Evolution of a Family of Plant Genes with Regulatory Functions Picea abies

Evolution of a Family of Plant Genes with Regulatory Functions Picea abies
Comprehensive Summaries of Uppsala Dissertations
from the Faculty of Science and Technology 573
_____________________________
_____________________________
Evolution of a Family of Plant
Genes with Regulatory Functions
in Development; Studies on Picea
abies and Lycopodium annotinum
BY
MATS SVENSSON
ACTA UNIVERSITATIS UPSALIENSIS
UPPSALA 2000
Dissertation for the Degree of Doctor of Philosophy in Physiological Botany
presented at Uppsala University in 2000
Abstract
Svensson, M., 2000. Evolution of a Family of Plant Genes with Regulatory Functions
in Development; Studies on Picea abies and Lycopodium annotinum. Acta
Universitatis Upsaliensis. Comprehensive Summaries of Uppsala Dissertations from
the Faculty of Science and Technology 573. 45 pp. Uppsala. ISBN 91-554-4826-7.
This work is focused on the molecular genetic basis for morphological change in
evolution. Genes belonging to the MADS-box gene family, which includes members,
that determine angiosperm floral organ identity, were isolated and characterised from
two non-angiosperm plants; Norway spruce (Picea abies) and the club moss
(Lycopodium annotinum).
The exon/intron organisation of the isolated genes was determined, and its
significance as an independent test of the position of a gene within the gene family tree
evaluated.
Norway spruce genes that are closely related to the angiosperm floral organ
identity genes were identified. One Norway spruce gene, DAL2, is an ortholog to
angiosperm C-class MADS-box genes that specify stamen and carpel identity. The
expression of DAL2 in male and female cones suggests that orthologous genes in
conifers and angiosperms determine the identities of pollen- and seed-bearing
structures. Constitutive expression of DAL2 in the angiosperm Arabidopsis resulted in
homeotic conversions very similar to those resulting from constitutive expression of
the Arabidopsis C-class gene.
Angiosperm B-class MADS-box genes determine petal and stamen identity. The
isolated Norway spruce B-class orthologs: DAL11, DAL12, and DAL13 are expressed
in the developing male cones exclusively, suggesting a conserved function of B-class
related genes in the determination of pollen forming organs among seed plants.
No orthologs to the floral organ identity genes could be isolated from the club
moss, suggesting that the origin of these gene classes may be coupled to the origin of
the pollen and the seed.
The club moss MADS-box genes, LAMB2, LAMB4 and LAMB6, conform
structurally to plant type MADS-box genes, whereas LAMB1 is divergent in details.
The genes LAMB3 and LAMB5 encode shorter proteins.
LAMB1 expression is restricted to reproductive structures, whereas LAMB2,
LAMB4, LAMB5 and LAMB6 are broadly expressed. The implications from these
expression patterns on the ancestral function of plant type MADS-box genes are
discussed.
Mats Svensson, Department of Physiological Botany, EBC, Uppsala University,
SE-752 36, Uppsala, Sweden
© Mats Svensson 2000
ISSN 1104-232X
ISBN 91-554-4826-7
Printed in Sweden by University Printers, Uppsala 2000
…Lycopodium? That’s a good one!
E. MEYEROWITZ
3
This thesis is based on the following papers, which will be referred
to in the text by their Roman numerals:
I.
Tandre K., Svenson M., Svensson M. E., Engström P. (1998)
Conservation of gene structure and activity in the regulation of
reproductive organ development of conifers and angiosperms.
Plant J. 15:615-623.
II.
Sundström J., Carlsbecker A., Svensson M. E., Svenson M.,
Johanson U., Theißen G., Engström P. (1999) MADS-box
genes active in developing pollen cones of Norway spruce (Picea
abies) are homologous to the B-class floral homeotic genes in
angiosperms. Dev. Genet. 25:253-266.
III.
Svensson M. E., Johannesson H., Engström P. (2000) The
LAMB1 gene from the clubmoss, Lycopodium annotinum, is a
divergent MADS-box gene, expressed specifically in sporogenic
structures. Gene 253:31-43.
IV.
Svensson M. E., Engström P. Closely related MADS-box genes
in the club moss, Lycopodium annotinum, show broad expression
patterns and are structurally similar to typical seed plant MADSbox genes. (in manuscript)
4
Table of contents
Abbreviations
Homeotic genes and the evolution of plant developmnet
6
7
1.i
Interaction of floral homeotic genes, the ABC-model
8
1.ii
Floral homeotic genes belonging to the MADS-box gene family
9
1.iii The expression patterns of MADS-box genes reflect their function
9
1.iv Conservation of function between orthologous genes
9
1.v
The MADS-box genes constitute a large gene family in eukaryotes
10
1.vi The MADS-domain
10
1.vii Subfamilies of MADS-box genes
11
1.iix Functional domains of MIKC-genes
11
1.ix Exon/intron organisation of MIKC-genes (I, II, IV)
12
1.x
Divergent MADS-box genes with a MIKC exon/intron pattern (IV)
13
1.xi The role of MIKC-genes in angiosperm development
15
1.xii Phylogeny of function in plant MADS-box genes
16
2.
MADS-box genes in gymnosperms (I, II)
16
2.i
DAL2 – a Norway spruce C-class gene (I, II)
17
2.ii
DAL2 can act as a C-class gene in an angiosperm context (I)
17
2.iii DAL2 and the functional diversification of the AG-clade
18
2.iv Putative B-class genes from Norway spruce (II)
19
2.v
Exon/intron organisation of DAL11, DAL12, and DAL13 (II)
20
2.vi Expression of Norway spruce B-class genes in male cones (II)
20
2.vii. B- and C-genes in Gnetum
21
3.
MADS-box genes in non seed plants, e.g. Lycopsids (III, IV) 21
3.i
Lycopsids constitute the sister group to all other extant vascular plants
22
3.ii
LAMB1 is a divergent MIKC-gene from Lycopodium annotinum (III)
22
3.iii Typical MIKC-genes from L. annotinum (IV)
24
3.iv The phylogenetic position of the L. annotinum MADS-box genes (III,IV) 25
3.v
Exon/intron organisation of LAMB2, LAMB3, LAMB4, and LAMB6 (IV) 25
3.vi A leucine-zipper in the K- and C-region of LAMB4 (IV)
26
3.vii Expression patterns of L. annotinum MIKC-genes (III, IV)
26
3.iix. MADS-box genes in ferns
27
3.ix Evolutionary considerations (III,IV)
27
4.
Homeotic genes and organ homology
28
4.i
Homology as a neo-Darwinian paradigm
30
4.ii
Homeotic genes and homology
30
4.iii Homeotic gene expression does not define morphological homology
30
4.iv Homologies between reproductive structures of seed plants
32
4.v
B- and C genes and seed plant reproductive organ homologies (I, II)
33
4.vi ABC-genes and floral organ homology
34
4.vii Lycopsid microphyll homology
35
5.
Concluding remarks and suggestions
35
Acknowledgements
37
References
38
1.
5
Abbreviations
aa
amino acids
bp
base pairs
mya
million years ago
RACE
rapid amplification of cDNA ends
RT-PCR
reverse transcriptase polymerase chain reaction
6
1. Homeotic genes and the evolution of plant development
When the discipline of molecular evolution emerged, two important notions were
made. Firstly, there appeared to be no necessary connection between the number of
genes, and the morphological complexity of an organism (Ohno, 1972). A recent
comparison between the number of genes in the nematode, Caenorhabditis elegans
and the fruit fly, Drosophila melanogaster genomes, ca 20000 vs. 12000
(Meyerowitz, 1999), is in accordance with this notion. Secondly, evolutionary
change in coding sequences appeared to a large extent to be neutral with respect to
protein function (Kimura, 1977), and hence have no effect on morphology. These
findings suggest that change in the cis-regulatory elements of genes, rather than in
their coding sequences would be the predominant mechanism for morphological
evolution (Doebley and Lukens, 1998).
Development can be modeled as composed of modules showing varying
degree of connection /dissociation relative each other (e.g. Raff, 1996). That
developmental modules are (more or less) dissociated from each other is of great
evolutionary importance, because mutations may consequently affect one module
without having effects on the entire organism. For naturalists such as Cuvier, the
belief that the organism represents an integrated whole, and that no individual detail
could be changed without affecting this whole, served as the major argument against
the evolutionary theory (see Gould, 1992). At the morphological level a
developmental module may be represented by for example cell types, tissues or
structures such as a meristem or different organ primordia in plants, or tagma,
segments, and developing organs in arthropods. Evolutionary change of a module
may involve quantitative changes in the interaction between components within the
module, or the recruitment of submodules into the module, probably by the changes
in cis-regulatory elements.
Large phenotypic changes are less likely to confer a selective advantage,
than small (e.g. Fisher, 1930). Doebley and Lukens (1998) argue that a transcription
factor, unlike many other proteins, generally is confined to a particular
developmental module, rather than having global developmental roles. This suggests
that mutations directly affecting transcription factor activity would be less
pleiotropic than many other types of mutations. Accordingly, Doebley and Lukens
demonstrate that mutations in transcriptional regulators often have less pleiotropic
effects than mutations in signal transduction genes. As a synthesis of the presented
arguments, Doebley and Lukens propose that the cis-regulatory elements of
transcription factors are likely to be most important in the evolution of novel
phenotypes.
One type of transcription factors that is especially interesting from the
perspective of morphological evolution determines organ specific gene expression.
Mutations in such genes may result in homeotic changes, i.e. that one structure in an
organism becomes replaced by another structure. One of the most exciting biological
discoveries in the second half of the 20th century biology was that of the Drosophila
melanogaster homeotic mutants ultrabithorax and antennapedia affecting genes that
indeed encode transcription factors. Both these genes encode proteins with a
conserved DNA binding motif, the homeodomain. Since, a large number of
transcription factors with key regulatory roles possessing this motif have been
isolated (for a review on these discoveries, see Gehring, 1994). Hence, within D.
7
melanogaster, and in other species, a family of homeobox genes exists, and different
members of this gene family have different developmental roles.
Members of a gene family within a species are derived from a common
ancestor through gene duplication events, and are referred to as paralogous genes
(homologous by gene duplication, Fitch, 1970). The same paralogs in two different
species are referred to as orthologous genes (homologous by speciation, Fitch,
1970).
A duplication event of a homeotic gene may lead to functional divergence
between the resulting paralogs, a process that might be coupled to the formation of
two dissociated developmental modules. For example, repeated morphological
features, may end up being controlled by different developmental regulators, and
hence evolve independently from each other. Lewis already in 1978 proposed a
similar scheme for the evolution of the insect body from a centipede-like ancestor,
by the diversification of homeotic genes. Although wrong in detail (see Akam et al.,
1994), Lewis general idea is still attractive. Thus, even if changes in cis-regulatory
elements may very well be the main cause of phenotypic change, gene duplication
and diversification do occur, and the diversification of the homeotic transcription
factor encoding genes, may be connected to the origin of new morphological
complexity.
Reconstructing the evolutionary history of a class of homeotic genes, may
thus be an important contribution to the study of the evolution of developmental
processes and hence of morphology.
1.i
Interaction of floral homeotic genes, the ABC-model
Homeotically altered plants of different kinds have been known to botanists for
centuries (e.g. Svedelius, 1909), and an important task for plant geneticists was to
characterise the molecular genetic basis for homeosis in plants, as had been done in
Drosophila melanogaster.
When the modern work on homeosis in Arabidopsis, Arabidopsis thaliana
and snapdragon, Antirrhinum majus began, three important types of floral homeotic
mutants were recognised, the A-, B- and C-mutants (e.g. Bowman et al., 1991; Coen
and Meyerowitz, 1991). In the A-mutants, sepals and petals do not form. Instead,
carpels and stamens respectively are formed in their place (but see below). In the Bmutants, sepals and carpels replace petals and stamens respectively, and in the Cmutants, petals and sepals replace stamens and carpels. The C-mutants furthermore
have a “flower within a flower” phenotype, new whorls of petals, petals, sepals and
so forth, are formed in the floral center. There is a difference between A. majus and
Arabidopsis here: in Arabidopsis the inner sepals actually belong to the inner flower,
whereas in A. majus a true fourth whorl of more or less sepaloid organs is present
(for discussion, see Coen and Meyerowitz, 1991 and Davies et al., 1999). A model,
based on these mutants, and of their double- and triple-mutant combinations,
explains how the corresponding A-, B-, and C-function genes interact in a
combinatorial manner to determine organ identity. A-genes direct sepal
development, A+B genes direct petal development, B+C genes direct stamen
development, and C-genes alone direct carpel development. Furthermore, A- and Cfunctions act as each others antagonists. In the A-mutant, C-function is ectopically
active within the first two whorls, and vice versa (Coen and Meyerowitz, 1991).
8
The A function is somewhat loosely defined. In accordance with the ABCmodel, the described A-mutant phenotype is found also in gain of function mutants
of the A. majus C-class gene (Bradley et al., 1993; Lönnig and Saedler, 1994). Lossof-function mutants in genes referred to as A-class genes may, however, also show
other phenotypes. Mutant alleles of the A-class gene APETALA2 (AP2) in
Arabidopsis display a range of floral phenotypes, some conforming to the described
A-mutant phenotype, but more common is the lack of second whorl organs
(Bowman et al., 1991). Plants carrying mutant alleles of the gene APETALA1 (AP1)
in Arabidopsis do not form sepals and petals, and it is hence an A-class gene, but
bracts and secondary flowers, respectively, may be formed in their place (Mandel et
al., 1991).
1.ii
Floral homeotic genes belonging to the MADS-box gene family
In 1990 it was recognised that the homeotic genes AGAMOUS from Arabidopsis,
and DEFICIENS from A. majus, both contained a region of 168 bp that was highly
similar in sequence. Similar regions were also found in MCM1 from Saccharomyces
cerevisiae, and SRF from Homo sapiens. This element was called the MADS-box
(Schwarz-Sommer et al., 1990), an acronym referring to these four genes, in which
it was first identified. MCM1 was known to be a DNA-binding transcription factor
(Passmore, 1988), and evidence suggests that all MADS-box genes encode
transcription factors. Of the genes that have been shown to act in the ABC-model of
floral organ identity determination, all, but the A-gene AP2 from Arabidopsis
(Jofuku et al., 1994), are MADS-box genes.
Two B-genes have been identified in A. majus, DEFICIENS, DEF (Sommer
et al., 1990) and GLOBOSA, GLO (Tröbner et al., 1992), as well as in Arabidopsis,
where the corresponding genes are called APETALA3, AP3 (Jack et al., 1992) and
PISTILLATA, PI (Goto and Meyerowitz, 1994). On the other hand, there is only one
C-class gene in Arabidopsis, AGAMOUS (AG) (Yanofsky et al., 1990). PLENA
(PLE) in A. majus confers the same function (Bradley et al., 1993).
1.iii The expression patterns of MADS-box genes reflect their function
The expression patterns of MADS-box genes within the floral primordia reflect their
function; for example C-genes are expressed in stamen- and carpel primordia, and
later in developing stamens and carpels. The B-class gene expression patterns are
also consistent with a role in the development of the organs, which are affected in
their mutant phenotypes: AP3 and PI are mainly expressed in whorls two and three.
Constitutive expression of the floral homeotic genes in plants, show phenotypes
expected from the ABC-model (Krizek and Meyerowitz, 1996; Mandel et al., 1992;
Mizukami and Ma, 1992). For example ectopic expression in Arabidopsis of the Cclass gene AG results in an A-mutant like phenoptype: sepals transformed to carpels
in the first whorl and staminoid, or absent, organs in the second whorl (Mizukami
and Ma, 1992).
