Year: 2005
Type: article
Abstract: The Fabaceae, or legumes, constitute the third largest family of flowering plants, comprising more than 650 genera and 18,000 species (Polhill and Raven, 1981). Economically, legumes represent the second most important family of crop plants after Poaceae (grass family), accounting for approximately 27% of the world's crop production (Graham and Vance, 2003). On a worldwide basis, legumes contribute about one-third of humankind's protein intake, while also serving as an important source of fodder and forage for animals and of edible and industrial oils. One of the most important attributes of legumes is their unique capacity for symbiotic nitrogen fixation, underlying their importance as a source of nitrogen in both natural and agricultural ecosystems. Legumes also accumulate natural products (secondary metabolites) such as isoflavonoids that are beneficial to human health through anticancer and other health-promoting activities (Dixon and Sumner, 2003). Dendrogram depicting phylogenetic relationships of Papilionoideae legumes. (Figure reprinted from Choi et al. [2004b], based on figure 5 of Doyle and Luckow [2003].) Chromosome number and genome size of major model and crop legumesa Data were from the Plant DNA C-values Database (http://www.rbgkew.org.uk/cval/homepage.html). Chromosome number and genome size of major model and crop legumesa Data were from the Plant DNA C-values Database (http://www.rbgkew.org.uk/cval/homepage.html). An immediate goal of legume genomics is to transfer knowledge between model and crop legumes. Accordingly, an in-depth understanding of conservation of genome structure among legume species is a prerequisite to achieving this goal. The idea that conserved genome structure can facilitate transfer of knowledge among related plant species is best addressed in grasses in which genome macrosynteny and microsynteny have been extensively maintained (Bennetzen, 2000; Devos and Gale, 2000). These studies, however, also revealed many exceptions to the conserved synteny, with frequent local genic rearrangements including gene inversion, duplication, translocation, and insertion/deletion. Although the degeneracy of local genome microstructure has been widely documented, it is less clear the extent to which such alterations to genome microstructure contribute to the divergence of genome function. In this review, we focus on recent results of comparative genome analysis between model and crop legumes, and also highlight the recent successes of using comparative genomics tools for cross-species gene isolation. It has been estimated that upwards of 80% of all angiosperms are likely to have a polyploid origin (Masterson, 1994). It is unlikely that legumes are an exception to this. Soybean, for example, has long been known to be an ancient polyploid with putative homoeologous chromosomal regions readily identified by genetic mapping (Shoemaker et al., 1996; Lee et al., 1999, 2001) and by characterization of homoeologous bacterial artificial chromosome (BAC) clones (Foster-Hartnett et al., 2002; Yan et al., 2003). And recently, segmental duplications within the soybean genome were visualized by fluorescence in situ hybridization of BACs (Pagel et al., 2004). Segmental duplications also were identified in the M. truncatula and L. japonicus genomes through high-throughput genome sequencing (Zhu et al., 2003; N. Young, personal communication). Extensive expressed sequence tag (EST) collections exist for two legumes representing distinct phylogenetic clades, soybean (tropical legumes) and M. truncatula (cool season legumes). Much of a genome's evolutionary history can be read in these transcripts. Schlueter et al. (2004) identified duplicate transcripts from the EST collections of both of these legumes and estimated genetic distances of the pairs using synonymous substitution measurements. It was estimated that soybean probably underwent two major genome duplications events: one at 15 million years ago (MYA) and another at 44 MYA. A genome duplication event also was estimated to have occurred in M. truncatula at approximately 58 MYA. A subsequent analysis using a multigene approach concluded that the more ancient duplication events probably represent a single event that occurred before soybean and Medicago diverged (Pfeil et al., 2005). If this is true, then approximately 7,000 other legumes share the same genome duplication event (Pfeil et al., 2005). Genome duplications often are followed by gene loss, rearrangements, tandem gene or segmental duplications, and divergence of duplicated gene sequences. All of these events are involved in the process of diploidization (Ohno, 1970) and complicate the interpretation of comparative genomic data. Macrosynteny generally refers to conserved gene order between species revealed by comparative genetic mapping of common DNA markers or in silico mapping of homologous sequences. Early comparative studies of legume genomes were focused on closely related species of the same genus or tribe, based primarily on comparative mapping of common RFLP markers. Weeden et al. (1992) first reported conserved gene order between pea and lentil, accounting for approximately 40% of the lentil genome. Later, Menancio-Hautea et al. (1993) demonstrated that mungbean and cowpea (Vigna unguiculata) also exhibited a high