marina, suggesting that these genes were lost
before the lineages that led to extant Zosteraceae
and Lemnoideae separated. Given the relatively
frequent loss of FLC-like genes throughout
angiosperm evolution (Gramzow and Theißen
2015), however, it also appeared quite possible
that the absence of FLC-like genes in both species traces back to two independent gene-loss
events. However, even though evidence was
provided, that the first genome sequence of a
duckweed contained at least 90% of genic
sequences (Wang et al. 2014), incompleteness
and inaccuracies of any genome sequence make
it difficult to exclude false negative results.
9.5 Lost in Miniaturization: Setback
of MADS-Box Genes
in Duckweed Genomes
To test our initial conclusions concerning the loss
of MIKC-type genes in duckweeds, we here use
the additional genomic resources that have
become available since the time of our previous
study. These novel data include the genome
sequence of another strain of Spirodela polyrhiza
(strain 9509), and the genomes of Lemna gibba
and Lemna minor (Michael et al. 2017; van
Hoeck et al. 2015; Ernst 2016). Furthermore, we
reanalysed the Zostera marina MADS-box genes
(Olsen et al. 2016) and extended our analyses to
MIKC*-group genes. MADS-box genes were
identified, their phylogeny was reconstructed and
the presence and absence of MADS-box gene
clades were determined essentially as previously
described (Gramzow and Theißen 2015).
Redundant genes from the same species were
removed using Jalview and a redundancy
threshold of 95 (Waterhouse et al. 2009).
9.5.1 Conserved Clades of MIKC-Type
Genes in Alismatales
Our analyses identified 11 out of 17 clades of
MIKC
C -group genes and both (S and P) clades of
MIKC*-group genes in all duckweed species
analysed and in Zostera marina, suggesting that
these clades are conserved throughout Alismatales (Table 9.1; Fig. 9.1). The conserved
clades of MIKC
C -group genes are essentially the
ones also identified in our previous study of
flowering plants (Gramzow and Theißen 2015)
and include most of the clades which are formed
by genes with important functions in flower
development (AG-, AGL2-, AGL6-, DEF-,
GGM13-(B sister ), GLO-, SQUA-, STK-like
genes). Furthermore, some clades comprised of
genes that in functional terms are less well
understood, StMADS11- and AGL17-like genes,
are also conserved in Alismatales and hence in
all other flowering plants studied. Additionally,
TM8-like genes were found to be present
throughout the Alismatales species studied
(Table 9.1), but not throughout flowering plants
in general (Gramzow and Theißen 2015). In our
previous study, we found TM3-like genes a clade
of MIKC
C -group genes that had not been lost in
any of the species analysed. In our current study,
the genes previously classified as TM3-like genes
from Spirodela polyrhiza strain 7498 now cluster
among the FLC-like genes, however, as do their
closest relatives from the other duckweed species
and strains. Hence, assignment of genes to the
clade of FLC- or TM3-like genes is ambiguous,
and additional analyses will be required to reveal
whether the genes identified from duckweeds
belong to the clade of TM3- or FLC-like genes.
Only then one can determine which of the two
clades was actually lost in duckweeds. So either
FLC-like genes have been lost in the stem group
of extant Alismatales (the scenario shown in
Fig. 9.1), or TM3-like genes have been lost in the
stem group of extant duckweeds.
9.5.2 Clades of MIKC
C -Group Genes
Missing Completely
in Alismatales
Representatives of one or two clades of MIKC
C -
group genes may be absent from all Alismatales
species analysed (Table 9.1). In contrast to previous analyses (Gramzow and Theißen 2015), in
our current study, AGL15-like genes have not
been identified in Alismatales (Table 9.1) and
9 Stranger than Fiction: Loss of MADS-Box Genes …
97
before the lineages that led to extant Zosteraceae
and Lemnoideae separated. Given the relatively
frequent loss of FLC-like genes throughout
angiosperm evolution (Gramzow and Theißen
2015), however, it also appeared quite possible
that the absence of FLC-like genes in both species traces back to two independent gene-loss
events. However, even though evidence was
provided, that the first genome sequence of a
duckweed contained at least 90% of genic
sequences (Wang et al. 2014), incompleteness
and inaccuracies of any genome sequence make
it difficult to exclude false negative results.
9.5 Lost in Miniaturization: Setback
of MADS-Box Genes
in Duckweed Genomes
To test our initial conclusions concerning the loss
of MIKC-type genes in duckweeds, we here use
the additional genomic resources that have
become available since the time of our previous
study. These novel data include the genome
sequence of another strain of Spirodela polyrhiza
(strain 9509), and the genomes of Lemna gibba
and Lemna minor (Michael et al. 2017; van
Hoeck et al. 2015; Ernst 2016). Furthermore, we
reanalysed the Zostera marina MADS-box genes
(Olsen et al. 2016) and extended our analyses to
MIKC*-group genes. MADS-box genes were
identified, their phylogeny was reconstructed and
the presence and absence of MADS-box gene
clades were determined essentially as previously
described (Gramzow and Theißen 2015).
Redundant genes from the same species were
removed using Jalview and a redundancy
threshold of 95 (Waterhouse et al. 2009).
9.5.1 Conserved Clades of MIKC-Type
Genes in Alismatales
Our analyses identified 11 out of 17 clades of
MIKC
C -group genes and both (S and P) clades of
MIKC*-group genes in all duckweed species
analysed and in Zostera marina, suggesting that
these clades are conserved throughout Alismatales (Table 9.1; Fig. 9.1). The conserved
clades of MIKC
C -group genes are essentially the
ones also identified in our previous study of
flowering plants (Gramzow and Theißen 2015)
and include most of the clades which are formed
by genes with important functions in flower
development (AG-, AGL2-, AGL6-, DEF-,
GGM13-(B sister ), GLO-, SQUA-, STK-like
genes). Furthermore, some clades comprised of
genes that in functional terms are less well
understood, StMADS11- and AGL17-like genes,
are also conserved in Alismatales and hence in
all other flowering plants studied. Additionally,
TM8-like genes were found to be present
throughout the Alismatales species studied
(Table 9.1), but not throughout flowering plants
in general (Gramzow and Theißen 2015). In our
previous study, we found TM3-like genes a clade
of MIKC
C -group genes that had not been lost in
any of the species analysed. In our current study,
the genes previously classified as TM3-like genes
from Spirodela polyrhiza strain 7498 now cluster
among the FLC-like genes, however, as do their
closest relatives from the other duckweed species
and strains. Hence, assignment of genes to the
clade of FLC- or TM3-like genes is ambiguous,
and additional analyses will be required to reveal
whether the genes identified from duckweeds
belong to the clade of TM3- or FLC-like genes.
Only then one can determine which of the two
clades was actually lost in duckweeds. So either
FLC-like genes have been lost in the stem group
of extant Alismatales (the scenario shown in
Fig. 9.1), or TM3-like genes have been lost in the
stem group of extant duckweeds.
9.5.2 Clades of MIKC
C -Group Genes
Missing Completely
in Alismatales
Representatives of one or two clades of MIKC
C -
group genes may be absent from all Alismatales
species analysed (Table 9.1). In contrast to previous analyses (Gramzow and Theißen 2015), in
our current study, AGL15-like genes have not
been identified in Alismatales (Table 9.1) and
9 Stranger than Fiction: Loss of MADS-Box Genes …
97