RESEARCH ARTICLE1978

Development 139, 1978-1986 (2012) doi:10.1242/dev.077073© 2012. Published by The Company of Biologists Ltd

INTRODUCTIONThe flower is an evolutionary innovation that contributes to thesuccess of angiosperms. Dicotyledonous flowers are composed offour major types of organs: sepal, petal, stamen and carpel. Fourmajor classes of homeotic genes, A, B, C and E, specify the fourfloral organ types in a combinatorial manner (reviewed by Krizekand Fletcher, 2005). A-class genes, APETALA1 and APETALA2(AP1 and AP2), together with the E-class genes, confer sepal identityin the first whorl. Petal identity is determined by the activities of A-, B- and E-class genes in the second whorl. The C-class gene,AGAMOUS (AG), together with B- and E-class genes, specifiesstamen identity in the third whorl. Carpel identity is conferred by C-and E-class activities in the fourth whorl. With the exception of AP2,all floral homeotic genes encode MADS domain-containing proteinsfor which DNA binding and dimerization/multimerizationspecificities have been extensively characterized (reviewed byImmink et al., 2010). By contrast, the activity of AP2 as atranscription factor in flower development is poorly understood. Inaddition to promoting sepal and petal identities, AP2 restricts AGexpression to the inner two whorls (Bowman et al., 1991; Drews etal., 1991). Previous studies using a GUS reporter system have shownthat the 3-kb AG second intron contains sequence elements requiredfor its proper expression, including responsiveness to repression byAP2 (Bomblies et al., 1999; Deyholos and Sieburth, 2000).

AP2 is the founding member of a family of 144 genes thatencode at least one AP2 DNA-binding domain in Arabidopsis; thebiological functions of this family range from development tostress and defense responses (Jofuku et al., 1994; Weigel, 1995;Okamuro et al., 1997; Riechmann and Meyerowitz, 1998). ThisDNA-binding domain was thought to be plant specific butcomputational analysis has identified this DNA binding domain inspecies such as Tetrahymena (Wuitschick et al., 2004) andPlasmodium (Yuda et al., 2009). Plant AP2 domain-containinggenes were categorized into five subfamilies (Sakuma et al., 2002).Members of the AP2-like subfamily generally contain two AP2DNA-binding domains, AP2R1 and AP2R2 (Kim et al., 2006;Shigyo et al., 2006). In the second to fifth subfamilies, the ERF-like, DREB-like, RAV-like and others, members contain one AP2DNA-binding domain and are involved in abiotic and biotic stressresponses.

The DNA-binding properties of AP2 domain proteins fromvarious subfamilies have been studied. Members of the ERF-likeand DREB-like subfamily bind well-documented GC-rich motifs(Ohme-Takagi and Shinshi, 1995; Stockinger et al., 1997; Hao etal., 1998; Liu et al., 1998). The AP2 domain in a RAV-like familymember binds a CAACA motif (Kagaya et al., 1999). ANT is theonly protein in the AP2-like subfamily for which DNA-bindingproperties have been studied (Nole-Wilson and Krizek, 2000;Krizek, 2003). Despite the crucial role of AP2 in flowerdevelopment and the diversity of its targets (Yant et al., 2010), theDNA binding specificity of AP2 has never been characterized.

To better understand the molecular functions of AP2, we soughtto determine its binding consensus sequence in vitro andcharacterize the relevance of the sequence in vivo. Here, we reportthat AP2R2 specifically binds the TTTGTT or AACAAA motif invitro. We show that these motifs within the 2nd intron of AG areimportant for restricting AG expression to the inner two whorls invivo. In silico analysis of 2nd intron sequences from AG orthologs

1Department of Botany and Plant Sciences, Center for Plant Cell Biology, Institute ofIntegrative Genome Biology, University of California, Riverside, CA 92521, USA.2Max Planck Institute for Developmental Biology, Department of Molecular Biology,D-72076 Tübingen, Germany. 3Howard Hughes Medical Institute, University ofCalifornia, Riverside, CA 92521, USA.

*Present address: Department of Organismic and Evolutionary Biology, HarvardUniversity, Cambridge, MA 02138, USA‡Author for correspondence (xuemei.chen@ucr.edu)

Accepted 19 March 2012

SUMMARYCell fate specification in development requires transcription factors for proper regulation of gene expression. In Arabidopsis,transcription factors encoded by four classes of homeotic genes, A, B, C and E, act in a combinatorial manner to control properfloral organ identity. The A-class gene APETALA2 (AP2) promotes sepal and petal identities in whorls 1 and 2 and restricts theexpression of the C-class gene AGAMOUS (AG) from whorls 1 and 2. However, it is unknown how AP2 performs these functions.Unlike the other highly characterized floral homeotic proteins containing MADS domains, AP2 has two DNA-binding domainsreferred to as the AP2 domains and its DNA recognition sequence is still unknown. Here, we show that the second AP2 domain inAP2 binds a non-canonical AT-rich target sequence, and, using a GUS reporter system, we demonstrate that the presence of thissequence in the AG second intron is important for the restriction of AG expression in vivo. Furthermore, we show that AP2 bindsthe AG second intron and directly regulates AG expression through this sequence element. Computational analysis reveals thatthe binding site is highly conserved in the second intron of AG orthologs throughout Brassicaceae. By uncovering a biologicallyrelevant AT-rich target sequence, this work shows that AP2 domains have wide-ranging target specificities and provides a missinglink in the mechanisms that underlie flower development. It also sets the foundation for understanding the basis of the broadbiological functions of AP2 in Arabidopsis, as well as the divergent biological functions of AP2 orthologs in dicotyledonous plants.

KEY WORDS: AGAMOUS, ANT, AP2, APETALA2, DNA binding, Flower development

The floral homeotic protein APETALA2 recognizes and actsthrough an AT-rich sequence elementThanh Theresa Dinh1, Thomas Girke1, Xigang Liu1, Levi Yant2,*, Markus Schmid2 and Xuemei Chen1,3,‡

DEVELO

PMENT

1979RESEARCH ARTICLEAP2 binds an AT-rich sequence

uncovers strong conservation of this element in the Brassicaceaefamily. Furthermore, we found that AP2 directly regulates AG inyoung flowers through these elements. Curiously, the AP2 full-length protein binds DNA with no apparent specificity in vitro,suggesting that other factors influence its DNA binding specificityin vivo. These findings establish a missing link in the mechanismsunderlying flower development, shed light on the molecularfunction of AP2, and set the foundation for further appreciation ofthe molecular basis for the broad biological functions of AP2.

MATERIALS AND METHODSPlasmid constructionTo express the AP2R1, AP2R2 and AP2R1R2 domains of AP2 in E. coli,the corresponding coding regions from the AP2 cDNA were amplified byPCR (supplementary material Table S1) and cloned into the pET21-Avector using BamHI and EcoRI sites (Novagen). To express the full-lengthAP2 protein, the entire coding region of AP2 was cloned in-frame to an N-terminal MBP and His tag using BamHI and EcoRI sites in the pmCSG7XF0510 MBP-LIC vector (a gift from Dr Xiaofeng Cao, Institute ofGenetics and Developmental Biology, Beijing, China).

For in vivo analysis of the AG 2nd intron, the region of the AG 2ndintron in the KB31 construct (Bomblies et al., 1999) was amplified andcloned into PCR2.1 (Invitrogen). Site-directed mutagenesis was performed(supplementary material Table S1) to introduce mutations into each of thetwo AP2-binding sites. The wild-type and mutant KB31 fragments werethen cloned into pD991 (Tilly et al., 1998) using BamHI and HindIII sites.The 35S::AP2m3-GR construct was generated as described previously(Yant et al., 2010).

