initial marker-assisted selection in rice breeding at cuu long delta ...

Omonrice 17: 8-21 (2010)INITIAL MARKER-ASSISTED SELECTION IN RICE BREEDING AT CUU LONGDELTA RICE RESEARCH INSTITUTENguyen Thi Lang 1 and Bui Chi Buu 2 .1 Cuu Long Delta Rice Research Institute (CLRRI)2 Institute of Agricultural Sciences for Southern Vietnam ( IAS)ABSTRACTVietnam has made a big leap in agricultural production with annual increase rate of 4.3%,out of which food production increased by 5.8% per annum. As a result, Vietnam haschanged from a country with chronic hunger to one of the world largest rice exporters, andfarmers’ living standards considerably improved. However, a large proportion of farmersliving in coastal salt affected areas of the Mekong and Red River deltas have not benefitedmuch from these developments, owing to the low productivity of these areas caused bypersisting abiotic stresses. In addition, population is rapidly increasing and agriculturallands are diminishing due to industrial expansion. Land and water resources in the Vietnamare being exhausted at increasing rate due to soil erosion, land degradation and othercauses. Since most agronomic traits composed of more than two genes and, in extremecases, ten to more genes, breeding strategies to combine them is similar to a “hit or miss”approach, making it extremely difficult to select the most preferable alleles for each of thegenes corresponding to a phenotype. To increase and stabilize food productivity andenhance livelihood and food security in salt-affected areas, the following objectives arebeing targeted: (1) to develop and distribute varieties possessing enhanced tolerance of saltand drought stresses together with accompanying management options, for dissemination,with the potential to double the yield under stress conditions in target areas; (2) to provideNARES with varieties combining tolerance drought and salinity for coastal areas. Threemain approaches are being followed to increase and sustain agriculture productivity inVietnam: (1) increase the area under cultivation by exploring less productive marginallands, (2) increasing yields through development of high yielding varieties tolerant toprevailing stresses, together with proper management practices, and (3) increasing thevalues of agricultural products through introduction of more adapted high value cropsKeywords: Amylose content; Aquaculture; Aroma; QTL mapping; Rice; Salinity.INTRODUCTIONMekong Delta is the biggest granary of Vietnam. Itproduces more than 50% of rice in the country.Intensive cultivation of rice in the delta has led toincreased threat due to continuous changingdisease races and insect biotypes. Moreover, alarge part of the delta is severely affected by theacid sulfate soil conditions. Thus, pests anddiseases as well as abiotic stresses lower riceproductivity in the Mekong Delta. Breedingdisease resistant rice genotypes using markerassisted selection was reported (Lang et al. 2004a).One of the main objectives of plant breeders is toimprove existing cultivars, which are deficient inone or more traits by crossing such cultivars withlines that possess the desired trait. A conventionalbreeding program thus involves crossing wholegenomes followed by selection of the superiorrecombinants from among the several segregationproducts. Such a procedure is laborious and timeconsuming, involving several crosses, severalgenerations and careful phenotypic selection. Inaddition, tight linkage of the desired loci withundesired loci may make it difficult to achieve thedesired objective. Recombinant DNAmethodology can help to overcome a fewlimitations, but genetic engineering approaches arealso limited by the lack of sufficient number ofcloned genes and the lack of availability ofOMONRICE 17 (2010)

Initial marker-assisted selection in rice breeding at Cuu Long delta rice research institute9standardized transformation protocols in manycrop species. Moreover, polygenic traits aredifficult to manipulate by genetic engineeringprocedures. With the advent of DNA markertechnology, several types of DNA markers arenow available to plant breeders and geneticists,helping them to overcome many of the problemsfaced during conventional breeding. This paperreports on some of the applications of DNAmarker technology in our research, for geneticanalysis and characterization of various riceaccessions, fingerprinting for purity tests, as wellas for use in marker assisted backcrossing(MASC) to breed rice varieties and hybridspossessiong particular traits such as specific grainquality and tolerance to particular biotic andabiotic stresses.Segregation analysis and germplasm/parentalsurveysThe most important step in segregation analysis isthe selection of the appropriate parental lines. Theparents should be genetically divergent enough toexhibit sufficient polymorphism. The parents arecrossed to produce the segregating population,which could be an F2, backcross, recombinantinbred lines (RILs), or double haploids (DH). Forexample polymorphism between parents isidentified using PCR-based markers. We made 10crosses (Table1) to evaluate the geneticbackgrounds and introgression of new varietiesfrom donor varieties. The graphical genotypes ofthe population were constructed using RM13simple sequence repeat (SSR marker). Five allelicconditions of the plants susceptible to