PLANTARUM PHYSIOLOGIA 121: 149162. 2004 Printed in Denmark all rights reserved
Copyright # Physiologia Plantarum 2004
Functional characterization of SEPALLATA3 and AGAMOUS orthologues in silver birch
Juha Lemmetyinen*, Minna Hassinen, Annakaisa Elo, Ilkka Porali, Kaija Keinonen, Hannu Makela and Tuomas Sopanen
Department of Biology, University of Joensuu, P.O.Box 111, FIN-80101 Joensuu, Finland *Corresponding author, e-mail: [e-mail]
Received 18 August 2003; revised 17 November 2003
The development of flowers is regulated by a complex network of transcriptional activators and repressors, many of which belong to the MADS box gene family. In this study, we describe two MADS box genes of silver birch (Betula pendula Roth), BpMADS1 and BpMADS6, which are similar to SEPALLATA3 and AGAMOUS in Arabidopsis thaliana, respectively. In situ hybridization showed that BpMADS1 was expressed in the inflorescence meristem at a very early stage, but not later. Both genes were expressed in developing carpels, ovules and stamens but not in tepals or scales. Ectopic expression of BpMADS1 in Arabidopsis resulted in a reduced number of floral organs or whole whorls and in petaloid or carpelloid sepals, a phenotype reminiscent of that of fil mutants. 35S::BpMADS6 caused very early flowering in
Arabidopsis. In tobacco, both 35S::BpMADS1 and 35S::BpMADS6 accelerated flowering and, in addition, 35S::BpMADS6 caused changes in sepals and petals. In some transgenic birch plants, 35S::BpMADS1 antisense resulted in the development of both male and female organs in the axil of a single bract and in a change of some inflorescences into vegetative shoots. In two plants, either 35S::BpMADS6 sense or antisense constructs resulted in an increase in the number of tepals and in complete lack of stamens in some male inflorescences. These results suggest that BpMADS1 participates both in inflorescence and in flower formation and BpMADS6 participates in flower formation and that they are functional homologues to SEPALLATA3 and AGAMOUS, respectively.
Introduction
Much of our understanding of basic features of flower development comes from the analysis of mutations affecting flowering or flower structure in Arabidopsis and Antirrhinum (Yanofsky 1995). Many other plants are now being analysed to extend or refine general principles in flower development and to uncover origins of specific modifications. Flowering is induced by both internal and external factors. This is particularly obvious in many cases in which induction is effective only after the plant has changed from the juvenile phase to the mature phase. Many genes involved in the induction have been identified (Hempel et al. 2000), but little is known about the molecular changes during the transition from the juvenile to the mature phase, although this phenomenon is very important in the life cycle of trees (Longman 1976, Poethig 1990). In most plants, induction leads to a change of vegetative meristems into inflorescence meristems, which then give rise to floral meristems. The chain of events leading to the formation of floral meristems and floral organs involves a network of genes mainly encoding transcriptional regulators. The identity of floral organs is partly determined by ABC homeotic genes (Coen and Meyerowitz 1991, Weigel and Meyerowitz 1994). In Arabidopsis, the A function genes AP1 and AP2 specify sepals, AP1 and AP2 together with the B function genes AP3 and PI determine the identity of petals. Stamens are
Abbreviations SSC, saline-sodium citrate; SSPE, saline-sodium phosphate-EDTA; WPM, woody plant medium (Lloyd and Mc Cown 1980) The nucleotide sequence data of BpMADS6 reported here is in the EMBL Nucleotide Sequence Databases under the accession number AJ252071.
Physiol. Plant. 121, 2004
149
determined by the B function genes together with the C function gene AGAMOUS (AG), and the identity of the carpel is determined by AG. Several other organ identity genes have recently been identified, for example, the three SEPALLATA (SEP) genes, SEP1, SEP2 and SEP3 (Pelaz et al. 2000, Honma and Goto 2001). In Arabidopsis the SEP1/2/3 proteins function redundantly and at least one of them is needed for the functioning of the ABC function proteins (Pelaz et al. 2000). Many of the genes regulating flower development belong to the MADS box gene family. A typical MADS domain protein in plants consists of a highly conserved MADS domain, a less conserved I region, a relatively well-conserved K domain and the least conserved C terminal region. MADS proteins are putative transcriptional regulators and they bind to various promoters as homo- or heterodimers (Davies et al. 1996b, Fan et al. 1997) or larger complexes (Egea-Cortines et al. 1999). We have been interested in genes regulating flower development in silver birch (Betula pendula), which is, with its near relatives, a very important broad-leaved forest tree in the temperate region. The development of birch inflorescences starts about a year before the opening of flowers but the anthesis takes place only during the following spring, after overwintering (Atkinson 1992). The structure of the birch flower is different from that of the model plants (e.g. Arabidopsis and Antirrhinum) and has features typical of many windpollinated plants. Male and female flowers are on separate inflorescences (catkins) on the same plant. One catkin contains 200300 small flowers (Dahl and Fredrikson 1996). The flowers are in groups of three in the axils of three-lobed scales. The male flower consists of a reduced perianth with two or three tepals and of two stamens, whereas the female flower consists of a pistil with two stigmata (Atkinson 1992). There is no clearly detectable development of the organs of the opposite sex in either type of flowers, in contrast to, for example, Silene and maize (for a review see Lebel-Hardenack and Grant 1997). The genetic regulation of unisexuality must therefore differ in these plants. We study genes regulating the early stages of inflorescence and flower development in order to use the knowledge obtained in the production of non-flowering birches or in the acceleration of flowering for breeding purposes. The use of non-flowering trees as recipients of economically important transgenes would be one way to prevent the spreading of transgenes into natural tree populations. We have previously isolated the birch homologue of SEP3 (BpMADS1) and demonstrated that transgenic plants containing BARNASE, a cytoplasmic ribonuclease gene from Bacillus amyloliquefaciens, under the control of the BpMADS1 promoter do not produce flowers or produce sterile ones (Lemmetyinen et al. 2001). In another study, we have shown that BpMADS3, 4 and 5 all result in early flowering when ectopically expressed in tobacco (Elo et al. 2001). In addition to practical applications, we hope to gain more information
150
to understand the start of inflorescence formation and the development of unisexuality in birch. The aim of the present study is to characterize BpMADS1 in more detail as well as isolate and characterize the AG homologue from birch, to study their expression during male and female inflorescence development and to investigate their functions in order to evaluate their suitability for bioengineering. AG homologues are also evident choices for the manipulation of fertility.
Materials and methods
Plant material For the isolation of RNA, male and female inflorescences of silver birch (Betula pendula Roth) were collected from wild trees. In addition, shoot tips (0.5 cm), leaves and roots were collected from in vitro grown birches (clone JR1/4). The micropropagated birch plants were grown on MS, MS (Murashige and Skoog 1962) or WPM media as described in Lemmetyinen et al. (1998). Tobacco plants (Nicotiana tabacum L. cv. Petit havana SR1) and transgenic birches were cultivated in an ordinary greenhouse under long-day conditions (day 16 h, night 8 h) or under continuous illumination. Arabidopsis thaliana (Wassilewskija) seedlings were grown under long-day conditions on a mixture of vermiculite and fertilized peat (1 : 1, v/v) in plastic mini-greenhouses in ordinary greenhouse conditions. The cDNA cloning of BpMADS6 The full-length cDNA clone of BpMADS6 was isolated from the same lambda ZAP II library as the cDNA clone of BpMADS1 (Lemmetyinen et al. 2001). An Arabidopsis AG cDNA EcoRI fragment from pCIT565 containing the whole coding region (Yanofsky et al. 1990) was used as a probe for the screening in low-stringency hybridization conditions (hybridization 37 C: 50% formamide, 5 Denhardts solution, 6 SSPE, 0.2% SDS, 100 mg ml1 denatured DNA; final wash 37 C: 1 SSC, 0.1% SDS). Both strands of the cDNA clone were sequenced by using the dideoxy-nucleotide chain termination method (T7Sequencing Kit; Amersham Pharmacia Biotech, Uppsala, Sweden). RNA and DNA gel blot analysis To avoid cross hybridization, the 30 end region of the BpMADS6 cDNA clone (776 bp, nucleotides 3491125 for northern and 406 bp, nucleotides 573979 for Southern hybridization) was used. Hybridizations were performed as described in Lemmetyinen et al. (2001). In situ hybridization In situ hybridization studies were carried out using a modified protocol (Di Laurenzio et al. 1996).
