Genetics of meiotic prophase I in plants

Detailed recombination studies along chromosome 3B provide new insights on crossover distribution in wheat (Triticum aestivum L.)

In wheat (Triticum aestivum L.), the crossover (CO) frequency increases gradually from the centromeres to the telomeres. However, little is known about the factors affecting both the distribution and the intensity of recombination along this gradient. To investigate this, we studied in detail the pattern of CO along chromosome 3B of bread wheat. A dense reference genetic map comprising 102 markers homogeneously distributed along the chromosome was compared to a physical deletion map. Most of the COs (90%) occurred in the distal subtelomeric regions that represent 40% of the chromosome. About 27% of the proximal regions surrounding the centromere showed a very weak CO frequency with only three COs found in the 752 gametes studied. Moreover, we observed a clear decrease of CO frequency on the distal region of the short arm. Finally, the intensity of interference was assessed for the first time in wheat using a Gamma model. The results showed m values of 1.2 for male recombination and 3.5 for female recombination, suggesting positive interference along wheat chromosome 3B.

The RAD51 gene family in bread wheat is highly conserved across eukaryotes, with RAD51A upregulated during early meiosis

The RADiation sensitive protein 51 (RAD51) recombinase is a eukaryotic homologue of the bacterial Recombinase A (RecA). It is required for homologous recombination of DNA during meiosis where it plays a role in processes such as homology searching and strand invasion. RAD51 is well conserved in eukaryotes with as many as four paralogues identified in vertebrates and some higher plants. Here we report the isolation and preliminary characterisation of four RAD51 gene family members in hexaploid (bread) wheat (Triticum aestivum L.). RAD51A1, RAD51A2 and RAD51D were located on chromosome group 7, and RAD51C was on chromosome group 2. Q-PCR gene expression profiling revealed that RAD51A1 was upregulated during meiosis with lower expression levels seen in mitotic tissue, and bioinformatics analysis demonstrated the evolutionary linkages of this gene family to other eukaryotic RAD51 sequences. Western blot analysis of heterologously expressed RAD51 from bread wheat has shown that it is detectable using anti-human RAD51 antibodies and that molecular modelling of the same protein revealed structural conservation when compared with yeast, human, Arabidopsis and maize RAD51A orthologues. This report has widened the knowledge base of this important protein family in plants, and highlighted the high level of structural conservation among RAD51 proteins from various species.

Sexual devolution in plants: apomixis uncloaked?

There are a growing number of examples where naturally occurring mutations disrupt an established physiological or developmental pathway to yield a new condition that is evolutionary favored. Asexual reproduction by seed in plants, or apomixis, occurs in a diversity of taxa and has evolved from sexual ancestors. One form of apomixis, diplospory, is a multi-step development process that is initiated when meiosis is altered to produce an unreduced rather than a reduced egg cell. Subsequent parthenogenetic development of the unreduced egg yields genetically maternal progeny. While it has long been apparent from cytological data that meiosis in apomicts was malfunctional or completely bypassed, the genetic basis of the phenomenon has been a long-standing mystery. New data from genetic analysis of Arabidopsis mutants in combination with more sophisticated molecular understanding of meiosis in plants indicate that a weak mutation of the gene SWI, called DYAD, interferes with sister chromatid cohesion in meiosis I, causes synapsis to fail in female meiosis and yields two unreduced cells. The new work shows that a low percentage of DYAD ovules produce functional unreduced egg cells (2n) that can be fertilized by haploid pollen (1n) to give rise to triploid (3n) progeny. While the DYAD mutants differ in some aspects from naturally occurring apomicts, the work establishes that mutation to a single gene can effectively initiate apomictic development and, furthermore, focuses efforts to isolate apomixis genes on a narrowed set of developmental events. Profitable manipulation of meiosis and recombination in agronomically important crops may be on the horizon.

Towards molecular breeding of reproductive traits in cereal crops

The transition from vegetative to reproductive phase, flowering per se, floral organ development, panicle structure and morphology, meiosis, pollination and fertilization, cytoplasmic male sterility (CMS) and fertility restoration, and grain development are the main reproductive traits. Unlocking their genetic insights will enable plant breeders to manipulate these traits in cereal germplasm enhancement. Multiple genes or quantitative trait loci (QTLs) affecting flowering (phase transition, photoperiod and vernalization, flowering per se), panicle morphology and grain development have been cloned, and gene expression research has provided new information about the nature of complex genetic networks involved in the expression of these traits. Molecular biology is also facilitating the identification of diverse CMS sources in hybrid breeding. Few Rf (fertility restorer) genes have been cloned in maize, rice and sorghum. DNA markers are now used to assess the genetic purity of hybrids and their parental lines, and to pyramid Rf or tms (thermosensitive male sterility) genes in rice. Transgene(s) can be used to create de novo CMS trait in cereals. The understanding of reproductive biology facilitated by functional genomics will allow a better manipulation of genes by crop breeders and their potential use across species through genetic transformation.

