Rice Genetics IV

Contents.

Foreword.

Acknowledgments.

Opening address.

From being a poor cousin to maize, wheat, and tomato for genetic knowledge as recently as the 1980s, rice has become a model plant for molecular genetic research. Numerous scientists in laboratories worldwide have helped make rice a favored higher plant for molecular and cellular genetic studies. Below are some of the major advances in this rapid progress in rice genetics:.

• Van der Stok in 1908 for the first time reported Mendelian segregation in rice.

• Kuwada established the basic chromosome number of rice to be 24 in 1910.

• The first linkage in rice was reported by Parnell et al in 1917.

• Dr. K. Ramiah advocated the standardization of gene symbols in rice.

• Kadam and Ramiah published a review of gene symbols for the first time in 1943.

• Gene symbolization was discussed by the International Rice Commission working party on rice breeding at its Sixth Session in Penang, Malaysia, in 1955.

• Shastry et al numbered the chromosomes in decreasing order of length at the pachytene stage of meiosis in 1960.

• Rules for gene symbolization were reviewed during the symposium on rice genetics and cytogenetics held at IRRI in 1963.

• Nagao and Takahashi proposed the 12 linkage groups of rice in 1963.

• Regeneration of haploids from anther culture was reported by Niizeki and Oono in 1968.

• Independence of linkage groups was tested by Iwata and Omura in Japan and by G.S. Khush et al at IRRI in 1984 through trisomic analysis.

• Publication of the Rice Genetics Newsletter began in 1984 under the editorship of H.I. Oka and G.S. Khush. The First International Rice Genetics Symposium was held at IRRI in 1985 and the Rice Genetics Co-operative was established for international collaboration in rice genetics.

• The Rockefeller Foundation established an International Program on Rice Biotechnology in 1985.

• Yamada et al obtained regeneration from protoplasts in 1985.

• McCouch et al constructed the first molecular genetic linkage map in 1988.

• Transgenic rice plants were produced first by three groups: Toriyama et al, Zhang and Wu, and Zhang et al in 1988.

• The Second International Rice Genetics Symposium was held at IRRI in 1990 and a uniform chromosome numbering system was established.

• The Rice Genome Research Program (RGRP) began at Tsukuba in 1991.

• Ahn and Tanksley constructed comparative linkage maps of the rice and maize genomes in 1993.

• The yeast artificial chromosome (YAC) library in rice was established by the RGRP and a bacterial artificial chromosome (BAC) library by Wang et al in 1995.

• The first agronomically important gene in rice, Xa21, was cloned by Song et al through map-based cloning in 1995.

• The Third International Rice Genetics Symposium was held in Manila in 1995 and the correct orientation of morphological and molecular genetic maps was established.

• An international network on rice genome sequencing was established in 1998 under the leadership of the RGRP.

• Projects on functional genomics began in 1999.

The application of Mendelian genetics has clearly led to many breeding advances in rice as well as in other crops. In this chapter, we emphasize economically important traits for which segregation ratios can be distinguished without elaborate progeny testing or molecular markers. Four general groups of traits are reviewed: agronomic and physiological traits, grain quality, pest resistance, and resistance to abiotic stresses. The single most important trait has been semidwarfism, conferred by the sd1 gene. Other important agronomic and physiological traits are photoperiod sensitivity, glabrous hulls, gold hull color, and purple leaf. Among grain quality characters, amylose content is the most important, ranging from waxy types with essentially no amy-lose to temperate japonica short- and medium-grain types with 16–18% amy-lose, to tropical japonica long grains with 21–24% amylose, and to many indica types with up to 28% amylose. Another significant grain quality trait is aroma, which often appears to be under simple genetic control but which is difficult to recover in high-yielding backgrounds. Breeding for pest resistance, including both diseases and insects, has been one of the most successful examples of the use of major genes in crops, and yet it is a recurring challenge. The most important disease example is blast resistance, which has been a focus of breeding efforts for decades. The exploitation of major genes for bacterial blight resistance has been more successful than for blast. Many successful cases of major gene resistance for brown planthopper, green leaf-hopper, and gall midge have been reported in the past three decades. In general, resistance to abiotic stresses has been polygenic. Exceptions have been low-temperature-induced chlorosis at the seedling stage and submergence tolerance.