Physiol. Plant. 121, 2004
Digoxigenin-labelled antisense RNA probes were made according to the manufacturers instructions (Roche Biochemicals, Mannheim, Germany). Corresponding sense strands were synthesized as negative controls. The probes were treated by alkaline hydrolysis to obtain 50 200 nt long fragments and quantified according to the manufacturers guide (The DIG system users guide for filter hybridization). The developing birch inflorescences were fixed in PBSbuffered 4% paraformaldehyde, dehydrated in ethanol, cleared in xylene and embedded in paraffin. Consecutive 710 mm tissue sections were attached on sets of Menzel Superfrost Plus microscope slides (Gerhard Menzel GmbH, Braunschweig, Germany) so that comparable sections could be hybridized with different probes. The sections were deparaffinized in xylene and rehydrated in a decreasing ethanol series. After permeabilization in 0.2 M HCl and 10 mg ml1 Proteinase K, the sections were post-fixed in 4% PFA/PBS (paraformaldehyde), acetylated in 0.5% acetic anhydride and dehydrated. The hybridization with labelled probe (0.050.5 mg ml1 kb1 probe complexity) was done in a moistened atmosphere (50%-formamide) for 16 h at 45 C. After hybridization, the sections were washed and treated with RNase A. The signal detection was carried out using the antidigoxigeninalkaline phosphatase conjugate essentially following the manufacturers instructions (Dig Nucleic Acid Detection Kit; Roche Molecular Biochemicals). The slides were then rapidly dehydrated, mounted with Depex, coverslipped and inspected using a Zeiss Axioplan II Imaging microscope (Carl Zeiss GmbH, Jena, Germany). Digital photographs were taken with Zeiss Axiocam HR digital camera. Sequence comparison Sequence analyses were done using mainly the GCG software package (versions 810, Genetics Computer Group, Madison, WI, USA). Sequence alignments were first done by using the PILEUP (GCG) and then the CLUSTAL X programme. Phylogenetic analyses were computed either by using the Neighbour-Joining (CLUSTAL X or PHYLIP), parsimony (PHYLIP) or maximum-likelihood (PUZZLE) algorithm. The nucleotide and amino acid sequences were obtained from the GenBank or EMBL databases except AOM1 (Caporali et al. 2000) and BpMADS7, which were sequenced by our group (P. Jarvinen, unpublished results). Constructs and transformation The cDNA clones containing the entire coding region of BpMADS1 or BpMADS6 either in sense or antisense orientation were ligated to the plasmid pHTT602 under the control of CaMV 35S promoter (Elo et al. 2001). Tobacco, Arabidopsis and birch (the early flowering clones BPM2 and BPM5) were transformed as described earlier (Lemmetyinen et al. 1998, 2001, Keinonen 1999). The phenotypes of tobacco were studied both on the primary transformants (T0) and T1 seedling generations. The phenotypes of 35S::BpMADS1 and 35S::BpMADS6
Physiol. Plant. 121, 2004
transformants of Arabidopsis were studied on T2 and T1 seedlings, respectively. The phenotypes of birches were studied on primary transformants (T0). The phenotypes of transformants in birches were compared with the phenotypes of non-transformed plants. In addition, they were compared to the phenotypes of plants containing a GUS construct (J. Lemmetyinen, K. Keinonen, T. Sopanen, unpublished results). Microscopy For scanning electron microscopy, plant tissues were fixed in the FAA fixative (50% ethanol, 5% acetic acid, 2% formaldehyde), critical point dried and coated with gold. Specimens were examined and photographed with the JEOL JSM-35CF scanning electron microscope (JEOL Ltd, Tokyo, Japan). For light microscopy, plant tissues were fixed in PBS-buffered 4% paraformaldehyde (as for in situ hybridization) and stained with toluidine blue.
Results
BpMADS6 is similar to AG In order to isolate the birch AG homologue, we screened a birch cDNA library using Arabidopsis AG as a heterologous probe. Several cDNAs, apparently derived from the same gene, were isolated and the gene was named BpMADS6. The longest cDNA clone of BpMADS6 was 1154 bp long plus a poly (A) tail. It had a 112-bp long UTR 50 end, a 315bp long UTR 30 end and a poly (A) tail. The putative protein of 242 amino acids encoded by BpMADS6 was a typical MADS domain protein of the AG group (Riechmann and Meyerowitz 1997) containing 16 amino acids in front of the MADS domain (Fig. 1). BpMADS6 was homologous to AG (73% identity and 81% similarity). A genomic DNA blot analysis with four different restriction enzymes and a probe containing the less conserved 30 end of the cDNA clone revealed only one hybridizing band suggesting that, in birch there were no other genes that were very similar to BpMADS6 (data not shown). Phylogeny Phylogenetic analyses with several algorithms were made (neighbour-joining, parsimony or maximum-likelihood) using various regions (MADS, MIK, K-box or the whole coding region) of the nucleotide or putative amino acid sequences of BpMADS1 and BpMADS6 and genes similar to them. Various analyses resulted in a similar tree topology but none of them could resolve the grouping of all the genes analysed (data not shown). The trees in Fig. 2a and B were obtained using neighbour-joining with the entire putative amino acid sequences. Although the alignment of the entire sequences was difficult and possibly not optimal, it gave results rather similar to those obtained using the K-domain, but with higher bootstrap values. Several different analyses showed that BpMADS1 clearly belonged to the SEP3 cluster of the AP1/SEP3
151
MADS
BpMADS6 CaMADS1 AG SHP1 BpMADS6 CaMADS1 AG SHP1 BpMADS6 CaMADS1 AG SHP1 BpMADS6 CaMADS1 AG SHP1 BpMADS6 CaMADS1 AG SHP1 ~ ~ H ~ C C C C Q Q Q Q H H H Q Q Q Q Q ~ ~ F ~ K K K K Q Q Q Q N N N H D D D D ~ ~ L ~ R R R R E E E E N N D N Q Q Q Q ~ ~ Q ~ R R R R A A S A N N N N I M T P ~ ~ L ~ N N N N A A A S Q Q Q M P A A P ~ ~ L ~ G G G G K K K K I F I Y L L L L ~ ~ Q ~ L L L L L L L L L L L L Q Q Q Q ~ ~ I ~ L L L L R R R R R R R R L L L L ~ ~ S ~ K K K K G G Q R A A A A V V V V ~ ~ Y ~ K K K K Q Q Q Q K K K K ~ ~ F ~ A A A A I I I I I I I I ~ ~ P ~ Y Y Y Y R R I R A A A A ~ ~ E ~ E E E E S S S D E E E E ~ ~ N ~ L L L L V V I I N N N G ~ ~ H ~ S S S S Q Q Q Q E E E A ~ ~ F ~ V V V V N D N N R R R R ~ ~ P ~ L L L L S S S S . . . L ~ ~ K ~ C C C C N N N N . . . N ~ ~ K ~ D D D D R R R R . . . P ~ ~ N ~ A A A A H H Q H N N N D ~ ~ K ~ E E E E L M L I Q Q N Q ~ ~ T ~ I I V V L L M V Q Q P Q ~ ~ F ~ A A A A G G G G N N S E ~ ~ P ~ L L L L E E E E L L I S ~ ~ F ~ I I I V A A T S N N S S ~ ~ V ~ V V V I L L I L V V L V ~ ~ L ~ F F F F S S G G M M M I ~ ~ L ~ S S S S E E S S P P P Q ~ ~ P ~ S S S T L L M L G G G G ~ ~ P ~ R R R R N N S N G G G T ~ ~ T ~ G G G G F F P F G G S T ~ ~ A ~ R R R R K K K K N N N V ~ ~ I ~ L L L L E E E E Y Y Y Y M M T M Y Y Y Y L L L L E E E E E E A E E E E E K K R K . . Q S F F Y . Y Y Y Y N N N N L L L G Q Q Q E A A S A L L L L M M M V N N S G N N N N E E E E . . P . Q Q E G N N N N I K G G . . P S . . L S S S . . K N R R . . P S S S G S S S S S L L L L . . Q H M M G H V V V V E E E E . . T D S S D D K K K R K K R K Q Q Q Q V V S A T T G G G G S G S S S S S S S E T T T T I I I I Q Q Q Q P P P S I I I I N N T S S S P H Q Q L S E E E E K R R R Y F F Y R R R K R R R R I I I V D D D . K K K K Y Y Y Y R R R R S S S N L L S L K K K K S S S S R R R R G G G G K K K K K K K K T N N N R R R R A A A A K K K K Y Y Y Y G G G G C C I C N N N N F F F I K K K K A A S S E E E E Q Q Q P I I I I E D D D L L L L V V V V E E E E S S N A L L L L D D A N I I I I S S S V F L F V A A A L K K K K N N N N A A S A L L L L R R R R S S T P E E E E Q Q Q E I I I I G G G P I I I I P P P P E E E E S S S S E E D E . . N . N N N N V V V V Y Y Y Y N N N N T T T T S S A T M M M M H H H Q T T T T E E E E Q H Q Q H H H Q N N N N A A I A K K K K Y Y Y F R R R R N N N N R R R R P P S S Q Q Q Q T T A T E E E E . . S G V V V V Q Q Q Q A V V M . . A . T T T T F F Y Y E D D E . . G . F F F F Y Y Y Y L L L L R R R . 36 36 70 36 106 106 139 105 176 176 209 175 233 233 276 239
K-domain
* 242 * 242 * 285 * 248
Fig. 1. Prettybox presentation of BpMADS6 with AG, CaMADS1 and SHP1.
