JXB Advance Access originally published online on August 7, 2006
Journal of Experimental Botany 2006 57(12):3099-3107; doi:10.1093/jxb/erl081
RESEARCH PAPER |
Unique and redundant functional domains of APETALA1 and CAULIFLOWER, two recently duplicated Arabidopsis thaliana floral MADS-box genes
Laboratorio de Genética Molecular, Desarrollo y Evolución de Plantas, Departamento de Ecología Funcional, Instituto de Ecología, UNAM, Ap. Postal 70–275, México DF 04510, México
*To whom correspondence should be addressed. E-mail: elena.alvarezbuylla{at}gmail.com
Received 6 April 2006; Accepted 5 June 2006
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
APETALA1 (AP1) and CAULIFLOWER (CAL) are closely related MADS box genes that are partially redundant during Arabidopsis thaliana floral meristem determination. AP1 is able to fully substitute for CAL functions, but not vice versa, and AP1 has unique sepal and petal identity specification functions. In this study, the unique and redundant functions of these two genes has been mapped to the four protein domains that characterize type-II MADS-domain proteins by expressing all 15 chimeric combinations of AP1 and CAL cDNA regions under control of the AP1 promoter in ap1-1 loss-of-function plants. The ‘in vivo’ function of these chimeric genes was analysed in Arabidopsis plants by expressing the chimeras. Rescue of flower meristem and sepal/petal identities was scored in single and multiple insert homozygous transgenic lines. Using these chimeric lines, it was found that distinct residues of the AP1 K domain not shared by the same CAL domain are necessary and sufficient for complete recovery of floral meristem identity, in the context of the CAL protein sequence, while both AP1 COOH and K domains are indispensable for complete rescue of sepal identity. By contrast, either one of these two AP1 domains is necessary and sufficient for complete petal identity recovery. It was also found that there were positive and negative synergies among protein domains and their combinations, and that multiple-insert lines showed relatively better rescue than equivalent single-insert lines. Finally, several lines had flowers with extra sepals and petals suggesting that chimeric proteins yield abnormal transcriptional complexes that may alter the expression or regulation of genes that control floral organ number under normal conditions.
Key words: AP1, CAL, MADS-box genes, protein domains, redundancy
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Gene duplication is an important source of evolutionary novelties (Ohno, 1970). After duplication, (i) one paralogue may become a pseudogene, (ii) both paralogues may diverge functionally (neofunctionalization), (iii) both copies may keep subsets of the original functions and together maintain the total capacity of the ancestral gene (subfuctionalization), or (iv) both may keep their original complete functions thus making the organism more robust to mutations in either orthologue (Force et al., 1999; Gu, 2003; Kramer et al., 2004; Causier et al., 2005). Mutations in the regulatory and coding regions may underlie the evolutionary scenarios of the first three cases. This paper focuses on AP1 and CAL, two closely related genes (Alvarez-Buylla et al., 2000) that originated from a recent duplication event <60 million years ago (Lawton-Rauh et al., 1999), with very similar sequences and expression patterns and that have partially redundant functions during flower development in Arabidopsis thaliana (herein Arabidopsis) (Mandel et al., 1992; Gustafson-Brown et al., 1994; Kempin et al., 1995).
When a plant becomes florally induced, the apical meristem switches from a vegetative to an inflorescence meristem. While the former only produces leaves as lateral organs, the inflorescence meristem produces flowers that arise from its flanks in a spiral arrangement. The four sepal primordia are the first to arise from the flower meristem; subsequently, four petal and six stamen primordia are initiated almost simultaneously, and the remaining floral meristem interior to the three other whorls comprises the gynoecial primordium (Bowman, 1993).
At least four genes are necessary for the specification of floral meristem identity in Arabidopsis: LEAFY (LFY), CAULIFLOWER (CAL), APETALA1 (AP1), and FRUITFULL (FUL) (Mandel et al., 1992; Weigel et al., 1992; Kempin et al., 1995, Ferrandiz et al., 2000a). The latter three are MADS-box genes and they are functionally redundant in specifying flower meristem identity (Ferrandiz et al., 2000a). These are typical type-II MADS box genes (Alvarez-Buylla et al., 2000) that code for proteins with a conserved DNA-binding domain, a more divergent intervening I region, a conserved K domain which may participate in protein interaction, and a divergent COOH (see below) (Krizek and Meyerowitz, 1996a; Riechmann et al., 1996; Fan et al., 1997; Kaufmann et al., 2005).
The 60 aa MADS-domain is at the N-terminus and it has been shown that this motif is involved in specific binding to DNA sequences (CArG boxes) conforming the consensus sequence CC(A/T)6GG in both animals and plants (Pellegrini et al., 1995; Riechmann et al., 1996). In type-II MADS-domain proteins, the formation of dimers that are capable of DNA binding requires the I-region (aa 60–86); this domain is a key determinant for the specificity of DNA-binding dimer formation (Riechmann et al., 1996).
