Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Molecular Analyses of MADS-Box Genes Trace Back to Gymnosperms the Invention of Fleshy Fruits Alessandro Lovisetto,1 Flavia Guzzo,2 Alice Tadiello,1 Ketti Toffali,2 Alessandro Favretto,1 and Giorgio Casadoro1,* 1 Department of Biology, University of Padua, Padua, Italy Department of Biotechnology, University of Verona, Verona, Italy *Corresponding author: E-mail: [email protected], [email protected]. Associate editor: Neelima Sinha 2 Abstract Key words: gymnosperm fruits, MADS-box genes, Ginkgo biloba, Taxus baccata. Introduction The seed habit is the most complex and evolutionarily successful method of sexual reproduction found in vascular plants, and development of the seed habit represents one of the most significant evolutionary events in the history of vascular plants (quoted from Taylor et al. 2009). Its invention has boosted the success of seedproducing plants (i.e., Gymnosperms and Angiosperms) owing to their increased possibility of spreading into new habitats. Also, the invention of fruits constitutes a remarkable evolutionary advantage because they represent the instrument used by plants to disperse the seed into the environment. In the botanical jargon, a fruit derives from an ovary following an event of fertilization. When an ovary develops into a fruit, its wall becomes the pericarp which, on the basis of its characteristics, can be distinguished into either a parenchymatic fleshy type or a sclerenchymatic dry type (Esaù 1965). Dry and fleshy fruits use different strategies to disperse seeds. In particular, fleshy fruits are generally edible and have attractive colors, so they direct the attention of frugivorous animals toward their juicy pericarp and, by becoming a source of food, they entrust the animals with the task of dispersing their seeds. Animals are mobile and are therefore quite efficient in dispersing the seeds into the environment together with their own excrements. Angiosperms are characterized by the production of flowers; hence, they have ovaries that can be transformed into fruits after fertilization. Having neither flowers nor ovaries to be transformed into a fruit, how do Gymnosperms manage to disperse their seeds? Besides sclerenchymatic strobili that open at maturity to release the seeds, many Gymnosperms produce fleshy eatable structures around the seed and use endozoochory for seed dispersal as Angiosperms do. The fleshy tissues that surround the seeds of Gymnosperms are therefore functionally equivalent to the angiosperm fruit (Herrera 1989). All living orders of Gymnosperms (Cycadales, Ginkgoales, Coniferales, and Gnetales) have species producing fruit-like structures, with fleshy tissues that can originate from a variety of anatomical structures, depending on the species: seed sarcotesta (Cycas and Ginkgo), aril (Taxus and Cephalotaxus), bracts (Juniperus and Ephedra), and seed stalk (Podocarpus). The origin of Gymnosperm fruit-like structures is considered to be polyphyletic (Herrera 1989 and references herein) but also Angiosperm fruits are considered to have a polyphyletic origin in spite of the common carpellar derivation (Knapp 2002 and references herein). Interestingly, the habit of surrounding the seed with a fleshy tissue is not restricted to the Gymnosperms since also some Angiosperms have maintained it. To the latter purpose, worth remembering are the seeds of pomegranate, magnolia, peony, and others, © The Author 2011. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution. All rights reserved. For permissions, please e-mail: [email protected] Mol. Biol. Evol. 29(1):409–419. 2012 doi:10.1093/molbev/msr244 Advance Access publication October 4, 2011 409 Research article Botanical fruits derive from ovaries and their most important function is to favor seed dispersal. Fleshy fruits do so by attracting frugivorous animals that disperse seeds together with their own excrements (endozoochory). Gymnosperms make seeds but have no ovaries to be transformed into fruits. Many species surround their seeds with fleshy structures and use endozoochory to disperse them. Such structures are functionally fruits and can derive from different anatomical parts. Ginkgo biloba and Taxus baccata fruit-like structures differ in their anatomical origin since the outer seed integument becomes fleshy in Ginkgo, whereas in Taxus, the fleshy aril is formed de novo. The ripening characteristics are different, with Ginkgo more rudimentary and Taxus more similar to angiosperm fruits. MADS-box genes are known to be necessary for the formation of flowers and fruits in Angiosperms but also for making both male and female reproductive structures in Gymnosperms. Here, a series of different MADS-box genes have been shown for the first time to be involved also in the formation of gymnosperm fruit-like structures. Apparently, the same gene types have been recruited in phylogenetically distant species to make fleshy structures that also have different anatomical origins. This finding indicates that the main molecular networks operating in the development of fleshy fruits have independently appeared in distantly related Gymnosperm taxa. Hence, the appearance of the seed habit and the accompanying necessity of seed dispersal has led to the invention of the fruit habit that thus seems to have appeared independently of the presence of flowers. MBE Lovisetto et al. · doi:10.1093/molbev/msr244 all of them surrounded by fleshy tissues that attract animals when the fruit containing them is split open. Other Angiosperms, like strawberry- and pome-bearing species, produce false fruits derived from nonovary tissues. The activity of MADS-box transcription factors (TFs) belonging to the C (i.e., AGAMOUS) and E (i.e., SEPALLATA) types is necessary for the development of an ovary. (Pelaz et al. 2000; Theissen and Saedler 2001; Krizek and Fletcher 2005). Recently, TFs of the same type have been shown to be important also for the subsequent transformation of the ovary into a fleshy fruit. In tomato, both TAGL1 (C type) and RIN (E type) genes are involved in the development of a ripe berry (Vrebalov et al. 2002, 2009; Itkin et al. 2009), whereas MADS-box genes of the SEPALLATA type have been shown to have a role in the development of the fleshy strawberry tissues (Seymour et al. 2011). Finally, a C-type gene from peach, when ectopically expressed in tomato, caused the transformation of the normally leafy sepals into ectopic fruits (Tadiello et al. 2009). MADS-box genes are present also in Gymnosperms, and similarly to what occurs in Angiosperms, some of them are involved in the formation of the reproductive structures. In particular, the