The Rockefeller Foundation's design of a long-term program on rice biotechnology was the product of a 2-year intensive survey and analysis of the genetic prospects for the world's major food crops conducted in the early 1980s. In late 1984, the Foundation's Board of Trustees approved a strategy for a 10–15-year program. That program was highly speculative and indicated substantial risk with regard to the status at that time of cereal plant molecular biology and rice in particular. During the first 5–7 years, projects supported by the Foundation laid the scientific basis for "rice biotechnology" as we know it today. Early successes were the first DNA molecular marker map of rice, the regeneration and transformation of rice, the use of rice pest genomic information to unravel age-old riddles of host-plant resistance, and numerous other discoveries that changed the way rice geneticists viewed breeding objectives such as insect resistance, abiotic stress tolerance, and hybrid rice. These discoveries culminated in the revelation of rice's pivotal genomic position in the evolution of cereal species. Over the ensuing 7–8 years, the program shifted its focus to the transfer of the resulting biotechnologies to institutions in rice-producing and -consuming countries. This task required the strengthening of both physical and human resources in cooperation with national and international rice research systems in Asia, Africa, and Latin America. The Foundation's program management sought to support further technology generation and application while promoting the program's greatest asset, international collaborative research-cum-training. This "win-win" component of the program linking fledgling national rice biotechnology efforts directly to advanced research institutes in the United States, Europe, Japan, and Australia became the hallmark of the Foundation's management strategy. During the program's 17-year lifetime, more than 400 (primarily Asian) rice scientists were trained in this manner. The successful linkage of research in cutting-edge biotechnology with the training of rice scientists often produced long-term collaborative relationships that outgrew dependence on Foundation support and continue today (such as the IRRI-managed Asian Rice Biotechnology Network). Some of these successes were undoubtedly a consequence of the basic research progress in rice plant molecular genomics, which brought greater financial support for rice-centered research as rice became the "model cereal" for genomic research, rivaling even Arabidopsis.

This chapter intends to explore some implications of rice evolution from the viewpoints of genetics and ecology. Core issues are (1) what genetic changes are associated with differentiation among and within species of cultivated rice and their wild relatives and (2) what factors are responsible for the domestication process. First, genetic diversity among and within AA genome species is summarized. Second, four directions of differentiation within the Asian AA genome gene pool are clarified: differentiation from wild to cultivated type (domestication), ecotype differentiation from perennials to annuals in wild races, geographical variation in wild races, and indica-japonica differentiation in cultivars. Third, the genetic basis of the domestication syndrome is discussed. Our recent study demonstrated that mapped genomic locations of quantitative trait loci (QTLs) tended to cluster, reflecting the domestication syndrome as well as the indica-japonica syndrome. This phenomenon was explained by "multifactorial linkages."Domestication might be a process driven by conscious and unconscious selection of adaptive gene blocks distributed over the genome.

In the past decade, a wealth of data provided by molecular markers, together with phenotypic, ecological, and archaeological data, significantly increased our evolutionary understanding of the genus Oryza. The target species dealt with in this chapter are diploid AA genome species—cultivated rice and its wild relatives. Several important problems such as the genetic basis of reproductive isolation are not included, but some new information obtained from our recent studies is discussed.

The past 15 years have seen an intense research drive to apply the new molecular biology to rice. Initiatives such as the Rockefeller Foundation's International Program on Rice Biotechnology, begun in the mid-1980s, have been underpinned by the in-depth corporate knowledge of the crop built up by research organizations in Southeast Asia such as the International Rice Research Institute and accelerated by the application of a vast, and previously uncoordinated, research capacity in national programs in the area. Moreover, unlike the other two 500-million-ton crops, wheat and maize, rice has a small tractable genome, and the development of genetic and genomic tools not available in any other cereal has ensured the promotion of rice as a favored research target. On top of all this, the discovery that gene content and gene order—genome colinearity—have been maintained over the whole grass family, which includes all cereals and many forage crops, has elevated rice still further to the status of a "model" organism. The initiation of genomic DNA sequencing efforts in the public and private sector will further ensure rice's central position in plant science.

In this chapter, we will describe the ways in which rice genomic tools and knowledge of the rice genome are already being applied in research on the other major cereals, wheat and maize. Moreover, many aspects of rice genetics can be transferred to the "orphan" crops, the several minor economic grass species that have not themselves warranted extensive research and breeding.