Protein expression and purificationThe pET21A-AP2R1, AP2R2 and AP2R1R2, and the MBP-AP2 full-lengthprotein plasmids were transformed into E. coli BL21. Protein expression andpurification were carried out as previously described (Smith et al., 2002;Husbands et al., 2007) and purified proteins were quantified against BSA.

Selection affinity and amplification binding (SAAB) assayEither 200 ng or 500 ng of doubly affinity-purified (with Ni2+ beads and T7antibody) and desalted AP2R2, AP2R1 or AP2R1R2 was subjected to aSAAB assay as previously described (Smith et al., 2002). Briefly, the protein-bead mixture was divided into six tubes. In the first tube, a pool of random,double-stranded oligonucleotides (supplementary material Table S1) wasadded and incubated for 4 hours with the protein-bead mixture. The DNAbound by the protein-bead complex was eluted and PCR was performed toamplify the bound sequences. An aliquot of the PCR reaction was added tothe second tube of the protein-bead mixture to allow protein-DNA bindingto occur. This process of affinity binding and amplification was reiterated atotal of six times. The PCR product from either cycle 5 and/or 6 was clonedvia TA cloning and sequenced. The sequences were analyzed with the motiffinding program MEME to identify consensus motifs.

Electrophoretic mobility shift assays (EMSAs)EMSAs were performed as described (Husbands et al., 2007) with somemodifications. For EMSAs shown in Figs 1, 2 and in supplementarymaterial Fig. S4B, Figs S5, S7, probes were generated by annealing 100pmol of sense and antisense oligonucleotides (supplementary materialTable S1) and 1-2 pmol of probe was used in each reaction. For gel-shiftsshown in Fig. 3B,C and supplementary material Fig. S8, the DNA fragmentwas amplified by PCR from wild-type or mutant versions of KB31, or fromCol genomic DNA, and 0.1-0.2 pmol of labeled probe was used in eachreaction. Probes were prepared as previously described (Broitman-Maduroet al., 2005).

Gel shift reactions were conducted at 4°C in 20% glycerol, 20 mM Tris(pH 8.0), 10 mM KCl, 1 mM DTT, 12.5 ng poly dI/C, 6.25 pmol ofrandom, single-stranded oligonucleotides, Herring sperm DNA, BSA andthe probe in the amount specified above.

All samples involving AP2R2 were loaded on an 8% gel, whereas asthose involving the AP2 full-length protein were loaded on a 6% gel toresolve protein-DNA complexes. Gels were then dried and either exposed

to X-ray films or imaged and quantified using the TyphoonPhosphorImager. For quantification analyses in Fig. 2 and supplementarymaterial Fig. S5, the percentage bound was calculated by dividing theshifted amount after incubation with the mutated probes by the shiftedamount after incubation with the -probe.

In reactions with cold competitors, 5-40� unlabeled probes wereincluded in the reactions. Furthermore, anti-His and anti-T7 antibodieswere added to some reactions at 1-2� the amount of the AP2R2 protein toobtain super-shifts.

GUS staining and microscopyInflorescences were stained for GUS activity and processed for sectioningas previously described (Sieburth and Meyerowitz, 1997). Slides wereviewed under a Leica DMR compound microscope and images were takenwith a Spot digital camera (Diagnostic Instruments).

Induction and expression analysis of 35S::AP2m3-GRThe inflorescences of 35S::AP2m3-GR ap2-2 plants were treated once witha solution of 10 M dexamethasone (DEX)/0.015% Silwet with or without10 M cyclohexamide (CHX) (Fisher). Six hours later, the treatedinflorescences were dissected to remove stage 8 and older flowers. TotalRNA was isolated from the dissected inflorescences and subjected toDNaseI treatment and reverse transcription. RT-PCR was performed on thecDNAs using primers specific for AG and UBQ5 (supplementary materialTable S1). Real-time RT-PCR was performed on the same cDNAs using areal-time PCR SYBR Green system (BioRad). Three technical replicateswere performed for each real-time RT-PCR. Three biological replicates ofDEX induction and real-time RT-PCR were performed. Error bars representthe standard deviation from three technical replicates.

Chromatin immunoprecipitationChromatin immunoprecipitation (ChIP) experiments were performed ontwo biological replicates following previously described protocols (Gomez-Mena et al., 2005; Mathieu et al., 2009; Yant et al., 2010). The input andChIP samples were subjected to real-time PCR (BioRad). Three technicalreplicates were performed. The data were analyzed as previously described(Wierzbicki et al., 2008).

Sequence analysisAll 2nd intron sequences from AG orthologs were downloaded fromGenBank (supplementary material Table S2). The start and end positions,provided by Hong et al. (Hong et al., 2003) and those specified in theGenBank annotations files, were used to parse the introns from their sourcesequences and to bring them into their proper sense orientation(supplementary material Table S2). Sequence manipulations and analyseswere performed with custom scripts that are based on the Biostrings packageof the statistical programming environment R (Morgan et al., 2009; RDevelopment Core Team, 2010). Multiple sequence alignments (MSAs) werecomputed with the diaglign2-2 software from Morgenstern (Morgenstern,2004) using the default parameters in the DNA mode. A sliding windowanalysis was performed to visualize the degree of conservation in the finalMSA. For this, the relative conservation of each base was calculated at eachposition where a value of 1.0 indicates perfect conservation of one base(disregarding gaps) and a value of 0 indicates equal representation of all fourbases. For plotting purposes, these conservation values were smoothed bycalculating their mean for a sliding window size of 20 nucleotides along allMSA positions. Pattern searches were performed with the matchPatternfunction of the Biostrings package (Morgan et al., 2009).

RESULTSAP2R2 binds a novel consensus sequenceTo begin uncovering the molecular mechanisms underlying the roleof AP2 in development, we sought to identify its binding sequenceby performing a SAAB assay with AP2R1, AP2R2 or AP2R1R2doubly purified based on their N- and C-terminal tags.

When the SAAB assay was performed for the purified AP2R2(supplementary material Fig. S1A), amplified DNA from the boundfraction could be detected starting from cycle 4 (supplementary D

EVELO

PMENT

1980

material Fig. S1B). After cloning and sequencing of the AP2R2-bound DNA after cycle 6, 20 unique sequences were obtained, 17of which contained two perfect AT-rich motifs – AACAAA or thecomplementary TTTGTT (supplementary material Fig. S1C) –whereas the other three clones had either one site (clone 11), asingle nucleotide change at each site (clone 12) or a singlenucleotide change at one of the sites (clone 7) (supplementarymaterial Fig. S1C). Therefore, all recovered clones contained atleast one copy of the consensus sequence with no more than onenucleotide change.

When the same procedure was applied to purified AP2R1(supplementary material Fig. S2A), no bound DNA was detectableby PCR, indicating that AP2R1 did not bind DNA in vitro(supplementary material Fig. S2B). The lack of DNA recovered fromthe AP2R1 SAAB assay was not due to loss of the protein during theprocedure because the AP2R1 protein was present on the beadsthroughout the experiment (supplementary material Fig. S2C). TheSAAB assay was also performed for purified AP2R1R2(supplementary material Fig. S3A). DNA bound to AP2R1R2 wasdetectable from cycles 3 to 6 (supplementary material Fig. S3B);however, sequencing of cloned DNA bound to AP2R1R2 from cycle6 did not reveal any obvious consensus sequence (supplementarymaterial Fig. S3C). Ten out of 25 unique sequences contained one of

RESEARCH ARTICLE Development 139 (11)

the sites bound by AP2R2 (supplementary material Fig. S3C; anddata not shown), one had a site with one nucleotide change (clone3), and some of the other sequences were GC rich.

