Duplication and Diversification in the APETALA1/FRUITFULL Floral Homeotic Gene Lineage: Implications for the Evolution of Floral Development

Corresponding author: Amy Litt, 266 Whitney Ave., Yale University, New Haven, CT 06520-8104., amy.litt{at}yale.edu (E-mail)

Communicating editor: D. WEIGEL

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Phylogenetic analyses of angiosperm MADS-box genes suggest that this gene family has undergone multiple duplication events followed by sequence divergence. To determine when such events have taken place and to understand the relationships of particular MADS-box gene lineages, we have identified APETALA1/FRUITFULL-like MADS-box genes from a variety of angiosperm species. Our phylogenetic analyses show two gene clades within the core eudicots, euAP1 (including Arabidopsis APETALA1 and Antirrhinum SQUAMOSA) and euFUL (including Arabidopsis FRUITFULL). Non-core eudicot species have only sequences similar to euFUL genes (FUL-like). The predicted protein products of euFUL and FUL-like genes share a conserved C-terminal motif. In contrast, predicted products of members of the euAP1 gene clade contain a different C terminus that includes an acidic transcription activation domain and a farnesylation signal. Sequence analyses indicate that the euAP1 amino acid motifs may have arisen via a translational frameshift from the euFUL/FUL-like motif. The euAP1 gene clade includes key regulators of floral development that have been implicated in the specification of perianth identity. However, the presence of euAP1 genes only in core eudicots suggests that there may have been changes in mechanisms of floral development that are correlated with the fixation of floral structure seen in this clade.

THE products of MADS-box genes have been implicated in the regulation of a variety of plant developmental mechanisms and have been shown to be particularly important in the specification and development of the angiosperm flower (COEN and MEYEROWITZ 1991 ; ANGENENT et al. 1995 ; ROUNSLEY et al. 1995 ; ALVAREZ-BUYLLA et al. 2000A ; FERRANDIZ et al. 2000 ). In Arabidopsis thaliana and other core eudicot species, MADS-domain-containing proteins are required for the proper transition from an inflorescence meristem to a floral meristem and for the correct specification of the identity of the four types of floral organs. The specification of floral organ identity has been codified in the ABC model (COEN and MEYEROWITZ 1991 ), which postulates three gene functions, A, B, and C, that act in overlapping concentric domains of the meristem to specify the floral organs. According to this model, based on work in the two model species A. thaliana and Antirrhinum majus, A-function specifies sepal identity in the outer domain of the meristem, A + B specifies petal, B + C specifies stamen (male reproductive organs), and C-function specifies carpel identity (female reproductive organs) in the innermost domain. Nearly all of the A-, B-, and C-function genes belong to the MADS-box family. Thus, understanding how different floral morphologies and developmental mechanisms evolved requires a determination of how these genes may have changed during the course of angiosperm diversification.

The history of the MADS-box gene family in plants is characterized by duplication events and subsequent divergence. For instance, phylogenies of the MADS-box gene family show that two lineages, which include the Arabidopsis B-function genes APETALA3 (AP3) and PISTILLATA (PI), arose by duplication from a single ancestral gene lineage and that the A-, B-, and C-function lineages themselves are probably all products of duplication events (DOYLE 1994 ; PURUGGANAN et al. 1995 ; TANDRE et al. 1995 ; HASEBE and BANKS 1997 ; KRAMER et al. 1998 ; KROGAN and ASHTON 2000 ; THEISSEN et al. 2000 ). In addition to these duplications that preceded or occurred in conjunction with the origin of the angiosperms, MADS-box gene lineage duplications have also occurred within individual angiosperm lineages (KRAMER and IRISH 1999 ; LOWMAN and PURUGGANAN 1999 ). The frequency of these events suggests that any comparative study of MADS-box genes requires as its foundation a comprehensive gene phylogeny that can be used to identify gene clades and to determine orthology (relationship through speciation) and paralogy (relationship through duplication) of various genes and lineages. Such a phylogeny provides a basis for defining orthologous genes for comparison and thereby provides a framework for comparative studies of gene structure, expression, and function across the angiosperms. We undertook a phylogenetic analysis of the APETALA1/FRUITFULL (AP1/FUL) MADS-box gene lineage (also called the SQUA lineage; e.g., KROGAN and ASHTON 2000 ; THEISSEN et al. 2000 ), members of which have been identified as key regulators of floral development in several model species, to identify duplication and sequence divergence events that occurred during the history of this gene lineage across the angiosperms.

In Arabidopsis, severe apetala1 mutants have sepals transformed into bract-like structures that subtend secondary flowers. Petals are absent. The inner two whorls of organs, the stamens and carpels, are essentially normal. This pattern may be repeated in the secondary flowers with the formation of tertiary nested floral structures (IRISH and SUSSEX 1990 ; BOWMAN et al. 1993 ). On the basis of this phenotype AP1 has been implicated in the specification of floral meristem identity as well as of sepal and petal identity (A-function; e.g., IRISH and SUSSEX 1990 ; BOWMAN et al. 1993 ; WEIGEL and MEYEROWITZ 1994 ).

The Arabidopsis genome contains two genes, CAULIFLOWER (CAL) and FUL, that are closely related to AP1 and that share redundant functions for floral meristem specification. CAL has no phenotype on its own, but the ap1 cal double mutant shows an enhancement of the repeated branching pattern seen in ap1 (BOWMAN et al. 1993 ; KEMPIN et al. 1995 ). An even more severe branching phenotype is seen in the Arabidopsis ap1 cal ful triple mutant, in which essentially all floral meristem character is lost and flowers are not formed (FERRANDIZ et al. 2000 ). In contrast to CAL, FUL also has separate and nonredundant functions and is required for proper fruit and leaf development (GU et al. 1998 ). In the Arabidopsis ful mutant, a lack of proper cell differentiation in the fruit walls abolishes fruit elongation, causing some fruits to rupture prematurely as the seeds develop. In addition, the cauline leaves of ful mutants are broader and rounder and have fewer cell layers than wild type.

To date little information is available regarding the function of members of the AP1/FUL gene family in other angiosperm species. CAL appears to be the result of a duplication specific to Brassicaceae (PURUGGANAN 1997 ; PURUGGANAN and SUDDITH 1998 ) and has been implicated in the cauliflower phenotype of Brassica oleracea (KEMPIN et al. 1995 ). Putative FUL orthologs are widespread throughout angiosperms, but their roles in other species have not yet been defined. The loss-of-function mutation of the Antirrhinum AP1 ortholog, SQUAMOSA (SQUA), shows a more complete loss of floral meristem identity than ap1 shows and rarely produces flowers, a phenotype similar to the ap1 cal ful triple mutant. Notably, when squa flowers are produced, the specification of organ identity is normal (HUIJSER et al. 1992 ). The proliferating inflorescence meristem phenotype in Pisum sativum, caused by a mutation in the AP1 ortholog PEAM4, has been described as similar to that of squa mutants (TAYLOR et al. 2002 ). To date, AP1 is the only gene in this lineage that has been shown to confer A-function in its native species.

We investigated the history of the AP1/FUL lineage by constructing a phylogeny that included sequences from a variety of angiosperm species. Previous analyses had suggested that AP1 and FUL themselves belong to separate closely related gene clades that were the result of a duplication event that occurred sometime after the divergence of the monocot lineage (HASEBE and BANKS 1997 ). The AP1 and FUL clades have been included in a single gene family, generally called the AP1 or SQUA family (e.g., SOUTHERTON et al. 1998 ; HASEBE 1999 ; THEISSEN et al. 2000 ; PELUCCHI et al. 2002 ), which has been further grouped with the SEPALLATA genes in the AP1/AGL9 family (e.g., PURUGGANAN et al. 1995 ; PURUGGANAN 1997 ; BUCHNER and BOUTIN 1998 ; MOON et al. 1999 ; LAWTON-RAUH et al. 2000 ). AP1 and FUL genes share significant sequence similarity, and thus the orthology of published genes is often difficult to ascertain. Our goal therefore was to determine to which group individual sequences belong in order to clarify orthology and paralogy and to provide a framework for gene comparisons. We also wanted to ascertain where, with respect to angiosperm phylogeny, the AP1-FUL duplication occurred. The results of our phylogenetic analysis indicate that there were several duplications in the evolution of the AP1/FUL gene family and that the AP1-FUL duplication is correlated with the diversification of the core eudicots and the concurrent fixation of floral structure. Sequence comparisons also identify conserved amino acid motifs that allow us to differentiate AP1-like and FUL-like sequences. These data allow us to formulate hypotheses regarding the evolution of floral developmental mechanisms across the angiosperms.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Unique AP1- and FUL-like sequences available during the course of this study were identified by BLAST searches (ALTSCHUL et al. 1997 ) and were included in the analysis (see supplemental data at http://www.genetics.org/supplemental/ for accession numbers). SEPALLATA-like sequences and Arabidopsis AGL6 and gymnosperm DAL1-like sequences, identified in published analyses as most closely related to the AP1/FUL lineage (PURUGGANAN 1997 ; HASEBE 1999 ; THEISSEN et al. 2000 ), were included as outgroups.