1.iv Conservation of function between orthologous genes
Coen and Meyerowitz (1991) recognised that the floral homeotic genes with the
same functions in Arabidopsis and A. majus were those that showed the highest
sequence similarity. Of the genes mentioned above, AG is most similar in sequence
to PLE, AP3 to DEF, and PI to GLO. This suggests a conservation of function
9
between orthologous genes. Subsequent studies in e.g. Petunia gave consistent
results. However, the Petunia DEF-ortholog pMADS1, was found to be essential for
petal development only, but redundant for stamen development (Angenent et al.,
1992; 1995; Kush et al., 1993; Tsuchimoto et al., 1993; van der Krol et al., 1993;
van der Krol and Chua, 1993). Hence, although the ABC-model was found to be
valid for a range of angiosperm species, differences in details are apparent.
Phylogenetic analyses on MADS-box genes from several angiosperm species
confirmed that the MADS-box genes with the corresponding function in different
species were orthologs. Thus, the gene duplication events resulting in the floral
homeotic genes predate the origin of these angiosperm species (Doyle, 1994; Tandre
et al., 1995).
1.v
The MADS-box genes constitute a large gene family in eukaryotes
A large number of MADS-box genes have been isolated in particular from flowering
plants, and shown to function in various developmental processes (e.g. Ma et al.,
1991; Huang et al., 1995; Rounsley et al., 1995; Riechmann and Meyerowitz, 1997).
MADS-box genes have also been isolated from animals and fungi (for a review on
non-plant MADS-box genes, see Shore and Sharrocks, 1995). In animals, MADSbox genes of the SRF and MEF2-classes have roles in the control of cellular
differentiation. MADS-box genes in the budding yeast, Saccharomyces cerevisiae,
participate in the regulation of processes as diverse as arginine metabolism (ARG80,
first described as ARGR1, Dubois et al., 1987) and the mating type switch (MCM1).
MADS-box genes have not so far been described in protists or prokaryotes, although
the UspA-like genes from Escherichia coli and Haemophilus influenzae have been
suggested as possible relatives to MADS-box genes (for a discussion, see AlvarezBuylla et al., 2000; Theißen et al., 2000).
1.vi The MADS-domain
The MADS-box is a DNA-sequence element, generally defined as an 168 bp motif,
as mentioned, or as an 180 bp motif (e.g. Theißen et al., 1996), that encodes a DNAbinding protein motif, the MADS-domain.
The interaction between the MADS-domain and DNA is best understood in
the human proteins SRF, and MEF2A, as well as the yeast protein MCM1, since the
structures of these DNA-protein complexes have been determined (Pellegrini et al.,
1995; Tan et al., 1998; Huang et al., 2000; Santelli and Richmond, 2000). The
proteins form homodimers that bind specifically to a sequence element called the
CArG-box, with the sequence CC (A/T) 6GG (Phan-Dinh-Tuy et al., 1988). The
MADS-domains of the SRF, MEF2A, and MCM1 protein homodimers bind the
center of the CArG sequence through one amphipatic α-helix from each monomer
making contact predominantly with the minor groove. These α-helices contain some
of the most conserved amino acids in the MADS-domain, indicating that DNA
binding of different members within the gene family occurs in a similar manner. The
binding causes a bending of DNA, which in the case of MCM1 has been shown to
enhance interaction of other transcription factors with the adjacent major grooves.
(Tan et al., 1998) The DNA-binding α-helices are also involved in dimerisation,
together with a more C-terminal β-sheet structure (Pellegrini et al., 1995, Tan et al.,
1998).
10
1.vii Subfamilies of MADS-box genes
Outside the MADS-box, MADS-box genes from plants, animals and fungi display
different structural characteristics. Most MADS-box genes characterised from plants
conform to an architecture that distinguishes them from known MADS-box genes in
other organisms. These plant MADS-box genes possess a second sequence element,
the K-box (named after similarities to keratin, Ma et al., 1991), encoding a 65 aa
motif. The K-box shows a high degree of conservation, although lower than the
MADS-box. An intervening region, the I-region (also called the linker region, Lregion), separates the K-box from the MADS-box. The I-region is less conserved
than the K-domain and also varies somewhat in length, between 27 and 42 amino
acids on the protein level (using the definition of the MADS-domain as consisting of
56 amino acids, Riechmann and Meyerowitz 1997). Downstream of the K-domain
follows the C-terminal region which is the least conserved part of these transcription
factors both in length and sequence (Ma et al., 1991; Pnueli et al., 1991; Riechmann
and Meyerowitz, 1997).
Genes conforming to the described architecture are referred to as MIKCgenes (Münster et al., 1997). Many MIKC-proteins have the MADS-domain at the
N-terminal end, but in AG-like genes, an N-terminal region is present, typically
extending 16 or 17 aa upstream the MADS-domain (Ma et al., 1991; Mandel et al.,
1992; Tsuchimoto et al., 1993; Davies et al., 1999).
A number of Arabidopsis MADS-box genes, have been shown to lack a Kbox, and most of these group in a clade separate from the MIKC-genes (AlvarezBuylla et al., 2000, III). A recent study suggests that this clade is most closely
related to the genes MCM1 and ARG80 from yeast and the SRF genes from animals.
The MIKC-genes, on the other hand, appear to be related to the MEF2-type genes
from animals and the genes YBR1245 and RLM1 from yeast (Alvarez-Buylla et al.,
2000). These two clades are characterised by MADS-domains containing diagnostic
amino acids, referred to as TypeI and TypeII MADS-domains. MIKC-genes encode
TypeII MADS-domains. Hence, the split between genes with TypeI and TypeII
MADS-boxes predate the separation of the lineage leading to fungi and animals
from that leading to plants.
1.iix Functional domains of MIKC-genes
Like MCM1 and SRF, MIKC-proteins bind DNA as dimers at CArG-boxes. Both
homo- and heterodimerisation of MADS-domain proteins have been demonstrated
(e.g. Schwarz-Sommer et al., 1992, Tröbner et al., 1992; Huang et al., 1993; Huang
et al., 1996, Davies et al., 1996; Riechmann et al., 1996). The MADS-domain and a
region extending towards the C-terminal, variable in length from encompassing only
the first amino acids of the I-region to a portion of the K-domain, constitutes the
core that is needed for DNA-binding and dimerisation (Huang et al., 1996;
Riechmann et al., 1996). This is similar to the situation in SRF, where the MADSdomain and the adjacent downstream region constitute the DNA-binding and
dimerisation core of the protein (Pellegrini et al., 1995).
The major function of the I-region appears to be that of determining
dimerisation specificity, as demonstrated by domain swapping experiments with
AP1, AP3, PI and AG (Riechmann et al., 1996). To be precise, a region slightly
larger than the I-region is needed for specific dimerisation, encompassing also the
11
C-terminal region of the MADS-domain (i.e. AG homodimerisation) or the Nterminal region of the K-domain (i.e. AP3/PI heterodimerisation).
A similar function was expected for the K-domain, since its predicted coiledcoil structure with hydrophobic faces implies a role in protein-protein interaction
(Ma et al., 1991). It has been demonstrated that the K-domain of AG is sufficient to
mediate specific dimerisation with other K-domains (Fan et al., 1997). However, the
K-domains of certain proteins with which AG is known to dimerise could not
interact in these experiments. Thus, the K-domain appears to be involved in
dimerisation, but in a manner not well understood.
The highly variable C-region has not until recently been functionally
characterised. In AP1, the C-terminal region has been demonstrated to function as a
transcription activation domain (Cho et al., 1999). The C-region appears to be of
importance in the formation of a ternary complex between DEF/GLO heterodimers
and SQUAMOSA homodimers in A. majus (Egea-Cortines et al., 1999). It thus
appears that the C-region is involved in the interaction with ternary factors, being
either other MADS-domain proteins or factors closer to the basal transcriptional
machinery.
1.ix Exon/intron organisation of MIKC-genes (I, II, IV)
Typically, the coding region of MIKC-genes contains six or seven introns, that I
refer to as introns M, A, B, C, D, E, and F (I, II, IV). Intron M is situated within the
I-region, close to the end of the MADS-box. Together with intron A, situated within
the I-region, close to the 5’ end of the K-box, this M defines the exon that
corresponds to the major part of the I-region. The K-box of AG, and most other
MIKC-genes is built up by three exons. In AG the exon defined by intron A and B is
62 bp long, that by B and C 100 bp long, and that defined by C and D 42 bp. The Kbox is followed by an exon also spanning 42 bp, defined by introns D and E. Intron
F is positioned further toward the 3’-end, in AG this intron is situated at the 3’-end
of the coding sequence (Yanofsky et al., 1990).
Certain genes may lack some of these introns, for example PI lacks intron M,
AGL5 lacks intron B, and AGL2 lacks intron D. Genes possessing the same introns
as AG, may differ in the exact positions of these. Some of these changes appear to be
unique to a particular gene, but other are clade-specific. This indicates that the
details in the exon/intron organisation may be useful as an independent test of the
phylogenetic association of a particular gene. In this study, I focus on introns M, A,
B, C, D and E (see fig. 1, see I, II, IV for sources of data).
Independent origins of the introns M, A, B, C, D and E is highly unlikely. A
phylogenetic position of a gene that would imply loss and regain of any of these
introns is hence likely to be incorrect. However, shifts in the exact positions of these
introns may be prone to homoplasy. Alternative splicing was encountered among the
genes presented in this report (II, IV) and a shift in intron position might hence in
certain cases be regarded as a quantitavtive, rather than qualitative change:
acceptor/donor sites used at low frequency become the “correct” sites. This simple
mechanism suggests that identical shifts in intron positions may occur
independently. Alternative splicing has also been documented to occur in other
MIKC-genes, e.g. the A. majus gene FAR (Davies et al., 1999).
The genes AP1, SQUA, and AGL8 share a shift of intron A three bp toward
the 5’ end relative its position in AG. They further have intron M shifted three bases
12
toward the 3’-end relative AG. An identical shift in intron M occurs in AGL9, AGL2,
AGL4, and AGL3 that belong to a more inclusive clade together with AP1, SQUA
and AGL8. This more inclusive clade also includes the closely related genes AGL6
and AGL13, which, however, conform to AG in the position of intron M. The
relation between the AP1, AGL9 and AGL6 subclades differs among analyses or is
unresolved (e.g. I, II, III, IV). If the AP1-like position is derived, and the AG-like
position ancestral, the intron M position in itself suggests a basal position for AGL6,
because independent acquisitions of this position would otherwise be demanded in
the AP1- and AGL9 subclades. The present phylogenetic analyses indicate that
identical shifts in the position of intron M have occurred several times. For example,
the AGL17 clade has members with both AP1- and AG-like intron M positions. In
short, homoplasies in the shifts of intron positions exist to some extent. The close
paralogs AGL2 and AGL4 are united in the lack of intron D. The paralogous genes
AGL17, ANR1, F2809.80 and F20D10.60 are united by a dramatic shift of intron E,
42bp towards the 3’-end, doubling the size of the exon defined by D and E, relative
AG. This shift is also present in their homolog DEFH125 from A. majus.
The B-class genes AP3, PI, DEF and GLO are united by a shift of intron E
three bases to the 3’ end, and apart from PI that lacks intron M, a shift of intron M
six bases toward the 3’-end (see section 2.v).
The AG-like pattern of intron positions occurs in genes closely related to AG:
AGL1, and AGL11. Within the AG-clade some deviations are documented; AGL5
lacks intron B.
The AG-like pattern also occurs in AGL24, AGL6, F7H19.130, and AGL20
that do not appear to be closely related to AG, or (except F7H19.130 and AGL20), to
each other. That the AG-pattern is suitable as a reference is probably not a
coincidence; it may represent an ancestral exon/intron organisation. In this work,
data supporting this hypothesis is presented.
1.x
Divergent MADS-box genes with a MIKC exon/intron pattern (IV)
The exon/intron organisations of genes encoding TypeII MADS-domains, AGL12,
AGL27 (=F22K20.15), FLF (=AGL25), but reportedly lacking K-boxes (AlvarezBuylla et al., 2000), as well as of the divergent gene F28K20.7 are similar to that of
MIKC-genes. AGL12 has a exon/intron organisation conforming to that of AG,
regarding introns M, A, B, C, D, and E. The introns B, C, D, and E of AGL27 also
conform to the positions in e.g. AG. Intron A, however, is lacking in this gene, and
intron M is shifted by four bp toward the 5'-end, compared to e.g. AG. In FLF,
introns A, B, C, D, and E conform to the AG-pattern, whereas intron M is shifted by
three bp toward the 3'-end in relation to it position in AG. In F28K20.7, introns B
and E are missing, intron A is shifted by 12 bp toward the 5'-end, and intron D is
shifted by 21 bp toward the 3'-end. Intron M is shifted by 6 bp towards the 3'-end in
this gene. The position of Intron C, however, conforms to the corresponding position
in e.g. AG.
The similarities in exon/intron organisation within the K-domain region
(introns A-E) to typical MIKC-genes, show that the genes AGL12, AGL27
(=F22K20.15), FLF (=AGL25) and F28K20.7, do have regions homologous to Kdomains. However, the exon/intron pattern of F28K20.7 is highly divergent.
13
t ANR1
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
n
----
t
-6
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t AGL17
+42
t
t
+3
t
+3
t
t
t
t
+21
t
E
t
C
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
100
62
B
t
K box
+42
AGL13
AGL6
AGL3
AGL9
AGL2
AP1
+27
DAL12
AGL17
CLADE
B-CLASS
CLADE
GLO
AP3
DAL13
FLF
AGL27
AGL12
F28K20.7
LAMB1
LAMB6
LAMB4
AG
t
42
D
t
42
C region
t
t
t
t
Figure 1. Exon/intron organisation of MIKC-genes. Intron positions are
indicated by triangles. Exon lengths are indicated by numbers above the
AG-sequence, and shifts in position relative AG by positive
(downstream) or negative (upstream) numbers above each triangle,
corresponding to a shifted intron. Clades are indicated to the right.
14
t
+3
t
+3
t
t
t
-6
+3
t
+3
t
-3
+3
t
t
-9
t
+6
t
+6
+3
t
-4 t
-4
t
+6
t
t
t
t
t
t
t
t
t
-12
t
+6
t
+3
M
t
t
t
t
Linker
-8
A
t
t
1.xi The role of MIKC-genes in angiosperm development
A large number of MIKC-genes have been isolated from angiosperms, and their
function is best known in Arabidopsis thaliana, Antirrhinum majus (e.g, SchwarzSommer et al., 1990; Huijser et al., 1992; Tröbner et al., 1992; Lönnig and Saedler,
1994; Zachgo et al., 1997; Egea-Cortines et al., 1999), Tomato, Lycopersicon
esculentum (e.g. Pnueli et al., 1991; 1994; Hareven et al., 1994), Petunia (e.g.,
Angenent et al., 1992; 1994; 1995; Kush et al., 1993; Tsuchimoto et al., 1993; van
der Krol et al., 1993; van der Krol and Chua, 1993; Colombo et al., 1995), and
Gerbera hybrida (Yu et al., 1999). Unless otherwise stated, the following discussion
will refer to work done in Arabidopsis.
