SEARCH

SEARCH BY CITATION

Keywords:

  • evolution;
  • Picea abies;
  • APETALA3;
  • PISTILLATA;
  • yeast two-hybrid system;
  • in situ hybridization;
  • transcription factor

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The Norway spruce MADS-box genes DAL11, DAL12 and DAL13 are phylogenetically related to the angiosperm B-function MADS-box genes: genes that act together with A-function genes in specifying petal identity and with C-function genes in specifying stamen identity to floral organs. In this report we present evidence to suggest that the B-gene function in the specification of identity of the pollen-bearing organs has been conserved between conifers and angiosperms. Expression of DAL11 or DAL12 in transgenic Arabidopsis causes phenotypic changes which partly resemble those caused by ectopic expression of the endogenous B-genes. In similar experiments, flowers of Arabidopsis plants expressing DAL13 showed a different homeotic change in that they formed ectopic anthers in whorls one, two or four. We also demonstrate the capacity of the spruce gene products to form homodimers, and that DAL11 and DAL13 may form heterodimers with each other and with the Arabidopsis B-protein AP3, but not with PI, the second B-gene product in Arabidopsis. In situ hybridization experiments show that the conifer B-like genes are expressed specifically in developing pollen cones, but differ in both temporal and spatial distribution patterns. These results suggest that the B-function in conifers is dual and is separated into a meristem identity and an organ identity function, the latter function possibly being independent of an interaction with the C-function. Thus, even though an ancestral B-function may have acted in combination with C to specify micro- and megasporangia, the B-function has evolved differently in conifers and angiosperms.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Developmental biology during the past decade has identified and characterized molecular mechanisms that are central to the regulation of plant form. Many of these pathways have been found to be regulated at the level of transcription. Our interest is in the evolution of reproductive organs in plants and the possible connection between the evolution of complexity in form of these organs, and the evolution of the regulatory genes which control their development. We approach this problem by a comparative analysis of such regulatory genes in flowering plants (angiosperms) and their distant relatives, the conifers.

Molecular data suggest that extant seed plants (gymnosperms and angiosperms) share a last common ancestor, about 285 million years ago in the Late Carboniferous (Savard et al., 1994). Conifers, which belong to the gymnosperm group of seed plants, have exposed seeds and monosporangiate pollen- and seed-cones formed from separate shoot meristems. Angiosperms have seeds enclosed by a carpel, and also differ from conifers and other gymnosperms in morphology, in that their reproductive organs are organized in flowers – whorls of sterile sepals and petals surrounding whorls of male stamens and female carpels.

A model proposed by Coen and Meyerowitz (1991) suggests that in the angiosperm flower three gene functions, A, B and C, act in combination to specify the identities of the different organs: A alone specifies sepals; A with B, petals; B with C, stamens; and C alone specifies carpels. The model also implies a negative regulation of A by C and vice versa. This model was originally based on analyses of mutants from two eudicot species, Arabidopsis thaliana and Antirrhinum majus (Bowman et al., 1991; Meyerowitz et al., 1991; Schwarz-Sommer et al., 1990; Sommer et al., 1990). In Arabidopsis, genes belonging to the MADS-box gene family of transcription factors carry out all or part of each of these activities. APETALA1 (AP1) is part of the A activity (Gustafson-Brown et al., 1994; Mandel et al., 1992). The B-class genes are APETALA3 (AP3) (Jack et al., 1992) and PISTILLATA (PI) (Goto and Meyerowitz, 1994), and the C function is mediated by the AGAMOUS (AG) gene (Bowman et al., 1991). The ABC model has been further elaborated by Pelaz et al. (2000), who showed that the activities of the B and C organ-identity genes also require the activities of three closely related and functionally redundant MADS-box genes, SEPALLATA1, 2 and 3 (SEP1, 2 and 3). The sep1, 2 and 3 triple mutant forms sepaloid organs in all four whorls of organs in the flower, and it has been suggested that the activity of SEP1/2/3 may be mediated by the formation of ternary complexes with the B and C organ-identity genes (Egea-Cortines and Davies, 2000; Honma and Goto, 2001).

