BPC transcription factors and a Polycomb Group protein confine the expression of the ovule identity gene SEEDSTICK in Arabidopsis

This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1111/tpj.14673 This article is protected by copyright. All rights reserved DR VERONICA GREGIS (Orcid ID : 0000-0003-1876-9849) Article type : Original Article BPC transcription factors and a Polycomb Group protein confine the expression of the ovule identity gene SEEDSTICK in Arabidopsis Rosanna Petrella1, Francesca Caselli1, Irma Roig-Villanova1,2, Valentina Vignati1, Matteo Chiara1, Ignacio Ezquer1, Luca Tadini1, Martin M. Kater1, Veronica Gregis1* 1 Università Degli Studi di Milano, Dipartimento di Bioscienze. Via Celoria 26, 20133, Milan, Italy. 2 Department of Agri-Food Engineering and Biotechnology, Barcelona School of Agricultural Engineering, UPC. Esteve Terrades 8, Building 4, 08860 Castelldefels, Spain. Contact information: *Email of corresponding author Veronica Gregis: [email protected] Running title: BPCs, SVP and LHP1 confine STK expression Keywords: MADS-box, BPCs, homeotic genes, STK, LHP1, PRC, Arabidopsis thaliana, transcription factors Accepted Article This article is protected by copyright. All rights reserved Summary The BASIC PENTACYSTEINE (BPC) GAGA (C-box) binding proteins belong to a small plant transcription factor family. We previously reported that BPCs of class I bind directly to C-boxes in the SEEDSTICK (STK) promoter and the mutagenesis of these cis-elements affects STK expression in the flower. The MADS-domain factor SHORT VEGETATIVE PHASE (SVP) is another key regulator of STK. Direct binding of SVP to CArG-boxes in the STK promoter are required to repress its expression during the first stages of flower development. Here we show that BPCs of class II directly interact with SVP and that MADS-domain binding sites in the STK promoter region are important for the correct spatial and temporal expression of this homeotic gene. Furthermore, we show that BPCs of class I and II act redundantly to repress STK expression in the flower, most likely by recruiting TERMINAL FLOWER 2/LIKE HETEROCHROMATIN PROTEIN 1 (TFL2/LHP1) and mediating the establishment and the maintenance of H3K27me3 repressive marks on the DNA. We investigate the role of LHP1 in the regulation of STK expression. Besides providing a better understanding of the role of BPC transcription factors in the regulation of STK expression, our results suggest the existence of a more general regulatory complex composed of BPCs, MADSdomain factors and PRCs, that cooperate to regulate gene expression in reproductive tissues. We believe that our data along with the molecular model herein described could provide significant insights for a more comprehensive understanding of gene regulation in plants. Accepted Article This article is protected by copyright. All rights reserved Introduction Transcription factors (TFs) are regulators of gene expression; they act at multiple levels to orchestrate developmental processes. TFs bind specific DNA sequences and they can cooperate through genetic and epigenetic mechanisms. TFs act in multimeric complexes that can include members of different TFs families and other proteins. The composition of these complexes determines their binding specificity and their activity on target gene regulation (Martinez and Rao, 2012). Although in the last decades different classes of plant TFs have been characterised, the molecular mechanisms by which they act and the complexes they are part of, are yet to be fully understood. Recently, a new class of transcription factors, named BASIC PENTACYSTEINE/ BARLEY B RECOMBINANT (BPC/BBR), has been identified (Santi et al., 2003). BPCs bind the RGARAGRRA consensus site, also called GAGA or C-box, to regulate their target genes (Meister et al., 2004; Kooiker et al., 2005; Simonini et al., 2012; Simonini and Kater, 2014; Hecker et al., 2015; Mu et al., 2017; Shanks et al., 2018; Theune et al., 2019; Roscoe et al., 2019; Wu et al., 2019). BPCs have been described in different plant species including monocots (Oryza sativa (rice) and Hordeum vulgare (barley)) and dicots (Glycine max (soy-bean) and Arabidopsis thaliana) (Sangwan and O’Brian, 2002; Santi et al., 2003; Kooiker et al., 2005; Monfared et al., 2011; Berger and Dubreucq, 2012; Simonini et al., 2012; Simonini and Kater, 2014; Hecker et al., 2015; Mu et al., 2017; Xiao et al., 2017; Shanks et al., 2018; Theune et al., 2019; Roscoe et al., 2019; Wu et al., 2019). In Arabidopsis, BPCs are divided into three subfamilies: class I (containing BPC1 to BPC3), class II (containing BPC4 to BPC6), and class III (containing only BPC7) (Meister et al., 2004; Monfared et al., 2011). Except for BPC5, which is a pseudogene, all the other BPCs are ubiquitously expressed. Combinations of multiple bpc mutants show strong phenotypes with a wide range of defects, addressing an important role during plant development (Monfared et al., 2011). Previously, we have identified the MADS-box gene SEEDSTICK (STK) as a direct target of BPCs belonging to the class I (Kooiker et al., 2005; Simonini et al., 2012). STK is specifically expressed during ovule and seed development and has a wide range of functions in these tissues (Favaro et al., 2003; Pinyopich et al., 2003; Brambilla et al., 2007; Losa et al., 2010; Mizzotti et al., 2014; Mendes et al., 2016; Balanzà et al., 2016; Ezquer et al., 2016; Herrera-Ubaldo et al., 2019). Accepted Article This article is protected by copyright. All rights reserved During carpel development, STK expression is confined to placental tissues and ovule primordia; in mature ovules, it is expressed strongly in the funiculus and in integuments that will later form the seed coat (Mizzotti et al., 2014). STK acts redundantly with two other MADS-box factors named SHATTERPROOF 1 (SHP1) and SHATTERPROOF 2 (SHP2) in the determination of ovule identity (Favaro et al., 2003; Pinyopich et al., 2003). BPCs of class I form homoand heterodimers and bind C-boxes in the promoter of STK inducing DNA loop formation (Kooiker et al., 2005). C-boxes are important for STK regulation since mutations in these sequences result in the ectopic expression of the homeotic gene in the flower (Kooiker et al., 2005; Simonini and Kater, 2014). The MADS-domain factor SHORT VEGETATIVE PHASE (SVP) is another key regulator of STK. SVP acts redundantly with APELATA1 (AP1) and AGAMOUS-LIKE24 (AGL24) to repress STK expression during early stages of flower development, by binding directly to its promoter (Simonini et al., 2012; Gregis et al., 2013). Furthermore, BPCs of class I and SVP directly interact to repress STK expression in the floral meristem, and C-boxes are important to facilitate the binding of SVP to the STK promoter region (Simonini et al., 2012). Recently, members of the BPCs family have been shown to be implicated in the recruitment of histone-modifying complexes that can inactivate gene expression, like the Polycomb Repressive Complexes (PRCs) (Hecker et al., 2015; Mu et al., 2017; Xiao et al., 2017; Roscoe et al., 2019; Wu et al., 2019). BPCs of class II directly interact with LHP1, a component of plant PRC1 that is associated with genes marked by trimethylation of histone H3 lysine 27 (H3K27me3) (Hecker et al., 2015). Furthermore, LHP1 also acts as a component of PRC2 to establish H3K27me3 and has a role in maintaining this mark at PRC2 target genes (Zhang et al., 2007b; Derkacheva et al., 2013; Wang et al., 2016). Interestingly, it was demonstrated that SVP can form heterodimers with LHP1 to modulate H3K27me3 deposition on the SEPALLATA3 (SEP3) locus (Liu et al., 2009). Furthermore, BPCs can physically interact with the PRC2 subunit SWINGER (SWN) to repress the expression of their target ABSCISIC ACID INSENSITIVE4 (ABI4) during root development by the trimethylation of Histone H3 Lysine 27 (Mu et al., 2017); moreover a close proximity of BPC6 with VERNALIZATION2 (VRN2) has been reported (Hecker