A second conserved domain, denominated K (87–150 aa), is postulated to form three 
-helices referred to as K1, K2, and K3 that potentially form coiled-coils, with K1 and K2 helices located entirely within the K domain, while K3 helix spans the K domain–C domain boundary (Yang and Jack, 2004). The K-domain is assumed to generate a surface important for protein–protein interaction (Davies et al., 1996; Fan et al., 1997). Finally, the C-terminal domain is a length-variable amino-acid stretch that may have several functions. For example, it is thought to participate in higher-order MADS interactions (Honma and Goto, 2001), it is required for functional specificity (Lamb and Irish, 2003), it may be involved in transcriptional activation (Huang et al., 1995), and can enhance/stabilize interactions that are mediated by the K-domain (Fan et al., 1997; Pelaz et al., 2001).
AP1 and CAL MADS and K regions differ only in a few amino acids and most are conservative; in their K domain only six non-conservative differences are found (Fig. 1A). Loss-of-function alleles for AP1 (ap1-1) and CAL (cal-5) have been characterized (Mandel et al., 1992; Kempin et al., 1995). In ap1-1, flower meristems are partially converted into inflorescence meristems, and the presence, position, and identity of the outer two whorls of floral organs (sepals and petals) are altered. By contrast, cal-5 single mutant flowers are indistinguishable from wild-type flowers. In an ap1-1 mutant background, however, cal-5 enhances the ap1-1 mutant phenotype producing inflorescences within inflorescences to yield a structure with a massive proliferation of meristems that closely resembles the commercial vegetable cauliflower (Kempin et al., 1995).
|
The above results show that AP1 and CAL have partially redundant functions. AP1 can fully substitute for CAL functions and CAL seems to be redundant with AP1 in specifying flower meristems, but CAL is unable to substitute for AP1 in specifying sepal and petal identity or fully rescuing floral meristem identity. Functional redundancy among developmentally important genes is common in plant and non-plant systems (Smyth, 2000; Avendano et al., 2005), but in very few cases has the molecular basis of this redundancy been mapped to different regions of the duplicate genes (Krizek and Meyerowitz, 1996a, b; Riechmann et al., 1996; Lamb and Irish, 2003). Over-expression lines of AP1 and CAL demonstrated that the differences in the phenotype are the result of changes in the coding region of these two genes and not in their expression patterns (Savidge, 1996; Liljegren et al., 1999). This suggestion was also supported in the molecular evolution analysis of Lawton-Rauh et al. (1999) who documented distinct rates of non-synonymous sites in these two genes and documented differences in evolution between protein domains. These studies suggest that important aspects of the unique and shared functions of these two MADS-domain proteins depend on differences between their functional protein domains.
Redundancy seems to be common among MADS-box genes (see review in Smyth, 2000), especially in closely related ones that originated from a recent duplication. This is the case of SHATTERPROOF1 and 2 (SHP1, SHP2) and STK (SEEDSTICK) in the AGAMOUS clade (Ferrandiz et al., 2000b; Pinyopich et al., 2003), and of the three SEPALLATA genes and FUL, CAL, and AP1 in the APETALA1 clade (Ferrandiz et al., 2000a; Pelaz et al., 2001).
Type-II MADS-domain transcription factors are postulated to function in multimers of at least four proteins of the same family (Honma and Goto, 2001; Pelaz et al., 2001) whose composition determines target specificity and in which some members are interchangeable and others are not (Honma and Goto, 2001). Shortly after duplication, the two products should interact with the same proteins, but eventually some of them may be lost and others gained by either protein. The number of common interacting partners may be taken as a measure of functional overlap and this will depend on the evolution of domains critical for determining protein–protein interaction specificity. If the number of shared proteins is the basis for the functional overlap in the case of AP1 and CAL, it would be expected that the putative protein–protein interaction domain would be critical in determining the shared and unique functions of these two genes.
Here the different roles of AP1 and CAL during flower meristem and floral organ development are mapped to the different regions of these two genes, by assaying the recovery potential of chimeric cDNA constructs for all possible combinations of the four main regions of these two genes when expressed under the AP1 native promoter in ap1-1 loss-of-function mutant plants. With this specific promoter, the two functions of AP1 that are determined by different levels of AP1 activity can be assayed; a high amount of AP1 activity is required for organ identity function, and a lower concentration of this protein is necessary for floral meristem identity function (Krizek et al., 1999).
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Construction of the AP1/CAL chimeric proteins
APETALA1 and CAULIFLOWER domains as defined in this study are indicated in Fig. 1A. All possible combinations of the four regions characteristic of AP1 and CAL (Fig. 1B) were constructed using a PCR-based approach. In first-round independent PCR reactions the four regions of each gene were amplified. For these first-round PCRs, 5' sense primers homologous to CAL or AP1 with a ‘tail’ of an additional 21 base pairs homologous to AP1 or CAL, respectively, and 3' antisense primers homologous to either AP1 or CAL sequences were used. The sequences of primers used are as follows (with sequences homologous to CAL cDNA underlined).