the diseasewere detected: homozygotes for tolerance allele,homozygotes for susceptible allele andheterozygotes.Table 1: Allelic condition of RM13 locus in F2 populations (with bacterial leaf blight reaction)No. Designation Allele Size (bp)1 IR64 B 1502 IR64 x C53 A, B 300, 1503 IR64 x Jasmine85 B, E 100-1504 IR64 x OM2514 A B5 IR24 B 1506 IR24 x IR64 B 1507 IR24 x IR36 B, C 120-1508 IR36 C 1209 IR36 x IR24 C, B 120-15010 IR36 x Jasmine85 C 12011 IR36 x OM2514 C, D 120-14012 C53 A 30013 C53 x OM2514 A, D 300-14014 Jasmine85 E 10015 Jasmine85 x IR64 E, B 100-150Anther culture shortens the breeding cycle byrapid generation of homozygous lines in a singlegeneration. Anther culture lines from the crossesof C53/Doc Phung and C53/Pokkali, have beendeveloped and genotyped by RM223 (Table 2)OMONRICE 17 (2010)

10Nguyen Thi Lang et al.Table 2. Allelic variation of RM 223 locus in anther culture-derived populations (A2)S. No. Designation Allele Size (bp)1 IR28 A 1602 Pokkali B 1403 C53/Doc phụng-1 A 1604 C53/Doc phụng-2 A 1605 C53/Doc phụng-3 A 1606 C53/Doc phụng-17 B 1407 C53/Doc phụng-19 B 1408 C53/Pokkali-1 A 1609 C53/Pokkali-2 A 16010 C53/Pokkali-3 A 16011 C53/Pokkali-5 B 14012 C53/Pokkali-11 B 14013 C53/Pokkali-25 A 16014 C53/Pokkali-26 A 16015 C53/Pokkali-27 B 14016 C53/Pokkali-42 B 14017 C53/Pokkali-43 B 14018 C53/Pokkali-44 AB 140-160DNA markers can also be used for fingerprinting,and this approach has recently been usedextensively for determining seed purity such as theexample of the genotypes, Ngang thom Cho Daoand Nang Nhen (Table 3).Table.3. Application of SSR markers for fingerprinting of Nang Nhen variety detected a mixture of 30lines.Marker Numbers of allele Chromosome PCR productRM13 3 5 23RM223 4 2 23RM106 3 8 28Information on the genetic diversity in a cropspecies is important for selection of parental lines,for construction of populations, for variouspurposes and also for studying genetic variabilitywithin a species. for example from wild rice O.officinalis is shown in Fig. 1130bp 140bp 150bp 120bp 110bpFigure 1. PCR – based DNA characterization of some Oryza officinalis populations at locus RM270.OMONRICE 17 (2010)

Initial marker-assisted selection in rice breeding at Cuu Long delta rice research institute11Table 4. Genotyping of selected lines of O. officinalis using marker RM270.No. Designation Accession Allele Size (bp)M Marker1 O. officinalis C 130 bp2 O. officinalis C 130 bp3 O. officinalis 1934 O. officinalis 195 C 130 bp5 O. officinalis 196 B 140 bp6 O. officinalis 216 C 130 bp7 O. officinalis 236 B 140 bp8 O. officinalis 235 B 140 bp9 O. officinalis 231 B 140 bp10 O. officinalis 234 C 130 bp11 O. officinalis 230 A 150 bp12 O. officinalis 233 C 130 bp13 O. officinalis 22914 O. officinalis 232 D 120 bp15 O. officinalis 228 E 110 bp16 PTB33 A 150 bpUnderstanding germplasm relationshipsUnderstanding and management of the naturalvariation present within the domestic cultivars andwild relatives of a plant species is very importantin the establishment of an efficient program aimedat crop improvement. Exploiting natural variationis very important for several reasons: geneticuniformity in crops is undesirable because it tendsto make the crop vulnerable to epidemics andenvironmental disasters resulting in yield loss.Many wild relatives of crop plants contain geneswhich confer resistance to biotic stresses such aspests and diseases and tolerance to abiotic stressessuch as drought, cold, and salinity (Lang et al,2005a). When such traits are incorporated intoeconomically important varieties, large yieldlosses can be avoided. In addition, the breeder alsoaims at improving certain desired characters suchas grain quality and yield for specific end uses. Apre-requisite for improving the overall plantcharacteristics is an understanding of the structureof the germplasm collection, which in turn willallow a systematic sampling of the germplasm forbreeding and conservation purposes.The genetic diversity and varietal identities of 100local varieties with bacterial leaf blight resistancewas studied. This genetic divesity ranged from0.24 to 0.60 with an average of 0.40. DNAmarkers have been used to quantify the geneticdiversity and determine phenetic relationships inseveral rice species. Cluster analysis was usefulfor studying the relationships among closelyrelated accessions. The diversity of the 100traditional varieties from Mekong delta wasassessed using 34 polymorphic SSR markers andquantitative morphological characters. Moleculardiversity analysis revealed genetic diversity amongthe 100 traditional varieties, which generated fiveclusters at 0.72 similarity coefficient. Some withthe same variety names were classified intodifferent clusters. Though they belong to the samecluster based on morphological markers,molecular analysis showed that they weregenetically different. The genetic diversity valueranged from 0.116 to 0.894 with an average of0.724 (Table 5)OMONRICE 17 (2010)