There are about 80 MADS-box genes in the A. thaliana genome (Riechmann and Ratcliffe, 2000). This large family of transcription factors is thought to have evolved through a series of gene duplications followed by functional divergence. Phylogenetic reconstruction of the MADS-box gene family in different angiosperm species shows that the genes group into functional clades (Tandre et al., 1995; Theissen et al., 1996). Thus genes responsible for C-function from different species are more closely related to each other than to other MADS-box genes in the same species. This conclusion is based on mutant data and gene expression analyses from a wide range of angiosperm species, including core eudicot and monocot species (for a review see Theissen et al., 2000). Tandre et al. (1998) and others (e.g. Rutledge et al., 1998) provide evidence to suggest that a conservation of gene function between phylogenetically related regulatory genes might not be limited to the angiosperms, but might be present even among very distantly related plant species. Functional analyses of the conifer C-like MADS-box gene DAL2 (DEFICIENS AGAMOUS LIKE-2) from Picea abies demonstrated that this gene could functionally substitute for the Arabidopsis C-class gene under ectopic expression conditions in transgenic Arabidopsis plants (Tandre et al., 1998). These results, together with the specific expression of DAL2 in sporogenous tissues of both pollen cones (Sundström et al., 1999) and seed cones (Tandre et al., 1998) in spruce, indicate that the function of the C-like genes in the last common ancestor of conifers and angiosperms was to specify the identity of the reproductive organs.

The phenotypic effects observed in Arabidopsis plants expressing DAL2, together with phylogenetic data, suggest that other parts of the regulatory system might also be conserved between angiosperms and conifers, most notably the B-function (Baum, 1998; Tandre et al., 1995). This hypothesis was confirmed by the cloning of the spruce genes DAL11, DAL12 and DAL13 which in a phylogenetic analysis group basal to or within the B-class clade of angiosperm MADS-box genes. These B-like conifer MADS-box genes are specifically expressed in pollen cones, with a cell type-specific expression pattern suggesting that they have roles related to different aspects of pollen cone development (Sundström et al., 1999). These results have led us to propose that the evolution of B- and C-type genes coincided with the evolution of separate microspores (male) and megaspores (female) in the ancestor of seed plants.

In this report we have used transgenic technology in Arabidopsis and the yeast two-hybrid system to show that the conifer B-like genes DAL11, DAL12 and DAL13 are related not only structurally, but also functionally, to the angiosperm B-class genes. We also examine the expression pattern of these genes in very early stages of pollen cone development. Together these data suggest that the spruce genes, even though functionally related to angiosperm organ identity genes, are active not only in the pollen-bearing organs but also during early stages of reproductive shoot development.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

DAL11, DAL12 and DAL13 are differentially expressed in male bud meristems

Early stages of pollen cone development are characterized by the formation of an apical dome in which three distinct zones can be recognized: the apical zone containing the bud meristem; the peripheral zone; and the central pith (Figure 1a). Shortly after apical dome formation, lateral organs (microsporophylls) start to develop in the peripheral zone. The acropetal initiation of microsporophylls eventually results in a terminate strobilus with spirally arranged microsporophylls, each microsporophyll harbouring two microsporangia. mRNA in situ hybridization experiments, using sections of pollen cones at the stages of apical dome formation and initial microsporophyll development, demonstrated that transcripts of DAL11 and DAL12 were present throughout the pollen cone bud, in the apical and the peripheral zones as well as in the central pith (Figure 1b,c). In similar experiments using a DAL13 probe, a hybridization signal was detected only in the peripheral zone in cells localized basal to the emerging microsporophyll primordia (Figure 1d,e). As the first microsporophylls became visible, the DAL13 hybridization signal was confined to the developing lateral organs (data not shown). In control hybridizations using DAL11, DAL12 or DAL13 sense probes, no hybridization signal was detected (data not shown). Experiments performed with sections of pollen cones collected after meristem termination showed differential expression of all three genes (Figure 1f–h; Sundström et al., 1999). In these experiments, high levels of DAL11 transcript were detected in the central pith and the surrounding vascular tissue. A signal was also present in the microsporophylls, but no signal above background was detected in the pollen mother cells of the microsporangia (Figure 1f). At these stages of development, a weak signal of DAL12 could be detected only in the vascular tissues surrounding the central pith (Figure 1g). Expression of DAL13 was detectable in the microsporophylls, the signal being confined to the cell layers surrounding the microsporangia, but no signal was present in the pollen mother cells, in the vascular tissue, or in the central pith.