OEAB1: 5'-GTATTCGAAGAGTTTCCCCTT-3' (antisense MADS AP1)
OEAB2: 5'-CGACCAGTTTGTATTGACGTC-3' (antisense I AP1)
OEAB3: 5'-CTCCTTGATCTGTTTAGAAAG-3' (antisense K AP1)
OEAB4: 5'-AAGGGCAAATTGTTCGAGTACTCCACTGATTCTTGTATGGAG-3' (sense MADS)
OEAB5: 5'-AATGCACAGACGAACTGGTCAATGGAGTATAACAGGCTTAAG-3' (sense I)
OEAB6: 5'-CTTACCAAACAGATAAAGGAGAGGGAAAAAATTCTTAGGGCT-3' (sense K)
OEAB7: 5'-GTACTCGAACAATTTGCCCTT-3' (antisense MADS CAL)
OEAB8: 5'-TGACCAGTTCGTCTGTGCATT-3' (antisense I CAL)
OEAB9: 5'-CTCCTTTATCTGTTTGGTAAG-3' (antisense K CAL)
OEAB10: 5'-AAGGGGAAACTCTTCGAATACTCCTCTGAATCTTGCATGGAG-3' (sense MADS)
OEAB11: 5'-GACGTCAATACAAACTGGTCGATGGAGTATAGCAGGCTTAAG-3' (sense I)
OEAB12: 5'-CTTTCTAAACAGATCAAGGAGAGGGAAAACATCCTAAAGACA-3' (sense K)
Additionally, the sequences of the AP1 forward and reverse external primers used are, respectively:
AP15': 5'-ATGGGAAGGGGTAGG
S2AP1: 5'-AACCTTGGCTGCTTCGCCGCATGA
The sequences of the CAL 5' and 3' external primers used are, respectively:
OEAB13: 5'-AGAAATGGGAAGGGGTAGGGTTGAA-3'
OEAB14: 5'-TACCTTGGCTGTTACGCCGCTTGA-3'
All external primers had BamH1 restriction sites at borders to be used as cloning sites. Plasmid DNAs from pAM571 and pBS85 were used as templates for AP1 and CAL cDNAs, respectively, and were obtained from the laboratory of Marty Yanofsky. Once the four regions (MADS, I, K, and COOH) of each gene were amplified carrying a tail at their 5' end homologous to the sequence of the upstream corresponding region of the other gene, the appropriate combination of these fragments for each chimera (Fig. 1B) were used as templates for a second-round PCR. The PCR was made using the 5' and 3' primers of AP1 or CAL depending if the AP1 or CAL MADS or COOH was the extreme of each construct. The products of the second-round PCRs were cloned into BamH1 sites of Promega PGM-Teasy or PGM, or into Invitrogen's TA-cloning kits according to the supplier's instructions. The cloned AP1/CAL chimeras were sequenced and clones with identical sequences to the corresponding AP1 and CAL reported sequences were kept for the next cloning step.
Cloning into a binary plasmid with the wild-type AP1 promoter
Chimeric fragments were released from the PGM-Teasy or PGM plasmids with BamH1 and cloned into the BamH1 site of the binary plasmid pAM571 downstream of the AP1 promoter sequence. The orientation of the insert was confirmed by PCR.
Plant transformation, T3 selection for single-insert lines, and scoring of the phenotypes observed
Agrobacterium tumefaciens (MP5-1) was transformed with the corroborated binary plasmids carrying each one of the constructs. Arabidopsis thaliana (Columbia ecotype) loss-of-function ap1-1 line (Mandel et al., 1992) was transformed with A. tumefaciens carrying each chimeric plasmid by the dipping method (Clough and Bent, 1998). Transformants were selected in kanamycin plates and at least 30 independent lines per construct were transplanted to soil and grown to maturity. Their seeds were recovered for future characterizations. The segregating T2 generation of each independent T1 line was planted in soil and observed. For each construct, lines with similar rescue phenotypes, among which were those with the strongest recovery phenotypes, were followed to the next generation. T2 seeds of selected lines were plated in kanamycin and three lines with a 3:1 segregation ratio were selected out of at least 10 with similar and strong recovery phenotypes. Some lines with 3:1 segregation could have inserts in tandem; however, by choosing from at least 10 lines with similar phenotypes, the probability of selecting lines with two inserts in tandem was negligible. Furthermore, several lines were found in each construct that stood out for having a clearly stronger recovery phenotype in comparison to the rest. At least four lines per construct with these characteristics were selected and the number of inserts established by segregation analyses, confirming that they had more than one construct in all cases.