Some MADS-box genes participate in the regulation of vegetative
development. AGL12, AGL14, AGL17 (Rounsley et al., 1995), and ANR1 (Zhang
and Forde, 1998) have all been shown to be expressed in the root. Of these, ANR1 is
functionally best characterized and it is involved in the formation of lateral roots as a
response to environmental [NO3-] (Zhang and Forde, 1998). The gene AGL3 in
Arabidopsis is expressed in all above ground organs (Huang et al., 1995), but its
function in these tissues is poorly understood.
A broad array of MADS-box genes are known to have regulatory functions
in floral development, from the onset of flowering, the establishment of floral
meristem identity, the determination of floral organ identity, as well as in fruit and
seed development. One of the earliest changes in MADS-box gene expression
related to the transition to flowering is the down regulation of FLF that encodes a
repressor of flowering. FLF has been demonstrated to be regulated by genes in
flowering induction pathways (Sheldon et al., 1999).
The MADS-box gene AP1, together with other genes, mediates the
determination of floral meristem identity. One of these other genes is LEAFY, that
encodes a transcription factor unrelated to MADS-proteins (Weigel et al., 1992).
The lfy mutation shows transition of flowers into shoots, although abnormal flowers
may arise late in development. The double mutant lfy ap1 shows an enhanced
vegetative phenotype (Bowman et al., 1993). Plants ectopically expressing AP1
flower early and have shoots transformed into flowers (Mandel and Yanofsky,
1995). Hence, AP1 is important in the establishment of floral meristem identity. The
gene CAL is very similar in sequence to AP1, and a cal ap1 double mutant has, in
the place of flowers, "cauliflowers" with a reiterated floral meristem (Kempin et al.,
1995), but may finally form ap1-like flowers (described above). A third MADS-box
gene closely related to AP1and CAL is FUL (=AGL8). FUL is expressed in the
inflorescence stem and meristem, as well as in the cauline leaves. In the floral
primordia it becomes downregulated by AP1, that initially is expressed throughout
the floral meristem (Gu et al., 1998). A triple mutant ap1calful shows a very
dramatic vegetative phenotype (Ferrandiz et al., 2000), that under normal growth
conditions never flowers. In short, transition to flowering involves a network of
interacting transcription factors of MADS- and other types, showing partial
redundancy.
At later stages, the onset of AG expression (the C-function gene) in the two
inner whorls downregulates AP1, that specifically becomes expressed in whorl one
and two. Here, AP1 promotes sepal and petal development, and in this sense
functions as an A-gene.
15
Temporal and spatial expression patterns of the gene AGL9 in Arabidopsis,
and the orthologous genes TM5 in tomato (Pnueli et al., 1994) and FBP2 in Petunia,
(Angenent et al., 1994) suggest a function in the developmental sequence between
the activity of floral identity genes and the B- and C- genes of the inner three
whorls. In a recent study it was shown that agl2agl4agl9 triple mutants form
indeterminate flowers with all whorls occupied by sepals (hence the genes are also
referred to as SEPALLATA 1, 2, 3), but that the initial expression of B- and C-genes
remains unaffected in the triple-mutant (Pelaz et al., 2000).
The gene FUL (AGL8), whose role in inflorescence- and flower development
is mentioned above, also has a function in carpel development. The loss of function
mutant show reduced growth of the siliques (Gu et al., 1998).
The Arabidopsis MADS-box gene AGL11 is specifically expressed in
developing ovules and associated placental tissue (Rounsley et al., 1995). In petunia,
the AGL11-orthologs FBP7 and FBP11 have been shown to be necessary and
sufficient, within the floral context, for the initiation of ovule formation, and have
been proposed to represent D-function genes (Angenent et al., 1995, Colombo et al.,
1995). Also other MADS-box genes are expressed in the ovule. The B-class gene
AP3 is expressed in the integuments, but viable seeds are produced by ap3 mutants,
and hence the gene is not essential for ovule development (Jack et al., 1992).
1.xii Phylogeny of function in plant MADS-box genes
In the discussion above, examples of the conservation of function between
orthologous genes were encountered, e.g. the function in the inner three whorls of
the orthologous genes AGL9, TM5, and FBP2. Paralogous genes may also have
similar functions. The partial redundancy in function between the close paralogs
AP1, CAL, and FUL, as well as between the close paralogs AGL2, AGL4 and AGL9,
provide examples of gene duplication products that probably originally had identical
functions, but that are undergoing functional divergence. The entire MADS boxgene family most likely has evolved in this manner, through gene duplication and
subsequent functional divergence of the resulting genes. This leads to the question
of what function(s) the common ancestor(s) of the different clades/functional groups
recognised among angiosperm MIKC-genes had. Do orthologs to e.g. the C-class
MADS-genes that define stamen and carpel organ identity genes exist only in plants
with stamens and carpels, or do such genes exist also in other plants? When did the
different paralogs of MADS-box genes defining structures unique to angiosperms
evolve, or in more general terms, in what respect is the evolution of organ identity
genes related to the evolution of morphological novelties?
Such questions constitute the basis for the work presented here.
2.
MADS-box genes in gymnosperms (I, II)
A major breakthrough in the MADS-box gene research was the demonstration by
Tandre et al., 1995 that the conifer Norway spruce, Picea abies, contains MADSbox genes, which belong to the clades that correspond to functional groups in
angiosperms. The gene DAL1 was shown to group close to the AGL6/ AGL2-like
genes, and DAL3 grouped together with the gene TM3 from tomato, a gene
subsequently shown to be orthologous to AGL14 in Arabidopsis (II). Most
interesting, however, was the finding that DAL2 appeared to be closely related to
AG, and other angiosperm C-class genes.
16
MADS-box genes have later been isolated also from other conifer species, e.g. Pinus
radiata (Mouradov et al., 1996; 1997; 1998; 1999), some of, which are included in
the phylogenetic analysis, presented in (II).
2.i
DAL2 – a Norway spruce C-class gene (I, II)
The gene DAL2 (Tandre et al., 1995) shows extensive sequence similarity to C-class
MADS-box genes, e.g. AG from Arabidopsis. Within the MADS-domain (defined as
56 aa), only one amino acid differs between DAL2 and AG (Tandre et al., 1995).
Phylogenetic analyses support the association between DAL2 and the angiosperm Cclass genes. As an independent test of the phylogenetic position of DAL2, a partial
characterisation of its exon/intron organisation was conducted. This showed that the
introns A, B, C, D and E conform to the AG-like positions. However, this pattern is
likely to be ancestral among MIKC-genes and hence of limited phylogenetic
significance (see below). DAL2 is also closely related to the Arabidopsis genes
AGL1, AGL5, AGL11, and their orthologs in other angiosperm species (Tandre et
al., 1995, I, II, III, IV ). All these genes have in common a role in gynoecial
development. AGL11 is expressed in ovules (Rounsley et al., 1995), AGL1 and its
close paralog AGL5 are specifically expressed in the carpel (Ma et al., 1991,
Flanagan et al., 1996, Savidge et al., 1995).
The gynoecial expression of AG and all its close paralogs, and the expression
of DAL2 in female cones (I), suggest a conservation of a function in the
development of ovuliferous structures between DAL2 and the entire AG-clade. The
expression of DAL2 also in the pollen forming organs, the male cones (Tandre et al.,
1995, II), however, is specifically similar to the staminal expression of AG and its
orthologs in other angiosperm species, but not to the AG paralogs within the clade
(and their orthologs in other species).
In situ hybridisation experiments were made in order to get a
morphologically detailed expression pattern of DAL2 (I, II). DAL2 in situ
hybridisation signal was detected in the ovuliferous scale, but no expression was
found in the bract, subtending the ovuliferous scale, in the primary axis of the cone,
or in the apical meristem (I). DAL2 expression could be detected also at early stages
of the development of the ovuliferous scale, in the axil of bracts, with no discernible
scale primordia. Cones with these developing ovuliferous scales present close to the
apical meristem also have developmentally older ovuliferous scales further toward
the base. At late stages, DAL2 expression was found to be restricted to the part of the
scale on which ovules later will develop (I). In situ hybridisation experiments
conducted on male cones (II) revealed expression of DAL2 specifically in he prepollen mother cells on the abaxial side of the microsporophylls.
2.ii
DAL2 can act as a C-class gene in an angiosperm context (I)
To study if DAL2 could act developmentally in a manner similar to AG, a construct,
of DAL2 positioned in the sense orientation downstream the 35S promoter was
introduced in Arabidopsis via Agrobacterium tumefaciens mediated transformation
(I). DAL2 lacks the N-terminal extension characteristic of C-class genes in
angiosperms, and in order to elucidate whether this motif is of importance for a
functional C-class gene, a fusion construct between the N terminal of the Brassica
napus AG-homolog, BAG1, and DAL2 was made, called NDAL2.
17
Arabidopsis plants constitutively expressing DAL2 constructs show either a
weak phenotype with minor alterations in floral organ shape or a strong phenotype
with homeotic alterations similar to those of plants transformed with a
corresponding AG-construct (I). More specifically, ovule-like structures and
stigmatic tissue were formed on the margins of the sepals, and the petals were
transformed into filamentous or stamen-like organs. This was seen both in two
plants transformed with the 35S::DAL2 construct, and one plant transformed with
the 35S::NDAL2 construct.
Similar results were obtained in a separate study, where the DAL2-ortholog,
from black spruce, SAG1 (Picea mariana) was ectopically expressed in Arabidopsis
(Rutledge et al., 1998). This indicates a highly conserved tertiary structure between
DAL2 (and SAG1) and AG, suggesting that the context in which DAL2 acts during
cone development, may be very similar to that of AG during flower development.
2.iii DAL2 and the functional diversification of the AG-clade
Two genes that are structurally similar but that have distinct functions in different
parts of the plant, may still confer the same phenotype when constitutively
expressed. If constitutive expression of a sister gene to both these genes, derived
from a different species, again result in the same phenotype, it can not be determined
whether this sister gene acts in a developmental context similar to the first, or to the
second gene.
Phylogenetic reconstructions suggest that DAL2 is a sister gene to the entire
AG clade, including AGL1, AGL5, as well as AGL11 (I, II). AGL1 and AGL5 was
only recently functionally characterised (Liljegren et al., 2000). These two genes are
closely related and show functional redundancy, with no dramatic single gene
mutant phenotypes. An agl1agl5 mutant lacks dehiscence zone development (and
hence the genes are now known as SHATTERPROOF 1 and 2), but has no effect on
floral organ identity. Constitutive expression of AGL1 or AGL5 causes floral
homeotic alterations, similar to those observed to result from constitutive expression
of AG, although their wild type function is restricted to the carpel.
Constitutive expression of AGL1 and of AGL5 results in homeotic
alterations, that are similar to those resulting from constitutive expression of AG, in
spite of the fact that AGL1 and AGL5 have developmental functions distinct from
those of AG. This makes a conclusion of conserved function between DAL2 and AG,
based upon the same homeotic conversions somewhat less convincing. However,
ectopic expression of AGL1 and AGL5 also results in alterations restricted to the fruit,
and not seen upon ectopic expression of AG, or DAL2. This suggests that the
function of AG and DAL2 indeed may be conserved, whereas the AGL1 and AGL5
function may be derived.
The diversification of the AG-family within the angiosperms may reflect new
developmental processes associated with the origin of the angiosperm gynoecium
(I). It can not, however, at present be excluded that one or more of the paralogs may
exert functions distinct from AG, but similar to functions of DAL2 in spruce
development. Hence the diversification within the AG-family in angiosperms may
also reflect a "division of labor", rather than the origin of new functions. AGL1 and
AGL5 appear to regulate an angiosperm-specific function, whereas the D-function of
AGL11 (as inferred from the D-function of the Petunia orthologs FBP7 and FBP11,
and the ovule specific expression pattern of AGL11) regulates the formation of a
18
structure shared between angiosperms and other seed plants. Accordingly, Theißen
et al. (2000) proposed that DAL2 may be considered to be a C/D- function gene, and
the AG/AGL11 split would thus represent the "division of labor" alternative. No
expression of DAL2 was observed in ovules however, which would be expected
from this hypothesis.
The picture is still a bit more complex, because several species have been
shown to contain more than one AG paralog. In Petunia, tobacco, and A. majus,
respectively, the AG homologs FBP6, NTPLE and PLE form a clade separate from a
second clade of AG-homologs, PMADS3, NAG, and FAR, from the same species
(Davies et al.,1999, for similar examples in other species see e.g. Kater et al., 1998
and Theißen et al., 1995). These close paralogs may have different functions. For
example, ectopic expression of PLE in tobacco plants gives carpels in whorl one and
stamens in whorl two (Davies et al., 1996). Constitutive expression of FAR,
however, generally leaves the first whorl unaffected, but causes a more extreme
conversion of whorl two into stamens. The ple-mutant is a typical C-mutant,
whereas the far-mutant does not lead to homeotic changes, but to male sterility
(Davies et al., 1999). This indicates that constitutive expression of DAL2 in different
angiosperm species may result in different phenotypes, and the choice of
Arabidopsis, in this perspective is of course arbitrary.
2.iv Putative B-class genes from Norway spruce (II)
The phylogenetic analyses placing the conifer MADS-box genes DAL1, DAL2 and
DAL3, within clades of angiosperm genes, suggested that orthologs also to other
angiosperm MADS-box genes, such as the B-class genes, should exist in conifers
(Tandre et al., 1995, I).
Three conifer MADS-box genes DAL11, DAL12 and DAL13, were isolated using a
screen of a Norway spruce cDNA library, and/or RACE. In a phylogenetic analysis,
based on MADS- and K-box sequences, the genes appear in the B-class gene clade
(II). However the B-clade is unstable among analyses that differ in details of the
matrix, as compared to, for example, the C-clade, that is less affected by such
differences. This is also reflected in the low bootstrap support for the B-clade in the
tree (II). In general, the comparatively low support for the B-class clade, may be
attributed to a higher evolutionary rate within this lineage (see Purugganan, 1997).
An independent source of data that supports the association of DAL11,
DAL12 and DAL13 to the angiosperm B-class clade is their possession of specific
protein sequence motifs in the C-terminal region. These motifs were identified in an
extensive comparative study of angiosperm B-class genes, and are referred to as the
PI and AP3 motifs respectively (Kramer et al., 1998). The AP3 motif can be divided
into the EuAP3-motif and the PaleoAP3 motif. The PaleoAP3 motif
(YGxHDLRLA) occurs in AP3-like genes of basal angiosperms, whereas the
derived EuAP3-motif is a characteristic of “higher” eudicots (Kramer et al., 1998).
The PI motif (MPFxFRVQPxQPNLQE) is present in all angiosperm B-class genes
(and hence supports the monophyly of these). DAL11, DAL12 and DAL13 all
possess the PI-motif, whereas a PaleoAP3 motif is present only in DAL12. In itself,
this suggests DAL12 to be a member of the AP3 clade, and DAL11 and DAL13 to be
members of the PI-clade. Most phylogenetic analyses support a close relation
between DAL11 and DAL13, whereas the topology uniting these two genes, DAL12
and the angiosperm B-class genes varies. The tree presented in (II) suggests that all
19
three spruce genes belong to the PI-clade, a position that would require the existence
(or loss) of unknown AP3-like spruce genes, and hence is not very likely.
2.v
Exon/intron organisation of DAL11, DAL12, and DAL13 (II)
Another source of data we have suggested as an independent test of the
phylogenetic position of MIKC-genes is details in the exon/intron organisation (see
section 1.ix). Angiosperm B-class genes have a diagnostic exon/intron organisation.