New species were selected for inclusion in the analysis according to phylogenetic position (Fig 1) and availability of floral bud material. Species used and genes cloned from each are listed in Table 1. The species sampled include core eudicots as well as a variety of non-core eudicots and non-eudicots. Total RNA was extracted from 1 g of floral buds of varied ages using the standard Trizol (Invitrogen, Carlsbad, CA) protocol. For P. sativum, the RNeasy kit (QIAGEN, Valencia, CA) was used; for Heuchera americana and Corylopsis sinensis, Concert Plant RNA reagent (Invitrogen) was used to eliminate starch coprecipitation. Poly(A)⁺ RNA was isolated from total RNA using Magnetight particles (Novagen, Madison, WI). The purification procedure was performed twice on each RNA sample for cleaner separation of poly(A)⁺ RNA. cDNA was synthesized using Superscript II (Invitrogen) according to the manufacturer's instructions.

View larger version (34K): In this window In a new window Download PPT slide   Figure 1. Outline of angiosperm phylogeny. Arabidopsis is a rosid and Antirrhinum is an asterid. Species used in this study are indicated according to taxonomic group. This simplified phylogeny is based on analyses from ANGIOSPERM PHYLOGENY GROUP (1998), SAVOLAINEN et al. 2000 , and SOLTIS and SOLTIS 2000 .

Amplification of target genes was carried out in two stages using a protocol designed to recover all possible genes belonging to the AP1/FUL gene lineage. First, a forward degenerate primer (AP1MDS3, GTNCARYTNARRMGNATNGARAAYAAGAT), designed to anneal to the MADS-box of AP1- and FUL-like sequences, was used with a poly(T) reverse primer [poly(T), GACTCGAGTCGACATCGA(T)₁₇V]. The reaction was run for 30–35 cycles with an annealing temperature of 42° on a GeneAmp 2400 thermocycler (Perkin-Elmer/Applied Biosystems, Norwalk, CT). Products with discrete bands of 500–1000 bp were cloned (TOPO-TA cloning kit, Invitrogen). In addition, the product of this first amplification reaction was diluted 1:25 and used as template in successive PCR reactions. These reactions used combinations of three nested forward primers (AP1MDS1, GCICWTGARMTNTCNRTNYTNTGYGATGC; AP1MDS2, TGGNYTKNTSAAGAARGCTCATGA; SQUA, TCWGTKCTTTGTGATGCTGAAGT) and three nested reverse primers (AP1R2, ATASASTGGTTCCAGMGTWAGGTC; SQUAR, GCAAAGCATCCMAKATGGCATG; AGL8R, AGRTGRYKAASCATCCAIKGIGGCA) as well as the two primers used in the initial amplification reaction. These reactions were run for 30–40 cycles at an annealing temperature of 46°. All products showing a 500- to 1000-bp band on an agarose gel were cloned.

At least 50 clones were sequenced for most species. AP1-, FUL-, and SEP-like sequences from each species were aligned using GeneWorks (Oxford Molecular, Springfield, VA) or CLUSTALX (THOMPSON et al. 1994 ) to determine how many different sequences were present, and representative samples were sequenced on the reverse strand. Sequences from several species that were not exhaustively assessed were also included. These species include Petunia hybrida, Lycopersicon esculentum, Clarkia concinna, P. sativum, Syringa vulgaris, Ranunculis acris, and Papaver nudicaule. The following unpublished sequences were used with permission: Lilium regale LrSQA, LrSQB (A. Kanno, Tohoku University), P. sativum PM9 (F. Madueno, Universidad Politecnica de Valencia-CSIC), and Paeonia suffruticosa PsMADS9 (E. Kramer, Harvard University).

It was not possible to determine whether slightly different sequences represented alleles of a single gene or were in fact different genes; therefore the observed pattern of nucleotide variability was used to make this assessment. Groups of similar sequences were compared at variable sites; if a group could be divided into subgroups such that members of each subgroup shared the same nucleotide at each site, the subgroups were treated as separate genes and were all included in the analysis. If a group of similar sequences could not be so subdivided, and members showed nucleotide differences in a variable pattern across the gene, those sequences were taken to represent alleles of the same gene. In these cases a consensus sequence was used in the analysis.

Attempts to align nucleotide sequences produced inconsistent and significantly variable results. Amino acid sequences gave more reproducible results but contained insufficient information to produce well-resolved phylogenies. Therefore putative amino acid sequences were aligned in CLUSTALX and aa2dna (http://www.bio.psu.edu/People/Faculty/Nei/Lab/software.htm) was used to substitute the nucleotide sequences for the amino acids, thus producing a matrix of aligned nucleotide sequences (supplemental data at http://www.genetics.org/supplemental/). Mega version 2.1 (KUMAR et al. 2001 ) was used to translate the alignment into a format for phylogenetic analysis. Phylogenetic analyses were performed in PAUP 4.0b10 (SWOFFORD 2000 ) with Arabidopsis AGL6 used to root the tree on the basis of previously published analyses (PURUGGANAN 1997 ; HASEBE 1999 ; THEISSEN et al. 2000 ). Parsimony analyses were performed with heuristic search replicates (100 repetitions of random stepwise taxon addition with TBR branch swapping). Bootstrap support (FELSENSTEIN 1985 ) for clades was estimated with 1000 heuristic search replicates (random stepwise taxon addition and TBR branch swapping). Analyses were performed using different representatives of the SEPALLATA and AGL6 gene clades as outgroups (not shown); these different outgroup samples produced no appreciable changes in the topology of the ingroup. An analysis was also performed in which 200 nucleotides from the most variable region of the 3' end of the sequences were eliminated from the aligned data matrix. The results (not shown) were the same except for the placement of AtFL and the monophyly of the non-core eudicot clade.

Core eudicot gene clades were named according to the Arabidopsis gene belonging to that clade (the euAP1 and euFUL clades). Non-core eudicot and non-eudicot gene clades were designated "FUL-like" on the basis of the similarity of the sequences in these clades to those in the euFUL clade. Genes were named according to species and to gene clade membership; thus SvAP1 is the euAP1 gene isolated from the core eudicot S. vulgaris (lilac), PaFUL is the euFUL gene isolated from the core eudicot Phytolacca americana (pokeweed), and MfFL is the FUL-like gene isolated from the magnoliid Michelia figo.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AP1/FUL gene phylogeny is congruent with angiosperm phylogeny:
To generate a phylogeny of the AP1/FUL genes, we cloned representatives of this gene lineage from 19 species representing major clades from across the angiosperms (Fig 1). The sequences were used in a parsimony analysis, along with AP1/FUL and outgroup (SEP, DAL1-like, and AGL6) sequences available in GenBank. The analysis found two most parsimonious trees that differ only in the relative positions of the three Brassicaceae AP1 sequences. The consensus of the two trees is shown in Fig 2 and in simplified form in Fig 3. The structure of the monophyletic AP1/FUL clade in general mirrors angiosperm phylogeny (Fig 1), with successive branches leading to clades that consist of genes from successive branches of the angiosperm phylogenetic tree. The results of the bootstrap analysis (Fig 2) show that although there is strong support for most of the major clades in the AP1/FUL phylogeny, there is little support (<50%) for the arrangement of these clades relative to each other. This suggests that conclusions that rely on the order of branching of the major gene clades must be made with caution. However, the congruence of the most parsimonious gene trees with established angiosperm phylogeny provides corroborating evidence for this topology.

View larger version (34K): In this window In a new window Download PPT slide   Figure 2. Consensus of the two most parsimonious trees. The data included 133 total sequences. Bootstrap values of 50% are indicated for clades discussed in text. Dotted lines indicate bootstrap support of <50% for clades discussed in text. The individual trees were 13,666 steps; CI = 0.16 and RI = 0.58. The only disagreement between the two trees was the relationship of three Brassicaceae genes (black hexagon). Gene clades are indicated by color: green, euAP1; blue, euFUL; purple, core eudicot FUL-like; pink, non-core eudicot FUL-like; red, magnoliid FUL-like; light and dark brown, monocot FUL-like; light gray, SEPALLATA; dark gray, gymnosperm DAL1-like; black, angiosperm AGL6-like. Taxonomic affiliations of species from which individual gene sequences were obtained are indicated by font color: black, core eudicot; brown, ranunculid, orange, other non-core eudicot (Pachysandra); blue, magnoliid; green, monocot; gray, gymnosperm. The black circle indicates non-core eudicot clade, and the single arrow points to the Arabidopsis sequence that groups within this clade. Double arrows point to two Antirrhinum sequences with characteristics of FUL-like genes. Black squares indicate Solanaceae gene clades, X's indicate caryophyllid gene clades, black diamonds indicate asterid gene clades, and the black star indicates rosid euAP1 clade. Asterisks denote sequences generated in this study. See text for details.

View larger version (26K): In this window In a new window Download PPT slide   Figure 3. Summary of the results of the phylogenetic analysis. Major gene clades are represented by triangles. The outgroups consist of SEPALLATA sequences and AGL6-like sequences from gymnosperms and angiosperms.

Examination of the consensus tree turns up only one inconsistency in the correlation between the topology of the AP1/FUL gene tree and the angiosperm phylogenetic tree: the presence of an Arabidopsis sequence (AtFL) in the midst of an otherwise non-core eudicot clade (Fig 2). Analysis of the predicted protein sequence of this open reading frame, which is represented in GenBank only as a genomic fragment, shows that it is most likely a highly divergent paralog. In addition, four of the putative AP1/FUL sequences generated in this study (MfAGL6A and MfAGL6B from M. figo, RbAGL6 from R. bulbosus, and SvAGL6 from S. vulgaris) proved to be more similar to Arabidopsis AGL6 than to the AP1/FUL genes.