For each construct, three single-insert lines with the strong rescue phenotype were kept, and two of these lines were selected for quantitative analyses, scanning electron microscopy (SEM), and dissecting light microscope imaging. For each construct a few lines with segregation scores that indicated two, three, or more inserts were also kept for comparison with single-insert lines.
For two individuals of each one of the two selected lines per construct the number of flowers per pedicel was quantified in flowers 1–5 and 11–15, counting from the shoot base towards the apex of the plant. For each one of these flowers the number of sepals and petals was scored, as well as the number of other types of organs in each one of the two first whorls. Average and standard errors for these numbers are reported.
Scanning electron microscopy
Before scoring the plants, they were observed in a dissecting light microscope and photographed. In preparation for SEM, samples were treated as in Ferrandiz et al. (2000a) and analysed on a Cambridge 360 SEM (Cambridge Instruments, Cambridge, UK) at 10 kV accelerating voltage. The light and SEM photographs were processed in Adobe Photoshop V5, only modifying their contrast level.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Results are shown for lines for which segregation analyses suggest that they are single copy and homozygous for the transgene. In all cases the background phenotype is ap1-1. As expected, the positive control line that expressed the complete AP1 cDNA under its own promoter (line 1; see Fig. 1B) rescued the ap1-1 phenotype to wild-type condition in both floral meristem and flower organ identity (Figs 2–4). By contrast, the negative control that also expressed the full-length cDNA of CAL under the AP1 promoter (line 16; see Fig. 1B) did not rescue either the floral meristem or floral organ identity defects of ap1-1 mutants (Figs 2, 3), and this line is indistinguishable from an ap1-1 plant. Also as expected, ap1-1 cal5 double mutants recovered the ap1-1 single mutant phenotype when the full-length CAL cDNA was expressed under the AP1 promoter (data not shown; however, for the ap1-1 phenotype see Fig. 2B), and these plants recovered the wild-type phenotype when transformed with the AP1 full-length cDNA under this gene's native promoter (Fig. 2A, line 1).
|
|
|
Of the lines expressing chimeric AP1-CAL cDNAs, one group that are called strong lines showed complete rescue of the floral meristem identity phenotype (lines 2–6) and did not have secondary flowers; in all cases there was only one flower per pedicel (Figs 2, 3). A second group, herein called intermediate lines, had incomplete but significant levels of floral meristem identity rescue (lines 7–12) (Figs 2, 3). Finally, a third group, referred to as weak lines, possessed very little or no rescue of floral meristem identity and closely resembled line 16 or ap1-1 plants (lines 13–15) (Figs 2, 3). Interestingly, the lines of the first group all shared the K domain of AP1. In fact, line 4, which had only the AP1 K domain and all other CAL domains, showed complete floral meristem identity rescue with no secondary flowers.
The results summarized above clearly show that the differences between the AP1 and CAL proteins in the K domain (Fig. 1A) are critical for the distinct functions of these two genes during floral meristem determination. This conclusion is reinforced by the fact that the line expressing construct 9, that had all AP1 regions except the K domain, does not show complete rescue to wild-type flower meristem identity function (Fig. 2B).
However, among the lines with intermediate degrees of floral meristem identity rescue, there are two lines, 10 and 12, that also harboured the AP1 K region and still did not show complete rescue of flower meristem identity function (Fig. 2A). This suggests that the effect of one domain on the phenotype is not independent of the effect of other domains or of their combinations. The only thing that these two lines have in common is that their constructs combine MADS and I regions from AP1 and CAL, thus suggesting that, in these lines, the MADS–I combination from the two genes caused a negative synergy in floral meristem rescue, even in the presence of the AP1 K domain. It is also noteworthy that lines with constructs that lack both the AP1 K and COOH regions (lines 13–15) showed no significant rescue of floral meristem identity with respect to the ap1-1 phenotype, suggesting that AP1 COOH-specific residues are also important for floral meristem identity function.
Flower numbers per pedicel were also counted for flowers 11–15 and these showed a similar tendency to those observed for flowers 1–5 but with proportionately fewer secondary flowers (flowers per pedicel) in comparison to the early-rising flowers (data not shown). In fact, weak lines 13–15 had a similar mean number of flowers per pedicel as intermediate lines. This suggests that the AP1-specific MADS and I residues have a positive synergistic effect in flower meristem identity rescue that is evident when the rest of the protein domains (K and COOH) come from the CAL gene. Hence, only the lines expressing the full-length CAL protein had a mean number of secondary flowers that did not differ significantly to that observed in the ap1-1 lines.
