MADS-box genes and floral development: the dark side

Journal of Experimental Botany, Vol. 63, No. 15, pp. 5397 5404, 2012 doi:10.1093/jxb/ers233 Advance Access publication 21 August, 2012 Flowering Newsletter Review MADS-box genes and floral development: the dark side Klaas Heijmans 1, Patrice Morel 2 and Michiel Vandenbussche 2, * 1 Plant Genetics, IWWR, Radboud University Nijmegen, 6525AJ Nijmegen, The Netherlands 2 UMR 5667 CNRS-INRA-ENS Lyon-Université Lyon I, RDP laboratory, ENS Lyon, 69364 Lyon Cedex 07, France * To whom correspondence should be addressed. E-mail: michiel.vandenbussche@ens-lyon.fr Received 29 June 2012; Revised 29 June 2012; Accepted 25 July 2012 Abstract The origin of the flower during evolution has been a crucial step in further facilitating plants to colonize a wide range of different niches on our planet. The >250 000 species of flowering plants existing today display an astonishing diversity in floral architecture. For this reason, the flower is a very attractive subject for evolutionary developmental (evo-devo) genetics studies. Research during the last two decades has provided compelling evidence that the origin and functional diversification of MIKC c MADS-box transcription factors has played a critical role during evolution of flowering plants. As master regulators of floral organ identity, MADS-box proteins are at the heart of the classic ABC model for floral development. Despite the enormous progress made in the field of floral development, there still remain aspects that are less well understood. Here we highlight some of the dark corners within our current knowledge on MADS-box genes and flower development, which would be worthwhile investigating in more detail in future research. These include the general question of to what extent MADS-box gene functions are conserved between species, the function of TM8-clade MADS-box genes which so far have remained uncharacterized, the divergence within the A-function, and post-transcriptional regulation of the ABC-genes. Key words: ABC model, floral architecture, floral evolution, flower development, MADS-box gene family, Petunia. MADS-box gene function and black boxes MADS-domain proteins are transcription factors that are present throughout the eukaryotic section of the tree of life. Mammals possess several MADS-box genes, but in plants this family has greatly expanded and 107 members have been identified in Arabidopsis (Parenicova et al., 2003). In this species, 46 members belong to the type II subfamily of MADS-box genes, whereas 61, the majority, belong to the type I subfamily, which is very poorly characterized in comparison with the type II subfamily. In fact, only five type I genes have been functionally characterized and all of them play a role in the development of the female gametophyte and/or the early seed (Köhler et al., 2003; Portereiko et al., 2006; Bemer et al., 2008; Colombo et al., 2008; Kang et al., 2008; Steffen et al., 2008), and in both of these tissues, most type I genes are preferentially expressed (Bemer et al., 2010). All MADS-box proteins possess the eponymous MADS-box domain, which is the domain responsible for binding to the target DNA element [the CArG-box, CC(A/ T) 6 GG], carries a nuclear localization signal and is involved in dimerization and binding of other factors (De Bodt et al., 2003). Type II MADS-box genes can be further subdivided into MIKC* (also called Mò) and MIKC c classes (the c stands for classic) (Parenicova et al., 2003; Gramzow and Theissen, 2010). These two classes are distinguished via structural differences in their K-domains and the length of the I-domain, which is longer and encoded by more exons in the MIKC* class proteins (Henschel et al., 2002). Unlike the type I subfamily, MIKC proteins carry three additional characteristic regions: a weakly conserved intervening region (I), a well-conserved keratin-like domain (K), both of which are involved in protein interactions, and a more variable C-terminal region (C) which mediates higher order formation and confers regulatory activity on the functional complex of which the MADS-domain proteins are part. In plants, the MIKC c MADS-box genes are most famous for their roles in flower development, and members of this type spell out the ABC of floral development (Schwarz-Sommer et al., 1990; Coen and Meyerowitz, 1991). The phylogeny of MIKC c genes has been The Author [2012]. Published by Oxford University Press [on behalf of the Society for Experimental Biology]. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com