The overall topology of the AP1/FUL gene tree corresponds to angiosperm phylogeny, but within the two major clades of core eudicot AP1/FUL genes (the euAP1 and euFUL gene clades, Fig 3), this congruence breaks down. There are subclades composed of genes from specific angiosperm lineages, for instance, Solanaceae and Caryophyllidae (Fig 2); however, not all subclades are present in both the euAP1 and euFUL clades. For example, each has a subclade of asterid sequences, but only the euAP1 clade has a subclade of rosid sequences (Fig 2). In addition, the relative positions of some subclades differ; for instance, in the euFUL clade the caryophyllid sequences are nested within the clade, whereas in the euAP1 clade they are sister group to the rest of the clade (Fig 2). The lack of congruence between the topologies of the euAP1 and euFUL gene clades probably reflects uneven taxonomic sampling in the two clades, but may reflect gene loss in one or both clades.

Phylogeny of the AP1/FUL lineage shows several duplication events:
Inspection of the consensus tree reveals evidence for the occurrence of several duplications during the history of the AP1/FUL gene lineage. The monocot AP1/FUL sequences (monocot FUL-like genes) fall into two successively branching clades (Fig 2 and Fig 3). This suggests a duplication in the gene lineage either prior to the origin of the monocots, with loss of one of the paralogs in later branching angiosperm lineages, or within the monocots, with unequal rates of divergence in the two resulting gene clades (Fig 4). However, the bootstrap analysis shows poor support (<50%) not only for the placement of the two monocot gene clades relative to each other, but also for the monophyly of the larger of the two monocot clades.

View larger version (26K): In this window In a new window Download PPT slide   Figure 4. Relationship of duplication events to angiosperm phylogeny. The positions of the duplication events identified in the phylogenetic analysis of the AP1/FUL gene lineage are indicated on the angiosperm phylogeny shown in Fig 1. See text for explanation.

The ranunculid AP1/FUL genes (non-core eudicot FUL-like genes) also fall into two clades, but in this instance the two clades together form a monophyletic group (Fig 2). This topology suggests an AP1/FUL lineage duplication within the ranunculid lineage (Fig 4). Bootstrap support for the sister-group relationship of the two subclades is <50%; however, most of the ranunculid species sampled are represented by at least one sequence in each of the two clades, providing evidence for a duplication. The absence of a Papaver nudicaule sequence from one of the two clades may be due to incomplete sampling or to loss of one lineage in that species. The sequences of the Pachysandra terminalis group in one of the two ranunculid clades; this position is supported by a moderately high (82%) bootstrap value (Fig 2). Pachysandra is not a ranunculid but is in the same paraphyletic assemblage of non-core eudicots (Fig 1). The presence of Pachysandra genes in only one of the two clades may indicate that the duplication occurred within the ranunculids and that the sequences in the clade that lack the Pachysandra sequences diverged more rapidly from the ancestral preduplication eudicot sequence.

Within the core eudicots there is evidence for two duplications that produced three gene clades: the euAP1 clade, the euFUL clade, and the core eudicot FUL-like clade (Fig 2 Fig 3 Fig 4). Representatives of the euAP1 and euFUL clades have been identified from a wide variety of core eudicot species; however, to date core eudicot FUL-like genes have been identified in only six species. The core eudicot FUL-like clade has strong bootstrap support (93%) and includes genes from all major core eudicot lineages that were sampled for this analysis, suggesting that more intense sampling of other species may uncover additional members of this clade. The core eudicot FUL-like gene clade is sister group to a monophyletic group formed by the euAP1 and euFUL gene clades, but bootstrap analysis shows weak support (<50%) for this position.

Amino acid alignment defines conserved C-terminal motifs:
The predicted amino acid sequences of the genes included in this analysis have the typical "MIKC" structure of plant type II MADS-domain containing proteins (ALVAREZ-BUYLLA et al. 2000B ). Some residues in the M, I, and K domains appear to be diagnostic for the AP1/FUL clade as compared to the SEP, DAL1-like, and AGL6-like sequences. In general, the sequences included in this analysis are highly conserved throughout the MADS and K domains and somewhat more variable but still conserved in the I domain and N-terminal portion of the C terminus. In contrast, much of the C-terminal domain is widely divergent, even in sequences from closely related species. Most of the AP1/FUL sequences contain C-terminal regions rich in glutamine, but regions rich in proline, serine, or glycine are also common. At the very C terminus all the FUL-like and euFUL sequences show a highly conserved hydrophobic six-amino-acid sequence (FUL-like motif: L/MPPWML). This is generally followed by either two basic residues or one polar and one basic residue (Fig 5A). The number of amino acids between the FUL-like motif and the C terminus of the protein varies from five to seven. A related conserved motif (IPGWML) has been reported in SEPALLATA sequences (AMPOMAH-DWAMENA et al. 2002 ), and it can be seen in related DAL1-like sequences (MQGWMV; Fig 5A). The tryptophan in the fourth position is strictly conserved in all sequences included in this analysis, and the residue following the tryptophan is methionine in all but a few AP1/FUL and SEP sequences. The glycine in the third position is highly conserved in the DAL1-like and SEP sequences (see also TANDRE et al. 1995 ) but is replaced by a proline in most FUL-like and euFUL sequences. In the SEP and DAL1-like sequences this motif nearly always terminates the proteins, in contrast to the five- to seven-amino-acid extension that follows the FUL-like motif in FUL-like and euFUL sequences.

View larger version (52K): In this window In a new window Download PPT slide   Figure 5. Conserved C-terminal motifs. (A) C terminus of representative outgroup, FUL-like, euFUL, and euAP1 predicted protein sequences. Sequences marked with an asterisk were generated for this study; those sequences are incomplete at the C terminus, indicated by question marks. Conserved sequence motifs are in boldface type and boxed with a solid line (see text for details). Dotted box shows region of euAP1 sequences with a high percentage of acidic amino acids (see text for details). (B) Frameshift relationship between euFUL and euAP1 motifs. Representative sequences showing evidence that the farnesylation motif present in the predicted protein sequences of euAP1 genes may have evolved from the FUL-like motif through a translational frameshift. In the two examples presented, the top line is the nucleotide sequence, and the following two lines are two of the three possible translation frames. The predicted correct translations are boxed. In the case of CmFL2, the correct frame has the FUL-like motif, but the farnesylation motif can be seen in the incorrect frame. In the case of BpMADS3, the correct frame has the farnesylation motif but five of the six amino acids of the FUL-like motif can be seen in the incorrect frame.

The FUL-like motif, identified in the predicted protein products of monocot, magnoliid, ranunculid, and two groups of core eudicot genes (euFUL and FUL-like), is absent from the predicted products of the euAP1 gene clade. These euAP1 sequences instead have a distinct C terminus with two short conserved motifs, RRNaLaLT/NLa, where "a" is an acidic residue (euAP1 motif), and CFAT/A (farnesylation motif), which terminates the protein (Fig 5A). A variable number of additional acidic residues are just upstream of the euAP1 motif (Fig 5A), and in the case of several euAP1 proteins, this acidic region (including both the euAP1 motif and the upstream region) has been shown to have transcriptional activation properties when tested in yeast (CHO et al. 1999 ). The four terminal amino acids of euAP1 predicted proteins form a farnesylation motif, which is a signal for the attachment of a farnesyl moiety to the cysteine residue; farnesylation causes proteins to be targeted to membranes. YALOVSKY et al. 2000 showed that Arabidopsis AP1 is farnesylated in vivo and in vitro. The euFUL and FUL-like proteins, which lack this motif, are most likely not so modified.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The AP1/FUL genes are found only in angiosperms:
Many of the genes that have been shown in Arabidopsis to be key regulators of floral development (e.g., AP1, AP3, PI, AG, SEP1, SEP2, SEP3) belong to closely related paralogous lineages of the MADS-box gene family. These lineages appear to have arisen as a result of duplication events, although the exact relationship of the lineages to each other and the timing of the duplications is unclear (e.g., PURUGGANAN 1997 ; WINTER et al. 1999 ; KROGAN and ASHTON 2000 ; THEISSEN et al. 2000 ). The AP3/PI (Arabidopsis B-function) and AG (Arabidopsis C-function) lineages are both present in gymnosperms (TANDRE et al. 1995 ; HASEBE 1999 ; SUNDSTROM et al. 1999 ; BECKER et al. 2000 ; THEISSEN et al. 2000 ); thus the origins of the genes responsible for specifying the identity of the male and female reproductive organs predate the origin of the flower. This is consistent with the apparent homology of the male and female gametophytes of angiosperms with those of gymnosperms (ENDRESS 2001 ).

In contrast, our data and previous studies (e.g., TANDRE et al. 1995 ; HASEBE 1999 ; SUNDSTROM et al. 1999 ; BECKER et al. 2000 ; THEISSEN et al. 2000 ) indicate that the AP1/FUL genes, which have been shown in model species to be required for the specification of floral meristem identity (IRISH and SUSSEX 1990 ; HUIJSER et al. 1992 ; BOWMAN et al. 1993 ; FERRANDIZ et al. 2000 ), appear to be unique to angiosperms (Fig 2 and Fig 3). Although the gymnosperm DAL1-like genes share some sequence similarity with AP1/FUL genes, they are more similar to Arabidopsis AGL6, and in all analyses they group outside of the AP1/FUL clade (HASEBE and BANKS 1997 ; PURUGGANAN 1997 ; HASEBE 1999 ; WINTER et al. 1999 ; BECKER et al. 2000 ; THEISSEN et al. 2000 ). Thus AP1/FUL genes, which play a key role in floral specification in model species, appear to be angiosperm specific. This is not the case for the lineages of other genes that are implicated in the specification of floral meristem identity, such as Arabidopsis LEAFY (LFY). Members of the LFY lineage have been identified also in gymnosperms, where they appear to have a role in the specification of reproductive shoot identity (MELLEROWICZ et al. 1998 ; MOURADOV et al. 1998 ; SHINDO et al. 1999 ; FROHLICH and PARKER 2000 ). Thus the function of LFY in determining reproductive identity predates flowers and is not unique to angiosperms. In contrast, the correlation of the origin of the AP1/FUL lineage with the origin of flowers suggests a possible role for these genes in the evolution of this key angiosperm feature.