image

Figure 1. Anatomy and expression of DAL11, DAL12 and DAL13 in developing pollen cones of Norway spruce as detected by in situ hybridization.

(a) Schematic sketches of three pollen cones of different developmental stages. Left, an apical dome with the peripheral zone (pz) from which the microsporophylls are being initiated, showing the apical zone (az) including the meristem, the central pith (p) and a sterile bud scale (bs). Middle, a pollen-cone bud at a stage of development where the first microsporophylls (ms) are emerging. Right, a pollen cone after meristem termination when all microsporophylls have started to develop.

(b–d) Dark-field micrographs of longitudinal sections through pollen-cone buds show the hybridization signal of DAL11(b); DAL12(c); and DAL13(d) antisense probes. Arrows in (d) indicate positions of DAL13 signal. In dark-field micrographs, signal appears as white grains and aggregates of phenolic compounds appear as bright yellowish areas. Gene-specific probes of each gene were used, including the 3′ end of the K-box and the 3′ untranslated region.

(e) Light-field micrograph: high magnification (scale bar = 25 µm) of the boxed area in (d) where the DAL13 signal appears as dark spots.

(f–h) Dark-field micrographs of longitudinal sections of pollen cones after meristem termination showing the hybridization signal of DAL11(f); DAL12(g); and DAL13(h) antisense probes. Scale bar, 100 µm unless otherwise indicated.

Download figure to PowerPoint

DAL11, DAL12 and DAL13 each affect floral development when constitutively expressed in transgenic Arabidopsis

To examine whether the structural similarities between the spruce genes DAL11, DAL12 and DAL13 and the angiosperm class-B MADS-box reflect similarities in function, we constructed transgenic Arabidopsis plants which constitutively express DAL11, DAL12 or DAL13 under control of the cauliflower mosaic virus 35S promoter (CaMV35S), and examined phenotypic deviations from wild-type development in the plants.

Sixteen independent kanamycin-resistant plants harbouring DAL11 were raised. Neither of these plants differed from wild-type in the T1 generation. In the T2 generation, two independent transformant lines differed in floral development from the wild type in that organs in positions where the wild type bore sepals (whorl one) developed petaloid characters. Both organs fully converted to petals (Figure 2f), and organs with a mixture of petal and sepal cell types (Figure 2g) were observed. This phenotypic deviation was not observed in all flowers on each plant, but most commonly in the first six flowers formed on each inflorescence. Complete transformation of sepals into petals was found only in lateral first whorl organs (Figure 2e). Flowers formed at later stages were generally indistinguishable from wild-type flowers. Occasional flowers, however, had unfused carpels and stamens that failed to elongate (data not shown).

image

Figure 2. Transgenic Arabidopsis plants expressing DAL11 or DAL12, demonstrating homeotic transformations of sepals into petals.

Wild-type (wt) plants and plants from transgenic lines D11 : 10 and D12 : 5 expressing DAL11 and DAL12, respectively.