The genus Oryza, consisting of approximately 24 species, contains an enormous gene pool for the genetic improvement of rice cultivars. To realize the potential agricultural value of wild rice germplasm, continuous efforts have been made to understand the taxonomy, genomic composition, genetic diversity, and phylogeny of the Oryza species. Based on genome analyses using cytogenetic methods and genomic DNA hybridization, nine genome types (A, B, BC, C, CD, E, F, G, and HJ) were determined for various Oryza species. Recently, we amplified by polymerase chain reaction, cloned, and sequenced two nuclear genes, Adh1 and Adh2, and a chloroplast gene, matK, from all of the Oryza species. Phylogenetic relationships of rice genomes and species were reconstructed by comparing three gene phylogenies. The results supported previous recognition of nine genome types. A new genome type, HK, was recognized for Oryza schlechteri and Porteresia coarctata, suggesting that P. coarctata might be included in Oryza. The study further revealed that the EE genome species is most closely related to the DD genome progenitor that gave rise to the CD genome. In contrast to a single origin of three CCDD genome species, the BBCC genome species had different origins because their maternal parents had either a B or C genome. The G genome is the most basal lineage on the phylogeny. The AA genome group, which contains the two cultivated rice species, is a recently diverged and rapidly radiated lineage within the rice genus. Despite our better understanding of phylogenetic relationships of Oryza species based on the three genes, relationships of rice genomes and species still remain partially resolved or weakly supported. Finally, we discuss some remaining questions and future perspectives concerning studies on the phylogeny and evolution of the rice genus.

Miniature inverted repeat transposable elements (MITEs) are the most prevalent elements associated with the genes of maize and rice, in which they reside in introns and in 5′ and 3′ flanking regions. Several MITEs have recently been amplified in the genomes of maize and rice, thus giving rise to structurally distinct alleles. This MITE-mediated variation can be assayed and quantified using a modification of the amplified fragment length polymorphism method called transposon display. In this way, the genome-wide distribution of MITEs can be mapped and MITE markers isolated and added to geneticists' toolboxes.

A well-distributed set of 500 microsatellite markers has been genetically mapped onto the rice genome. These markers link the genetic and physical maps with the genomic sequence of rice, facilitating studies that seek to determine the relationship between the structure and function of genes and genomes. To facilitate the development of new microsatellite markers using publicly available DNA sequence information, a simple sequence repeat identification tool (SSRIT) has been developed for semiautomated identification of nonredundant simple sequence repeat (SSR) loci and for primer design (available at http://www.gramene.org/microsat/). Using this script, a total of 57.8 Mb of DNA sequence from rice was searched to determine the frequency and distribution of different SSRs in the genome. Because the length of the SSR unit in any single genome has proven to be a reasonably good predictor of overall polymorphism in related genotypes, SSR loci were categorized into two groups (class I and class II) based on the length of the repeat motif. Microsatellites with poly(AT)n repeats represented the most abundant and polymorphic class of SSRs but were frequently associated with the Micropon family of miniature inverted repeat transposable elements (MITEs) and were difficult to amplify. Estimates of total microsatellite frequencies in rice suggested that there were approximately 28,340 Class I and 70,530 Class II SSRs, or about 100,000 di-, tri-, and tetra-nucleotide SSR motifs in the rice genome. The distribution of SSR sequences showed that regions of the rice genome that were richer in expressed genes also tended to be richer in SSR sequences, underscoring their usefulness as genetic markers. Applications of SSRs in variety protection, diversity analysis, gene and QTL identification, marker-assisted selection, physical mapping, and gene isolation will be discussed.

Since the late 1980s, many loci controlling both qualitative and quantitative traits of rice have been mapped using DNA markers. This chapter focuses on the mapping of major genes in rice. The distinction between major and minor genes has become unclear because of advances in molecular-marker mapping. A major gene can be involved in determining both qualitative and quantitative traits. Here we define a major gene as a locus that results in discrete phenotypes in a segregating population or one that controls more than 50% of the phenotypic variation for a continuously distributed trait. The latter is common for major genes controlling traits that are highly influenced by environment. Resistances to diseases and insects have received the most attention for mapping because of their importance as breeding objectives. Molecular mapping of major genes is important in determining the allelism of genes conferring identical phenotypes, use as a selectable marker in a breeding program, and positional cloning of genes. Molecular markers will be most useful for selection when (1) the phenotype is difficult or expensive to measure directly, (2) genes of similar phenotype are being pyramided into a single line, or (3) markers are being used to select against the donor genome in a backcrossing program.

Advances in DNA markers and molecular linkage maps have stimulated a new area of molecular quantitative genetics through mapping of quantitative trait loci (QTLs). Several important questions regarding the types and number of QTLs, QTL × environment (QE) interactions, molecular dissection of trait correlation, and gene actions of QTLs are addressed based on results from many previous QTL mapping studies in rice and examples from two well-studied mapping populations and their related progenies. Two major types of QTLs, main-effect QTLs and epistatic QTLs, are recognized. Many QTLs are found to affect specific quantitative traits and they are widely distributed in the genome, though only a limited number of these loci are detectable in a mapping population largely because of epistasis and genotype × environment (GE) interactions. The effects (both main and epistatic) of individual QTLs affecting specific phenotypes may vary considerably. Most QTLs appear to be epistatic and complementary interaction appears to be the most common form of epistasis. Most QTLs tend to show varied degrees of QE interactions as a result of differential gene expression to biotic and abiotic stresses in different environments. QTLs differ greatly in their GE interactions and epistasis plays an important role in QE interactions. QTLs showing different gene actions appear to belong to different groups of genes and those exhibiting both additive and nonadditive gene actions are few. A new strategy is proposed for simultaneous QTL identification and transfer, and allele discovery through the development of introgression lines and use of DNA markers.