Compared to the positions in AG, intron M in AP3, DEF, and GLO is shifted six
bases toward the 3’-end (PI lacks this intron). In DEF, GLO, PI and AP3, the exon
between intron D and E has a length of 45 bp, rather than 42 bp, as in AG. In AP3
and DEF, as well as in AG, the exon separating introns C and D is 42bp long. In PI
and GLO this exon is 30 bp long, making introns D and E appear as shifted 12 bp
toward the 5’end (I), unless a corresponding gap is introduced in the exon between
intron C and D (II). In other words, PI, GLO, DEF, and AP3 differ from AG in that
intron E is shifted three bp toward the 3’ end.
Whereas in DAL13 all the introns M, A, B, C, D, and E, were characterised,
in DAL11, only introns C, D, and E were covered. The exon/intron organisation of
DAL11 and DAL13, does not conform to that of angiosperm B-class genes, but
rather to that of AG-like genes, including DAL2 (I). This is in conflict with the
position of DAL11, and DAL13 among the B-class genes, only if the AG-like
exon/intron pattern is a unique derived character (synapomorphy) of the C-clade, but
not if it is a shared ancestral, plesiomorphic, trait among MIKC-genes. The latter is
very likely, as will be discussed in greater detail below. Thus the exon/intron
organisation does not provide any information as to whether DAL11 and DAL13 are
related to the B-class genes, but given such a relation, it suggests that DAL11 and
DAL13 are positioned at the base of the B-class clade. DAL12 has a unique
exon/intron organisation, where intron D is missing, intron A is shifted 9 bp to the
5’-end, and intron E is shifted 27 bases toward the 3’ end, a position not shared with
any other MIKC-genes. Unlike DAL13, DAL12 has intron M at a position shared
with the B-class genes, shifted six bases toward the 3’-end relative the
corresponding position in AG. The position of intron M is the most variable among
the discussed introns, and homoplasies at this position are not unlikely.
2.vi The expression of conifer B-class genes is restricted to male cones (II)
Of all tissues examined, including vegetative buds, cambium and female cones,
DAL11, DAL12, and DAL13 are expressed exclusively in pollen cones. The
expression of DAL11 is rather uniform in the developing male cone both prior to and
after winter dormancy. In situ hybridisation experiments reveal that DAL11 is
expressed in the apical meristem, in the microsporophyll primordia and in the
central pith of recently initiated pollen cones, but not in the subtending bract or in
the basal "crown" region. At later stages, expression decreases in the pollen mother
cells. DAL12 and DAL13, however, show more restricted expression patterns, both
temporally and spatially. Northern blot analyses show that the expression of DAL12
is high in cones after meristem termination, before winter dormancy, but there
appeared to be no expression when growth resumes in spring. In situ hybridisation
experiments show that DAL12 is predominantly expressed in the procambium. A
Northern blot shows that the expression of DAL13 decreases earlier before winter
dormancy, at a stage earlier than that of DAL12, but unlike DAL12, the expression of
20
DAL13 increases again in spring. In pollen cones before apical meristem
termination, DAL13 was expressed in the apical meristem, and in the
microsporophyll primordia. In the microsporophyll primordia, expression was
lowest in the prepollen mother cells. After apical meristem termination, DAL13
expression is high only in the tissues surrounding the developing pollen mother
cells. After winter dormancy, the expression is restricted to the microsporangial
wall.
Hence, the genes DAL11, DAL12, and DAL13 are expressed specifically in
developing male cones. This, together with their phylogenetic position close to the
angiosperm B-class genes, indicates that the B-function in the determination of
pollen producing reproductive structures is conserved between angiosperms and
gymnosperms. The expression of DAL2 also in male cones (Tandre et al., 1995, II)
suggests that also the interaction between B- and C-genes in the determination of
pollen organ identity, might be conserved between gymnosperms and angiosperms.
2.vii B- and C-genes in Gnetum
A number of MIKC-genes have also been isolated the gnetophyte, Gnetum gnemon
(Winter et al., 1999). Two of these, GGM3 and GGM2 have been reported to be
orthologs of C- and B-class genes respectively, and their expression patterns are
consistent with such a notion. Interestingly, the Gnetum genes appear as being more
closely related to Norway spruce genes, than to angiosperm genes, thus
contradicting the anthophyte theory, where gnetophytes are considered the extant
closest sister group of angiosperms. However, because of the instability of the Bclass clade, among different analyses, it might be equivocal to draw such
conclusions based upon the phylogenetic position of GGM2.
3. MADS-box genes in non seed plants, e.g. Lycopsids (III, IV)
When orthologs to the floral homeotic genes were shown to exist in gymnosperms,
we asked whether these genes emerged in the common ancestor of seed plants
specifically, or earlier in evolution. Purugganan (1997) assumed, from molecular
clock estimates, that the divergence of the floral homeotic gene lineages dated back
to approximately 480 mya, in the common ancestor of all land plants, including
liverworts.
However, if the origin of floral homeotic genes is tightly coupled to the
origin of ovules and pollen bearing organs, B- or C- orthologs would not be
expected to be present in non-seed vascular plants, e.g. ferns and lycopsids.
Alternatively, such orthologs may be present, but have other developmental roles,
perhaps related to those in seed plants, such as in sporogenesis, or, alternatively, in
the determination of male and female gemetophyte structures.
In order to investigate the MADS-box gene family in an earlier branch of the
plant evolutionary tree, and throw light on the question of ancestral expression
patterns of MIKC-genes, I here present the isolation of MIKC-genes from a
lycopsid, Lycopodium annotinum.
21
3.i
Lycopsids constitute the sister group to all other extant vascular plants
The earliest sporophytes of vascular plants known are those of rhyniophytes of the
late Silurian. They have protostelic stems (similar to roots in ”higher” plants) that
ultimately end in sporangia. They lack roots and leaves (Stewart and Rothwell, 1993;
Kenrick and Crane, 1997). The Devonian gametophyte Sciadophyton, with its
branching stem containing vascular tissue, ultimately ending in the reproductive
organs, suggests that the earliest vascular plant may have had a close to isomorphic
life cycle (Kenrick and Crane, 1997).
In the early Devonian, vascular plants were split in to two main groups. One
group maintained the plesiomorphic condition of terminal sporangia, this is the
branch leading to e.g. angiosperms. The other group bore lateral sporangia. Early
representatives of this group, the zosterophylls, had in common with the
rhyniophytes the absence of roots and leaves, and the stems were protostelic and
branched dichotomously (Kenrick and Crane, 1997). The zosterophylls became
extinct in the early Carboniferous (Kenrick and Crane, 1997). Today the only
representatives of the lateral sporangiate group are the lycopsids, which appeared
already in the lower Devonian (Kenrick and Crane, 1997). The lycopsids differ from
the zosterophylls in that they have roots and leaf like structures, called microphylls.
The position of lycopsids as the sister group to all other extant vascular
plants is supported by cladistic analyses on morphological datasets (Kenrick and
Crane, 1997), as well as by molecular phylogenies based on cox3 (Malek et al.,
1996) and mitochondrial 19S rDNA sequences (Duff and Nickrent, 1999), and the
gene order in a 30-kb region of the cpDNA. Lycopsids (including Lycopodium,
Selaginella and Isoetes) share the gene order with bryophytes, whereas other
vascular plants are united by the reverse gene order (Raubeson and Jansen, 1992).
We have chosen to isolate MADS-box genes from Lycopodium annotinum,
the most common lycopsid in Sweden. Only sporophytes have been available for
study. The shoot system of the Lycopodium annotinum sporophyte consists of a
creeping main axis, which dichotomises unequally to produce erect side branches.
After up to 7 or, in Arctic areas, 12 years of growth these terminate in the
reproductive structure, a strobilus bearing sporangia (Kukkonen, 1994). Roots are
formed continuously along the main axis.
3.ii
LAMB1 is a divergent MIKC-gene from Lycopodium annotinum (III)
A screen of a Lycopodium annotinum strobili cDNA library for MADS-box genes,
using a fragment of a MADS-box isolated through PCR using degenerate MADSprimers, resulted in the isolation of a cDNA containing a region identical in
sequence to the probe. The MADS-box gene, corresponding to the cDNA was
named LAMB1 (Lycopodium annotinum MADS Box gene 1). The MADS-box of
LAMB1 was found to be different from those previously identified, by the presence
of an inserted amino acid within its C-terminal end. The exact position of this
insertion is uncertain. It may be the A, P, A or T at positions 49, 50, 51, or 52,
respectively. These positions are consistent with an insertion that would not disrupt
the MADS-domain structure, because they correspond to a region that in MEF2A
and SRF has been shown to form a loop within the β-sheet structure involved in
MADS-domain dimerisation (Pellegrini et al., 1995, Huang et al., 2000, see also
section 1.vi).
22
LAMB1 was shown to contain a region with high sequence similarity to Kdomains. However, outside a central core region the similarity to other K-domains
was low, and in some clustalW analyses gaps appeared within these less conserved
regions. The I-region of LAMB1, with its length of 171 nucleotides (taking into
account that the inserted amino acid within the MADS-domain suggests a 61 aa
MADS-domain, rather than 60 aa, and assuming that no gaps occur within the 5’region of the K-box) is longer than those of all other MIKC-genes.
PCR, using genomic DNA as a template was performed in order to
investigate the exon/intron organisation of LAMB1. The analyses revealed seven
introns. Of these, intron 1 is situated in the 5’ untranslated region. Intron 2 is
situated close to the end of the MADS-box, and hence corresponds to the conserved
intron M of angiosperm and conifer MADS-box genes. The I-region contains three
additional introns, compared to other MIKC-genes. Interestingly, Clustal W analyses
at the amino acid level produce gaps in typical MIKC-genes corresponding
approximately in position to these introns. Hence, instead of one exon as in typical
MIKC-genes, four exons are residing entirely within the I-region, of 48, 13, 68, and
28 bp respectively. Intron 6 is situated in the I-region, nine bases 5' of the putative
K-box, and thus corresponds exactly to the position of intron A in AG. This supports
the alignment of the LAMB1 K-domain, to other K-domains in the less conserved Nterminal region. Intron 7 is located within the K-box, and corresponds to the
conserved intron B. The conserved introns C, D, E, and F that occur at the 3' end of
the K-box, and within the C-terminal of other MIKC-genes appear to be absent in
LAMB1. However, structural features of the C-terminal suggests that it starts after a
K-domain of typical length (see below), with no gaps in the less conserved 3' end of
the LAMB1 K-box.
Is the exon/intron organisation of LAMB1 primitive or derived? If the
organisation of LAMB1 is primitive, the shorter I-regions of other MIKC-genes may
have evolved through the loss of exons by use of the acceptor/donor sites of two
neighboring introns, including the intervening I-region exons into a larger intron.
Intron M is very long in most MIKC-genes, and this might perhaps reflect such an
evolutionary history of this intron.
Secondary structure prediction analyses of the I- as well as the K-region of
LAMB1 revealed that this entire region is likely to have a mostly α-helical structure.
Plotting the amino acids into a helical wheel showed that these helices would be
likely to contain hydrophobic as well as hydrophilic faces, consistent with a function
in dimerisation, as in other MIKC-genes (see section 1.iix).
Between amino acid positions 80 and 101 the I-region shows limited, but
notable, similarities to a different group of plant transcription factors, the HD-ZIP
proteins (Ruberti et al., 1991; Mattsson et al., 1992; Schena et al., 1992). Although
the similarity occurs in a region that is a putative amphipathic helix, the region is not
leucine-zipper-like. Leucine zippers are built up by a series of amino acid heptads,
each given a number starting from the N-terminal part. Within the heptads, the
amino acids are given the letters a-g. Applying this nomenclature on the LAMB1
sequence, the following amino acids appear to be shared between LAMB1 and the
HD-ZIP genes (whose corresponding position is given within brackets, Mattsson
1995): the Q at position 80 (1c), the D at position 84 (1g), EY position 86-87 (2b-c),
K position 89 (2e). The latter four amino acids are characteristic of the Class 2 HDZip genes, whereas the first two occur in all HD-Zip genes. Three additional amino
23
acids appear in a second constellation identical between LAMB1 and Class 2 HDZips, the E at position 92 (3b), the N at position 98 (4a) and L at position 101 (4d).
The significance of this similarity is not clear, but may reflect a common
evolutionary origin of these regions in LAMB1 and the HD-Zip genes. It should be
mentioned that a HD-Zip gene recently was isolated from L. annotinum (my
unpublished results).
The region C-terminal to the K-box in LAMB1 spans 855 bp, and hence it is
longer than in any other MIKC-gene known. In contrast to the C-termini of other
known MADS-box genes, it is built up by repeated sequence elements. The entire Cterminus can be divided into four regions that upon alignment at the amino acid
level show similarities to each other. These repeats 1 to 4 are of different lengths:
repeat 1 is 38 amino acids long, repeats 2 and 3 are 99 amino acids long each, and
repeat 4 is 49 amino acids long. No amino acid positions are completely conserved
among all four repeats, but all repeats share sequence similarities with one or more
of the other sequences.
Compared to plant MADS-domain proteins, long C-terminal regions are also
found in the non-plant MEF2 class of MADS domain proteins. In MEF2C this long
C-terminus has been shown to function as a transcriptional activation domain
(Molkentin et al., 1996). The C-terminus of AP1 has also been shown to function in
transcriptional activation (Cho et al.,1999), and a similar role for the LAMB1 Cterminal is therefore possible. The region is glutamine-rich, a characteristic of
certain transactivation domains (Latchman, 1998). Glutamine and asparagine
together constitute 16% of the amino acids in the C-terminal region. Acidic amino
acids comprise 14% of the C-terminal amino acids. However, preliminary
experiments examining the capacity of fusions between the GAL4 DNA-binding
domain and segments of the LAMB1 cDNA to activate transcription of the lacZ
reporter gene in a yeast system has not supported a transactivating function of the Cterminal part of LAMB1, and its function remains unknown.
3.iii Typical MIKC-genes from L. annotinum (IV)
Degenerate PCR on genomic DNA from Lycopodium annotinum was undertaken in
order to isolate additional MADS-boxes. Based on the sequence of four fragments
obtained in these degenerate PCR reactions, primers were designed and used in 3'RACE experiments. Products from these experiments, in combination with
subsequent 5'-RACE and RT-PCR experiments were used to reconstruct cDNA
sequences of five novel MIKC-genes from Lycopodium annotinum: LAMB2,
LAMB3, LAMB4, LAMB5, and LAMB6. Three of these genes, LAMB2, LAMB4, and
LAMB6, conform very closely in architecture to seed plant MIKC-genes. The Iregions of LAMB2 and LAMB6 are 34 amino acids, and that of LAMB4 is 30
amino acids long. These lengths are very similar to for example the I-region of AG,
that is 31 aa long, but different from the 57 aa I-region of LAMB1. LAMB3 and
LAMB5 represent shorter proteins, but are highly similar in sequence to LAMB2,
LAMB4, and LAMB6. The LAMB5 deduced protein ends eight amino acids
downstream the MADS-domain. The LAMB3 deduced protein is longer, and
includes a 30 aa I-region, as well as a region corresponding to the first exon of the
K-box. A phylogenetic analysis confirms that LAMB2, LAMB3, LAMB4, LAMB5,
and LAMB6 are more closely related to each other, than to other MADS-box genes.