(a) A wild-type Arabidopsis flower. (b–d) Scanning electron micrographs of (b) a wild-type inflorescence; (c) sepal cells; (d) petal cells. (e) Inflorescence derived from DAL11-expressing transgenic plant, arrow indicates position of a transformed sepal; (f) flower of a DAL11 plant with a transformed sepal (arrow); (g) SEM of a petaloid sepal of a DAL11-expressing plant, arrows point to sepal and petal cell types; (h) inflorescence of a DAL12-expressing plant, arrow indicates position of a transformed sepal; (i) flower of a DAL12-expressing plant, arrow points to a transformed sepal; (j) SEM of a petaloid sepal of a DAL12-expressing plant, arrows point to sepal and petal cell types. Scale bars, 20 µm unless otherwise noted.

Download figure to PowerPoint

Fourteen transgenic Arabidopsis lines expressing DAL12 were raised. Three independent lines had phenotypic alterations as compared to the wild type, similar to those seen in the transgenic DAL11 plants. In the DAL12-expressing plants, the floral phenotype with homeotic changes of sepals into petals was found already in the T1 generation, and retained in the T2 generation in one transformant line (Figure 3h–j). No phenotypic deviations from wild-type or kanamycin-resistant control plants grown on selective medium were found in vegetative parts of DAL11- or DAL12-expressing plants.

image

Figure 3. Arabidopsis flowers expressing DAL13.

Flowers of a DAL13-expressing plant from the transgenic line 13 : 7, with anther tissues in (a,b) first whorl organs; (c,d) second whorl organs; (e,f) fourth whorl organs. Arrows indicate position of anther tissue.

Download figure to PowerPoint

Twenty-one transgenic Arabidopsis lines expressing DAL13 were raised. Three lines had altered floral morphology as compared to the wild-type in the T1 generation. Flowers in these plants opened prematurely, and flower organs had aberrant shapes due to ectopic formation of anther tissue on organs in whorl one (Figure 3a,b), whorl two (Figure 3c,d), or whorl four (Figure 3e,f). These plants were sterile due to low or no pollen formation. Cross-pollination with wild-type pollen did not restore seed production in these lines, indicating that the pistil also developed aberrantly. The inflorescence of the three DAL13 lines formed terminal flowers with sepals with carpelloid tissue in the first whorl and bent, and in some cases non-fused carpels in whorl four (data not shown). Vegetative growth of these plants was also affected. The stem length was reduced and leaves were small, pointed and dark green. An additional 14 DAL13 transformant plant lines had a vegetative phenotype with curled cotyledons and curled rosette leaves. This phenotype was retained in the T2 generation. In these plants, DAL13 expression did not cause any deviation in flower morphology.

In each set of transgenic plants, the fraction of plant lines that showed severe phenotypic deviations from wild-type was relatively low. The absence of phenotypic deviations in most transformant lines correlated with a low level of expression of the introduced gene in a majority of these lines, whereas the lines that differed from wild type in floral development had a higher level of expression of DALl11, DAL12 or DAL13, respectively (Figure 4).

image

Figure 4. Transcript levels of DAL11, DAL12 and DAL13 in the transgenic lines.

Transgene-expression levels were measured by RNA gel-blot analysis in representative plants from transgenic lines D11 : 10, DAL12 : 5 and DAL13 : 7, and compared to wild-type Arabidopsis plants and transgenic control plants, c, with no obvious mutant phenotype. Expression levels of (a) DAL11; (b) DAL12; (c) DAL13. Also shown are hybridizations to the 18S ribosomal subunit as a controls for equal loading.

Download figure to PowerPoint

Interactions between spruce B-like proteins and Arabidopsis B-class proteins

In Arabidopsis the AP3 and PI proteins bind DNA as an obligate heterodimer. The transgenic data presented in this report, at least in part, could be explained by interactions between the heterologous spruce proteins and the endogenous B-proteins. To test whether the spruce proteins had the capacity to dimerize with the Arabidopsis B-proteins, we sought evidence for interactions between the Arabidopsis B-proteins AP3 and PI, and the spruce B-like proteins DAL11, DAL12 and DAL13, using a yeast two-hybrid system.