A duplication at the base of the core eudicots produced the euAP1 and euFUL clades:
Genes of the euAP1 clade are found only in core eudicot species, and these species also possess euFUL genes, thus providing evidence for a duplication that coincided with the origin of this angiosperm clade (Fig 2 Fig 3 Fig 4). Core eudicots, which comprise the majority of extant angiosperm species, have a fixed floral architecture, in contrast to earlier diverging angiosperms, which are more plastic in their floral structure (ENDRESS 1992 , ENDRESS 1994 ; DRINNAN et al. 1994 ; ALBERT et al. 1998 ; SOLTIS et al. 2003 ). In non-eudicot and non-core eudicot species, floral organs may be arranged in discrete whorls, in continuous spirals, or in a combination of both (e.g., whorled perianth but spiral reproductive organs). Particularly in species with spiral phyllotaxy, the number of organs of each type may be variable from flower to flower. In addition, flowers of non-eudicot and non-core eudicot species may have only one type of sterile perianth organ (tepals), rather than a bipartite perianth of differentiated sepals and petals. In the core eudicots these elements of floral structure become fixed; thus all core eudicot flowers have, as a basic plan, a whorled arrangement of four distinct organ types with a fixed number of organs in each whorl.

The fixation of floral structure in the core eudicots suggests that there may have been changes in floral developmental mechanisms that occurred in conjunction with the origin of this angiosperm group. It is thus notable that the duplication event in the AP1/FUL lineage that produced the euAP1 gene clade occurred at the base of the core eudicots and furthermore that the predicted euAP1 amino acid sequences contain novel C-terminal motifs that are postulated to confer new functional capabilities on the euAP1 proteins. The correlation of the origin of the euAP1 gene clade with the fixation of floral structure in the core eudicots suggests that this new protein structure may have played a role in the evolution of the core eudicot flower.

Similar duplications have been identified in the lineages of other MADS-box floral development genes, suggesting that multiple individual gene duplications or a genome-wide duplication event may have played a role in the evolution of core eudicot floral structure. KRAMER et al. 1998 studied the phylogeny of the AP3 (Arabidopsis B-function) gene lineage and identified a duplication event that parallels what is seen in the AP1/FUL lineage. They found two gene clades (euAP3 and TM6) within the core eudicots, whereas outside of the core eudicots they found only one lineage (paleoAP3). Furthermore, they identified a conserved motif in the predicted paleoAP3 amino acid sequences that was maintained in the core eudicot TM6 clade, but that was lost in the euAP3 clade and was replaced by a novel motif. Examination of the phylogeny of the AG lineage (Arabidopsis C-function) also reveals evidence of a duplication that occurred after the divergence of the monocots (HASEBE and BANKS 1997 ; DAVIES et al. 1999 ; E. KRAMER, personal communication). Thus in the lineages of three key regulators of floral development, AP1, AP3, and AG, we can identify significant evolutionary events that are correlated with the diversification of the core eudicots. This suggests that there may have been changes in the genetic mechanisms regulating floral development that occurred in conjunction with the origin of the core eudicots. Associating specific sequence motif changes with the evolution of particular morphological novelties will require functional analyses of these genes in core eudicot and noncore eudicot species (e.g., LAMB and IRISH 2003 ).

Additional duplications occurred during the evolution of the AP1/FUL lineage:
The topology of the most parsimonious trees found in this analysis indicates that there have been at least three other duplications within the AP1/FUL lineage. In addition to the euAP1 and euFUL core eudicot gene clades, the phylogeny shows a third clade of core-eudicot sequences, the core eudicot FUL-like genes. The presence of three clades of core eudicot genes suggests that there were two AP1/FUL lineage duplication events within the core eudicots (Fig 2 Fig 3 Fig 4). However, bootstrap support for this position of the core eudicot FUL-like gene clade is weak (<50%; Fig 2). Preliminary analyses based on a slightly smaller data set suggested that the additional duplication occurred at the base of the eudicots, rather than within the core eudicots (results not shown); not surprisingly, bootstrap support for that topology was also low (<50%).

The presence of two clades of monocot FUL-like genes is evidence of another duplication. The observed topology (Fig 2 and Fig 3) of successively branching monocot FUL-like gene clades implies that the duplication occurred prior to the origin of the monocots. This requires that both resulting paralogous lineages were maintained in the monocots, but that one of the lineages was lost in later branching angiosperm groups. An alternative explanation is suggested by the uneven taxon representation in the two clades. The smaller, earlier branching clade is composed of genes from three species (Tradescantia, Oryza, and Hordeum), whereas the larger clade is composed of genes from these three species and seven more. Tradescantia, Oryza, and Hordeum are all members of one lineage of monocots, the commelinoids, suggesting that the duplication may have occurred within the commelinoid monocots. Under this scenario, the successive branching pattern of the two monocot FUL-like gene clades would most likely be due to a higher rate of divergence in the smaller clade. AP1/FUL genes from earlier branching angiosperm lineages are needed to clarify the position of this duplication.

The results of our analysis also indicate a duplication within the ranunculids. The ranunculid FUL-like genes group in two subclades, one of which is moderately well supported (82%) but one of which has weak (<50%) bootstrap support (Fig 2). The two clades together form a weakly supported (<50%) monophyletic group, with most species being represented in both clades. This topology suggests a single duplication at the base of the ranunculids. Evidence of duplication events within the ranunculids was also seen in phylogenetic analyses of the AP3 and PI gene family (KRAMER and IRISH 1999 ). The results of that study point to several separate duplication events within different ranunculid lineages. In contrast, our analysis of the AP1/FUL lineage suggests a single duplication at the base of the ranunculid clade.

Phylogenetic analysis clarifies orthology and paralogy:
The phylogenetic analysis presented here provides a framework for the assessment of the orthology and paralogy of AP1/FUL genes by identifying duplication events in the history of this gene lineage and by defining the resulting paralogous gene clades. Previous studies have not had a basis for determining orthology or paralogy of newly identified AP1/FUL genes, although differences between euAP1 and euFUL genes have been noted (ELO et al. 2001 ; GOCAL et al. 2001 ; JANG et al. 2002 ). As a result, comparisons have been made among paralogous genes (e.g., KYOZUKA et al. 1997 ; POUTEAU et al. 1997 ; IMMINK et al. 1999 ; JIA et al. 2000 ; GOCAL et al. 2001 ). Our analysis clearly demonstrates that the euAP1 sequences form a derived clade within the AP1/FUL family (Fig 2, Fig 3, and Fig 5A). Thus non-core eudicot and non-eudicot AP1/FUL genes, such as those of monocots, magnoliids, and ranunculids, are most appropriately compared with other FUL-like and with euFUL genes, and not with euAP1 genes. Likewise, comparisons between core eudicot euAP1 and euFUL genes should be done in the context of their belonging to paralogous gene lineages and possessing different sequence motifs.

The presence of two distinct clades of core eudicot genes with FUL-like sequence characteristics (euFUL and core eudicot FUL-like; Fig 2 and Fig 3) suggests that designation of genes as FUL orthologs (members of the euFUL clade) may be difficult in the absence of a phylogenetic analysis. On the basis of sequence examination alone it is difficult to determine if a given core eudicot gene belongs to the euFUL or FUL-like clade. For instance, MULLER et al. 2001 suggested DEFH28 as the probable Antirrhinum FUL ortholog on the basis of analysis of the coding sequence and promoter region, the similarities of expression pattern, and the overexpression phenotype in Arabidopsis. In our analysis, however, DEFH28 groups with the core eudicot FUL-like genes, not with the euFUL clade. We cloned, in addition to SQUA and DEFH28, a third Antirrhinum gene (here designated AmFUL) that groups with the euFUL clade and thus can be considered the ortholog of FUL. This obviously does not preclude the possibility that DEFH28 has functional roles similar to those of FUL, as suggested by MULLER et al. 2001 . DEFH28 and AmFUL may have redundant functions, a suggestion supported by the close sequence similarity of the euFUL and core eudicot FUL-like genes.

Diversification of C-terminal domains:
The C-terminal domain of AP1/FUL predicted proteins is highly variable, as is characteristic of plant MADS-domain-containing proteins. Nonetheless, there is a strongly conserved hydrophobic six-amino-acid motif at the end of all FUL-like and euFUL proteins. This FUL-like motif can be seen in the outgroup (SEP, DAL1-like, and AGL6) sequences (Fig 5A), although the exact residue composition is not strictly conserved. The high degree of conservation of this motif is a strong indication that it is functionally important and suggests that its loss and replacement with a different motif in euAP1 proteins may result in altered functional capabilities of the euAP1 proteins.