24
3.iv The phylogenetic position of the L.annotinum MADS box-genes (III, IV)
Phylogenetic analyses suggest that LAMB1 may represent an early branch in the
phylogenetic tree of MIKC-genes. One analysis (III) is based on a nucleotide matrix
of MADS- and K-box sequences that includes a majority of the available
Arabidopsis MADS-box genes, also those encoding what subsequently was defined
as TypeI MADS-domains (Alvarez-Buylla et al., 2000). The non-plant MADS-box
genes included in this analysis were MCM1, and human and Drosophila SRF-genes,
all encoding TypeI MADS-domains. In the analysis the Arabidopsis genes encoding
a TypeI MADS-domain appear as a sister clade of the MIKC-genes. One of the Kbox lacking genes, TM021BO4.16 (=AGL39), that by Alvarez-Buylla et al., 2000,
was suggested to possibly represent a "mixed" type MADS-domain, group among
the MIKC-genes. In the tree, LAMB1 is situated at a branch between the TypeI genes
and the MIKC-genes, indicating a basal position of LAMB1 among the latter.
The phylogenetic analysis in (IV) is based on an amino acid matrix of
MADS- and K-domains. The analysis is based only on the deduced peptide
sequences from genes encoding TypeII-domains, including MEF2-genes from man
and Drosophila, as well as the yeast genes YBR1245, and RLM1. Also in this
analysis, the consensus tree of which is shown in fig. 2, LAMB1 appears at the very
base of the MIKC-genes. F28K20.7 appear at a branch, positioned between LAMB1
and an unresolved clade including the rest of the MIKC-genes. LAMB2, LAMB3,
LAMB4, LAMB5, and LAMB6 appear as a clade (the LAMB2 clade) with unresolved
relation to other MIKC-genes. That LAMB1 and the LAMB2 clade appear at
different positions in the phylogeny in relation to the fern and seed plant MIKCgenes indicates that the split between LAMB1 and the LAMB2 clade occurred in the
common ancestor of lycopsids and other vascular plants. If the position of LAMB1 at
the very base of the MIKC-gene tree is correct, it may imply that LAMB1 represents
a lineage lost in the ancestor of seed-plants, i.e. Arabidopsis, the only plant from
which a close to full set of MADS-box genes have been identified. Alternatively,
orthologs to LAMB1 exist, but the phylogenetic analyses have failed in identifying
these.
In the analyses, sequences interpreted to represent K-boxes were included
from genes that by Alvarez-Buylla et al. (2000), was interpreted not to encode Kdomains. Arguments that these genes do possess regions homologous to K-boxes are
presented in section 1.x., and in (IV). A highly derived putative K-box is found in
the gene F28K20.7.
3.v
Exon/intron organisation of LAMB2, LAMB3, LAMB4, and LAMB6 (IV)
Two RT-PCR clones of LAMB2 differed in length as compared to the other LAMB2
clones. One clone contained an insertion of 41 bp at a position two bases
downstream from the MADS-box, and the second contained a seven bp deletion at
the corresponding position. This position corresponds to the intron M, matching the
corresponding position in AG perfectly. These variants are hence likely to be
products of alternative splicing. Sequences spanning over positions corresponding to
the conserved introns were amplified from genomic DNA. However, no introns
were present in the resulting PCR-products. This apparent paradox may be explained
by the existence of both intron-less and intron containing LAMB2 copies in the
genome, the former being more easily amplified in these PCR-experiments.
25
The exon/intron organisation of the LAMB4 gene was examined in a region
spanning the 3'-end of the I-region and the entire C-region within the LAMB4
sequence. The experiments revealed the existence of six introns, A, B, C, D, E, and
F. An alignment with other MIKC-genes revealed that the positions of the introns A,
B, C, D, and E perfectly matched those of the corresponding introns in AG .
This finding is consistent with the AG-like exon/intron organisation being an
ancestral state among MIKC-genes, with important implications as regards the
significance of using exon/intron organisation as a test of the phylogenetic position
of MIKC-genes (see sections 1.ix, 1.x, and 2.v). The ancestral position of the AGpattern is further supported by the fact that introns A-E, as well as intron M, in the
phylogenetically diverse Arabidopsis genes AGL6, AGL12, AGL15, F7H19.130,
AGL20, and AGL24, match this pattern. The ancestral state of intron M is equivocal,
however, for example LAMB1 conforms to the AP1-like, rather than the AG-like,
position. The present phylogenetic analyses indicate that identical shifts in the
position of intron M have occurred several times (see fig. 1).
LAMB3 encodes a shorter protein than LAMB2, LAMB4, and LAMB6, that
does not contain a complete K-domain. Hence this gene lacks the introns B, C, D,
and E, that in other genes reside within and downstream the K-box. However, intron
A is present within the 3’-end of the I-region, at a position corresponding to that in
LAMB4, and AG.
LAMB6 differs in its structure from LAMB3 and LAMB4, but is similar to
LAMB2 in that it contains an insertion within the 3’-end of the I-region,
corresponding to four amino acids. Interestingly, intron A resides within this
insertion and is shifted eight bp toward the 5’-end relative the position in LAMB4.
3.vi A leucine-zipper in the K- and C- region of LAMB4 (IV)
In LAMB4, the last exon of the K-box, and the first exon of the C-region, both 42 bp
in length, together encode a leucine zipper-like structure. Corresponding leucine
zippers, encoded by the corresponding exons appear also to exist also in other
MIKC-genes, in e.g. AGL14, in Arabidopsis. This indicates that these two exons
form a functional unit, and that the boundary between the K-box, and the C-region is
somewhat artificially assigned, at least in this case.
3.vii Expression patterns of L. annotinum MIKC-genes (III, IV)
To study the expression of the LAMB-genes, main shoot apices, side shoot apices,
strobili, and roots were collected at intervals of 5-20 days during the growth season,
from May to September. Strobili were sectioned and examined with light
microscopy in order to determine the developmental stage of the structure of the
different dates. In June, masses of sporogeneous cells are present in the developing
sporangia. In the samples from the July 10th, different stages of meiosis could be
observed, and spore tetrads are found in samples from July 22 nd.
Since the quantities of RNA obtained from several samples were low, RTPCR was chosen to examine the expression patterns. The result showed a striking
difference between LAMB1, on the one hand, and LAMB2, LAMB4, LAMB5, as well
as LAMB6, on the other. Whereas expression of LAMB1 could be detected in
developing strobili only, the other genes showed broader expression patterns.
LAMB2, LAMB4, and LAMB6 were all more strongly expressed in vegetative
tissues, including roots, than in strobili. Especially for LAMB2, the expression in
26
strobili appeared to be low. LAMB5, however, showed the strongest signal in strobili
and main apices. The broad expression patterns of the LAMB2, LAMB4, LAMB5,
and LAMB6 can be taken to support the notion that the broad expression patterns of
MIKC-genes, represents an ancestral condition (Theißen et al., 2000). The
expression pattern of LAMB1, however, being restricted to strobili, does not only
show that MIKC-genes with a more restricted expression pattern also occur in
lycopsids. It also suggests that it may indeed be an ancestral condition, because
LAMB1 generally group at the very base of the MIKC-genes in phylogenetic
analyses.
3.iix MADS-box genes in ferns
A number of genes have been isolated from ferns, mainly from Ceratopteris
richardii and C. pteroides, which are leptosporangiate ferns, but also from the
eusporangiate fern Ophioglossum pedunculosum (Münster et al., 1997; Hasebe et
al., 1998). The Ceratopteris genes generally group into three distinct clades in
phylogenetic analyses, the CRM1, CRM3 and CRM6-like genes (Theißen et al.,
2000, note that CRM6=CerMADS2). None of the genes appear to be orthologs to
any seed plant genes, but the fern clades generally appear as dispersed among the
angiosperm clades, rather than forming a separate monophyletic group. Some
structural features have been suggested to associate certain Ceratopteris genes with
specific angiosperm genes. For example, CRM6-like genes have an N-terminal
extension (Hasebe et al., 1998; Theißen et al., 2000), as previously reported for
genes closely related to AG (see section 1.vii). This is weak phylogenetic evidence,
however, especially since DAL2 lacks such an extension. The CRM3 gene has on
similarly rather weak grounds been suggested to be related to the B-class genes,
because of its possession of a sequence element with some similarities to the
PaleoAP3 Motif (Kramer et al., 1998). Apart from the genes OPM3 and OPM4, the
Ophioglossum genes appear to be closely related to the Ceratopteris genes (Theißen
et al., 2000). In contrast to most seed plant MADS-box genes, the Ceratopteris
genes generally appear to have very broad expression patterns, including both
sporophytic, and gametophytic tissue. CRM9 and CMADS1, however, show mainly
sporophytic expression (Hasebe et al., 1998; Theißen et al., 2000), and the
expression of CMADS4 within the sporophyte is concentrated to roots (Hasebe et al.,
1998).
3.ix. Evolutionary considerations (III, IV)
The reproductive expression of LAMB1 is interesting from the perspective that many
of the Arabidopsis MIKC-genes appear to participate in processes related to floral
development, and, thus, this expression pattern may reflect a retention of an
ancestral function of MIKC-genes in reproductive development. The complex
morphologies seen in vascular plants might be regarded as structures added on the
sporophyte, from the ancestral state of a simple sporangium, as found in extant
charophycean algae (see Albert, 1999). Hence, an ancestral function in reproductive
development for a class of morphological key regulators in land plant sporophytes is
an attractive hypothesis. However, LAMB1 does not appear to be more closely
related to angiosperm MIKC-genes with specific functions in reproductive
development, than to genes involved in other developmental processes. On the
contrary, LAMB2, LAMB3, LAMB4, LAMB5, and LAMB6 may be more closely
27
related to the floral homeotic genes in Arabidopsis reproductive development, than
LAMB1. Hence, the hypothesis that the angiosperm MIKC-genes, involved in
reproductive development represents a retention of a primitive (though diversified)
function shared with LAMB1, is at the present equivocal. An alternative hypothesis
states that a broad expression pattern is a primitive feature of MIKC-genes, as have
been previously proposed based upon the fern MIKC-gene expression patterns
(Theißen et al., 2000). According to this hypothesis, the function of the MIKC-genes
have evolved from controlling more general developmental processes, to more
restricted ones, as the gene family has diversified (Theißen et al., 2000).
Furthermore, the fossil evidence suggests that the earliest vascular plants
may have had a close to isomorphic life cycle (section 3.ii). It is hence very likely
that sporophyte developmental programs were recruited from gametophyte specific
programs in the common ancestor of extant charophycean algae, bryophytes, and
vascular plants rather than invented exclusively within the sporophyte context. This
implies that to a large extent, the same key regulatory genes probably were active in
both the gametophyte and sporophyte of primitive vascular plants. The expression
pattern in both phases of the life cycle of Ceratopteris may be a retention of this
ancestral condition. If so, it is likely that also the L. annotinum MIKC-genes are
expressed in the gametophyte generation. The species Lycopodium cernuum and
Lycopodium obscurum have spores that are easy to germinate (Bruce, 1976;
Whittier, 1976), and hence would provide an oppurtunity for studying the
gametophyte generation.
4. Homeotic genes and organ homology
The expression patterns of MADS-box genes have been proposed to reflect organ
homology (e.g. Doyle, 1994; Albert et al., 1998; Winter et al., 1999; Theißen et al.,
2000). In this section, I will discuss this idea, first in general terms, and then its
relation to the data presented in this work.
The concept of homology was first put forward by Richard Owen in the 19th
century. Owen belonged to a school of morphologists that claimed that different
morphologies are functional variations, of a non-functional archetypal idea,
manifesting itself as "homology" between organs (see e.g. Young, 1992; Gould,
1993). Johann Wolfgang von Goethe, some decades earlier, supported a similar idea,
that all the organs in a flower, the sepals, petals, stamens and carpels constitute
modified leaves, by observations of homeotic mutants (see Singer, 1959). Thus,
homologies were not only seen between organisms, but also, in serial homology,
between structures within a single organism. Since Darwin (1859), homologies are
defined as similarities in structure caused by a common ancestry (as opposed to
structures being similar due to independent adaptation to a similar environment). In
cladistic terms, homology is equivalent to synapomorphy. Homologies are
corroborated when intermediate stages between structures, predicted from the
homology proposition, are found, for example in the fossil record. A classical
example is the intermediate forms between the reptilian jaw joint and the
mammalian middle ear (Crompton and Parker, 1978). Various assignments of
homology, especially in plants, have been highly controversial, for example the
integuments of the ovule, the derived structures of grass embryos, the angiosperm
floral organs (sepals, petals, stamens, and carpels), or the microphylls of lycopsids.
28
AGL14
CLADE
AP1/
AGL6 / AGL9
CLADE
AGCLADE
B-CLASS
CLADE
AGL17
CLADE
YBR1245
RLM1
d MEF2
MEF2A
LAMB1
LAMB2
LAMB5
LAMB6
LAMB3
LAMB4
CMADS1*
CerMADS3*
CerMADS2*
DAL3***
AGL14
F7H19.130
AGL20
CerMADS1*
CRM1*
CMADS2*
CMADS4*
CRM3*
DAL1***
GGM9**
GGM11**
AGL3
AGL9
DEFH72
DEFH200
AGL2
AGL4
DEFH49
AGL6
AGL13
AGL8
AP1
CAL
SQUA
DAL2***
GGM3**
AG
AGL1
AGL5
PLE
FAR
AGL11
DAL11***
DAL13***
GGM2**
AP3
DEF
PnAP3.2
PI
GLO
PnPI.1
DAL12***
GGM13**
FLF
F22K20.15
AGL12
AGL15
AGL17
F20D10.60
ANR1
DEFH125
F2809.80
F14M13.6
AGL24
F2809.240
F28K20.7
Figure 2. A phylogenetic analysis on the deduced peptide sequences of MADS-box
genes from Saccharomyces cerevisiae (YBR 1245 and RLM1), Homo sapiens
(MEF2) and Drosophila melanogaster (dMEF2). Lycopodium annotinum genes are
marked with bold face. *) MADS-box genes from Ceratopteris richardii. **)
MADS-box genes from the gymnosperm: Gnetum gnemon. The Picea abies genes
are bold and marked with ***. This analysis is presented in (IV).
29
4.i.
Homology as a neo-Darwinian paradigm
The neodarwinian model of evolution, which implies that new form evolves through
gradual modification of pre-existing form, suggests that homologies should be
possible to identify. If we do not find the intermediate forms, this might be
explained to an incompleteness of the fossil record. Several authors (e.g. McKinney
and McNamara, 1991) have proposed that most morphological evolution in animals
could be attributed to changes in the relative growth rate of body parts
(heterochrony), a hypothesis consistent with homologous structures gradually
changing in form. The hypothetical processes suggested to explain evolutionary
changes in plant form (overtopping, recurvation, planation etc.) in the telome theory
(Zimmermann, 1965) are conceptually similar. If, on the other hand, evolutionary
novelties originated in a non-Darwinian, saltatory manner, structures could originate
without being homologous to a pre-existing structure. This has sometimes been
proposed to be of importance in "macro evolution". For example, Theißen et al.
(2000), in their review on MADS-box gene evolution, assumes that "novel structures
or complete new body plans" have originated in a non-Darwinian manner.
4.ii
Homeotic genes and homology
If evolutionary changes at the morphological level are due to differences in relative
growth of body parts, most evolutionary changes at the genetic level would involve
regulatory genes affecting these processes in a quantitative manner. This probably
involves mutations at cis-regulatory elements causing quantitative changes in gene
expression, most likely confined to an organ specific module (see section 1).
Changes often referred to as heterochrony, resulting in different proportions of the
body in animals, or branching patterns in plants, may involve genes affecting a
larger part of the organism, however. Such changes are likely to also involve
expression levels of proteins directly participating in hormone production/sensation,
and not only transcription factors.