12Nguyen Thi Lang et al.Table 5. Genetic divesity value of 100 local accessions using 34 polymorphic SSR markers.Cluster HI 0.1667II 0..1171III 0.6055IV 0.3633V 0.8947Average 0.7242Both the morphological and the SSR markers wereable to classify the rice varieties into these agroecologicalgroups, while (principal componentanalysis) provides a more complete representationof the relationships among major groups. In a“core” collection, the individuals, varieties, oraccessions are classified into a limited number ofentities based on their degree of similarity and alimited number of genotypes can efficientlyrepresent a much larger group. A “core” collectionthus represents most of the diversity in thegermplasm collection and allows one toextrapolate conclusions to the entire collection.DNA fingerprinting has been found to extensiveapplications in assessment of seed purity,resolution of uncertainties, parentage, legalprotection of improved varieties and geneticdiagnostics.Construction of genetic linkage mapsDNA markers in plants have been used for thedevelopment of detailed linkage maps. In order toefficiently use the innumerable polymorphisms asgenetic markers, knowledge of their individualgenomic locations is necessary and thisinformation can be obtained by constructing agenetic linkage map. Thus, a genetic linkage mapgraphically represents the arrangement of theinnumerable loci, which include morphologicaltraits, isozymes, as well as DNA markers, alongthe chromosome. The distance between these lociis expressed in centimorgan (cM) which representsthe recombination rates between the loci (1 cM=10% recombination). Examples of genetic linkagemaps constructed in rice at the CLRRI formapping specific traits are shown in Table 6.Table 6. Construction of genetic linkage some maijor genes in rice at CLRRITarget Trait Chromosomelocation ofQTLs1 QTL forsalt stressLinked markers Population Reference12 C560-C747 ,R3156- C563,C1454 , C397,R1684Tenasai 2 / CBLang et al . 2001a2 Salt Tol 8 RM223 IR28/ Doc Phụng Lang et al. 2001b3 Salt Tol 1 RM315 IR64/OMCSBuu.and.Lang.2000//OMCS2000; 2004aIR64/OM 1706//Om17064 Salt Tol 2 OSR1 IR64/OM 1706//Om 1706 Buu and Lang2004b5 P12 OM 2395/ AS 996 Lang et al , 2005deficiencytolerance6 Droughttolerance9 RM201 linkage 0,4cMOM 1490/WAB880-1-38-18-20-P1-HBLang et al., 2007OMONRICE 17 (2010)

Initial marker-assisted selection in rice breeding at Cuu Long delta rice research institute13Target Trait Chromosomelocation ofQTLs7 BrownplanthopperLang et al.,19998 Brownplanthopper ResLinked markers Population Reference12 RM457F-R ,linkage at 3.5 cM12 RM270 , linkage0.2 cM RM260, 5cMPTB33 / TN1IR64/ Hoa LaiLang et al., 2005b9 Aroma 8 RM223, RG28F-R IR64/ Hoa Lai Lang et al 2004c10 Amylose 6 RM42, RM276 IR 64/ KhaoDawMali 105 Lang et al 2004dContent11 Blast R 6 RG64:1,0cM OM 1308 / Te tep Lang et al., 20033,8 cM - 4,8 and Soc Nau / OM99712 GrainProtein7 RM234 BC 2 F 2 IR64 / Khao DawkMali 105 .Lang et al., 2005bMarker Assited Selection (MAS) for major genesMAS is based on the concept that it is possible toinfer the presence of a gene from the presence of amarker that is tightly linked to the gene. If themarker and the gene are located far apart then thepossibility, they will be transmitted together to theprogeny individuals. They will be reduced due todouble crossover recombination events. Hence, aprerequisite to using markers in such selection isthat they should be tightly linked to the gene ofinterest. For this purpose, saturation of specificregions containing the gene of interest on thegenetic linkage map is necessary.Grain qualityThis study was undertaken with the view to taggene(s) controlling grain protein content (GPC)using molecular markers in rice. Genotype IR64with low protein content (7.5%) was crossed togenotype Nang thom Cho Dao with high proteincontent (10.62%) and BC3F2 population of 149lines was derived. The parental genotypes andBC3F2 population were analyzed with RM234.One primer pair for the locus RM234 showedassociation with protein content. This was furtherconfirmed through selective genotyping. The cosegregationdata on the molecular marker andprotein content on 149 BC3F2 lines was analysedusing single marker linear regression approach.The results showed that the linked QTL accountedfor 18.1% of the variation for protein contentbetween the parents. Significant regressionsuggested linkage between RM234 and a QTL forprotein content on chromosome 7. The resultsshowed that this marker-linked QTL accounted for8.73% of the variation for protein content betweenthe parents. The marker has been located onchromosome arm 7 (Lang et al., 2005b)Amylose content is one of the importantcharacteristics of grain quality of rice verities. Apolymorphic microsatellite sequence closelylinked to the Wx gene was reported. .The genotypeHoa lai and Khao Daw Mali 105 with low amylosecontent (16.8% and 19.2%, respectively) werecrossed with IR64, with medium amylose content(24.5%), and 120 BC2F2 lines were derived.These BC populations showed normal distributionfor amylose content. One locus associated withamylose content was identified near the Wx genein both populations (P < 0.000, R2= 17.8%, for IR64/ Hoa Lai and; and P < 0.0000 , R2 = 19.7% forIR64/KhawdawMali 105). The parental genotypeswere analysis with 20 primer pairs for SSRmarkers on Chromosome 6. This QTL might beuseful in marker- assisted breeding for theimprovement of rice amylose content.We also conducted genetic analysis of two traits:amylose content (AC) and gel consistency (GC),as the most important traits for cooking and eatingOMONRICE 17 (2010)