et al., 2015). Here we clarify the molecular mechanisms by which BPCs of class II and SVP act in the regulation of STK expression. We show that MADS-domain binding sequences in the STK promoter region Accepted Article This article is protected by copyright. All rights reserved are important for the correct spatial and temporal expression of the ovule identity gene. Our data indicate that both BPCs of class I and II redundantly control the expression of STK, by modulating the deposition and/or the maintenance of H3K27me3 marks. Our results provide insights into the molecular mechanisms that drive transcription regulation in plants and investigate the involvement of a protein complex in which BPCs, MADS-domain factors and LHP1 can cooperate to orchestrate the expression of homeotic genes during plant development. Results STK is deregulated during flower development in the bpc1-2 bpc2 bpc3 bpc4 bpc6 mutant To gain more insights into the role of class I and class II BPCs in the regulation of STK expression during flower development we generated the bpc1-2 bpc2 bpc3 bpc4 bpc6 quintuple mutant (henceforth called bpcV). In contrast with the previously published quintuple mutant, this mutant includes the bpc1-2 allele which leads to a complete knock-out of the gene (Monfared et al., 2011; Simonini and Kater, 2014). The contribution of class I and II BPCs to the correct regulation of STK was analysed by in-situ hybridisation assays (Figure 1). In wild-type plants, STK expression was confined to ovules and the placenta and was never observed in flowers before stage 8 neither in inflorescences nor in floral meristems (Figure 1a, b). The knock-out of all the BPCs of class I (Figure 1c, d) or class II (Figure 1e, f) did not affect STK expression in the flower. In contrast, in the bpcV mutant the expression of STK was not only observed in ovules and placenta, but also in floral meristems, young flowers and developing petals (Figure 1h). STK expression was also detectable in flower organ primordia (Figure 1g). The in-situ hybridisation controls, using a H4 genespecific probe (confirming the integrity of the tissue) and a STK sense probe (Figure S1; Fobert et al., 1994; Favaro et al., 2003), were performed to confirm the in-situ data. These results clearly demonstrated the redundant role that class I and II BPCs have in the regulation of STK expression during flower development. Phenotypical characterization of the bpcV mutant and 35S:STK lines To further investigate the role of BPCs in plant development, we performed a phenotypical analysis of the bpcV mutant. The quintuple mutant plants were shorter when compared to wildAccepted Article This article is protected by copyright. All rights reserved type plants and were characterized by both vegetative and reproductive defects (Figure 2a, b and Figure S2b and c). The knockout of the five BPC genes caused a drastic phenotype in the siliques. In wild-type, upon successful fertilization, from 3 to 12 days after pollination (dap), the siliques elongate to reach their maximum length. In contrast, in the bpcV mutant no silique elongation was registered (Figure 2b). As also pointed out in Monfared et al. (2011), these results suggest that BPCs are involved in different aspects of plant development. To analyse the phenotypic effects of the deregulation of STK and to compare them with the phenotypes observed in the bpcV mutant, we transformed wild-type plants with a chimeric gene construct in which the CDS of STK was fused to the Cauliflower mosaic virus (CaMV) 35S promoter (Favaro et al., 2003). STK expression was analysed by quantitative Real-Time PCR in three lines, where we could detect statistically significant upregulation of STK expression (Figure S2a). The line that showed the highest upregulation (henceforth called 35S:STK) was propagated and in the next generations used for further analysis. Intriguingly, also this plant was shorter compared to the wild-type (Figure 2a) and showed defects that phenocopy the bpcV mutant, including the silique phenotype (Figure 2b) which is consistent with an upregulation of STK in the BPCV mutant. The MADS-domain factor STK is a master player in ovules and seeds development (Favaro et al., 2003; Pinyopich et al., 2003; Mizzotti et al., 2014; Ezquer et al., 2016). To determine whether the constitutive expression of STK affected seed development, seed area was analysed in 35S:STK plants and bpcV mutants. As a control, wild-type, stk and arf2-8 seeds were used. Our results confirmed that stk had smaller seeds, as previously reported by Pinyopich et al. (2003) whereas arf2-8 seeds were larger (Schruff et al., 2006). Interestingly, both bpcV and 35S:STK plants showed a wider seed area when compared to the wild-type and the stk mutant, even though bpcV mutant seeds were larger than those of 35S:STK (Figure 2c). Our results support the hypothesis that BPCs regulate STK expression in the gynoecium and in seeds. BPCs of class II interact with SVP Accepted Article This article is protected by copyright. All rights reserved Previously, we have shown that class I BPCs act together with SVP in the control of STK expression (Simonini et al., 2012). The in-situ analysis (Figure 1) suggests that BPCs of class II have an important role in regulating STK expression. To understand whether BPCs of class II (BPC4 and BPC6) interact with SVP, different protein interaction assays were performed. We confirmed by yeast two-hybrid assays and bimolecular fluorescence complementation assays (BiFC) in tobacco leaves (Nicotiana benthamiana) that BPC4 and BPC6 can form homoand heterodimers (Wanke et al., 2011; Figure S3a, c). Furthermore. we showed, using yeast twohybrid assays, that both BPC4 and BPC6 can interact with SVP (Figure 3a). To confirm the interactions between SVP and the BPC4 and BPC6 factors, a co-Immunoprecipitation (Co-IP) assay was performed, using SVP-GFP in combination with BPC4-RFP and BPC6-RFP fusion proteins, transiently co-expressed under the control of the Cauliflower mosaic virus (CaMV) promoter in Nicotiana benthamiana leaves. These Co-IP experiments all revealed coprecipitation of the BPC and SVP proteins, respectively (Figures 3b and Figure S3b), suggesting that BPC4 and BPC6 are able to interact with SVP in vivo. Further validation of these results in planta was obtained by BiFC assays in tobacco (Nicotiana benthamiana) leaves. The combination SVP-YFPN BPC4-YFPC showed a clear nuclear interaction between BPC4 and SVP (Figure 3c). All the other combinations that were tested (BPC4-YFPN SVPYFPC, SVP-YFPN BPC6-YFPC, BPC6-YFPN SVP-YFPC) resulted in an interaction in the cytoplasm (Figure 3c, all the control experiments are reported in Figure S3-S5). Although this result was unexpected, Immink et al. (2002) previously showed that some MADS-domain proteins need to dimerise with another MADS-domain factor for their nuclear localisation. SVP interacts with the MADS-domain protein AP1 during floral development and therefore it might facilitate the nuclear location of SVP-BPC dimers (Pelaz et al., 2002; de Folter et al., 2005). To test this hypothesis, we co-expressed SVP-BPC4 and SVP-BPC6 dimers with an AP1-RFP fusion protein in tobacco leaves. As shown in Figure 3d, the presence of AP1 facilitates the nuclear localisation of the BPC4-SVP and BPC6-SVP dimers. To determine whether BPCs of class II could directly interact with AP1, a BIFC interaction assay was performed which showed no interaction between BPCs of class II and AP1, as reported in Figure S3c. Taken together these results clearly show that AP1 is sufficient for the translocation of class II BPCs-SVP heterodimers to the nucleus. Accepted Article This article is protected by copyright. All rights reserved Finally, we also analysed the subcellular localization of BPC4/6-RFP, SVP-GFP and AP1-RFP in tobacco leaves (Nicotiana benthamiana). As shown in Figure S4b, BPC4, BPC6 and AP1 all localized in the nuclei whereas SVP localization