SWI1 is required for meiotic chromosome remodeling events

The Arabidopsis dsy10 mutant was previously identified as being defective in the synapsis of meiotic chromosomes resulting in male and female sterility. We report here the molecular analysis of the mutation and show that it represents a T-DNA insertion in the third exon of the SWI1 gene. Four mutations have now been identified in SWI1, several of which exhibit different phenotypes. For example, the swi1-1 and dyad mutations only affect meiosis in megasporocytes, while the swi1-2 and dsy10 mutations block both male and female meiosis. Furthermore, as part of a detailed cytological characterization of dsy10 meiocytes, we identified several differences during male meiosis between the swi1-2 and dys10 mutants, including variations in the formation of axial elements, the distribution of cohesin proteins and the timing of the premature loss of sister chromatid cohesion. We demonstrate that dsy10 represents a complete loss-of-function mutation, while a truncated form of SWI1 is expressed during meiosis in swi1-2 plants. We further show that dys10 meiocytes exhibit alterations in modified histone patterns, including acetylated histone H3 and dimethylated histone H3-Lysine 4.

Meiotic genes and proteins in cereals

We review the current status of our understanding and knowledge of the genes and proteins controlling meiosis in five major cereals, rye, wheat, barley, rice and maize. For each crop, we describe the genetic and genomic infrastructure available to investigators, before considering the inventory of genes and proteins that have roles to play in this process. Emphasis is given throughout as to how translational genomic and proteomic approaches have enabled us to circumvent some of the intractable features of this important group of plants.

Nuclear architecture and chromosome dynamics in the search of the pairing partner in meiosis in plants

The formation of haploid gametes in organisms with sexual reproduction requires regular bivalent chromosome pairing in meiosis. In many species, homologous chromosomes occupy separate territories at the onset of meiosis. To be paired at metaphase I, they need to be brought into a close proximity for interactions that include homology recognition and the establishment of some form of bonds. How homologues find each other is one of the least understood meiotic events. Plant species with large or medium sized genomes, such as wheat or maize, are excellent materials for the cytological analysis of chromosome dynamics at early meiosis, but genes that control meiosis have been identified mainly in small genome species such as Arabidopsis thaliana. This review is focused on the contribution studies on plants are providing to the knowledge of the initial steps of the meiotic process.

Pairing and synapsis in wild type Arabidopsis thaliana

A spreading technique was used to perform a structural analysis of prophase I nuclei in pollen mother cells (PMCs) of wild-type Arabidopsis thaliana. In leptotene, all chromosomes developed fully axial elements before a presynaptic alignment was observed. Pairing and synapsis start in regions close to the telomeres at early zygotene. Interstitial synaptonemal complex (SC) stretches were found to occur at several sites per bivalent at mid zygotene. Within individual bivalents, extensive regions of SC formation often existed at the same time as other extensive regions that were unsynapsed. Also in the same nucleus, one bivalent might have several SC segments, while other bivalents have only a few. The classical bouquet was not so evident as in other plant species. Length measurements of the five pachytene bivalents have allowed the elaboration of a pachytene karyotype. Pachytene chromatin compaction in Arabidopsis was significantly less than that observed in the other species analysed and this is paralleled with a higher recombination rate (centimorgans per megabase).

Homologous chromosome pairing is completed in crossover defective atzip4 mutant

In transposon-tagged lines of Arabidopsis, we found a mutant that was defective in meiotic chromosome segregation. This mutant, named atzip4-4, was due to a novel mutant allele of AtZIP4, which has sequence similarity to yeast ZIP4/SPO22, which codes a ZMM protein that is a proposed unit of the synapsis initiation complex. The chiasma distribution in atzip4-4 differed from that in the wild-type, involved in a deficiency of interfering crossovers in the mutant genome. On the other hand, FISH staining of loci on two independent chromosomes in mutant meiocytes indicated that homologous chromosome pairing to synapse progresses normally until the pachytene stage, yet homologous chromosomes often separated abruptly at diplotene and diakinesis. These results suggest that AtZIP4 plays an important role in normal crossover formation and meiotic chromosome segregation, but not in homolog search. The relationship of AtZIP4 and other related proteins in meiotic events is discussed and compared with that in yeast.

