state at its target sites.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Plant material

Plants were grown on agar-solidified MS plates. Long-day (LD) conditions were 16 h light, 8 h dark under cool white fluorescent tubes (100 μmol m−2 sec−1); short-day conditions (SD) were 8 h light, 16 h dark using the same light conditions. Gibberellin (GA) treatment was by addition of 10−6 m GA3 to the growth media. Plant material was fixed in 20 ml 1% formaldehyde in PBS for 1 h under vacuum, after which 1.6 ml 1 m glycine was added; samples were then frozen in liquid nitrogen. Arabidopsis transformation was via the floral dip method (Clough and Bent, 1998); transformed seed were selected on MS agar plates supplemented with 100 mg l−1 kanamycin.

Expression constructs

A sequence encoding the 3xFLAG epitope was constructed by annealing two oligonucleotides (5′-GGGGGGATCCGATTACAAAGATCATGATGGTGACTATAAGGACCACGACATCG and 5′-GGGGGAGCTCTCATTTATCATCATCATCTTTGTAATCGATGTCGTGGTCCTTATAG) and filling the single-stranded overhangs in with Taq DNA polymerase to produce an 89 bp fragment that was ligated into pGEM T-easy (Promega, Madison, WI, USA) and sequenced to verify that the correct sequence was present. The 3xFLAG sequence excised from pGEM T-easy with BamHI and HindIII was inserted into pBI121.1 (Jefferson, 1987) cut with BamHI and HindIII. The resulting plasmid was cut with XbaI, made blunt by filling in with Pfu DNA polymerase, and ligated with the GatewayTM (Invitrogen, Carlsbad, CA, USA) rfC cassette. The resulting vector allows insertion of GatewayTM-adapted gene fragments by an LR recombination reaction producing a fusion protein with a linker derived from the GatewayTM attB2 site followed by a 3xFLAG epitope at the C-terminus (additional amino acid sequence shown in Figure 1) – a total of 36 amino acids including the 22 amino acid 3xFLAG peptide. The pCH252 vector for expression of genomic 3xFLAG constructs in plants contains the rfC-3xFLAG-Tnos fragment from pCH183 cloned into the pZP211 vector (Hajdukiewicz et al., 1994). The full-length FLC cDNA was amplified with GatewayTM attB1 and attB2 sequences added to the PCR primers and cloned into an ampicillin-resistant pDONR201 derivative (Invitrogen). This clone was used in a GatewayTM LR recombination reaction with pCH183 to produce the 35S::FLC-FLAG construct.

The full-length FLC genomic clone was made in two steps. pENTR-gFLC-A was constructed by PCR amplifying a 2477 bp fragment using Pfx polymerase (Invitrogen) using primers P72 5′-CACCAACATAAATGCATAGAAACAATCTGG and P35 5′-ATTAAGTAGTGGGAGAGTCACCGG and recombining into pENTR/D-TOPO (Invitrogen) according to the manufacturer's instructions. pENTR-gFLC-A was digested with NotI and blunt-ended by Klenow fill-in reaction. A partial digestion with SpeI retaining a 2403 bp region of FLC (+3210 to +5613 relative to the start ATG) generated pENTR-gFLC-A*, and finally a 5235 bp SmaI–SpeI restriction fragment of a lamba clone (FLC from C24 ecotype Sheldon et al., 1999), spanning the region −2026 to +3209 of FLC, was ligated into pENTR-FLC-A* yielding pENTR-gFLC (no stop codon) spanning −2026 bp to +5613 of the FLC gene. The resulting clone was used in a GatewayTM LR recombination reaction with pCH252 to make the gFLC-FLAG construct.

Chromatin immunoprecipitation (ChIP)