was registered in the cytoplasm as well, suggesting that in the BiFC assays described above, SVP is in most combinations tested the critical factor for cytoplasmatic localization. Molecular mechanism of SVP-class I BPCs binding to the regulatory region of STK To clarify the mechanism by which BPCs and SVP interact with the STK regulatory region, we performed a series of Chromatin Immunoprecipitation (ChIP) experiments in different mutant backgrounds. As shown in Figure 4a, SVP binds CArG-boxes that are surrounded by C-boxes in the regulatory region of STK. As previously shown, SVP, AP1 and AGL24 redundantly determine the identity of the floral meristem through direct repression of floral homeotic genes (Gregis et al., 2006; Gregis et al., 2008; Gregis et al., 2009). In fact, in the svp agl24 ap1-12 triple mutant, STK is ectopically expressed in floral meristems and young flowers (Simonini et al., 2012). To determine whether SVP, AP1 and AGL24 are required for BPCs binding to the promoter of STK, three independent ChIP assays using specific antibodies against class I BPCs were performed. The experiments were conducted using svp agl24 ap1-12 triple mutant inflorescences. Furthermore, inflorescences from wild-type and bpc1-2 bpc2 bpc3 triple mutant plants were used as a positive and negative control, respectively. In our ChIP experiments, no enrichment was detected in the svp agl24 ap112 triple mutant in the region containing C-box 12 and the region containing C-box 4 and 5 (region B) (Figure 4b). These results demonstrate that SVP, AP1 and AGL24 are necessary for the binding of class I BPCs to the STK promoter. Subsequently, the role of class I and class II BPCs in the binding of SVP to the promoter of STK was investigated by crossing the bpcV mutant, described above, with pSVP:SVP-GFP svp plants. In subsequent generations, plants homozygous for the svp and bpc1-2 bpc2 bpc3 bpc4 bpc6 mutations containing the pSVP:SVP-GFP construct were selected. ChIP experiments using commercial antibodies against GFP were performed. Inflorescences from pSVP:SVP-GFP svp Accepted Article This article is protected by copyright. All rights reserved plants were used as a positive control, whereas wild-type was used as a negative control. An enrichment was detected when binding to the consensus regions for SVP was tested in the bpcV mutant background (Figure 4c). These results suggest that BPCs of class I and class II are not necessary for SVP binding to STK promoter. Taken together, the results obtained by these ChIP assays are consistent with a model where SVP binds the STK promoter independently of BPCs, whereas BPCs of class I require MADS-domain factors for the correct binding to the STK regulatory region. CArG-boxes drive the correct temporal and spatial expression of STK and are important for SVP and BPCs of class I binding to the promoter of STK To further characterise the role of SVP in STK regulation, we decided to perform a functional characterisation of the CArG-boxes contained in STK regulatory region; these regions were identified based on the MADS-domain factor consensus binding sequences located in the STK locus where SVP binding was detected by ChIP-seq (Gregis et al., 2013). Considering our previous experiments using a STK promoter with mutated C-boxes or CArG-boxes (Simonini et al., 2012; Mendes et al., 2016), we suspected that the 12 CArG-boxes in the regulatory region of STK could be redundant. Therefore, a mutated version of the STK promoter was used in which 11 out of the 12 CArG-boxes were altered, considering the following criteria: (i) preserving the DNA conformation, introducing only 4 to 5 transitions to each consensus; (ii) avoiding the mutation of C-boxes; (iii) preventing the formation of new CArG-boxes (see Table S1). The mutagenized STK promoter was fused to the uidA reporter gene that encodes for beta-glucuronidase (GUS) and the resulting pSTK_CArGm:GUS construct was used to transform Arabidopsis wild-type plants and pSVP:SVP-GFP svp plants. As a positive control, the wild-type STK promoter (pSTK_CArGwt:GUS), which drives specific expression in the placenta and all stages of ovule development, was used (Figure 5a). Out of the 39 plants transformed with the pSTK_CArGwt:GUS construct, 36 showed a correct spatial and temporal expression of the GUS reporter, reflecting the endogenous expression of STK (Figure 5b-d), whereas the other three plants did not show any GUS activity. In contrast, out of 41 plants transformed with the pSTK_CArGm:GUS construct, two plants did not show any GUS activity, whereas 14 (36%) showed strong deregulation of GUS Accepted Article This article is protected by copyright. All rights reserved the normal STK expression domain). The binding of the STK repressive complex might not only be dependent on the association of SVP with the promoter but other cofactors are expected to be involved in recruiting the complex to the DNA. In the absence of SVP binding, the interaction of the complex with the mutated STK promoter could be less stable and influenced by for instance fluctuations in environmental conditions. It remains important to underline that all our experiments demonstrate that BPC and MADSdomain factors are together essential for the correct expression of STK and that, binding of SVP alone is not per se sufficient to repress STK expression in the floral meristems. We previously revealed that BPCs of class I can interact with each other (Simonini et al., 2012); moreover, BPCs of class II form homo and heterodimers with members of class I (Wanke et al., 2011; Simonini et al., 2012 and Figure S3). BPC protein-protein interactions studies suggest that BPC factors of class I and II can act synergistically and redundantly to regulate the expression of their targets as we demonstrated for STK. An example has been provided by Mu et al. (2017), who showed that mutations in BPCs of class I and II increased ABI4 expression in roots. To further investigate the molecular and functional relationships between the MADS-domain factor SVP and BPCs of class II, we tested their ability to form heterodimers in planta. We revealed that SVP interacts with BPC4 and BPC6, but the dimers are mainly retained in the cytoplasm. An interesting observation was that the MADS-domain protein AP1, an interactor of SVP, facilitated the co-localisation of SVP-BPC4 (and BPC6) to the nucleus. These data further clarify the role of AP1 in the regulation of STK (Simonini et al., 2012). The Role of BPCs during seed development Pinyopich et al. (2003) reported that STK has also a role during seed development since the stk mutant presented smaller seeds compared to the wild-type. In this work we found that BPCs can restrict the expression of STK in certain spatiotemporal window since in the bpcV mutant the expression of STK was extended to other tissues in floral organs. The analysis of the 35S:STK line here presented, further explored the effects of the deregulation of STK throughout reproductive development. The defects in seeds size registered in the bpcV and 35S:STK suggest that BPCs might control STK expression later during Accepted Article This article is protected by copyright. All rights reserved development, in seeds. The observation that in the bpcV mutant seeds are bigger than in 35S:STK lines, suggests an addictive role of BPCs during seed development. It will be interesting to further investigate the role of BPCs in such an important aspect of plant development which has enormous implications in agronomical species. BPCs of class II and SVP recruit LHP1 for the regulation of STK Farkas et al. (1994) have first characterised the GAGA Associated Factor of Drosophila melanogaster (dGAFs). Even though GAFs and BPCs are phylogenetically unrelated, they present several similarities. BPCs can bind to (GA)n sequences (Berger and Dubreucq, 2012) to control the expression of their targets (Meister et al., 2004; Berger et al., 2011; Simonini et al., 