Differential gene expression in Arabidopsis wild-type and mutant anthers: insights into anther cell differentiation and regulatory networks

In flowering plants, the anther contains highly specialized reproductive and somatic cells that are required for male fertility. Genetic studies have uncovered several genes that are important for anther development. However, little information is available regarding most genes active during anther development, including possible relationships between these genes and genetically defined regulators. In Arabidopsis, two previously isolated male-sterile mutants display dramatically altered anther cell differentiation patterns. The sporocyteless (spl)/nozzle (nzz) mutant is defective in the differentiation of primary sporogenous cells into microsporocytes, and does not properly form the anther wall. The excess microsporocytes1 (ems1)/extrasporogenous cells (exs) mutants produce excess microsporocytes at the expense of the tapetum. To gain additional insights into microsporocyte and tapetum differentiation and to uncover potential genetic interactions, expression profiles were compared between wild-type anthers (stage 4-6) and those of the spl or ems1 mutants. A total of 1954 genes were found to be differentially expressed in the ems1 and/or spl anthers, and these were grouped into 14 co-expression clusters. The presence of genes with known and predicted functions in specific clusters suggests potential functions for other genes in the same cluster. To obtain clues about possible co-regulation within co-expression clusters, we searched for shared cis-regulatory motifs in putative promoter regions. Our analyses were combined with data from previous studies to develop a model of the anther gene regulatory network. This model includes hypotheses that can be tested experimentally to gain further understanding of the mechanisms controlling anther development.

OsDMC1 is required for homologous pairing in Oryza sativa

OsDMC1 is the rice homologue of the yeast DMC1 gene. Here, we analyzed the function of OsDMC1 in meiosis using an RNA interference approach. The OsDMC1-RNAi lines grew normally during their vegetative phase but showed spikelet and pollen sterility. The sterility phenotypes were associated with down-regulated OsDMC1 transcript and protein levels mediated by RNAi. Further cytological observations of male meiocytes revealed that knock-down of OsDMC1 led to defects in bivalent formation and subsequent unequal chromosome segregation and irregular spore generation, and induced changes in male meiotic progression. Fluorescent in situ hybridization experiments revealed that the OsDMC1-RNAi lines were defective in homologous pairing. These data indicate that OsDMC1 is essential for rice meiosis and plays an important role in homologous pairing.

Expression and functional analysis of TaASY1 during meiosis of bread wheat (Triticum aestivum)

Background

Pairing and synapsis of homologous chromosomes is required for normal chromosome segregation and the exchange of genetic material via recombination during meiosis. Synapsis is complete at pachytene following the formation of a tri-partite proteinaceous structure known as the synaptonemal complex (SC). In yeast, HOP1 is essential for formation of the SC, and localises along chromosome axes during prophase I. Homologues in Arabidopsis (AtASY1), Brassica (BoASY1) and rice (OsPAIR2) have been isolated through analysis of mutants that display decreased fertility due to severely reduced synapsis of homologous chromosomes. Analysis of these genes has indicated that they play a similar role to HOP1 in pairing and formation of the SC through localisation to axial/lateral elements of the SC.

Results

The full length wheat cDNA and genomic clone, TaASY1, has been isolated, sequenced and characterised. TaASY1 is located on chromosome Group 5 and the open reading frame displays significant nucleotide sequence identity to OsPAIR2 (84%) and AtASY1 (63%). Transcript and protein analysis showed that expression is largely restricted to meiotic tissue, with elevated levels during the stages of prophase I when pairing and synapsis of homologous chromosomes occur. Immunolocalisation using transmission electron microscopy showed TaASY1 interacts with chromatin that is associated with both axial elements before SC formation as well as lateral elements of formed SCs.

Conclusion

TaASY1 is a homologue of ScHOP1, AtASY1 and OsPAIR2 and is the first gene to be isolated from bread wheat that is involved in pairing and synapsis of homologous chromosomes.