We have conducted a series of studies to elucidate the biological mechanism of heterosis using rice as the model system. We analyzed the genetic basis of heterosis using an F_2:3 population and an "immortalized" F₂ population derived from a cross between Zhenshan 97 and Minghui 63. Our results demonstrated the involvement of large numbers of two-locus interactions, or epistasis, underlying the genetic basis of quantitative traits and heterosis. We assessed the relationship between gene expression and heterosis by assaying the patterns of differential gene expression in hybrids relative to their parents in a diallel cross. The analysis revealed that differentially expressed fragments occurring in only one parent of the cross were positively correlated with heterosis, and fragments detected in F₁s but not in the respective parents were negatively correlated with heterosis. For further analysis, 384 fragments were recovered from gels and arrayed onto nylon membranes. Hybridization with RNAs from seedling and flag leaf tissues detected an overall elevated level of gene expression in the hybrid compared with the parents. Several fragments showed much higher expression in the highly heterotic hybrid than in other hybrids. Many of these fragments were sequenced and mapped to the rice linkage map, which provided insights into the understanding. We believe that analyses combining genetic and molecular approaches will eventually lead to the characterization of the biological mechanisms of heterosis.

The rice genome sequencing project has been pursued as a national project in Japan since 1998. At the same time, a desire to accelerate the sequencing of the entire rice genome led to the formation of the International Rice Genome Sequencing Project (IRGSP), initially comprising five countries. The sequencing strategy is the conventional clone-by-clone shotgun method using P1-derived artificial chromosome/bacterial artificial chromosome (PAC/BAC) libraries from rice variety Nipponbare as a common template resource. As of September 2000, ten countries from this international collaboration had already contributed about 30 Mb of the rice genome sequence. Analysis of the rice genome should facilitate a better understanding of the concept of inheritance in the rice plant and the development of new research endeavors in physiology and biochemistry. Crucial information from nucleotide sequences will be useful for improving breeding technology as one of the ultimate goals of rice genome research.

Finishing genomic sequence is the process of assembling and refining raw sequence data to produce a complete and accurate final sequence. The goals of finishing are to achieve contiguity, delineate insert-vector junctions, and resolve sequence discrepancies and ambiguities such that the error rate of the sequence is less than one in 10,000 bases. Such a level of accuracy is often difficult to establish due to a repetitive sequence or other secondary structures that require the use of more complicated finishing techniques. Because of such complexities, the finishing process is the bottleneck in the sequencing queue, and as such it is extremely beneficial to make the draft data publicly available before the sequence is completed. Much of the valuable information contained in the sequence can be determined from data at this earlier stage.

The CCW Rice Genome Sequencing Consortium is funded to sequence and annotate the short arms of rice chromosomes 10 and 3—approximately 30 Mb of DNA. To efficiently sequence the rice genome, the Clemson University Genomics Institute has developed a framework consisting of two deep-coverage bacterial artificial chromosome (BAC) libraries and BAC fingerprint, genome-anchoring, and sequence-tagged connector databases. These resources have been provided to the International Rice Genome Sequencing Project and are being used extensively. In this chapter, we summarize the framework project and its use to sequence the short arm of chromosome 10.

Recent progress in rice genome analysis has made it possible to analyze naturally occurring allelic variation underlying complex traits. Using heading date as a model for complex traits, we detected and characterized quantitative trait loci (QTLs) and identified genes at QTLs at the molecular level. QTLs for heading date were mapped by using several types of progeny derived from a cross between varieties Nipponbare and Kasalath. Nine QTLs were mapped precisely as single Mendelian factors by the use of advanced backcross progeny. Nearly isogenic lines of QTLs were also developed by marker-assisted selection and were used to identify the function of each detected QTL. Combining two QTLs into the genetic background of Nipponbare allowed us to investigate epistatic gene interactions among QTLs. We analyzed a large segregating population by genetic and physical mapping to narrow down candidate genomic regions for target QTLs. These analyses revealed 10- to 50-kb regions as candidate regions for the target genes. We identified genes of the most probable candidates for the photoperiod-sensitivity loci Hd1 and Hd6. Naturally occurring allelic variation could be a new resource for the functional analysis of rice genes.