14quality of rice grains. The material used in theanalysis included five groups: A0 (duration from90-95 days) , A1 (duration from 95-100 days), A2( duration from 105-110-days), abiotic stresstolerant, and genotypes with specific traits such asaroma, together with IR64 ,Khao Daw Mali 105 aschecks A microsatellite repeat that is part of thedetected with gene amylose with marker RG42and GC for RM276. Three new varieties with lowamylose content such as: OM4900, Hau Giang 2,OM3536 . The higher significantly betweenphenotype and genotype of two traits .The aroma or good flavor of cooked rice has beenshown to be caused mainly by formaldehydes. Thegenomic clone RG28, which is tighly linked to frggene in rice, provides opportunities to performmarker aided selection in rice breeding program.This study was conducted to identify the targetgene using SSR markers (RM223 ) and STS (RG28FL-RB ) markers linked to fgr gene in rice.RG28 can be converted by sequencing into STSand used as marker for PCR amplification ofgenomic DNA from rice varities differing in theiraroma . Genotypes of 16 local varieties and 49improved varieties were determined by performingprogeny testing for fgr. The results indicated anaccuracy of close to 100% in identifying aromaticrice plants, which were similar to that usingRG28FL-RB. Germplasm survey was conductedfor parents selection for use in breeding (Lang etal. 2006) .Phytic acid, myo-isositol 1, 2, 3, 4, 5, 6 –hexakisphosphate (IP 6) is a major storagecompound of phosphorous (P) in plants. Rice graincontains anti-nutritional factors, which reduce thebioavailability of iron and zinc. Phytate has beenknown to lower the absorption of these cations andother minerals in humans and non-ruminantanimals. Induced mutations in the most importantcereals have been utilized to reduce the phytatecontent of rice grains. All efforts to developmutants with low phytic acid content have beenrecognized in the development of four ricegenotypes from different wild types asOMCS2000 and OM1490. Their rice grains weretreated with gamma radiation (20 kr) then screenedfor high levels of free phosphate in order toidentify low phytic acid (LPA) mutants. The LPAphenotype was analyzed in mutants, wild types,Nguyen Thi Lang et al.101 improved varieties and 600 landraces. LPAmutants derived from OM1490, OMCS2000 inM3, M4 and M5 showed uniform agronomic traits(Lang et al. 2007)The advent of marker assisted selection permitsrapid identification of individuals that containgenes for low phytic acid. The presence or absenceof the associated molecular marker indicates, at anearly stage, the presence or absence of the desiredtarget gene. Segregating populations were used toconfirm co-segregation between SSR markers andthe gene for LPA. Polymorphic marker RM207detected LPA in OM1490 mutants. Lane 1 (IR64)exhibited band of 200 bp and lane 2 showed Lpa-1at 220 bp. Lane 3-7 presented high phytic acidallele from the wild type OM1490, lane 8-12indicated LPA of landraces as Nep than, Nep Hạtto, Nep Ao vang , Nep Thom and Nep Mau Luon.Our findings may help rice breeders to developlow phytic acid genotypes with improvednutritional value to overcome anemia syndromeand meet the demand of biofortificationBiotic stressBacterial Blight:Bacterial leaf blight (BB) caused by Xanthomonasoryzae pv. oryzae (Xoo) is one of the majordiseases of rice in the world. In some areas ofAsia, it can reduce crop yield by 50% (Khush etal., 1989) or even up to 80% (Singh et al.1977 ).The disease caused by bacteria, also brings aboutreduction in the rice production. Plants genesgoverning highly effective resistance to microbialattack has been identified before, and often, theseresistance genes are clustered in specific region ofthe plant genome (Richter et al., 1995). Theseresistance genes belong to multigene families, andare closely linked. Resistance to bacterial blightwas transferred from wild species O.longistaminata to the cultivated rice variety IR24generating the introgression line IR-BB21 and thelocus Xa-21 was found to confer resistance to allknown Xanthomonas oryzae pv. oryzae in Indiaand the Philippines (Khush et al., 1990). Abouthundred and sixty six local accessions, and 25parental lines of hybrid rice were screened for leafblight resistance using 10 international bacterialraces in comparison with check varieties. Fivelocal cultivars were identified as resistant of theOMONRICE 17 (2010)

Initial marker-assisted selection in rice breeding at Cuu Long delta rice research institute15bacterial races as IRBB21, 3 cultivars showedresistant reaction as IRBB5 and 58 cultivars wereresistant to the race No 4 and No 6 as IRBB13.These cultivars were subsequence genotyped usingRG556, RG136 and PTA248 for detecting xa-5,xa-13 and Xa-21 genes. PCR reactions usingRG556 and PTA248 failed to detect xa-5 and Xa-21. Marker RG136 was successful in selecting 5local rice accessions and 3 parental lines of hybridrice containing xa-13 gene. A susceptible moderncultivar (IR24) and the resistant local genotype(Nang Som) were used for developing a backcrosspopulation to transfer recessive xa-13 gene.Bacterial blight resistance gene, xa-13, was notdetected in the first BC generation. Screening of130 plants with IR24 phenotype in BC1F2population identified 20 plants that were resistantto at least 6 races including race 4 and race 6(typical resistance phenotype of xa-13 gene).These plants were genotyped with 5 microsatellitemarkers (RM21, RM114, RM122, RM164, andRM190), of