interaction of B- and C-type proteins leads to the development of the male strobili, whereas the sole C-type proteins appear involved in the formation of the female structures (Theissen and Melzer 2007; Wang et al. 2010). The role played by these TFs has indirectly been confirmed by Zhang et al. (2004) who were able to complement the agamous Arabidopsis mutant by using an AGAMOUS gene from Cycas. Also, an AGAMOUS-like gene from Picea exhibited C-type activity in Arabidopsis and caused the homeotic transformation of sepals into carpel-like structures (Tandre et al. 1998). As regards the SEPALLATA genes, so far, they have not been found in extant Gymnosperms (Nam et al. 2003; Zahn et al. 2005), however, in comprehensive phylogenetic trees, they form a higher degree clade together with the AGL6 genes, and Gymnosperms do have AGL6 genes that are expressed also in their reproductive structures (Mouradov et al. 1998; Winter et al. 1999). Similarly to what occurs in Angiosperms, MADS-box genes might have been employed by Gymnosperms also to make their own fruit-like structures. Should this be the case, the conclusion would be that the basic molecular mechanisms necessary to make a fleshy fruit have actually been invented by Gymnosperms. In this work, we have investigated the involvement of various MADS-box genes in the development of fleshy fruits in Ginkgo biloba and Taxus baccata, species that represent two different developmental models owing to the fact that in Ginkgo, it is the outermost seed tegument (sarcotesta) that grows and becomes fleshy, whereas in Taxus, the fleshy aril is formed de novo at the base of the ovule. Genes notoriously involved in the formation of colors and in softening of angiosperm fruits have also been studied in order to understand whether the molecular mechanisms involved in fruit ripening in Gymnosperms are like those present in Angiosperms. 410 Materials and Methods Plant Material and RNA Extraction Plant material came from the Botanic Garden of Padua. In Ginkgo, initial samples consisted of whole ovules, then the fleshy sarcotesta could be analyzed separately at various development stages indicated by dates in the figures. In Taxus, the aril develops de novo as a collar at the base of the ovule (supplementary data 1, Supplementary Material online). Then it becomes visible as a green leafy structure (LA), which grows thick and fleshy (FA) and starts to gradually develop a reddish color (Breaker: B1, B2) that becomes bright and evenly distributed at the ripe stage (R). Total RNA was extracted from different tissues according to Chang et al. (1993). For each tissue, the extraction was made using a pool of samples and was repeated at least twice. RNA yield and purity were checked by means of ultraviolet (UV) absorption spectra, whereas RNA integrity was ascertained by electrophoresis in agarose gel. cDNA Isolation, Sequencing, and Analysis All the cDNAs isolated de novo for this work were obtained by reverse transcription polymerase chain reaction (RT-PCR) using primers constructed following alignment of known sequences. Primers are listed in supplementary data 2, Supplementary Material online. In particular, for ripeningrelated genes, the template consisted of RNA extracted from ripening pulp samples (Ginkgo: pulp of 19 July 2006; Taxus: breaker-2 arils), whereas for MADS-box genes, the template RNA was extracted from young ovules. The sequences obtained were registered in the Genebank database with the following accession numbers: JF440955 (G. biloba expansin); JF519740 (G. biloba pectate lyase [PL]); JF519741 (G. biloba phytoene synthase); JF519743 (G. biloba ACC oxidase); JF519744 (T. baccata PL); JF519745 (T. baccata expansin); JF519747 (T. baccata phytoene synthase); JF519748 (T. baccata chalcone synthase); JF519750 (T. baccata ACO oxidase); JF519751 (T. baccata ethylene receptor); JF519754 (T. baccata AGAMOUS); JF519755 (T. baccata AGL6); and JF519756 (T. baccata TM8). DNA sequencing was performed at BMR Genomics (Padua). Sequence manipulations, analyses, and alignments were performed using the ‘‘Lasergene’’ software package (DNASTAR). Construction of a MADS-Box Phylogenetic Tree A set of MADS-box protein sequences downloaded from GeneBank plus sequences obtained in this work were used to construct a phylogenetic tree. The Genbank accession numbers are AT4G18960 (Arabidopsis thaliana AG); NM130127 (A. thaliana AGL6); NM115976 (A. thaliana AGL13); PRU42400 (Pinus radiata PrMADS2); AB029470 (G. biloba GbMADS8); AB029463 (G. biloba GbMADS1); U76726 (P. radiata PrMADS3); AB022665 (Gnetum parvifolium GpMADS3); L18924 (Zea mays ZAG1); AY114304 (G. biloba AG); AY295079 (Cycas edentata AG); AJ132215 (Gnetum gnemon GGM9); AJ132217 (G. gnemon GGM11); AB359029 (Cryptomeria japonica AGL6); AB359027 (C. japonica M09 TM8); AB359028 (C. japonica O23 MBE Gymnosperm Fleshy Fruits · doi:10.1093/molbev/msr244 TM8); AB029468 (G. biloba GbMADS6); AB029469 (G. biloba GbMADS7); AB029473 (G. biloba GbMADS11); XM002283880 (Vitis vinifera TM8); X60760 (Lycopersicum esculentum LeTM8); AB046596 (Cucumis sativus ERAF17); X79280 (Picea abies DAL2); AB035567 (Chara globularis CgMADS1); AB359031 (C. japonica AG); PMU69482 (Picea mariana SAG1A); AJ132209 (G. gnemon GGM3); JF519754 (T. baccata TbAG); JF519755 (T. baccata TbAGL6); and JF519756 (T. bacccata TbTM8). In particular, the tree was constructed using the ‘‘MIK domain,’’ that is, the MADS domain (60 aa) plus most of the I and K domains (90 aa) similarly to Winter et al. (1999), Becker and Theissen (2003), and Melzer et al. (2010). The amino acid sequences were aligned with the CLUSTALW program, and the obtained alignments were used to construct the tree with the MEGA 4.0.2 program. The tree was constructed with the neighbor joining method (Saitou and Nei 1987) and evaluated by bootstrap analysis. Analysis of Gene Expression This analysis was performed by standard real-time PCR. Six micrograms of total RNA was pretreated with 2 U of DNase I (Promega). The first-strand cDNA was synthesized from 3 lg of the DNase I–treated RNA by means of the High-Capacity cDNA Archive Kit (Applied Biosystems, Foster City, CA), using random hexamers as primers. Primer sequences for the selected genes are listed in supplementary data 2, Supplementary Material online. The internal standards consisted of the internal transcribed spacer of the ribosomal RNA sequences specific for each species (supplementary data 2, Supplementary Material online). PCR was carried out with the Gene Amp 7500 Sequence Detection System (Applied Biosystems). The obtained CT values were analyzed by means of the Q-gene software by averaging three independently calculated normalized