Interestingly, line 4, with only the K domain of the AP1 protein, showed complete rescue in floral meristem identity but did not show a complete rescue of floral organ identity specification (Figs 3A, 4). In this line, sepals were either absent or had leaf-like traits such as ramified trichomes with cells typical of leaves rather than sepals, although petals have almost a wild-type phenotype. All other strong lines showed good rescue of petal identity, but line 2 that lacked the AP1 COOH domain did not show total rescue of sepal identity (Fig. 3A). Intermediate lines that had either the AP1 K or COOH domains showed partial rescue of sepal identity, suggesting that both AP1 K and COOH protein domains are required for normal sepal identity function (Fig. 3A). By contrast, petals could be rescued to almost normal identity and number when either the AP1 COOH (lines 7 and 9) or only the API K (line 4) domains were present (Figs 3B, 4). As with floral meristem identity function the combination of AP1 and CAL MADS or I domains caused negative synergies in the rescue of petal identity (see lines 10 and 12 in Fig. 3B).
The weak lines (13–15), where neither the AP1 K nor the COOH domains were present, showed much less or no rescue at all of floral organ identity rescue. But line 13 that had both AP1 MADS and I domains with the CAL K and COOH domains showed a slight degree of sepal and petal identity rescue (Fig. 3). This suggests again that interactions among domains are important in determining levels of rescue. It can nonetheless be concluded that both the AP1 K and COOH domains are critical for complete rescue of sepal identity, but the presence of either one of these two domains in the context of a CAL protein are sufficient to rescue petal identity of ap1-1 mutant plants.
Some additional remarkable observations are summarized here. Although single insert lines suggested that specific AP1 domains are critical for its specific floral meristem and organ identity functions with respect to CAL function, it was observed that at least four lines per construct type, which had multiple inserts, showed a greater degree of rescue to that observed in the corresponding single insert lines, and the degree of rescue seemed to be correlated to the number of inserts (data not shown but available upon request). This supports the idea that the copy number and the expression are positively correlated as has been suggested before, at least below a certain transgene number (Schubert et al., 2004). Also, it is noteworthy that some transgenic lines (lines 3 and 7) produced up to six sepals and/or petals, suggesting aberrant interactions with proteins or abnormal regulation of target genes that are critical for the determination of the correct number of floral organs. Finally, only lines 1 and 3 rescue to wild-type condition the morphology of trichomes in sepals, lines 2 and 5–11 have both simple and complex trichomes, and lines 4 and 12–16 only have complex trichomes, suggesting that they have a leaf-like morphology.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Understanding the molecular basis of the redundant and unique functions of two recently duplicated genes is fundamental for interpreting the evolutionary fate of duplicate developmental genes in multicellular organisms. The role of changes in the spatio-temporal regulation of key transcription factors has been stressed as an important molecular basis of morphological evolution (Doebley and Lukens, 1998). However, it is shown here that cDNA differences between two recently duplicated genes underlie their contrasting roles in flower development. Transcription factors that are encoded by recently duplicated genes may share target genes that will depend on DNA binding or transactivation functions that in turn sometimes depend on the composition of the multimeric complexes that may be formed by the transcription factors. This indeed seems to be the case of the MADS-domain factors (Honma and Goto, 2001).
Although chimeric studies had been performed before (Krizek and Meyerowitz, 1996a, b; Honma and Goto, 2001), most had used the 35S constitutive promoter rather than a native promoter (although see exception in Krizek et al., 1999) and none had pursued a mapping analysis such as the one presented here for two closely related genes, in which all possible combinations of functional domains identified were assayed. The present approach using cDNA chimeras is a first step towards dissecting the molecular basis of the unique and redundant functions of recently duplicated genes. It also provides a starting point for the study of sub/neofunctionalization (Kramer et al., 2004; Causier et al., 2005) at the molecular level following gene duplication in the important MADS-box gene family. It was found that the putative protein–protein interaction K domain of AP1 is necessary, and largely sufficient in the context of the CAL protein, to recover flower meristem identity of ap1-1 mutant plants. By contrast, sepal identity can be fully rescued only if both the AP1 K and COOH domains are present, while petal identity is recovered to wild-type if either of these two AP1 domains is present in the context of an otherwise CAL protein sequence. Thus this study also helps to establish which domains are equivalent and/or different between AP1 and CAL proteins. It seems, for example, that in lines 3 and 5 the MADS and I domains of these two proteins are interchangeable in terms of their ability to rescue either floral meristem or floral organ identity. However, transgenic lines expressing chimeras that had the MADS, I or both of the AP1 cDNA in the context of the CAL cDNA was not able to rescue either identity function.
A recent two hybrid-based protein–protein interaction map for most MADS-domain proteins suggests that AP1 and CAL share only five member proteins of this family: SOC1, AGL24, SVP, SEP3, and AGL74N (Pelaz et al., 2001; de Folter et al., 2005) that are the only interacting proteins found for CAL, while FUL, which is the sister MADS-domain protein to AP1 and CAL, also shares four of these partners (AGL24, SOC1, SEP3, and AGL74N) and, in addition, shares another four with AP1 (SEP1, SEP4.2, AGL6, and AGL21). In addition to these eight proteins, AP1 has gained interactions with another six MADS-domain proteins (AGL15, AGL16, AGL39, AGL86, AGL92, and AGL97) that have not been functionally characterized, and lacks two with which FUL interacts (AG and AGL14).