Yeast cells co-transformed with DAL11 inserted into the binding vector, and either DAL11, DAL13 or AP3 inserted into the activation vector, grew on standard dropout media (Figure 5a,c,d), indicating that the DAL11 protein can form homodimers as well as heterodimers with DAL13 and AP3. Cells harbouring DAL11 together with DAL12 or PI did not grow under these conditions (Figure 5b,e). Similarly, yeast cells co-transformed with DAL13 in the binding vector and DAL11, DAL13 or AP3 inserted into the activation vector grew (Figure 5f,h,i), whereas cells harbouring DAL13 together with DAL12 or PI did not (Figure 5g,j). This indicates that both DAL11 and DAL13 are able to heterodimerize with the Arabidopsis protein AP3 but not with PI, and also that DAL11 and DAL13 can form homodimers as well as heterodimers with each other. Yeast cells co-transformed with AP3 or PI inserted into the respective DNA-binding vector; activation vector constructs of DAL11, DAL12, DAL13, AP3 and PI, respectively, confirm these results. Growth was confirmed in cells co-transformed with AP3 and DAL11 or DAL13 (Figure 5l,n) and in cells co-transformed with AP3 or PI in either expression vector (Figure 5o,p). Yeast cells did not grow when co-transformed with PI in the binding vector and PI, DAL11, DAL12 or DAL13 in the activation vector (Figure 5q–t). We note that cells transformed with AP3 in both binding and activation vectors did grow on selective medium (Figure 5k). This result demonstrates the capacity of AP3 to form a homodimer in yeast. It has previously been shown that AP3 binds DNA only in a heterodimeric complex with PI (Riechmann et al., 1996a), but AP3 homodimer formation has been previously reported to occur in immunoprecipitation experiments (Riechmann et al., 1996a). The results described above were confirmed by use of a LacZ assay (data not shown).

image

Figure 5. Interactions between DAL11, DAL12, DAL13, PI and AP3 in yeast as detected by the two-hybrid system.

Serial dilutions of 104−102 Y190 cells co-transformed with binding and activation vectors containing different plasmid combinations as indicated, grown on standard dropout medium minus tryptophan, leucine and histidine.

Download figure to PowerPoint

The DAL12-binding vector construct alone showed some activity in the LacZ assays. To discriminate between this background activity and activation due to protein–protein interactions, a quantitative β-galactosidase assay, using O-nitrophenyl β-d-galactopyranoside (ONPG) as a substrate, was applied (Figure 6). In this experiment yeast cells co-transformed with DAL12 in both vectors had three times higher β-galactosidase activity than yeast co-transformed with DAL12 and DAL11, DAL13, PI or AP3, respectively; and 20 times higher activity than control cells transformed with the DAL12-binding vector only.

image

Figure 6. β-galactosidase assay of interactions between DAL12 inserted into the binding vector and DAL11, DAL12, DAL13, AP3 or PI inserted into the activation vector (n = 3).

Activity calculated using the formula: 1000 × OD420/(OD600 × assay time in min × volume in ml).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The angiosperm B-function, as defined by the ABC model, is dual (Coen and Meyerowitz, 1991). B-class genes act in combination with A-genes to specify petal identity, and with C-genes to specify stamen identity. In Arabidopsis, the B-function is mediated by two MADS-domain proteins, AP3 and PI, which act as heterodimers (Riechmann et al., 1996a) and are dependent on each other for nuclear import (McGonigle et al., 1996) and DNA-binding (Riechmann et al., 1996b). We and others have previously shown that genes that are phylogenetically related to the angiosperm B-genes are also present in gymnosperms and in conifers (Mouradov et al., 1999; Sundström et al., 1999) as well as in Gnetum (Winter et al., 1999). This suggests that the angiosperm B-genes have evolved from an ancestral B-type gene present in the unknown last common ancestor of conifers and angiosperms.