5398 Heijmans et al. extensively studied, and shows that they can be subdivided into 13 major gene clades, of which seven probably existed already in the most recent common ancestor of angiosperms and gymnosperms ~300 million years ago (Becker and Theissen, 2003). Due to its undeniable advantages as a plant model species, MIKC c genes have been most extensively studied in Arabidopsis, for which for many clades all members have been functionally characterized, while for others at least one or more members have been analysed. In angiosperm genomes, one or more representatives of each clade are usually found, suggesting their functional conservation and importance for the plant life cycle. A simplified phylogeny (using only one representative gene per species per clade) based on MADS-box proteins from Arabidopsis and Petunia is shown in Fig. 1 to illustrate this. Remarkably, however, Arabidopsis seems to have lost one of the ancient clades called the TM8 lineage (Fig. 1). This has been known for a while (Becker and Theissen, 2003), but now that more and more plant genomes are available, it becomes clear that the absence of a TM8 homologue as in Arabidopsis is an exception rather than a rule. Gene loss is not uncommon among the angiosperm MADS-box genes. For example, Arabidopsis seems to have lost the paleoap3 subclade of the DEF/GLO clade (Kramer and Irish, 2000), and the lineage that led to extant gymnosperms may have lost the SEPALLATA (SEP)-clade genes (Becker and Theissen, 2003). However, the TM8 clade represents a special case since Arabidopsis contains no members of this clade at all and none of the TM8-clade genes has been functionally characterized in any species, whereas we currently know the function of at least one representative from all the other MIKC c clades (from Arabidopsis, and only in some instances from multiple other species). Therefore, the functional characterization of this gene clade is of special interest in order to complete our understanding of MIKC c gene function. While so far no loss-offunction phenotypes have been described, at least the expression data available for tomato and grapevine TM8 genes suggest a role during flower development (Pnueli et al., 1991; Díaz-Riquelme et al., 2009). In addition, one should keep in mind that to a large extent, our knowledge on MIKC c gene function is based primarily Fig. 1. Simplified phylogeny of the major MIKC c clade genes in Arabidopsis and Petunia. Neighbor Joining tree of MIKC c proteins from Arabidopsis thaliana and Petunia hybrida (prefix Ph- and shaded grey). The tree has been constructed using one representative of each clade in both species. For the TM8 clade, representatives have been added from some other species to illustrate its wide taxonomic distribution (Sl, Solanum lycopersicon; Am, Antirrhinum majus; Pt, Populus trichocarpa; Vv, Vitis vinifera). Numbers along the branches are bootstrap values exceeding 50%, and branches with <50% bootstrap support were collapsed. Black boxes represent clades of which there are, to our knowledge, no functional data from species other than the Brassicaceae, or of which there are no functional data at all (in the case of the TM8 clade). The Arabidopsis type I MADS-box protein AGL64 was used as an outgroup. The tree was generated as described previously (Vandenbussche et al., 2003).

MADS-box genes and floral development: the dark side 5399 upon research in Arabidopsis, and functional data from other species are often lacking. Of course, in some instances, gene function has been shown to be relatively conserved (as in the case of the B- and C-function genes in flower development) but, in other cases, data from Arabidopsis do not necessarily translate directly to other species, as illustrated below. Therefore, to truly get a grasp on a gene s function in plants, functional data from multiple phyla are highly desirable, and of course necessary to assess a gene s functional conservation or divergence. As shown by the black boxes in Fig. 1, there is still some work to do in this respect. Moreover, the majority of the MADS-box subfamilies have been shaped by a complex history of gene duplications followed by random subfunctionalization events, of which the final outcome might differ significantly, even between closely related species. Therefore, when performing an interspecies comparison of individual gene functions isolated from their own specific gene subfamily context, it might be very hard to distinguish true fundamental differences in gene function from differences simply caused by a divergent degree of redundancy and subfunctionalization with the remaining subfamily members. Consequently, the function of all subfamily members should be determined and any possible redundancies uncovered (by multiple mutant analysis) before any hard conclusions can be drawn. The dark side of the ABC model A typical dicotyledonous flower, such as an Arabidopsis flower, consists of four different organs, arranged in four whorls or concentric rings of tissue, from outside to inside; the sepals, petals, stamens, and the carpels. The ABC model, initially based on floral mutant phenotypes in Arabidopsis and Antirrhinum, postulates that three different gene classes, or functions, specify the identity of these organs in the four floral whorls. In the outermost whorl, the A-function specifies sepal identity, while the A- and B-function genes together control petal identity in the second whorl. In the third whorl, the B- and C-function genes are required for stamen development, while the C-function by itself specifies carpel identity in the inner whorl and terminates floral meristem activity. In addition, the A- and C-functions antagonize each other, enforcing the