degree of linkage conservation, whereas chromosomal rearrangements have occurred since the divergence of the two species. Comparative mapping among mungbean, common bean, and soybean in the Phaseoleae tribe indicated that mungbean and common bean linkage groups were highly conserved, but synteny with soybean was limited only to the short linkage blocks (Boutin et al., 1995). A more recent study, however, using Arabidopsis (Arabidopsis thaliana) as a bridging species revealed that homoeologous segments of soybean chromosomes showed a higher degree of synteny with chromosomes of common bean and mungbean than previously thought (Lee et al., 2001). The most in-depth analysis of legume macrosynteny recently was reported by Choi et al. (2004a, 2004b) using M. truncatula as a central point of comparison. This research took advantage of abundant EST sequence information from the model legume M. truncatula to develop cross-species genetic markers where locus orthology was tested through phylogenetic analysis. Gene-specific PCR primers were designed to anneal to highly conserved exon sequences that span predicted introns, which allowed for efficient PCR amplification across species and for developing single nucleotide polymorphism markers for linkage mapping in multiple taxa. These putatively orthologous markers were mapped in M. truncatula, alfalfa, pea, mungbean (Choi et al., 2004a, 2004b), and chickpea (H. Zhu and D. Cook, unpublished data). In addition, 60 markers developed based on homology to mapped genetic markers of soybean were mapped in M. truncatula. Furthermore, the macrosyntenic relationship between M. truncatula and L. japonicus was evaluated based on 63 pairs of sequenced BAC clones, representing putatively orthologous loci with known genetic position in both species. A simplified consensus map for eight legume species. The figure is based on figure 5 of Choi et al. (2004b) with modification. Mt, M. truncatula; Ms, alfalfa; Lj, L. japonicus; Ps, pea; Ca, chickpea; Vr, mungbean; Pv, common bean; Gm, soybean. S and L denote the short and long arms of each chromosome in M. truncatula. Syntenic blocks are drawn to scale based on genetic distance. By contrast, macrosyntenic relationships between M. truncatula and Phaseoloid legumes were more complicated and less informative. Twenty-nine of the 38 (approximately 76%) markers mapped between M. truncatula and mungbean revealed evidence of conserved gene order, whereas the remaining markers mapped to nonsyntenic positions. Similarly, 23 of the 60 mapped markers identified 11 syntenic blocks between M. truncatula and soybean. The finding that synteny was limited only to small genetic intervals between more distantly related legumes suggests correlation between the frequency of chromosomal rearrangement and divergence time, which also is reflected by the differences in chromosome number between Galegoid and Phaseoloid legumes. In the case of soybean, duplication (polyploidization) followed by gene loss and segmental reshuffling (diploidization) may make it difficult to identify lengthy stretches of syntenic chromosome segments between soybean and related legumes. In contrast with macrosynteny, microsynteny often refers to conserved gene content and order at sequence level over a short, physically defined DNA contig. Nearly all the interspecies analyses of microsynteny reported so far have been based on comparisons of a limited number of specific regions, and the conclusions drawn therein may not be extended to the global level because genome microstructure is highly dynamic and the level of conservation varies with different parts of a genome. Yan et al. (2003) estimated the level of microsynteny between M. truncatula and soybean using a hybridization strategy involving BAC contigs. Twenty-seven of 50 soybean contigs (54%) were shown to possess some level of microsynteny with M. truncatula. Sequence analysis of regions around the putatively orthologous apyrase genes between M. truncatula and soybean also revealed conserved gene order, with at least 6 genes in common over 70 kb (Cannon et al., 2003). Similar comparison was conducted between the rgh1 locus of soybean and the putatively orthologous region of M. truncatula (Choi et al., 2004b). From a total of 29 distinct genes identified in M. truncatula and soybean within the syntenic interval, 14 (approximately 48%) were conserved between the two genomes. More extensive analysis of microsynteny between M. truncatula and L. japonicus was facilitated by the ongoing genome sequencing efforts in both species. Sixty-three pairs of the sequenced BAC clones analyzed shared an average of nine microsyntenic gene pairs (Choi et al., 2004b). Results from detailed analysis of 10 of the 63 clone pairs with broadly spaced genetic positions in the two genomes showed that approximately 82% of identified genes were syntenic between M. truncatula and L. japonicus. Tandem duplication accounts for a 12% and a 17% increase in the number of predicted genes in L. japonicus and M. truncatula, respectively, with only one case that the same homolog duplicated in both species. This observation suggests that the majority of tandem duplication events occurred independently after the divergence of the two species. Intriguingly, in many cases, the regions that were conserved between legumes were also conserved with one