Next, to confirm the SAAB assay results, we performed EMSAswith the AP2R2 protein. Sense and antisense strands correspondingto one of the sequences obtained from the SAAB assay containingboth AACAAA (termed ‘’ site) and TTTGTT (termed ‘’ site)were used as a probe (Fig. 1A). DNA binding as revealed by a shiftin mobility was observed with as little as 50 ng of AP2R2 (data notshown), and with 100 or 200 ng of AP2R2, which yielded morestrong and consistent binding (Fig. 1B, lanes 1 and 2, arrow). Toconfirm the observed binding, 20-fold cold competitor was addedto the binding reaction. Indeed, the intensity of the shift was greatlydiminished (Fig. 1B, lane 3). As the experiments were conductedwith AP2R2 purified from E. coli, there was a possibility that theobserved shift may be due to a contaminating protein instead ofAP2R2 (although AP2R2 was the only protein detected byCoomassie staining in the protein fraction; supplementary materialFig. S1A). If the observed binding was specifically caused byAP2R2, the addition of His and T7 antibodies would generate asupershift as AP2R2 had both a T7 and a His tag. We observed thatthe inclusion of the His and T7 antibodies resulted in super-shiftedbands when compared with AP2R2 alone (Fig. 1B, compare lanes4 and 5 to lane 2, stars), confirming that it was indeed AP2R2 itselfthat bound the probe. As a control, we also added 400 ng His andT7 antibodies to the reaction in the absence of AP2R2 and we didnot observe any of the same band shifts as seen in the presence ofAP2R2 (Fig. 1B, lane 6).

Next, we tested whether both sites were necessary for AP2R2binding. We mutated the AACAAA site to AGGTGA and theTTTGTT site to TCCACT (Fig. 1A). The EMSA showed thatAP2R2 was still able to bind probes that had one intact site (Fig.1C, lanes 2 and 5, arrow). The shift was lost upon addition of thecorresponding cold competitor (Fig. 1C, lanes 3 and 6). Loss ofboth sites, however, completely abolished binding (Fig. 1C, lane8). These results demonstrate that AP2R2 binds AACAAA and/orTTTGTT in vitro.

ANT, the only protein characterized in terms of DNA-bindingproperties in the AP2-like subfamily of AP2 domain-containingproteins, has been shown to bind a loose and long consensus sequence,

Fig. 1. AP2R2 binds the TTTGTT and/or AACAAA motif in vitro.(A)The sequences of the DNA probes (only one strand is shown). Theconsensus sites are underlined. , a probe containing two sites; or, probes containing only one site. (B)AP2R2 binds the probecontaining both sites. The shifted band that represents binding (lanes 1and 2) is indicated by an arrow. The binding was lost upon the additionof 20� cold competitor (lane 3). The + sign indicates that 200 ng ofAP2R2 protein was included; for lane 1, 100 ng AP2R2 was used. Inlanes 4 and 5, 200 ng and 400 ng, respectively, of T7 and Hisantibodies were added to the reactions. The bands indicated by thestars and double stars probably represent the super-shifts. As anegative control, 400 ng His and T7 antibodies were added withoutAP2R2 (lane 6). Lane 7 represents the free probe lane. (C)One site issufficient for binding by AP2R2. A shift was observed, as indicated bythe arrow, with probes containing one site (lanes 2 and 5). Binding waslost upon addition of the cold competitors (lanes 3 and 6). AP2R2binding was abolished when both sites were mutated (lane 8). Lanes 1,4 and 7 are the free probe lanes.

Fig. 2. Dissection of the AP2R2 consensus sequence. EMSAs wereperformed upon mutational perturbations at each position of the six-mer consensus sequence to the other three nucleotides while keepingthe other five positions unchanged. Green boxes represent the originalsite; yellow and orange boxes represent medium and minimal binding,respectively; red boxes indicate complete loss of binding. Medium andminimal binding are defined as 20-30% and 1-5% of binding incomparison with the wild-type sequence, respectively.

DEVELO

PMENT

gCAC(A/G)N(A/T)TcCC(a/g)ANG(c/t) (Nole-Wilson and Krizek,2000). We wondered how tight the AP2R2 consensus bindingsequence might be. As AP2R2 was able to bind probes containing onesite, probes containing just one site were used for testing thenucleotide specificity at each position. The nucleotide at each positionwas converted into one of the other three nucleotides, whereas theother five positions remained unchanged. All possible perturbationswere made and EMSAs were performed with AP2R2. Our resultsindicated that the binding site was extremely tight as any mutation atany position greatly compromised binding by AP2R2 (Fig. 2,supplementary material Fig. S4). Notably, any change at the fourthposition resulted in loss of binding (Fig. 2, supplementary materialFig. S4).

Proteins in the ERF-like and DREB-like subfamilies containinga single AP2 DNA-binding domain have been shown to bindspecific GC-rich consensus sequences (Ohme-Takagi and Shinshi,1995; Stockinger et al., 1997; Hao et al., 1998; Liu et al., 1998).Thus, we wanted to test whether AP2R2 could bind theconventional GCC-box (Hao et al., 1998), GCCGCC or otherrandom sequences. We performed EMSAs with a GCC-box, tworandom probes (supplementary material Fig. S5A) and the probe as a positive control. We observed binding of AP2R2 to the probe (supplementary material Fig. S5B, lane 2) but not to thecanonical GCC-box or two other random sequences(supplementary material Fig. S5B, lanes 5, 7 and 9).

AP2 full-length protein lacks obvious DNA-bindingspecificity in vitroNext, to determine whether the full-length AP2 protein could bindthe consensus sequence TTTGTT, we cloned the AP2 full-lengthprotein into a vector containing an N-terminal His and MBP tag(MBP-AP2), and purified and desalted it (supplementary materialFig. S6A). Owing to the large size of MBP-AP2, 400 ng was usedto perform the EMSA. Using the same probes as in the AP2R2EMSAs (Fig. 1A), we found that MBP-AP2 was able to bindprobes containing one or both consensus sequences (supplementarymaterial Fig. S7A, lanes 2 and 6, arrow). Interestingly, MBP-AP2was also able to bind the probe with both sites mutated(supplementary material Fig. S7A, lane 9, arrow). To assess thebinding specificity of the AP2 full-length protein further, weperformed EMSAs with a probe containing the canonical GCC-boxand two random probes (supplementary material Fig. S5A).Interestingly, MBP-AP2 could bind all of the probes in vitro(supplementary material Fig. S7B, lanes 2, 5 and 8, arrow). Toaddress whether the MBP tag may be binding the probes, an equalamount of MBP was added to the reactions for the EMSA(supplementary material Fig. S6B). MBP alone did not bind anyDNA sequences (supplementary material Fig. S7A, lanes 3, 7 and10; S7B, lanes 3, 6 and 9).

AP2R2 binds AG 2nd intron in vitroConsidering that AP2R2, but not AP2 full-length protein,specifically bound the consensus sequence, we sought to testwhether the AP2R2-binding site had biological relevance. Previousstudies have shown that AP2 represses AG expression (Drews etal., 1991), but it is still unknown whether AG is a direct target ofAP2, although AP2 has been shown to bind the AG 2nd intron invivo in our previous ChIP-seq analysis (Yant et al., 2010).Characterization of the AG 2nd intron has identified a 750 bpregion that, when fused to the GUS reporter in a construct termedKB31, confers AP2 responsiveness to GUS (Bomblies et al., 1999;Deyholos and Sieburth, 2000). In silico analysis indicated that the

1981RESEARCH ARTICLEAP2 binds an AT-rich sequence

KB31 region contained two AP2R2-binding sites at the 3� end,which we termed A and B (Fig. 3A). Thus, we proceeded to testwhether AP2R2 was able to bind this region of the AG secondintron in vitro. Using primers encompassing this region(supplementary material Table S1), we generated a 167 bp probe(Fig. 3A) to perform an EMSA. Indeed, a shift was found withAP2R2 and the binding was stronger as we added increasing