clade (Fig. 2A). In addition to the Hamamelidae, this cluster now has representatives in the Dilleniidae (SEP3 and SaMADSD), Asteridae (FBP2, TM5, NsMADS3, DefH72 and DefH200), Rosidae (peaMTF1 and EGM1) and possibly in monocot species, too (e.g. ADOMADS). The closely related SEP1 and SEP2 with their apparent counterparts in other species formed a separate cluster. In many analyses, AGL3 was also associated with this cluster although the bootstrap values were always low. BpMADS6 was very similar (94% identity at the putative amino acid level) to the Corylus avellana gene CaMADS1 (Rigola et al. 1998) (Fig. 1). Because Corylus is phylogenetically closely related to Betula (e.g. Chen et al. 1999b), this was not surprising. Although BpMADS6 grouped in most analyses together with AG, PLENA, etc. (Fig. 2B), rather than SHATTERPROOF 1 (SHP1, former AGL1) and SHATTERPROOF 2 (SHP2, former AGL5), the whole cluster was quite unstable and, as an example, when only the K-domain (at the amino acid level) was analysed by the neighbour-joining method, BpMADS6 and CaMADS1 grouped together with SHP1 and SHP2. Most putative amino acid sequences in both AG and SHP1/2 groups have an N-terminal sequence in front of the MADS domain. Putative amino acid sequences of genes in the AGL11 group as well as the monocot sequences weakly resembling AGL11 (Fig. 2B) do not seem to have this N-terminal peptide. Putative AG-like sequences from gymnosperms do not have it either. Genes from both conifers and monocots formed separate clusters and had no clear tendency to join any special cluster in the whole AG/SHP1/SHP2/AGL11 clade although some monocot genes grouped together with AGL11 (Fig. 2B). BpMADS6 is expressed in stamens and carpels To determine where and when BpMADS6 was expressed, RNA gel blot analysis was first used (Fig. 3). Transcripts of BpMADS6 were detected in both male and female inflor152
escences, but not in roots, leaves or shoot tips. At the early phases of development, the levels of expression were quite low. The first very weak expression of BpMADS6 was detected in male inflorescences collected in summer (in inflorescences 7 mm long, not visible in the figure). In female inflorescences, a weak expression (not visible in the figure) could be detected in autumn (4 mm long). When the expression of BpMADS6 was first detected, small initials of stamens or pistils were visible in the axils of the already more developed scales. In both types of inflorescences, the expression of BpMADS6 continued until flowering and in female inflorescences the expression continued during seed development. At the inflorescence level, the expression pattern was quite similar to that of BpMADS1 (Lemmetyinen et al. 2001). However, BpMADS1 seemed to be expressed slightly earlier than BpMADS6 and the first strong expression was detected earlier with BpMADS1 than with BpMADS6. The localization of the expression of BpMADS6 in inflorescences was also studied with in situ hybridization (Fig. 4AF). The first signs of BpMADS6 expression were detected both in the epidermal cell layer (L1) and in layers below it (L2 and L3) when initials of the scales were formed. In female inflorescences, BpMADS6 was detected only in a clearly defined area at the distal part of the pistil (Fig. 4DF). This might suggest that the lower part of the pistil actually represents the peduncle. Later, the expression was detected in the integuments (data not shown) as well as in the embryo sac (Fig. 4F). In male inflorescences, the expression of BpMADS6 was detected even in cells from which stamen formation apparently starts later (data not shown) and the expression was maintained later during anther and filament development (Fig. 4B and C). The expression of BpMADS6 was not detected in tepals, scales or axis. BpMADS1 is expressed in inflorescence meristems and later in stamens and carpels BpMADS1 was active during a short period in the newly formed inflorescence meristem before flower
Physiol. Plant. 121, 2004
A
AP1 (Arabidopsis thaliana) PreMADS1 (Pinus resinosa) 1000 PRMADS2 (Pinus radiata ) 1000 PrMADS3 (Pinus radiata) DAL1 (Picea abies) ZAG5 (Zea mays ) 1000 981 OsMADS6 (Oryza sativa) 1000 TaMADS12 (Triticum aestivum ) 882 MdMADS11 (Malus x domestica) AGL6 (Arabidopsis thaliana) GRCD1 (Gerbera hybrida ) SbMADS1 (Sorghum bicolor) 1000 OsMADS8 (Oryza sativa) 604 M9 (Hordeum vulgare) M79 (Oryza sativa) 1000 1000 OsMADS45 (Oryza sativa) 999 OsMADS7 (Oryza sativa) 725 DOMADS1 (Dendrobium ) ADOMADS (Aranda deborah) 673 MpMADS13 (Magnolia praecocissima ) AoM1 (Asparagus officinalis) 593 1000 SEP3 (Arabidopsis thaliana ) 705 SaMADS D (Sinapis alba) 632 MTF1 (Pisum sativum) BpMADS1 (Betula pendula) VvMADS4 (Vitis vinifera ) 707 1000 NsMADS3 (Nicotiana sylvestris ) 743 705 FBP2 (Petunia hybrida ) 984 543 DEFH200 (Antirrhinum majus ) DEFH72 (Antirrhinum majus ) 852 EgM1 (Eucalyptus grandis ) TM5 (Lycopersicon esculentum ) DOMADS3 (Dendrobium) 1000 OsMADS1 (Oryza sativa) OsMADS5 (Oryza sativa) CMB1 (Dianthus caryophyllus ) DEFH49 (Antirrhinum majus) 1000 1000 FBP5 (Petunia hybrida ) TM29 (Lycopersicon esculentum ) 837 SEP1 (Arabidopsis thaliana ) 1000 SEP2 (Arabidopsis thaliana ) VvMADS2 (Vitis vinifera ) 895 CAGL2 (Cucumis sativus) 668 MAGL4 (Populus tremuloides ) MdMADS9 (Malus domestica) 1000 574 984 MdMADS1 (Malus domenstica) MdMADS8 (Malus domestica) AGL3 (Arabidopsis thaliana ) 1000 FBP9 (Petunia hybrida ) 999 NtMADS4 (Nicotiana tabacum) 998 CanMADS1 (Capsicum annuum ) 788 FBP23 (Petunia hybrida ) EgM3 (Eucalyptus grandis ) 960 1000 MdMADS3 (Malus domestica) MdMADS7 (Malus domestica) 0.1 1000
B
GgM3 (Gnetum gnemon) AF023615 (Pinus radiata) 520 1000 PreMADS (Pinus resinosa) DAL2 (Picea abies) 534 SMADS42A (Picea mariana) OsMADS13 (Oryza sativa) 1000 1000 ZmM1 (Zea mays ) ZAG2 (Zea mays ) 1000 FBP11 (Petunia hybrida ) FBP7 (Petunia hybrida ) 941 BpMADS7 (Betula pendula) 993 AGL11 (Arabidopsis thaliana) CAG1 (Cucumis sativus) VvMADS5 (Vitis vinifera) 648 MdMADS10 (Malus x domestica ) 997 HAG1 (Hyacinthus orientalis) PeMADS1 (Phalaenopsis equestris) 532 999 HvAG2 (Hordeum vulgare) ZAG1 (Zea mays ) 987 HvAG1 (Hordeum vulgare) 996 MZEAG (Zea mays ) 813 OsMADS3 (Oryza sativa) SHP2 (Arabidopsis thaliana) 1000 1000 SHP1 (Arabidopsis thaliana ) BnSHP1 (Brassica napus) CsM1 (Cucumis sativus) 944 992 MASAKOD1 (Rosa rugosa) MdMADS14 (Malus x domestica) CUS1 (Cucumis sativus ) FBP6 (Petunia hybrida ) 882 PLE (Antirrhinum majus ) 933 LAG (Liquidambar styraciflua ) VvMADS1 (Vitis vinifera ) 1000 BpMADS6 (Betula pendula ) CaMADS1 (Corylus avellana) CUM1 (Cucumis sativus) MdMADS15 (Malus x domestica) 985 RAG (Rosa hybrida) 1000 StAG1 (Fragaria x ananassa) AG (Arabidopsis thaliana ) 1000 BAG1 (Brassica napus ) GAG2 (Panax ginseng ) 770 1000 GAGA1 (Gerbera hybrida) GAGA2 (Gerbera hybrida ) FAR (Antirrhinum majus ) 860 TAG1 (Lycopersicon esculentum ) 998 NAG1 (Nicotiana tabacum ) 795 PMADS3 (Petunia x hybrida) 822 SLM1 (Silene latifolia ) RAP1 (Rumex acetosa ) GhAP3 (Gossypium hirsutum) 996 PtAG1 (Populus balsamifera ) 786 PtAG2 (Populus balsamifera ) 0.1 1000 AP1 (Arabidosis thaliana)