Functional analysis of maize RAD51 in meiosis and double-strand break repair

In Saccharomyces cerevisiae, Rad51p plays a central role in homologous recombination and the repair of double-strand breaks (DSBs). Double mutants of the two Zea mays L. (maize) rad51 homologs are viable and develop well under normal conditions, but are male sterile and have substantially reduced seed set. Light microscopic analyses of male meiosis in these plants reveal reduced homologous pairing, synapsis of nonhomologous chromosomes, reduced bivalents at diakinesis, numerous chromosome breaks at anaphase I, and that >33% of quartets carry cells that either lack an organized nucleolus or have two nucleoli. This indicates that RAD51 is required for efficient chromosome pairing and its absence results in nonhomologous pairing and synapsis. These phenotypes differ from those of an Arabidopsis rad51 mutant that exhibits completely disrupted chromosome pairing and synapsis during meiosis. Unexpectedly, surviving female gametes produced by maize rad51 double mutants are euploid and exhibit near-normal rates of meiotic crossovers. The finding that maize rad51 double mutant embryos are extremely susceptible to radiation-induced DSBs demonstrates a conserved role for RAD51 in the repair of mitotic DSBs in plants, vertebrates, and yeast.

ATM-mediated transcriptional and developmental responses to gamma-rays in Arabidopsis

ATM (Ataxia Telangiectasia Mutated) is an essential checkpoint kinase that signals DNA double-strand breaks in eukaryotes. Its depletion causes meiotic and somatic defects in Arabidopsis and progressive motor impairment accompanied by several cell deficiencies in patients with ataxia telangiectasia (AT). To obtain a comprehensive view of the ATM pathway in plants, we performed a time-course analysis of seedling responses by combining confocal laser scanning microscopy studies of root development and genome-wide expression profiling of wild-type (WT) and homozygous ATM-deficient mutants challenged with a dose of γ-rays (IR) that is sublethal for WT plants. Early morphologic defects in meristematic stem cells indicated that AtATM, an Arabidopsis homolog of the human ATM gene, is essential for maintaining the quiescent center and controlling the differentiation of initial cells after exposure to IR. Results of several microarray experiments performed with whole seedlings and roots up to 5 h post-IR were compiled in a single table, which was used to import gene information and extract gene sets. Sequence and function homology searches; import of spatio-temporal, cell cycling, and mutant-constitutive expression characteristics; and a simplified functional classification system were used to identify novel genes in all functional classes. The hundreds of radiomodulated genes identified were not a random collection, but belonged to functional pathways such as those of the cell cycle; cell death and repair; DNA replication, repair, and recombination; and transcription; translation; and signaling, indicating the strong cell reprogramming and double-strand break abrogation functions of ATM checkpoints. Accordingly, genes in all functional classes were either down or up-regulated concomitantly with downregulation of chromatin deacetylases or upregulation of acetylases and methylases, respectively. Determining the early transcriptional indicators of prolonged S-G2 phases that coincided with cell proliferation delay, or an anticipated subsequent auxin increase, accelerated cell differentiation or death, was used to link IR-regulated hallmark functions and tissue phenotypes after IR. The transcription burst was almost exclusively AtATM-dependent or weakly AtATR-dependent, and followed two major trends of expression in atm: (i)-loss or severe attenuation and delay, and (ii)-inverse and/or stochastic, as well as specific, enabling one to distinguish IR/ATM pathway constituents. Our data provide a large resource for studies on the interaction between plant checkpoints of the cell cycle, development, hormone response, and DNA repair functions, because IR-induced transcriptional changes partially overlap with the response to environmental stress. Putative connections of ATM to stem cell maintenance pathways after IR are also discussed.

Meiosis-driven genome variation in plants

Meiosis includes two successive divisions of the nucleus with one round of DNA replication and leads to the formation of gametes with half of the chromosomes of the mother cell during sexual reproduction. It provides a cytological basis for gametogenesis and nheritance in eukaryotes. Meiotic cell division is a complex and dynamic process that involves a number of molecular and cellular events, such as DNA and chromosome replication, chromosome pairing, synapsis and recombination, chromosome segregation, and cytokinesis. Meiosis maintains genome stability and integrity over sexual life cycles. On the other hand, meiosis generates genome variations in several ways. Variant meiotic recombination resulting from specific genome structures induces deletions, duplications, and other rearrangements within the genic and non-genic genomic regions and has been considered a major driving force for gene and genome evolution in nature. Meiotic abnormalities in chromosome segregation lead to chromosomally imbalanced gametes and aneuploidy. Meiotic restitution due to failure of the first or second meiotic division gives rise to unreduced gametes, which triggers polyploidization and genome expansion. This paper reviews research regarding meiosis-driven genome variation, including deletion and duplication of genomic regions, aneuploidy, and polyploidization, and discusses the effect of related meiotic events on genome variation and evolution in plants. Knowledge of various meiosis-driven genome variations provides insight into genome evolution and genetic variability in plants and facilitates plant genome research.