Several studies have investigated the significance of the C-terminal domain and the conserved motifs of euAP1 proteins. KRIZEK and MEYEROWITZ 1996 made constructs consisting of the MI region of AP1 and the KC region of AGAMOUS and found that under the control of the CaMV 35S promoter the construct was able to partially rescue the strong ap1-1 mutant. Constructs consisting only of the AP1 MI domains failed to produce an overexpression phenotype in wild-type Arabidopsis. KRIZEK and MEYEROWITZ 1996 concluded that although K and C domains are required for proper AP1 protein function, the specific motifs present in the AP1 K and C domains are not completely required and can be partially replaced by those of AG.

In contrast, other studies have localized specific functions of Arabidopsis AP1 and Antirrhinum SQUA to the C-terminal domain (CHO et al. 1999 ; EGEA-CORTINES et al. 1999 ; YALOVSKY et al. 2000 ). A domain rich in acidic residues (Fig 5A) has been identified in Arabidopsis AP1 and the putative euAP1 orthologs from Raphanus sativus, Nicotiana sylvestris, and N. tabacum as a transcription activation domain (CHO et al. 1999 ). In Arabidopsis AP1 and its ortholog from the closely related Raphanus this domain was found to function synergistically with an upstream glutamine-rich region, the two regions together producing a higher activation level than the sum of that produced by the two regions separately. This upstream region is lacking from the Nicotiana sequences, which correspondingly showed lower levels of activation. The euFUL and FUL-like predicted proteins lack an acidic domain, but most have regions rich in glutamine or proline. Glutamine- and proline-rich regions have been shown to confer transcription activation activity (GERBER et al. 1994 ), but this function has not been tested in these proteins. The high degree of conservation of the acidic domain in euAP1 sequences, however, suggests that these proteins will show strong transcription activation activity.

The final four amino acids of the predicted proteins of euAP1 genes conform to a farnesylation signal (CaaX), which has been shown to be functional in AP1 and to be required to produce an AP1 overexpression phenotype in Arabidopsis (YALOVSKY et al. 2000 ). Research in other species, however, is equivocal as to whether this farnesylation motif is required for proper functioning of these proteins. The predicted product of PEAM4, the euAP1 gene from P. sativum, ends after the second amino acid of the farnesylation motif. Nonetheless, when overexpressed in Arabidopsis in the strong ap1-1 mutant background, PEAM4 is able largely to rescue the mutant phenotype (BERBEL et al. 2001 ). NtMADS11, from N. tabacum (JANG et al. 2002 ), and LtMADS2, from Lolium temulentum (GOCAL et al. 2001 ), are FUL-like genes, the predicted products of which lack the farnesylation motif entirely, yet, when overexpressed in Arabidopsis, they also show partial rescue of strong ap1 mutants. In all these cases the heterologous gene is able partially to replace AP1 function, but not completely. One can therefore interpret this as meaning either that the farnesylation motif is unimportant, because proteins that lack it can partly substitute for AP1 functionally, or that the farnesylation motif is required, because proteins that lack it are not able to completely substitute for AP1. However, the high degree of conservation suggests that this motif is likely to play an important role in modulating euAP1 gene function.

New motifs in euAP1 proteins may have arisen via translational frameshift:
Inspection of different possible translation frames of ranunculid FUL-like and various core eudicot euAP1 sequences shows that the change in C-terminal motifs may have arisen at least in part via a simple translational frameshift. For instance, the translation of CmFL2, a noncore eudicot FUL-like gene identified from Chelidonium (Fig 1), in one frame terminates with RLCPPGCFIT, the final four amino acids of which form a canonical farnesylation motif. However, in the correct frame the translation is LMPGWMLHH, which lacks the farnesylation motif but has the expected FUL-like motif (Fig 5B). Evidence for this frameshift can be seen in the different translation frames of some genes from non-core eudicot species, which have only FUL-like genes (as in the Chelidonium example above), as well as in genes from core eudicot species, which possess euAP1, euFUL, and FUL-like genes. For example, the correct translation of the Betula euAP1 gene BpMADS3 terminates with CHLGCFAT, whereas one of the two alternative translations terminates with MSPWMLCH, which contains five of the six residues characteristic of the FUL-like motif, including the strictly conserved tryptophan (Fig 5B). Thus the farnesylation motif characteristic of the predicted protein products of the euAP1 genes may have been derived by insertion of a single nucleotide or by loss of two nucleotides upstream of the FUL-like motif of an ancestral FUL-like gene.

Implications for the ABC model of floral organ specification:
Arabidopsis AP1 and Antirrhinum SQUA are members of the euAP1 clade, and likewise Arabidopsis AP3 and its Antirrhinum ortholog DEF are members of the euAP3 clade (KRAMER et al. 1998 ); these clades are core-eudicot specific, and genes with those sequence properties are not found in species outside of the core eudicots. AP1/SQUA and AP3/DEF are among the key regulatory genes upon which the ABC model is based; however, non-core eudicot species such as ranunculids, monocots, and magnoliids do not have euAP1 or euAP3 genes. By extension, these species lack proteins with the conserved motifs characteristic of euAP1 and euAP3 proteins. The correlation between the origin of these derived genes and the origin of the core eudicots suggests that at that point in angiosperm evolution there were changes in the mechanisms regulating floral development. Thus we must consider the possibility that outside of the core eudicots the molecular mechanisms underlying floral development may differ from what is seen in Arabidopsis and Antirrhinum.

Arabidopsis remains the only species identified so far in which a mutant for a gene belonging to the AP1/FUL lineage results in a misspecification of organ identity (IRISH and SUSSEX 1990 ; MANDEL et al. 1992 ; BOWMAN et al. 1993 ). However, Arabidopsis has three closely related genes, AP1, FUL, and CAL, with redundant roles in floral meristem specification (FERRANDIZ et al. 2000 ); thus it is likely that even strong ap1 mutants show only a partial loss of function for this role. In strong ap1 mutants, the meristem shows some floral identity, particularly in the inner two whorls, in which the reproductive organs develop nearly normally. The outer two whorls show some transition to floral character in the whorled arrangement of the organs, but their leaf-like nature and the presence of secondary flowers demonstrate a persistent inflorescence character. Thus the loss of proper organ identity may be a consequence of the incomplete nature of the transition from inflorescence to floral meristem, particularly in the outer two whorls.

According to this interpretation of the role of the AP1 gene, there is no discrete A-function; rather, the apparent misspecification of floral organ identity in the Arabidopsis ap1 mutant is a consequence of the incomplete specification of floral meristem identity. If sepal production represents the ground state function of a florally determined meristem, floral organ identity can be adequately specified with only the equivalent of the B- and C-function of the ABC model, as articulated by SCHWARZ-SOMMER et al. 1990 (see also EGEA-CORTINES and DAVIES 2000 ; THEISSEN et al. 2000 ). Thus the lack of examples of A-function mutants, coupled with the data presented here regarding the restriction of the euAP1 genes to the core eudicots, suggests that the universality and perhaps the concept of A-function should be reevaluated.

	FOOTNOTES

Sequence data from this article have been deposited with the GenBank Data Library under accession nos. AY306138, AY306139, AY306140, AY306141, AY306142, AY306143, AY306144, AY306145, AY306146, AY306147, AY306148, AY306149, AY306150, AY306151, AY306152, AY306153, AY306154, AY306155, AY306156, AY306157, AY306158, AY306159, AY306160, AY306161, AY306162, AY306163, AY306164, AY306165, AY306166, AY306167, AY306168, AY306169, AY306170, AY306171, AY306172, AY306173, AY306174, AY306175, AY306176, AY306177, AY306178, AY306179, AY306180, AY306181, AY306182, AY306183, AY306184, AY306185, AY306186, AY306187, AY306188, AY306189, AY306190, AY306191, AY306192, AY306193, AY306194, AY306195.

	ACKNOWLEDGMENTS

The authors thank Akira Kanno, Francisco Madueño, and Elena Kramer for permission to use unpublished sequences, Jongmin Nam for use of the aa2dna program, and our colleagues, in particular Lena Hileman, for comments on the manuscript. This work was supported by grant no. 2001-35304-09901 from the U.S. Department of Agriculture Cooperative State Research, Education, and Extension Service to A.L.

Manuscript received March 21, 2003; Accepted for publication June 11, 2003.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ALBERT, V. A., M. H. G. GUSTAFSSON and L. DILAURENZIO, 1998 Ontogenetic systematics, molecular developmental genetics, and the angiosperm petal, pp. 349–374 in Molecular Systematics of Plants II, edited by D. E. SOLTIS, P. S. SOLTIS and J. J. DOYLE. Kluwer Academic Publishers, New York.

ALTSCHUL, S. F., T. L. MADDEN, A. A. SCHAFFER, J. ZHANG, and Z. ZHANG et al., 1997  Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25:3389-3402.[Abstract/Free Full Text]

ALVAREZ-BUYLLA, E. R., S. J. LILJEGREN, S. PELAZ, S. E. GOLD, and C. BURGEFF et al., 2000a  MADS box gene evolution beyond flowers: expression in pollen, endosperm, guard cells, roots and trichomes. Plant J. 24:457-466.[Medline]

ALVAREZ-BUYLLA, E. R., S. PELAZ, S. J. LILJEGREN, S. E. GOLD, and C. BURGEFF et al., 2000b  An ancestral MADS-box gene duplication occurred before the divergence of plants and animals. Proc. Natl. Acad. Sci. USA 97:5328-5333.[Abstract/Free Full Text]

AMPOMAH-DWAMENA, C., B. A. MORRIS, P. SUTHERLAND, B. VEIT, and J.-L. YAO, 2002  Down-regulation of TM29, a tomato SEPALLATA homolog, causes parthenocarpic fruit development and floral reversion. Plant Physiol. 130:605-617.[Abstract/Free Full Text]

ANGENENT, G. C., J. FRANKEN, M. BUSSCHER, A. VAN DIJKEN, and J. L. VAN WENT et al., 1995  A novel class of MADS box genes is involved in ovule development in Petunia. Plant Cell 7:1569-1582.[Abstract]

An ordinal classification for the families of flowering plants. (1998) Ann. Mo. Bot. Gard. 85:531-553.