which 2 markers RM21 and RM190,showed polymorphisim with the accuracy of 55%and 50%, respectively. In comparison with BBpathogen reaction, these plants do not have xa-13gene but their resistance was caused by multiplegenes at other loci.Marker-assisted backcrossing (MABC) has beensuccessfully used by Huang et al. (1997) forpyramiding four resistance genes into IR24background. DNA marker-assisted selection wasemployed to select xa-13 and xa-5 bacterial blightresistance genes. Genotypes with both genes wereselected from NIL populations involving indica xindica crosses. DNA marker-assisted selection wasemployed to select xa-13, xa-5 and Xa-4 bacterialblight resistance genes. Genotypes with both geneswere selected from four BC4F4 populations fromIR24/ Base (local Vietnamese variety). With theassistance of PCR-based markers, 60 true breedinglines were identified from CLRRI and 100 BC4F4populations were also subjected to marker assistedselection for xa-13 locus. Plants were analyzedwith the STS marker RG136, which showedpolymorphism between IR24 and other parentallines involved in the crosses. PCR analysis wasconducted using oligos RG136 in an BC4F4population segregating for xa-13 locus.OMCS2000 showed banding pattern identical tothat of its xa-13 parent, Based on this, these plantswere therefore, assumed to carry xa-13 gene inhomozygous state. The homozygotes andheterozygotes were scored and goodness of fit wastested. One hundred plants were raised from eachselected BC4F4 plant. Sixty (60) lines raised fromeach selected line were tested against 11 races ofbacterial blight pathogen, and all the plantsshowed resistance to BB pathogen. The resistancereaction was also confirmed using SSR and STSmarkers for xa-13, xa-5 and Xa-4 resistance genes,and superior plants carrying Xa-5 gene wereselected (OM 2517 and OM 5636). Besides, fivelines carrying Xa-4 gene (OM 2718, AS996,OM2514 , OMCS2000 and DS2002) wereselected. In total, 60 phenotypically superior plantscarrying genes for resistance were selected (Phucand Lang, 2005).Blast resistanceBlast is caused by Pyricularia grisea Cav., is oneof the major fungal diseases of rice in Vietnam.Local varieties have been considered as geneticsources of disease resistance. This study aims atusing DNA marker assisted selection to breedvarieties that have blast resistance genes such asPi-2(t), Pi-5(t) and Pi-3(t). STS marker RG64 wasused to screen 100 local varieties for the presenceof Pi-2(t) in chromosome 6, and SSR markerRM21 was used for Pi-5(t) and Pi-3(t) genes inchromosome 4. Phenotypic evaluation was used tocompare selected lines with checks to evaluate theaccuracy of genotypic selection using thesemarkers, and the data showed that MAS has anaccuracy of 100% for the STS marker RG64 and99.49% for SSR marker RM21. These markers,are therefore sufficiently accurate for use inbreeding to select breeding lines that have blastresistance genes. The local genotypes NangHuong, Lem Bui, Soi Da, Nang Tra, Nang Tra RanDoc, Soc Nau, and Te Tep have 3 blast resistancegenes Pi-2(t), Pi-5(t), and Pi-3(t), and areconsidered as valuable material for pyramidingresistance genes to develop varieties with durableresistance.Development of blast resistant rice varietiesThe resistance genes in CLRRI rice varieties wereexamined based on reaction patterns of isolates ofblast Pyricularia grisea Sacc., from CLRRI. ToOMONRICE 17 (2010)

16confirm the genes, genetic analyses were carriedout using BC1F2 progenies derived from crossesof IRRI varieties with IR24 and IR36 . This studydemonstrated the utility of differential system inelucidating the genetic constitution of CLRRI andIRRI varieties for blast resistance (Table 8). Abackcross population consisting of 118 BC2F2lines derived from IR24 / OM2514 was analysedfor blast resistance genes and genotyped with 14simple sequence repeat (SSR) markers. One SSRmarker (RM21) was significantly associated withNguyen Thi Lang et al.blast resistance in rice (P=0.01). These markersaccounted for phenotypic variation ranging from9.6% to 20.5% and contributed to 66% of the totalvariation of percentage diseased leaf area (DLA)observed under natural infection. To evaluate thegenetic backgrounds and introgression ofresistance genes from donor varieties, thegraphical genotypes of the population wereconstructed using 3 SSR markers, RM44, RM111,RM483 and marker φX174 – HaeIII (Fig 2). Thesematerials will further be used for breeding.Table 7. Reaction pattern of F2 lines from IR24/OM2514 and IR36/OM2514 to blast isolates fromCLRRI (OMP -1)Designation Scorereaction 0Scorereaction 1Scorereaction 3Scorereaction 5Scorereaction 7Scorereaction 9IR24 30OM2514 30F1 50 32 27 44 27 75BC1F2 32 45 22 44 27 75IR36 / OM2514IR36 10OM2514 30F1 50BC1F2 12 40 34 27 50 25A set of 28 differential varieties is a useful tool toidentify blast resistance genes in rice and tocharacterize the new varieties. The reactions of 28lines to the blast isolates from CLRRI wereconfirmed using markers RM44, RM111, RM483together with marker φX174 – Hae III .Introgression of a resistance gene for brown planthopper ( BPH)Several wild species of rice with high degree ofresistance to pests have been identified at IRRI.Similarly, O. rufipogon, a wild rice is tolerant toacid sulfate soils that occurs in Dong thap Muoi,Vietnam. The wild species thus offer greatpotential to transfer