or more regions of the Arabidopsis genome, despite at a lower level. The conserved genome structure between M. truncatula and crop legumes has allowed for map-based cloning of genes required for nodulation in crop legumes, using M. truncatula as a surrogate genome (Endre et al., 2002; Limpens et al., 2003). One example is a nodulation receptor kinase (NORK) gene that is required for both bacterial and fungal symbiosis (Endre et al., 2002). Three loci with similar nonnodulation mutant phenotypes were mapped to syntenic locations of M. truncatula, pea, and alfalfa. The closely linked flanking markers in alfalfa were used as probes to pull out M. truncatula BACs that cover the orthologous locus in M. truncatula. Map-based cloning and a complementation test were performed in M. truncatula and eventually led to the simultaneous cloning of three orthologous genes (i.e. DOES NOT MAKE INFECTION2 [DMI2] in M. truncatula, NORK in alfalfa, and SYM19 in pea). At the same time, the ortholog of L. japonicus (called symbiosis receptor-like kinase, or SYMRK), which is located in a syntenic region of M. truncatula, alfalfa, and pea, also was isolated (Stracke et al., 2002, 2004). A similar strategy also was successful for cloning the pea SYM2 orthologous genes of M. truncatula (Limpens et al., 2003). Pea SYM2 is a putative Nod-factor entry receptor involved in the rhizobial infection process (Geurts et al., 1997). Map-based cloning of SYM2 in pea was difficult due to its large genome and the lack of efficient transformation methods. However, the pea SYM2 region is highly syntenic with M. truncatula (Gualtieri et al., 2002). The tightly linked markers flanking the SYM2 in pea were used to identify M. truncatula BACs, and a physical contig (approximately 300 kb) covering the SYM2 orthologous region of M. truncatula was sequenced to identify candidate genes. Using the RNA interference reverse genetic tool, Limpens et al. (2003) showed that two LysM-domain receptor kinases were specifically involved in infection thread formation, and, therefore, are potential orthologs of the SYM2 in pea. The degree of conservation of genome structure between legumes and Arabidopsis is less straightforward. Grant et al. (2000) reported substantial macrosynteny between soybean and Arabidopsis, while comparison between M. truncatula and Arabidopsis revealed a lack of extended macrosynteny between the two genomes (Zhu et al., 2003). Nevertheless, it is obvious that synteny is frequently maintained over small chromosomal segments. In cases of localized synteny, genetically linked loci in M. truncatula often are collinear with several segments of Arabidopsis, consistent with the fact that the Arabidopsis genome has experienced extensive segmental duplication and reshuffling accompanied by selective gene loss (Vision et al., 2000; Bowers et al., 2003). Sequence analyses also revealed networks of microsynteny that are often highly degenerate, similar to that reported by Ku et al. (2000). The erosion of microsynteny could be ascribed to either the selective gene loss from duplicated loci or the absence of close homologs of legume genes in Arabidopsis. Microsynteny of the MtDMI2/LjSYMRK regions and comparison with four segments of Arabidopsis. Lines indicate significant homology matches between predicted genes. The orientations of predicted genes are indicated by arrows. The maps are drawn to scale. The fact that NORK orthologs are extremely conserved in terms of sequence similarity (87%–97%), function, and genomic location among multiple legumes suggests that such divergence of genome microstructure has occurred before the divergence of legume family. The two closest homologs of MtDMI2/LjSYMRK in Arabidopsis are At1g67720 and At2g37050 with a sequence identity of approximately 33%. A TBLASTn search of the M. truncatula Gene Index using the sequences of At1g67720 and At2g37050 identified highly conserved genes from M. truncatula, suggesting that At1g67720 and At2g37050 likely are not the ancient orthologs of MtDMI2/LjSYMRK. Therefore, MtDMI2/LjSYMRK likely is absent in the lineages giving rise to Arabidopsis. Similar divergence of genome microstructure also was observed in comparisons of maize (Zea mays), sorghum (Sorghum bicolor), and rice (Oryza sativa), where a cluster of maize zein storage protein genes are conserved with those of sorghum (kafirin), while the homologs of the storage protein genes are completely missing from the rice orthologous segment and elsewhere of the rice genome (Song et al., 2002). These observations indicate that selective gene loss/retention have played an important role in the divergence of species-specific or family specific phenotypes (e.g. nodulation in legumes). Another legume receptor-like kinase gene required for regulation of the root nodule number recently was cloned from soybean (GmNARK for G. max nodule autoregulation receptor kinase; Searle et al., 2003) and L. japonicus (LjHAR1 for hypernodulation aberrant roots; Krusell et al., 2002; Nishimura et al., 2002). The putative M. truncatula ortholog MtSUNN (supernumerary nodules) also was identified and mapped to the syntenic regions of GmNARK/LjHAR1 (Penmetsa et al., 2003; Schnabel et al., 2003; Choi et al., 2004a). Similar to that observed in the MtDMI2/LjSYMRK regions, the GmNARK/LjHAR1 regions can be aligned with at least seven