proper domains of activity. A later addition to the model are the E-function genes, which are required for all the other functions, since they encode proteins necessary for the formation of the higher order complexes in which most of the ABC-function proteins are active. Interestingly, except for one of the A-function genes, APETALA2 (AP2), all the ABCE-functions are encoded by one or more type II MADS-box genes, and a detailed overview of their function and regulation is given in Krizek and Fletcher (2005). Functional analyses of B- and C-lineage MADS-box protein genes in a diverse range of species (covering several dicot and monocot species) show that in general the B- and C-functions are very well conserved at the molecular level. If there are differences at all, they are usually found at the level of subfunctionalization/specialization between paralogues in the same gene lineage. In contrast, how the A-function is encoded in species other than Arabidopsis still remains one of the very dark corners of flower development. The A-function, required for the specification of sepals and petals (together with B-function genes) and for restriction of C-function genes to the inner whorls, is proposed to be composed of two genes in Arabidopsis: APETALA1 (AP1) (Irish and Sussex, 1990; Mandel et al., 1992; Bowman et al., 1993) and AP2 (Komaki et al., 1988; Bowman et al., 1989, 1991; Kunst et al., 1989; Drews et al., 1991; Jofuku et al., 1994). While different ap1 and ap2 mutants show highly variable phenotypes, in a broad sense they follow the predictions of the ABC model, since perianth organ identity is generally disrupted. Recently, however, researchers have started to question if these genes really should be thought of as perianth organ identity genes as defined in the classical ABC model, in the sense that they actively direct the formation of the perianth organs (Litt, 2007); perhaps more probably, their activity indirectly leads to sepal and petal development by establishing the floral meristem [which is in fact another function of these genes (Schultz and Haughn, 1993)] and restricting the C-function to the two inner whorls. This view is corroborated by the phenotypes of some ap1 and ap2 flowers. For both mutants, sepals can be converted to leaves or bracts, suggesting loss of floral meristem identity, or to carpels, indicating ectopic expression of the Arabidopsis C-gene AGAMOUS (AG). In the second whorl, secondary flowers may arise, also suggesting a loss of floral meristem identity, or petals are converted to stamens, as a result of ectopic expression of AGA. In other words, within the context of floral organ specification (as opposed to floral meristem specification), the A-function genes may have no function other than repressing the C-function. In recognition of this fact, Wollmann et al. (2010) propose a scenario, based on detailed expression analyses of mir172, AP2, and AG, in which AP2 acts predominantly as a cadastral gene, indirectly influencing organ fate, which is dependent upon the balance between AG and the AG-repressing activity of AP2. Thus, we will focus on their role in C-gene regulation when discussing the A-function genes in other species below. Homologues of AP2 have been found in a broad range of species, including, but not limited to, Antirrhinum, Amborella, Petunia, tobacco, tomato, maize, rice, Norway spruce, pine, and hybrid larch, and this gene may very well be ubiquitous within the seed plants. Although AP2 genes from other species are poorly characterized, the available data suggest that its function has been conserved to some extent. In maize, two AP2 homologues, INDETERMINATE SPIKELET1 (IDS1) and SISTER OF INDETERMINATE SPIKELET1 (SID1), play a role in floral meristem development and spikelet determinacy, but also in repression of C-function, since loss of function of these genes leads to the formation of carpelloid bracts (Chuck et al., 2008), indicating a high degree of conservation between Arabidopsis and maize. When a rice AP2 homologue is silenced, flowers carry fewer stamens and multiple stigmas, which is also observed in some Arabidopsis ap2 alleles (Zhao et al., 2006). Via heterologous expression, a spruce AP2-like gene has been shown to delay flowering and lead to a loss of determinacy in Arabidopsis (Nilsson et al., 2007), which also occur when the endogenous AP2 is overexpressed in Arabidopsis. Interestingly, orthologues of AP2 from either Petunia or Antirrhinum do not appear to play a role in repression of C-class genes in the outer whorls, but the Antirrhinum AP2-like genes do play a role in floral development,