Formaldehyde cross-linked (1% for 1 h) whole seedlings (1 g) were ground in liquid nitrogen and added to 5 ml lysis buffer (50 mm HEPES pH 7.5, 150 mm NaCl, 1% v/v Triton X100, 0.1% w/v deoxycholic acid, 0.1% w/v sodium dodecyl sulfate) with 50 μl plant protease inhibitor mix (Sigma, St Loius, MO, USA). The mixture was sonicated 4 times for 15 sec and centrifuged at 20 500 g for 40 min to pellet debris. A further 50 μl plant protease inhibitor mix was added to the supernatant recovered after centrifugation. For each ChIP reaction, 1 ml of extract was pre-cleared with 60 μl protein A agarose (Upstate, Lake Placid, NY, USA) for 2 h at 4°C with rotation. Antibody [2 μg anti-FLAG M2 (Sigma) or 1 μg affinity-purified anti-FLC] was added to the supernatant from the pre-clearing step and incubated for 2 h at 4°C with rotation. After this, 60 μl protein G agarose (anti-FLAG M2) or 60 μl protein A agarose (anti-FLC) was added and the samples rotated at 4°C for a further 2 h. The protein A or G agarose was then washed by rotating twice for 10 min with 1 ml lysis buffer at 4°C, once for 10 min with 1 ml LNDET buffer (0.25 m LiCl, 1% v/v NP40, 1% w/v deoxycholic acid, 1 mm EDTA, 10 mm Tris, pH 8.0) at 4°C, and three times for 5 min with TE (10 mm Tris, pH 8.0, 1 mm EDTA) at 4°C for the first two washes and room temperature for the final wash. The complexes were eluted in two 150 μl volumes of elution buffer (1% w/v sodium dodecyl sulfate, 0.1 m NaHCO3) incubated at room temperature for 15 min. Cross-linking was reversed by addition of 12 μl 5 m NaCl and incubation at 65°C for 4 h. The DNA was extracted from the immunoprecipitates by treatment with 20 μg proteinase K, followed by extraction with phenol:chloroform (1:1) and ethanol precipitation. ChIP experiments were carried out in triplicate and data from the replicates averaged.

Real-time PCR

Real-time PCR was carried out using SYBR Green Jumpstart ready mix (Sigma) with a Corbett Rotor-Gene RG3000A PCR machine (Corbett Research, Sydney, Australia). Primers used for real-time PCR analysis are shown in Supplementary Table 1. Analysis of the data in Figure 3 was carried out by comparison of threshold cycles of the samples compared with a standard curve of C24 genomic DNA or CArG::GUS genomic DNA for the CArG::GUS and ΔCArG::GUS samples. The data analysis for the primer sets in Figure 5 was carried out using the relative quantification function in the Corbett Rotor-Gene 6 software. The input amount of tissue was the same for the samples in each pair, and the approximately equal amounts of AMP1 genomic DNA in the sample pairs (Figure 5) confirms that the same quantity of background DNA was present in each sample. This allows direct comparison of the abundance of the candidate regions without normalization.

Semi-quantitative RT-PCR

cDNA was synthesized from 3 μg total RNA using Superscript III reverse transcriptase (Invitrogen) according to the manufacturer's instructions. PCR reactions were carried out using cDNA equivalent to 50 ng total RNA per reaction using Amplitaq DNA polymerase (Roche, Mannhein, Germany) according to the manufacturer's instructions with 1.5 mm MgCl2, with an annealing temperature of 60°C. PCR primers and number of cycles were: FT, ACTATATAGGCATCATCACCGTTCGTTACTCG and ACAACTGGAACAACCTTTGGCAATG, 30 cycles; AGL18, GAAATGAAGCTGTGTTGCGAAATG and TTCTTCCAATGCTTTTTTCTCCTG, 30 cycles; SHP2, AATACATGCAAAAAAGGGAAATCG and AGGAAAGAAGCTTATGTTAGACTG, 35 cycles; actin, AGAGATTCAGATGCCCAGAAGTCTTGTTCC and AACGATTCCTGGACCTGCCTCATCATACTC, 25 cycles.

Size exclusion chromatography (SEC) analysis

Protein complexes were extracted from 0.2 g whole seedlings in a lysis buffer containing 50 mm Tris–HCl (pH 7.4), 150 mm NaCl, 1% Triton X-100, 5 mm EDTA, 2 mm MgCl2, 1 mm dithiothreitol, 2 mm phenylmethanesulfonyl fluoride and a 1:200 dilution of protease inhibitor mixture. Sonicated samples were centrifuged at 20 500 g for 10 min and filtered through a 0.22 μm polycarbonate filter. Extract or size standards (BioRad, Hercules, CA, USA) were loaded onto a Sephadex 200 (300/10) column (Amersham, Little Chalfont, UK) using lysis Buffer without protease inhibitors as the running buffer. Fractions were precipitated with 3 volumes of acetone and resuspended in loading buffer (1% SDS, 50 mm Tris, pH 6.8, 10% glycerol, 50 mm dithiothreitol, 0.01% bromophenol blue), heated at 80°C for 3 min and separated on a 10% SDS–PAGE gel, blotted to Immobilon-P membrane (Millipore, Billerica, MA, USA) and probed with FLAG M2 antibody (1:20 000, Sigma) and sheep antimouse:HRP conjugate (1:16 000; Silenus, Melbourne, Australia) and visualized by chemiluminescence.