A collection of IR64 mutant populations has been established using diepoxybutane, fast neutron, and gamma ray mutagenesis. Phenotypic screening is being conducted on about 12,000 M₃ or M₄ lines for morphological variations and altered response to biotic and abiotic stresses. About 6% of the mutants exhibit morphological variations in the vegetative and reproductive stages. Disease-response mutants are recovered at approximately 0.3%. Based on the frequency of visible and conditional mutants observed under biotic stresses, the number of mutated sites per genome is estimated to be high (>10) in these mutants.

Mutants are screened for gain and loss of resistance against bacterial blight, blast, sheath blight, and tungro viruses. We have identified mutants with a gain in resistance to tungro viruses, blast, and bacterial blight. Lesion mimic mutants with enhanced resistance to both blast and bacterial blight are found. To dissect the defense pathways, double disease-response mutants have been produced to examine the epistatic interactions and expression profiles of host defense genes. The mutant collection is being evaluated for tolerance of submergence and salinity. Mutants are also being screened for altered response under water stress at the flowering stage.

Nucleotide-binding site and leucine-rich repeat sequences are used to detect genomic changes in loss-in-resistance mutants; however, the efficiency of this approach is limited by the number of available candidate genes. To efficiently assign DNA sequences to mutant phenotypes, we are developing a gene array-based screening strategy using subtractive cDNA/expressed sequence tag libraries enriched for stress-response genes. We are also producing additional mutant populations to generate a range of deletion sizes that are suitable for array-based and high-throughput PCR screening. As a first step, we are using stress-response and morphological mutants with phenotypic and genetic descriptions to develop a mutant database to be linked to sequence databases. We expect the mutant database to grow as the research community continues to use this mutant resource.

T-DNA insertions have produced approximately 30,000 transgenic lines of rice. Polymerase chain reaction and genomic DNA gel-blot analyses have shown that approximately 65% of the population contains more than one copy of the inserted T-DNA. Hygromycin resistance tests determined the number of T-DNA inserts to be an average of 1.4 genetic loci per plant. From this, it can be estimated that at least 40,000 taggings have been generated. The binary T-DNA vector used in the insertion contained the promoterless β-glucuronidase (gus) reporter gene with an intron as well as multiple splicing donors and acceptors immediately next to the right border. This gene trap vector is designed to detect a gene fusion between gus and the endogenous gene that is tagged by the T-DNA. The leaves, roots, mature flowers, and developing seeds of the transgenic rice plants were subjected to histochemical GUS assays. The results showed that 1.6–2.1% were GUS-positive in the tested organs and that their gus expression patterns were organ- or tissue-specific or ubiquitous in all parts of the plant. This large population of T-DNA tagged lines will be useful for identifying insertional mutants in various genes and for discovering new genes in rice.

Genome and expressed sequence tag sequencing in rice provides a vast resource of gene sequences whose functions need to be determined by reverse genetics methods for expression and mutational analysis. To develop insertional mutagenesis strategies in rice, we transformed japonica and indica cultivars with maize transposon constructs, for knockout and gene detection insertions. A green fluorescent protein (GFP) excision assay developed enabled the visualization of transposon excision in a variety of tissues. Surprisingly, early Ac excision was observed directly after transformation from a construct containing the strong double CaMV enhancer element adjacent to the Ac promoter. We identified genotypes with Ac amplification events and with a forward transposition rate of 15–50% that are useful for generating lines containing multiple transposons. The sequence of DNA flanking transposed Ac provided a resource of Ac-tagged sites, which represented about 50% in the target region, indicating insertional specificity appropriate for the identification of mutants of sequenced genes. Clustered Ac transposition was revealed by six insertions in 70 kb of chromosome 6. Gene detection Ac-Ds enhancer trap and activation tag transformants revealed active transposition in about half the lines. These resources for functional genomics are developed by an EU-funded consortium and will be made available to rice researchers worldwide.

Five endogenous active retrotransposons have been found in rice. Among them, the most active one, called Tos17, was characterized in detail. Tos17 is silent under normal conditions and becomes active only under tissue culture conditions. Five to 30 transposed Tos17 copies were found in each plant regenerated from culture. Tos17 was shown to transpose preferentially into low-copy-number, gene-rich regions, indicating that Tos17 can be used as an efficient insertional mutagen. A collection of 32,000 regenerated rice lines carrying about 256,000 insertions was generated, and these lines are being used for forward and reverse genetic analyses. By using a transposon-tagging strategy, causative genes for viviparous, dwarf, semidwarf, brittle culm, pale green, and narrow leaf mutations, among others, have been cloned. For reverse genetic studies, two strategies are being employed. One is the polymerase chain reaction (PCR) screening of mutants of the gene of interest. We screened 12,000 lines and found mutants of 15 genes, including MAPK, MADS-box, and P450 genes, among the 47 genes analyzed. This suggests that at least 37,000 lines are required for saturation mutagenesis. Another important strategy is the random sequencing of mutated genes by isolating the sequences flanking transposed Tos17. The flanking sequences are amplified by TAIL (thermal asymmetric interlaced)- and suppression-PCR and directly sequenced. Until now, 7,376 independent flanking sequences from 2,134 lines have been determined and mutants of different classes of genes have been identified.