Some components of a specific module are likely to be highly conserved
among species, particularly those that are important for the dissociation of one
module from other modules. Therefore these components may serve as indicators of
modular/organ homology, and may reveal organ homology between highly diverged
structures. Candidates for such components are the homeotic genes encoding
transcription factors that determine organ identity. Homologous homeotic genes
would thus be expressed in homologous organs. This logic is evidently applicable to
orthologous genes, but it may as well be significant for paralogous genes. Serially
homologous organs, or “paralogous” organs may be developmentally regulated by
paralogous (or the same) genes.
4.iii Homeotic gene expression does not define morphological homology
The use of homologous genes to address organ homology may not always be
straightforward. Homologous regulatory genes may act in homologous processes in
non-homologous structures. This is illustrated by the orthologs to the gene distalless being involved in the formation of non-homologous appendages in vertebrates,
arthropods and echinoderms (Abouheif et al., 1997; Panganiban et al., 1997). So,
even if distal-less may have a conserved, homologous, function in the formation of
protruding structures, it can not be used to identify whether particular appendages
are homologous.
30
Pax-6 homologs have been shown to regulate eye formation in both insects
and vertebrates. This might support homology between all eyes. Gehring and Ikeo
(1999) propose, in a discussion of general interest, that the evolution of eyemorphologies, have occurred through the intercalation of new developmental steps
in an ancestral simple regulatory pathway in which Pax-6 control rhodopsinformation. Obviously, these intercalated steps would be different in different
evolutionary branches. This clearly illustrates that an organ might be homologous at
one level, but not another, and that use of homeotic genes to address homologies
must be precise with respect to this level (for a discussion, see also Bolker and Raff,
1996). For example, the involvement of Pax-6 homologs in formation of eyes in
cephalopods and vertebrates does not indicate that the camera architecture of the
two eyes is due to homology.
The notion that a number of developmental regulatory pathways show
redundancy (e.g. Pickett et al., 1995; Tauz, 1992) indicates that it is possible for a
regulatory gene to be replaced by a non-homologous in evolution, via a state where
they exist in parallel. The partially redundant interaction between transcription
factors of unrelated types in the A-function and the determination of floral meristem
identity, LFY, AP2, AP1, may represent a system with the potential to evolve in this
manner. The existence of overlapping regulatory pathways in plants may be
explained by the partly similar, partly distinct responses by plant development on
environmental factors. If this explanation is correct, one might expect redundancy to
be more frequent in developmentally plastic processes, such as onset of flowering,
than in processes defining non-plastic traits, such as flower morphology. This
hypothesis has not been tested. It is tempting to speculate that the A-function appear
to be less conserved among angiosperm species (see e.g. Theißen et al., 2000), than
the B- and C-functions, because it is more closely associated with the plastic process
of the onset of flowering.
Homeotic changes have generally been considered not to be an important
process in evolution, because such changes often have deleterious effects on the
survival of the individual. Using a very broad definition of homeosis, including also
cell-types, or physiological features (Sattler, 1988), homeosis has already been
suggested to be a component in the evolution of novelties (as in the incorporation of
sub modules within a module). However, it is very likely that homeotic changes in
the more strict sense also have played some role in evolution. Homeosis may result
in "hybrid", "chimeric" or "intermediate" organs making assignments of homology
between structures difficult or even pointless (Sattler, 1988; 1991). Homeosis does
not necessarily mean saltational, non-Darwinian evolution, however. In certain
instances, a homeotic change might represent a slight morphological change in an
adaptive perspective, and the formation of "intermediate" organs is conceivable as a
gradual process. One example from the animal kingdom is the demonstration by
Averof and Patel (1997) that the evolution of thoracic limbs into feeding appendages
in crustaceans correlates with a reduced expression pattern of the homeotic genes
Ubx and AbdA.
Because of the morphological simplicity and great plasticity of plant
development, rearrangements between plant parts probably have a higher chance of
being maintained in evolution compared to animals. Hempel et al. (1998) showed
that primordia on the Arabidopsis stem, are not strictly committed to, but rather
biased toward a particular developmental state. Continuous light was shown to be
31
able to induce floral development of already initiated primordia, that would
otherwise form side-shoots at such a late stage, that chimeric structures were
formed. Tomato plants grown at low temperature display floral homeotic changes,
coupled to changes in floral homeotic MADS-box gene expression levels (Lozano et
al., 1998). These two examples demonstrate that homeosis may lie within the
reaction norm of wild type plant development.
The perianths of lilies (and several other monocotyledons) that contain two
whorls of petal like organs (tepals) have probably evolved through the extension of
the B-function into the first whorl (Theißen et al., 2000). In Rumex (a eudicot) the
perianth has undergone the opposite transition into two sepaloid whorls, by
withdrawal of the B-function from whorl two (Ainsworth et al., 1995). Hence both
the lily, and the Rumex perianth appear to have evolved through homeosis (Theißen
et al., 2000).
A striking example of cooption of a MADS-box gene to an apparently
unrelated function must be mentioned; NMH7 from alfalfa (Medicago sativa). This
gene is phylogenetically nested within the homeotic B-class genes (Kramer et al.,
1998; Theißen et al., 1996; II), and has a consistent exon/intron organisation (acc
no. AF042068), but is involved in root-nodule formation (Heard and Dunn, 1995), a
structure apparently with no homology to floral organs. It would be of great interest
to see if there are any similarities in the downstream genes being regulated by
NMH7 and by floral B-class genes.
Gene-expression data in questions of morphological homology should be
used as a tool for proposing/testing hypothesis, not as a definition. If two conflicting
hypothesis state that structure A is homologous to structure B or structure C,
respectively, the expression of an orthologous homeotic gene in A and B, but not in
C, would support that A is homologous to B, but not to C.
4.iv Homologies between reproductive structures of seed plants
Pollen sacks and ovules in angiosperms are born on stamens and carpels
respectively, both within the flower, while in conifers they are born on separate
cones. The female cones of conifers consist of spirally arranged ovuliferous scales,
each subtended by a bract. This organisation suggests that the bract may be a leaf,
and the ovuliferous scale a branch in its axil. Florin (1951) supported such a view,
by proposing homology between the ovuliferous scale with the ovuliferous branches
of the extinct Cordaitales, a plant group that is known from the Carboniferous and
Permian. The cordaitean female cones consisted of two orders of branches, of which
the second bore ovules. According to Florin the ovuliferous scale had evolved from
secondary branches that had undergone reduction in the number of ovules, from
several to e.g. two. Recurvation of the ovules made their micropyle point toward,
rather than away from, the primary axis. Finally, the fusion of the constituent parts
of the secondary branch made it condense and scale-like, rather than forming a loose
branching structure. In the Voltziales of the Carboniferous-Triassic periods, several
possible intermediate stages in the evolution of a branching system to an ovuliferous
scale were recognised.
The male conifer cones, on the other hand, consist of a simple determinate
shoot. A bract subtends this shoot that bears spirally arranged microsporophylls.
Since micro- and megasporangia (i.e. pollen and seed) are derived from a single
monosporangiate structure, seed- and pollen cones may be homologous at some
32
level. The male cone, because of the branching order, have been proposed to be
homologous to the ovuliferous scale, rather than to the entire female cone (Banks,
1972).
The origin of the angiosperm flower has remained obscure. It has not been
possible to test the different variants and combinations of the two main hypothesis,
whether the stamens and carpels are simple sporophylls or compound branching
systems (the euanthium vs. pseudanthium hypothesis, see Friis and Endress, 1990,
for a review).
4.v
B- and C genes and seed plant reproductive organ homologies (I, II)
The discussion above on the B- and C-functions being conserved between
angiosperms and conifers is based on their expression in organs on which
homologous structures form; pollen and ovules respectively. The notion of a
conserved, homologous function, is not based upon a proposal of homology between
the organs per se.
The expression of DAL11, DAL12, and DAL13 in the male cone, and PI and
AP3 orthologs in developing angiosperm stamens indicate homology between these
organs. Because these DAL genes differ in their exact expression patterns (see
above), the data do not provide suggestions as to whether the stamen is homologous
to the microsporophyll, or to the entire male cone. Thus, the expression patterns do
not provide arguments in the pseudanthium/euanthium controversy.
The expression of DAL2 in the ovuliferous scale (II) might be taken as an
indication of homology between this structure and the angiosperm carpel. This
would support the pseudanthial theory of carpel origin: the carpel being homologous
to a branching structure. The interpretation is problematic, however, because of the
diversification of the AG-family within angiosperms (see above).
Extending the comparison to serial homologies, the expression of DAL2 in
the microsporophyll of the male cone (II), and in the ovuliferous scale of the female
cone (I), might be taken to suggest homology between these structures. This is in
apparent conflict with the ovuliferous scale being homologous to the entire male
cone, rather than to the microsporophyll. However, late DAL2 expression is
strongest in the part of the ovuliferous scale on which ovules are to be formed (I),
and this may be the region homologous to the microsporophyll. According to one,
highly speculative, interpretation consistent with both Florin and this homology
proposition based on the DAL2 expression, the ovuliferous scale may be a fused
organ between a branching structure and a sporophyll (II).
If the ancestral function of a C-class gene, or C/D-class gene (Theißen et al.,
2000) is to promote ovule development (or pollen development in concert with Bclass genes) on the developing organ on which it is expressed, evolutionary change
in the expression pattern of the C-class gene may confer this property to an other
part of the plant. Such a homeotic change needs not to be deleterious to the plant,
and may have played a role in the evolution of seed plant reproductive structures. A
disjunction between orthologous gene expression and organ homology is
conceivable in this case, i.e. the pollen- and ovule bearing organs of conifers and
angiosperms may well be non-homologous, although they express particular
orthologous homeotic genes.
33
4.vi ABC-genes and floral organ homology
Inference of serial homology within the angiosperm flower from MADS-box gene
expression would suggest that stamens and carpels are homologous (C-function),
that stamens and petals are homologous (B-function) and that petals and sepals are
homologous (A-function). Homology, at some level of organisation, between
stamens and carpels, like the male and female cone, is reasonable. Micro- and
macrosporangia of seed plants are likely to be derived from one single type of
sporangium in a homosporous ancestor. Fossil progymnosperms support this
hypothesis (Stewart and Rothwell, 1993).
A- and B-gene expression, indicate that core eudicot petals may be
homologous to both sepals and stamens, and hence may be "hybrid" organs. Baum
(1998; 1999) presented a similar hypothesis, that from a primitive flower with a
sepal-like perianth, petals may have evolved through the extension of B-function
into the perianth. Albert et al. (1998) state that petals are homologous to both sepals
and stamens, but not clearly as a hypothesis of evolution (they use their term
"process homology"). Instead, they speculate that the basal angiosperm flower had a
simple, petaloid perianth expressing both A- and B-genes. It now appears as if the
genus Amborella is the most basal angiosperm identified, followed by water lilies,
Nymphaeaceae (Matthews and Donoghue, 1999; Soltis et al., 1999; Qiu et al.,
1999). This supports a simple perianth as a basal condition among angiosperms, and
in at least Nymphaeaceae it is clearly petaloid.
The hypothesis that some basal angiosperms may have petals derived from
sepaloid organs (bracteopetals, Thaktajan, 1991), but unrelated to stamens, would be
corroborated if an absence of B-function in these could be proven.
B-mutants in the grasses, maize and rice, conform to the ABC-model
(Schmidt et al., 1998; Kang et al., 1998; Theißen et al., 2000). This is consistent with
homology between the derived structures of the grass flower, the palea and the
lodicule, and sepals and petals respectively. It further suggests that the ABC-system
was established before the split between eudicots and monocots.
A mutant form of Anemone nemorosa fails to form petals and stamens, but
instead cauline leaves are present in their place (Svedelius, 1909). This is probably a
B-mutant (Svensson, 1996), with the cauline leaves behaving like sepals. These
cauline leaves may be homologous to sepals, determined by an A-function organ
identity gene. These sepals may have become enlarged, possibly by recruiting
developmental pathways specifically involved in the formation of vegetative leaves,
enabling floral photosynthesis in this rhizomic plant. Alternatively, the organs may
be non-homologous to sepals, but are ordinary vegetative leaves, that form
ectopically in the mutant flower in the absence of other organ identity signals.
Detailed study of this, and other known mutant forms of A. nemorosa, may provide
information on the ABC-system within basal eudicots, which recently has been
given some attention by expression studies of B-class genes (Kramer and Irish,
1999).
The B-class gene AP3 is expressed in the integuments (see section 1.xi.). The
significance of a homology proposal based on this expression patterns depends on
whether the regulatory function of AP3 in ovule development is distinct from, or
related to that in flower development. If the latter is the case this may be a reflection
of flowers and ovules having evolved as developmental modules interconnected as a
reiterative series (see Albert et al., 1998).
34
4.vii Lycopsid microphyll homology
Homology between microphylls of lycopsids and megaphylls (leaves of ferns and
seed plants) is in conflict with their phylogenetic distribution (Crane and Kenrick,
1996; Kenrick and Crane, 1997). Microphylls are generally believed to have evolved
from enations, as found on the stems of several zosterophyll genera (Bower 1935,
reviewed in e.g. Gifford and Foster, 1989; Stewart and Rothwell, 1993; Raven et al.,
1999). A recent hypothesis is that microphylls evolved from sporangia. Microphylls
and sporangia, unlike enations, show similar distribution on the stem and are
vascularized. Furthermore, enations are not found on the zosterophylls (a
paraphyletic group) most closely related to lycopsids (Crane and Kenrick, 1996;
Kenrick and Crane, 1997). The hypothesis that microphylls and sporangia are
homologous appears to be the most likely at the present.
Studies on genetic mechanisms for organ identity determination may
contribute to the evaluation of these conflicting hypothesis. For example, is the
development of microphylls controlled by genes homologous to those that control
development of true leaves or to those that control sporangial development?
5. Concluding remarks and suggestions
The present data it suggests that the floral homeotic B- and C-class genes emerged
in the common ancestor of all seed plants, but after this lineage split from ferns, the
closest extant relatives of seed plants. The expression patterns of the conifer B-class
orthologs DAL11, DAL12, DAL13 are restricted to male cones. The conifer C-class
ortholog DAL2 is expressed in both male and female cones. These data suggest that
the function of B- and C-class genes as key regulators of male- and female
sporophytic structures respectively, was established in the common ancestor of
extant seed plants.
The structurally divergent MIKC-gene LAMB1, isolated from the club moss,
Lycopodium annotinum, may represent one of the earliest lineages in the MIKCgene family. This gene has an expression pattern restricted to strobili, suggesting
that a function in reproductive development may have been ancestral for MIKCgenes. The genes LAMB2, LAMB4, LAMB5, and LAMB6 from L. annotinum have a
more typical MIKC-organisation. These genes have broad expression patterns, like
MIKC-genes from ferns. It remains unclear whether the reproductive expression
patterns of floral homeotic genes represent an ancestral condition shared with
LAMB1, or whether the valid hypothesis is that the diverse functions of MIKC-genes
are derived from more general developmental roles, reflected by broad expression
patterns.
Different clades of MIKC-genes have diagnostic details in their exon/intron
organisation. DAL2 conforms to other C-class genes (e.g. AG) in this respect.
However, also DAL11 and DAL13 conform to this pattern, a finding that is not in
conflict with the association of these genes to the B-class genes, if the pattern is
ancestral. LAMB4 has an exon/intron organisation that in its details resembles that of
AG, indicating that this organisation might be ancestral.