Co-immunoprecipitation

Plant material (1 g) was ground in liquid nitrogen and added to 5 ml of lysis buffer (50 mm Tris–HCl, pH 7.4, 150 mm NaCl, 1 mm EDTA, 1% Triton X100 with 50 μl protease inhibitor cocktail) and sonicated for 4 × 15 sec. Debris was pelleted by centrifugation at 20 500 g for 40 min. The supernatant was retained and a further 50 μl protease inhibitor cocktail added. A 1 ml aliquot of extract was then incubated with 60 μl EZview anti FLAG M2 agarose (Sigma) for 2 h. The agarose beads were washed 1× with 1 ml of lysis buffer for 5 min at 4°C and 2× with LNDET buffer for 5 min at 4°C. The anti-FLAG M2 agarose beads were then incubated at 95°C for 5 min in 100 μl SDS sample buffer (0.125 m Tris–HCl, pH 6.8, 4% w/v sodium dodecyl sulfate, 20% v/v glycerol, 2% v/v 2-mercaptoethanol). Samples were separated on 12% SDS–PAGE gels and blotted to Immobilon-P membrane and probed with FLAG M2-HRP antibody (1:5000) or FLC antiserum (1:10 000) followed by antirabbit antibody (1:5000; Amersham) and visualized by chemiluminescence.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