As rice genomics data continue to accumulate at a rapid rate, databases are becoming more valuable for storing and providing access to large and rigorous data sets. This chapter gives an overview of available resources on rice bioinformatics and their role in elucidating and propagating biological and genomic information on rice. Of particular focus here is the informatics infrastructure developed at the Rice Genome Research Program (RGP) following an extensive rice genome analysis. The database named INE (INtegrated Rice Genome Explorer) integrates genetic and physical mapping information with the genome sequence being generated in collaboration with the International Rice Genome Sequencing Project (IRGSP). Database links are initially evaluated using a query tool to explore and compare data across the rice and maize genome databases and for potential application to multiple-crop database querying. A proposed logistics for interlinking these resources is presented to integrate, manipulate, and analyze information on the rice genome. One of the biggest challenges of rice bioinformatics lies in the emerging role of rice as a model system among grass crop species. In view of the importance of comparative genomics in the formulation of new knowledge on plant genome structure and function, bioinformatics remains an essential strategy for gaining new insights into the needs and expectations of rice genomics.

Pi-ta-mediated resistance in rice controls strains of the rice blast fungus that express avirulent alleles of AVR-Pita in a gene-for-gene manner. Map-based cloning of AVR-Pita identified a gene predicted to encode a neutral zinc metalloprotease. We also cloned Pi-ta, a centromere-linked resistance gene on rice chromosome 12, using a map-based cloning strategy. It encodes a predicted 928-amino acid cytoplasmic receptor with a centrally localized nucleotide-binding site. Pi-ta is a single-copy gene and it shows constitutive low-level expression in both resistant and susceptible lines of rice. Susceptible rice lines contain pi-ta^- alleles encoding proteins that share in common one amino acid difference relative to the Pi-ta resistance protein, serine instead of alanine at position 918. Direct interaction of the Pi-ta and AVR-Pita proteins was demonstrated using the yeast two-hybrid system and in vitro binding assays. Our current hypothesis is that the Pi-ta protein is an intracellular receptor that binds to the mature AVR-Pita protease inside the host cell, initiating Pi-ta-mediated defense responses. Understanding how pathogen recognition occurs provides the opportunity to manipulate recognition specificity. Population analysis of AVR gene structure and dynamics will aid in the deployment of resistance genes. The relationship between the Pi-ta and Pi-ta² genes is discussed.

We study signaling pathways involved in the interaction of rice and the rice blast fungus, Magnaporthe grisea, by using molecular and genetic methods. We recently demonstrated that the rice homologue of the small GTPase Rac is an important molecular switch for resistance gene–mediated disease resistance against rice blast infection. Furthermore, it was shown to activate the production of reactive oxygen species as well as cell death, both of which are often observed during the resistance response to avirulent pathogens in many plant systems. In another approach, we study lesion-mimic mutants of rice showing increased resistance to the rice blast fungus. Biochemical analysis of these mutants allowed us to identify genes that may play an important role in disease resistance of rice.

The recent cloning and characterization of several rice genes with resistance to pathogens represent a breakthrough in our understanding of the molecular basis of disease resistance and also provide a starting point for dissecting the resistance pathway in rice. The first resistance gene cloned in rice was Xa21, a gene introgressed from the wild rice Oryza longistaminata. It encodes a putative receptor-like kinase consisting of leucine-rich repeats (LRRs) in the extracellular domain and serine/threonine kinase in the intra-cellular domain. Sequence analysis of seven members of the gene family at the locus suggests that duplication, recombination, and transposition have occurred during the evolution of this gene family. Experiments with a truncated member indicate that the LRR domain determines race-specific recognition and is subject to adaptive evolution. To identify additional components in the Xa21-mediated resistance pathway, both the yeast two-hybrid screen and mutagenesis approaches are being used. Several defense-related genes were found to interact with the Xa21 protein in yeast when the kinase domain was used in the screen. Using diepoxybutane and fast-neutron mutagenesis, we recovered 31 mutants that have changed from resistant to fully susceptible (10) or partially susceptible (21) to nine races of the bacterial blight pathogen in the Philippines. All fully susceptible mutants showed changes at the Xa21 locus as detected by polymerase chain reaction and Southern hybridization. For the partially susceptible mutants, no detectable changes were found at the Xa21 locus, suggesting that these mutations occur at other loci controlling the Xa21-mediated defense pathway.