Understanding the character of the regulatory function of this ancestral B-type gene is central to our understanding of the evolution of reproductive organs in plants, as the B-function in angiosperms is in part responsible for directing the development of organs that distinguish angiosperms from gymnosperms – the petals. It is an attractive hypothesis that this aspect of the B-gene function, in specifying the identity of the sterile petals, has evolved specifically in the angiosperm lineage, and thus is derived from an ancestral B-gene function which would have been restricted to specification of the identity of the microspore-bearing organs. This second, reproductive aspect of B-function would thus be conserved between angiosperms and their last common ancestor with gymnosperms. Similarly, the function of B-like genes in the conifers may also have evolved independently, and thus may deviate in different respects from the ancestral state of the B-function.

The results presented here demonstrate that the conifer B-like genes, when active in transgenic Arabidopsis plants, specifically affect development of organs that are related to B-function. These results, in a general sense, are consistent with the hypothesis that the reproductive organ aspect of B-function is conserved between angiosperms and conifers. In fact, DAL11 and DAL12 both cause developmental abnormalities in the first whorl organs that are highly similar to those observed in plants expressing the endogenous B-gene PI under similar conditions (Krizek and Meyerowitz, 1996). In developing wild-type Arabidopsis flowers, AP3 is initially expressed in the first whorl organ primordia, as well as second and third whorls (Jack et al., 1992). Continued expression of AP3 is dependent on an autoregulatory feedback by the AP3/PI heterodimer (Hill et al., 1998; Jack et al., 1992), and expression of AP3 in the first whorl organs is reduced in later stages of flower development. The conversion of first whorl organs into petaloid sepals in plants ectopically expressing PI is consistent with the observed expression pattern of AP3, which allows for the formation of an active AP3/PI heterodimer in the young sepal primordia. Similarly, our data on plants expressing DAL11 and DAL12 indicate that both these genes either have the capacity to functionally interact with downstream targets of AP3/PI as homodimeric complexes or, alternatively, may form heterodimeric complexes with AP3, functionally substituting for the AP3/PI heterodimer. In both cases, a possible target could be the promotor of AP3 and/or PI. In the case of DAL11, both these possibilities are supported by our yeast two-hybrid data which demonstrate the capacity of the DAL11 protein both to form homodimers and to interact with AP3. DAL12 also forms homodimers in yeast, but appears to lack the capacity to form heterodimers with AP3.

We note that expression of DAL11 and DAL12 in transgenic Arabidopsis plants affects the development of sepals and petals, organs that have no obvious counterpart in conifers. This might be a reflection of the conservation of promotor-sequence recognition between the conifer B-like proteins and the angiosperm B-proteins, even though the downstream target genes may have evolved differently in the angiosperm and conifer lineages. Alternatively, this may be a consequence of the ability of spruce proteins to interact with AP3 and not with PI. In either case, these data demonstrate a conservation of B-protein function between angiosperms and conifers.

The tissue distribution of DAL11 and DAL12 expression within the developing cone is consistent with this hypothesis. Both genes are active in all tissues of the developing pollen cone, including the meristem, during the first stages of pollen cone development. After apical meristem termination this general expression pattern is retained in the case of DAL11, which remains active throughout pollen cone development. The activity of DAL12 in later stages of pollen cone development is more elusive and, as indicated by Northern blot experiments, ceases shortly after meristem termination (Sundström et al., 1999).