Fig. 3. AP2R2 and AP2 full-length protein bind AG 2nd intron invitro. (A)A diagram of a partial AG genomic region. Solid dark grayboxes indicate the 2nd and 3rd exons; lines represent introns; the solidlight-gray box indicates the 750 bp region (KB31) that still responds tonegative regulation by AP2. Within KB31, a 167 bp region, for whichthe sequence is shown, contains two AP2R2 consensus sites(rectangles). This 167 bp region was amplified with primers (arrows)and used as a probe for EMSA. Several nucleotides in the twoconsensus sequences were mutated (in gray) to generate the ABprobe. (B)EMSA with AP2R2 on the 167 bp wild-type (AB) or ABprobe. AP2R2 bound the wild-type probe (lanes 2-7; arrow) and thebinding was lost upon addition of the cold competitor (lanes 8 and 9).c20� cold competitor; C40� cold competitor. The protein amountsare as follows: 35 ng (lane 2), 75 ng (lanes 3 and 11), 107 ng (lane 4),160 ng (lanes 5 and 12) and 240 ng (lanes 6 and 13). Binding wasminimal for the AB probe containing mutated versions of both sites(lanes 11-13). Free0.2 pmol/reaction (lanes 1 and 10). (C)EMSA withMBP-AP2 full-length protein (labeled as MBP-AP2) on the 167 bpfragment. MBP-AP2 full-length protein bound both the wild-type (lanes3-8) and mutated (lanes 12 and 13) probes, and the binding was lostupon addition of the cold competitor (lanes 9 and 14). Binding wasobserved starting at 243 ng of the protein (lane 6), and was mostobvious with 355 ng (lane 7) and 429 ng (lane 8) of the protein. MBP(M) alone did not bind any probe (lanes 2 and 11). Free0.1pmol/reaction (lanes 1 and 10). Triangles depict increasing amounts ofprotein added. The arrow marks the shift. D

EVELO

PMENT

1982 RESEARCH ARTICLE Development 139 (11)

amounts of the protein (Fig. 3B, lanes 2-7, arrow). Furthermore thebinding was lost upon the addition of 20� or 40� cold competitor(Fig. 3B, lanes 8 and 9, respectively) to the binding reaction. Totest whether the observed binding required the two sites within the167 bp sequence, we performed site-directed mutagenesis to mutateboth sites (Fig. 3A). With the AB probe, binding was diminishedgreatly (Fig. 3B), showing that AP2R2 binds the 167 bp region viathe two elements in vitro. In addition, and consistent with priorresults, gel shifts showed binding of the MBP-AP2 full-lengthprotein to both the wild-type and the AB probe (Fig. 3C, lanes 3-8, 12 and 13, arrow). This binding was abolished upon addition of40� cold competitor (Fig. 3C, lanes 9 and 14) and MBP alone didnot bind either probe (Fig. 3C, lanes 2 and 11).

The AP2R2 binding sites are important for therestriction of AG expression in vivoTo evaluate the importance of the AP2R2 binding sites in the AG2nd intron in vivo, we used the KB31 GUS reporter, which hadbeen shown to report faithfully the endogenous domains of AGexpression and to respond to the regulation by AP2 (Bomblies etal., 1999; Deyholos and Sieburth, 2000). We cloned the 750 bpKB31 region from the AG 2nd intron containing either the wild-type or mutant (AB) sites into a GUS expression vector with aminimal (–60) 35S promoter (Tilly et al., 1998). The constructswere introduced into rdr6-11 to prevent post-transcriptional genesilencing of the transgenes (Dalmay et al., 2000; Mourrain et al.,2000). For the wild-type construct, GUS staining ofinflorescences from 99 independent T1 transgenic plants showedthat 74 recapitulated the proper AG expression patterns (Fig.4A,C,E). Twenty-two inflorescences did not show any GUSstaining and three did not recapitulate the proper AG expressionpatterns. For the AB construct, however, 71% of the 27independent transformants showed expansion of the GUSexpression domain to the outer two whorls (Fig. 4B,D,F) in allstages of flower development, with the remainder showing thecorrect expression patterns. Therefore, the A and B sites areimportant for the restriction of AG expression to the inner twofloral whorls.

AP2 directly regulates AG in young flowersthrough the binding sitesThe fact that the AP2R2 binding sites in the 2nd intron of AG areimportant for the restriction of AG expression to the inner twowhorls implies that AP2 is a direct regulator of AG. To addresswhether AP2 acts on AG directly, we used a rat glucocorticoidreceptor (GR)-induction system that has been widely used inArabidopsis as a method to establish direct relationships betweena transcription factor and its targets (Sablowski and Meyerowitz,1998; Wagner et al., 1999; Ito et al., 2004; William et al., 2004).Because AP2 is targeted by miR172 and transgenes containingmiRNA target sites are readily silenced in vivo, we fused amiR172-resistant AP2 cDNA (AP2m3) (Chen, 2004) to GR. The35S::AP2m3-GR construct was transformed into the progeny ofap2-2/+ plants. After obtaining single-locus insertiontransformants of the ap2-2/+ genotype, single and continuoustreatments of 10 M DEX were performed to determine thefunctionality of the transgene. A single DEX treatment ofinflorescences was not sufficient to induce the AP2m3 phenotype(data not shown) (Chen, 2004). Continuous treatments (once aday for 1 week), however, led to the induction of the AP2m3phenotype, thus showing that the transgene was functional (Fig.5A,B).

To determine whether AP2 directly represses AG expression, wesubjected 35S::AP2m3-GR ap2-2 inflorescences to a singletreatment of cyclohexamide (CHX) with or without DEX. After 6hours, inflorescence tissue was micro-dissected to remove stage 8and older flowers. RT-PCR was performed to measure AG mRNAlevels. We found that upon induction of AP2m3-GR, AG mRNAlevels decreased in young flowers (Fig. 5C). Real-time RT-PCR ofthree biological replicates revealed a 50% decrease in AG transcriptlevels upon AP2m3-GR induction (Fig. 5D). Therefore, AG islikely to be a direct target of AP2.

Next, we sought to determine whether AP2 regulates AG throughthe two AP2R2-binding sites. If AP2 acts through the two sites, wewould expect KB31, but not KB31AB, to be repressed by AP2.KB31 and KB31AB transgenic lines harboring a single transgenelocus were identified and crossed into the 35S::AP2m3-GR ap2-2background. Homozygous KB31 35S::AP2m3-GR ap2-2 orKB31AB 35S::AP2m3-GR ap2-2 inflorescences were treated withDMSO or DEX for 6 hours and GUS expression was determinedby real-time RT-PCR. DEX induction caused a decrease in GUSmRNA levels in KB31 but not in KB31AB (Fig. 5E,F). Therefore,AP2 represses AG through the two AP2R2-binding sites in vivo.

AP2 binds AG 2nd intron in vivoTo test whether AP2 is associated with the AG 2nd intron in vivo, weperformed ChIP assays using anti-AP2 antibodies (Mlotshwa et al.,2006). The antibodies were directed against a C-terminal region ofAP2 that is predicted to be absent in the ap2-2 mutant. From a ChIP-seq experiment conducted with these antibodies on wholeinflorescences, genome-wide AP2-binding sites were uncovered(Yant et al., 2010). The ChIP-seq effort identified a region in the 5�end of the KB31 fragment that was bound by AP2, which we namedregion II (Fig. 6A). This region did not overlap with the regioncontaining our binding sites, AB (Fig. 6A). To specifically test

Fig. 4. The AP2R2-binding sites are essential in restricting AGexpression in floral meristems. Either the wild-type KB31 fragmentor a mutant version lacking the two AP2R2 binding sites (KB31AB)was fused to GUS and GUS expression was evaluated in vivo in T1transgenic lines. Representative images of typical T1 inflorescences areshown. (A,B)Stages 3-4 floral meristems. (C,D)Stages 6-7 flowers.(E,F)Stages 8-12 flowers. (A,C,E) Wild-type KB31. (B,D,F) Mutant KB31lacking both A and B sites (KB31AB).