We thank Sue Allen, Anna Wielopolska and Robyn East for technical assistance, Frank Gubler for assistance with affinity purification of the FLC antibody, Greg Tanner for assistance with FPLC analysis, Carol Andersson for RT-PCR primers and George Coupland for providing the SOC1 CArG box lines.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Borner, R., Kampmann, G., Chandler, J., Gleissner, R., Wisman, E., Apel, K. and Melzer, S. (2000) A MADS domain gene involved in the transition to flowering in Arabidopsis. Plant J. 24, 591599.
  • Brizzard, B.L., Chubet, R.G. and Vizard, D.L. (1994) Immunoaffinity purification of FLAG epitope-tagged bacterial alkaline phosphatase using a novel monoclonal antibody and peptide elution. Biotechniques, 16, 730734.
  • Clough, S.J. and Bent, A.F. (1998) Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant J. 16, 735743.
  • de Folter, S., Immink, R.G.H., Kieffer, M. et al. (2005) Comprehensive interaction map of the Arabidopsis MADS box transcription factors. Plant Cell, 17, 14241433.
  • Hajdukiewicz, P., Svab, Z. and Maliga, P. (1994) The small, versatile pPZP family of Agrobacterium binary vectors for plant transformation. Plant Mol. Biol. 25, 989994.
  • Helliwell, C.A., Chin-Atkins, A.N., Wilson, I.W., Dennis, E.S. and Chaudhury, A.M. (2001) The Arabidopsis AMP1 gene encodes a putative glutamate carboxypeptidase. Plant Cell, 13, 21152125.
  • Helliwell, C.A., Wesley, S.V., Wielopolska, A.J. and Waterhouse, P.M. (2002) High throughput vectors for efficient gene silencing in plants. Funct. Plant Biol. 29, 12171225.
  • Hepworth, S.R., Valverde, F., Ravenscroft, D., Mouravdov, A. and Coupland, G. (2002) Antagonistic regulation of flowering-time gene SOC1 by CONSTANS and FLC via separate promoter motifs. EMBO J. 21, 43274337.
  • Honma, T. and Goto, K. (2001) Complexes of MADS-box proteins are sufficient to convert leaves into floral organs. Nature, 490, 525529.
  • Huang, H., Mizukami, Y., Hu, Y. and Ma, H. (1993) Isolation and characterization of the binding sequences for the product of the Arabidopsis floral homeotic gene AGAMOUS. Nucleic Acids Res. 21, 47694776.
  • Jefferson, R.A. (1987) Assaying chimeric genes in plants: the GUS gene fusion system. Plant Mol. Biol. Rep. 5, 387405.
  • Koornneef, M., Blanketijn-de Vries, H., Hanhart, C., Soppe, W. and Peeters, T. (1994) The phenotype of some late-flowering mutants is enhanced by a locus on chromosome 5 that is not effective in the Landsberg erecta wild-type. Plant J. 6, 911919.
  • Lee, I., Micheals, S.D., Masshardt, A.S. and Amasino, R.M. (1994) The late-flowering phenotype of FRIGIDA and LUMINIDEPENDENS is suppressed in the Landsberg erecta strain of Arabidopsis. Plant J. 6, 903909.
  • Lee, H., Suh, S.-S., Park, E., Cho, E., Ahn, J.H., Kim, S.-G., Lee, J.S., Kwon, Y.M. and Lee, I. (2000) The AGAMOUS-LIKE 20 MADS domain protein integrates floral inductive pathways in Arabidopsis. Genes Dev. 14, 23662376.
  • Lee, H., Habas, R. and Abate-Shen, C. (2004) Msx1 cooperates with histone H1b for inhibition of transcription and myogenesis. Science, 304, 16751678.
  • Michaels, S.M. and Amasino, R.M. (1999) FLOWERING LOCUS C encodes a novel MADS domain protein that acts as a repressor of flowering. Plant Cell, 11, 949956.
  • Michaels, S.D., Himelblau, E., Kim, S.Y., Schomburg, F.M. and Amasino, R.M. (2005) Integration of flowering signals in winter-annual Arabidopsis. Plant Physiol. 137, 149156.
  • Moon, J., Suh, S.-S., Lee, H., Choi, K.-R., Hong, C.B., Paek, N.-C., Kim, S.-G. and Lee, I. (2003) The SOC1 MADS box gene integrates vernalization and gibberellin signals for flowering in Arabidopsis. Plant J. 35, 613623.
  • Pařenicová, L., de Folter, S., Kieffer, M. et al. (2003) Molecular and phylogenetic analyses of the complete MADS-box transcription factor family in Arabidopsis: new openings to the MADS world. Plant Cell, 15, 15381551.
  • Ratcliffe, O.J., Nadzan, G.C., Reuber, T.L. and Riechmann, J.L. (2001) Regulation of flowering in Arabidopsis by an FLC homologue. Plant Physiol. 126, 122132.
  • Ratcliffe, O.J., Kumimoto, R.W., Wong, B.J. and Riechmann, J.L. (2003) Analysis of the Arabidopsis MADS AFFECTING FLOWERING gene family: MAF2 prevents vernalization by short periods of cold. Plant Cell, 15, 11591169.
  • Riechmann, J.L., Krizek, B.A. and Meyerowitz, E.M. (1996) Dimerization specificity of Arabidopsis MADS domain homoetic proteins APETALA1, APETALA3, PISTILLATA, and AGAMOUS. Proc. Natl Acad. Sci. USA, 93, 47934798.
  • Rouse, D.T., Sheldon, C.C., Bagnall, D.J., Peacock, W.J. and Dennis, E.S. (2002) FLC, a repressor of flowering, is regulated by genes in different inductive pathways. Plant J. 29, 183191.
  • Sheldon, C.C., Burn, J.E., Perez, P.P., Metzger, J., Edwards, J.A., Peacock, W.J. and Dennis, E.S. (1999) The FLF MADS box gene: a repressor of flowering in Arabidopsis regulated by vernalization and methylation. Plant Cell, 11, 445458.
  • Sheldon, C.C., Rouse, D.T., Finnegan, E.J., Peacock, W.J. and Dennis, E.S. (2000) The molecular basis of vernalization: the central role of FLOWERING LOCUS C (FLC). Proc. Natl Acad. Sci. USA, 97, 37533758.
  • Shindo, C., Aranzana, M.J., Lister, C., Baxter, C., Nicholls, C., Nordberg, M. and Dean, C. (2005) Role of FRIGIDA and FLC in determining variation in flowering time of Arabidopsis. Plant Physiol. 138, 11631173.
  • Shiraishi, H., Okada, K. and Shimura, Y. (1993) Nucleotide sequences recognized by the AGAMOUS MADS domain of Arabidopsis thaliana in vitro. Plant J. 4, 385398.
  • Shore, P. and Sharrocks, A.D. (1995) The MADS-box family of transcription factors. Eur. J. Biochem. 229, 113.
    Direct Link:
  • Tilly, J.J., Allen, D.W. and Jack, T. (1998) The CArG boxes in the promoter of the Arabidopsis floral organ identity gene APETALA3 mediate diverse regulatory effects. Development, 125, 16471657.
  • Wang, H., Tang, W., Zhu, C. and Perry, S. (2002) A chromatin immunoprecipitation (ChIP) approach to isolate genes regulated by AGL15, a MADS domain protein that preferentially accumulates in embryos. Plant J. 32, 831843.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Table S1 PCR primers used for ChlP experiments

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
TPJ_2686_sm_TableS1.xls16KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.