2012; Simonini and Kater, 2014; Mu et al., 2017; Theune et al., 2019; Roscoe et al., 2019; Wu et al., 2019). They also present a highly conserved zinc finger like DNA-binding domain, similar to Trl of Drosophila (Wanke et al., 2011). Interestingly, cooperative binding of BPC1 proteins to GA-rich motifs in the STK promoter region leads to condensation and looping of DNA (Kooiker et al., 2005), similar to what has been described for dGAF from Drosophila. Recent works in Arabidopsis revealed an intriguing interaction among BPCs and Polycomb group proteins, similar to those described in animals for dGAF, which can cooperate with Polycomb Group factors (PcG) to repress gene expression (Horard et al., 2000). PcG complexes have paramount roles in cell fate determination and differentiation both in plants and in animals. These proteins have been identified in Drosophila more than 40 years ago as key repressors of homeotic genes (Hox) throughout embryonic development (Lewis, 1978). Besides, the sequences and functions of PcG genes are highly conserved between animals and plants. Several publications recently showed that BPCs can interact with proteins belonging to PRC1 and PRC2, suggesting that it could be a mechanism to repress the expression of their target genes (Wanke et al., 2011; Mu et al., 2017; Xiao et al., 2017). Our results provide further insights into the connection between BPCs and PRC members for the regulation of target genes. We suggest that BPCs of class II and SVP recruit LHP1 and act redundantly with the class I members to establish and maintain H3K27me3 repressive mark on the regulatory region of STK in reproductive tissue. In fact, we registered a reduction of Accepted Article This article is protected by copyright. All rights reserved H3K27me3 in the bpcV mutant. In the lhp1 background, we detected increased levels of STK, moreover its expression is localized also in the inflorescences and in the floral meristems as well as in the first floral buds. In contrast to our results in the bpcV, no signal was detected in other floral structures at maturity, thus suggesting that BPCs of class I and II might repress STK expression during flower development also via other mechanisms that do not involve LHP1 activity. Recently, several BPC targets have been discovered. Most of them are also associated with PRC mediated silencing: the KNOX gene BREVIPEDICELLUS (BP) is repressed by BPCs throughout flower development (Simonini et al., 2012). The expression of BP is directly regulated by the recruitment of the EMBRYONIC FLOWER (EMF) complex by ASYMMETRIC LEAVES 1 and 2 (AS1 and AS2), which triggers H3K27me3 deposition (Lodha et al., 2013). BP was also identified in our computational analysis of regions enriched in binding sites for MADS-domain and BPC family members and resulted decorated with H3K27me3 marks (Data S3). Also, FUS3 has recently been characterised as a BPC target (Roscoe et al., 2019; Wu et al., 2019) and already reported to be a target of PRCs (Makarevich et al., 2006; Zhang et al., 2007a; Bouyer et al., 2011; Yang et al., 2013; Xiao et al., 2017). Interestingly, our computational analysis showed that its regulatory region could also be bound by MADS-domain factors, suggesting a possible conserved mechanism for target regulation. The MADS-domain factor SVP interacts with LHP1 and is required to recruit the PRC1 factor to the promoter of SEPALLATA3 (SEP3), acting as a pioneer factor (Liu et al., 2009). In accordance to this hypothesis, H3K27me3 deposition on the SEP3 locus is reduced in lhp1 background. Our ChIP assays confirm binding of LHP1 to the 3’ end of STK, indicating a direct regulation of the homeotic gene by this factor. Interestingly, we previously reported that SVP binds the 3’UTR of many of its targets, among which STK (Gregis et al., 2013). Therefore, it is likely that SVP recruits LHP1 on the STK locus and subsequently repress the expression of the ovule identity gene via PRC2 recruitment, as previously shown for SEP3 (Liu et al., 2009). PRC2 components are required for H3K27me3 deposition to the target locus (Wang et al., 2016). Three different PRC2 complexes regulate plant development by targeting a subset of genes. LHP1 has been reported to associate with several PRC2 members (Derkacheva et al., 2013; Wang et al., 2016), to mediate their recruitment to the target locus. Previously was shown that the EMF Accepted Article This article is protected by copyright. All rights reserved complex played a role in the repression of AGAMOUS (AG) and SEP3 in the flower (Yoshida et al., 2001; Kinoshita et al., 2001; Chanvivattana et al., 2004; Calonje et al., 2008). Notably, Derkacheva et al. (2013) reported that LHP1 is directly associated with the EMF complex. We previously characterised SEP3 and AG, as targets of BPCs and SVP (Gregis et al., 2009; Simonini et al., 2012); as matter of fact, the upregulation of AG registered in the lhp1 single mutant is increased in the lhp1 bpc4 bpc6 triple mutant, confirming a LHP1-class II BPCs interplay in seedlings (Hecker et al., 2015). Considering all these observations it is tempting to speculate that BPCs and SVP might regulate STK expression by the recruitment of LHP1. Then LHP1 as PRC1 member could interact with the EMF complex to mediate the correct deposition of the H3K27me3. Furthermore, LHP1 could assure the maintenance and the spreading of the repressor marks on the STK locus throughout flower development (Figure 9). A general regulatory mechanism in plants Understanding the molecular mechanisms through which BPCs and SVP containing complexes act, is important, since it is likely that the mechanism by which these factors regulate STK can be extended to many other genes during plant development. This is based on the following observations: (i) many genes contain both C-boxes and CArG-boxes in their putative promoter regions; (ii) BPCs are ubiquitously expressed in plants while MADS-domain factors are specifically expressed in all the fundamental developmental stages; and (iii) combination of bpc alleles showed pleiotropic phenotypes (Monfared et al., 2011). Furthermore, Berger et al. (2011) identified three cis-elements required for LEAFY COTYLEDON2 (LEC2) repression: C-boxes, CArGboxes and PRE-like elements, corroborating the idea that the understanding of the synergistic interaction between MADS-domain factors and BPCs is an important key to decode gene regulation in plants. Several key developmental factors that are worth to be tested as putative direct targets of both MADS-domain and BPC factors are reported in Data S3. In fact, they were identified in our computational analysis of regions enriched in binding sites for both MADS and BPC family members. Notably, the analyses of ChIP-seq data available for selected MADS-domain and BPCs in vegetative tissues reveal an highly significant levels of overlap of their binding Accepted Article This article is protected by copyright. All rights reserved profiles in vivo, which suggest a cooperative role of BPCs-MADS factors also during vegetative stages that will be interesting to investigate. The regulatory mechanism through which BPCs act is of course not restricted to Arabidopsis. Several GAGA binding proteins have been discovered in crops and several targets have already been characterised (Sangwan and O’Brian, 2002; Santi et al., 2003; Meister et al., 2004; Gong et al., 2018). Therefore, a better understanding of the mechanisms by which these factors act in Arabidopsis may provide knowledge to be used for future crop improvement. Experimental procedures Plant Material and Growth Conditions Arabidopsis thaliana ecotype Columbia was used in this study; the plants were directly sown on soil and kept under short-day conditions for 2 weeks (22°C, 8 h light and 16 h dark) and then moved to long-day conditions (22°C, 16 h light and 8 h dark). The agl24 svp ap1-12 triple mutant and the pSVP:SVP-GFP svp line were previously described by Gregis