High-throughput analysis of expressed genes, achieved by cataloguing expressed sequence tags (ESTs) and monitoring hybridization patterns by microarrays, has recently become possible in rice. As the first results become available, the value of these technologies can be gauged. Through ESTs and microarrays, we can obtain a more complete view than in the past of plant gene complexity, tissue specificity, and developmental or externally affected expression patterns. In particular, EST and microarray analyses can have tremendous impact in plant breeding, based on accelerated identification of complex traits such as those controlling plant responses to abiotic stresses. Owing to the novelty and lack of refinement in the use of microarray technology, we discuss advantages and limitations. We demonstrate responses to salt stress in rice (Oryza sativa) monitored by microarray analysis.

Programmed cell death has been observed in various developmental processes in higher plants. The presence of air-filled spaces known as aerenchyma in rice is considered to be an important anatomical adaptive feature necessary for plant survival under flooded conditions. Ample evidence has been presented to show that this system provides a diffusion path for the transportation of oxygen from aerial plant parts to roots. In addition, the molecular dissection of cell death is discussed.

Cyclin-dependent protein kinases (CDKs) are serine/threonine protein kinases that are involved in the regulation of the eukaryotic cell cycle. The growing list of CDKs in plant cells suggests that, during development, each CDK may play a specific role at a specific time in the cell cycle. Rice plants also express different kinds of CDKs. We present the outcome of our work indicating that the division of rice cells might be regulated by different sets of CDK/cyclin, leading to cell proliferation.

Apomixis is asexual reproduction through seed. It occurs in more than 300 plant species but appears to be absent from the genus Oryza. One of the most important applications of apomixis in rice would be to develop true-breeding hybrid genotypes that would allow poor farmers to benefit from hybrid yield advantage. Although considerable progress has been made in mapping loci that control apomixis in other members of the family Poaceae, no gene capable of inducing apomixis has yet been isolated for introduction into rice. This chapter reports on the development of molecular tools to achieve a synthetic apomixis in rice, in which the sexual embryo would be replaced by an asexual embryo induced in the nucellus. We cover two objectives of the research: induction of the nucellar embryo and ablation of the zygotic embryo. We emphasize the tasks of identifying embryo-inducing genes, placing them under the control of nucellus-specific promoters, and assaying their efficacy in rice.

Plant viruses together with their insect vectors cause considerable losses in rice production. Several viruses attack rice plants. Resistance sources are limited or the available resistance genes are either difficult to transfer into cultivated rice or are under threat of being overcome by evolving viral strains. Pathogen-derived resistance (PDR), the expression of pathogen-derived transgenes in plants to interrupt the virus infection cycle, has been employed as an alternative strategy. This approach has given resistance in rice against rice stripe virus (RSV), rice tungro spherical virus (RTSV), rice tungro bacilliform virus (RTBV), rice ragged stunt virus (RRSV), rice hoja blanca virus (RHBV), and rice yellow mottle virus (RYMV). PDR for viruses was originally achieved using gene constructs designed to express wild-type or dysfunctional viral proteins. In many cases, however, these PDR genes appear to be operating at the RNA level associated with posttranscriptional gene silencing. Our latest research shows that viral immunity can be efficiently obtained using constructs that produce dsRNA. This approach has great promise for producing virus resistance in rice. Our experiences and those of other research groups in engineering virus resistance are also discussed.

Transgenic approaches offer new opportunities to improve tolerance for dehydration stress in rice by incorporating genes that are involved in stress tolerance. In this chapter, we present the results of transforming rice separately with plasmids containing genes that encode a group 3 LEA (late embryogenesis abundant) protein, a group 2 LEA protein, a group 1 LEA protein, and a delta 1-pyrroline-5-carboxylate synthetase (P5CS). In each experiment, the growth of transgenic plants under dehydration stress was shown to be faster than that of nontransformed control plants. To maximize gene expression and plant growth, we compared the results of using a constitutive promoter and a stress-inducible promoter in driving the expression of the P5CS gene. Next, the effect of using a matrix attachment region sequence on transgene expression was tested. Finally, we compared two commonly used methods for transforming rice.