760
*
849
*
*
Fig. 2. Neighbour-joining analysis of putative MADS proteins related to SEP3 and AG. The alignments used and the bootstrap-tree were made by the ClustalX program and whole putative amino acid sequences were used in the analysis. (A) SEP3 and (B) AG groups. The sequences from gymnosperms are indicated by vertical dashed lines, monocot sequences by vertical continuous lines and birch sequences by *. The bootstrap values are indicated when over 500.
Male inflorescences
Female inflorescences
BpMAD36
28S
Fig. 3. RNA gel blot analysis of BpMADS6. Buds denote terminal buds possibly containing very early stages of male inflorescences. Female inflorescences bearing seeds are collected about 40 days after anthesis. For other details see Materials and methods.
Physiol. Plant. 121, 2004
formation started (Fig. 5A and C), but at later stages no expression could be detected in the inflorescence meristem. Later, the expression of BpMADS1 was localized to the same tissues (pistils and stamens, but not tepals, scales or axis) (Fig. 5CH) as that of BpMADS6. However, slight differences existed and later the tissues with the highest activities were different with BpMADS1 and BpMADS6. In general, BpMADS1 activity was strongest in the proximal and central parts of the female flower whereas BpMADS6 activity was strongest in the distal parts (Figs 4DF and 5G and H). Interestingly, in the inflorescence or floral meristem the first BpMADS1 expression, differently from the expression of BpMADS6, was mostly localized to L2 (Fig. 5B) and equally or less to L3, L1 being generally but not always without a signal. The increase in the expression levels of BpMADS1 and BpMADS6 during inflorescence development seen in RNA gel blot analysis (Lemmetyinen et al. 2001, this study) is mainly due
153
Fig. 4. BpMADS6 RNA localization in birch inflorescence by in situ hybridization. (A) sense probe (BF) antisense probe. In parentheses the day of the collection of inflorescences. A, axis; C, carpel; Im, inflorescence meristem; L, leaf; O, ovule; Sc, scale; St, stamen; T, tepal. (A) Female inflorescence (10.05) control. (B) Male inflorescence (04.07). BpMADS6 RNA is detected in stamens. (C) Male inflorescence (02.08). BpMADS6 RNA is detected in stamens. (D) Female inflorescence (14.09) BpMADS6 RNA is detected in carpels. (E) Female inflorescence (17.05). The distal parts of carpels are hybridized most strongly. (F) Female inflorescence (24.05). BpMADS6 RNA becomes even more restricted to the distal parts of carpels and, on the other hand, to ovules.
to the increase of the relative proportion of stamens or carpels in the inflorescence.
The ectopic expression of BpMADS1 results in a mixed floral organ identity in Arabidopsis In Arabidopsis, the expression of 35S::BpMADS1 resulted in phenotypes with altered flower structures in 15 of the 46 kanamycin-resistant lines obtained and in reduced plant size or viability in 24 lines. In the flowers, the main phenotype was a mixed organ identity and a reduction in the numbers of organs and/or whorls. In some flowers of these plants, petals and sepals were either reduced in size or were completely missing (Fig. 6H and I). There was also a continuum from the normal green sepals to white petaloid sepals (Fig. 6M). Sometimes there was stigmatic tissue at the tip of sepals (Fig. 6K). In the same inflorescence, there were flowers with both a normal and an altered phenotype and sometimes clusters of flowers with some organs missing (Fig. 6IK and M). Some stamens were thin and had small anthers (Fig. 6JM), sometimes these filament-like structures ended with narrow petaloid blades or had stigmatic papillae or even ovules at their end (Fig. 6J). Some floral organs looked like intermediates between stamens and carpels (Fig. 6I). Carpels in the fourth whorl remained unaffected in Arabidopsis expressing BpMADS1. We have also observed some aberrant flowers in control plants but not in such young plants.
Physiol. Plant. 121, 2004
The ectopic expression of BpMADS1 accelerates flowering in tobacco We studied the possible functions of BpMADS1 and BpMADS6 by expressing the corresponding birch cDNAs constitutively with the 35S promoter in tobacco, Arabidopsis and birch. In tobacco, 14 independent transformants were obtained with 35S::BpMADS1. The presence of transgene in the lines studied in more detail (11 lines) was confirmed with PCR and RNA gel blot analysis using leaf tissues (data not shown). Some of the transgenic tobacco lines flowered earlier than controls, the two earliest ones had about 13 leaves before the first flowers, whereas wild-type (WT) plants formed flowers after about 25 leaves (Table 1). The flowers of the 35S::BpMADS1 tobacco lines looked normal and after self-pollination all lines produced viable seeds. Early flowering caused by the ectopic expression of BpMADS1 was also observed in the offspring confirming that this phenotype was not a result of a non-genotypic effect caused by the culturing in in vitro conditions.
154
Fig. 5. BpMADS1 RNA localization in birch inflorescence by in situ hybridization. In parentheses the day of the collection of inflorescences. A, axis; C, carpel; Im, inflorescence meristem; L, leaf; O, ovule; Sc, scale; St, stamen; T, tepal. (A) A terminal bud (01.05). In the apical and older inflorescence meristem BpMADS1 RNA is detected only weakly, but in the axillary inflorescence meristem the expression is stronger. (B) A terminal bud (01.05). The expression of BpMADS1 can be seen in a place, where the flower development apparently starts. (C) A terminal bud (01.06). A developing male inflorescence with some developing flowers, which contain BpMADS1 RNA. Another inflorescence with much BpMADS1 RNA in inflorescence meristem is starting to develop at the base of inflorescence. (D) In male inflorescence (04.07) BpMADS1 RNA is localized in developing stamens but not in tepals. (E) Male inflorescence (30.07). The strongest signal of BpMADS1 RNA located in anthers but some RNA is also in the middle of filaments and in the receptacle. (F) Female inflorescence (22.08) prepared out of a bud. BpMADS1 RNA is localized to the flowers in the axis of bracts. (G) Female inflorescence (17.05). BpMADS1 RNA is localized to the carpel. (H) Female inflorescence (24.05). In carpel BpMADS1 RNA becomes more localized to ovules, especially to integuments.