et al. (2008; 2009); genotyping of the bpc1-2 bpc2 bpc3 bpc4 bpc6 mutants was done according to Simonini and Kater (2014) and Monfared et al. (2011). Seeds from the lhp1 (previously named tfl2-1 (Larsson et al., 1998)), arf2-8 and stk mutant in Columbia background were obtained from the Nottingham Arabidopsis Stock Centre. Generation of quintuple mutants and marker lines The bpc 1-2 bpc2 bpc3 bpc4 bpc6 quintuple mutant was obtained by crossing the bpc 1-2 bpc2 bpc3 triple mutant (Simonini and Kater, 2014) and bpc4 bpc6 double mutant (Monfared et al., 2011); the pSVP:SVP-GFP svp bpc1-2 bpc2 bpc3 bpc4 bpc6 was obtained crossing the line previously described by Gregis et al. (2009) and the bpcV. Generation of 35S:STK line Arabidopsis plants were transformed with the chimeric gene construct in which the CDS of STK was fused to the Cauliflower mosaic virus (CaMV) 35S promoter (Favaro et al., 2003) using the Agrobacterium tumefaciens–mediated floral dip method (Clough and Bent, 1998). Transformant Accepted Article This article is protected by copyright. All rights reserved plants were sown on MS medium and selected by hygromycin (20 mg/L) resistance; presence of the construct was assessed by genotyping and analysis of STK expression. STK promoter constructs and plant transformation The mutated version of the STK promoter (pSTK_CArGm) was synthesised by Twin Helix. The synthetic DNA fragment, like the wild-type version of the STK promoter, were cloned in pUC57Simple (GenScript). The two fragments were digested with AccI and KpnI and cloned in pDONR207 entry clone (Invitrogen), and successively into pGWB3 binary vector containing the GUS reporter gene. Arabidopsis plants were transformed with these constructs using the Agrobacterium tumefaciens–mediated floral dip method (Clough and Bent, 1998). Transformant plants were sown on MS plates and selected by hygromycin (20 mg/L) resistance; presence of the construct was assessed by PCR. GUS staining GUS assays were performed as described previously by Liljegren et al. (2000). The samples were mounted in lactic acid and subsequently observed using a Zeiss Axiophot D1 microscope equipped with differential interference contrast optics. Images were captured on an Axiocam MRc5 camera (Zeiss) using the Axiovision program (version 4.1). In-situ hybridisation assay Arabidopsis flowers were collected, fixed and embedded in paraffin as described by Huijser et al. (1992). Plant tissue sections were probed with STK antisense RNA, described in Brambilla et al. (2007); STK-sense and H4 histone gene were used as controls (Fobert et al., 1994). Hybridisation and immunological detection were executed as described previously by Coen et al. (1990). ChIP assay ChIP assays were performed as described by Gregis et al. (2009) using for SVP-GFP the commercial antibody GFP:Living Colors full-length (Clontech), and for BPCs of class I a polyclonal Accepted Article This article is protected by copyright. All rights reserved antibody as described by Simonini et al. (2012); HA antibody Anti-HA (Roche) were used as negative control in one of the experiments. Quantitative Real-Time PCR assays were performed to determine the enrichment of the fragments. The detection was performed in triplicate using the iQ SYBR Green Supermix (Bio-Rad) and the Bio-Rad iCycler iQ Optical System (software version 3.0a), with the primers listed in Table S2. ChIP-quantitative Real-Time PCR experiments and relative enrichments were calculated as reported by (Matias-Hernandez et al. (2010). We employed the following formulas to calculate the fold enrichment: dCT.tg = CT.i-CT.tg and dCT.nc =CT.i-CT.nc. Ct.tg is target gene mean value, Ct.i is input DNA mean value, and Ct.nc is ACTIN 7 (negative control) mean value: dCT.tg = CT.i-CT.tg and dCT.nc =CT.i-CT.nc. The propagated error values of these CTs are calculated using dSD.tg = sqrt((SD.i)^2+ (SD.tg^2)/sqrt(n) and dSD.nc = sqrt((SD.i)^2+ (SD.nc^2)/sqrt(n), n = number of replicate per sample. Fold-change over negative control was calculated finding the “delta delta CT” of the target region as follows: ddCT = dCT.tgdCT.nc and ddSD = sqrt((dSD.tg)^2+ (dSD.nc)^2. The transformation to linear “fold-change” values is obtained as follows: FC = 2^(ddCT) and FC.error = ln(2)*ddSD*FC. All the experiments were performed in three biological replicates. ChIP-based analysis of H3K27me3 histone modification For ChIP-based analysis of histone modifications, the following antibodies were used for immunoprecipitation: Anti-H3K27me3 Rabbit Polyclonal Antibody (Merck 07-449) and Rabbit anti-histone H3 (Sigma-Aldrich H0164). 0,8 mg of grinded and fixed material from unfertilized flowers from wild-type and bpcV mutant was collected. ChIP experiments were performed in a modified version of a previously reported protocol (Mizzotti et al., 2014). The quantitative RealTime PCR assay was conducted in triplicate on four different biological replicates, with three technical replicates for each sample, and was performed in a Bio-Rad iCycler iQ optical system (software version 3.0a). Quantitative Real-Time PCR assays were performed on input and immunoprecipitated samples and % of input was calculated. The signal obtained after precipitation with anti-H3K27me3 antibody (as indicated in Figure 6b) was normalized to actin levels. AGAMOUS region was used as a reference as it carries the H3K27me3 mark (Li et al., 2015). Relative enrichment of AT2G22560 was included as negative control for the H3K27me3 mark (Li et al., 2015). Sequences of oligonucleotides used for ChIP analyses are listed in Table S2. Accepted Article This article is protected by copyright. All rights reserved Yeast two-hybrid assay The two-hybrid assays were performed at 28°C in the yeast strain AH109 (Clontech). The coding sequences of BPC4, BPC6 and SVP were cloned into pDONR207 (Life Technologies) and successively transferred to the Gateway vector GAL4 system (pGADT7 and pGBKT7; Clontech). Yeast two-hybrid assays were performed on selective yeast synthetic dropout medium lacking Leu, Trp, Ade, and His supplemented with different concentrations of 3-aminotriazole (1, 2.5, and 5 mM of 3-AT). BiFC assay The BPC4, BPC6 and SVP coding sequences were first cloned into pDONR207 (Life Technologies) and subsequently transferred to the pYFPN43 and pYFPC43 vectors by Gateway recombination; while the AP1 coding sequence was cloned into pDONR207 (Invitrogen) and then transferred to pB7RWG2, purchased from the Flanders Interuniversity Institute for Biotechnology (Gent, Belgium); the previously described formation of VERDANDI-VALKYRIE heterodimers was used as positive control, whereas VERDANDI-VERDANDI combination was used as negative control (Figure S4A; Mendes et al. (2016); all the controls are reported in Figures S3, S4 and S5. BiFC assays were performed injecting Agrobacterium expressing viral suppressor p19/experimental constructs as described by Belda-Palazón et al. (2012). The abaxial surfaces of infiltrated tobacco (Nicotiana benthamiana) leaves were imaged 3 days after inoculation. Co-Immunoprecipitation (Co-IP) Protocol The coding sequences of BPC4, BPC6 and SVP were cloned into pDONR221 and then transferred to pB7RWG2 and pB7FWG2, both purchased from the Flanders Interuniversity Institute for Biotechnology (Gent, Belgium). Nicotiana benthamiana leaves were infiltrated with Agrobacterium tumefaciens, as previously described. 