Structural studies have suggested that regions of the MADS domain in animal type-I (human serum response factor) and type-II proteins (myocyte enhancer factor) are important for protein–protein interactions. No MADS-domain type-II plant protein has been crystallized, but in these proteins the fairly conserved K domain has been identified as important for protein–protein interactions (Davies et al., 1996; Fan et al., 1997; Pelaz et al., 2001). The results presented here suggest that the differences between AP1 and CAL K domains are critical for the unique and indispensable roles of AP1 during floral organ and meristem fate determination. Furthermore, it seems that it is the interaction of AP1 but not CAL with floral organ fate determination proteins that determines this specific AP1 function (SEP1, SEP4.2, and AGL6).
Conservation within the K domain is significantly higher among proteins capable of interacting with the same partners than among proteins which interact with other proteins (Davies et al., 1996). The K domain (87–150 aa; Fig. 1A) is postulated to form three -helices referred to as K1, K2, and K3 (Yang and Jack, 2004) that potentially form coiled-coils. Comparing the AP1 and CAL proteins and the putative -helices coiled-coil they formed, only two amino acids were found in the K3 region of the CAL protein that could change and disturb the helix formed within the K3 region of the AP1 protein (data not shown). Thus it may be hypothesized that these changes disrupt and prevent the interaction among proteins because it has been reported that the central and last part of the K-domain (K2–K3) is necessary for interactions between MICK-type proteins in Arabidopsis (Yang and Jack, 2004; Kaufmann et al., 2005). Future studies may use site-directed mutagenesis and a two-hybrid system or in vivo assays to test this and related hypotheses.
Since the K domain is critical for establishing specific protein–protein interactions, it could be that the amino acid differences between AP1 and CAL are critical for the privative protein interactions of AP1, while the shared residues are sufficient for interacting with the common partners. Protein–protein interactions could then be the basis for the enhanced flower meristem identity phenotype of the ap1-1 cal-5 phenotype in comparison to the single ap1-1 phenotype, and for the fact that cal-5 mutants are indistinguishable from wild type. In the case when only AP1 is missing, CAL is still able to interact with the shared partners and only the privative AP1 partners are left out, but when both AP1 and CAL are absent all the protein interactions that are shared and privative of these two genes will be left out, thus yielding an enhanced phenotype.
On the other hand, sepal fate also depends on specific AP1 COOH residues, while petal fate may be recovered with either the AP1 K or COOH domains alone in the context of the CAL protein sequence, suggesting that the critical interacting proteins needed to rescue petal identity may also rely on the COOH and not only on the K domain. Indeed, it has been shown that, although the K domain alone is sufficient for interactions between SEP3, SOC1, SVP, and AGL24 with CAL and AP1 baits, these interactions are stronger when approximately half the C-terminal domain is added to the K domain (Pelaz et al., 2001). The C-terminal domain apparently contains sites required for interaction between proteins or in the formation of multimers (Honma and Goto, 2001).
Chimeras that rescue different aspects of the ap1-1 mutant could now be used in two-hybrid assays to discern which protein interaction partners are critical for the floral meristem and flower organ identity functions of this gene. This could also be the basis to discern if specific AP1 partners are responsible for the distinctive role of this protein in floral organ specification that is not shared with CAL. The present results and the available protein interaction data (Pelaz et al., 2001; de Folter et al., 2005) suggest that the specific residues of the AP1 K domain are needed for the interaction with other complexes that are key during flower meristem determination and that the I, K, and COOH domains participate mainly in floral organ identity.
It is noteworthy that both positive and negative synergistic effects were observed among the different protein domains in rescuing flower meristem identity and organ identity functions of ap1-1 mutants. For example, the incomplete rescue of lines 10 and 12 may be due to the fact that they both combined the MADS and I regions of AP1 and CAL. This combining of domains may cause protein misfolding that may yield spurious protein–protein interactions. A positive synergistic interaction was observed in those cases in which sepal and petal rescue was improved in lines harbouring chimeras with both the AP1 MADS and I domains.
Constructs could have been transformed into ap1-1 cal-5 double mutants to map the molecular basis of redundancy between AP1 and CAL. However, since CAL and AP1 full-length cDNAs under the AP1 promoter rescue the double mutant to ap1-1 and wild-type phenotype, respectively, and cal-5 is indistinguishable from wild-type, such analyses would not provide additional information to that obtained with the present mapping analysis done in ap1-1 mutants.