DEVELO

PMENT

whether AP2 binds the AB region, especially in young flowers, weperformed ChIP experiments with dissected inflorescencescontaining stages 7 and younger flowers and used region II as thepositive control. We were able to find enrichment of AP2 within theAB region as well as region II in two biological replicates (Fig. 6;data not shown). As our ChIP assays showed that AP2 bound bothsites but the genome-wide study did not show enrichment in our ABsite (Yant et al., 2010), we further dissected the II region to test whichpart of that region may be bound by AP2. We divided the region intofour parts (1-4; supplementary material Fig. S8A and Table S1),including a region directly upstream and downstream. We found thatAP2R2 bound part 3 of region II (supplementary material Fig. S8B)and upon further inspection, the 5� end of part 3 contained an AP2R2consensus sequence with one nucleotide change (TTTGTT toTTTGTG). Consistent with our site mutagenesis analysis (Fig 2;supplementary material Fig. S4), AP2R2 is still able to bind theTTTGTG sequence in a medium manner. Binding was not visiblyobserved with this probe (p3) until 400 ng of protein was added(supplementary material Fig. S8, lane 5) when compared with 200ng of the wild-type probe (supplementary material Fig. S8, lane 2).Consistent with all of our previous results, MBP-AP2 was able tobind all four probes (supplementary material Fig. S8C).

1983RESEARCH ARTICLEAP2 binds an AT-rich sequence

The consensus sequence TTTGTT is highlyconserved in BrassicaceaeThe repression of AG expression by AP2 is specific to Arabidopsisas the AP2 orthologs in Antirhinnum, LIPLESS 1 and 2 (LIP1 andLIP2), do not negatively regulate C-class function (Keck et al.,2003). However, it is possible that AP2-mediated restriction of AGexpression is conserved in species closer to Arabidopsis. Thus, wesought to determine whether the AP2R2-binding consensussequence was conserved in Brassicaceae. Using AG 2nd intronsequences from 29 Brassicaceae species (Hong et al., 2003), weperformed multiple sequence alignment as well as a slidingwindow conservation analysis to identify regions in the 2nd intronthat are conserved both in sequence and in position. Although theTTTGTT (or AACAA) motif was present multiple times in each ofthe introns (supplementary material Fig. S10), the A-site (Fig. 3A)was embedded within a short region that was conserved throughoutBrassicaceae both in sequence and in position within the introns,as revealed by the sliding window analysis (supplementary materialFigs S9, S10). The B-site was not found at invariant positionsamong the introns (supplementary material Fig. S10). At the A-site,28 of the 29 Brassicaceae species showed a perfect match to theTTTGTT pattern. The apparent exception appeared to beThysanocarpus (AY253255) with a single nucleotide change in themotif. When the analyses included the 2nd intron of AG homologsfrom Antirrhinum majus (AY935269), Lycopersicon esculentum(AY254705) or Cucumis sativus (AY254702 and AY254704)belonging to Veronicaceae, Solanaceae or Cucurbitaceae,respectively, the divergence in these sequences was too high tocompute reliable multiple sequence alignments of the introns, thusprecluding any conclusions on the conservation of this motifoutside of Brassicaceae.

DISCUSSIONDNA binding specificities of AP2 domain proteinsGenes encoding one or more AP2 DNA-binding domains arecategorized under five subfamilies: DREB-like, ERF-like, RAV-like, AP2-like and others (Sakuma et al., 2002). The AP2-likesubfamily, which can be further divided into two lineages, ANT andeuAP2, is the only subfamily that contains two AP2 domains.

Fig. 5. Induction of AP2 leads to decreased AG mRNA levels inyoung floral meristems. (A)An uninduced 35S::AP2m3-GR flower.(B)A 35S::AP2m3-GR flower after continuous DEX induction. (C)RT-PCR analysis of AG mRNA levels from 35S::AP2m3-GR ap2-2inflorescences after 6 hours of treatment with CHX alone (6C) or CHX+ DEX (6DC). Thirty-three and 25 cycles of PCR were performed for AGand UBQ5 (an internal control), respectively. (D)Real-time RT-PCRanalysis of three biological replicates of 35S::AP2m3-GR ap2-2 C- orDC-treated inflorescence tissue. For both C and D, dissectedinflorescences containing stages 7 and younger flowers were used.(E)Real-time RT-PCR analysis of KB31 35S::AP2m3-GR ap2-2 C- or DC-treated inflorescence tissue. (F)Real-time RT-PCR analysis of KB31AB35S::AP2m3-GR ap2-2 C- or DC-treated inflorescence tissue. Primerscorresponding to two regions in the GUS-coding region (GUSp1 andGUSp3) were used for the real-time RT-PCR. For E and F, threebiological replicates were performed and yielded similar results. Onerepresentative image is shown. Error bars represent s.d. from threetechnical replicates.

Fig. 6. AP2 binds AG 2nd intron in vivo. (A)Diagram of a partial AGgenomic region. Dark-gray boxes indicate the 2nd and 3rd exons,whereas the line represents the 2nd intron. (B)AP2 is associated withthe AG 2nd intron in vivo. ChIP was performed on dissectedinflorescence tissue from wild type and ap2-2 using anti-AP2antibodies. Real-time PCR was performed on input and bound fractionsand the percentage of bound DNA relative to input was calculated. TheAB region, as well as region II (a positive control), was found to beenriched in the Ler versus ap2-2 sample. eIF4A1 was used as a negativecontrol. Error bars represent s.d. from three technical replicates.

DEVELO

PMENT

1984

Single AP2 domain containing proteins of the other subfamiliesbind to highly specific, mostly GC-rich sequence motifs (Ohme-Takagi and Shinshi, 1995; Stockinger et al., 1997; Hao et al., 1998;Liu et al., 1998). Only the target sequence of a single member ofthe AP2-like subfamily, ANT, has been reported. ANT binds a longand loose consensus sequence that is also GC rich (Nole-Wilsonand Krizek, 2000). In contrast to conventional GC-rich targetsequences of these characterized AP2-domain proteins, AP2R2 ishighly specific for the AT-rich consensus sequences TTTGTT orAACAAA. The AP2 domain in RAV1 also binds a non-GC richsequence CAACA (Kagaya et al., 1999). Two AP2 domain proteinsfrom Plasmodium were found to bind the consensus sequencesTGCATGCA and GTGCAC, which are different from the targetsequences of all plant AP2 domain proteins characterized to date(De Silva et al., 2008). Collectively, these studies show that AP2domains have wide ranging target specificities. Consistent withthis, the AP2 domain of AtERF1 and the AP2 domains of ANTappear to use largely non-conserved amino acids for DNA binding(Allen et al., 1998; Krizek, 2003) (supplementary material Fig.S11). The fact that AP2R1 does not appreciably bind any DNAsequences in vitro raises the possibility that some AP2 domainsfunction in processes other than DNA binding.

It is useful to compare and contrast the DNA-bindingspecificities of ANT and AP2 as representatives of the two lineageswithin the AP2-like subfamily. In vitro selection of DNA sequencesbound by ANT-AP2R1R2 led to the identification of a longconsensus sequence (Nole-Wilson and Krizek, 2000; Krizek,2003). In our study, we found that AP2R1R2 bound DNA in vitro,but no consensus sequence could be identified. We note that bothANT-AP2R1R2 and AP2R1R2 have poor DNA-bindingspecificities (as exemplified by the loose ANT consensus sequenceand the lack of obvious consensus motifs from AP2R1R2-boundsequences). We also note that ANT-AP2R1R2 and AP2R1R2 havedifferences in their binding preferences. Although ANT-AP2R1R2binds GC-rich sequences, AP2R1R2 probably prefers the TTTGTTor AACAAA motif as this motif was in 10 out of 25 clones fromthe SAAB assay. Moreover, not all clones from the SAAB assaywere GC rich. Consistently as well, this could be explained by theaddition of 10 amino acids in ANTR1 versus AP2R1 and a singleamino acid addition in ANTR2 versus AP2R2 (supplementarymaterial Fig. S11).