expression values for each sample. Expression values are given as the mean of the normalized expression values of the triplicates, calculated according to equation (2) of the Q-gene software (Muller et al. 2002). In Situ Hybridization Analysis Samples were fixed with 2% formaldehyde and 0.25% glutaraldehyde in phosphate-buffered saline buffer, pH 7.5. Clearing, infiltration, embedding, sectioning, prehybridization treatments, and posthybridization washes were performed according to Drews (1998), with minor modifications. PCR fragments of C domain and/or 3#untranslated region were isolated, cloned, and checked by sequencing. The list of genes examined and the relative primers for the PCR are listed in supplementary data 3, Supplementary Material online. The PCR products were cloned and the sequence validated. The DIG-labeled RNA sense and antisense probes were synthesized with the DIG RNA labeling Kit SP6/T7 (Roche diagnostics, Germany), the immunological detection was performed with the anti-Digoxigenin-AP-Fab fragments antibody (Roche Diagnostics), and the SIGMA FAST BCIP/ NBT tablets (Sigma) were used as phosphatase substrate, all according to the manufacturer instructions. Ethylene Treatments Arils at the stage breaker-2 were either used immediately after harvesting (control) or subjected to treatments with exogenous ethylene (100 ll/l) in air as follows: a pool of samples was left in air for 24 h, another pool was flushed 12 h with ethylene in an air-tight chamber and allowed to stand in air for 12 more hours. A group of samples was kept in air for 48 h while another pool was flushed 24 h with ethylene and moved afterward to air for 24 more hours. Results MADS-Box Genes In young ovules of both species, the expression of various MADS-box genes was studied by in situ hybridization. When fleshy tissues could be separated by the developing seed, gene expression was studied by real-time PCR. AGAMOUS-Like AGAMOUS (AG) is necessary to make female reproductive structures in Angiosperms (Rijpkema et al. 2010), and it is also involved in the development of fleshy fruits like tomato (Itkin et al. 2009; Vrebalov et al. 2009) and peach (Tadiello et al. 2009). Since AGAMOUS has been shown to be expressed also in Gymnosperm female reproductive structures (Tandre et al. 1998; Zhang et al. 2004; Melzer et al. 2010), it appeared an interesting candidate gene for this study. In particular, AGAMOUS had already been isolated in Ginkgo (Jager et al. 2003), whereas no MADS-box gene was known for Taxus, so an AG cDNA had to be isolated by RT-PCR. Its sequence was used together with other AG sequences to construct a phylogenetic tree, which showed that it grouped with the other gymnosperm AG genes (fig. 1). In particular it clustered with that of Cryptomeria (Cupressaceae), probably because Taxaceae are phylogenetically closer to Cupressaceae than to Pinaceae (Judd et al. 2008). In situ hybridization analysis revealed that in Ginkgo, AGAMOUS was strongly expressed both in the nucellus and in the not yet differentiated integuments of the young ovules. Strong expression was also seen where the ovule joins the peduncle, but the expression faded out moving away from the ovule (fig. 2A and D). Upon differentiation of the three zones that will give rise to the outer fleshy sarcotesta, to the sclerotesta, and to the thin innermost seed coat, respectively, AG was still strongly expressed throughout the ovule (fig. 2C). This expression pattern was also evident when a thick fleshy coat started to be visible (fig. 2F). In Taxus, AG was homogeneously expressed throughout the female reproductive structure, including ovule, distal portion of the peduncle and the outermost sterile scales (fig. 2G). Later on the AG expression was particularly localized to the young developing aril, immediately below the ovule (fig. 2I and K, arrows). The expression of AGAMOUS obtained by real-time PCR (fig. 3A and B) confirmed previous data for Ginkgo about the presence of transcripts in male strobili and ovules (Jager 411 Lovisetto et al. · doi:10.1093/molbev/msr244 FIG. 1. Phylogenetic tree showing subfamilies of the AGAMOUS, AGL6, TM8 MIKC–type MADS-box proteins. Sequences relative to the genes characterized in this work are marked by different arrows according to subfamily type. et al. 2003). As regards the fleshy sarcotesta, AG transcripts reached very high levels during the phase of fast growth (26 May), gradually decreased thereafter and, though not much, increased again during the ripening stage (29 August). In Taxus, AGAMOUS was expressed in both male and female structures, although at much higher levels in the ovule. As regards the aril, the expression was high in the green leafy one, continued to slightly increase during the acquisition of fleshyness, and reached a maximum at the start of ripening. Thereafter, there was a little decrease but the transcripts remained high until the red ripe stage. AGL6-Like In G. gnemon, two different AGL6 genes are expressed in the reproductive structures, although one gene seems to be more ovule specific (Winter et al. 1999). Accordingly, AGL6 genes are important for our study. Two full-length cDNAs coding for two different AGL6 proteins (GbMADS1 and GbMADS8, respectively) in Ginkgo were singled out by a search in public databases. As regards Taxus, we had to clone ourselves the AGL6 cDNA. Yet, in spite of considerable efforts, only one AGL6 cDNA could be isolated. 412 MBE The sequences from Ginkgo and from Taxus were used together with other AGL6 sequences to construct a phylogenetic tree (fig. 1). The two Ginkgo AGL6 genes fell into two separate clades, each together with another sequence from P. radiata, whereas the sequence from Taxus formed a separate clade with a sequence from C. japonica. The in situ hybridization analysis showed that in young Ginkgo ovules, both GbMADS1 and GbMADS8 were strongly expressed in the integuments, nucellus, at the base of the ovules, whereas a weaker signal was present in the peduncle region adjoining the ovule (fig. 2P and Q). In Taxus, the AGL6 gene was strongly expressed in the young developing aril and in the ovule and less expressed in the innermost scales and within the peduncle (fig. 2L). In the latter region, the AGL6 expression faded out upon moving away from the ovule. The real-time PCR analysis of the Ginkgo AGL6 genes showed important differences in their expression profile (fig. 3C). In general, GbMADS8 had a much higher level of expression than GbMADS1 and seemed to be more specific for the reproductive structures since GbMADS1 was expressed also in young leaves. For both genes, the expression was significantly higher in ovules than in male