BpMADS1 has an effect on flower formation in birch In birch, we analysed total of 26 transgenic lines with 35S::BpMADS1 sense and 56 lines with 35S::BpMADS1 antisense constructs. In 14 lines (in four sense and in 10 antisense lines) we detected two distinct, altered phenotypes. Because the phenotypes were similar both in sense
Physiol. Plant. 121, 2004
and antisense lines, we suppose that the effect of the sense construct was in most cases due to co-suppression, but this could not be verified because of limited phenotypical inflorescence material available. In the first phenotype (obtained with one sense and two antisense plants), floral organs of both sexes were formed in the axil of
155
Table 1. Flowering of wild-type (WT) and transgenic tobacco plants. The data are presented as the mean SD. Transgenic lines contain individuals which are either heterozygous or homozygous to the transferred genes. 1Number of individual plants per genotype; 2number of leaves before flowers; 3height of plants were measured when flower buds appeared; 4days of was flowering defined as the time from seed sowing to appearance of the flower buds. Data accompanied by different letter differ significantly (P , 0.05) (Multiple range analysis of LSD). Transgenic line WT 35S::BpMADS1, 35S::BpMADS1, 35S::BpMADS6, 35S::BpMADS6, 35S::BpMADS6, line line line line line 1 2 3 4 5 n1 Leaves2 10 8 5 8 9 8 24.5 1.6a 12.8 1.3c 13.4 4.2c 16.9 1.2b 17.6 1.0b 17.1 1.9b Height3 44.7 4.1a 21.0 3.7c 19.4 9.3c 27.1 5.1b 27.1 4.7b 29.8 5.5b Days to flowering4 101.6 4.1a 84.3 4.6c 91.8 11.1b 92.3 4.2b 93.4 3.7b 94.5 2.6b
more than half of the transgenic lines, the transferred gene also had some effects on flower structure, especially on sepals and petals. In typical weak phenotypes, the size of the tube and limbs was reduced and the edges of petal tips were curved outwards (Fig. 6A). In stronger phenotypes, corollas were split deeply (Fig. 6B and C) or petals had additional small anther-like structures (Fig. 6B). Sometimes filament-like structures were formed on the outer surface of petals (Fig. 6D). In some transgenic lines, sepals formed a tube resembling the carpel. The tips of sepals also had dark green stigma-like structures (Fig. 6E). In several transgenic lines there were liquidsecreting nectaries as an orange zone at the outside of the base of flowers (Fig. 6F). In some lines, seed production was much reduced. The ectopic expression of BpMADS6 causes extremely early flowering phenotypes in Arabidopsis With the 35S::BpMADS6 construct, we obtained 19 kanamycin-resistant Arabidopsis lines. All these lines flowered very early without producing any normal vegetative leaves and their entire shoots had an altered structure (Fig. 6N). Even the cotyledons were narrow and malformed and the four rosette leaves formed thereafter were small, grooved and had pale malformed tips. The length of the inflorescence stem was greatly reduced (reaching only 0.5 cm in extreme cases) and an apparent loss of inflorescence indeterminacy resulted in a cluster of flowers with reduced pedicels. The flowers lacked petals (Fig. 6O) and some sepals had stigmatic tissue at the tip. In a few cases, ovule-like structures were observed on the margins of the sepals (Fig. 6P). The sense and antisense constructs of BpMADS6 in birch In birch, we analysed a total of 47 lines containing sense constructs and 15 lines containing antisense constructs of BpMADS6, but a clear phenotype was only detected in two lines (one sense and one antisense) (Fig. 7F and 7G and J). In these lines, several male inflorescences without stamens were formed (Fig. 7J). In extreme cases, similarly to the phenotype of BpMADS1 antisense transformants, some male inflorescences looked normal in the beginning but then started to produce leaves and the flowers at the distal end of the inflorescence looked composite-like (Fig. 7G). In less extreme cases, single flowers began to develop in a more inflorescence-like manner containing scales that were similar to those in male inflorescences. On the other hand, the main scales were more like those in female inflorescences (Fig. 7J). The number of tepals was also increased.
some scales (Fig. 7E). This phenotype is also seen on WT early flowering birch clones (Stern 1961), but in the BPM2 and BPM5 clones we use (Lemmetyinen et al. 1998) this phenomenon is rare and is restricted to those lower parts of inflorescences which stay for a very long time inside the bud before emerging. In the second phenotype (obtained with three sense and eight antisense plants), the development of the terminal male inflorescences looked normal until they were about 8 mm long but then the inflorescences started to produce leaves (Fig. 7C and D). These leaves did not originate from the scales (which are modified bracts) because the scales were still present in these flowers. Noticeably, these scales looked more tepal-like and the number of actual tepals was greater than normally. Stamens were not formed but leaves were formed in the middle of the flower (Fig. 7I). Later these new inflorescences continued to develop like normal branches or shoots although the internodes were much shorter. The later development of other inflorescences in the same plants was usually normal, which made it difficult to prove that the effect obtained was the result of co-suppression in the case of sense transformants. The effects of BpMADS1 sense or antisense constructs on female inflorescences could not be studied because these inflorescences require overwintering to emerge. The ectopic expression of BpMADS6 has an effect on sepal and petal formation in tobacco With the 35S::BpMADS6 construct, we obtained 23 independent kanamycin-resistant tobacco lines. The presence of transgene in some lines with a clear phenotype was confirmed by means of PCR and RNA gel blot analysis which showed that the transgene was expressed in the leaves. The plants were fertile when self-pollinated, and the properties caused by the transgene were also observed in the offspring (F1 plants). Like 35S::BpMADS1, 35S::BpMADS6 also caused early flowering in tobacco, the earliest lines flowered after having produced about 17 leaves (Table 1). In
156
Discussion
The development and the structure of inflorescences and flowers in birch are quite different from those in, for example, Arabidopsis and Antirrhinum. So it is interesting
Physiol. Plant. 121, 2004
Fig. 6. Effect of ectopic expression of BpMADS1 and BpMADS6 on tobacco (AE) and Arabidopsis (GP). (A) Left, wild type (WT); right, 35S::BpMADS6, weak phenotype. (B) 35S::BpMADS6, strong phenotype, petals have additional stamen-like structures. (C) 35S::BpMADS6, moderate phenotype, deeply split corolla. (D) 35S::BpMADS6, moderate phenotype, additional filament-like structures starting from the petal tube and ending as a petal blade. Sepals removed. (E) 35S::BpMADS6, strong phenotype, sepals forming stigma-like structures (arrow). (F) 35S::BpMADS6, moderate phenotype, liquid secreting nectaries as an orange zone (arrow). (G) WT Arabidopsis. (H) 35S::BpMADS1. Flower consisting of carpels and stamens. (I) 35S::BpMADS1. Carpelloid stamen indicated by an arrow. No sepals or petals. (J) 35S::BpMADS1. Ovule formation (arrow) on the tip of sepaloid-carpelloid stamen, filamentous structures. (K) 35S::BpMADS1. Stigmatic tissue at the tip of sepals (arrow), filamentous structures. (L) 35S::BpMADS1. Stamens fused with sepals (arrow), narrow petals. (M) 35S::BpMADS1. A cluster of flowers, petaloid sepals, filamentous structures. (N) 35S::BpMADS6. Inflorescence formation. (O) 35S::BpMADS6. Flower without petals. White bar on the left 100 mm. (P) 35S::BpMADS6. Ovule-like structure on the margins of sepals (arrow).White bar on the left 100 mm.
to study differences in the genes regulating the development of flowers and, thus, which genes cause differences in morphology between species. On the other hand, by expressing birch genes heterologously in such distantly related species as Arabidopsis and tobacco, we can also get some information on the universality of the functional capability of genes between species.