genes for tolerance to bioticand abiotic stresses into rice cultivars. CLRRI hasgenerated a series of hybrids, and introgressionlines from the crosses of elite breeding lines of ricewith several wild species. Genes for resistance tobrown plant hopper (BPH), bacterial leaf blightand blast have been transferred from several wildrelatives into cultivated rice. The BPH resistantline from O. sativa / O. officinalis have beenreleased as commercial variety for cultivation inMekong Delta such as AS 996. Some of the genesintrogressed from wild species have been taggedwith molecular markers. IRRI and CLRRI havestrong on-going collaboration on the evaluationand utilization of wide cross progenies. Under theRockefeller Foundation (RF) support, tagging ofBPH resistance loci was conducted withmicrosatellite markers at Texas A&M Universityfrom Oryza officinalis / IR50 (Buu et al., 2005). Itshowed that the genes for BPH resistance (biotype4) are linked with RM18 on chromosome 7 at adistance of 1.3 cM, and RM168 on chromosome 3at a distance of 1.9 cM.Backcross progenies from crosses with IR64 / O.rufipogon and IR64 / O. officinalis were evaluatedfor BPH resistance. The results indicated that O.officinalis was a good donor resistant to BPH(Table 8)OMONRICE 17 (2010)

Initial marker-assisted selection in rice breeding at Cuu Long delta rice research institute17Table 8. Reaction patterns of BC2F1 lines to BPH from VietnamNumbers Designation Numbers ofsuceptibleplantsBph RGSV RRSV ??Numbers Numbers of Numbers R f Numbers forof resistant suceptible (medium) Plants R resistance)plants plants1 IR64/ O.Rufipogon 56 170 26 226 0 2262 IR 64/Nang ThomCho Dao 356 12 15 353 10 3583 IR 64/ O.officinalis 30 91 0 121 0 121R: resistanceBrown plant hopper causes direct damage bysucking plant sap, and it also tranmits several viraldiseases such as rice grassy stunt (RGSV; Riveraet al 1966) and ragged stunt (RRSV; Ling et al.,1978). About 121 lines derived from the crossIR64 / O. officinalis were evaluated. It indicatedthat all lines were resistant to RGSV and RRSVwhile lines derived from IR64 / Nang Thom ChoDao and IR64 /O. rufipogon were less resistant.The survey revealed three markers RM270 linkedto target regions on chromosome 12 that containedgenes for BPH resistance. O. officinalis had largeeffects on BPH resistance. The five lociindependently acted of each other in determiningthe resistance. However, the record did not detectpolymorphism for Oryza rufipogon.Abiotic stressEnvironmental stresses such as drought, salinity,submergence, and phosphorus deficiency aremajor factors limiting plant productivity. Plantshave developed different physiological andbiochemical strategies to adapt or tolerate to thesestress conditions in response to variousenvironments. Tolerances of plants toenvironmental stresses sometimes involve theaccumulation of compatible low-molecular weightosmolytes such as sugar alcohols, special aminoacids, and glycine betaine as an adaptationmechanism, especially under water stress.Drought toleranceIn most rice growing areas, yield reduction due todrought have been observed. To overcome thisproblem, A marker-assisted backcrossing (MABC)breeding programme was initiated to improve theroot morphological traits, and thereby, droughttolerance of the Vietnam upland rice varieties. Therecurrent parent in the advanced bacrossing hadnot previously been used for quantitative traitlocus (QTL) mapping. The donor parents usedwere WAB 880-1-38-18-20-P1, IR65195-3B-2-2-2-2 and WAB881 SG9 from IRRI. These lineswere crossed with OM1490 and OM4495 (indicavarieties). A linkage map was constructed withSSR markers spanning 260.4 cM alongchromosome 9 with average interval of 16.13 cMand used for QTL mapping with MapMarker/QTL. LOD score of 3.0 was used as the thresholdto identify the putative QTLs. Both parents containfavorable QTLs affecting this trait, suggesting thelikelihood of recovering transgressive segregants.Phenotypic and genotypic evaluations using 20markers on a total of 229 lines was completed.BC2F2 lines were evaluated for root length (RL),spikelet fertility (SF), drought recovery score(DRS) and yield (Y) at CLRRI. A target segmenton chromosome 9 (RM201) signifcantly increasedroot length and DT under drought stresstreatments, confirming that this root length QTLfrom OM1490 / WAB 880-1-38-18-20-P1;OM1490/WAB881 SG9, OM4495/IR65195-3B-2-2-2-2. The data suggest that drought tolerance foryield components is largely associated withgenetic and physiological factors independentfrom those determining the traits per se. Theimplications of these results for developing anOMONRICE 17 (2010)

18efficient strategy of marker-assisted selection fordrought tolerances are discussed.Phosphorus deficiency toleranceA molecular map was constructed according topublished microsatellites data from CornellUniversity and from Japan. About 116microsetellite markers were assigned to linkagegroups using MapMarker. Although there are afew gaps of more than 50 cM, the linkage map hada total map length of 2,905.50 cM. The averageinterval size was 23.05 cM, the smallest size wasin chromosome 12 and chromosome 9 (12.50cM)and the largest size was in chromosome 3. Thereare a few gaps larger than 50 cM. It indicated thatthe genetically related parents cause the low turnof polymorphism for microsatellite markers.A mapping population of 225 RIL lines(recombinant inbred lines) was derived from across between OM2395 / AS996 by single seeddescent method. This population was used todetect quantitative trait loci (QTL) associated withP-deficiency tolerance. From the random sampleand SMA (single maker analysis), F-tests weresignificant, indicating markers associated with P-deficiency tolerance. The results showed thatindividual putative QTLs explained an average ofphenotypic variation ranging from 11.01% to11.67%. RM247 and RM235 showed the highestF-value ( P3.0. Three putative QTLs were detectedwith percentage of variance explaining between11.2% - 9.13% of the variation in root length androot dry weight, respectively.Salinity toleranceSalinity is one of the major problems inagriculture, limiting crop growth and production inNguyen Thi Lang et al.many parts of the world. We investigated thegenetic basis of salinity tolerance and evaluatedimproved varieties using microsatellite markers.Phenotypes were evaluated by visual scores of saltinjury at vegetative and reproductive stages undersalt stress of EC = 12 dS/m in phytotron. Toexamine the power of the identified SSR marker inpredicting the phenotype of the salt locus, wedetermined the genotypes of 93 improved varietiesat RM223 locus. The results indicated an accuracyof more than 95 % in identifying the tolerantplants, which was similar to that using RM223.The results of the germplasm survey usinggenotypes such as OM5900, OM5930, OM6036,OM6037, OM6043, OM6041, OM6042, AS996,OM3729, OM4675 will be useful for the selectionof parents in breeding programs aimed attransferring these genes from one background toanother and use in marker assisted selection. Thesenewly developed varieties gave significantlyhigher yields over three years of testing and resultsindicated its superiority over the control varietiesPokkali and Doc Do. Results of evaluation of thesegenotypes at 12 locations showed higher yieldsunder field conditions, confirming their superiorityin saline areas.Method which has been used to identify markerstightly linked to a gene of interest makes use ofmore population (F2, BC,….) which differ in thepresence or absence of the target gene and a smallregion flanking the target gene. If the source of thegene and the recurrent parent are sufficientlydivergent, markers will reveal polymorphismsbetween the population and the recurrent parents.In this strategy, markers need not be mapped bythe usual segregation analysis in order to belocalized near the target gene. A breakthrough hasbeen achieved in breeding, some promisingcombination with high yield from 5-7 ton/ha(Table 9)OMONRICE 17 (2010)

Initial marker-assisted selection in rice breeding at Cuu Long delta rice research institute19Table 9. Some examples of lines successfully developed using MAS.No Designation Parentage Marker Trait Released Year1 OM4495 IR64/OM1706//IR64 RM42 Good grain quality 20052 OM4498 IR64/CS2000//IR64 RM223 Good grain quality, salt and 2007BPH tolerant and high yield3 OM4900 C53/Jasmine 85// RG28, Good grain quality, aroma promisingJasmine 85 RM264,RM42and high yield4 OM5239 IR64/OM2395 RM42 high yield 20075 OM5240 IR64/BusoK RM42 high yield promising6 OM6055 OMCS 2000/D43 RM223 high yield, short days promising7 OM5636 D26/IR68//IR68 RM241 P def. tolerance promising8 OM5635 D20/IR68//IR68 RM42 high yield, short days9 OM5634 D19/IR68//IR68 waxy high yield, short days10 OM5900 AS996/IR50404 RM315 Salt tolerance11 OM5936 OM1490-55/C53 RM223 high yield, short days promising12 OM6162 C50/Jasmine 85 RG28,RM42,RM270Good grain quality highyield, short dayspromising13 OM6073 C3/D3//C3 RM270 high yield, short days promisingCONCLUSIONMolecular markers reveal polymorphisms at theDNA level. With the rapid progress intechnological innovations, a lot of new molecularresources and tools such as molecular markers forprecise genetic mapping, and, high-quality genomesequence for comprehensive molecular analysis ofgenome structure and function, have alreadybecome available for rice. Diversity within thegenomic components is very important for findingnew alleles for breeding novel varieties and forunderstanding the molecular bases of the variationin different traits. One of the main objectives forplant breeders is to improve existing cultivars,which are deficient in one or more traits bycrossing such cultivars with lines that possess thedesired trait. With the advent of DNA markertechnology, several types of DNA markers andmolecular breeding strategies are now available toplant breeders and geneticists, helping them toovercome many of the problems faced duringconventional breeding. The applications of DNAmarker technology in the genetic analysis andimprovement of crop plants is now more feasible.The impact can be observed not only by the greatprogress in the field of molecular biology, but alsoin many achievements in other related fields ofscience. Currently, there is a greater possibility toboost rice production to meet the increasingdemand for food and overcoming the problem ofmalnutrition than ever before.AcknowledgmentsThe authors would like to thank the Ministry ofAgriculture and Rural Development (MARD) forfinancial support to work on rice. We alsoappreciated career development grant provided bythe Rockefeller Foundation for supporting theresearch on upland rice, the CGIAR ChalengeProgram on Water and Food (Project 7) forsupport the work on salinity tolerance. We wouldlike to thank our IRRI sponsorers as Drs NingHuang, Zhikang Li, DS Brar, GS Khush, AMIsmail, Hei Leung, CM Vera-Cruz, and CLRRIcolleagues as Drs LM Chau, PV Du for their kindsupports and useful suggestions.REFERENCESBuu BC and NT Lang. 2004. QTL for salttolerance gene. Final report of projectsponsorred by The Ministry of Agriculture andRural Development . .OMONRICE 17 (2010)

