All living organisms evolved from a common ancestor that lived more than 3.5 billion years ago, and the accumulation of mutations in their genomes has resulted in the present biodiversity. Traces of the evolutionary process are found in the genomes of extant organisms. By comparing the gene sequences and gene networks of different organisms, we can infer (1) the phylogenetic relationships of extant organisms and (2) the genetic changes that caused the evolution of morphology and development. The inferred phylogenetic relationships provide important insights into problems in various fields of evolutionary biology. Our group focuses on biogeography, the evolution of morphological traits, and systematics in a wide range of taxa. Concerning the evolution of morphology and development, we hope to explore the genetic changes that led to the evolution of the plant body plan. We have selected Arabidopsis (angiosperm), Gnetum (gymnosperm), Ginkgo (gymnosperm), Ceratopteris (pteridophyte), Physcomitrella (bryophyte), and some green algae as models to compare the functions of genes involved in the development of both reproductive and vegetative organs in land plants.

I. Origin of the Plant Cell

The first green alga cell evolved via symbiosis between an ancestral non-photosynthetic eukaryote and a cyanobacterium. Cyanobacteria now exist as chloroplasts in the host cell. The factors and mechanisms of chloroplast movement are being investigated to reveal the molecular mechanisms used to "domesticate" cyanobacteria as organelles. Analyses of cytosolic calcium icon concentration and cytoskeleton organization during chloroplast movement in the moss Physcomitrella patens is in progress by a team directed by Y. Sato.

II. Evolution from unicellular to multicellular organisms

The first evolutionary step from unicellular to multicellular organisms is to form two different cells from a single cell via asymmetric cell division. The first cell division of a protoplast isolated from the protonemata of the moss Physcomitrella patens is asymmetric regarding to its shape and nature, and gives rise to an apical meristematic cell and a differentiated non-meristematic cell. A systematic overexpression screening for genes involved in asymmetric cell division of protoplasts in P. patens is in progress by a team directed by T. Fujita. We constructed three full-length cDNA libraries from non-treated, auxin-treated, and cytokinin-treated protonemal cells of P. patens, then determined the sequences of more than 40,000 cDNAs from the both ends (Nishiyama, Fujita et al. 2003). We used these clones as materials for the overexpression screening. Individual cDNAs were selected based on their sequence, subcloned under a constitutive promoter and introduced into the protoplasts of P. patens for transient expression. We observed and categorized phenotypes of the regenerating protoplasts. Thus far we identified many cDNAs, whose overexpression resulted in symmetric cell division rather than asymmetric cell division, isotropic outgrowth with no polarity or curved cells showing incorrect direction of growth. These preliminary results indicate that some of these genes likely function for polarity formation and/or asymmetric cell division of the protoplasts. Functional analyses of these genes should help us to understand molecular mechanisms of how plants generate distinct cell lineages to build their multicellular bodies.

(Figure 1) Asymmetric cell division of a Physcomitrella protoplast. Regular asymmetric cell division (1, 2) and defective cell division with overexpression (3, 4). Cell wall of the cell in (1) is visualized with calcoflor in (2).

III. Evolution from cells to tissues

The most prominent difference between plant and animal cells is that plant cells have a cell wall and do not move during development. Therefore, the plane of cell division and the direction of cell elongation, which are regulated by cortical microtubules, determine the morphology of differentiated tissues and organs.

Organization of g-tubulin

g-tubulin is a protein that is essential for the formation of microtubules in animal cells. We found that g-tubulin is located at the end of cortical microtubules, and a loss of g-tubulin due to gene silencing causes a malformed organ with irregularly shaped cells. In vitro experiments using isolated plasma membrane/microtubule complexes suggested that g-tubulin attaches onto the side of existing cortical microtubules, and initiates a new cortical microtubule from it. The nucleus is known to initiate microtubules after cell division. We hypothesize that microtubules formed around the nucleus elongate to cell surface, and trigger initiation of cortical microtubules via attachment of g-tubulin. Once cortical microtubules are formed, they can turnover without microtubules from the nucleus. Factor(s) responsible for attachment of g-tubulin onto the side of microtubules is a key element responsible for the difference between plant and animal cells. Isolation of the factor(s) responsible for attachment of g-tubulin onto the side of microtubules by biochemical and other approaches is inprogress by a team directed by T. Murata.

Dynamics of actin filaments and microtubules in the moss P. patens

Cells of the moss P. patens gametophytes are an excellent model to study dynamics of cytoskeleton because of the easiness for observation and the feasibility of gene-targeting. Transformants with reporter constructs by fusing a GFP with P. patens a-tubulin or an actin binding domain of mouse talin were established by Y. Sato and collaborators to visualize microtubules and actin filaments. These transformants will be useful to investigate the dynamics of microtubules and actin filaments in the processes of cell division, cell elongation, and chloroplast movement.