5400 Heijmans et al. as loss of function of two AP2 homologues, LIPLESS1 (LIP1) and LIPLESS2 (LIP2), leads to sepal-to-leaf transformations and defects in petal development (Keck et al., 2003). Loss of function of an AP2 homologue in Petunia did not result in a floral phenotype (Maes et al., 2001). However, the genome sequence of Petunia (and Antirrhinum) was not known at the time of analysis, so it cannot yet be excluded that some of the floral functions are redundantly encoded by unknown closely related AP2 homologues, and therefore have been missed. So while part of the function of AP2 has been shown to be conserved in some species, especially with respect to its role in floral meristem identity specification, its role in C-gene repression remains to be further investigated. Functional conservation of AP1 is similarly limited. PROLIFERATING INFLORESCENCE MERISTEM (PIM) from pea, LeMADS_MC from tomato, and SQUAMOSA (SQUA) from Antirrhinum all appear to be involved in floral meristem identity establishment (Huijser et al., 1992; Berbel et al., 2001; Taylor et al., 2002; Vrebalov et al., 2002), but the role of AP1 in C-gene repression in Arabidopsis (Gregis et al., 2006) does not seem to be conserved in other species (as reviewed in Litt, 2007). In addition to AP1 and AP2, several other genes in Arabidopsis have a role in repressing AG, and thus may be argued also to contribute to the A-function: STERILE APETALA, RABBIT EARS, BELLRINGER, CURLY LEAF, AINTEGUMENTA, LEUNIG, SEUSS, SHORT VEGETATIVE PHASE, and AGAMOUS-LIKE24 (Liu and Meyerowitz, 1995; Goodrich et al., 1997; Byzova et al., 1999; Krizek et al., 2000, 2006; Franks et al., 2002; Bao et al., 2004; Gregis et al., 2006). Whereas STYLOSA, the Antirrhinum homologue of LEUNIG, has been shown to play a role in C-gene regulation (Motte et al., 1998), for other genes and species, these genes constitute largely unknown territory. Nevertheless, mutants with a partial A-function phenotype do exist in Petunia and Antirrhinum, called blind (bl) and fistulata (fis), respectively. In both mutants, the petals are partially transformed into antheroid tissues (the petal tubes develop normally), whereas the sepals are often wild type in appearance, only occasionally carrying carpelloid tips (Vallade et al., 1987; McSteen et al., 1998; Motte et al., 1998). As expected, mrna of C-function genes is ectopically present in these mutants (McSteen et al., 1998; Tsuchimoto et al., 2000). Both alleles were found to be loss-of-function alleles of members of a conserved plant microrna family, MIR169 (Cartolano et al., 2007). This family targets NF-YA transcription factors (Jones-Rhoades and Bartel, 2004), which are proposed to regulate target genes via binding to CCAAT-boxes (Mantovani, 1999). These CCAAT-boxes have been found to be conserved in the large regulatory intron of the C-function genes in a number of species (Hong et al., 2003). Together, these observations suggest a model in which BLIND and FISTULATA regulate C-gene expression levels by post-transcriptional repression of the NF-YA genes, which would act as C-gene activators. However, it remains to be proven that NF-YA genes are required for C-activity in the flower. Remarkably, it seems that both BLIND and FISTULATA do not simply function as C-repressors in the outer two whorls, since they are also expressed within the native C-expression domain. In fact, it has increasingly become clear that the establishment of the C-function domain, and the resulting reproductive organ development, should be seen as the outcome of the simultaneous and spatially overlapping action of both activators and repressors, instead of a simple activation in the inner whorls, and repression in the outer whorls. With its broad and overlapping expression pattern, the BLIND NF-YA module would fit in such a model. Based on data from Antirrhinum, during activation of the C-function genes, FIS may be temporarily repressed in the centre of the flower by FIMBRIATA (FIM), which would allow for increased activity of NF-YA and, as a result, increased expression of the C-function genes. After derepression of FIS, NF-YA activity will decrease, presumably dampening the expression level of C-function genes within their native expression domain and thereby preventing ectopic C-function. In support of such a scenario, it has been shown that increasing C-gene expression in the centre of the flower can lead to the transformation of perianth organs into reproductive organs (Cartolano et al., 2009), highlighting the importance of controlling C-gene expression levels in its own domain. As mentioned earlier, both Petunia and Antirrhinum possess AP2 homologues, and mir172 family members [which regulate the expression of APETALA2 in Arabidopsis (Chen, 2004; Wollmann et al., 2010)] are present at least in Petunia (unpublished data). It would be interesting to find out why the AP2 homologues in these species do not seem to play a role in C-gene repression. Conversely, does the BLIND NF-YA module play a role in C-gene repression in Arabidopsis or, indeed, in other species? Cartolano et al. (2009) have already noted that a lot of the elements involved in this mechanism of C-gene regulation are at least present in Arabidopsis, but that the function of FIM, the homologue of UNUSUAL FLORAL ORGANS (UFO) from Arabidopsis, has diverged. On the other hand, the tandem CCAAT-boxes are essential for proper expression of AG in Arabidopsis and are conserved in other plant species (Hong et al., 2003), which, in turn, suggests that regulation by NF-YA, which is proposed to bind to these elements, is required for establishment of C-gene activity. To conclude, while the B- and