Physiol. Plant. 121, 2004
Functions of BpMADS1 BpMADS1 seems to resemble other characterized SEP3like genes both structurally and functionally. It seems to function both in inflorescence and floral meristems. The first expression of SEP3 has been detected in flower primordia at the late stage 2 (Mandel and Yanofsky 1998), but its near homologue SaMADSD in Sinapis
157
Fig. 7. Effect of the ectopic expression of BpMADS1 and BpMADS6 in birch. (A) Male and female inflorescences of wild-type birch. (B) Male inflorescences of wild type early flowering birch clone BPM5. (C, D) 35S::BpMADS1-antisense. Apical inflorescences of transgenic birch lines. (E) 35S::BpMADS1-sense. A scale with stamens (St) and carpels (C). (F) 35S::BpMADS6-antisense. Apical inflorescence of transgenic birch lines. (G) 35S::BpMADS6-antisense. The flowers at the distal end of the inflorescence. (H) Male inflorescences of wild-type early flowering birch clone BPM5. Sc, scale; T, tepal; St, stamen. (I) 35S::BpMADS1-antisense. (J) 35S::BpMADS6-antisense scale-like tepal indicated by an arrow.
has been detected even earlier, already at stage 1, in a region just below the inflorescence meristem (Bonhomme et al. 1997). AOM1 in Asparagus officinalis and DOMADS1 in Dendrobium are also expressed in inflorescence meristems (Caporali et al. 2000, Yu and Goh 2000). The expression of BpMADS1 in the inflorescence meristem also explains why the BpMADS1::BARNASE construct totally prevented inflorescence formation in Arabidopsis and tobacco (Lemmetyinen et al. 2001) as well as in birch (Lemmetyinen et al. 2004). The functioning of the isolated promoter in the same tissues where the gene is active suggests that there are no essential regulatory elements outside the promoter region used. The only effect of the ectopic expression of BpMADS1 in tobacco seemed to be the acceleration of flowering
158
time, resembling the effect of the rice genes, OsMADS7 and 8, in tobacco (Kang et al. 1997). The ectopic expression of SEP3 in Arabidopsis has also resulted in an early flowering phenotype (Honma and Goto 2001, Pelaz et al. 2001). Both the localization of the early expression of BpMADS1 and some other genes in the SEP3 group in the inflorescence meristem and the acceleration of flowering caused by over-expression are consistent with the suggestion that the genes of the SEP3 group might have a role in the developing inflorescence meristem (Bonhomme et al. 1997). Because the tested SEP3-like genes have only relatively slight effects on flowering time, they may not alone play any major role in the initiation of flowering. Recently, it has been shown that SEP3 expressed ectopically together with AP1 resulted in
Physiol. Plant. 121, 2004
an earlier flowering phenotype in Arabidopsis than either of them alone (Pelaz et al. 2001). The ectopic expression of birch genes in tobacco, either BpMADS3 (similar to AP1), BpMADS4 or BpMADS5 (similar to FUL), had a much stronger effect than BpMADS1 on the flowering time and had no effect on flower structure (Elo et al. 2001). In birch, BpMADS3, BpMADS4 and BpMADS5 were also expressed similarly to BpMADS1 from a very early phase of inflorescence development until the developing seed stage (Elo et al. 2001, Lemmetyinen et al. 2001). However, BpMADS3 and especially BpMADS4 were also expressed in vegetative tissues and the expression of BpMADS4 and BpMADS5 was strong even in the very early phases of inflorescence development. Later we have shown that these two genes are expressed in the inflorescence meristem (M. Hassinen and T. Sopanen, unpublished results). The expression of BpMADS1 in putative floral meristem cells even before any anatomical signs of flower development suggests that BpMADS1 might also have a role in the initial development of the flower. This type of role has already been suggested for SaMADSD (Bonhomme et al. 1997). The appearance of the clusters of flowers in Arabidopsis ectopically expressing BpMADS1 (Fig. 6M) is quite consistent with this. During flower development the expression of SEP3 and many genes in the SEP3 group is mainly restricted in the three inner whorls (see, e.g. Theissen et al. 1996) but in some species some gene activity was also detected in sepals, e.g. SAMADS D in Sinapis (Bonhomme et al. 1997). The absence of BpMADS1 expression in the tepals suggests that in birch male flowers tepals are homologous to sepals rather than to petals. This suggestion is further supported by the finding that B function genes in birch have no expression in tepals either (M. Hassinen, S. Vepsalainen, T. Sopanen, unpublished results). In spite of the apparent absence of petals in birch, the tested promoter region of BpMADS1 (BpMADS1::GUS) was active in the petals of transgenic tobacco (Lemmetyinen et al. 2001). It seems therefore, that despite the simplified flower structure and absence of petals in birch, BpMADS1 still contains regulatory elements needed for its activation in petals. The ectopic expression of BpMADS1 in Arabidopsis also resulted in changes in the number and identity of floral organs. This indicates that BpMADS1 may have a role in the organization of flower meristems and in the determination of the number of whorls as was also suggested by Pnueli et al. (1994) for TM5. Whereas transgenic tomatoes carrying an antisense gene construct of TM5 had additional whorls (Pnueli et al. 1994), the transgenic Arabidopsis carrying the 35S::BpMADS1 sense construct, in accordance, had missing whorls. Some inflorescences in transgenic birches containing the antisense (or sense) construct of BpMADS1 had organs of both sexes, which means that they have activated the development of the fourth whorl that does not normally exist in male inflorescences. This suggests that BpMADS1 might be involved in the development of the
Physiol. Plant. 121, 2004
unisexuality of birch flowers. Interestingly, the ectopic expression of SEP3 in Arabidopsis does not lead to homeotic changes in floral organs (Honma and Goto 2001) suggesting some differences in regulation between birch and Arabidopsis. Several Arabidopsis mutants with a decreased number of organs within a whorl have been reported, e.g. wus (Laux et al. 1996) and fil (previously called FL54 mutant) (Chen et al. 1999a, Sawa et al. 1999a). The phenotype resulting from the ectopic expression of BpMADS1 in Arabidopsis had several properties similar to those of the fil mutant (e.g. aberrant number, shape and arrangement of floral organs as well as filament-like floral organs) although filament-like flowers were not detected. Therefore, there may be some connection between FIL and SEP3. Another argument for this is that the expression patterns of FIL and SEP3 are at least partly mutually complementary (see Fig. 5E in Sawa et al. 1999b and Fig. 3EH in Mandel and Yanofsky 1998). The prevention of the stamen but not of tepal development in birch by the suppression of BpMADS1 suggests that BpMADS1 is necessary for the development of stamens but not for the development of tepals. This further suggests that tepals evolved from sepals. When the expressions of TM5 (Pnueli et al. 1994), TM29 (Ampomah-Dwamena et al. 2002) or FBP2 (Angenent et al. 1994) have been reduced using the antisense technique or co-suppression, the three inner whorls of the flowers of tomato or petunia were affected and their characteristics were changed towards the characteristics of the outer whorl. In Arabidopsis expressing ectopically BpMADS1, the characteristics of the three outer whorls were changed towards those of the inner whorl (Fig. 6IM). In some flowers, sepals and petals were completely missing or there were carpelloid sepals and staminoid petals (Fig. 6K and L). However, the sep3 single mutant or SEP3 antisense plants show only a subtle phenotype in which petals are partially transformed into sepals (Pelaz et al. 2001). Because the phenotypes obtained with antisense were more extreme than those observed for sep3 single mutants, it was concluded that part of the effects in SEP3 antisense plants was perhaps due to the suppression of SEP1 and/or SEP2 as well. Possibly in other plant species, such as petunia, tomato and birch, distributions of the tasks of SEP homologues with other genes are organized slightly differently. Because the phenotype obtained either with TM5, FBP2 or BpMADS1 constructs were stronger than those obtained with SEP single mutants, it seems that the SEP homologues in petunia, tomato and birch are less redundant than in Arabidopsis. However, recently it was shown that the phenotype obtained by co-suppression of FBP2 is caused by the down-regulation of both FBP2 and FBP5, which seem to be functionally redundant (Ferrario et al. 2003). Because the SEP gene products, or the products of their homologues in other species, form complexes with several ABC gene products (Honma and Goto 2001), and they are necessary for the activity of B and C genes (Pelaz et al. 2000), it is not surprising that suppression of