The basal branching pattern of MIKC-genes is equivocal, perhaps due to a
very rapid rate of gene duplication events occurring simultaneously with the
separation of the major lineages of land plants. Isolation and characterisation of
MIKC-genes from lineages branching off earlier in the evolution of plants, such as
35
bryophytes and charophycean algae, may help both to resolve the phylogeny and to
address the structural uniqueness of LAMB1. Isolation of genes from other
lycopsids, such as Selaginella that belong to a lineage separated from Lycopodium
since the Devonian (Stewart and Rothwell, 1993) may also be informative. The
basal branching pattern of the MADS-box gene family as a whole is also
problematic: for example, TypeI MADS-box genes need to be isolated from more
plant species, and protists need to be examined.
Within a not too far future, we anticipate the genomes, or very inclusive
cDNA-collections, of several grasses (maize, rice, and other crop species),
commercially important eudicots, including a range of floral morphologies and
growth habits, from weeds to trees. Phylogenetically important species may be
thoroughly investigated because their simple morphologies make them useful model
systems, such as the moss Physcomitrella, the liver wort, Marchantia, or algae, such
as Chlamydomonas and Volvox. Hence we might expect as large samples of MADSbox genes, from several species, as at the present is only available in Arabidopsis.
Furthermore, expression studies of these genes, and other functional analyses, will be
conducted at a very different scale. The study of evolution of developmental
mechanisms and its connection to the origin of morphological novelties is a field
that will be understood at much greater depth than at the present.
36
Acknowledgements
I would like to thank professor Peter Engström for taking me on as a PhD student
and sustaining the funding of the projects.
I deeply want to thank Charlotta Thornberg for excellent technical assistance and
fikapauser.
I am also greatful to Afsaneh, Eva and Marie, who have been sequencing quite a lot
of samples from me. Thanks!
Jan Bergsten, Anders Rådén and Anders Lindholm, thanx for driving the car to the forest!
All the Funky Lab/ MADS-people are thanked:
Karolina Tandre, Annelie Carlsbecker (thanks a lot for the support!), Jens
Sundström, Anders Kvarnheden, Victor Albert, Deyue Yu, Liz Izquerdo, Henrik Johannesson,
Yan Wang, and Jelka L.
All the other PhD students are of course warmly thanked for putting some color to
the world and so on:
Alessia, Anna, Coby, Eva, Joel, Johannes, Katarina, Mattias (thanks for support and
interesting discussions, and the very much needed boat trip), Ingela, and Sandra.
I also thank Annika Sundås-Larsson and Eva Sundberg, as well as…
…I simply thank the present and former workers at the department, in particular
Cellskapet. All my friends at the other departments at EBC are warmly thanked.
Birgitta is thanked for taking care of us all.
Gun-Britt is thanked for her contribution to the courses.
Agneta is thanked for taking care of e.g. the shopping list.
Marie Svenson, sorry Marie Englund, is thanked for always being nice and supporting.
Thanks to Jin-Long for the calligraphy kit!
Thanks to Tage for his interest in knowledge.
Björn and Kauko, thank you.
I thank my long time biology friends,
Tim Olsson (thanks for all interesting e-mails!), Olle Israelsson (thanks for the
Xenoturbella-adventures), Anders Normann, and Anders Lindroth.
I thank all my new and old friends, and my family…
This work is dedicated to the memory of my mother, Gunnel Svensson.
37
REFERENCES
Abouheif, E., Akam, M., Dickinson, W. J., Holland P. W. H., Meyer, A., Nipam,
H.P., Raff, R. A., Roth, V. L., Wray, G. A. (1997) Homology and
developmental genes. Trends Genet. 13:432-433.
Ainsworth, C., Crossley, S., Buchanan-Wollaston, V., Thangavelu, M., Parker, J.
(1995) Male and female flowers of the dioecious plant sorrel show different
patterns of MADS box gene expression. Plant Cell 7:1583-1598.
Akam, M., Averof, M., Castelli-Gair, J., Dawes, R., Falciani, F., Ferrier, D. (1994)
The evolving role of Hox genes in arthropods. Dev Suppl 1994:209-215.
Albert, V. A. (1999) Shoot apical meristems and floral patterning: an evolutionary
perspective. Trends Plant Sci 4:84-86.
Albert, V. A., Gustafsson, M. H. G., Di Laurenzio, L. (1998) Ontogenetic
systematics, molecular developmental genetics, and the angiosperm petal. In:
Soltis DE, Soltis PS, Doyle JJ (eds), Molecular Systematics of Plants II, pp.
349–374. Kluwer Academic Publishers, Boston.
Alvarez-Buylla, E. R., Pelaz, S., Liljegren, S. J., Gold, S. E., Burgeff, C., Ditta,
G. S., Ribas de Pouplana, L., Martinez-Castilla, L., Yanofsky, M. F. (2000)
An ancestral MADS-box gene duplication occurred before the divergence of
plants and animals. Proc. Natl. Acad. Sci. U S A., 97:5328-5333.
Angenent, G. C., Busscher, M., Franken, J., Mol, J. N., van Tunen, A. J. (1992)
Differential expression of two MADS box genes in wild-type and mutant petunia
flowers. Plant Cell 4:983-993.
Angenent, G. C., Franken, J., Busscher, M., Weiss, D., Van-Tunen, A. J. (1994)
Co-suppression of the petunia homeotic gene fbp2 affects the identity of the
generative meristem. Plant J. 5:33-44.
Angenent, G.C., Franken, J., Busscher, M., van Dijken, A., van Went, J. L.,
Dons, H. J., van Tunen, A. J. (1995) A novel class of MADS box genes is
involved in ovule development in petunia. Plant Cell 7:1569-1582.
Averof, M., Patel, N. H. (1997) Crustacean appendage evolution associated with
changes in Hox gene expression. Nature 388:682-686.
Banks, H. P. (1972) Evolution and plants of the past. Macmillan press, London.
Baum, D. A. (1998) The evolution of plant development. Curr. Opin. Plant Biol.
1:79-86.
Baum, D. A., Whitlock, B. A. (1999) Plant development: Genetic clues to petal
evolution. Curr. Biol. 1999:R525-527.
Bolker, J. A., Raff, R. A. (1996) Developmental genetics and traditional homology.
Bioessays. 18:489-494.
Bower, F. O. (1935) Primitive land plants. London: Macmillan.
Bowman, J. L., Smyth, D. R., Meyerowitz, E. M. (1991) Genetic interactions among
floral homeotic genes of Arabidopsis. Development 112:1-20.
Bowman, J. L., Alvarez, J., Weigel, D., Meyerowitz, E. M., Smyth, D. R. (1993)
Control of flower development in Arabidopsis thaliana by APETALA1 and
interacting genes. Development 119:721-743.
Bradley, D., Carpenter, R., Sommer, H., Hartley, N., Coen, E. (1993)
Complementary floral homeotic phenotypes result from opposite orientations of a
transposon at the plena locus of Antirrhinum. Cell 72:85-95.
38
Bruce, J. G. (1976) Gametophytes and subgeneric concepts in Lycopodium. Amer.
J. Bot. 63:919-924.
Cho, S., Jang, S., Chae, S., Chung, K. M., Moon, Y. H., An, G., Jang, S. K.
(1999) Analysis of the C-terminal region of Arabidopsis thaliana APETALA1 as a
transcription activation domain. Plant Mol. Biol. 40:419-429.
Coen, E.S., Meyerowitz, E.M. (1991) The war of the whorls: genetic interactions
controlling flower development. Nature 353:31-37.
Colombo, L., Franken, J., Koetje, E., van Went, J., Dons, H. J., Angenent, G.
C., van Tunen, A. J. (1995) The petunia MADS box gene FBP11 determines
ovule identity. Plant Cell 7:1859-1868.
Crane, P. R., Kenrick, P. (1997) Diverted development of reproductive organs: a
source of morphological innovation in land plants. Pl. Syst. Evol. 206:161-174.
Crompton, A.W., Parker, P. (1978). Evolution of the mammalian masticatory
apparatus. American Scientist. 66:192-201.
Davies B., Egea-Cortines, M., de Andrade Silva, E., Saedler, H., Sommer, H.
(1996) Multiple interactions amongst floral homeotic MADS box proteins.
EMBO J. 15:4330-4343.
Davies, B., Motte, P., Keck, E., Saedler, H., Sommer, H., Schwarz-Sommer, Z.
(1999) PLENA and FARINELLI: redundancy and regulatory interactions between
two Antirrhinum MADS-box factors controlling flower development. EMBO J.
18:4023-4034.
Doebley, J., Lukens, L. (1998) Transcriptional regulators and the evolution of plant
form. Plant Cell 10:1075-1082.
Doyle, J.J. (1994) Evolution of a plant homeotic multigene family: toward
connecting molecular systematics and molecular developmental genetics. Syst.
Biol. 43:307-328.
Dubois, E., Bercy, J., Messenguy, F. (1987) Characterization of two genes, ARGRI
and ARGRIII required for specific regulation of arginine metabolism in yeast.
Mol. Gen. Genet. 207:142-148.
Duff, R. J., Nickrent, D. L. (1999) Phylogenetic relationships of land plants using
mitochondrial small-subunit rDNA sequences. Am. J. Bot. 86:372-386.
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.
18:5370-5379.
Endress, P. K. (1994) Floral structure and evolution of primitive angiosperms:
recent advances. Pl. Syst. Evol. 192:79-97.
Fan, H. Y., Hu, Y., Tudor, M., Ma, H. (1997) Specific interactions between the Kdomains of AG and AGLs, members of the MADS domain family of DNA
binding proteins. Plant J. 12:999-1010.
Ferrandiz, C., Gu, Q., Martienssen, R., Yanofsky, M. F. (2000) Redundant
regulation of meristem identity and plant architecture by FRUITFULL,
APETALA1 and CAULIFLOWER. Development. 127:725-734.
Fisher, R. A. (1930) The genetical theory of natural selection. Calerdon press,
Oxford.
Fitch, W. M. (1970) Distinguishing homologous from analogous proteins.
Systematic Zoology 19:99-113.
39
Flanagan, C. A., Ma, H. (1994) Spatially and temporally regulated expression of
the MADS-box gene AGL2 in wild-type and mutant Arabidopsis flowers. Plant
Mol. Biol., 26:581-95.
Flanagan, C. A., Hu Y., Ma H. (1996) Specific expression of the AGL1 MADSbox gene suggests regulatory functions in Arabidopsis gynoecium and ovule
development. Plant J. 10:343-353.
Florin, R. (1951) Evolution in cordaites and conifers. Acta Horti Bergiani 15:285388.
Friis, E. M., Endress, P. K. (1990) Origin and evolution of angiosperm flowers. In:
Advances in botanical research. Ed. J. A. Callow. 17:100-162.
Gehring, W. J. (1994) A history of the Homeobox. In: Denis Duboule (ed.)
Guidebook to the homeobox genes. pp. 3-9. Oxford university press, New York.
Gehring, W. J., Ikeo K. (1999) Pax 6: mastering eye morphogenesis and eye
evolution. Trends Genet. 15:371-377.
Gifford, E.M., Foster, A. S. (1989) Morphology and evolution of vascular plants.
3rd ed. New York: Freeman.
Goto, K., Meyerowitz, E. M. (1994) Function and regulation of the Arabidopsis
floral homeotic gene PISTILLATA. Genes Dev 8:1548-1560.
Gould, S. J. (1992) Ontogeny and phylogeny, revisited and reunited. Bioessays.
14:275-279.
Gould, S. J. (1993) Eight little piggies, reflections in natural history. W. W. Norton
& Company, New York, London.
Gu, Q., Ferrandiz, C., Yanofsky, M. F., Martienssen, R. (1998) The FRUITFULL
MADS-box gene mediates cell differentiation during Arabidopsis fruit
development. Development. 125:1509-1517.
Hareven, D., Gutfinger, L., Pnueli, L., Bauch, L., Cohen, O., Lifschitz, E.(1994)
The floral system of tomato. Euphytica 79:235-243.
Hasebe, M., Wen, C. K., Kato, M., Banks, J. A.. (1998) Characterization of
MADS homeotic genes in the fern Ceratopteris richardii. Proc. Natl. Acad. Sci.
U S A. 95:6222-6227.
Heard, J., Dunn, K. (1995) Symbiotic induction of a MADS-box gene during
development of alfalfa root nodules. Proc. Natl. Acad. Sci. U S A 92:5273-5277
Hempel, F. D., Zambryski, P. C., Feldman, L. J. (1998) Photoinduction of flower
identity in vegetatively biased primordia. Plant Cell 10:1663-1676.
Huang, K., Louis, J. M., Donaldson, L, Lim, F. L, Sharrocks, A. D., Clore, G.
M. (2000) Solution structure of the MEF2A-DNA complex: structural basis for
the modulation of DNA bending and specificity by MADS-box transcription
factors. EMBO J. 19:2615-2628.
Huang, H., Tudor, M., Su, T., Zhang, Y., Hu, Y., Ma, H. (1996) DNA binding
properties of two Arabidopsis MADS domain proteins: binding consensus and
dimer formation. Plant Cell 8:81-94.
Huang, H., Tudor, M., Weiss, C. A., Hu, Y., Ma, H. (1995) The Arabidopsis
MADS-box gene AGL3 is widely expressed and encodes a sequence-specific
DNA-binding protein. Plant Mol. Biol. 28:549-567.
Huijser, P., Klein, J., Lonnig, W. E., Meijer, H., Saedler, H., Sommer, H. (1992)
Bracteomania, an inflorescence anomaly, is caused by the loss of function of the
MADS-box gene squamosa in Antirrhinum majus. EMBO J. 11:1239-1249.
40
Jack, T., Brockman, L. L., Meyerowitz, E. M. (1992) The homeotic gene
APETALA3 of Arabidopsis thaliana encodes a MADS box and is expressed in
petals and stamens. Cell 68:683-697.
Jack, T., Fox, G. L., Meyerowitz, E. M. (1994) Arabidopsis homeotic gene
APETALA3 ectopic expression: transcriptional and posttranscriptional regulation
determine floral organ identity. Cell 76:703-716.
Kang, H. G., Jeon, J. S., Lee, S., An, G. (1998) Identification of class B and class
C floral organ identity genes from rice plants. Plant Mol Biol. 38:1021-1029.
Kater, M. K., Colombo, L., Franken, J., Busscher, M., Masiero, S., Van
Lookeren Campagne, M., Angenent, G. C. (1998) Multiple AGAMOUS
homologs from cucumber and petunia differ in their ability to induce
reproductive organ fate. Plant Cell 10:171-182.
Kempin, S. A., Savidge, B., Yanofsky, M. F. (1995) Molecular basis of the
cauliflower phenotype in Arabidopsis. Science 267:522-525.
Kenrick, P., Crane, P. R. (1997) The origin and early diversification of Land
Plants. Smithsonian institution press, Washington and London.
Kimura, M. (1977) Preponderance of synonymous changes as evidence for the
neutral theory of molecular evolution. Nature 267:275-276.
Kramer, E. M., Dorit, R. L, Irish, V. F. (1998) Molecular evolution of genes
controlling petal and stamen development: duplication and divergence within the
APETALA3 and PISTILLATA MADS-box gene lineages. Genetics. 149:765-83.
Kramer, E. M., Irish, V. F. (1999) Evolution of genetic mechanisms controlling
petal development. Nature 399:144-148.
Krizek, B. A., Meyerowitz, E. M. (1996) The Arabidopsis homeotic genes
APETALA3 and PISTILLATA are sufficient to provide the B class organ identity
function. Development 122:11-22.