strobili. GbMADS1 showed very low and decreasing transcript amounts in the pulp as to be barely detectable in the summer samples. Also the expression of GbMADS8 showed a decreasing pattern until the last sampling when a marked increase was measured concomitant with ripening. The Taxus AGL6 gene was not expressed in leaves therefore, like the Ginkgo GbMADS8, it seems specific for the reproductive structures (fig. 3D). In arils, the expression was generally higher than in male strobili and in ovules. In particular, a marked increase occurred during the transition from leafy to fleshy stages. Then, there was some decrease at the breaker-1 stage followed by a further increase leading to maximum expression at the breaker-2 stage, that is when arils are in full ripening process. TM8-Like Degenerate primers constructed following the protocol described in Hileman et al. (2006) were also used to try to isolate a second AGL6 gene from Taxus. Instead of the long sought AGL6 gene, an unexpected TM8-like gene was obtained. An informatic search was therefore done, and three TM8-like sequences from Ginkgo could be retrieved from public databases. The known TM8-like sequences were then used to construct a phylogenetic tree (fig. 1). Interestingly, angiosperms and gymnosperms sequences grouped together into a general TM8 clade. Within such clade, three subgroups could be distinguished that reflected the phylogenetic relationships among the species considered: one was formed by the sole angiosperm sequences, the second groups contained the sequences from Taxus and Cryptomeria (both of them Coniferales), whereas the third group was formed by the three Ginkgo sequences. In situ hybridization analysis evidenced that in Ginkgo, the three TM8-like genes had overlapping expression in ovules. Strong GbMADS6 and GbMADS11 expression was Gymnosperm Fleshy Fruits · doi:10.1093/molbev/msr244 MBE FIG. 2. Expression of AGAMOUS (panels A, C, D, F, G, I, and K), AGL6-like (GbMADS1: panel P; GbMADS8: panel Q; and TbAGL6: panel L), and TM8-like (Taxus TM8-like: panel N; GbMADS6: panels R; GbMADS11: panel S; and GbMADS7: panel T) analyzed by in situ hybridization in young Ginkgo (A, C, D, F, P, Q, R, S, and T) and Taxus (G, I, K, L, and N) female reproductive structures. Panels B, E, H, J, M, O, and U: sense control in Ginkgo (B, E, and U) and Taxus (H, J, M, and O). pc, pollen chamber; in, integuments; es, embryo sac; oi, outer integument; and ii, inner integument. Arrows: developing arils. AG 5 AGAMOUS; AGL6 5 AGL6-like; TM8 5 TM8-like; s 5 sense probe; and a 5 antisense probe. observed throughout the ovule, whereas a weaker signal was visible in the peduncle region adjoining the ovule (fig. 2R and S). GbMADS7 expression was very weak compared with the other two TM8-like genes but showed a similar pattern (fig. 2T). In Taxus, TM8 was strongly expressed in the ovule and in the young developing arils and weakly expressed in the distal part of the peduncle and in the more apical sterile scales (fig. 2N). Real-time PCR analysis revealed that of the three Ginkgo TM8-like sequences, GbMADS7 was always expressed at very low amounts compared with the other two (GbMADS6 and GbMADS11). In particular, the highest expression level was observed in young leaves, followed by the ovules, whereas in the pulp, its transcripts were always very low (fig. 3E). GbMADS6 was the most highly expressed of the three genes, and similarly to the other two, the maximum transcript amount was found in young leaves. Also male strobili and ovules exhibited high transcript amounts comparable to those present in the young sarcotesta undergoing a fast increase in thickness. Thereafter, in the latter tissue, the expression slightly decreased albeit remaining quite high until the last sampling when ripening occurred. The transcripts of the GbMADS11 gene were high in leaves, ovules, and male strobili. In the fleshy sarcotesta, the initially low expression (26 May) showed an increase during the subsequent stage of fast growth (16 June), an 413 Lovisetto et al. · doi:10.1093/molbev/msr244 MBE FIG. 3. Relative expression profiles of Ginkgo biloba (Gb) and Taxus baccata (Tb) MADS-box genes: AG (AGAMOUS), AGL6, and TM8. Gb samples: YL, young leaf; #, male strobili; $, young ovules; 26 May, 16 June, 19 July, 01 August, and 29 August sampling dates (2006 season) of pulps at different developmental stages. Tb samples: YL, young leaf ; #, male strobili; $, young ovules; LA, leafy arils; FA, fleshy green arils; B1, breaker-1 arils; B2, breaker-2 arils; and R, red ripe arils. Values (means of the normalized expression) have been obtained by RT-PCR analyses. Bars are the standard deviations from the means. unexpected drop in early summer (19/7) that was followed by a significant increase to higher amounts when the fruit was undergoing the ripening process. Similarly to Ginkgo, also the Taxus TM8-like gene was expressed in young leaves, male strobili, and ovules (fig. 3F). However, contrary to Ginkgo, the highest expression was found in the arils and not in the leaves. In particular, there was a significant transcript amount during growth of the leafy arils, followed by a further increment that led to a maximum concomitant with the development of fleshyness. Thereafter, the expression slowly decreased, though remaining substantially high throughout the entire ripening process. Ripening Syndrome The process of ripening transforms fleshy fruits into structures that are attractive for animals. In short, fruits become more or less soft, their initially green color is substituted by other colors that catch the eye of animals, while the 414 production of characteristic aromas make them an appealing source of food. Such processes constitute the ripening syndrome, and in this work, attention has been given to softening and changes in color, both of them well known at the molecular level in many angiosperm fruits. The formation of aroma has not been considered because it is extremely variable and far less known. Softening In general, softening occurs as a consequence of pectin and hemicellulose degradation, and different enzymes have been shown to act on the different components (Brett and Waldron 1996; Brummell and Harpster 2001). We have chosen to analyze genes coding for PL and expansin whose role has been shown in strawberry (Jiménez-Bermùdez et al. 2002) and tomato (Brummell et al. 1999), respectively. Pectate Lyase. By aligning various known sequences, a particularly conserved region was identified, and degenerate primers were prepared. RT-PCR experiments allowed Gymnosperm Fleshy Fruits · doi:10.1093/molbev/msr244 MBE the cloning of two cDNA fragments coding for PL in Ginkgo and