4 days after infiltration, leaf disks (16 mm diameter) were collected and homogenised in 1 ml of immunoprecipitation (IP) buffer (30 mM HEPES-KOH pH 8.0, 200 mM NaCl, 60 mM KOAc, 10 mM MgOAc, 0,5% [v/v] Nonidet P-40 and proteinase inhibitor cocktail [cOmplete™, COEDTAF-RO, Roche]). Samples were incubated in ice for 15 min to allow membrane solubilisation and subjected to a centrifugation step (10 min at Accepted Article This article is protected by copyright. All rights reserved 16,000 g). Supernatants were incubated (2 h, at 4°C) with 20 µl RFP-Trap®_MA (ChromoTek) or GFP-Trap®_MA (ChromoTek). Beads were then washed 3 times for 10 min with 1 ml of IP buffer and eluted with Laemmli sample buffer. Protein samples were fractionated on SDS–PAGE (10% [w/v] acrylamide (Schägger and von Jagow, 1987) and then transferred to polyvinylidene difluoride (PVDF) membranes. Filters were immuno-decorated with specific antibodies; the Coomassie Brilliant Blue (CBB) staining of the gel was performed as loading control. The anti-GFP antibody was purchased from Thermo Fisher Scientific while the anti-RFP antibody was obtained from ChromoTek. Gene expression analysis Quantitative Real-Time PCR experiments were performed using cDNA obtained from inflorescences. Total RNA was extracted using lithium chloride. The Ambion TURBO DNA-free DNase kit was used to remove genomic DNA contaminations, according to the manufacturer’s instructions (http://www.ambion.com/). The ImProm-IITM reverse transcription system (Promega) was used to retrotranscribe the treated RNA. Transcripts were detected using a Sybr Green Assay (iQ SYBR Green Supermix; Bio-Rad) using UBIQUITIN as a reference gene. Assays were done in in triplicate using a Bio-Rad iCycler iQ Optical System (software version 3.0a). The enrichments were calculated normalising the amount of mRNA against housekeeping gene fragments. The expression of different genes was analysed using specific oligonucleotides primers (Table S2). Microscopy and imaging Images of plants, cauline and rosette leaves were acquired using a Canon EOS 6D camera whereas images of siliques were taken using a Leica® MZ 6 stereomicroscope. For in-situ experiments sections were analysed using a Zeiss Axiophot D1 microscope supplied with differential interface contrast (DIC) optics and Axiocam MRc5 camera (Zeiss) using the AXIOVISION program (version 4.4). Scanning Accepted Article This article is protected by copyright. All rights reserved Seed area size were analysed by using SMART-GRAIN software. ANOVA and post-hoc Tukey HSD (honestly significant difference) test were used for wild-type versus other genotypes comparison. Computational analyses DAP-seq peaks data were obtained in the form of narrowpeaks files from from http://neomorph.salk.edu/dap_web/pages/index.php. Narrow-peaks files were concatenated and overlapped genomic regions were merged by the means of the bedtools merge utility. Finally, candidate binding regions showing a positive hit for the majority (that is n/2+1, if profiles for n family members were available) of the members of a family were retained to form the "consensus" family profile. ChIP-seq peaks for BPC1 (GSE84483), BPC6, SVP (GSE54881), SOC1 (GSE45846), FLC (GSE54881) and FLM (GSE48082) were retrieved directly from their respective entries in the GEO database. Data of selected MADS-box ChIP-seq were chosen based on the tissue in which the experiments were performed: seedling or vegetative tissues as for ChIP-seqs available for both BPC1 and BPC6. Intersection of peaks coordinates were performed using the bedtools intersect program (Quinlan and Hall, 2010) using default parameters; peaks with an overlap of 1 bp were considered coincident. The "-u" option was used in order to collapse peaks showing multiple overlaps. Statistical significance of overlaps was assessed by using the hypergeometric distribution. Annotation of selected DAP-seq peaks was performed by the means of the annotatePeaks program from the Homer suite (Heinz et al., 2010) using the reference TAIR10 annotation. Identification of enriched sequence motifs and identification of closely related motifs from publicly available dataset of were performed by the means of the findMotifsGenome utility in Homer. Functional enrichment analyses were performed by using the web interface of the DAVID suite (Huang et al., 2009). Accession numbers Sequence data from this article can be found in the Arabidopsis Genome Initiative or GenBank/EMBL databases under the following accession numbers: STK (AT4G09960), BPC1 (AT2G01930), BPC2 (AT1G14685), BPC3 (AT1G68120), BPC4 (A T2G21240), BPC6 (AT5G42520), AGL24 (AT4G24540), SVP (AT2G22540), AP1 (AT1G69120), LHP1 Accepted Article This article is protected by copyright. All rights reserved Krajewski P, Coupland G. (2015). Combinatorial activities of SHORT VEGETATIVE PHASE and FLOWERING LOCUS C define distinct modes of flowering regulation in Arabidopsis. Genome Biol. 16:31. Khan, A., Fornes, O., Stigliani, A., Gheorghe, M., Castro-Mondragon, J. A., van der Lee, R., Bessy, A., Chèneby, J., Kulkarni, S. R., Tan, G., Baranasic D, Arenillas DJ, Sandelin A, Vandepoele K, Lenhard B, Ballester B, Wasserman WW, Parcy F, Mathelier A. (2018). JASPAR 2018: update of the open-access database of transcription factor binding profiles and its web framework. Nucleic Acids Res. 46:D260– D266. Kinoshita, T., Harada, J. J., Goldberg, R. B., and Fischer, R. L. (2001). Polycomb repression of flowering during early plant development. Proc. Natl. Acad. Sci. U. S. A. 98:14156–61. Kooiker, M., CA, A., A, L., PS, M., L, F., MM, K., and L., C. (2005). BASIC PENTACYSTEINE1, a GA Binding Protein That Induces Conformational Changes in the Regulatory Region of the Homeotic Arabidopsis Gene SEEDSTICK. PLANT CELL ONLINE 17:722–729. Kotake, T., Takada, S., Nakahigashi, K., Ohto, M., and Goto, K. (2003). Arabidopsis TERMINAL FLOWER 2 Gene Encodes a Heterochromatin Protein 1 Homolog and Represses both FLOWERING LOCUS T to Regulate Flowering Time and Several Floral Homeotic Genes. Plant Cell Physiol. 44:555–564. Lafos, M., Kroll, P., Hohenstatt, M. L., Thorpe, F. L., Clarenz, O., and Schubert, D. (2011). Dynamic Regulation of H3K27 Trimethylation during Arabidopsis Differentiation. PLoS Genet. 7:e1002040. Larsson, A. S., Landberg, K., and Meeks-Wagner, D. R. (1998). The TERMINAL FLOWER2 (TFL2) gene controls the reproductive transition and meristem identity in Arabidopsis thaliana. Genetics 149:597–605. Lewis, E. B. (1978). A gene complex controlling segmentation in Drosophila. Nature 276:565–570. Li, C., Chen, C., Gao, L., Yang, S., Nguyen, V., Shi, X., Siminovitch, K., Kohalmi, S. E., Huang, S., Wu, K., et al. (2015). The Arabidopsis SWI2/SNF2 Chromatin Remodeler BRAHMA Regulates Polycomb Function during Vegetative Development and Directly Activates the Flowering Repressor Gene SVP. PLOS Genet. 11:e1004944. Liljegren, S. J., Ditta, G. S., Eshed, Y., Savidge, B., Bowman, J. L., and Yanofsky, M. F. (2000). SHATTERPROOF MADS-box genes control seed dispersal in Arabidopsis. Nature 404:766–770. Liu, C., Xi, W., Shen, L., Tan, C., and Yu, H. (2009). Regulation of Floral Patterning by Flowering Time Accepted Article This article is protected by copyright. All rights reserved Genes. Dev. Cell 16:711–722. Lodha, M., Marco, C. F., and Timmermans, M. C. P. (2013). The ASYMMETRIC LEAVES complex maintains repression of KNOX homeobox genes via direct recruitment of Polycomb-repressive complex2. Genes Dev. 27:596–601. Losa, A., Colombo, M., Brambilla, V., and Colombo, L. (2010). Genetic interaction between AINTEGUMENTA (ANT) and the ovule identity genes SEEDSTICK (STK), SHATTERPROOF1 (SHP1) and SHATTERPROOF2 (SHP2). Sex. Plant Reprod. 23:115–121. Makarevich, G., Leroy, O., Akinci, U., Schubert, D., Clarenz, O., Goodrich, J., Grossniklaus, U., and Köhler, C. (2006). Different Polycomb group complexes regulate common target genes in Arabidopsis. EMBO Rep. 7:947–952. Malley, R. C. O., Carol, S., Song, L., Galli, M., Ecker, J. R., Malley, R. C. O., Huang, S. C., Song, L., Lewsey, M. G., Bartlett, A., et al. (2016). Cistrome and Epicistrome Features Shape the Regulatory DNA Landscape Resource Cistrome and Epicistrome Features Shape the Regulatory DNA Landscape. Cell 165:1280–1292. Martinez, G. J., and Rao, A. (2012). Cooperative Transcription Factor Complexes in Control. Science (80-. ). 