C-function genes have been shown to be highly conserved in many angiosperms, the A-function may tell a different story. On the one hand, there simply is a lack of data. A systematic comparative analysis of A -function genes in other species is therefore needed. On the other hand, the data available suggest that the regulation of the C-function genes is controlled by different mechanisms, not only within one species (as indicated by the many C-gene regulators in Arabidopsis), but, more interestingly, also between different species, as illustrated by the examples of Petunia and Antirrhinum. Finally, not only are there differences in the set of genes involved, there are also differences with respect to the regulatory logic employed. While AP2-mediated control of the C-function primarily prevents expression from the outer whorls of the flower, the BLIND FIS/NF-YA module seems to control the C-function primarily within its native expression domain. Based on the difficulties in defining a general A-function discussed above, Causier et al. (2009) have recently proposed a revised (A)BC model in which the (A)-function comprises a plethora of genes that ultimately allow the B- and C-genes to determine floral organ identity. In this model, the (A)-function is composed of such disparate genes as C-gene repressors, SEP

MADS-box genes and floral development: the dark side 5401 genes, LEAFY, WUSCHEL, and others. On one hand, this model succeeds much better than the traditional model in incorporating the variability observed within the classical A-function and in recognizing that only the B- and C-function genes seem to be bona fide, well-conserved organ identity genes. On the other hand, one might argue that defining the (A)-function as a function that sets the stage for the organ identity genes to function remains somewhat too vague to be useful. Nevertheless, this model, which is essentially an expanded version of an old BC model (then referred to as the AB model) proposed by Schwarz-Sommer et al. (1990), seems better suited to the current data. Post-transcriptional control of the MADS-box ABC-genes? Given that the ABC-genes control the identity of the floral organs, their expression patterns, both in time and in space, are of prime importance, and, as such, they are subject to tight regulation. Most research in this regard has focused on transcriptional regulation. It is still unclear, however, to what extent post-transcriptional control directly influences the activity of the MADS-box genes. Some observations suggest that at least some of the type II MADS-box genes themselves might be directly regulated at the post-transcriptional level. Jack et al. (1994) have described 35S:AP3 lines in Arabidopsis, which, in accordance with the predictions of the ABC model, develop stamens in their fourth whorl. However, because AP3 functions in a complex with PISTILLATA (PI), the predicted transformation from sepals to petals does not occur, since PI is not expressed in the first whorl (Goto and Meyerowitz, 1994). Interestingly, the authors have noted that while AP3 mrna is present in the first whorl of 35S:AP3 flowers (as expected), the protein product is not. This discrepancy between mrna and protein levels could suggest that post-transcriptional regulation of AP3 is taking place. Alternatively, the turnover rate of the protein may be very high in the first whorl, possibly because of the absence of PI, which might stabilize AP3 levels via protein protein interaction. Such a view would also correspond to the discrepancy observed between AP3 mrna and AP3 protein levels in pi mutants, in which AP3 mrna is present in the second whorl of stage 5 10 flowers, while the protein is not (Jack et al., 1992, 1994). Nevertheless, to our knowledge, it has never been actually established whether these observations reflect active, PI-dependent, post-transcriptional regulation of AP3, or whether they are simply a result of reduced stability of AP3 in the absence of PI. Indirect indications of post-transcriptional regulation of ABCs come from studies in Petunia. Whereas Arabidopsis 35S:AG flowers show transformations as predicted by the ABC model, a conversion of petals to stamens and sepals to carpels (Mizukami and Ma, 1992), analogous experiments in Petunia produce very different results. Ectopic overexpression of either PMADS3 or FBP6, the two Petunia C-function genes, results in only partial petal-to-stamen transformation, while sepals remain virtually unchanged, only sometimes producing carpelloid tips (Tsuchimoto et al., 1993; and unpublished results). In contrast, when PMADS3 is expressed in tobacco via the 35S promoter, complete petal-to-stamen and sepal-to-carpel transformations occur, showing that this gene is capable of specifying reproductive organ fate in the perianth whorls (Tsuchimoto et al., 1993). Since RNA of the transgenes is present throughout the flower in the Petunia overexpression lines, the lack of full homeotic conversion may suggest that the Petunia C-function genes are post-transcriptionally regulated in the two outer floral whorls. Alternatively, cofactors or protein interaction