duplicated segments of the Arabidopsis genome, but none of them contain the GmNARK/LjHAR1 ortholog. The ortholog also was missing in a paralogous region of L. japonicus. In this case, however, the closest homolog of GmNARK/LjHAR1, AtCLAVATA1, is located in a nonsyntenic region of Arabidopsis. Despite the uncertainty of their orthology, AtCLAVATA1 and GmNARK/LjHAR1 share both highly conserved sequence similarity and gene structure, indicating that they are descendants of a common ancestor (Krusell et al., 2002; Nishimura et al., 2002; Searle et al., 2003). Apparently, these homologous genes have experienced functional divergence between plant families, with AtCLAVATA1 controlling stem cell proliferation in Arabidopsis, whereas GmNARK/LjHAR1 functions in the leaf and exerts long-distance control of nodulation (Searle et al., 2003). Such functional divergence appears to have been associated with its relocation in either or both of the lineages leading to Arabidopsis or legumes, and this event should have occurred before the divergence of legume family. As there are two copies of AtCLAVATA1-like genes in soybean (GmCLAVATA1-A and GmCLAVATA1-B; Yamamoto et al., 2000), it was suggested that GmNARK (GmCLAVATA1-B) is likely a duplicated version of AtCLAVATA1 and has been recruited into nodulation function in legumes (Searle et al., 2003). The recruitment of the GmNARK into its current roles is associated with changes in gene expression, leading to differential tissue-specific expression to AtCLAVATA1 and function in controlling the nodule number (Searle et al., 2003). Alternatively, the ancestor of CLAVATA1 might be bifunctional, and the descendent duplicated genes evolve to partition the ancestral function by showing organ- and/or time-specific expression (Adams et al., 2003). Thus, losing one copy of the duplicated and subfunctionalized gene during speciation eventually may lead to species-specific phenotypes. The divergence of plant phenotypes that distinguish one species from another is a big puzzle for plant biologists. In the case of legumes, an interesting question we often ask is as follows: Are genes required for nodulation and nitrogen fixation legume specific? Recent studies have shown that some of the genes required for nodulation (e.g. DMI1 and DMI3 in M. truncatula) are highly conserved between legumes and non-legumes, and their orthologs can be unambiguously defined in non-legumes such as rice and/or Arabidopsis through microsyntenic analysis (Ane et al., 2004; Levy et al., 2004). These results suggested that at least some of the genes involved in nodulation are evolved from broadly conserved pathways of plant development (Parniske and Downie, 2003; Szczyglowski and Amyot, 2003). This view was further supported by Fedorova et al. (2002), who identified 340 tentative consensus sequences with nodule-specific expression patterns, approximately 40% of which shared sequence homology to sequences from non-legumes. It will be a challenging task to undertake the discovery and dissection of the function of the genes that are required for nodulation but are conserved across plant families (e.g. DMI1 and DMI3) in non-legumes. In addition to the recruitment of broadly conserved genes for novel legume functions, legumes also may have evolved novel genes that are involved in legume-specific functions. By comparing unigene sets from M. truncatula, L. japonicus, and soybean to non-legume unigene sets and to the genomic sequences of rice and Arabidopsis, Graham et al. (2004) were able to identify more than 2,500 legume-specific EST contigs, accounting for approximately 6% of genes in legume unigene sets. Motif analysis identified three major gene families from this putative legume-specific gene set, including F-box-related proteins, Pro-rich proteins, and Cys cluster proteins. In particular, the more than 300 Cys cluster proteins, with predicted similarity to defensins, represent approximately 1% of expressed genes in Medicago, primarily from nodules (Fedorova et al., 2002; Mergaert et al., 2003; Graham et al., 2004). Work is under way to characterize these putative legume-specific genes using both forward and reverse genetic tools (K. VandenBosch, personal communication). The past several years have seen tremendous progress in the study of legume genomics, thanks to the development of abundant genetic and genomic resources for two model legumes, M. truncatula and L. japonicus, and the important crop legume soybean. Undoubtedly, model systems will continue to play a critical role in contributing to our understanding of the mechanisms underlying nodulation and symbiotic nitrogen fixation, the most conserved phenotype in legumes. There is still much to learn about the genomic organization of crop legumes such as soybean, groundnut, and common bean. A major challenge for comparative legume genomics is to translate information gained from model species into improvements in crop legumes. The complexity of that challenge may well be defined by the structural and functional similarities and dissimilarities among these very fascinating genomes. We thank Dr. Steve Cannon and Dr. Kathryn VandenBosch for their on the This is with the of the of the
Source: PLANT PHYSIOLOGY
Language: English
Subfield: Plant Science
Domain: Life Sciences
SDG: Zero hunger
Open Access status: bronze
OpenAlex ID: W2159171298