338:891–892. Matias-Hernandez, L., Battaglia, R., Galbiati, F., Rubes, M., Eichenberger, C., Grossniklaus, U., Kater, M. M., and Colombo, L. (2010). VERDANDI Is a Direct Target of the MADS Domain Ovule Identity Complex and Affects Embryo Sac Differentiation in Arabidopsis. Plant Cell 22:1702–1715. Meister, R. J., Williams, L. A., Monfared, M. M., Gallagher, T. L., Kraft, E. A., Nelson, C. G., and Gasser, C. S. (2004). Definition and interactions of a positive regulatory element of the Arabidopsis INNER NO OUTER promoter. Plant J. 37:426–438. Mendes, M. A., Guerra, R. F., Castelnovo, B., Silva-Velazquez, Y., Morandini, P., Manrique, S., Baumann, N., Groß-Hardt, R., Dickinson, H., and Colombo, L. (2016). Live and let die: a REM complex promotes fertilization through synergid cell death in Arabidopsis. Development 143:2780–2790. Mizzotti, C., Ezquer, I., Paolo, D., Rueda-Romero, P., Guerra, R. F., Battaglia, R., Rogachev, I., Aharoni, A., Kater, M. M., Caporali, E., et al. (2014). SEEDSTICK is a Master Regulator of Development and Metabolism in the Arabidopsis Seed Coat. PLoS Genet. 10:e1004856. Monfared, M. M., Simon, M. K., Meister, R. J., Roig-Villanova, I., Kooiker, M., Colombo, L., Fletcher, J. C., Accepted Article This article is protected by copyright. All rights reserved and Gasser, C. S. (2011). Overlapping and antagonistic activities of BASIC PENTACYSTEINE genes affect a range of developmental processes in Arabidopsis. Plant J. 66:1020–1031. Mu, Y., Zou, M., Sun, X., He, B., Xu, X., Liu, Y., Zhang, L., and Chi, W. (2017a). BASIC PENTACYSTEINE proteins repress Abscisic Acid INSENSITIVE 4 expression via direct recruitment of the polycombrepressive complex 2 in arabidopsis root development. Plant Cell Physiol. 58:607–621. Mu, Y., Zou, M., Sun, X., He, B., Xu, X., Liu, Y., Zhang, L., and Chi, W. (2017b). Basic Pentacysteine Proteins Repress Abscisic Acid Insensitive4 Expression via Direct Recruitment of the PolycombRepressive Complex 2 in Arabidopsis Root Development. Plant Cell Physiol. 58:pcx006. O’Malley, R. C., Huang, S. C., Song, L., Lewsey, M. G., Bartlett, A., Nery, J. R., Galli, M., Gallavotti, A., and Ecker, J. R. (2016). Cistrome and Epicistrome Features Shape the Regulatory DNA Landscape. Cell 165:1280–1292. Pelaz, S., Gustafson-Brown, C., Kohalmi, S. E., Crosby, W. L., and Yanofsky, M. F. (2002). APETALA1 and SEPALLATA3 interact to promote flower development. Plant J. 26:385–394. Pinyopich, A., Ditta, G. S., Savidge, B., Liljegren, S. J., Baumann, E., Wisman, E., and Yanofsky, M. F. (2003). Assessing the redundancy of MADS-box genes during carpel and ovule development. Nature 424:85–88. Posé, D., Verhage, L., Ott, F., Yant L., Mathieu, J., Angenent G.C., Immink R.G., Schmid M. (2013). Temperature-dependent regulation of flowering by antagonistic FLM variants. Nature 21;503(7476):414-7. Quinlan, A. R., and Hall, I. M. (2010). BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics 26:841–842. Roscoe, T. J., Vaissayre, V., Paszkiewicz, G., Clavijo, F., Kelemen, Z., Michaud, C., Lepiniec, L., Dubreucq, B., Zhou, D.-X., and Devic, M. (2019). Regulation of FUSCA3 Expression During Seed Development in Arabidopsis. Plant Cell Physiol. 60:476–487. Sangwan, I., and O’Brian, M. R. (2002). Identification of a soybean protein that interacts with GAGA element dinucleotide repeat DNA. Plant Physiol. 129:1788–94. Santi, L., Wang, Y., Stile, M. R., Berendzen, K., Wanke, D., Roig, C., Pozzi, C., Müller, K., Müller, J., Rohde, W., et al. (2003). The GA octodinucleotide repeat binding factor BBR participates in the transcriptional regulation of the homeobox gene Bkn3. Plant J. 34:813–826. Accepted Article This article is protected by copyright. All rights reserved Schägger, H., and von Jagow, G. (1987). Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 166:368–79. Schruff, M. C., Spielman, M., Tiwari, S., Adams, S., Fenby, N., and Scott, R. J. (2006). The AUXIN RESPONSE FACTOR 2 gene of Arabidopsis links auxin signalling, cell division, and the size of seeds and other organs. Development 133:251–61. Shanks, C. M., Hecker, A., Cheng, C.-Y., Brand, L., Collani, S., Schmid, M., Schaller, G. E., Wanke, D., Harter, K., and Kieber, J. J. (2018). Role of BASIC PENTACYSTEINE transcription factors in a subset of cytokinin signaling responses. Plant J. 95:458–473. Simonini, S., and Kater, M. M. (2014). Class I BASIC PENTACYSTEINE factors regulate HOMEOBOX genes involved in meristem size maintenance. J. Exp. Bot. 65:1455–1465. Simonini, S., Roig-Villanova, I., Gregis, V., Colombo, B., Colombo, L., and Kater, M. M. (2012). BASIC PENTACYSTEINE Proteins Mediate MADS Domain Complex Binding to the DNA for Tissue-Specific Expression of Target Genes in Arabidopsis. Plant Cell 24:4163–4172. Theune, M. L., Bloss, U., Brand, L. H., Ladwig, F., and Wanke, D. (2019). Phylogenetic Analyses and GAGA-Motif Binding Studies of BBR/BPC Proteins Lend to Clues in GAGA-Motif Recognition and a Regulatory Role in Brassinosteroid Signaling. Front. Plant Sci. 10:466. Turck, F., Roudier, F., Farrona, S., Martin-Magniette, M.-L., Guillaume, E., Buisine, N., Gagnot, S., Martienssen, R. A., Coupland, G., and Colot, V. (2007). Arabidopsis TFL2/LHP1 Specifically Associates with Genes Marked by Trimethylation of Histone H3 Lysine 27. PLoS Genet. 3:e86. Wang, H., Liu, C., Cheng, J., Liu, J., Zhang, L., and He, C. (2016). Arabidopsis Flower and Embryo Developmental Genes are Repressed in Seedlings by Different Combinations of Polycomb Group Proteins in Association with Distinct Sets of Cis-regulatory Elements Advance Access published 2016, doi:10.1371/journal.pgen.1005771. Wanke, D., Hohenstatt, M. L., Dynowski, M., Bloss, U., Hecker, A., Elgass, K., Hummel, S., Hahn, A., Caesar, K., Schleifenbaum, F., et al. (2011). Alanine zipper-like coiled-coil domains are necessary for homotypic dimerization of plant GAGA-factors in the nucleus and nucleolus. PLoS One 6:e16070. Wu, J., Petrella, R., Dowhanik, S., Gregis, V., and Gazzarrini, S. (2019). Spatiotemporal restriction of FUSCA3 expression by class I BPC promotes ovule development and coordinates embryo and Accepted Article This article is protected by copyright. All rights reserved endosperm growth. bioRxiv Advance Access published April 19, 2019, doi:10.1101/612408. Xiao, J., Jin, R., Yu, X., Shen, M., Wagner, J. D., Pai, A., Song, C., Zhuang, M., Klasfeld, S., He, C., et al. (2017). Cis and trans determinants of epigenetic silencing by Polycomb repressive complex 2 in Arabidopsis. Nat. Genet. 49:1546–1552. Yang, J., Lee, S., Hang, R., Kim, S.-R., Lee, Y.-S., Cao, X., Amasino, R., and An, G. (2013). OsVIL2 functions with PRC2 to induce flowering by repressing O s LFL 1 in rice. Plant J. 73:566–578. Yoshida, N., Yanai, Y., Chen, L., Kato, Y., Hiratsuka, J., Miwa, T., Sung, Z. R., and Takahashi, S. (2001). EMBRYONIC FLOWER2, a novel polycomb group protein homolog, mediates shoot development and flowering in Arabidopsis. Plant Cell 13:2471–81. Zhang, X., Clarenz, O., Cokus, S., Bernatavichute, Y. V, Pellegrini, M., Goodrich, J., and Jacobsen, S. E. (2007a). Whole-Genome Analysis of Histone H3 Lysine 27 Trimethylation in Arabidopsis. PLoS Biol. 5:e129. Zhang, X., Germann, S., Blus, B. J., Khorasanizadeh, S., Gaudin, V., and Jacobsen, S. E. (2007b). The Arabidopsis LHP1 protein colocalizes with histone H3 Lys27 trimethylation. Nat. Struct. &Amp; Mol. Biol. 14:869. Figures legends. Figure 1. Mutation of BPCs of class I and class II affects STK expression in the flower. In-situ hybridisation on wild-type [(a) and (b)], bpc1-2 bpc2 bpc3 [(c) and (d)], bpc4 bpc6 [(e) and (f)] and bpcV [(g) and (h)] inflorescences using a STK-specific antisense probe. IM: inflorescence meristem; P: petal; numbers represent flower stages. Scale bars=50 µm. Figure 2. Overexpression of STK affects vegetative and reproductive development. Accepted Article This article is protected by copyright. All rights reserved (a) From left to right: wild-type, bpcV and 35S:STK plants; plants were photographed six weeks after sowing; scale bars=1 cm. (b) Fruit morphology and length in wild type, bpcV and 35S:STK (from top to bottom); scale