Kukkonen, I., (1994) Notes on the treatment of the family Lycopodiaceae for Flora
Nordica. Ann. Bot. Fenn. 31:197-202.
Kush, A., Brunelle, A., Shevell, D., Chua, N. H.. (1993) The cDNA sequence of
two MADS box proteins in Petunia. Plant Physiol. 102:1051-1052.
Latchman, D. S. (1998). Eukaryotic Transcription Factors. 3d ed. Academic Press,
London.
Lewis, E. B. (1978) A gene complex controlling segmentation in Drosophila.
Nature 276:565-570.
Liljegren, S. J., Ditta, G. S., Eshed, Y., Savidge, B., Bowman, J. L., Yanofsky,
M. F. (2000) SHATTERPROOF MADS-box genes control seed dispersal in
Arabidopsis. Nature 404:766-770.
Lönnig, W. E., Saedler, H. (1994) The homeotic Macho mutant of Antirrhinum
majus reverts to wild-type or mutates to the homeotic plena phenotype. Mol. Gen.
Genet. 245:636-643.
Lozano, R., Angosto, T., Gomez, P., Payan, C., Capel, J., Huijser, P., Salinas, J.,
Martinez-Zapater, J. M. (1998) Tomato flower abnormalities induced by low
temperatures are associated with changes of expression of MADS-Box genes.
Plant Physiol. 117:91-100.
Ma, H., Yanofsky, M. F., Meyerowitz, E. M. (1991) AGL1-AGL6, an Arabidopsis
gene family with similarity to floral homeotic and transcription factor genes.
Genes Dev. 5:484-495.
41
Malek, O., Lättig K. , Hiesel R. , Brennicke A., Knoop, V. (1996). RNA editing
in bryophytes and a molecular phylogeny of land plants. EMBO J. 15:1403-1411.
Mandel, M. A., Bowman, J. L., Kempin, S. A., Ma, H., Meyerowitz, E. M.,
Yanofsky, M. F. (1992) Manipulation of flower structure in transgenic tobacco.
Cell 71:133-143.
Mandel, M. A., Gustafson-Brown, C., Savidge, B., Yanofsky, M. F. (1992)
Molecular characterization of the Arabidopsis floral homeotic gene APETALA1.
Nature 360:273-277.
Mandel, M. A., Yanofsky, M. F. (1995) A gene triggering flower formation in
Arabidopsis. Nature 377:522-524.
Mathews, S., Donoghue, M. J. (1999) The root of angiosperm phylogeny inferred
from duplicate phytochrome genes. Science 286:947-950.
Mattsson, J., Söderman, E., Svenson, M., Borkird, C., Engström, P. (1992). A
new homeobox-leucine zipper gene from Arabidopsis thaliana. Plant Mol. Biol.
18:1019-1022.
Mattsson, J. (1995). Isolation and characterization of homeodomain-leucine zipper
encoding genes in plants. Comprehensive Summaries of Uppsala Dissertations
from the Faculty of Science and Technology, Acta Universitatis Upsaliensis,
Uppsala.
McKinney, M. L., McNamara, K. J. (1991) Heterochrony. Plenum Press, New
York and London.
Meyerowitz, E. M. (1999) Plants, animals and the logic of development. Trends
Cell Biol. 9:M65-68.
Mizukami, Y., Ma, H. (1992) Ectopic expression of the floral homeotic gene
AGAMOUS in transgenic Arabidopsis plants alters floral organ identity. Cell
71:119-131.
Mizukami, Y., Huang, H., Tudor, M., Hu, Y., Ma, H. (1996) Functional domains
of the floral regulator AGAMOUS: characterization of the DNA binding domain
and analysis of dominant negative mutations. Plant Cell 8:831-845.
Molkentin, J. D., Black, B. L., Martin, J. F., Olson, E. N. (1996). Mutational
analysis of the DNA binding, dimerisation and transcriptional activation domains
of MEF2C. Mol. Cell. Biol. 16:2627-2636.
Mouradov, A., Glassick, T., Vivian-Smith, A., Teasdale, R. (1996) Isolation of a
MADS-box gene family from Pinus radiata. Plant physiol. 110:1047.
Mouradov, A., Hamdorf, B., Teasdale, R. D. (1997) Isolation and characterization
of a new MADS-box cDNA from Pinus radiata. Plant Physiol 113:664.
Mouradov, A., Glassick, T. V., Hamdorf, B. A., Murphy, L. C., Marla, S. S.,
Yang, Y., Teasdale, R. D. (1998) Family of MADS-Box genes expressed early
in male and female reproductive structures of monterey pine. Plant Physiol.
117:55-62.
Mouradov, A., Hamdorf, B., Teasdale, R. D., Kim, J. T., Winter, K. U.,
Theissen, G. (1999) A DEF/GLO-like MADS-box gene from a gymnosperm:
Pinus radiata contains an ortholog of angiosperm B class floral homeotic genes.
Dev Genet. 25:245-252.
Münster, T., Pahnke, J., Di Rosa, A., Kim, J. T., Martin, W., Saedler, H.,
Theißen, G. (1997) Floral homeotic genes were recruited from homologous
MADS-box genes preexisting in the common ancestor of ferns and seed plants.
Proc. Natl. Acad. Sci. U S A 94:2415-2420.
42
Ohno, S. (1972) An argument for the Genetic Simplicity of Man and other
Mammals. Journal of human evolution 1:651-662.
Panganiban, G., Irvine, S. M., Lowe, C., Roehl, H., Corley, L. S., Sherbon, B.,
Grenier, J. K., Fallon, J. F., Kimble, J., Walker, M., Wray, G. A., Swalla, B.
J., Martindale, M. Q., Carroll, S. B. (1997)The origin and evolution of animal
appendages.Proc. Natl. Acad. Sci. U S A 94:5162-5166.
Passmore, S., Maine, G. T., Elble, R., Christ, C., Tye, B.-K. (1988)
Saccharomyces cerevisiae protein involved in plasmid maintenance is necessary
for mating of MATα cells. J. Mol. Biol. 204:593-606.
Phan-Dinh-Tuy, F., Tuil, D., Schweighoffer, F., Pinset, C., Kahn, A., Minty, A.
(1988) The 'CC.Ar.GG' box. A protein-binding site common to transcriptionregulatory regions of the cardiac actin, c-fos and interleukin-2 receptor genes.
Eur. J. Biochem. 173:507-515.
Pelaz, S., Ditta, G. S., Baumann, E., Wisman, E., Yanofsky, M. F. (2000) B and
C floral organ identity functions require SEPALLATA MADS-box genes. Nature
405:200-203.
Pellegrini, L., Tan, S., Richmond, T. J. (1995) Structure of serum response core
bound to DNA. Nature 376, 490-497.
Pickett, F. B. and Meeks-Wagner, D. R. (1995) Seeing double: appreciating
genetic redundancy. The Plant Cell 7:1347-1356.
Pidkowich, M. S., Klenz, J. E., Haughn, G. W. (1999) The making of a flower:
control of floral meristem identity in Arabidopsis. Trends Plant Sci. 4:64-70.
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: 175-186.
Pnueli, L., Abu-Abeid, M., Zamir. D., Nacken. W., Schwarz-Sommer, Z.,
Lifschitz, E. (1991) The MADSbox gene family in tomato: temporal expression
during floral development, conserved secondary structures and homology with
homeotic genes from Antirrhinum and Arabidopsis. Plant J., 1:255-266
Purugganan, M. D. (1997) The MADS-box floral homeotic gene lineages predate
the origin of seed plants: phylogenetic and molecular clock estimates. J. Mol.
Evol. 45:392-396.
Qiu, Y. L., Lee, J., Bernasconi-Quadroni, F., Soltis, D. E., Soltis, P. S., Zanis,
M., Zimmer, E. A., Chen, Z., Savolainen, V., Chase, M. W. (1999) The
earliest angiosperms: evidence from mitochondrial, plastid and nuclear genomes.
Nature 402:404-407.
Raff, R. A. (1996) The Shape of Life. The University of Chicago press, Chicago
and London.
Raubeson, L. A., Jansen, R. K. (1992) Chloroplast DNA Evidence on the Ancient
Evolutionary Split in Vascular Land Plants. Science 255:1697-1699.
Raven, P. H., Evert, R. F., Eichhorn, S. E. (1999) Biology of plants. 6th ed. W. H.
Freeman and Company. New York.
Riechmann, J. L., Krizek, B. A., Meyerowitz, E. (1996) Dimerization specificity
of Arabidopsis MADS domain hometotic proteins APETALA1, APETALA3,
PISTILLATA, and AGAMOUS. Proc. Natl. Acad. Sci. USA 93: 4793-4798.
Riechmann, J. L., Meyerowitz, E. M. (1997) MADS domain proteins in plant
development. Biol. Chem.378:1079-1101.
43
Rounsley, S. D., Ditta, G. S., Yanofsky, M. F. (1995) Diverse roles for MADS-box
genes in Arabidopsis development Plant Cell 7:1259-1269.
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 homolog from the conifer black spruce (Picea mariana) that
produces floral homeotic conversions when expressed in Arabidopsis. Plant J.
15:625-634.
Santelli, E., Richmond, T. J. (2000) Crystal structure of MEF2A core bound to
DNA at 1.5 Å resolution. J. Mol. Biol. 297:437-449.
Sattler, R. (1988) Homeosis in plants Amer. J. Bot. 75:106-117.
Sattler, R. (1991) Process morphology: structural dynamics in development and
evolution, Can. J. Bot. 70:708-714.
Savidge, B., Rounsley, S. D., Yanofsky, M. F. (1995) Temporal relationship
between the transcription of two Arabidopsis MADS box genes and the floral
organ identity genes. Plant Cell 7:721-733.
Schwarz-Sommer, Z., Huijser, P., Nacken, W., Saedler, H., Sommer, H. (1990)
Genetic Control of Flower Development by Homeotic Genes in Antirrhinum
majus, Science 250:931-936.
Sheldon, C. C., Burn, J. E., Perez, P. P., Metzger, J., Edwards, J. A., Peacock,
W. J., Dennis, E. S. (1999). The FLF MADS box gene: a repressor of flowering
in Arabidopsis regulated by vernalization and methylation. Plant Cell 11:445458.
Shore, P., Sharrocks, A. D. (1995) The MADS-box family of transcription factors.
Eur J Biochem, 229:1-13.
Singer, C. (1959) A short history of scientific ideas to 1900. Oxford university
press, London.
Soltis, P. S., Soltis, D. E., Chase, M. W. (1999) Angiosperm phylogeny inferred
from multiple genes as a tool for comparative biology. Nature 402:402-404.
Sommer, H., Beltran, J. P, Huijser, P., Pape, H., Lönnig, W. E., Saedler, H.,
Schwarz-Sommer, Z. (1990) Deficiens, a homeotic gene involved in the control
of flower morphogenesis in Antirrhinum majus: the protein shows homology to
transcription factors. EMBO J. 9:605-613.
Stewart, W. N., Rothwell, G. W. (1993) Paleobotany and the Evolution of Plants
2nd. ed. Cambridge University Press, Cambridge.
Svedelius, N. (1909) Om några svenska monstrositetsformer af Anemone nemorosa.
Svensk Bot. Tidskr. 3:47-63.
Svensson, M. (1996) Homeotiska mutanter kastar nytt ljus över blommorfologin hos
vitsippa, Anemone nemorosa Svensk Bot. Tidskr. 90:231-237.
Takhtajan, A. L. (1991) Evolutionary trends in flowering plants. Columbia
University Press, New York.
Tan, S., Richmond, T.J. (1998) Crystal structure of the yeast
MATalpha2/MCM1/DNA ternary complex, Nature 391:660-666.
Tandre, K., Albert, V. A., Sundås, A., Engström, P. (1995) Conifer homologs to
genes that control floral development in angiosperms. Plant Mol Biol. 27:69-78
Tauz, D. (1992) Redundancies, development and the flow of information.
BioEssays. 14:263-266.
44
Theißen, G., Strater, T., Fischer, A., Saedler, H. (1995) Structural
characterization, chromosomal localization and phylogenetic evaluation of two
pairs of AGAMOUS-like MADS-box genes from maize. Gene 156:155-166.
Theißen, G., Kim, J., 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:
484–516.
Theißen, G., Becker, A., Di Rosa, A., Kanno, A., Kim, J. T., Münster, T.,
Winter, K. U., Saedler, H. (2000) A short history of MADS-box genes in plants.
Plant Mol. Biol. 42:115-149.
Tröbner, W., Ramirez, L., Motte, P., Hue, I., Huijser, P., Lonnig, W. E., Saedler, H.,
Sommer H., Schwarz-Sommer, Z. (1992) GLOBOSA: a homeotic gene which
interacts with DEFICIENS in the control of Antirrhinum floral organogenesis.
EMBO J. 11: 4693-4704.
Tsuchimoto, S., van der Krol, A. R., Chua, N. H. (1993) Ectopic expression of
pMADS3 in transgenic petunia phenocopies the petunia blind mutant. Plant Cell
5: 843-853.
van der Krol, A. R., Brunelle, A., Tsuchimoto, S., Chua, N. H. (1993) Functional
analysis of petunia floral homeotic MADS box gene pMADS1. Genes Dev.
7:1214-1228.
van der Krol, A. R., Chua, N. H. (1993) Flower development in Petunia. Plant
Cell 5:1195-1203.
Weigel D., Alvarez, J., Smyth, D. R., Yanofsky, M. F., Meyerowitz, E. M. (1992)
LEAFY controls floral meristem identity in Arabidopsis. Cell 69:843-859.
Whittier, D. P., (1976) Gametophytes of Lycopodium obscurum as grown in axenic
culture. Can. J. Bot. 55:563-567.
Winter, K. U., Becker, A., Münster, T., Kim, J. T., Saedler, H., Theißen, G.
(1999) MADS-box genes reveal that gnetophytes are more closely related to
conifers than to flowering plants. Proc. Natl. Acad. Sci. U S A 96:7342-7347.
Yanofsky, M. F, Ma, H., Bowman, J. L., Drews, G. N., Feldmann, K. A.,
Meyerowitz E. M. (1990) The protein encoded by the Arabidopsis homeotic
gene AGAMOUS resembles transcription factors. Nature 346:35-39.
Young, D. (1992) The discovery of evolution. Natural History Museum
Publications, London and Cambridge University Press, Cambridge.
Yu, D., Kotilainen, M., Pollanen, E., Mehto, M., Elomaa, P., Helariutta, Y.,
Albert, V. A., Teeri, T. H. (1999) Organ identity genes and modified patterns of
flower development in Gerbera hybrida. Plant J. 17:51-62.
Zachgo, S., Saedler, H., Schwarz-Sommer, Z. (1997) Pollen-specific expression
of DEFH125, a MADS-box transcription factor in Antirrhinum with unusual
features. Plant J. 11:1043-1050.
Zhang, H., Forde, B. G. (1998) An Arabidopsis MADS Box gene That Controls
Nutrient-Induced Changes in Root Architecture. Science 279:407-409.
Zimmermann, W. (1965) Die Telomtheorie. Gustav Fischer Verlag, Stuttgart.
45
Errata list
p. 34
line 6: macrosporangia should be megasporangia
line 11: Baum (1999) should be Baum and Whitlock (1999)
line 19: ”This supports a simple perianth as a basal condition among angiosperms,
and in at least Nymphaeaceae it is clearly petaloid.” should be ”This supports a
spiral perianth as a basal condition among angiosperms, and in at least
Nymphaeaceae it contains clearly petaloid organs.”
Errors can be discussed by e-mail: [email protected]
46
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement