Table of Contents

EVOLUTIONARY PHASE: BROADENING GENE POOL OF RICE CULTIVARS

Wide Hybridization

Gene transfer from wild species to rice

Cytoplasmic diversification

Tagging alien genes with molecular markers

Somaclonal Variation

Genetic Engineering

Transgenic rice for modifying yield potential

Transgenic rice for insect resistance

Transgenic rice for disease resistance

Transgenic rice for abiotic stress tolerance

Transgenic rice for improved nutritional quality

EVALUATION PHASE: INCREASING SELECTION EFFICIENCY

Anther Culture

Molecular Marker Technology

Molecular Maps

Molecular marker aided selection (MAS)

Gene tagging in rice

QTL mapping

Pyramiding major genes

Synteny relationships

Map based cloning of genes

Functional genomics

Potential impact

Rice is the most important food crop and a staple food for 40% of the world population. More than 90% of rice is produced and consumed in Asia. It is grown under a wide range of agro-climatic conditions. Major advances have been made in increasing rice production worldwide as a result of large-scale adoption of modern high-yielding rice varieties and improved cultural practices. The rapid rice production increase was mainly achieved through the application of principles of classical Mendelian genetics and conventional plant breeding methods. However, rice production must increase by 50 percent to meet the growing demand of this ever increasing population in 2030. To meet this goal, rice varieties with higher yield potential, durable resistance to diseases and insects, and tolerance to abiotic stresses are needed.

Plant breeding consists of two phases: the evolutionary phase and the evaluation phase. Recent advances in cellular and molecular biology and rice biotechnology have produced new tools to increase the efficiency of both phases (Khush and Brar, 1998).

EVOLUTIONARY PHASE: BROADENING GENE POOL OF RICE CULTIVARS

Genetic variability for agronomic traits is the key component of breeding programs to broaden the gene pool of rice and other crops. In conventional plant breeding approaches mainly the genetic variability available within O. sativa germplasm has been exploited. However, genetic variability for many traits such as tolerance to stemborer, tungro, sheath blight and salt stress is limited in the cultivated germplasm. Breeders, therefore, search for genetic variability in other gene pools involving wild relatives of Oryza and apply new techniques for creation and transfer of variability through somaclonal variation and genetic engineering.

Wide Hybridization

Wide hybridization involving hybridization between rice and related wild species to broaden the gene pool of rice. The genus Oryza is comprised of 24 species representing AA, BB, CC, BBCC, CCDD, EE, FF, GG, HHJJ and HHKK genomes. Of these species, O. sativa (2n=24 AA) is cultivated worldwide whereas O. glaberrima (2n=24 AA) is grown in West Africa. The wild species are an important reservoir of useful genes. However, several barriers are encountered in transferring useful genes from wild species to cultivated rice (Brar and Khush 1997). The barrier most commonly encountered is lack of crossability because of chromosomal and genic differences. Biotechnology tools such as embryo rescue and protoplast fusion have been employed to overcome the crossability barriers, and several interspecific hybrids have been produced. More recently, molecular techniques have been employed in the precise monitoring of alien gene introgression.

Gene transfer from wild species to rice

Hybrids between rice and distantly related wild species are usually difficult to produce. Low crossability and abortion of hybrid embryos are the common features of such crosses. Hybrids have been produced through embryo rescue between elite breeding lines or varieties of rice and several accessions of wild species representing BBCC, CC, CCDD, EE, FF, GG, HHJJ, and HHKK genomes. A number of useful genes for resistance to brown planthopper (BPH), whitebacked planthopper (WBPH), bacterial blight (BB), blast and tungro disease have been transferred from wild species to rice (Khush et al. 1990, Brar and Khush 1997) (Table 1). The first example of transfer of a useful gene from wild species is the introgression of a gene for grassy stunt virus resistance from O. nivara to cultivated rice (Khush 1977). Grassy stunt resistant varieties, IR28, IR29 and IR30 were released for cultivation in 1974 and IR32, IR34, IR36 were released in later years. Grassy stunt resistance gene has now been incorporated into numerous varieties.

Cytoplasmic diversification

The wild species with AA genome have been an important source of cytoplasmic male sterility (CMS), the major tool for breeding commercial rice hybrids. Lin and Yuan (1980) reported the development of male sterile line having cytoplasm of wild species (O. Sativa l. F. Spontanea) and nuclear genome of rice. This cytoplasmic source has been designated as wild abortive (WA) refering to a male sterile wild rice plant having abortive pollen. About 95% of the male sterile lines used in commercial production in China and other countries have WA type of cytoplasm. A new CMS source from O. Perennis was transferred into indica rice (Dalmacio et al. 1995). Newly developed CMS line has been designated as IR66707A. It has the cytoplasm of O. Perennis and the nuclear genome of cultivated rice IR64. Genetic studies show that male sterility source of IR66707A is different from that of WA. Another CMS line (IR 69700A) having cytoplasm of O. Glumaepatula and nuclear genome of IR64 has been developed. Both IR66707A and IR69700A are completely stable for male sterility. Search for restorers of these CMS sources is underway.

Tagging alien genes with molecular markers

Alien genes for resistance to BPH, BB, and blast have been tagged with molecular markers. RFLP analysis of the introgression lines derived from O. sativa x O. officinalis showed introgression of the chromosome segments in 11 of the 12 chromosomes of O. officinalis (Jena et al. 1992). Most introgressed segments were detected by single RFLP markers and the flanking markers were negative for introgression. RFLP analysis was carried out on the introgression line IR65482-4-136-2-2 with resistance to three BPH biotypes derived from O. sativa x O. australiensis. Of the 14 probes, previously mapped to chromosome 12, only RG457 detected introgression from O. australiensis. Co-segregation for BPH resistance and RG457 showed that the gene for resistance to BPH is linked to molecular marker RG457, with a crossover value of 3.68 + 1.29% (Ishii et al. 1994). Such tight linkage should facilitate selection for BPH resistance during the transfer of this resistance to other elite breeding lines of rice.

Somaclonal Variation

Somaclonal variation refers to the variation arising through tissue culture in regenerated plants and their progenies. The technique consists of growing callus or cell suspension cultures for several cycles and regenerating plants from such long term cultures. The regenerated plants and their progenies are evaluated in order to identify individuals with a new phenotype. Some useful somaclonal variants, including those for disease resistance and male sterility, have been isolated. Heszky and Simon-Kiss (1992) tested several somaclonal variants of anther culture origin. Of these, one was released as a variety named Dama. This variety is resistant to Pyricularia and has good cooking quality.

Genetic Engineering

Introduction of alien genes into crop species allows plant breeders to accomplish breeding objectives considered impossible until a decade ago. Transformation of rice is now possible through several techniques, e.g. electroporation, polyethylene glycol induced uptake of DNA into protoplasts, and microprojectile bombardment. More recently Agrobacterium mediated transformation has become feasible. Transgenic rices carrying agronomically important genes for resistance to stemborer, fungal pathogens, and herbicide tolerance have been produced in japonica as well as in indica rices (Table 2). Several laboratories have produced transgenic rices mainly through protoplast mediated DNA transformation but also via microprojectile bombardment. Hiei et al. (1994) and Cheng et al. (1998) obtained a large number of fertile transgenic plants through Agrobacterium mediated transformation. Chen et al. (1998) produced transgenic rice carrying multiple transgenes. Transgenic plants were produced after cobombarding embryogenic callus and cell suspensions with a mixture of 14 different pUC-based plasmids. Eighty-five percent of the Ro plants contained more than 2 and 17% more than nine of the target genes. Plants containing multiple transgenes had normal morphology and 63% set viable seeds.

Transgenic rice for modifying yield potential

ADP-glucose pyrophosphorylase (ADPGPP) is critical enzyme in regulating starch biosynthesis in plant tissues. Starch levels and dry matter accumulation were enhanced in potato tubers of plants transformed with glgC¹⁶ gene from E. coli encoding ADPGPP (Stark et al. 1992). We have introduced glgC¹⁶ gene into rice. Yield potential of these lines will be evaluated shortly. Also, the transfer of C₄ traits into C₃ rice is being explored to improve photosynthetic efficiency. Ku et al. (1999) used Agrobacterium-mediated transformation and introduced into rice a gene for phosphoenolpyruvate carboxylase (PEPC) from maize, which catalyzes the initial fixation of atmospheric CO₂ in C₄ plants. Most transgenic rice plants showed high-level expression of the maize gene; the activities of PEPC in leaves of some transgenic plants were two-to threefold higher than those in maize, and the enzyme accounted for up to 12% of the total leaf soluble protein. Level of expression of the maize PEPC in transgenic rice plants correlated with the amount of transcript and the copy number of the inserted maize gene. The transgenic rice plants exhibited reduced O₂ inhibition of photosynthesis and photosynthetic rates comparable to those of untransformed plants. These findings demonstrate a successful strategy for introducing the key biochemical component of the C₄ pathway of photosynthesis into rice.

Transgenic rice for insect resistance

As early as 1987, genes coding for toxins from Bacillus thuringiensis (Bt) were transferred to tomato, tobacco and potato to provide protection against lepidopteran insects. A major target for Bt deployment in transgenic rice is the yellow stemborer, Scirpophaga incertulas (Walker). The pest is widespread in Asia. Fujimoto et al. (1993) introduced a truncated _-endotoxin gene, cry1A(b) into rice. Transgenic plants in the R₂ generation expressing the cry1A(b) protein had increased resistance to striped stemborer and leaffolder. Wunn et al. (1996) introduced cry1A(b) gene into IR58 through particle bombardment. The transgenic plants in R₀, R₁ and R₂ generation showed significant insecticidal effect on several lepidopteran insect pests. Feeding studies showed up to 100% mortality for yellow stemborer and striped stemborer larvae. Nayak et al. (1997) transformed IR64 through particle bombardment using cry1A(c) gene. Six independent transgenic lines showing high expression of insecticidal crystal protein were identified. The transferred synthetic cry1A(c) gene was stably expressed in the T2 of these lines and the transgenic rice plants were highly toxic to yellow stemborer larvae. Cheng et al. (1998) produced over 2600 transgenic rice plants in nine strains through Agrobacterium-mediated transformation. The plants were transformed with two synthetic cry1A(b) and cry1A(c) coding sequences from Bacillus thuringiensis. Bioassays in R₁ indicated that the transgenic plants were highly toxic to striped stemborer and yellow stemborer. Maqbool et al. (1998) transformed, Basmati 370 and M7 cultivars using the novel endotoxin cry2A gene.

Transgenic rice for disease resistance

Sources of resistance to some diseases have been identified within cultivated rice germplasm. However, sources of resistance to sheath blight are not available and only a few donors for resistance to tungro disease. A highly successful strategy, termed coat protein (CP) mediated protection, has been employed against certain viral diseases such as tobacco mosaic virus in tobacco and tomato. A coat protein gene for rice stripe virus was introduced into two japonica varieties by electroporation of the protoplasts (Hayakawa et al. 1992). The resultant transgenic plants expressed the CP at high levels (up to 0.5% of the total soluble proteins) and exhibited a significant level of resistance to virus infection. The resistance was inherited to the progenies.

Zhang et al. (1998) introduced Xa21 using projectile bombardment of cell suspensions of elite indica rice varieties; IR64, IR72, Minghui 63, and BG90-2. Six of the 55 R₀ lines carrying Xa21 showed high level of resistance to BB in subsequent generations. Sheath blight of rice is caused by the fungus Rhizoctonia solani, which has a wide host range. About six chitinase genes have been identified in rice and are being manipulated to increase the level of resistance to fungal diseases (Zhu and Lamb 1991). Lin et al. (1995) introduced 1.1 kb rice genomic DNA fragment containing a chitinase gene through PEG mediated transformation.

Transgenic rice for abiotic stress tolerance

Genetic engineering approaches hold great promise in developing rice cultivars with higher levels of tolerance to abiotic stresses. Sakamoto and Murata (1998) introduced codA gene for choline oxidase from Arthrobacter globiformis into rice. The codA gene was inherited to the second generation of transgenic rice and its expression was stably maintained at levels of the mRNA, the protein and enzyme activity. Levels of glycinebetaine were estimated as high as 1 and 5 _mol g^-l fresh leaves in two types of transgenic plants (chl COD and cyt COD plants) in which choline oxidase was targeted to the chloroplasts and cytosol respectively. Preliminary results showed that the transgenic plants grew better than the control plants during the recovery from treatment with 0.15 M NaCl for 7 days. Further analysis of transgenic plants showed their ability to synthesize betaine conferring enhanced tolerance to salt and cold.

Transgenic rice for improved nutritional quality

Millions of rice consumers who depend on rice for a large proportion of their calories suffer from vitamin A deficiency. Ye et al. (2000) produced transgenic rice (golden rice) with the provitamin A (_-carotene) biosynthetic pathway engineered into its endosperm. Agrobacterium mediated transformation was used to introduce three genes; phytoene synthase (psy), phytoene desaturase (crt1) and lycopene cyclase (lcy). HPLC analysis showed presence of _-carotene in transgenic seeds. The transformed rice Taipei 309 is no longer cultivated. Efforts are now underway at IRRI to transfer the genes for _-carotene into IR64 and others.

Goto et al. (1999) introduced entire coding sequence of the soybean ferritin gene into kita-ake, a rice cultivar, via Agrobacterium mediated transformation. The introduced ferritin gene was regulated by the rice seed storage protein glutelin promoter, GluB-1 and terminated by the Nos polyadenylation signal. Synthesis of soybean ferritin protein was confirmed in each of the transformed rice seeds by western blot analysis. Specific accumulation in endosperm was determined by immunological tissue printing. The iron content of T1 seeds was upto three times more than the untransformed seeds.

EVALUATION PHASE: INCREASING SELECTION EFFICIENCY

Anther Culture

As early as 1968, Niizeki and Oono reported the production of haploids from anther culture of rice. Anther culture is important technique to develop true breeding lines in the immediate generation, thereby shortening the breeding cycle of new varieties. The selection efficiency in doubled haploid (DH) lines is higher, especially when dominance variation is significant. Seeds of such populations can be distributed to workers in other laboratories, and populations can be grown repeatedly in many different environments, greatly facilitating the additional mapping of both DNA markers and genetic loci controlling traits of agronomic importance. In rice, mapping populations (DH lines) produced through anther culture of IR64 x Azucena (indica/japonica cross), and CT9993 x IR62266 are being used in molecular mapping of genes including quantitative trait loci.

A number of varieties and improved breeding lines have been developed through anther culture in China, Republic of Korea, Japan and U.S.A. Most of the anther culture derived varieties are japonicas. Indica rices are generally regarded as recalcitrant for anther culture. At IRRI, anther culture technique is being employed to obtain DH lines for different rice breeding objectives. One of the anther culture derived lines, IR51500-AC-11-1, has been named as a variety (PSBRc50) in Philippines. Many DH lines produced through anther culture are now being used as parents in breeding programs. DH lines are being produced to overcome sterility in the progenies of O. sativa x O. glaberrima crosses (Enriquez et al. 2000).

Molecular Marker Technology

A series of molecular markers, e.g. random fragment length polymorphism (RFLP), random amplified polymorphic DNA (RAPD), amplified fragment length polymorphism (AFLP) and microsatellites have become available. These markers offer new opportunities for various studies in genetic and breeding research particularly in the construction of saturated molecular maps, gene tagging, QTL mapping, marker-assisted breeding, gene pyramiding, physical mapping of genome, map-based gene cloning, alien introgression and DNA finger printing of pathogen populations (Tanskley et al. 1989, Paterson et al. 1991, Mohan et al. 1997).

Molecular Maps

The availability of comprehensive molecular genetic map in rice comprising of more than 2300 DNA markers has been a major advance in rice genetics. A molecular genetic map of rice, based on RFLPs, was developed at Cornell University, Ithaca, USA, in collaboration with IRRI (McCouch et al. 1988). This map was based on F₂ population derived from an indica x japonica cross and mapped sequences were cloned from a genomic library derived from the indica variety IR36. Later many new markers were added to the map. A comprehensive molecular genetic map consisting of 2275 DNA markers has been developed at the Rice Genome Research Program in Japan (Harushima et al. 1998). Singh et al. (1996) mapped centromeres on the molecular genetic map of rice and determined correct orientation of linkage groups.

Molecular marker aided selection (MAS)

Most genes of economic importance behave in a dominant or recessive manner and require time consuming efforts to transfer. Sometimes the screening procedures are cumbersome and expensive and require large field space. If such genes can be tagged by tight linkage with DNA or isozyme markers, time and money can be saved in transferring these genes from one varietal background to another. A molecular marker very closely linked to the target gene can act as a "tag" which can be used for indirect selection of the gene(s) in breeding program.

Gene tagging in rice

Development of varieties with durable resistance to rice blast caused by Pyricularia oryzae, and bacterial blight caused by Xanthomonas oryzae pv. Oryzae is the focus of a co-ordinated effort at IRRI and National Agricultural Research Systems (NARS) using molecular marker technology. Several genes for bacterial blight and for blast resistance have been tagged with molecular markers (Table 3).

Resistance, based on single major genes, has generally been overcome in a few years in blast prone environments. Resistance in some cultivars, however, has been more long lasting or durable. In several cases, durable resistance to blast is believed to be associated with quantitative or polygenic inheritance. Under these conditions, there is little or no gain in fitness for a pathogen variant to overcome only a fraction of the polygenes. Breeders now aim at incorporating quantitative or polygenic resistance into rice varieties. Wang et al. (1994) identified a gene Pi5 conferring complete resistance including nine regions of the genome with quantitative effects on blast resistance. These latter were considered to be putative QTL for blast resistance. Three of the QTL found to be associated with partial resistance in this study had been previously identified as being linked to genes for complete resistance. Analyses of major resistance genes has also been carried out using near isogenic lines (NILs). The NILs have been used for mapping of the genes Pi2 and Pi4.

Efforts to detect markers closely linked to bacterial blight resistance genes have similarly taken advantage of the availability of a set of ten NILs. These lines each contain a single gene for resistance and, in the case of Xa21, the gene has been introgressed from the wild species of rice, O. longistaminata (Khush et al. 1990). Segregating populations have been used to confirm co-segregation between the RFLP markers and Xa1, Xa2, Xa3, Xa4, xa5, Xa10, xa13 and Xa21.

Genes for aroma, wide compatibility, fertility restoration and thermosensitive genetic male sterility, BPH resistance, tolerance to tungro virus and submergence have been tagged with molecular markers (Table 3). PCR based markers have been developed for some of these traits to facilitate selection. Nair et al. (1995, 1996) developed PCR-based markers for gall midge resistance genes Gm2 and Gm4 which are useful in MAS for developing gall midge resistant varieties. Ichikawa et al. (1997) developed PCR based marker for Rf1, a fertility restorer gene. The emphasis now is on tagging genes of economic importance by developing PCR based markers.

QTL mapping

Genetic differences affecting such as yield, quality and tolerance to various abiotic stresses (drought, salinity, submergence, etc.) are controlled by a relatively large number of loci, each of which can make a small positive or negative contribution to the final phenotypic value of the traits. Such loci are termed "quantitative trait loci or QTL". The genes governing such traits called polygenes or minor genes also follow Mendelian inheritance but are greatly influenced by the environments. The advent of molecular markers has made it possible to map the QTL having large genotypic effect on phenotype. These molecular markers allow transformation of QTL into Mendelian or quasi-Mendelian entities that can be manipulated in classical breeding programmes. Through linkage analysis of large segregating populations polymorphic for molecular markers and QTL for yield and yield components, locations of the QTL can be determined. Several QTL for traits such as blast resistance (Wang et al. 1994), root traits (Redoña and Mackill 1996, Yadav et al. 1997, Zheng et al. 2000), submergence tolerance (Nandi et al. 1997) and yield components (Xiao et al. 1996, 1998), have been mapped with molecular markers.

Wang et al. (1994) mapped Pi5 and Pi7 for blast resistance on chromosomes 4 and 11 respectively. Nine QTL having quantitative resistance to isolate PO6-6 of blast were also identified. Li et al. (1995) identified 6 QTL contributing to resistance to Rhizoctonia solani. These QTL are located on 6 of the 12 rice chromosomes and accounted for 60% of the genotypic variation in Lemont x Tequing cross. Nandi et al. (1997) used AFLP markers and identified 4 QTL for submergence tolerance on chromosomes 6, 7, 11 and 12 of rice. In addition, a major gene Sub1 for submergence tolerance was localized on chromosome 9. Xiao et al. (1996, 1998) analyzed BC₂ test cross families from the interspecific cross (O. sativa x O. rufipogon) and found that O. rufipogon alleles at marker loci RM5 on chromosome 1 and RG256 on chromosome 2 were associated with an 18 and 7% increase in grain yield per plant. A total of 68 significant QTL were identified, and of these 35 (51%) had beneficial alleles derived from the phenotypically inferior O. rufipogon parent. Moncada et al. (2001) used advanced backcross QTL analysis on O. sativa x O. rufipogon derivatives and found that certain regions of rice genome harbor genes useful over range of environments.

Molecular markers can be used to identify QTL from wild species responsible for transgressive segregation. Comparative mapping among different cereals has also increased the efficiency of mapping orthologous QTL. Future research should focus on (1) identification of QTL which could be exploited in different environments, (2) exploitation of complementary QTL to isolate transgressive segregants particularly from interspecific crosses, (3) identification of orthologous QTL among different species, as the conservation of such QTL among species may provide new opportunities for manipulation of economic traits, (4) high resolution of QTL to determine whether QTL are single genes or clusters of tightly linked genes and whether overdominance plays a significant role in heterosis, (5) cloning of QTL based on high resolution mapping.

Pyramiding major genes

The development of saturated molecular maps and the possibility of finding tight linkage of target genes with molecular markers such as RFLPs and conversion of these markers to PCR-based markers have provided new opportunities to use MAS in rice breeding. Protocols for PCR based MAS in rice have been developed (Zheng et al. 1995). In MAS individuals carrying target genes are selected in a segregating population based on tightly linked markers rather than on their phenotypes. Thus, the populations can be screened at early seedling stage and in various environments. MAS helps overcome interference from interactions between different alleles of a locus or of different loci. MAS was applied for pyramiding four genes for BB resistance e.g. Xa4, xa5, xa13, and Xa21. Breeding lines with two or three genes were also developed. The pyramided lines showed a wider spectrum and a higher level of resistance than lines with only a single gene (Huang et al. 1997). Yoshimura et al. (1995) developed lines carrying Xa4 + xa5 and Xa4 + Xa10 using RFLP and RAPD markers linked to the BB resistant genes. These lines were evaluated for reaction to 8 strains of BB, representing eight pathotypes and three genetic lineages. Lines carrying Xa4 + xa5 were more resistant to isolates of race 4 than were either of the parental lines.

Sanchez et al. (2000) used sequence-tagged site (STS) markers to pyramid three genes for BB resistance in an elite breeding line of rice. The pyramided lines having three or four genes in combination showed an increased and wider spectrum of resistance to bacterial blight than those having a single resistance gene. Singh et al. (2001) also used MAS to pyramid genes for BB resistance into a high-yielding indica rice cultivar, PR106, that is susceptible to BB. MAS was also employed to pyramid genes for resistance to blast (Hittalmani et al. 2000) and gall midge (Katiyar and Bennett 2001).

Molecular marker analysis is used to identify pathogen populations with wide diversity, which are then employed for screening and for developing durable resistant genotypes. Gene deployment strategies are being employed based on understanding of pathogen population genetics and on the genetic basis of durable resistance (Leung et al. 1993).

Synteny relationships

The development of molecular genetic maps has helped in understanding the homoeologous relationships between the genomes of crop plants. Kurata et al. (1994a) analyzed synteny relationships between the genomes of rice and wheat and found that many wheat chromosomes contained homoeologous genes and genomic DNA fragments in the same order as in rice. Comparative genome mapping in rice, maize, wheat, barley, sorghum, foxtail millet, and sugarcane demonstrates that gene content and order are highly conserved at both the map and megabase levels between different species within the grass family, but the amount and organization of repetitive sequences have diverged considerably (Devos and Gale 1997).

Based on comparative mapping information, gene location in one species can be used to predict the presence and location of orthologous loci in other species. The synteny relationships among cereals have resulted in the discovery of common genes such as dwarf phenotype in maize (D8) and wheat (Rht1) based on genomic information derived from rice.

Map based cloning of genes

The high density molecular genetic map coupled with the development of BAC and YAC libraries has led to the isolation of rice genes such as Xa1, Xa21, and Pib. Wang et al. (1995a) used BAC library and identified clones linked to Xa21 gene for BB resistance. Song et al. (1995) isolated Xa21 gene by positional cloning. The isolated gene has been introduced into rice cultivars through transformation. Similarly, Xa1 for BB resistance has been cloned through map-based cloning strategy (Yoshimura et al. 1998). The expression of Xa1, unlike previously studied resistance genes, was induced by wound and pathogen infection.