Finally, the present results suggest that protein–protein interactions may be dosage dependent because rescue was better in lines with multiple inserts in comparison to those with single inserts, suggesting that increased levels of cDNA and protein may override protein–protein or DNA–protein interaction barriers when aberrant residues are present. This has been suggested before and it has been hypothesized that the four whorls may have differential sensitivity to protein activity (Zachgo et al., 1995; Yang et al., 2003).
|
Acknowledgements |
|---|
Important parts of this study were done in M Yanofsky's laboratory and his advice and support are greatly acknowledged. Special thanks are due to Moses Salgado and Nathan Chu for tremendous help in endless laboratory tasks, to Alejandra Mandel for the AP1 promoter construct, and to the other members of the Yanofsky laboratory, especially Cristina Ferrandiz, Soraya Pelaz, Sherry Kempin, Gary Ditta, Sarah Liljegren, and Medard Ng, for much help and advice. Discussions with Francisco Vergara-Silva were helpful while designing the study and analysing its results. Technical collaboration by Rigoberto Pérez and the assistance of Aida Navarrete in several tasks are acknowledged. This work was done with support from the PEW Foundation and Grants from PAPIIT (Programa de Apoyo a Proyectos de Investigación e Innovación Tecnológica), UNAM (Universidad Nacional Autónoma de México; IN 230002 and IN212995), and CONACYT (Consejo Nacional de Ciencia y Tecnología; 41848-Q and 31871-N) to ERA-B, and we also acknowledge support from the Santa Fe Institute (New Mexico, USA) International Fellowship awarded to ERA-B which helped complete this work.
|
Footnotes |
|---|
These authors contributed equally to this work. 
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Alvarez-Buylla ER, Pelaz S, Liljegren SJ, Gold SE, Burgeff C, Ditta GS, Ribas de Pouplana L, Martinez-Castilla L, Yanofsky MF. (2000) An ancestral MADS-box gene duplication occurred before the divergence of plants and animals. Proceedings of the National Academy of Sciences, USA 97:5328–5333.
Avendano A, Riego L, DeLuna A, et al. (2005) Swi/SNF-GCN5-dependent chromatin remodelling determines induced expression of GDH3, one of the paralogous genes responsible for ammonium assimilation and glutamate biosynthesis in Saccharomyces cerevisiae. Molecular Microbiology 57:291–305.[CrossRef][ISI][Medline]
In Bowman J (Ed.). Arabidopsis: an atlas of morphology and development (1993) (Springer-Verlag, Berlin).
Causier B, Castillo R, Zhou J, Ingram R, Xue Y, Schwarz-Sommer Z, Davies B. (2005) Evolution in action: following function in duplicated floral homeotic genes. Current Biology 15:1508–1512.[CrossRef][ISI][Medline]
Clough SJ and Bent AF. (1998) Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. The Plant Journal 16:735–743.[CrossRef][ISI][Medline]
Davies B, Egea-Cortines M, de Andrade Silva E, Saedler H, Sommer H. (1996) Multiple interactions amongst floral homeotic MADS box proteins. EMBO Journal 15:4330–4343.[ISI][Medline]
de Folter S, Immink RG, Kieffer M, et al. (2005) Comprehensive interaction map of the Arabidopsis MADS box transcription factors. The Plant Cell 17:1424–1433.
Doebley J and Lukens L. (1998) Transcriptional regulators and the evolution of plant form. The Plant Cell 10:1075–1082.
Fan HY, Hu Y, Tudor M, Ma H. (1997) Specific interactions between the K domains of AG and AGLs, members of the MADS domain family of DNA binding proteins. The Plant Journal 12:999–1010.[CrossRef][ISI][Medline]
Ferrandiz C, Gu Q, Martienssen R, Yanofsky MF. (2000a) Redundant regulation of meristem identity and plant architecture by FRUITFULL, APETALA1 and CAULIFLOWER. Development 127:725–734.[Abstract]
Ferrandiz C, Liljegren SJ, Yanofsky MF. (2000b) Negative regulation of the SHATTERPROOF genes by FRUITFULL during Arabidopsis fruit development. Science 289:436–438.
Force A, Lynch M, Pickett FB, Amores A, Yan YL, Postlethwait J. (1999) Preservation of duplicate genes by complementary, degenerative mutations. Genetics 151:1531–1545.
Gu X. (2003) Evolution of duplicate genes versus genetic robustness against null mutations. Trends in Genetics 19:354–356.[CrossRef][ISI][Medline]
Gustafson-Brown C, Savidge B, Yanofsky MF. (1994) Regulation of the arabidopsis floral homeotic gene APETALA1. Cell 76:131–143.[CrossRef][ISI][Medline]
Honma T and Goto K. (2001) Complexes of MADS-box proteins are sufficient to convert leaves into floral organs. Nature 409:525–529.[CrossRef][Medline]
Huang H, Tudor M, Weiss CA, Hu Y, Ma H. (1995) The Arabidopsis MADS-box gene AGL3 is widely expressed and encodes a sequence-specific DNA-binding protein. Plant Molecular Biology 28:549–567.[CrossRef][ISI][Medline]
Kaufmann K, Melzer R, Theissen G. (2005) MIKC-type MADS-domain proteins: structural modularity, protein interactions and network evolution in land plants. Gene 347:183–198 Review.[CrossRef][ISI][Medline]
Kempin SA, Savidge B, Yanofsky MF. (1995) Molecular basis of the cauliflower phenotype in Arabidopsis. Science 267:522–525.