Taxus, respectively. In Ginkgo, the gene was mostly expressed in growing tissues like young leaves, young reproductive structures (male strobili and ovules), and young sarcotesta (fig. 4A). No measurable expression was observed during pulp ripening. On the contrary, in Taxus, the PL gene showed a ripeningspecific expression pattern (fig. 4B). Transcripts started to be measurable in green fleshy arils, then their amount showed a great and continuous increase up to a maximum in the breaker-2 arils and was finally followed by a slight decrease at the red ripe stage. Expansin (EXP). Expansins are known to cause cell wall loosening by breaking bonds between cellulose fibrils and matrix polysaccharides (Cosgrove 2000). Since cell wall loosening occurs both during growth-related tissue expansion and during fruit softening, the activity of expansins can be found in every type of cell separation events. The importance of a softening specific expansin has been shown in tomato (Brummell et al. 1999). Degenerate primers for RT-PCR experiments were prepared by aligning various EXP sequences, and an EXP cDNA fragment was obtained for both Ginkgo and Taxus. The Ginkgo expansin had the highest expression in young leaves (fig. 4C). Though much less, it was expressed also in male strobili and ovules. As regards the pulp, the gene was expressed mainly in the young growing sarcotesta and decreased thereafter until the early summer samples, then it showed a slight increase during ripening. In Taxus, the EXP gene was expressed at low levels in young leaves and ovules and almost 3-fold more in the male strobili (fig. 4D). Much higher amounts of transcript were found in the ripening arils where the gene reached its maximum expression in the breaker-2 arils. Change of Color Ripening fruits usually lose their green color and accumulate other pigments, thus signaling to animals that they have reached the eatable stage. But for betalains whose presence seems limited to Caryophyllales, the nongreen colors in plants, and particularly in fruits, are mostly due to either carotenoids or anthocyanins/flavonoids (Tanaka et al. 2008). The enzyme that specifically leads to the synthesis of carotenoids is phytoene synthase (PSY). Anthocyanins/flavonoids are phenols, and chalcone synthase (CHS) is an enzyme of the biosynthetic pathway that acts upstream of all these colored molecules (Taiz and Zeiger 2006). Phytoene Synthase. Degenerate primers were prepared to be used in RT-PCR experiments that led to the cloning of two cDNA fragments encoding PSY in Ginkgo and in Taxus, respectively. In Ginkgo, high amounts of PSY transcripts were found in young leaves (fig. 4E). Extremely, low levels of expression were observed in the yellowish male strobili, in the green ovules, and in the young green sarcotesta. In Taxus, some expression the PSY gene (fig. 4F) was observed in the young leaves, but the highest transcript amount was FIG. 4. Relative expression profiles of Ginkgo biloba (Gb) and Taxus baccata (Tb) genes related to softening (A–D), color (E–H), and ethylene biosynthesis (I, L). Samples are indicated as in figure 3. Panel (M) shows relative expression of an ethylene receptor gene (ETR) of T. baccata, whereas panel (N) shows the relative expression of the yew ACO gene in arils either kept in air or treated with exogenous ethylene (100 ll/l) in air (a: arils frozen immediately after sampling ; b: arils kept 24 h in air; c: arils kept 12 h in ethylene followed by 12 h in air; d: arils kept 48 h in air; and e: arils kept 24 h in ethylene followed by 24 h in air). Values (means of the normalized expression) have been obtained by RT-PCR analyses. Bars are the standard deviations from the means. 415 MBE Lovisetto et al. · doi:10.1093/molbev/msr244 found in the arils where the gene expression showed a great and continuous increase up to a maximum at the breaker-2 stage, followed by a decrease at the red ripe stage. Chalcone Synthase. Degenerate primers to be used in RTPCR experiments were prepared by aligning several cDNA sequences. One cDNA fragment coding for chalcone synthase was obtained for each species. In Ginkgo, the CHS gene was highly expressed in male strobili and in ovules (fig. 4G). Much lower amounts of transcript were visible in young leaves and in young growing sarcotesta. All the other samples showed extremely low levels of expression, although a slightly increasing pattern was visible during ripening in the summer samples. In Taxus, the highest CHS transcript amount was observed in the young leaves, followed by that present in the ovules. Arils had extremely low amounts of CHS transcripts, but for the green leafy ones that had a bit higher expression level (fig. 4H). Ethylene-Related Genes Angiosperm fruits may or may not have their ripening controlled by ethylene. In the first case, fruits are said to be climacteric (e.g., tomato) and show a peak of ethylene production (climacteric ethylene) at the onset of ripening. The synthesis of climacteric ethylene is autocatalytic; hence, the hormone is able to stimulate its own synthesis, contrary to what happens in young growing tissues where ethylene has not such capacity. In the second case, fruits do not produce any climacteric ethylene and are said to be nonclimacteric (e.g., grape). ACC oxidase (ACO) is the enzyme that does actually synthesize the hormone starting from the substrate ACC (1-aminocyclopropane-1-carboxylic acid). Since a relation has been found between the expression of ACO and the amount of measured ethylene, the expression of ACO is often used as an indirect measure of ethylene production. ACC Oxidase. By using degenerate primers, two cDNA fragments were obtained that encoded a Ginkgo and a Taxus ACO, respectively. In Ginkgo, the ACO gene was mostly expressed in young leaves and, at much lower levels, in male strobili, whereas its expression was hardly detectable in all the other samples (fig. 4I). Taxus behaved differently since the ACO gene had a clear ripening-specific pattern of expression. The transcript amount was extremely low in all the samples except in the arils that were becoming fleshy. Particularly striking was the high expression increase occurring during the late ripening stages (fig. 4L). Ethylene Receptor. Based on the pattern of ACO expression, ethylene seemed to be important for ripening in Taxus but not in Ginkgo. Accordingly, only for Taxus, a cDNA was cloned that codes for an ethylene receptor, and its expression analyzed in various tissues (fig. 4M). The ETR gene appeared to be expressed in all the examined tissues, and this is in agreement with the fact that basal levels of ethylene are made in plants by all living tissues (Taiz and Zeiger 2006). However, the expression of this gene showed a marked and significant increase in breaker-2 and red arils that paralleled the observed increases of ACO expression. 