bars=1.5 mm. (c) Average seeds area size of wild-type, arf2-8 (Schruff et al., 2006), stk (Pinyopich et al., 2001), 35S:STK and bpcV; error bars represent the standard error mean of replicates; ANOVA and post-hoc Tukey HSD (honestly significant difference) test were used, **P < 0.01 for wild-type versus other genotypes comparison. In the lower row, seeds of the analysed genotypes are shown. Figure 3. Class II BPCs interact with SVP in vivo. (a) Yeast two-hybrid interaction assay for SVP and BPCs of class II: positive interactions on selective media –W-L-H +5mM 3-AT. (b) Co-immunoprecipitation assays. Nicotiana benthamiana leaves were infiltrated with constructs carrying SVP-GFP together with BPC4-RFP and BPC6-RFP, as described in experimental procedures. Immunoprecipitation step was performed using RFP-trap on total protein leaf extract. Samples were probed with GFP and RFP antibodies. S/N: supernatant; IP: immunoprecipitation. (c) Bi-molecular fluorescence complementation (BiFC) assay. Nicotiana benthamiana epidermis cells were transiently transformed with the indicated YN and YC fusions. In the first and the second column yellow fluorescence and the merging in the bright field were shown, respectively. (d) Bi-molecular fluorescence complementation (BiFC) assay. N.benthamiana epidermis cells were transiently transformed with the indicated YN and YC fusions and AP1-RFP construct. In the first, the second and the third column yellow fluorescence, red fluorescence and the merging between the two channels in the bright field were shown, respectively. Scale bars=40 μm. Accepted Article This article is protected by copyright. All rights reserved Figure 4. ChIP experiments on different mutant backgrounds. ChIP experiments on different mutant backgrounds. (a) Schematic diagram of the STK locus indicating the regions analysed by chromatin immunoprecipitation (ChIP; black bars). Black boxes, exons; white boxes, promoters and introns; asterisks, C-boxes; grey boxes, CArG-boxes; scale bar=500 bp. (b) Quantitative Real-Time PCR analysis of ChIP assay using chromatin extracted from svp ap1-12 agl24, wild-type (as a positive control), and bpc1-2 bpc2 bpc3 (as a negative control) testing the C-12, B and NC box regions. Antibodies against BPCs of class I were used. (c) Quantitative Real-time PCR analysis of ChIP assay using chromatin extracted from pSVP:SVP-GFP svp bpcV, pSVP:SVP-GFP svp (as a positive control) and wild-type (as a negative control), testing C-12, B and NC box regions. For the IP, commercial antibodies against GFP were used. Error bars represent the propagated error value using three replicates. ChIP results of one representative experiment are shown. Positive binding site fragments were considered only if they were enriched compared with the controls in at least three independent experiments. Figure 5. Mutation of CArG-boxes interferes with SVP and class I BPCs binding to STK promoter. (a)Schematic representation of the STK promoter versions generated: dark grey squares represent CArG-boxes wild-type and mutated (crossed). (b)-(g) GUS staining on inflorescences from pSTK:GUSwt (b-d) and pSTK_CArGm:GUS (e-g): whole inflorescence [(b) and (e)]; mature flower [(c) and (f)]; inflorescence meristem (IM), floral meristems (FM) and young flowers [(d) and (g)]; scale bars in (c), (d), (f) and (g)=100 μm. (h) Quantitative Real-Time PCR analysis of ChIP assay using chromatin extracted from pSVP:SVP-GFP pSTK_CArGm svp showing deregulation of the reporter and pSTK_CArGm as a negative control, testing wild-type region, mutated region and NC box. For the IP, antibodies against GFP have been used. (i) Quantitative Real-Time PCR analysis of ChIP assay using chromatin extracted from pSVP:SVP-GFP pSTK_CArGm svp showing correct expression of the reporter and pSTK_CArGm as a negative control, testing wild-type region, mutated region and NC box. For the IP, antibodies against GFP have been used. (l) Quantitative Real-Time PCR analysis of ChIP assay using chromatin extracted from pSVP:SVP-GFP pSTK_CArGm testing wild-type region, mutated region and NC box. For the IP, antibodies against Accepted Article This article is protected by copyright. All rights reserved Class I BPCs have been used; for negative control commercial antibodies against HA was used. Error bars represent the propagated error value using three replicates. ChIP results of one representative experiment are shown. Positive binding site fragments were considered only if they were significantly enriched compared to the controls in at least three independent experiments. Figure 6. Epigenetic regulation of STK. (a) Schematic representation of the STK genomic region tested in ChIP assay. Black boxes indicate exonic regions. Black bars indicate the regions analysed by chromatin immunoprecipitation (ChIP); region1 is located in the H3k27me3 – enriched region published by (Li et al. 2015) spanning -2627 upstream STK-transcriptional start site to +2050 pb downstream STK-transcriptional start site, whereas region 2 is localized 3 pb downstream the stop codon of the gene. Black arrow indicates the STK-transcription start site. Scale bar= 500 bp. (b) ChIP-quantitative Real-Time PCR determining the levels of H3K27me3 across the STK locus in inflorescence tissue. Quantitative Real-Time PCR quantification of STK sequences in precipitated chromatin was used to infer the methylation of histone H3 at lysine 27 (H3K27me3) and histone H3 density. Ct values were used to calculate the IP/IN signal. ChIP enrichments are presented as the percentage (%) of bound/input signal normalized to actin levels in the relative regions. We tested the efficiency of IP on histone modifications by quantifying the presence of the H3K27me3 mark in AG region which carries the mark H3k27me3, reported in Li et al. (2015). H3K27me3 mark in AT2G22560 was used as negative control for H3K27me3 mark (Li et al. 2015). The data were normalized to actin, with error bars indicating standard deviations based on three independent technical replicates. Four, independent ChIP experiments were performed and similar results were obtained. Figure 7. LHP1 directly regulates STK during flower development. Accepted Article This article is protected by copyright. All rights reserved (a)-(d) In-situ hybridisation on wild-type [(a) and (b)] and lhp1 inflorescences [(c) and (d)] using a STK-specific antisense probe (Brambilla et al., 2007). IM: inflorescence meristem; FM: floral meristem; numbers represent flower stages; scale bars=50 µm. (e) Expression analysis of STK by quantitative Real-Time PCR in lhp1 and wild-type inflorescences. The expression of STK was normalized to that of ubiquitin and the expression level in wildtype was set to 1. Asterisk indicates P < 0.05 in a Student’s t-test. (f) Quantitative Real-Time PCR analysis of ChIP assay using chromatin extracted from pLHP1:LHP1-GFP (Kotake et al, 2003) and wild-type (as a negative control), testing the Region 1 and Region 2 (Figure 6a). For the IP, commercial antibodies against SVP were used. Error bars represent the propagated error value using three replicates. ChIP results of one representative experiment are shown. Positive binding site fragments were considered only if they were significantly enriched compared with the controls in at least three independent experiments. Figure 8. Analysis of BPC and MADS-box transcription factor families binding sites. Venn diagram displaying the number of DAP-seq peaks and the common number of peaks associated to the BPC and MADS transcription factor families according to our analysis of the data by O'Malley et al (see Experimental procedures). Enriched motifs, as recovered by Homer (p-value ≤ 1e-30), are displayed underneath. Figure 9. Model of the protein complex formed to represses gene expression during flower development. BPCs and SVP bind C-boxes (in dark purple) and CArG-boxes (in light purple) respectively, and recruit LHP1 to a subset of gene loci. Accepted Article