At IRRI, BAC library has been constructed from IR64 genomic DNA (Yang et al. 1997). The library consists of 18,432 clones. BAC clones carrying xa5 for BB resistance have been identified (Yang et al. 1998). Efforts to clone xa5 are underway. Once cloned, this gene would be transferred to elite rice varieties through transformation. Wang et al. (1999) cloned blast resistance gene Pib. The expression of Pib is induced upon altered environmental conditions such as altered temperature and darkness. Sanchez et al. (1999) identified BAC contigs flanking the xa13 locus for BB resistance. Hayano-Saito (2000) identified overlapping BAC clones flanking Stvb1, the gene for stripe virus resistance.

Functional genomics

Rice has become a model system for genomics research due to the comparatively its smaller size of genome, the synteny of its genome with those of other cereals, the availability of densely populated molecular maps containing more than 2,300 DNA markers, characterized YAC and BAC libraries, large-scale analysis of expressed sequence tags (ESTs), the vast amount of genetic resources, and the comparative ease of transformation.

Mutants induced through transposable elements, retrotransposons, miniature inverted repeat transposable elements (MITEs), and T-DNA insertions have provided a wealth of genetic resources for functional genomics. Some notable examples are T-DNA-tagged insertional mutants with 30,000 lines carrying 42,000 T-DNA inserts (An et al. 2001), Tos17 retrotransposon insertional mutants with about 30,000 lines carrying more than 250,000 Tos17s (Hirochika et al. 2001), and deletion mutant populations produced by fast-neutron, gamma radiation, and chemical mutagenesis (Leung et al. 2001). These plant materials have provided the necessary link between structural or sequence data and gene function through forward and reverse genetics. Furthermore, a large number of ESTs from various cDNA libraries have been produced for microarray expression experiments. One of the major challenges is to determine the function of previously unknown rice genes revealed by sequencing projects.

New high-throughput methods are being developed for expression analysis. Biochips are being used to follow changes in gene expression in response to abiotic stresses. Using gene chips or microarrays, the representative genes of rice can be analyzed on a glass slide and used in RNA hybridization to reveal gene expression patterns and identify pathways by association. A combination of mRNA and proteomics will precisely reveal the function of rice genes involved in the phenotypic expression of different agronomic traits.

Potential impact

Recent advances in cellular and molecular biology of rice offer new opportunities to enhance the efficiency of both evolutionary and evaluationary phases of rice breeding. Biotechnology is becoming important component of rice breeding (Tables 4, 5). Anther culture has been used by breeders to shorten the breeding cycle for developing rice varieties, fix recombinants by overcoming sterility in distant crosses. Several rice varieties have been developed through anther culture. Doubled haploid populations are important in mapping of genes governing agronomic traits including QTL.

Molecular markers have led to the tagging of numerous genes for tolerance to major biotic and abiotic stresses. MAS has become a tool for moving genes from one varietal background to other and for pyramiding genes and to develop durable pest resistant cultivars. Efforts on fine mapping of QTL should provide means to pyramid QTLs for tolerance to major abiotic stresses. Map based cloning has made it possible to isolate useful genes governing important agronomic traits and incorporation of these genes into elite rice cultivars through transformation. Advances in tissue culture and molecular marker technology have resulted in broadening of the gene pool of rice and enhanced the efficiency of introgression of useful genes from wild species across crossability barriers. Advances in genetic engineering have facilitated the introduction of novel genes into rice through transformation. Transgenic rices with enhanced resistance to diseases and insects and improved nutritional quality have been produced and will impact rice production and in providing more nutritious rice. Identification of genes and their manipulation would be another major breakthrough in rice genetics and breeding. As the foregoing discussion shows many rice cultivars have been developed through application of biotechnology tools. Future food and nutritional security will depend upon availability of rice cultivars with higher yield potential, durable resistance to diseases and insects, tolerance to abiotic stresses and higher levels of micronutrients in the grain. Conventional breeding methods and biotechnology tools will help us meet these challenges.

Table 1. Introgression of genes from wild Oryza species into cultivated rice.

Table 1.� Introgression of genes from wild Oryza species into cultivated rice.

Modified from Brar and Khush (2002).

Table 2. Some examples of transgenic rice plants with agronomically important genes.

Modified from Khush and Brar (2002).

Table 3. Some examples of rice genes of agronomic importance mapped with molecular markers

Modified from Khush and Brar (1998)

Table 4. Overcoming some of the constraints of conventional rice breeding using biotechnology tools

Table 5. Some examples of the biotechnology products in rice