416 Treatment of Arils with Ethylene. The quasi parallel increasing expression of the ACO and ETR genes during ripening suggested that the arils of Taxus might behave like climacteric fruits. Since ethylene can stimulate its own synthesis in climacteric fruits, ethylene was used to treat arils at the breaker-2 stage, that is, arils in which the synthesis of ethylene was apparently ‘‘on’’ but not yet at its maximum rate. At the end of the experiment, the expression of the ACO gene was measured in the different samples (fig. 4N). As expected, the breaker-2 arils kept 24 h in air showed an increased expression of ACO, but surprisingly, a comparable increment was found also in the ethylene-treated arils. Similar results were obtained with the arils altogether treated for 48 h, albeit in this case, the expression values were much higher. Discussion Different Ripening Syndromes Are Present in Ginkgo and Taxus Fruits The two cell separation–related genes had a ripeningspecific expression in Taxus but not in Ginkgo. Actually, albeit at very low levels, the Ginkgo expansin showed some increasing expression in late stages of fleshy sarcotesta development; therefore, it might be involved also in the process of softening. We cannot however exclude that other not yet cloned genes are involved in softening of the Ginkgo fruit-like structures. Alternatively, the softening in Ginkgo might occur, at least partly, by water loss through the cuticle followed by a loss of tissue stiffness, the latter a mechanism for softening suggested by Saladié et al. (2007). Also, the modality for changing color during ripening turned out to be different in the two species. While in the fleshy sarcotesta of Ginkgo, the phytoene synthase (PSY) transcripts were barely detectable; in Taxus, the PSY gene had a marked increasing expression in the ripening arils. Therefore, it can be concluded that carotenoids are de novo accumulated in ripening Taxus arils. The latter statement is also supported by the results of a chromatography where pigments extracted by means of diethyl ether from both leafy green arils and red ripe ones were compared. Interestingly, the red berries contained newly formed red carotenoids that were not present in the green leafy arils (supplementary data 4, Supplementary Material online). The expression of the chalcone synthase (CHS) gene seemed specific for ovules and young leaves in Taxus and for ovules and male strobili in Ginkgo. In particular, Ginkgo has deciduous leaves and forms its reproductive structures when the new foliage is still poorly developed; therefore, they might be easily damaged by UV radiations. To the latter purpose, an Arabidopsis mutant, incapable to make flavonoids because deprived of a functional CHS, could survive only in the absence of UV radiation (Li et al. 1993), thus evidencing the protective role performed by flavonoids. Accordingly, the high levels of CHS expression observed in Ginkgo reproductive structures, but also in Gymnosperm Fleshy Fruits · doi:10.1093/molbev/msr244 ovules and young leaves of Taxus, might be linked to the formation of protective compounds. Therefore, most of the yellowish color in the ripe and senescent Ginkgo fruit-like structures might actually be due to photosynthetic carotenoids becoming visible after chlorophyll breakdown, as it happens with senescent leaves. This idea seems supported by the results of a chromatography where carotenoids from either green or yellow pulps were compared and no apparent carotenoid difference was visible (supplementary data 4, Supplementary Material online). In Ginkgo, the expression of an ACO gene suggests that this gene is involved in the synthesis of growth-related hormone. In Taxus, the aril- and ripening-specific expression of the ACO gene indicates that ethylene is involved in the ripening of yew berries. A quasi parallel increasing expression of an ethylene receptor gene during ripening suggested that the aril might be like a climacteric fruit. Yet, this idea turned out to be untrue since treatments with exogenous ethylene were unable to stimulate the autocatalytic hormone synthesis characteristic of the truly climacteric fruits. Hence, it seems that the mechanism of climacteric fruit ripening has not yet appeared in Taxus. In general, between the two gymnosperms, the ripening of Taxus arils is the most similar to that of angiosperm fruits. Ginkgophyta are considered to have appeared in lower Permian (Zhou 2009), whereas Taxaceae appearance has been situated in middle Jurassic (Taylor et al. 2009). Accordingly, the apparently simpler ripening syndrome of the Ginkgo fruit-like structures might reflect the greater antiquity of Ginkgophyta compared with Taxaceae. MADS-Box Genes of the Same Type Are Involved in the Development of the Fleshy Fruit Habit in Gymnosperms The biochemical pathways that concur to establish the ripening syndrome of fleshy fruits are under the control of different TFs. For instance, a MYB TF is involved in the formation of colored molecules in ripening strawberries (Aharoni et al. 2001). The synthesis of the bilberry fruit pigments has been shown to be under the control of a SQUAMOSA MADS-box TF (Jaakola et al. 2010). In tomato, it has been shown that LeHB1, an HD-ZIP TF, can bind to the promoter of an ACO gene, thus regulating its expression (Lin et al. 2008). The accumulation of pigments in the tomato berry has been shown to be regulated by the TAGL1 MADS-box gene (Itkin et al. 2009; Vrebalov et al. 2009). The various pathways that form the ripening syndrome are situated at the end of the overall development of fleshy fruits and, before they can be activated, a given anatomical part must first undertake the developmental pathway leading to the fleshy fruit condition. Once the fleshy fruit habit has been established, the ripening syndrome typical for each species will be settled. This has been shown in tomato where, by ectopically expressing a C type MADS-box gene from peach, the normally green and leafy sepals acquired the characteristic of carpels and were thus able to develop as ectopic fruits that ripened like those derived from the ovary (Tadiello et al. 2009). Therefore, of paramount MBE importance is the function of those genes whose activity is able to switch on the fleshy fruit habit. MADS-box genes are interesting candidate genes because they are involved in the formation of reproductive structures in both Gymnosperms and Angiosperms (Theissen et al. 2000). In ovary-derived fruits, MADS-box genes