16. Evolution of Reproductive Organs in Vascular Plants
253
7 Future Prospects
Based on the above discussion, the increases of the number of MADS genes and the
recruitment of these genes in specific tissues are likely to have had an important
role in the evolution of reproductive organs in vascular plants (Fig. 3), although the
details are still unknown. Our present scenario for the evolution of MADS genes
and reproductive organs is as follows. Trimerophytophyta had at least two MADS
genes, which were likely to have been orthologous to the genes in the AG and
AGL2-AGL6-AP1 clades and ubiquitously expressed in both vegetative and reproductive organs. MADS genes were not recruited to be specifically expressed in
reproductive organs in the fern lineage, although the number of MADS genes was
increased. After the seed plant lineage diverged from the fern lineage, the number
of MADS genes dramatically increased in the seed plant lineage, and most members of the gene groups presently observed in angiosperms were established. Before
the divergence of angiosperms and living gymnosperms, some MADS genes were
recruited to be specifically expressed in reproductive organs. In the living gymnosperm lineage, the API ortholog was lost and the number of other MADS genes in
each gene group was increased. The loss of the API ortholog is likely to be the
reason for the simpler reproductive organs of living gymnosperms compared to
angiosperms. In the angiosperm lineage, the ancestral B-function gene was duplicated and AP3 and PI genes were established. Extensive gene duplications in each
gene group occurred. The time at which organ-specific expression of A-, B- and Cfunction genes was established is unknown. Comparisons between gene cascades of
angiosperm B- and C-function genes and those of their gymnosperm orthologs
should reveal when and how B- and C-function genes were recruited to be expressed in a tissue-specific manner, and became homeotic selector genes of floral
organ development. For example, it is important to know (1) whether gymnosperm
LFY genes can positively regulate gymnosperm B- and C-function gene orthologs,
(2) whether the gymnosperm UFO, if it exists, can regulate the expression of angiosperm B-function genes as the angiosperm UFO gene can do, (3) whether the
gymnosperm B-function gene orthologs can regulate the down-stream genes of the
angiosperm B-function gene, and (4) whether CURLY LEAF (Goodrich et al. 1997)
and APETALA2 (Drews et al. 1991; lufuku et al. 1994) genes known to be involved
in the regulation of the angiosperm C-function genes also regulate gymnosperm
MADS genes.
Furthermore, analyses of MADS genes in lower plants including bryophytes
and green algae will provide valuable information on the original function of MADS
genes in the most recent common ancestors of land plants and green plants, respectively.
253
7 Future Prospects
Based on the above discussion, the increases of the number of MADS genes and the
recruitment of these genes in specific tissues are likely to have had an important
role in the evolution of reproductive organs in vascular plants (Fig. 3), although the
details are still unknown. Our present scenario for the evolution of MADS genes
and reproductive organs is as follows. Trimerophytophyta had at least two MADS
genes, which were likely to have been orthologous to the genes in the AG and
AGL2-AGL6-AP1 clades and ubiquitously expressed in both vegetative and reproductive organs. MADS genes were not recruited to be specifically expressed in
reproductive organs in the fern lineage, although the number of MADS genes was
increased. After the seed plant lineage diverged from the fern lineage, the number
of MADS genes dramatically increased in the seed plant lineage, and most members of the gene groups presently observed in angiosperms were established. Before
the divergence of angiosperms and living gymnosperms, some MADS genes were
recruited to be specifically expressed in reproductive organs. In the living gymnosperm lineage, the API ortholog was lost and the number of other MADS genes in
each gene group was increased. The loss of the API ortholog is likely to be the
reason for the simpler reproductive organs of living gymnosperms compared to
angiosperms. In the angiosperm lineage, the ancestral B-function gene was duplicated and AP3 and PI genes were established. Extensive gene duplications in each
gene group occurred. The time at which organ-specific expression of A-, B- and Cfunction genes was established is unknown. Comparisons between gene cascades of
angiosperm B- and C-function genes and those of their gymnosperm orthologs
should reveal when and how B- and C-function genes were recruited to be expressed in a tissue-specific manner, and became homeotic selector genes of floral
organ development. For example, it is important to know (1) whether gymnosperm
LFY genes can positively regulate gymnosperm B- and C-function gene orthologs,
(2) whether the gymnosperm UFO, if it exists, can regulate the expression of angiosperm B-function genes as the angiosperm UFO gene can do, (3) whether the
gymnosperm B-function gene orthologs can regulate the down-stream genes of the
angiosperm B-function gene, and (4) whether CURLY LEAF (Goodrich et al. 1997)
and APETALA2 (Drews et al. 1991; lufuku et al. 1994) genes known to be involved
in the regulation of the angiosperm C-function genes also regulate gymnosperm
MADS genes.
Furthermore, analyses of MADS genes in lower plants including bryophytes
and green algae will provide valuable information on the original function of MADS
genes in the most recent common ancestors of land plants and green plants, respectively.