Attempts have been made to transfer C₄ traits to C₃ plants by introducing a chimeric gene construct containing cDNAs for C₄ enzymes under the control of active promoters in C₃ plants. However, the levels of transcripts and proteins in these transformants were far below those in C₄ plants. Our studies have demonstrated that the promoters for maize C₄-specific genes encode phosphoenolpyruvate carboxylase (PEPC) and pyruvate, orthophosphate dikinase (PPDK), which can drive the high-level expression of a reporter gene in transgenic rice plants in an organ-specific, mesophyll-specific, and light-dependent manner as in maize. These results suggest that the rice plant possesses the regulatory factors necessary for high-level expression of the C₄-specific genes, and imply that the introduction of the intact maize genes would lead to the high-level expression of the C₄ enzymes in rice leaves. The introduction of the intact maize C₄-specific genes containing all exons and introns and its own promoter and terminator sequences led to the high-level expression of the PEPC and PPDK proteins in the leaves of transgenic rice plants. The activities of PEPC and PPDK were 110- and 40-fold higher than those of nontransgenic rice, respectively. The high-level expression of each C₄ enzyme altered metabolism slightly but did not seem to increase the photosynthetic efficiency of transgenic rice leaves.

As in other direct DNA transfer systems, transgenes delivered by particle bombardment integrate into the cereal genome by illegitimate recombination assisted by short regions of homology. Exogenous plasmid DNA tends to concatemerize prior to integration, resulting in tandemly arranged contiguous copies. However, recent research has shown that such concatemers may positively influence the integration of further transgenes nearby, perhaps by recruiting DNA repair complexes to the site of the original break. Transgenic loci therefore tend to comprise clusters of contiguous copies interspersed with short regions of genomic DNA. The analysis of transgenic wheat chromosomes by fluorescence in situ hybridization (FISH) indicates a further level of organization, where several transgene clusters integrate in the same region of the chromosome arm, but individual clusters produce separable FISH signals at metaphase, suggesting they are interspersed by large segments of genomic DNA. A segregating transgene locus therefore has a hierarchical structure, showing up to three levels of organization.

The analysis of transgene integration shows that recombination plays a strong role in determining integrated transgene structure. Particularly, certain sites in the transformation plasmid can provide hotspots for recombination, leading to particular types of transgene rearrangement. Since the failure of transgene expression often reflects such rearrangements, the modification of transformation vectors to eliminate troublesome sequences should improve the efficiency of transformation. The "clean DNA" system takes this strategy to its logical extreme by removing all unnecessary backbone elements and using just minimal linear cassettes (promoter, coding region, terminator) for transformation. Such experiments consistently generate transgenic plants with simple integration patterns, lower transgene copy numbers, fewer rearrangements, and stable transgene expression. The clean DNA system also allows direct transformation with multiple genes, but the integration patterns remain simple and silencing occurs very rarely.

Silencing often correlates with DNA methylation, and it is of particular interest to identify sequences in plant transgenes that induce methylation and result in unstable expression. There is evidence that methylation can be induced by interactions between homologous transgene copies (or between a transgene and homologous endogene) or may reflect genomic position effects. Recent data suggest, however, that more complex underlying processes may be involved. In a line of plants containing a single-copy three-gene transgenic locus, different forms of silencing could occur de novo and affect individual transgenes, even though the three were separated by only a few nucleotides. Remarkably, two completely different modes of silencing, associated with different methylation patterns, could be established in adjacent transgenes. Transposon mutagenesis is one of the most powerful tools available to gain an understanding of factors underlying transgene expression and stability. We have a population of more than 500 independent indica rice transformants containing the maize Ac transposon, representing a total of more than 1,500 plants. We have observed active excision and re-insertion of the Ac element in this population, which can now be scaled up to look for genes affecting genomic methylation and transgene expression.

Genome surveillance systems of higher organisms protect not only against intragenomic parasites such as retroposons and transposons but also against invasive DNA introduced via genome transformation, resulting in (trans)gene silencing at both the transcriptional and posttranscriptional levels. Transcriptional gene silencing of both bialaphos resistance and Btt CryIIIA crystal toxin genes was encountered in experiments targeted to provide resistance to the rice water weevil. Extensive methylation of the transgenes, especially of the promoter elements, was evident. Germination of seedlings from silenced lines in the presence of 5-azacytidine (which prevents cytosine methylation) led to the reactivation of bialaphos resistance in progeny of silenced plants. The CaMV 35S promoter has been implicated in many instances of silencing and the possibility that some promoters are especially prone to silencing is supported by the finding that RCg2, a rice root-specific promoter, is silenced in more than 80% of transformants. Genes flanking RCg2/uidA are rarely silenced, showing that silencing can be highly targeted. Multicopy insertions are very susceptible to silencing and epistatic interactions between multicopy and single-copy inserts are documented. Several strategies to alleviate silencing have been considered, such as sequence diversification and flanking transgene inserts with matrix attachment regions.

Rice molecular breeding workshop.

Functional genomics workshop.

Bioinformatics workshop.

Rice genetics cooperative.