are further used for the subsequent transformation of the carpel into a fruit (Itkin et al. 2009; Tadiello et al. 2009; Vrebalov et al. 2009). In this work, MADS-box genes like AGAMOUS and AGL6 have been shown to be involved also in the formation of the gymnosperm fruit. Only one AGAMOUS gene is known in Gymnosperms, although a different situation is found regarding AGL6 genes. In various Gymnosperms, two AGL6 genes have been evidenced; however, in Soltis et al. (2007), three such genes were reported for Zamia fischeri and only one gene for Welwitschia mirabilis. A southern analysis carried out by us indicated that two AGL6 genes might be present also in Taxus (supplementary data 5, Supplementary Material online). However, in spite of considerable efforts, we were able to clone only one AGL6 gene in this species, and such a result is in agreement with the high levels of expression of this gene in the arils. Therefore, though it appears that we failed to clone a second AGL6 gene, the one characterized in this work is certainly a gene highly involved in the formation of the Taxus arils, and together with the results obtained for the Ginkgo GbMADS8 AGL6 gene, it indicates that genes belonging to the AGL6 group undoubtedly participate in the development and ripening of the fruit-like structures of both gymnosperm species. As summarized in Introduction, C- and E-type MADSbox genes are important for the development of angiosperm fruits. E-type (i.e., SEPALLATA) genes are not present in Gymnosperms that have instead the closely related AGL6 genes. In Petunia, it was recently demonstrated that an AGL6 gene had a SEPALLATA-like function in floral patterning (Rijpkema et al. 2009). Moreover, it was suggested in Melzer et al. (2010) that the gymnosperm AGL6 genes might carry out some functions performed in Angiosperms by the E-type genes. Accordingly, some sort of ‘‘SEPALLATA function’’ in gymnosperm fruits might be played at least by the AGL6 gene whose transcripts are continuously present in the developing fruits of both Ginkgo and Taxus. Besides AGAMOUS and AGL6, a TM8-like cDNA for Taxus and three cDNAs coding for the same gene type in Ginkgo were also obtained. TM8 is a poorly known gene isolated for the first time in tomato (Pnueli et al. 1991), where it was found that in fruits, it is expressed throughout development and ripening with an apparent maximum at the ‘‘yellow’’ stage (Hileman et al. 2006). Like in tomato, and in Ginkgo, the three TM8-like genes were expressed in both vegetative and reproductive tissues, and two of them were particularly expressed throughout development and ripening of the fleshy sarcotesta, thus indicating that they have a role in the formation of the Ginkgo fruit-like structures. In Taxus, the expression pattern of the TM8-like gene suggests that it might be more important for the aril formation and ripening than in vegetative tissues. 417 Lovisetto et al. · doi:10.1093/molbev/msr244 AGAMOUS, AGL6, and TM8-like MADS-box genes appear all to be involved in the development of the fleshy structures in both Ginkgo and Taxus. Therefore, distantly related taxa have recruited the same type of genes in order to achieve the fleshy fruit habit, and this in spite of the different anatomical structures actually employed for this purpose. Nothing is known about Gnetales fleshy fruits, but in Cycadales (C. edentata), it was evidenced that AGAMOUS was highly expressed also in the young sarcotesta (Zhang et al. 2004). The fruit-like structures of Gymnosperms have in common a close vicinity to the developing seed; hence, as observed in Ginkgo and in Taxus, genes normally involved in the development of ovule/seed have just to spread their expression to close anatomical territories for the development of a fleshy structure to be accomplished. Therefore, as far as MADS-box genes are concerned, the general molecular mechanism seems to be the same, although it appears to have developed various times as suggested by the different anatomical tissues toward which such genes had to shift their expression for the fruit-like structure to form. Angiosperms use AGAMOUS, TM8, and SEPALLATA genes, the latter absent in Gymnosperms, for the development of fleshy fruits. Given the phylogenetic vicinity of AGL6 and SEPALLATA genes, it seems that the main molecular networks underlying the formation of the fleshy fruit habit have developed independently of the presence of flowers and have basically been maintained in the flower-producing Angiosperms. This idea is in agreement with a recent finding (Chanderbali et al. 2010), which shows that genetic programs operating in gymnosperm female cones are also present in angiosperm carpels. Supplementary Material Supplementary data 1–5 are available at Molecular Biology and Evolution online (http://www.mbe.oxfordjournals.org/). Acknowledgments Dr E. Negrisolo is thanked for helpful advice regarding the construction of the phylogenetic tree. A. Pavanello is thanked for skillful technical help. All plant material used in this work was obtained by trees growing at the Botanic Gardens of the University of Padua. This work was supported by a grant from Ministero dell’Istruzione, dell’Università e della Ricerca (MIUR), Italy, to G.C. References Aharoni A, Ric De Vos CH, Wein M, Sun Z, Greco R, Kroon A, Mol JNM, O’Connell AP. 2001. The strawberry FaMYB transcription factors suppresses anthocyanin and flavonol accumulation in transgenic tobacco. Plant J. 28:319–332. Becker A, Theissen G. 2003. The major clades of MADS-box genes and their role in the development and evolution of flowering plants. Mol Phylogenet Evol. 29(3):464–489. Brett CT, Waldron KW. 1996. Physiology and biochemistry of plant cell walls, 2nd ed. London: Chapman & Hall. Brummell DA, Harpster MH. 2001. Cell wall metabolism in fruit softening and quality and its manipulation in transgenic plant. Plant Mol Biol. 47:311–339. 418 MBE Brummell DA, Harpster MH, Civello PM, Palys JM, Bennett AB, Dunsmuir P. 1999. Modification of expansin protein abundance in tomato fruit alters softening and cell wall polymer metabolism during ripening. Plant Cell 11:2203–2216. Chanderbali AS, Yoo M-J, Zahn LM, et al. (13 co-authors). 