Transcriptome profiling of maize anthers using genetic ablation to analyze pre-meiotic and tapetal cell types

Oligonucleotide arrays were used to profile gene expression in dissected maize anthers at four stages: after-anther initiation, at the rapid mitotic proliferation stage, pre-meiosis, and meiotic prophase I. Nearly 9200 sense and antisense transcripts were detected, with the most diverse transcriptome present at the pre-meiotic stage. Three male-sterile mutants lacking a range of normal cell types resulting from a temporal progression of anther failure were compared with fertile siblings at equivalent stages by transcription profiles. The msca1 mutant has the earliest visible phenotype, develops none of the normal anther cell types and exhibits the largest deviation from fertile siblings. The mac1 mutant has an excess of archesporial derivative cells and lacks a tapetum and middle layer, resulting in moderate transcriptional deviations. The ms23 mutant lacks a differentiated tapetum and shows the fewest differences from fertile anthers. By combining the data sets from the comparisons between individual sterile and fertile anthers, candidate genes predicted to play important roles during maize anther development were assigned to stages and to likely cell types. Comparative analyses with a data set of anther-specific genes from rice highlight remarkable quantitative similarities in gene expression between these two grasses.

T-loop phosphorylation of Arabidopsis CDKA;1 is required for its function and can be partially substituted by an aspartate residue

As in other eukaryotes, progression through the cell cycle in plants is governed by cyclin-dependent kinases. Phosphorylation of a canonical Thr residue in the T-loop of the kinases is required for high enzyme activity in animals and yeast. We show that the Arabidopsis thaliana Cdc2(+)/Cdc28 homolog CDKA;1 is also phosphorylated in the T-loop and that phosphorylation at the conserved Thr-161 residue is essential for its function. A phospho-mimicry T161D substitution restored the primary defect of cdka;1 mutants, and although the T161D substitution displayed a dramatically reduced kinase activity with a compromised ability to bind substrates, homozygous mutant plants were recovered. The rescue by the T161D substitution, however, was not complete, and the resulting plants displayed various developmental abnormalities. For instance, even though flowers were formed, these plants were completely sterile as a result of a failure of the meiotic program, indicating that different requirements for CDKA;1 function are needed during plant development.

The mismatch repair protein MLH1 marks a subset of strongly interfering crossovers in tomato

In most eukaryotes, the prospective chromosomal positions of meiotic crossovers are marked during meiotic prophase by protein complexes called late recombination nodules (LNs). In tomato (Solanum lycopersicum), a cytological recombination map has been constructed based on LN positions. We demonstrate that the mismatch repair protein MLH1 occurs in LNs. We determined the positions of MLH1 foci along the 12 tomato chromosome pairs (bivalents) during meiotic prophase and compared the map of MLH1 focus positions with that of LN positions. On all 12 bivalents, the number of MLH1 foci was approximately 70% of the number of LNs. Bivalents with zero MLH1 foci were rare, which argues against random failure of detecting MLH1 in the LNs. We inferred that there are two types of LNs, MLH1-positive and MLH1-negative LNs, and that each bivalent gets an obligate MLH1-positive LN. The two LN types are differently distributed along the bivalents. Furthermore, cytological interference among MLH1 foci was much stronger than interference among LNs, implying that MLH1 marks the positions of a subset of strongly interfering crossovers. Based on the distances between MLH1 foci or LNs, we propose that MLH1-positive and MLH1-negative LNs stem from the same population of weakly interfering precursors.

The road to crossovers: plants have their say

Crossovers involve the reciprocal exchange of large fragments of genetic material between homologous chromosomes during meiosis. In this way, crossovers are the basis of genetics. Remarkably, the number and distribution of crossovers on chromosomes are closely controlled. Data from various model organisms (notably Saccharomyces cerevisiae) show that the distribution of crossovers results from a series of tightly regulated events involving the formation and repair of double-strand breaks and interference. Recent advances in genetic and cytological tools, particularly for studying Arabidopsis thaliana, have enabled crossover control in plants to be studied in more detail. In this article, we discuss the contribution of plant studies to meiosis research, particularly to our understanding of crossover control and interference, and we evaluate models of interference.