The third B-related spruce MADS-box gene, DAL13, shows an activity pattern consistent with a role as an organ-identity determinant – activity confined to the developing pollen-bearing organs. This gene, when expressed in transgenic Arabidopsis plants, also causes B-related homeotic alterations of floral organs, but of a character different from those caused by constitutive expression of either DAL11, DAL12 or the Arabidopsis B-genes. The DAL13 effects on floral development in Arabidopsis indicate that the gene has the capacity to transform organs in any of the first, second or fourth whorls to staminoid identity – to specify male organ identity to any organ of the flower. We interpret these results to mean that DAL13 either has the capacity to direct male identity to these organs independently of the Arabidopsis C-function, or that DAL13, in addition to the B-type organ-identity determination function, also has the capacity directly or indirectly to activate the C-gene AG, and to act together with AG to direct stamen identity to the organs. This suggests that DAL13 may act independently of AP3 and PI, a notion which is partly supported by the yeast two-hybrid experiments in which DAL13 formed homodimers and heterodimers with AP3. In either case, the data are consistent with DAL13 being active as a determinant of identity of pollen-bearing organs, both when artificially active in Arabidopsis and in the developing pollen cone of Norway spruce. This conclusion relies on the assumption that the spatial pattern of expression of DAL13 reflects its function, as previously demonstrated for several angiosperm MADS-box genes with roles relating to floral development (Rounsley et al., 1995), as well as the C-type gene DAL2 in Norway spruce (Tandre et al., 1998). Following the same assumption, the expression patterns of DAL11 and DAL12 would indicate that these genes, rather than being determinants of organ identity, may act to specify the identity to the pollen cone meristem. If correct, this hypothesis suggests that the B-related function in spruce is complex, and involves both a specification of the male cone meristem and a separate specification of the identity of the organs produced by this meristem, the two functions being dependent on different MADS-box genes. This apparent division of labour between the conifer B-like MADS-box genes in the reproductive aspect of the B-function may have evolved specifically in the conifer lineage, or alternatively, may represent the organization of the ancestral B-function. As phylogenetic analysis of the MADS-box gene family does not allow a solid conclusion as to whether the gene duplications resulting in DAL11, 12 and 13 occurred before or after the split between conifers and angiosperms (Sundström et al., 1999), we cannot at present distinguish between these two possibilities. A further implication of this hypothesis is that the determination of sex in the truly unisexual conifer reproductive organs, rather than being coupled to the determination of the identity of the specific pollen- and ovule-bearing organs, would occur at the level of the shoot meristem, and therefore would depend on DAL11 and DAL12 rather than DAL13. This function has no obvious homologue in angiosperms, but may have been lost with the evolution of the hermaphrodite flower.

Morphologically, the reproductive organs of conifers and angiosperms differ in fundamental respects. The results presented here, together with previous results from our group (Sundström et al., 1999; Tandre et al., 1995; Tandre et al., 1998) and others (Mouradov et al., 1999; Rutledge et al., 1998) demonstrate that central components of the molecular mechanism which regulates reproductive development are conserved between conifers and angiosperms, and thus among all seed plants. This report further suggests that the specific functions of these components may have evolved differently in the angiosperm and gymnosperm lineages.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Expression analyses

In situ RNA hybridization was performed essentially as described by Jackson (1991). All tissues were fixed in 4% paraformaldehyde and 0.25% glutaraldehyde, and embedded in 100% Histowax (Histolab, Göteborg, Sweden). Sections (7 µm) were hybridized to 35S-labelled RNA probes obtained with a TransProbe T kit (Pharmacia Biotech, Uppsala, Sweden). Templates for in vitro transcription were obtained as described previously (Sundström et al., 1999). The slides were coated with NBT2 emulsion (Eastman Kodak Co., Rochester, NY) and exposed for 2–3 weeks. After development, sections were stained in 20% Gill's hematoxylin [BHD, Poole, UK].

CDNA templates

Spruce cDNAs used as templates for yeast and Arabidopsis experiments were DAL11-1 (Accession no. AF158539), DAL12-1 (Accession no. AF158541) and DAL13-1 (Accession no. AF156543). Arabidopsis RT–PCR fragments corresponding to APETALA3 (AP3) (Accession no. M86357) and PISTILLATA (PI) (Accession no. D30807) were amplified using the Access RT–PCR system (Promega, Madison, WI, USA) and RNA derived from floral buds. Total RNA was prepared as described (Chang et al., 1993). RT–PCR fragments were cloned into TOPO-TA vectors (Invitrogen, Leek, the Netherlands) and fully sequenced on both strands.