partners required for the formation of a functional complex may be absent from these tissues. However, this is seemingly contradicted by the fact that ectopically expressing an AG homologue of cucumber in Petunia produces nearly complete petal-to-stamen and very strong sepal-to-carpel transformations (Kater et al., 1998). On the other hand, the two Antirrhinum C-genes differ in their abilities to interact with E-function proteins (SEPs) (Airoldi et al., 2010). A single amino acid insertion in FARINELLI (FAR), one of the Antirrhinum C-function proteins, limits its interactions with the Arabidopsis SEPs to just SEP3, which is not expressed in the first whorl. As a result, ectopic expression of FAR in Arabidopsis flowers does not result in sepal-to-carpel transformation in Arabidopsis. Such a scenario may also play a role in Petunia, especially considering the fact it harbours six SEP genes with divergent expression patterns. However, this particular polymorphism does not play a role in Petunia, since neither of the two C-function genes contains this amino acid insertion. Nevertheless, these data show that small changes in protein sequence can dramatically alter protein function and perhaps the cucumber C-function protein is able to form a functional complex in the perianth whorls in Petunia, while the endogenous proteins are not, due to specialization among interaction partners. Another interpretation of these data is that some post-transcriptional repression mechanism in Petunia recognizes the mrnas of its own C-function genes and subsequently represses them in the sepals and petals. The cucumber homologue may be sufficiently divergent to avoid recognition by this repression mechanism, and would thus be able to accumulate as a protein and specify reproductive organ fate in the perianth whorls. Whatever the scenario, phenomena like these deserve further study, since it could lead to the discovery of novel regulatory mechanisms. A flourishing future for floral research? Over the years, enormous progress has been made in our understanding of floral development in a handful of model species. In general, these model species have quite simple flower structures, and they do not cover the mind-blowing diversity of floral architecture displayed in the >250 000 species of flowering plants. We are thus still far from identifying the key evolutionary molecular changes that generated this feast for the eye. So far, for technical reasons (e.g. no functional analysis possible), it has often been too challenging to start to work with more exotic non-model plants. However, we might have good hopes that this will change, perhaps already within the coming decade. Due to the unprecedented and continued progress made in next-generation sequencing technologies, sequencing the entire genome of large series of individual plants might become almost

5402 Heijmans et al. as trivial as sequencing a single gene. In such a case, it will become feasible to start forward and reverse genetics screens in any species that have a reasonable generation time and the possibility to self-fertilize. It will be sufficient to mutagenize a seedstock, grow populations, screen for interesting phenotypes, recover several independent alleles, and simply sequence the entire genome of these individuals to identify the mutated gene. Likewise, reverse genetics screens might be performed in silico by blasting a database composed of the individual genome sequences of all members of a mutagenized population. By starting to analyse a new series of plants species with interesting floral architectural characteristics, we can certainly hope for a flourishing future for floral research. References Airoldi CA, Bergonzi S, Davies B. 2010. Single amino acid change alters the ability to specify male or female organ identity. Proceedings of the National Academy of Sciences, USA 107, 18898 18902. Bao X, Franks RG, Levin JZ, Liu Z. 2004. Repression of AGAMOUS by BELLRINGER in floral and inflorescence meristems. The Plant Cell 16, 1478 1489. Becker A, Theissen G. 2003. The major clades of MADS-box genes and their role in the development and evolution of flowering plants. Molecular Phylogenetics and Evolution 29, 464 489. Bemer M, Heijmans K, Airoldi C, Davies B, Angenent GC. 2010. An atlas of type I MADS box gene expression during female gametophyte and seed development in Arabidopsis. Plant Physiology 154, 287 300. Bemer M, Wolters Arts M, Grossniklaus U, Angenent GC. 2008. The MADS domain protein DIANA acts together with AGAMOUS-LIKE80 to specify the central cell in Arabidopsis ovules. The Plant Cell 20, 2088 2101. Berbel A, Navarro C, Ferrándiz C, Cañas LA, Madueño F, Beltrán J-P. 2001. Analysis of PEAM4, the pea AP1 functional homologue, supports a model for AP1-like genes controlling both floral meristem and floral organ identity in different plant species. The Plant Journal 25, 441 451. Bowman JL, Alvarez J, Weigel D, Meyerowitz EM, Smyth DR. 1993. Control of flower development in Arabidopsis thaliana by APETALA1 and interacting genes. Development 119, 721 743. Bowman JL, Smyth DR, Meyerowitz EM. 1989. Genes directing flower development