Reinterpreting pericentromeric heterochromatin

In fission yeast, pericentromeric heterochromatin is directly responsible for the sister chromatid cohesion that assures accurate chromosome segregation. In plants, however, heterochromatin and chromosome segregation appear to be largely unrelated: chromosome transmission is impaired by mutations in cohesion but not by mutations that affect heterochromatin formation. We argue that the formation of pericentromeric heterochromatin is primarily a response to constraints on chromosome mechanics that disfavor the transmission of recombination events in pericentromeric regions. This effect allows pericentromeres to expand to enormous sizes by the accumulation of transposons and through large-scale insertions and inversions. Although sister chromatid cohesion is spatially limited to pericentromeric regions at mitosis and meiosis II, the cohesive domains appear to be defined independently of heterochromatin. The available data from plants suggest that sister chromatid cohesion is marked by histone phosphorylation and mediated by Aurora kinases.

Molecular assembly of meiotic proteins Asy1 and Zyp1 and pairing promiscuity in rye (Secale cereale L.) and its synaptic mutant sy10

Assembly of two orthologous proteins associated with meiotic chromosome axes in Arabidopsis thaliana (Asy1 and Zyp1) was studied immunologically at meiotic prophase of meiosis of wild-type rye (Secale cereale) and its synaptic mutant sy10, using antibodies derived from A. thaliana. The temporal and spatial expression of the two proteins were similar in wild-type rye, but with one notable difference. Unlike A. thaliana, in which foci of the transverse filament protein Zyp1 appear to linearize commensurately with synapsis, linear tracts of Asy1 and Zyp1 protein form independently at leptotene and early zygotene of rye and coalign into triple structures resembling synaptonemal complexes (SCs) only at later stages of synapsis. The sy10 mutant used in this study also forms spatially separate linear tracts of Asy1 and Zyp1 proteins at leptotene and early zygotene, and these coalign but do not form regular triple structures at midprophase. Electron microscopy of spread axial elements reveals extensive asynapsis with some exchanges of pairing partners. Indiscriminate SCs support nonhomologous chiasma formation at metaphase I, as revealed by multi-color fluorescence in situ hybridization enabling reliable identification of all the chromosomes of the complement. Scrutiny of chiasmate associations of chromosomes at this stage revealed some specificity in the associations of homologous and nonhomologous chromosomes. Inferences about the nature of synapsis in this mutant were drawn from such observations.

Origins and evolution of the recA/RAD51 gene family: evidence for ancient gene duplication and endosymbiotic gene transfer

The bacterial recA gene and its eukaryotic homolog RAD51 are important for DNA repair, homologous recombination, and genome stability. Members of the recA/RAD51 family have functions that have differentiated during evolution. However, the evolutionary history and relationships of these members remains unclear. Homolog searches in prokaryotes and eukaryotes indicated that most eubacteria contain only one recA. However, many archaeal species have two recA/RAD51 homologs (RADA and RADB), and eukaryotes possess multiple members (RAD51, RAD51B, RAD51C, RAD51D, DMC1, XRCC2, XRCC3, and recA). Phylogenetic analyses indicated that the recA/RAD51 family can be divided into three subfamilies: (i) RADalpha, with highly conserved functions; (ii) RADbeta, with relatively divergent functions; and (iii) recA, functioning in eubacteria and eukaryotic organelles. The RADalpha and RADbeta subfamilies each contain archaeal and eukaryotic members, suggesting that a gene duplication occurred before the archaea/eukaryote split. In the RADalpha subfamily, eukaryotic RAD51 and DMC1 genes formed two separate monophyletic groups when archaeal RADA genes were used as an outgroup. This result suggests that another duplication event occurred in the early stage of eukaryotic evolution, producing the DMC1 clade with meiosis-specific genes. The RADbeta subfamily has a basal archaeal clade and five eukaryotic clades, suggesting that four eukaryotic duplication events occurred before animals and plants diverged. The eukaryotic recA genes were detected in plants and protists and showed strikingly high levels of sequence similarity to recA genes from proteobacteria or cyanobacteria. These results suggest that endosymbiotic transfer of recA genes occurred from mitochondria and chloroplasts to nuclear genomes of ancestral eukaryotes.