AP2 DNA-binding specificities in vivoAP2 full-length protein was able to bind all probes that were testedin vitro (Fig. 3; supplementary material Figs S7, S8). This led usto question the specificity of AP2 DNA binding in vivo, especiallyin relation to its biological functions. AP2 has diverse biologicalfunctions such as seed development (Jofuku et al., 2005; Ohto etal., 2009), shoot apical meristem maintenance (Würschum et al.,2006), control of floral timing (Yant et al., 2010), preventingreplum overgrowth during fruit development (Ripoll et al., 2011),establishment of floral meristem identity (Schultz and Haughn,1993; Shannon and Meeks-Wagner, 1993), floral organspecification (Bowman et al., 1989; Kunst et al., 1989) and theregulation of homeotic gene expression (Drews et al., 1991).Perhaps the lack of strong inherent DNA-binding specificitiesunderlies the diverse biological roles of AP2. Whole-genome ChIP-seq experiments identified more than 2000 sites that are bound byAP2 in vivo (Yant et al., 2010), highlighting the potential of AP2in influencing the expression of a large number of genes. However,computational analyses failed to uncover a consensus sequence thatis enriched in regions bound by AP2 in vivo (data not shown). In

RESEARCH ARTICLE Development 139 (11)

addition, owing to its AT-rich nature, we were not able to state thatthe AP2R2-binding site was statistically significant among thetargets. However, it is interesting to note that in the sum sequencespace of the 2275 bound sites there are 445 instances of AACAAAand 473 of TTTGTT (Yant et al., 2010). The lack of ability to findan AP2 consensus sequence could be reflective of its diverse rolesin development. In addition, the discrepancy in the two sites (II andAB) that we found in the two ChIP experiments (Yant et al., 2010)(this study) could be due to tissue differences.

Despite the large number of in vivo binding sites, AP2 is stillselective in its DNA binding in vivo, in contrast to its largely non-specific DNA binding in vitro. One potential mechanismunderlying the in vivo specificity is that it might be conferred byother DNA-binding proteins that interact with AP2. In thisscenario, the largely non-specific DNA binding by AP2 enhancesthe binding of other transcription factors at specific sites. Thepromiscuous binding of MBP-AP2 to all DNA sequences in vitrolends itself to this hypothesis as it could be feasible that AP2 full-length itself, as a regulator of diverse functions, would havespecific binding abilities, depending on its protein-binding partners,that may modulate its activity in vivo. Another potentialmechanism is that other factors interact with AP2R1 to allowAP2R2 to specifically interact with DNA. We prefer thisexplanation as the AP2R2-binding sites in the AG 2nd intron areindeed important for the function of AP2 in vivo. Moreover, theAP2R2 consensus sequence was recovered in 10 out of 25 clonesin the AP2-R1R2 SAAB assay, suggesting that there is someinherent affinity of the AP2 domains for the TTTGTT orAACAAA consensus sequence. Both scenarios may occur in vivo,in which case the AP2R2 consensus sequence would only bepresent at some of the in vivo AP2-binding sites.

AP2 directly regulates AGAP2 has long been known to be essential in establishing theinner two whorl-specific pattern of AG expression (Drews et al.,1991). In ap2 loss-of-function mutants, AG expression expandsinto the outer two whorls. Using the GUS reporter system,elements responsive to AP2 regulation have been mapped to atleast two regions in the AG 2nd intron (Bomblies et al., 1999;Deyholos and Sieburth, 2000). However, it was not knownwhether AP2 regulates AG expression directly. We found that a750 bp AP2-responsive region contains two AP2R2 consensussequences. Site-directed mutagenesis experiments indicated thatthe two sites were important for the negative regulation of AGby AP2. In addition, we found that this negative regulation wasdirect through an inducible system (AP2m3-GR) as well as ChIPexperiments. Therefore, AP2 is a direct, negative regulator ofAG.

Our data also suggest that AP2 represses AG most effectivelyduring early stages of flower development. Initially, when AP2m3-GR whole inflorescences (composed of both young and oldflowers) were used in the induction experiments, no obviouschanges in AG mRNA levels were seen. However, upon micro-dissection of the inflorescences after induction to retain onlyflowers of stages 7 and younger, we observed a 50% decrease inAG mRNA levels upon AP2 induction. It is feasible that AP2 onlynegatively regulates AG during early stages of flower developmentas it has been shown that a myriad of other genes, such as CURLYLEAF (CLF), LEUNIG, SEUSS and RABBIT EARS also negativelyregulate AG (Goodrich et al., 1997; Liu et al., 1998; Franks et al.,2002; Krizek et al., 2006). It is possible that in the outer twowhorls, AP2 establishes the initial repression of AG, and other D

EVELO

PMENT

1985RESEARCH ARTICLEAP2 binds an AT-rich sequence

mechanisms, such as CLF-mediated histone modifications, areresponsible for the maintenance of the repressed state throughoutflower development.

In addition, it has been shown that AP2 acts through differentregions of the AG 2nd intron at different developmental timepoints(Bomblies et al., 1999; Deyholos and Sieburth, 2000). Thus, it isfeasible that region II and the AB site may both be important atdifferent time-points. In the Bomblies study, KB31 is sufficient toconfer AG expression and contain elements by which AP2negatively regulates AG, and this construct has both region II andthe AB site. Constructs that do not have region II or the AB site didnot provide much information because, although they did notrespond to the loss of AP2, there was no basal AG expression(KB28 encompasses p3 and the A site only). In this study, we showthat the AB site is functional, as loss of AB resulted in expansionof GUS expression (Fig. 4). Interestingly, AP2R2 was able to bindto both region II and the AB site, albeit at a higher affinity with theAB site.

The ‘A’-site is highly conserved in BrassicaceaeThe euAP2 lineage predates the divergence of gymnosperms andangiosperms, but the biological functions of AP2 and its orthologsdiffer amongst flowering plants (reviewed in Litt, 2007). InArabidopsis, AP2 specifies perianth identities and restricts C-function to the inner two whorls. However, characterized AP2orthologs from Antirrhinum and petunia do not appear to share therole of AP2 in flower development (Maes et al., 2001; Keck et al.,2003). For example, LIP1 and LIP2, AP2 orthologs in Antirrhinum,promote sepal identities but do not control petal identity or restrictthe expression of PLENA (C-class gene) (Keck et al., 2003). Infact, mutations with ectopic C function in the outer whorls inAntirrhinum and petunia map to a microRNA, miR169 (Cartolanoet al., 2007). Interestingly, the petunia ortholog of LIP/AP2,PhAP2A, was able to rescue the ap2-1 mutant when expressed inArabidopsis (Maes et al., 2001). The ability of the petunia AP2protein to regulate AG in the Arabidopsis context suggests that theDNA-binding properties of the petunia AP2 are similar to those ofArabidopsis AP2 and implies that divergence in C-class regulatorysequences or in AP2-interacting proteins may be responsible forthe divergence in the ability of AP2 to regulate C-class genes.

In this study, we show that AP2R2 recognizes an AT-richmotif in vitro and that two such motifs within the AG 2nd intronmediate the regulation of AG by AP2 in vivo. Given the ATrichness of introns, this sequence motif is present multiple timesin the introns of AG and AG orthologs from other species. Thepositions of the motifs relative to other transcription factorbinding sites may influence the ability of AP2 to act upon them.We show that the A site recognized by AP2R2 in the AG 2ndintron is conserved both in sequence and in position inBrassicaceae. This implies that AP2-mediated regulation of C-class gene expression is conserved in this family. Although themotif is present in the 2nd introns of AG orthologs from non-Brassicaceae species, the overall large divergence in 2nd intronsequence precluded confident alignments to determine whetherthe positions of the motifs are conserved.