in Arabidopsis. The Plant Cell 1, 37 52. Bowman J, Smyth D, Meyerowitz E. 1991. Genetic interactions among floral homeotic genes of Arabidopsis. Development 112, 1 20. Byzova MV, Franken J, Aarts MGM, de Almeida Engler J, Engler G, Mariani C, Van Lookeren Campagne MM, Angenent GC. 1999. Arabidopsis STERILE APETALA, a multifunctional gene regulating inflorescence, flower, and ovule development. Genes and Development 13, 1002 1014. Cartolano M, Castillo R, Efremova N, Kuckenberg M, Zethof J, Gerats T, Schwarz Sommer Z, Vandenbussche M. 2007. A conserved microrna module exerts homeotic control over Petunia hybrida and Antirrhinum majus floral organ identity. Nature Genetics 39, 901 905. Cartolano M, Efremova N, Kuckenberg M, Raman S, Schwarz Sommer Z. 2009. Enhanced AGAMOUS expression in the centre of the Arabidopsis flower causes ectopic expression over its outer expression boundaries. Planta 230, 857 862. Causier B, Schwarz Sommer Z, Davies B. 2009. Floral organ identity: 20 years of ABCs. Seminars in Cell and Developmental Biology 21, 73 79. Chen X. 2004. A microrna as a translational repressor of APETALA2 in Arabidopsis flower development. Science 303, 2022 2025. Chuck G, Meeley R, Hake S. 2008. Floral meristem initiation and meristem cell fate are regulated by the maize AP2 genes ids1 and sid1. Development 135, 3013 3019. Coen ES, Meyerowitz EM. 1991. The war of the whorls: genetic interactions controlling flower development. Nature 353, 31 37. Colombo M, Masiero S, Vanzulli S, Lardelli P, Kater MM, Colombo L. 2008. AGL23, a type I MADS-box gene that controls female gametophyte and embryo development in Arabidopsis. The Plant Journal 54, 1037 1048. De Bodt S, Raes J, Van de Peer Y, Theissen G. 2003. And then there were many: MADS goes genomic. Trends in Plant Science 8, 475 483. Díaz Riquelme J, Lijavetzky D, Martínez Zapater JM, Carmona MJ. 2009. Genome-wide analysis of MIKC C -type MADS box genes in grapevine. Plant Physiology 149, 354 369. Drews GN, Bowman JL, Meyerowitz EM. 1991. Negative regulation of the Arabidopsis homeotic gene AGAMOUS by the APETALA2 product. Cell 65, 991 1002. Franks RG, Wang C, Levin JZ, Liu Z. 2002. SEUSS, a member of a novel family of plant regulatory proteins, represses floral homeotic gene expression with LEUNIG. Development 129, 253 263. Goodrich J, Puangsomlee P, Martin M, Long D, Meyerowitz EM, Coupland G. 1997. A Polycomb-group gene regulates homeotic gene expression in Arabidopsis. Nature 386, 44 51. Goto K, Meyerowitz EM. 1994. Function and regulation of the Arabidopsis floral homeotic gene PISTILLATA. Genes and Development 8, 1548 1560. Gramzow L, Theissen G. 2010. A hitchhiker s guide to the MADS world of plants. Genome Biology 11, 214. Gregis V, Sessa A, Colombo L, Kater MM. 2006. AGL24, SHORT VEGETATIVE PHASE, and APETALA1 redundantly control AGAMOUS during early stages of flower development in Arabidopsis. The Plant Cell 18, 1373 1382. Henschel K, Kofuji R, Hasebe M, Saedler H, Münster T, Theißen G. 2002. Two ancient classes of MIKC-type MADS-box genes are present in the moss Physcomitrella patens. Molecular Biology and Evolution 19, 801 814. Hong RL, Hamaguchi L, Busch MA, Weigel D. 2003. Regulatory elements of the floral homeotic gene AGAMOUS identified by phylogenetic footprinting and shadowing. The Plant Cell 15, 1296 1309. Huijser P, Klein J, Lonnig WE, Meijer H, Saedler H, Sommer H. 1992. Bracteomania, an inflorescence anomaly, is caused by the loss of function of the MADS-box gene squamosa in Antirrhinum majus. EMBO Journal 11, 1239 1249.

MADS-box genes and floral development: the dark side 5403 Irish VF, Sussex IM. 1990. Function of the apetala-1 gene during Arabidopsis floral development. The Plant Cell 2, 741 753. Jack T, Brockman LL, Meyerowitz EM. 1992. The homeotic gene APETALA3 of Arabidopsis thaliana encodes a MADS box and is expressed in petals and stamens. Cell 68, 683 697. Jack T, Fox GL, Meyerowitz EM. 1994. Arabidopsis homeotic gene APETALA3 ectopic expression: transcriptional and posttranscriptional regulation determine floral organ identity. Cell 76, 703 716. Jofuku KD, den Boer BGW, Van Montagu M, Okamuro JK, 1994. Control of Arabidopsis flower and seed development by the homeotic gene APETALA2. The Plant Cell 6, 1211 1225. Jones Rhoades M, Bartel D. 2004. Computational identification of plant micrornas and their targets, including a stress-induced mirna. Molecular Cell 14, 787 799. Kang I H, Steffen JG, Portereiko MF, Lloyd A, Drews GN. 2008. The AGL62 MADS domain protein regulates cellularization during endosperm development in Arabidopsis. The Plant Cell 20, 635 647. Kater MM, Colombo L, Franken J, Busscher M, Masiero S, Van Lookeren Campagne MM, Angenent GC. 1998. Multiple AGAMOUS homologs from cucumber and petunia differ in their ability to induce reproductive organ fate. The Plant Cell 10, 171 182. Keck E, McSteen P, Carpenter R, Coen E. 2003. Separation of genetic functions controlling organ identity in flowers. EMBO Journal 22, 1058 1066. Köhler C, Hennig L, Spillane C, Pien S, Gruissem W, Grossniklaus U. 2003. The Polycomb-group protein MEDEA regulates seed development by controlling expression of