Kramer EM, Jaramillo MA, Di Stilio VS. (2004) Patterns of gene duplication and functional evolution during the diversification of the AGAMOUS subfamily of MADS box genes in angiosperms. Genetics 166:1011–1023.
Krizek BA and Meyerowitz EM. (1996a) Mapping the protein regions responsible for the functional specificities of the Arabidopsis MADS domain organ-identity proteins. Proceedings of the National Academy of Sciences, USA 93:4063–4070.
Krizek BA and Meyerowitz EM. (1996b) The Arabidopsis homeotic genes APETALA3 and PISTILLATA are sufficient to provide the B class organ identity function. Development 122:11–22.[Abstract]
Krizek BA, Riechmann JL, Meyerowitz EM. (1999) Use of the APETALA1 promoter to assay the in vivo function of chimeric MADS box genes. Sexual Plant Reproduction 12:14–26.[CrossRef]
Lamb RS and Irish VF. (2003) Functional divergence within the APETALA/PISTILLATA floral homeotic gene lineages. Proceedings of the National Academy of Sciences, USA 100:6558–6563.
Lawton-Rauh AL, Buckler 4th ES, Purugganan MD. (1999) Patterns of molecular evolution among paralogous floral homeotic genes. Molecular Biology Evolution 16:1037–1045.
Liljegren SJ, Gustafson-Brown C, Pinyopich A, Ditta GS, Yanofsky MF. (1999) Interactions among APETALA1, LEAFY, and TERMINAL FLOWER1 specify meristem fate. The Plant Cell 11:1007–1018.
Mandel MA, Gustafson-Brown C, Savidge B, Yanofsky MF. (1992) Molecular characterization of the Arabidopsis floral homeotic gene APETALA1. Nature 360:273–277.[CrossRef][Medline]
Ohno S. (1970) Evolution by gene duplication (Springer, Heidelberg).
Pelaz S, Gustafson-Brown C, Kohalmi SE, Crosby WL, Yanofsky MF. (2001) APETALA1 and SEPALLATA3 interact to promote flower development. The Plant Journal 26:385–394.[CrossRef][ISI][Medline]
Pellegrini L, Tan S, Richmond TJ. (1995) Structure of serum response factor core bound to DNA. Nature 376:490–498.[CrossRef][Medline]
Pinyopich A, Ditta GS, Savidge B, Liljegren SJ, Baumann E, Wisman E, Yanofsky MF. (2003) Assessing the redundancy of MADS-box genes during carpel and ovule development. Nature 424:85–88.[CrossRef][Medline]
Riechmann JL, Krizek BA, Meyerowitz EM. (1996) Dimerization specificity of Arabidopsis MADS domain homeotic proteins APETALA1, APETALA3, PISTILLATA, and AGAMOUS. Proceedings of the National Academy of Sciences, USA 93:4793–4798.
Savidge B. 1996. Floral meristem specification and floral organ development in Arabidopsis. PhD dissertation, University of California at San Diego, La Jolla, CA.
Schubert D, Lechtenberg B, Forsbach A, Gils M, Bahadur S, Schmidt R. (2004) Silencing in Arabidopsis T-DNA transformants: the predominant role of a gene-specific RNA sensing mechanism versus position effects. The Plant Cell 16:2561–2572.
Smyth D. (2000) A reverse trend-MADS functions revealed. Trends in Plant Science 5:315–317.[CrossRef][ISI][Medline]
Weigel D, Alvarez J, Smyth DR, Yanofsky MF, Meyerowitz EM. (1992) LEAFY controls floral meristem identity in Arabidopsis. Cell 69:843–859.[CrossRef][ISI][Medline]
Yang Y and Jack T. (2004) Defining subdomains of the K domain important for protein–protein interactions of plant MADS proteins. Plant Molecular Biology 55:45–59.[CrossRef][ISI][Medline]
Yang Y, Xiang H, Jack T. (2003) pistillata-5, an Arabidopsis B class mutant with strong defects in petal but not in stamen development. The Plant Journal 33:177–188.[CrossRef][ISI][Medline]
Zachgo S, Silva Ede A, Motte P, Trobner W, Saedler H, Schwarz-Sommer Z. (1995) Functional analysis of the Antirrhinum floral homeotic DEFICIENS gene in vivo and in vitro by using a temperature-sensitive mutant. Development 121:2861–2875.[Abstract]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  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]
|
|
|
|
 |
|
 Flowering Newsletter bibliography for 2006 J. Exp. Bot., April 20, 2007; (2007) erm028v2. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||