20Buu BC and NT Lang NT. 2005. Research andapply molecular marker to detect Bown planthopper resistance . Meeting Biotechnology atHanoi . Section 8.2Bonman JM and DJ Mackill D J . 1988. Durableresistance to rice blast disease . Oryza 25: 103– 110.Huang N, ER Angels, J Domingo, G Mangpantay,S Singh, G Zhang, N Kumar, BJ Vadivel, GSKhush. 1997. Pyramiding of bacterial blightresistance genes in rice: marker-assistedselection using RFLP and PCR. Theor ApplGenet 95: 313–320.Khush GS, E Bacalangco and T Ogawa. 1990. Anew gene for resistance to bacterial blightfrom Oryza longistaminata. Rice GenetNewslet 7:121 – 122.Ling KC, ER Tiongco and VM Aguiero. 1978.Rice ragged stunt, a new virus disease. PlantDis Rep 62: 70 – 75Mew TW. 1989. An overview of the worldbacterial bight situation.In: Bacterial Blight ofRice, pp. 7–12. International Rice ResearchInstitute, Los Banos, Laguna, Philippines.Mew TW and CM Vera-Cruz. 1979. Variability ofXanthomonas oryzae in infection of ricedifferential. Phytopathol 69: 152 – 155.Lang NT. 2002. Protocol for biotechnology.Agriculture publish HouseLang NT and BC Buu, 2004a. Molecular geneticanalysis and marker- assisted selection forrestore line and bacterial blight resistance inhybrid rice. SABRAO of breeding andgenetics .36(2) 83 – 93.Lang NT and BC Buu. 2004b. Genetic for salttolerance in bulk population. Agriculturepublish House 6: 824-826.Lang N T,.Li Z and Buu B C. 2001. Fine mappingfor blast resistance gene in rice using bulkedsegregant analysis. Omon rice 9: 1 – 9Lang NT, S Yanagihara, BC Buu. 2001a.Amicrosatellite marker for a gene conferring salttolerance on rice at the vegetative andNguyen Thi Lang et al.reproductive stages. SABRAO: Breeding &genetic 11 – 10Lang NT, S Yanagihara, BC Buu. 2001b. QTLanalysis of salt tolerance in rice. SABRAO:Breeding & Genetic:1:11 – 20Lang NT and BC Buu. 2003a.Genetic and physicalmaps of gene Bph-10 controling brown planthopper resistance in rice ( Oryza sativa L.).Omon rice 11: 35 – 41Lang NT , TB Thao,BC Buu. 2004. Genetic foraroma gene in rice. House agriculture publish6: 827 – 829.Lang NT and BC Buu. 2004c. Gen waxy (Wx) ingrain quality by marker molecular. Agriculturepublish House 9: 1170 – 1172.Lang NT and BC Buu . 2004d. Fine mapping withmicrosattelite marker to detected aroma inrice. Agriculture publish House 192 – 199.Lang NT, VH Dong, BC Buu. 2005a.Microsatellite marker linkage with Protein inChromosome 7 in rice (Oryza sativa L.)Biotechnology 3 (2): 231 – 236.Lang NT and BC Buu. 2005b. QTL mapping for Ptolerance in rice (Oryza sativa L.) .Biotechnology Meeting . Section 8.2Lang NT, TA Nguyet, NV Phang and BC Buu.2006. Breeding low phytic acid mutants in rice(Oryza sativa L.). Reported in Fianal Meetingof RAS, Ho chi Minh CityLang NT, and NV Phang. 2007. Low phytic acidgene evaluation on variation in rice .Agricultural Publisher, Ho chi Minh City,Vietnam Science and Technology Joural ofAgriculture & Rural Development Vol 1: 29 –40.Phuc NV and NT Lang. 2005. STS andmicrosatellite marker-assisted selection forbacterial blight resistance genes in rice, Oryzasativa L. Omone rice (?)Rivera.CT, SH, Ou and TT Lida. 1966. Grassystunt disease of rice and its transmission byNilaparavata lugens ( Stal). Plant Dis Rep 50:453 – 456.OMONRICE 17 (2010)

Initial marker-assisted selection in rice breeding at Cuu Long delta rice research institute21Richter TE, TJ Prjor, JL Bennetzien and SHHulburt. 1995. New rust resistancespecificity's associated with recombination inthe RP1. Genetics;141: 373 – 381.Singh GP, MK Srivastara, RV Singh and RMSingh. 1977. Variation in quantitative andqualitative losses caused by bacterial blight onrice varieties. Indian Phytopathol. 30: 180 –185.Bước đầu thực hiện MAS trong chọn tạo giống lúa tại CLRRIChon tạo giống lúa nhờ chỉ thị phân tử được thực hiện từ năm 1996 cho đến 2009 tại Viện Lúa ĐBSCLđã cho những kết quả bước đầu như sau:- Chọn giống lúa kháng rầy nâu từ nguồn gen của lúa hoang được áp dụng bởi marker: RM457F-R;RM270; RM260- Chọn giống lúa kháng đạo ôn từ nguồn gen của lúa bản địa được áp dụng bởi marker: RM21; RG64- Chọn giống lúa kháng bạc lá từ nguồn gen của lúa bản địa, lúa hoang được áp dụng bởi marker:RG136- Chọn giống lúa chống chịu thiếu P từ nguồn gen của lúa bản địa được áp dụng bởi marker: RM 247,RM 235.- Chọn giống lúa chống chịu thiếu mặn từ nguồn gen của lúa bản địa được áp dụng bởi marker: C560-C747, R3156- C563, C1454, C397, R1684, RM223, RM315, OSR1- Chọn giống lúa chống chịu khô hạn từ nguồn gen của lúa bản địa được áp dụng bởi marker: RM201- Chọn giống lúa có hàm lượng amylose thấp đến trung bình từ nguồn gen của lúa bản địa được ápdụng bởi marker: RM42, RM276- Chọn giống lúa có hàm lượng protein trong hạt cao từ nguồn gen của lúa bản địa được áp dụng bởimarker: RM234- Chọn giống lúa có mùi thơm từ nguồn gen lúa bản địa được áp dụng bởi marker: RM223, RG28F-ROMONRICE 17 (2010)