AcknowledgementsWe are grateful to Drs Aman Husbands, Harley Smith and Gina Maduro foradvice on SAAB assays and EMSAs. We thank Xiaofeng Cao from the Instituteof Genetics and Developmental Biology, Chinese Academy of Sciences for thepmCSG7 XF0510 MBP-LIC vector and Patricia Springer for sharing plasmidsand equipment. We thank Lijuan Ji, Shengben Li, Rae Eden Yumul andYuanyuan Zhao for comments on the manuscript.

FundingThe work was supported by grants from the National Institutes of Health[GM61146], the Howard Hughes Medical Institute and the Gordon and BettyMore Foundation to X.C. T.D. was supported by a National Science FoundationChemGen IGERT training grant [DGE0504249]. Deposited in PMC for releaseafter 6 months.

Competing interests statementThe authors declare no competing financial interests.

Supplementary materialSupplementary material available online athttp://dev.biologists.org/lookup/suppl/doi:10.1242/dev.077073/-/DC1

ReferencesAllen, M. D., Yamasaki, K., Ohme-Takagi, M., Tateno, M. and Suzuki, M.

(1998). A novel mode of DNA recognition by a beta-sheet revealed by thesolution structure of the GCC-box binding domain in complex with DNA. EMBOJ. 17, 5484-5496.

Bomblies, K., Dagenais, N. and Weigel, D. (1999). Redundant enhancersmediate transcriptional repression of AGAMOUS by APETALA2. Dev. Biol. 216,260-264.

Bowman, J. L., Smyth, D. R. and Meyerowitz, E. M. (1989). Genes directingflower development in Arabidopsis. Plant Cell 1, 37-52.

Bowman, J. L., Smyth, D. R. and Meyerowitz, E. M. (1991). Geneticinteractions among floral homeotic genes of Arabidopsis. Development 112, 1-20.

Broitman-Maduro, G., Maduro, M. F. and Rothman, J. H. (2005). Thenoncanonical binding site of the MED-1 GATA factor defines differentiallyregulated target genes in the C. elegans mesendoderm. Dev. Cell 8, 427-433.

Cartolano, M., Castillo, R., Efremova, N., Kuckenberg, M., Zethof, J., Gerats,T., Schwarz-Sommer, Z. and Vandenbussche, M. (2007). A conservedmicroRNA module exerts homeotic control over Petunia hybrida andAntirrhinum majus floral organ identity. Nat. Genet. 39, 901-905.

Chen, X. (2004). A microRNA as a translational repressor of APETALA2 inArabidopsis flower development. Science 303, 2022-2025.

Dalmay, T., Hamilton, A., Rudd, S., Angell, S. and Baulcombe, D. C. (2000).An RNA-dependent RNA polymerase gene in Arabidopsis is required forposttranscriptional gene silencing mediated by a transgene but not by a virus.Cell 101, 543-553.

De Silva, E. K., Gehrke, A. R., Olszewski, K., Leon, I., Chahal, J. S., Bulyk, M.L. and Llinas, M. (2008). Specific DNA-binding by apicomplexan AP2transcription factors. Proc. Natl. Acad. Sci. USA 105, 8393-8398.

Deyholos, M. K. and Sieburth, L. E. (2000). Separable whorl-specific expressionand negative regulation by enhancer elements within the AGAMOUS secondintron. Plant Cell 12, 1799-1810.

Drews, G. N., Bowman, J. L. and Meyerowitz, E. M. (1991). Negativeregulation of the Arabidopsis homeotic gene AGAMOUS by the APETALA2product. Cell 65, 991-1002.

Franks, R. G., Wang, C., Levin, J. Z. and Liu, Z. (2002). SEUSS, a member of anovel family of plant regulatory proteins, represses floral homeotic geneexpression with LEUNIG. Development 129, 253-263.

Gomez-Mena, C., de Folter, S., Costa, M. M., Angenent, G. C. andSablowski, R. (2005). Transcriptional program controlled by the floral homeoticgene AGAMOUS during early organogenesis. Development 132, 429-438.

Goodrich, J., Puangsomlee, P., Martin, M., Long, D., Meyerowitz, E. M. andCoupland, G. (1997). A Polycomb-group gene regulates homeotic geneexpression in Arabidopsis. Nature 386, 44-51.

Hao, D., Ohme-Takagi, M. and Sarai, A. (1998). Unique mode of GCC boxrecognition by the DNA-binding domain of ethylene-responsive element-bindingfactor (ERF domain) in plant. J. Biol. Chem. 273, 26857-26861.

Hong, R. L., Hamaguchi, L., Busch, M. A. and Weigel, D. (2003). Regulatoryelements of the floral homeotic gene AGAMOUS identified by phylogeneticfootprinting and shadowing. Plant Cell 15, 1296-1309.

Husbands, A., Bell, E. M., Shuai, B., Smith, H. M. and Springer, P. S. (2007).LATERAL ORGAN BOUNDARIES defines a new family of DNA-bindingtranscription factors and can interact with specific bHLH proteins. Nucleic AcidsRes. 35, 6663-6671.

Immink, R. G., Kaufmann, K. and Angenent, G. C. (2010). The ‘ABC’ of MADSdomain protein behaviour and interactions. Semin. Cell Dev. Biol. 21, 87-93.

Ito, T., Wellmer, F., Yu, H., Das, P., Ito, N., Alves-Ferreira, M., Riechmann, J. L.and Meyerowitz, E. M. (2004). The homeotic protein AGAMOUS controlsmicrosporogenesis by regulation of SPOROCYTELESS. Nature 430, 356-360.

Jofuku, K. D., den Boer, B. G., Van Montagu, M. and Okamuro, J. K. (1994).Control of Arabidopsis flower and seed development by the homeotic geneAPETALA2. Plant Cell 6, 1211-1225.

Jofuku, K. D., Omidyar, P. K., Gee, Z. and Okamuro, J. K. (2005). Control ofseed mass and seed yield by the floral homeotic gene APETALA2. Proc. Natl.Acad. Sci. USA 102, 3117-3122. D

EVELO

PMENT

1986 RESEARCH ARTICLE Development 139 (11)

Kagaya, Y., Ohmiya, K. and Hattori, T. (1999). RAV1, a novel DNA-bindingprotein, binds to bipartite recognition sequence through two distinct DNA-binding domains uniquely found in higher plants. Nucleic Acids Res. 27, 470-478.

Keck, E., McSteen, P., Carpenter, R. and Coen, E. (2003). Separation of geneticfunctions controlling organ identity in flowers. EMBO J. 22, 1058-1066.

Kim, S., Soltis, P. S., Wall, K. and Soltis, D. E. (2006). Phylogeny and domainevolution in the APETALA2-like gene family. Mol. Biol. Evol. 23, 107-120.

Krizek, B. A. (2003). AINTEGUMENTA utilizes a mode of DNA recognition distinctfrom that used by proteins containing a single AP2 domain. Nucleic Acids Res.31, 1859-1868.

Krizek, B. A. and Fletcher, J. C. (2005). Molecular mechanisms of flowerdevelopment: an armchair guide. Nat. Rev. Genet. 6, 688-698.

Krizek, B. A., Lewis, M. W. and Fletcher, J. C. (2006). RABBIT EARS is a second-whorl repressor of AGAMOUS that maintains spatial boundaries in Arabidopsisflowers. Plant J. 45, 369-383.