Transgenic plants

Full-length cDNA clones of DAL11, DAL12 and DAL13, respectively, were amplified using primer pairs with BamHI sites added, using AmpliTaq Gold (Perkin Elmer Inc., Foster City, NJ). The primers were located at nucleotide (nt) position 65–77 and 812–791 for DAL11, nt 4–28 and 914–894 for DAL12, and nt 40–58 and 912–933 for DAL13. The three PCR fragments were separately cloned into a TOPO-TA vector (Invitrogen) and their sequences were confirmed. The PCR-derived subclones were fused downstream of the 35S promotor in pHTT202 (Elomaa et al., 1993), and the constructs were introduced into Agrobacterium tumefaciens strain C58::pGV2260 by use of triparental mating (Shaw, 1988). Resulting A. tumefaciens strains were used to transform Arabidopsis thaliana ecotype Landsberg erecta (Ler) by an infiltration protocol (Bechtold et al., 1993), and transformants were selected on 50 µg ml−1 kanamycin. The phenotypic properties of T1 and T2 plants grown under long-day conditions at 20–22°C were studied. Tissue samples collected for SEM were fixed in ethanol, formaldehyde and acetic acid (3.7, 5 and 50%, respectively) and dehydrated through an ethanol series. Dehydrated tissue was dried in a PolaronCPD7501 critical-point dryer and coated with gold using a VG Microtech SC510 sputter coater. Specimens were mounted on stubs and examined at 5 kV in a Philips XL30 scanning electron microscope. Brightness and contrast were adjusted using photoshop (Adobe Inc., CA). Transgene transcript levels were measured by RNA gel-blot analysis. Total RNA from representative plants was extracted using TRIzol reagent (Gibco, Frederick, MD, USA) Samples of 5 µg total RNA were size-fractionated on a 1% denaturing formaldehyde/agarose gel and blotted onto Hybond-N membrane (Amersham Pharmacia Biotech, Uppsala, Sweden). Hybridiz ations with gene-specific probes were performed as described previously (Sundström et al., 1999);

Yeast two-hybrid analyses

Protein–protein interactions were assayed with the Matchmaker Two-hybrid System 2 (Clontech, Palo Alto, CA) with cDNAs inserted into the DNA binding-domain vector pACT2 or the transcription-activation vector pAS2-1. Truncated versions of DAL11, DAL12, DAL13, AP3 and PI spanning the linker, K-box and C-terminal, but omitting the DNA-binding MADS-box, were amplified using AmpliTaq Gold (Perkin Elmer). EcoR1/Nco1 sites or BamHI sites were added to the spruce and Arabidopsis primers used in the PCR reactions. All constructs had built-in stop codons after the last residue, and were fully sequenced from both strands. Co-transformations in the yeast strain Y190 of the different expression constructs were done essentially as described in the Clontech Matchmaker manual. Protein–protein interactions were assayed on standard dropout media without tryptophan, leucine and histidine, or by a LacZ activity assay. Quantitative estimations of β-galactosidase activity were carried out with cells harvested in mid-log phase and resuspended in Z buffer, using O-nitrophenyl β-d-galactopyranoside (ONPG) as a substrate.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

We thank Agneta Ottosson for assistance with the transgenic Arabidopsis plants, and Marie Lindersson for DNA sequencing. Annelie Carlsbecker, Johannes Hanson and Eva Sundberg are thanked for critical reading of the manuscript. Teemu Teeri is acknowledged for valuable discussions and technical advice. Vivian Irish is acknowledged for providing growth space and consumables during the finalization of this manuscript. This work was supported by grants from the Swedish Natural Science Research Council, the Swedish Foundation for Strategic Research, and the Wallenberg Foundation Consortium North to the Arabidopsis transgenic plant facility in Uppsala.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References