the MADS-box gene PHERES1. Genes and Development 17, 1540 1553. Komaki M, Okada K, Nishino E, Shimura Y. 1988. Isolation and characterization of novel mutants of Arabidopsis thaliana defective in flower development. Development 104, 195 203. Kramer EM, Irish VF. 2000. Evolution of the petal and stamen developmental programs: evidence from comparative studies of the lower eudicots and basal angiosperms. International Journal of Plant Sciences 161, S29 S40. Krizek BA, Fletcher JC. 2005. Molecular mechanisms of flower development: an armchair guide. Nature Reviews Genetics 6, 688 698. Krizek BA, Lewis MW, Fletcher JC. 2006. RABBIT EARS is a second-whorl repressor of AGAMOUS that maintains spatial boundaries in Arabidopsis flowers. The Plant Journal 45, 369 383. Krizek BA, Prost V, Macias A. 2000. AINTEGUMENTA promotes petal identity and acts as a negative regulator of AGAMOUS. The Plant Cell 12, 1357 1366. Kunst L, Klenz JE, Martinez Zapater J, Haughn GW. 1989. AP2 gene determines the identity of perianth organs in flowers of Arabidopsis thaliana. The Plant Cell 1, 1195 1208. Litt A. 2007. An evaluation of A-function: evidence from the APETALA1 and APETALA2 gene lineages. International Journal of Plant Sciences 168, 73 91. Liu Z, Meyerowitz E. 1995. LEUNIG regulates AGAMOUS expression in Arabidopsis flowers. Development 121, 975 991. Maes T, Van de Steene N, Zethof J, Karimi M, D Hauw M, Mares G, Van Montagu M, Gerats T. 2001. Petunia Ap2-like genes and their role in flower and seed development. The Plant Cell 13, 229 244. Mandel MA, Gustafson Brown C, Savidge B, Yanofsky MF. 1992. Molecular characterization of the Arabidopsis floral homeotic gene APETALA1. Nature 360, 273 277. Mantovani R. 1999. The molecular biology of the CCAAT-binding factor NF-Y. Gene 239, 15 27. McSteen PC, Vincent CA, Doyle S, Carpenter R, Coen ES. 1998. Control of floral homeotic gene expression and organ morphogenesis in Antirrhinum. Development 125, 2359 2369. Mizukami Y, Ma H. 1992. Ectopic expression of the floral homeotic gene AGAMOUS in transgenic Arabidopsis plants alters floral organ identity. Cell 71, 119 131. Motte P, Saedler H, Schwarz Sommer Z. 1998. STYLOSA and FISTULATA: regulatory components of the homeotic control of Antirrhinum floral organogenesis. Development 125, 71 84. Nilsson L, Carlsbecker A, Sundås Larsson A, Vahala T. 2007. APETALA2 like genes from Picea abies show functional similarities to their Arabidopsis homologues. Planta 225, 589 602. Parenicova L, de Folter S, Kieffer M, et al. 2003. Molecular and phylogenetic analyses of the complete MADS-box transcription factor family in Arabidopsis: new openings to the MADS world. The Plant Cell 15, 1538 1551. Pnueli L, Abu Abeid M, Zamir D, Nacken W, Schwarz Sommer Z, Lifschitz E. 1991. The MADS box gene family in tomato: temporal expression during floral development, conserved secondary structures and homology with homeotic genes from Antirrhinum and Arabidopsis. The Plant Journal 1, 255 266. Portereiko MF, Lloyd A, Steffen JG, Punwani JA, Otsuga D, Drews GN. 2006. AGL80 is required for central cell and endosperm development in Arabidopsis. The Plant Cell 18, 1862 1872. Schultz EA, Haughn GW. 1993. Genetic analysis of the floral initiation process (FLIP) in Arabidopsis. Development 119, 745 765. Schwarz Sommer Z, Huijser P, Nacken W, Saedler H, Sommer H. 1990. Genetic control of flower development by homeotic genes in Antirrhinum majus. Science 250, 931 936. Steffen JG, Kang I H, Portereiko MF, Lloyd A, Drews GN. 2008. AGL61 interacts with AGL80 and is required for central cell development in Arabidopsis. Plant Physiology 148, 259 268. Taylor SA, Hofer JMI, Murfet IC, Sollinger JD, Singer SR, Knox MR, Ellis THN. 2002. PROLIFERATING INFLORESCENCE MERISTEM, a MADS-box gene that regulates floral meristem identity in pea. Plant Physiology 129, 1150 1159. Tsuchimoto S, Mayama T, Van Der Krol A, Ohtsubo E. 2000. The whorl-specific action of a petunia class B floral homeotic gene. Genes to Cells 5, 89 99. Tsuchimoto S, van der Krol AR, Chua NH. 1993. Ectopic expression of pmads3 in transgenic petunia phenocopies the petunia blind mutant. The Plant Cell 5, 843 853. Vallade J, Maizonnier D, Cornu A. 1987. La morphogenèse florale chez le pétunia. I. Analyse d un mutant à corolle staminée. Canadian Journal of Botany 65, 761 764.

5404 Heijmans et al. Vandenbussche M, Theissen G, Van de Peer Y, Gerats T. 2003. Structural diversification and neo-functionalization during floral MADS-box gene evolution by C-terminal frameshift mutations. Nucleic Acids Research 31, 4401 4409. Vrebalov J, Ruezinsky D, Padmanabhan V, White R, Medrano D, Drake R, Schuch W, Giovannoni J. 2002. A MADS-box gene necessary for fruit ripening at the tomato Ripening-Inhibitor (Rin) locus. Science 296, 343 346. Wollmann H, Mica E, Todesco M, Long JA, Weigel D. 2010. On reconciling the interactions between APETALA2, mir172 and AGAMOUS with the ABC model of flower development. Development 137, 3633 3642. Zhao L, Xu S, Chai T, Wang T. 2006. OsAP2-1, an AP2-like gene from Oryza sativa, is required for flower development and male fertility. Sexual Plant Reproduction 19, 197 206.