2010. Conservation and canalization of gene expression during angiosperm diversification accompany the origin and evolution of the flower. Proc Natl Acad Sci U S A. 107:22570–22575. Chang S, Puyear J, Cairney J. 1993. A simple and efficient method for isolating RNA from pine trees. Plant Mol Biol Rep. 11:113–116. Cosgrove DJ. 2000. Loosening of plant cell walls by expansins. Nature 407:321–326. Drews GN. 1998. In situ hybridization. In: Martinez-Zapater JM, Salinas J, editors. Arabidopsis protocols. Totowa (NJ): Humana Press. p. 353–371. Esaù K. 1965. Plant anatomy, 2nd ed. New York: John Wiley & Sons. Herrera CM. 1989. Seed dispersal by animals: a role in angiosperm diversification? Am Nat. 133:309–322. Hileman LC, Sundstrom JF, Litt A, Chen M, Shumba T, Irish VF. 2006. Molecular and phylogenetic analyses of the MADS-box gene family in tomato. Mol Biol Evol. 23:2245–2258. Itkin M, Seybold H, Breitel D, Rogachev I, Meir S, Aharoni A. 2009. TOMATO AGAMOUS-LIKE 1 is a component of the fruit ripening regulatory network. Plant J. 60:1081–1095. Jaakola L, Poole M, Jones MO, et al. (12 co-authors). 2010. A SQUAMOSA MADS box gene involved in the regulation of anthocyanin accumulation in bilberry fruits. Plant Physiol. 153:1619–1629. Jager M, Hassanin A, Manuel M, Le Guyader H, Deutsch J. 2003. MADS-box genes in Ginkgo biloba and the evolution of the AGAMOUS family. Mol Biol Evol. 20:842–854. Jiménez-Bermùdez S, Redondo-Nevado J, Munoz-Blanco J, Caballero JL, López-Aranda JM, Valpuesta V, Pliego-Alfaro F, Quesada MA, Mercado JA. 2002. Manipulation of strawberry fruit softening by antisense expression of pectate lyase gene. Plant Physiol. 128:751–759. Judd WS, Campbell CS, Kellogg EA, Stevens PF, Donoghue MJ. 2008. Plant systematics. A phylogenetic approach, 3rd ed. Sunderland (MA): Sinauer Associates, Inc. Knapp S. 2002. Tobacco to tomatoes: a phylogenetic perspective on fruit diversity in the Solanaceae. J Exp Bot. 53(377):2001–2022. Krizek BA, Fletcher JC. 2005. Molecular mechanism of flower development: an armchair guide. Nat Rev Genet. 6:688–698. Li J, Ou-Lee TM, Raba R, Amundson RG, Last RL. 1993. Arabidopsis flavonoid mutants are hypersensitive to UV-B irradiation. Plant Cell 5:171–179. Lin Z, Hong Y, Yin M, Li C, Zhang K, Grierson D. 2008. A tomato HD-Zip homeobox protein, LeHB-1, plays an important role in floral organogenesis and ripening. Plant J. 55:301–310. Melzer R, Wang YQ, Theissen G. 2010. The naked and the dead: the ABCs of gymnosperm reproduction and the origin of the angiosperm flower. Semin Cell Dev Biol. 21:118–128. Mouradov A, Glassick TV, Hamdorf BA, Murphy LC, Marla SS, Yang Y, Teasdale RD. 1998. Family of MADS-Box genes expressed early in male and female reproductive structures of monterey pine. Plant Physiol. 117:55–62. Muller PY, Janovjak H, Miserez AR, Dobbie Z. 2002. Processing of gene expression data generated by quantitative realtime RT-PCR. Biotechniques 32:1372–1379. Nam J, dePamphilis CW, Ma H, Nei M. 2003. Antiquity and evolution of the MADS-box gene family controlling flower development in plants. Mol Biol Evol. 20:1435–1447. Pelaz S, Ditta GS, Baumann E, Wisman E, Yanofsky MF. 2000. B and C floral organ identity functions require SEPALLATA MADS-box genes. Nature 405:200–203. Pnueli L, Abu-Abeid M, Zamir D, Nacken W, Schwarz-Sommer Z, Lifschitz E. 1991. The MADS box gene family in tomato: Gymnosperm Fleshy Fruits · doi:10.1093/molbev/msr244 temporal expression during floral development, conserved secondary structures and homology with homeotic genes from Anthirrinum and Arabidopsis. Plant J. 1:255–266. Rijpkema AS, Vandenbussche M, Koes R, Heijmans K, Gerats T. 2010. Variations on a theme: changes in the floral ABCs in angiosperms. Semin Cell Dev Biol. 21:100–107. Rijpkema AS, Zethof J, Gerats T, Vandenbussche M. 2009. The petunia AGL6 gene has a SEPALLATA-like function in floral patterning. Plant J. 60:1–9. Saitou N, Nei M. 1987. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol. 4:406–425. Saladié M, Matras AJ, Isaacson T, et al. (15 co-authors). 2007. A reevaluation of the key factors that influence tomato fruit softening and integrity. Plant Physiol. 144:1012–1028. Seymour GB, Ryder CD, Cevik V, Hammond JP, Popovich A, King GJ, Vebralov J, Giovannoni JJ, Manning K. 2011. A SEPALLATA gene is involved in the development and ripening of strawberry (Fragaria ananassa Duch.) fruit, a non-climacteric tissue. J Exp Bot. 62:1179–1188. Soltis DE, Ma H, Frohlich MW, Soltis PS, Albert VA, Oppenheimer DG, Altman NS, dePamphilis C, Leebens-Mack J. 2007. The floral genome: an evolutionary history of gene duplication and shifting patterns of gene expression. Trends Plant Sci. 12:358–367. Tadiello A, Pavanello A, Zanin D, Caporali E, Colombo L, Rotino GL, Trainotti L, Casadoro G. 2009. A PLENA-like gene of peach is involved in carpel formation and subsequent transformation into a fleshy fruit. J Exp Bot. 60:651–661. Taiz L, Zeiger E. 2006. Plant physiology, 4th ed. Sunderland (MA): Sinauer Associates. Tanaka Y, Sasaki N, Ohmiya A. 2008. Biosynthesis of plant pigments: anthocyanins, betalains and carotenoids. Plant J. 54:733–749. Tandre K, Svenson M, Svensson ME, Engström P. 1998. Conservation of gene structure and activity in the regulation of reproductive organ development of conifers and angiosperms. Plant J. 15:615–623. MBE Taylor TN, Taylor EL, Krings M. 2009. Paleobotany. The biology and evolution of fossil plants, 2nd ed. Amsterdam: Elsevier. Theissen G, Becker A, Di Rosa A, Kanno A, Kim JT, Münster T, Winter KU, Saedler H. 2000. A short history of MADS-box genes in plants. Plant Mol Biol. 42:115–149. Theissen G, Melzer R. 2007. Molecular mechanisms underlying origin and diversification of the angiosperm flower. Ann Bot. 100:603–619. Theissen G, Saedler H. 2001. Plant biology: floral quartets. Nature 409:469–471. Vrebalov J, Pan IL, Arroyo AJM, et al. (11 co-authors). 2009. Fleshy fruit expansion and ripening are regulated by the tomato SHATTERPROOF gene TAGL1. Plant Cell 21:3041–3062. Vrebalov J, Ruezinsky D, Padmanabhan V, White R, Medrano D, Drake R, Schuch W, Giovannoni J. 2002. A MADS-box gene necessary for fruit ripening at the tomato ripening-inhibitor (rin) locus. Science 296:343–346. Wang YQ, Melzer R, Theissen G. 2010. Molecular interactions of orthologues of floral homeotic proteins from the gymnosperm Gnetum gnemon provide a clue to the evolutionary origin of ‘‘floral quartets’’. Plant J. 64:177–190. Winter KU, Becker A, Münster T, Kim JT, Saedler H, Theissen G. 1999. MADS-box genes reveal that gnetophytes are more closely related to conifers than to flowering plants. Proc Natl Acad Sci U S A. 96:7342–7347. Zahn LM, Kong H, Leebens-Mack JH, Kim S, Soltis PS, Landherr LL, Soltis DE, Depamphilis CW, Ma H. 2005. The evolution of the SEPALLATA subfamily of MADS-box genes: a preangiosperm origin with multiple duplications throughout angiosperm history. Genetics 169:2209–2223. Zhang P, Tan HTW, Pwee KH, Kumar PP. 2004. Conservation of class C function of floral organ development during 300 million years of evolution from gymnosperms to angiosperms. Plant J. 37:566–577. Zhou ZY. 2009. An overview of fossil Ginkgoales. Palaeoworld 18:1–22. 419