BECKER, A., K.-U. WINTER, H. SAEDLER, and G. THEISSEN, 2000  MADS-box gene diversity in seed plants 300 million years ago. Mol. Biol. Evol. 17:1425-1434.[Abstract/Free Full Text]

BERBEL, A., C. NAVARRO, C. FERRANDIZ, L. A. CANAS, and F. MADUENO et al., 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. Plant J. 25:441-451.[Medline]

BOWMAN, J. L., J. ALVAREZ, D. WEIGEL, E. M. MEYEROWITZ, and D. R. SMYTH, 1993  Control of flower development in Arabidopsis thaliana by APETALA1 and interacting genes. Development 119:721-743.[Abstract]

BUCHNER, P. and J.-P. BOUTIN, 1998  A MADS box transcription factor of the AP1/AGL9 subfamily is also expressed in the seed coat of pea (Pisum sativum) during development. Plant Mol. Biol. 38:1253-1255.[Medline]

CHO, S., S. JANG, S. CHAE, K. M. CHUNG, and Y.-H. MOON et al., 1999  Analysis of the C-terminal region of Arabidopsis thaliana APETALA1 as a transcription activation domain. Plant Mol. Biol. 40:419-429.[Medline]

COEN, E. S. and E. M. MEYEROWITZ, 1991  The war of the whorls: genetic interactions controlling flower development. Nature 353:31-37.[Medline]

DAVIES, B., P. MOTTE, E. KECK, H. SAEDLER, and H. SOMMER et al., 1999  PLENA and FARINELLI: redundancy and regulatory interactions between two antirrhinum MADS-box factors controlling flower development. EMBO J. 18:4023-4034.[Medline]

DOYLE, J. J., 1994  Evolution of a plant homeotic multigene family: toward connecting molecular systematics and molecular developmental genetics. Syst. Biol. 43:307-328.

DRINNAN, A. N., P. R. CRANE, and S. B. HOOT, 1994  Patterns of floral evolution in the early diversification of non-magnoliid dicotyledons (eudicots). Plant Syst. Evol. 8:93-122.

EGEA-CORTINES, M. and B. DAVIES, 2000  Beyond the ABCs: ternary complex formation in the control of floral organ identity. Trends Plant Sci. 5:471-476.[Medline]

EGEA-CORTINES, M., H. SAEDLER, and H. SOMMER, 1999  Ternary complex formation between the MADS-box proteins SQUAMOSA, DEFICIENS and GLOBOSA is involved in the control of floral architecture in Antirrhinum majus. EMBO J. 18:5370-5379.[Medline]

ELO, A., J. LEMMETYINEN, M. L. TURUNEN, L. TIKKA, and T. SOPANEN, 2001  Three MADS-box genes similar to APETALA1 and FRUITFULL from silver birch (Betula pendula). Physiol. Plant 112:95-103.[Medline]

ENDRESS, P. K., 1992  Evolution and floral diversity: the phylogenetic surroundings of Arabidopsis and Antirrhinum.. Int. J. Plant Sci. 153:S106-S122.

ENDRESS, P. K., 1994  Floral structure and evolution of primitive angiosperms: recent advances. Plant Syst. Evol. 192:79-97.

ENDRESS, P. K., 2001  Origins of flower morphology. J. Exp. Zool. 291:105-115.[Medline]

FELSENSTEIN, J., 1985  Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39:783-791.

FERRANDIZ, C., Q. GU, R. MARTIENSSEN, and M. F. YANOFSKY, 2000  Redundant regulation of meristem identity and plant architecture by FRUITFULL, APETALA1 and CAULIFLOWER. Development 127:725-734.[Abstract]

FROHLICH, M. W. and D. S. PARKER, 2000  The mostly male theory of flower evolutionary origins: from genes to fossils. Syst. Bot. 25:155-170.

GERBER, H. P., K. SEIPEL, O. GEORGIEV, M. HOFFERER, and M. HUG et al., 1994  Transcriptional activation modulated by homopolymeric glutamine and proline stretches. Science 263:808-811.[Abstract/Free Full Text]

GOCAL, G. F., R. W. KING, C. A. BLUNDELL, O. M. SCHWARTZ, and C. H. ANDERSEN et al., 2001  Evolution of floral meristem identity genes. Analysis of Lolium temulentum genes related to APETALA1 and LEAFY of Arabidopsis. Plant Physiol. 125:1788-1801.[Abstract/Free Full Text]

GU, Q., C. FERRANDIZ, M. F. YANOFSKY, and R. MARTIENSSEN, 1998  The FRUITFULL MADS-box gene mediates cell differentiation during Arabidopsis fruit development. Development 125:1509-1517.[Abstract]

HASEBE, M., 1999  Evolution of reproductive organs in land plants. J. Plant Res. 112:463-474.

HASEBE, M., and J. A. BANKS, 1997 Evolution of MADS gene family in plants, pp. 179–198 in Evolution and Diversification of Land Plants, edited by K. IWATSUKI and P. H. RAVEN. Springer, Tokyo.

HUIJSER, P., J. KLEIN, W.-E. LONNIG, H. MEIJER, and H. SAEDLER et al., 1992  Bracteomania, an inflorescence anomaly, is caused by the loss of function of the MADS-box gene squamosa in Antirrhinum majus.. EMBO J. 11:1239-1249.[Medline]

IMMINK, R. G. H., D. J. HANNAPEL, S. FERRARIO, M. BUSSCHER, and J. FRANKEN et al., 1999  A petunia MADS box gene involved in the transition from vegetative to reproductive development. Development 126:5117-5126.[Abstract]

IRISH, V. F. and I. M. SUSSEX, 1990  Function of the apetala-1 gene during Arabidopsis floral development. Plant Cell 2:741-753.[Abstract/Free Full Text]

JANG, S., K. AN, S. LEE, and G. AN, 2002  Characterization of tobacco MADS-box genes involved in floral initiation. Plant Cell Physiol. 43:230-238.[Abstract/Free Full Text]

JIA, H., R. CHEN, B. CONG, K. CAO, and C. SUN et al., 2000  Characterization and transcriptional profiles of two rice MADS-box genes. Plant Sci. 155:115-122.[Medline]

KEMPIN, S., B. SAVIDGE, and M. F. YANOFSKY, 1995  Molecular basis of the cauliflower phenotype in Arabidopsis. Science 267:522-525.[Abstract/Free Full Text]

KRAMER, E. M. and V. F. IRISH, 1999  Evolution of genetic mechanisms controlling petal development. Nature 399:144-148.[Medline]

KRAMER, E. M., R. L. DORIT, and V. F. IRISH, 1998  Molecular evolution of petal and stamen development: gene duplication and divergence within the APETALA3 and PISTILLATA MADS-box gene lineages. Genetics 149:765-783.[Abstract/Free Full Text]

KRIZEK, B. A. and E. M. MEYEROWITZ, 1996  Mapping the protein regions responsible for the functional specificities of the Arabidopsis MADS domain organ-identity proteins. Proc. Natl. Acad. Sci. USA 93:4063-4070.[Abstract/Free Full Text]

KROGAN, N. T. and N. W. ASHTON, 2000  Ancestry of plant MADS-box genes revealed by Bryophyte (Physcomitrella patens) homologues. New Phytol. 147:505-517.

KUMAR, S., K. TAMURA, I. JAKOBSEN and M. NEI, 2001 MEGA2: molecular evolutionary genetics analysis software, Arizona State University, Tempe, AZ.

KYOZUKA, J., R. HARCOURT, W. J. PEACOCK, and E. S. DENNIS, 1997  Eucalyptus has functional equivalents of the Arabidopsis AP1 gene. Plant Mol. Biol. 35:573-584.[Medline]

LAMB, R. S. and V. F. IRISH, 2003  Functional divergence within the PETALA3/PISTILLATA floral homeotic gene lineages. Proc. Natl. Acad. Sci. USA 100:6558-6563.[Abstract/Free Full Text]

LAWTON-RAUH, A. L., E. R. ALVAREZ-BUYLLA, and M. D. PURUGGANAN, 2000  Molecular evolution of flower development. Trends Ecol. Evol. 15:144-149.[Medline]

LOWMAN, A. C. and M. D. PURUGGANAN, 1999  Duplication of the Brassica oleracea APETALA1 floral homeotic gene and the evolution of domesticated cauliflower. Genetics 90:514-520.