159
genes in the SEP group results in phenotypic changes similar to those observed in some ABC mutants. Functions of BpMADS6 Several different results indicate that BpMADS6 is the birch orthologue to AG. (1) Phylogenetically, BpMADS6 is in most analyses, although with low bootstrap values, associated with AG, rather than with SHP1 or SHP2. (2) BpMADS6 is expressed in both stamens and carpels similarly to AG (Bowman et al. 1991), whereas SHP1 and SHP2, up-regulated by AG and down-regulated by FUL, are only expressed in carpels (Ma et al. 1991, Savidge et al. 1995, Flanagan et al. 1996, Ferrandiz et al. 2000). (3) The ectopic expression of BpMADS6 in tobacco had an effect on the first and second whorls similarly to the phenotype that was obtained when BAG1, a Brassica gene closely related to AG, was ectopically expressed in tobacco (Mandel et al. 1992). (4) Similarly to ag mutants (Bowman et al. 1989), when BpMADS6 was down-regulated with an antisense construct in birch, stamens did not develop and the number of tepals (or scales) had increased. The over-expression of BpMADS6 also affected the nectaries in tobacco. In tobacco, nectaries are normally at the base of the ovary (Mandel et al. 1992), but the over-expression of BpMADS6 caused a development of nectaries at the base of the outside of sepals, similarly to the effect of the AG homologues BAG1 and PLENA (Mandel et al. 1992, Davies et al. 1996a). This further supports the grouping of BpMADS6 together with other AG orthologues. Interestingly, in Arabidopsis, the development of nectaries, which are located in the third whorl does not depend on the ABC genes (Baum et al. 2001), indicating that the regulation of nectary formation differs in Arabidopsis and tobacco. Ectopic expression of BpMADS6 in Arabidopsis caused a strong phenotype resembling the phenotype caused by the ectopic expression of AG or SAG1, e.g. (Mizukami and Ma 1992, Rutledge et al. 1998) or the phenotype of emf1 and emf2 mutants (Sung et al. 1992, Yang et al. 1995, Chen et al. 1997). It has been shown by Chen et al. (1997) that EMF1 negatively regulates the expression of AG and therefore in emf1 mutant plants AG is expressed very early in shoot development. Therefore, it is not surprising that the ectopic expression of BpMADS6 leads to a phenotype resembling that of EMF1 mutant plants. BpMADS1 and BpMADS6 are expressed at the same time in birch inflorescences, from an early stage to a stage when seeds are developing but BpMADS1 is somewhat earlier. BpMADS1 might be necessary for BpMADS6 expression as seems to be the case with the corresponding genes (FBP2 and FBP6, respectively) in petunia (Angenent et al. 1994). In tomato, TM5 antisense does not have any effect on TAG1 expression (Pnueli et al. 1994) but this may be due to incomplete suppression of TM5. In the sep1 sep2 sep3 triple mutant, AG is activated normally (Pelaz et al. 2000), which indi160
cates that in Arabidopsis none of the SEP genes is needed for AG expression. On the other hand, SEP3 forms quaternary complexes with ABC proteins and thus the flower-specific expression of SEP3 is needed in the action of ABC genes in the flower development (Honma and Goto 2001). Practical applications From the practical point of view, it is interesting that the ectopic expression of BpMADS1 and BpMADS6 can accelerate the flowering of tobacco moderately. In addition, BpMADS1 did not affect flower morphology or fertility. Therefore, it might be feasible to use these BpMADS genes in molecular breeding in cases where a moderate acceleration of flowering is desirable. In birch, it seems that these genes do not have any essential role in the timing of flowering. We have already successfully applied the promoter of BpMADS1 to prevent flowering (Lemmetyinen et al. 2001, 2004, Lemmetyinen and Sopanen 2003). In our experiments BARNASE, regulated by the promoter of BpMADS1, either completely prevented the formation of inflorescences in tobacco and Arabidopsis or in weaker phenotypes led to formation of sterile flowers lacking stamens and carpels. We have not yet tested the promoter of BpMADS6 with BARNASE. It is known that AG contains regulatory elements preventing vegetative expression in its very long intron (Sieburth and Meyerowitz 1997), and this is likely to be true for its orthologue BpMADS6, too. If so, it is most probable that using 50 upstream sequences of BpMADS6 with BARNASE would lead to defects in the vegetative tissues. Both the antisense construct of BpMADS1 and that of BpMADS6 occasionally led to sterility when tested in our early flowering birch clone. However, the presence of fertile inflorescences on the same transgenic plants indicates inefficient attenuation of the target gene. In situ hybridization with the BpMADS6 antisense probe showed that, in lines ectopically expressing BpMADS6 antisense, there was still some natural BpMADS6 expression (M. Hassinen, J. Lemmetyinen, T. Sopanen, unpublished results). Efficient attenuation of BpMADS1 or BpMADS6 might be obtained, for example, by using an RNAi construct. The complete down-regulation of these genes would most likely prevent stamen and carpel formation completely. However, the silencing of BpMADS6 would probably not prevent inflorescence formation, but the developing flowers would be sterile and, on the other hand, vegetative defects should be very unlikely.
Acknowledgements We thank Leonard Medrano for providing us the AGAMOUS cDNA fragment pCIT565 and Teemu Teeri for providing the vector pHTT602 used for sense constructs made in his laboratory at the Institute of Biotechnology, University of Helsinki (A.E.). We are grateful to Liisa Tikka for her help in the initial phases of this study, Marja-Leena Turunen for DNA isolation, Mika Lannenpaa for Agrobacterium transformation of Arabidopsis and Riitta Pietarinen for her skilful technical assistance. We are also grateful to Yrjo Helariutta for his help in the setup of in situ hybridization. This research work was funded by the Academy of
Physiol. Plant. 121, 2004
Finland (J.L., T.S.), Tekes (as part of the Finnish Biodiversity Programme, FIBRE) (J.L.), the Faculty of Science, the University of Joensuu (J.L., A.E.), the Ministry of Education (A.E.), and the Graduate School of Biology and Biotechnology of Forest Trees, later the Graduate School of Forest Sciences.
References