IV. Evolution of molecular mechanisms in the development of vegetative organs

Meristem initiation and maintenance

Postembryonic growth of land plants occurs from the meristem, a localized region that gives rise to all adult structures. Meristems control the continuous development of plant organs by balancing the maintenance and proliferation of stem cells, and directing their differentiation. Meristem initiation and maintenance is a fundamental question in plant development research. However, the molecular mechanisms involved in meristem initiation and maintenance have not been studied in detail because most loss-of-function mutants are lethal. In the moss Physcomitrella patens, the developmental process of meristem is well defined at the cellular level, and gene targeting based on homologous recombination is feasible. Thus, meristem development in P. patens is used as a model system for studies of meristem development in land plants. We established approximately 20,000 gene- or enhancer-trap lines in P. patens to clone genes involved in meristem development. Seven lines, exhibiting reporter gene (uidA) expression preferentially in the apical cells, were isolated. Corresponding genes in three of these trap lines were identified as encoding kinesin- and ubiquitin-like proteins, and an unknown protein. Disruption of the gene encoding ubiquitin-like protein suggests that the gene be involved in cell division and elongation through microtubule organization. The functions of other genes in the meristem are currently under investigation by a team directed by Y. Hiwatashi.

The morphology of the shoot meristem in land plants varies. To investigate whether the molecular mechanisms of shoot development in angiosperms are conserved in other land plants, the functions of the KNOX homeobox, and the ZWILLE, NAC, and PIN genes, which are indispensable for shoot meristem development in angiosperms, are being studied in the fern Ceratopteris and the moss Physcomitrella.

Rhizoid differentiation

Rhizoids are multicellular filaments with similar functions to root hairs, and are observed in a wide range of plants. A team directed by K. Sakakibara, who was a graduate student in our lab, examined mechanisms underlying rhizoid development in the moss, Physcomitrella patens (Sakakinara et al. 2003).

Figure. 2 (1) a gametophyte of P. patens. (2) an auxin-treated gametophyte with plenty of rhizoids.

V. Origin and evolution of floral homeotic genes

The flower is the reproductive organ in angiosperms, and floral homeotic genes, such as MADS-box genes and FLO/LEAFY genes, regulate floral organ identity. To investigate the origin of floral homeotic genes, the functions of MADS-box genes and the FLO/LEAFY genes in gymnosperms (Gnetum, Ginkgo, and cycads), a fern (Ceratopteris), a moss (Physcomitrella), and three green algae (Chara, Coleochaete, and Closterium) are being analyzed.
Land plants are believed to have evolved from a gametophyte (haploid)-dominant ancestor without a multicellular sporophyte (diploid plant body); most genes expressed in the sporophyte probably originated from those used in the gametophyte during the evolution of land plants. To analyze the evolution and diversification of MADS-box genes in land plants, gametophytic MADS-box genes were screened using macroarray analyses for 105 MADS-box genes found in the Arabidopsis genome (Kofuji et al. 2003). Eight MADS-box genes were predominantly expressed in pollen, male gametophyte, and analyses of their function are in progress by a team directed by N. Aono.

VI. Evolution of life cycles

The mosses and flowering plants diverged more than 400 million years ago. The mosses have haploid-dominant life cycles, while the flowering plants are diploid-dominant. The common ancestors of land plants are inferred to have been haploid-dominant, suggesting that genes used in the diploid body of flowering plants were recruited from the genes used in the haploid body of their ancestors during the evolution of land plants.
To assess the evolutionary hypothesis that genes used in diploid body of flowering plants were recruited from the genes used in haploid body of the ancestors, a team directed by T. Fujita constructed an expressed sequence tag (EST) library of Physcomitrella, and T. Nishiyama mainly worked on the comparison of the moss transcriptome to the genome of Arabidopsis. We constructed full-length enriched cDNA libraries from auxin-treated, cytokinin-treated, and untreated gametophytes of Physcomitrella, and sequenced both ends of more than 40,000 clones. These data, together with the mRNA sequences in the public databases, were assembled into 15,883 putative transcripts. Sequence comparisons of Arabidopsis and Physcomitrella showed that the haploid transcriptome of Physcomitrella appears to be quite similar to the Arabidopsis genome, supporting the evolutionary hypothesis. Our study also revealed that a number of genes are moss specific and were lost in flowering plant lineage. Our full-length cDNA library will be a good resource for functional genomic studies in plants. Our EST data together with some information on the moss is open to public at PHYSCObase (http://moss.nibb.ac.jp).

VII. Molecular mechanisms of speciation

Reproductive isolation is the first step in speciation. To obtain insight into reproductive isolation, several receptors specifically expressed in the pollen tube are being studied to screen for the receptors that are involved in pollen tube guidance by a team directed by S. Miyazaki.

Polyploidization is a major mode of speciation in plants, although the changes that occur after genome duplication are not well known. Polyploid species are usually larger than diploids, but the mechanisms responsible for the size difference are unknown. To investigate these mechanisms, tetraploid Arabidopsis was established and its gene expression patterns are being compared to those of diploid wild-type plants using microarrays.

VIII. Phylogenetic analysis of land plants

We determined the complete nucleotide sequence of the chloroplast genome of the leptosporangiate fern, Adiantum capillus-veneris L. (Pteridaceae) as a collaboration work with P. Wolf and C. Rowe from Utah State University (Wolf et al. 2003). Phylogenetic analysis of basal land plants using the sequence is in progress by T. Nishiyama. Many unusual start codons and internal stop codons were found, suggesting that extensive number of RNA editing exists in the genome.