Kunst, L., Klenz, J. E., Martinez-Zapater, J. and Haughn, G. W. (1989). AP2gene determines the identity of perianth organs in flowers of Arabidopsisthaliana. Plant Cell 1, 1195-1208.

Litt, A. (2007). An evaluation of A-funciton: evidence from the APETALA1 andAPETALA2 gene lineages. Int. J. Plant Sci. 168, 73-91.

Liu, Q., Kasuga, M., Sakuma, Y., Abe, H., Miura, S., Yamaguchi-Shinozaki, K.and Shinozaki, K. (1998). Two transcription factors, DREB1 and DREB2, withan EREBP/AP2 DNA binding domain separate two cellular signal transductionpathways in drought- and low-temperature-responsive gene expression,respectively, in Arabidopsis. Plant Cell 10, 1391-1406.

Maes, T., Van de Steene, N., Zethof, J., Karimi, M., D’Hauw, M., Mares, G.,Van Montagu, M. and Gerats, T. (2001). Petunia Ap2-like genes and their rolein flower and seed development. Plant Cell 13, 229-244.

Mathieu, J., Yant, L. J., Murdter, F., Kuttner, F. and Schmid, M. (2009).Repression of flowering by the miR172 target SMZ. PLoS Biol. 7, e1000148.

Mlotshwa, S., Yang, Z., Kim, Y. and Chen, X. (2006). Floral patterning defectsinduced by Arabidopsis APETALA2 and microRNA172 expression in Nicotianabenthamiana. Plant Mol. Biol. 61, 781-793.

Morgan, M., Anders, S., Lawrence, M., Aboyoun, P., Pages, H. andGentleman, R. (2009). ShortRead: a bioconductor package for input, qualityassessment and exploration of high-throughput sequence data. Bioinformatics25, 2607-2608.

Morgenstern, B. (2004). DIALIGN: multiple DNA and protein sequence alignmentat BiBiServ. Nucleic Acids Res. 32, W33-W36.

Mourrain, P., Beclin, C., Elmayan, T., Feuerbach, F., Godon, C., Morel, J. B.,Jouette, D., Lacombe, A. M., Nikic, S., Picault, N. et al. (2000). ArabidopsisSGS2 and SGS3 genes are required for posttranscriptional gene silencing andnatural virus resistance. Cell 101, 533-542.

Nole-Wilson, S. and Krizek, B. A. (2000). DNA binding properties of theArabidopsis floral development protein AINTEGUMENTA. Nucleic Acids Res. 28,4076-4082.

Ohme-Takagi, M. and Shinshi, H. (1995). Ethylene-inducible DNA bindingproteins that interact with an ethylene-responsive element. Plant Cell 7, 173-182.

Ohto, M. A., Floyd, S. K., Fischer, R. L., Goldberg, R. B. and Harada, J. J.(2009). Effects of APETALA2 on embryo, endosperm, and seed coatdevelopment determine seed size in Arabidopsis. Sex. Plant Reprod. 22, 277-289.

Okamuro, J. K., Caster, B., Villarroel, R., Van Montagu, M. and Jofuku, K. D.(1997). The AP2 domain of APETALA2 defines a large new family of DNAbinding proteins in Arabidopsis. Proc. Natl. Acad. Sci. USA 94, 7076-7081.

R Development Core Team (2010). R: A language and environment for statisticalcomputing. Vienna, Austria.

Riechmann, J. L. and Meyerowitz, E. M. (1998). The AP2/EREBP family of planttranscription factors. Biol. Chem. 379, 633-646.

Ripoll, J. J., Roeder, A. H., Ditta, G. S. and Yanofsky, M. F. (2011). A novel rolefor the floral homeotic gene APETALA2 during Arabidopsis fruit development.Development 138, 5167-5176.

Sablowski, R. W. and Meyerowitz, E. M. (1998). A homolog of NO APICALMERISTEM is an immediate target of the floral homeotic genesAPETALA3/PISTILLATA. Cell 92, 93-103.

Sakuma, Y., Liu, Q., Dubouzet, J. G., Abe, H., Shinozaki, K. and Yamaguchi-Shinozaki, K. (2002). DNA-binding specificity of the ERF/AP2 domain ofArabidopsis DREBs, transcription factors involved in dehydration- and cold-inducible gene expression. Biochem. Biophys. Res. Commun. 290, 998-1009.

Schultz, E. A. and Haughn, G. W. (1993). Genetic analysis of the floral initiationprocess (FLIP) in Arabidopsis. Development 119, 745-765.

Shannon, S. and Meeks-Wagner, D. R. (1993). Genetic interactions thatregulate inflorescence development in Arabidopsis. Plant Cell 5, 639-655.

Shigyo, M., Hasebe, M. and Ito, M. (2006). Molecular evolution of the AP2subfamily. Gene 366, 256-265.

Sieburth, L. E. and Meyerowitz, E. M. (1997). Molecular dissection of theAGAMOUS control region shows that cis elements for spatial regulation arelocated intragenically. Plant Cell 9, 355-365.

Smith, H. M., Boschke, I. and Hake, S. (2002). Selective interaction of planthomeodomain proteins mediates high DNA-binding affinity. Proc. Natl. Acad.Sci. USA 99, 9579-9584.

Stockinger, E. J., Gilmour, S. J. and Thomashow, M. F. (1997). Arabidopsisthaliana CBF1 encodes an AP2 domain-containing transcriptional activator thatbinds to the C-repeat/DRE, a cis-acting DNA regulatory element that stimulatestranscription in response to low temperature and water deficit. Proc. Natl. Acad.Sci. USA 94, 1035-1040.

Tilly, J. J., Allen, D. W. and Jack, T. (1998). The CArG boxes in the promoter ofthe Arabidopsis floral organ identity gene APETALA3 mediate diverse regulatoryeffects. Development 125, 1647-1657.

Wagner, D., Sablowski, R. W. and Meyerowitz, E. M. (1999). Transcriptionalactivation of APETALA1 by LEAFY. Science 285, 582-584.

Wakem, M. P. and Kohalmi, S. E. (2003). Mutation in the ap2-6 allele causesrecognition of acryptic splice site. J. Exp. Bot. 54, 2655-2660.

Weigel, D. (1995). The APETALA2 domain is related to a novel type of DNAbinding domain. Plant Cell 7, 388-389.

Wierzbicki, A. T., Haag, J. R. and Pikaard, C. S. (2008). Noncoding transcriptionby RNA polymerase Pol IVb/Pol V mediates transcriptional silencing ofoverlapping and adjacent genes. Cell 135, 635-648.

William, D. A., Su, Y., Smith, M. R., Lu, M., Baldwin, D. A. and Wagner, D.(2004). Genomic identification of direct target genes of LEAFY. Proc. Natl. Acad.Sci. USA 101, 1775-1780.

Wuitschick, J. D., Lindstrom, P. R., Meyer, A. E. and Karrer, K. M. (2004).Homing endonucleases encoded by germ line-limited genes in Tetrahymenathermophila have APETALA2 DNA binding domains. Eukaryot. Cell 3, 685-694.

Würschum, T., Gross-Hardt, R. and Laux, T. (2006). APETALA2 regulates thestem cell niche in the Arabidopsis shoot meristem. Plant Cell 18, 295-307.

Yant, L., Mathieu, J., Dinh, T. T., Ott, F., Lanz, C., Wollmann, H., Chen, X. andSchmid, M. (2010). Orchestration of the floral transition and floral developmentin Arabidopsis by the bifunctional transcription factor APETALA2. Plant Cell 22,2156-2170.

Yuda, M., Iwanaga, S., Shigenobu, S., Mair, G. R., Janse, C. J., Waters, A. P.,Kato, T. and Kaneko, I. (2009). Identification of a transcription factor in themosquito-invasive stage of malaria parasites. Mol. Microbiol. 71, 1402-1414.

DEVELO

PMENT