Ampomah-Dwamena C, Morris BA, Sutherland P, Veit B, Yao J (2002) Down-regulation of TM29, a tomato SEPALLATA homolog, causes parthenocarpic fruit development and floral reversion. Plant Physiol 130: 605617 Angenent GC, Franken J, Busscher M, Weiss D, van Tunen AJ (1994) Co-suppression of the petunia homeotic gene fbp2 affects the identity of the generative meristem. Plant J 5: 3344 Atkinson MD (1992) Betula pendula Roth (B. verrucosa Ehrh.) and B. pubescens Ehrh. J Ecol 80: 837870 Baum SF, Eshed Y, Bowman JL (2001) The Arabidopsis nectary is an ABC-independent floral structure. Development 128: 46574667 Bonhomme F, Sommer H, Bernier G, Jacqmard A (1997) Characterization of SaMADS D from Sinapis alba suggests a dual function of the gene: in inflorescence development and floral organogenesis. Plant Mol Biol 34: 573582 Bowman JL, Drews GN, Meyerowitz EM (1991) Expression of the Arabidopsis floral homeotic gene AGAMOUS is restricted to specific cell types late in flower development. Plant Cell 3: 749758 Bowman JL, Smyth DR, Meyerowitz EM (1989) Genes directing flower development in Arabidopsis. Plant Cell 1: 3752 Caporali E, Spada A, Losa A, Marziani G (2000) The MADS box gene AOM1 is expressed in reproductive meristems and flowers of the dioecious species Asparagus officinalis. Sex Plant Reprod 13: 151156 Chen L, Cheng J-C, Castle L, Sung ZR (1997) EMF genes regulate Arabidopsis inflorescence development. Plant Cell 9: 20112024 Chen Q, Atkinson A, Otsuga D, Christensen T, Reynolds L, Drews GN (1999a) The Arabidopsis FILAMENTOUS FLOWER gene is required for flower formation. Development 126: 27152726 Chen Z-D, Manchester SR, Sun H-Y (1999b) Phylogeny and evolution of the Betulaceae as inferred from DNA sequences, morphology, and paleobotany. Am J Bot 86: 11681181 Coen ES, Meyerowitz EM (1991) The war of the whorls: genetic interactions controlling flower development. Nature 353: 3137 Dahl AE, Fredrikson M (1996) The timetable for development of maternal tissues sets the stage for male genomic selection in Betula pendula (Betulaceae). Am J Bot 83: 895902 Davies B, Di Rosa A, Eneva T, Saedler H, Sommer H (1996a) Alteration of tobacco floral organ identity by expression of combinations of Antirrhinum MADS-box genes. Plant J 10: 663677 Davies B, Egea-Cortines M, de Andrade Silva E, Saedler H, Sommer H (1996b) Multiple interactions amongst floral homeotic MADS box proteins. EMBO J 15: 43304343 Di Laurenzio L, Wysocka-Diller J, Malamy JE, Pysh L, Helariutta Y, Freshour G, Hahn MG, Feldmann KA, Benfey PN (1996) The SCARECROW gene regulates an asymmetric cell division that is essential for generating the radial organization of the Arabidopsis root. Cell 86: 423433 Egea-Cortines M, Saedler H, Sommer H (1999) Ternary complex formation between the MADS-box proteins SQUAMOSA, DEFICIENS and GLOBOSA is involved in the control of floral architecture in Antirrhinum majus. EMBO J 18: 53705379 Elo A, Lemmetyinen J, Turunen M-L, Tikka L, Sopanen T (2001) Three MADS-box genes similar to APETALA1 and FRUITFULL from silver birch (Betula pendula). Physiol Plant 112: 95103 Fan H-Y, 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. Plant J 12: 9991010 Ferrandiz C, Liljegren SJ, Yanofsky MF (2000) Negative regulation of the SHATTERPROOF genes by FRUITFULL during Arabidopsis fruit development. Science 289: 436438 Ferrario S, Immink RGH, Shchennikova A, Busscher-Lange J, Angenent GC (2003) The MADS box gene FBP2 is required for SEPALLATA function in petunia. Plant Cell 15: 914925
Physiol. Plant. 121, 2004
Flanagan CA, Hu Y, Ma H (1996) Specific expression of the AGL1 MADS-box gene suggests regulatory functions in Arabidopsis gynoecium and ovule development. Plant J 10: 343353 Hempel FD, Welch DR, Feldman LJ (2000) Floral induction and determination: where is flowering controlled? Trends Plant Sci 5: 1721 Honma T, Goto K (2001) Complexes of MADS-box proteins are sufficient to convert leaves into floral organs. Nature 409: 525529 Kang H-G, Jang S, Chung J-E, Cho Y-G, An G (1997) Characterization of two rice MADS box genes that control flowering time. Mol Cells 7: 559566 Keinonen K (1999) Towards genetic manipulation of silver birch (Betula pendula) and Scots pine (Pinus sylvestris). PhD Dissertation. University of Joensuu, Joensuu, Finland. Laux T, Mayer KFX, Berger J, Jurgens G (1996) The WUSCHEL gene is required for shoot and floral meristem integrity in Arabidopsis. Development 122: 8796 Lebel-Hardenack S, Grant SR (1997) Genetics of sex determination in flowering plants. Trends Plant Sci 2: 130136 Lemmetyinen J, Keinonen K, Sopanen T (2003) Prevention of the flowering of a tree, silver birch. Mol Breed, in press Lemmetyinen J, Keinonen-Mettala K, Lannenpaa M, von Weissenberg K, Sopanen T (1998) Activity of the CaMV 35S promoter in various parts of transgenic early flowering birch clones. Plant Cell Rep 18: 243248 Lemmetyinen J, Pennanen T, Lannenpaa M, Sopanen T (2001) Prevention of flower formation in dicotyledons. Mol Breed 7: 341350 Lemmetyinen J, Sopanen T (2003) Modification of flowering in forest trees. In: Kumar S, Fladung M (eds) Molecular Genetics and Breeding of Forest Trees. The Haworth Press, Inc., Binghamton, New York Lloyd GB, Mc Cown BH (1980) Commercially feasible micropropagation of mountain laurel (Kalmia latifolia) by use of shoottip culture. Proc Inter Plant Propagators Soc 30: 421437 Longman KA (1976) Some experimental approaches to the problem of phase change in forest trees. Acta Hortic 56: 8190 Ma H, Yanofsky MF, Meyerowitz EM (1991) AGL1-AGL6, an Arabidopsis gene family with similarity to floral homeotic and transcription factor genes. Gene Dev 5: 484495 Mandel MA, Yanofsky MF (1998) The Arabidopsis AGL9 MADS box gene is expressed in young flower primordia. Sex Plant Reprod 11: 2228 Mandel MA, Bowman JL, Kempin SA, Ma H, Meyerowitz EM, Yanofsky MF (1992) Manipulation of flower structure in transgenic tobacco. Cell 71: 133143 Mizukami Y, Ma H (1992) Ectopic expression of the floral homeotic gene AGAMOUS in transgenic Arabidopsis plants alters floral organ identity. Cell 71: 119131 Murashige T, Skoog F (1962) A revised medium for rapid growth and bio assays with tobacco tissue cultures. Physiol Plant 15: 473497 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: 200203 Pelaz S, Gustafson-Brown C, Kohalmi SE, Crosby WL, Yanofsky MF (2001) APETALA1 and SEPALLATA3 interact to promote flower development. Plant J 26: 385394 Pnueli L, Hareven D, Broday L, Hurwitz C, Lifschitz E (1994) The TM5 MADS Box gene mediates organ differentiation in the three inner whorls of tomato flowers. Plant Cell 6: 175186 Poethig RS (1990) Phase change and the regulation of shoot morphogenesis in plants. Science 250: 923930 Riechmann JL, Meyerowitz EM (1997) MADS domain proteins in plant development. Biol Chem 378: 10791101 ` ` Rigola DP, Pe ME, Fabrizio CM, Me G, Sari-Gorla M (1998) CaMADS1, a MADS box gene expressed in the carpel of hazelnut. Plant Mol Biol 38: 11471160 Rutledge R, Regan S, Nicolas O, Fobert P, Cote C, Bosnich W, Kauffeldt C, Sunohara G, Seguin A, Stewart D (1998) Characterization of an AGAMOUS homologue from the conifer black spruce (Picea mariana) that produces floral homeotic conversions when expressed in Arabidopsis. Plant J 15: 625634 Savidge B, Rounsley SD, Yanofsky MF (1995) Temporal relationship between the transcription of two Arabidopsis MADS Box genes and the floral organ identity genes. Plant Cell 7: 721733
161
Sawa S, Ito T, Shimura Y, Okada K (1999a) FILAMENTOUS FLOWER controls the formation and development of Arabidopsis inflorescences and floral meristems. Plant Cell 11: 6986 Sawa S, Watanabe K, Goto K, Liu Y-G, Shibata D, Kanaya E, Morita EH, Okada K (1999b) FILAMENTOUS FLOWER, a meristem and organ identity gene of Arabidopsis, encodes a protein with a zinc finger and HMG-related domains. Gene Dev 13: 10791088 Sieburth LE, Meyerowitz EM (1997) Molecular dissection of the AGAMOUS control region shows that cis elements for spatial regulation are located intragenically. Plant Cell 9: 355365 Stern K (1961) Uber den Erfolg einer uber drei Generationen gefuhrten Auslese auf fruhes Bluhen bei Betula verrucosa. Silvae Genet 10: 4851. Sung ZR, Belachew A, Shunong B, Bertrand-Garcia R (1992) EMF, an Arabidopsis gene required for vegetative shoot development. Science 258: 16451647 Theissen G, Kim JT, Saedler H (1996) Classification and phylogeny of the MADS-box multigene family suggest defined roles of
MADS-box gene subfamilies in the morphological evolution of eukaryotes. J Mol Evol 43: 484516 Weigel D, Meyerowitz EM (1994) The ABCs of floral homeotic genes. Cell 78: 203209 Yang C-H, Chen L-J, Sung ZR (1995) Genetic regulation of shoot development in Arabidopsis. Role of the EMF genes. Dev Biol 169: 421435 Yanofsky MF (1995) Floral meristems to floral organs: genes controlling early events in Arabidopsis flower development. Annu Rev Plant Phys Plant Mol Biol 46: 167188 Yanofsky MF, Ma H, Bowman JL, Drews GN, Feldmann KA, Meyerowitz EM (1990) The protein encoded by the Arabidopsis homeotic gene agamous resembles transcription factors. Nature 346: 3539 Yu H, Goh CJ (2000) Identification and characterization of three orchid MADS-box genes of the AP1/AGL9 subfamily during floral transition. Plant Physiol 123: 13251336
Edited by J. Almeida
162
Physiol. Plant. 121, 2004