MANDEL, M. A., C. GUSTAFSON-BROWN, and M. F. YANOFSKY, 1992  Molecular characterization of the Arabidopsis floral homeotic gene APETALA1. Nature 360:273-277.[Medline]

MELLEROWICZ, E. J., K. HORGAN, A. WALDEN, A. COKER, and C. WALTER, 1998  PRFLL—a Pinus radiata homologue of FLORICAULA and LEAFY is expressed in buds containing vegetative shoot and undifferentiated male cone primordia. Planta 206:619-629.[Medline]

MOON, Y. H., H. G. KANG, J. Y. JUNG, J. S. JEON, and S. K. SUNG et al., 1999  Determination of the motif responsible for interaction between the rice APETALA1/AGAMOUS-LIKE9 family proteins using a yeast two-hybrid system. Plant Physiol. 120:1193-1204.[Abstract/Free Full Text]

MOURADOV, A., T. GLASSICK, B. HAMDORF, L. MURPHY, and B. FOWLER et al., 1998  NEEDLY, a Pinus radiata ortholog of FLORICAULA/LEAFY genes, expressed in both reproductive and vegetative meristems. Proc. Natl. Acad. Sci. USA 95:6537-6542.[Abstract/Free Full Text]

MÜLLER, B. M., H. SAEDLER, and S. ZACHGO, 2001  The MADS-box gene DEFH28 from Antirrhinum is involved in the regulation of floral meristem identity and fruit development. Plant J. 28:169-179.[Medline]

PELUCCHI, N., F. FORNARA, C. FAVALLI, S. MASIERO, and C. LAGO et al., 2002  Comparative analysis of rice MADS-box genes expressed during flower development. Sex. Plant Reprod. 15:113-122.

POUTEAU, S., D. NICHOLLS, F. TOOKE, E. S. COEN, and N. BATTEY, 1997  The induction and maintenance of flowering in Impatiens.. Development 124:3343-3351.[Abstract]

PURUGGANAN, M. D., 1997  The MADS-box floral homeotic gene lineages predate the origin of seed plants: phylogenetic and molecular clock estimates. J. Mol. Evol. 45:392-396.[Medline]

PURUGGANAN, M. D. and J. I. SUDDITH, 1998  Molecular population genetics of the Arabidopsis CAULIFLOWER regulatory gene: evolution and naturally occurring variation in floral homeotic function. Proc. Natl. Acad. Sci. USA 95:8130-8134.[Abstract/Free Full Text]

PURUGGANAN, M. D., S. D. ROUNSLEY, R. J. SCHMIDT, and M. F. YANOFSKY, 1995  Molecular evolution of flower development: diversification of the plant MADS box regulatory gene family. Genetics 140:345-356.[Abstract]

ROUNSLEY, S. D., G. S. DITTA, and M. F. YANOFSKY, 1995  Diverse roles for MADS box genes in Arabidopsis development. Plant Cell 7:1259-1269.[Abstract]

SAVOLAINEN, V., M. W. CHASE, S. B. HOOT, C. M. MORTON, and D. E. SOLTIS et al., 2000  Phylogenetics of flowering plants based on combined analysis of plastid atpB and rbcL gene sequences. Syst. Biol. 49:306-362.[Medline]

SCHWARZ-SOMMER, Z., P. HUIJSER, W. NACKEN, H. SAEDLER, and H. SOMMER, 1990  Genetic control of flower development by homeotic genes in Antirrhinum majus. Science 250:931-936.[Abstract/Free Full Text]

SHINDO, S., M. ITO, K. UEDA, M. KATO, and M. HASEBE, 1999  Characterization of MADS genes in the gymnosperm Gnetum parvifolium and its implication on the evolution of reproductive organs in seed plants. Evol. Dev. 1:180-190.[Medline]

SOLTIS, E. D. and P. S. SOLTIS, 2000  Contributions of plant molecular systematics to studies of molecular evolution. Plant Mol. Biol. 42:45-75.[Medline]

SOLTIS, D. E., A. E. SENTERS, M. J. ZANIS, S. KIM, and J. D. THOMPSON et al., 2003  Gunnerales are sister to other core eudicots and exhibit some floral features of early-diverging eudicots. Am. J. Bot. 90:461-470.[Abstract/Free Full Text]

SOUTHERTON, S., H. MARSHALL, M. AIDYN, and R. D. TEASDALE, 1998  Eucalyptus MADS-box genes expressed in developing flowers. Plant Physiol. 118:365-372.[Abstract/Free Full Text]

SUNDSTRÖM, J., A. CARLSBECKER, M. E. SVENSSON, M. SVENSON, and U. JOHANSON et al., 1999  MADS-box genes active in developing pollen cones of Norway spruce (Picea abies) are homologous to the B-class floral homeotic genes in angiosperms. Dev. Genet. 25:253-266.[Medline]

SWOFFORD, D. L., 2000 PAUP*: Phylogenetic Analysis Using Parsimony (and Other Methods). Sinauer Associates, Sunderland, MA.

TANDRE, K., V. A. ALBERT, A. SUNDAS, and P. ENGSTROM, 1995  Conifer homologues to genes that control floral development in angiosperms. Plant Mol. Biol. 27:69-78.[Medline]

TAYLOR, S. A., J. M. I. HOFER, I. C. MURFET, J. D. SOLLINGER, and S. R. SINGER et al., 2002  PROLIFERATING INFLORESCENCE MERISTEM, a MADS-box gene that regulates floral meristem identity in Pea. Plant Physiol. 129:1150-1159.[Abstract/Free Full Text]

THEISSEN, G., A. BECKER, A. DI ROSA, A. KANNO, and J. T. KIM et al., 2000  A short history of MADS box genes in plants. Plant Mol. Biol. 42:115-149.[Medline]

THOMPSON, J. D., D. G. HIGGINS, and T. J. GIBSON, 1994  CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, positions-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22:4673-4680.[Abstract/Free Full Text]

WEIGEL, D. and E. M. MEYEROWITZ, 1994  The ABCs of floral homeotic genes. Cell 78:203-209.[Medline]

WINTER, K. U., A. BECKER, T. MUNSTER, J. T. KIM, and H. SAEDLER et al., 1999  MADS-box genes reveal that gnetophytes are more closely related to conifers than to flowering plants. Proc. Natl. Acad. Sci. USA 96:7342-7347.[Abstract/Free Full Text]

YALOVSKY, S., M. RODRIGUEZ-CONCEPCION, K. BRACHA, G. TOLEDO-ORTIZ, and W. GRUISSEM, 2000  Prenylation of the floral transcription factor APETALA1 modulates its function. Plant Cell 12:1257-1266.[Abstract/Free Full Text]

This article has been cited by other articles:

				V. F. Irish Evolution of petal identity J. Exp. Bot., July 1, 2009; 60(9): 2517 - 2527. [Abstract] [Full Text] [PDF]

				D. G. Howarth and M. J. Donoghue Duplications and Expression of DIVARICATA-Like Genes in Dipsacales Mol. Biol. Evol., June 1, 2009; 26(6): 1245 - 1258. [Abstract] [Full Text] [PDF]

				P. S. Soltis, S. F. Brockington, M.-J. Yoo, A. Piedrahita, M. Latvis, M. J. Moore, A. S. Chanderbali, and D. E. Soltis Floral variation and floral genetics in basal angiosperms Am. J. Botany, January 1, 2009; 96(1): 110 - 128. [Abstract] [Full Text] [PDF]

				D. E. Soltis, V. A. Albert, J. Leebens-Mack, C. D. Bell, A. H. Paterson, C. Zheng, D. Sankoff, C. W. dePamphilis, P. K. Wall, and P. S. Soltis Polyploidy and angiosperm diversification Am. J. Botany, January 1, 2009; 96(1): 336 - 348. [Abstract] [Full Text] [PDF]

				B. E. Thompson and S. Hake Translational Biology: From Arabidopsis Flowers to Grass Inflorescence Architecture Plant Physiology, January 1, 2009; 149(1): 38 - 45. [Full Text] [PDF]

				O. N. Danilevskaya, X. Meng, D. A. Selinger, S. Deschamps, P. Hermon, G. Vansant, R. Gupta, E. V. Ananiev, and M. G. Muszynski Involvement of the MADS-Box Gene ZMM4 in Floral Induction and Inflorescence Development in Maize Plant Physiology, August 1, 2008; 147(4): 2054 - 2069. [Abstract] [Full Text] [PDF]

				M. J. Carmona, J. Chaib, J. M. Martinez-Zapater, and M. R. Thomas A molecular genetic perspective of reproductive development in grapevine J. Exp. Bot., July 1, 2008; 59(10): 2579 - 2596. [Abstract] [Full Text] [PDF]

				M.-K. Chen, I-C. Lin, and C.-H. Yang Functional Analysis of Three Lily (Lilium longiflorum) APETALA1-like MADS Box Genes in Regulating Floral Transition and Formation Plant Cell Physiol., May 1, 2008; 49(5): 704 - 717. [Abstract] [Full Text] [PDF]

				D. V. Duclos and T. Bjorkman Meristem identity gene expression during curd proliferation and flower initiation in Brassica oleracea J. Exp. Bot., February 1, 2008; 59(2): 421 - 433. [Abstract] [Full Text] [PDF]

				S. Drea, L. C. Hileman, G. de Martino, and V. F. Irish Functional analyses of genetic pathways controlling petal specification in poppy Development, December 1, 2007; 134(23): 4157 - 4166. [Abstract] [Full Text] [PDF]

				L. P. Ronse De Craene Are Petals Sterile Stamens or Bracts? The Origin and Evolution of Petals in the Core Eudicots Ann. Bot., September 1, 2007; 100(3): 621 - 630. [Abstract] [Full Text] [PDF]

				R. Benlloch, A. Berbel, A. Serrano-Mislata, and F. Madueno Floral Initiation and Inflorescence Architecture: A Comparative View Ann. Bot., September 1, 2007; 100(3): 659 - 676. [Abstract] [Full Text] [PDF]

				D. E. Soltis, A. S. Chanderbali, S. Kim, M. Buzgo, and P. S. Soltis The ABC Model and its Applicability to Basal Angiosperms Ann. Bot., August 1, 2007; 100(2): 155 - 163. [Abstract] [Full Text] [PDF]

				N. Shitsukawa, C. Tahira, K.-i. Kassai, C. Hirabayashi, T. Shimizu, S. Takumi, K. Mochida, K. Kawaura, Y. Ogihara, and K. Murai Genetic and Epigenetic Alteration among Three Homoeologous Genes of a Class E MADS Box Gene in Hexaploid Wheat PLANT CELL, June 1, 2007; 19(6): 1723 - 1737. [Abstract] [Full Text] [PDF]

				C. Liu, J. Zhou, K. Bracha-Drori, S. Yalovsky, T. Ito, and H. Yu Specification of Arabidopsis floral meristem identity by repression of flowering time genes Development, May 15, 2007; 134(10): 1901 - 1910. [Abstract] [Full Text] [PDF]

				L. C. Hileman, J. F. Sundstrom, A. Litt, M. Chen, T. Shumba, and V. F. Irish Molecular and Phylogenetic Analyses of the MADS-Box Gene Family in Tomato Mol. Biol. Evol., November 1, 2006; 23(11): 2245 - 2258. [Abstract] [Full Text] [PDF]

				R. Benlloch, I. d'Erfurth, C. Ferrandiz, V. Cosson, J. P. Beltran, L. A. Canas, A. Kondorosi, F. Madueno, and P. Ratet Isolation of mtpim Proves Tnt1 a Useful Reverse Genetics Tool in Medicago truncatula and Uncovers New Aspects of AP1-Like Functions in Legumes Plant Physiology, November 1, 2006; 142(3): 972 - 983. [Abstract] [Full Text] [PDF]

				R. M. Bateman, J. Hilton, and P. J. Rudall Morphological and molecular phylogenetic context of the angiosperms: contrasting the 'top-down' and 'bottom-up' approaches used to infer the likely characteristics of the first flowers J. Exp. Bot., October 1, 2006; 57(13): 3471 - 3503. [Abstract] [Full Text] [PDF]

				M. M. Kater, L. Dreni, and L. Colombo Functional conservation of MADS-box factors controlling floral organ identity in rice and Arabidopsis J. Exp. Bot., October 1, 2006; 57(13): 3433 - 3444. [Abstract] [Full Text] [PDF]

				E. A. Kellogg Progress and challenges in studies of the evolution of development J. Exp. Bot., October 1, 2006; 57(13): 3505 - 3516. [Abstract] [Full Text] [PDF]

				J. C. Preston and E. A. Kellogg Reconstructing the Evolutionary History of Paralogous APETALA1/FRUITFULL-Like Genes in Grasses (Poaceae) Genetics, September 1, 2006; 174(1): 421 - 437. [Abstract] [Full Text] [PDF]

				G. de Martino, I. Pan, E. Emmanuel, A. Levy, and V. F. Irish Functional Analyses of Two Tomato APETALA3 Genes Demonstrate Diversification in Their Roles in Regulating Floral Development PLANT CELL, August 1, 2006; 18(8): 1833 - 1845. [Abstract] [Full Text] [PDF]

				Y. ZHAO, G. WANG, J. ZHANG, J. YANG, S. PENG, L. GAO, C. LI, J. HU, D. LI, and L. GAO Expressed Sequence Tags (ESTs) and Phylogenetic Analysis of Floral Genes from a Paleoherb Species, Asarum caudigerum Ann. Bot., July 1, 2006; 98(1): 157 - 163. [Abstract] [Full Text] [PDF]

				C. P. Scutt, M. Vinauger-Douard, C. Fourquin, C. Finet, and C. Dumas An evolutionary perspective on the regulation of carpel development J. Exp. Bot., July 1, 2006; 57(10): 2143 - 2152. [Abstract] [Full Text] [PDF]

				D. G. Howarth and M. J. Donoghue Phylogenetic analysis of the "ECE" (CYC/TB1) clade reveals duplications predating the core eudicots PNAS, June 13, 2006; 103(24): 9101 - 9106. [Abstract] [Full Text] [PDF]

				S. Ciannamea, J. Busscher-Lange, S. de Folter, G. C. Angenent, and R. G. H. Immink Characterization of the Vernalization Response in Lolium perenne by a cDNA Microarray Approach Plant Cell Physiol., April 1, 2006; 47(4): 481 - 492. [Abstract] [Full Text] [PDF]

				J. Leebens-Mack, L. A. Raubeson, L. Cui, J. V. Kuehl, M. H. Fourcade, T. W. Chumley, J. L. Boore, R. K. Jansen, and C. W. dePamphilis Identifying the Basal Angiosperm Node in Chloroplast Genome Phylogenies: Sampling One's Way Out of the Felsenstein Zone Mol. Biol. Evol., October 1, 2005; 22(10): 1948 - 1963. [Abstract] [Full Text] [PDF]

				A. Berbel, C. Navarro, C. Ferrandiz, L. A. Canas, J.-P. Beltran, and F. Madueno Functional Conservation of PISTILLATA Activity in a Pea Homolog Lacking the PI Motif Plant Physiology, September 1, 2005; 139(1): 174 - 185. [Abstract] [Full Text] [PDF]

				L. M. Zahn, J. Leebens-Mack, C. W. dePamphilis, H. Ma, and G. Theissen To B or Not to B a Flower: The Role of DEFICIENS and GLOBOSA Orthologs in the Evolution of the Angiosperms J. Hered., May 1, 2005; 96(3): 225 - 240. [Abstract] [Full Text] [PDF]

				A. Maizel, M. A. Busch, T. Tanahashi, J. Perkovic, M. Kato, M. Hasebe, and D. Weigel The Floral Regulator LEAFY Evolves by Substitutions in the DNA Binding Domain Science, April 8, 2005; 308(5719): 260 - 263. [Abstract] [Full Text] [PDF]

				L. M. Zahn, H. Kong, J. H. Leebens-Mack, S. Kim, P. S. Soltis, L. L. Landherr, D. E. Soltis, C. W. dePamphilis, and H. Ma The Evolution of the SEPALLATA Subfamily of MADS-Box Genes: A Preangiosperm Origin With Multiple Duplications Throughout Angiosperm History Genetics, April 1, 2005; 169(4): 2209 - 2223. [Abstract] [Full Text] [PDF]

				V. Hecht, F. Foucher, C. Ferrandiz, R. Macknight, C. Navarro, J. Morin, M. E. Vardy, N. Ellis, J. P. Beltran, C. Rameau, et al. Conservation of Arabidopsis Flowering Genes in Model Legumes Plant Physiology, April 1, 2005; 137(4): 1420 - 1434. [Abstract] [Full Text] [PDF]

				R. C. Moore, S. R. Grant, and M. D. Purugganan Molecular Population Genetics of Redundant Floral-Regulatory Genes in Arabidopsis thaliana Mol. Biol. Evol., January 1, 2005; 22(1): 91 - 103. [Abstract] [Full Text] [PDF]

				E. M. Eriksson, A. Bovy, K. Manning, L. Harrison, J. Andrews, J. De Silva, G. A. Tucker, and G. B. Seymour Effect of the Colorless non-ripening Mutation on Cell Wall Biochemistry and Gene Expression during Tomato Fruit Development and Ripening Plant Physiology, December 1, 2004; 136(4): 4184 - 4197. [Abstract] [Full Text] [PDF]

				P. S. Soltis and D. E. Soltis The origin and diversification of angiosperms Am. J. Botany, October 1, 2004; 91(10): 1614 - 1626. [Abstract] [Full Text] [PDF]

				W. S. Judd and R. G. Olmstead A survey of tricolpate (eudicot) phylogenetic relationships Am. J. Botany, September 1, 2004; 91(10): 1627 - 1644. [Abstract] [Full Text] [PDF]

				F. Fornara, L. Parenicova, G. Falasca, N. Pelucchi, S. Masiero, S. Ciannamea, Z. Lopez-Dee, M. M. Altamura, L. Colombo, and M. M. Kater Functional Characterization of OsMADS18, a Member of the AP1/SQUA Subfamily of MADS Box Genes Plant Physiology, August 1, 2004; 135(4): 2207 - 2219. [Abstract] [Full Text] [PDF]

				M. Calonje, P. Cubas, J. M. Martinez-Zapater, and M. J. Carmona Floral Meristem Identity Genes Are Expressed during Tendril Development in Grapevine Plant Physiology, July 1, 2004; 135(3): 1491 - 1501. [Abstract] [Full Text] [PDF]

				V. F. Irish and P. N. Benfey Beyond Arabidopsis. Translational Biology Meets Evolutionary Developmental Biology Plant Physiology, June 1, 2004; 135(2): 611 - 614. [Abstract] [Full Text] [PDF]

				A. V. Shchennikova, O. A. Shulga, R. Immink, K. G. Skryabin, and G. C. Angenent Identification and Characterization of Four Chrysanthemum MADS-Box Genes, Belonging to the APETALA1/FRUITFULL and SEPALLATA3 Subfamilies Plant Physiology, April 1, 2004; 134(4): 1632 - 1641. [Abstract] [Full Text] [PDF]

				E. M. Kramer, M. A. Jaramillo, and V. S. Di Stilio Patterns of Gene Duplication and Functional Evolution During the Diversification of the AGAMOUS Subfamily of MADS Box Genes in Angiosperms Genetics, February 1, 2004; 166(2): 1011 - 1023. [Abstract] [Full Text] [PDF]