Pyramiding of blast and bacterial leaf blight resistance genes into rice cultivar RD 6 using marker assisted selection

1 Department of Plant Science and Agricultural Resources, Faculty of Agriculture, Khon Kaen University Khon Kaen, 40002, Thailand. 2 Rice Gene Discovery Unit, Kasetsart University, Kanphaeng Saen, Nakhon Pathom, 73140, Thailand. 3 Center of Excellence on Agricultural Biotechnology: (AG-BIO PERDO-CHE), Bangkok 10900, Thailand and Agricultural Biotechnology Research Center for Sustainable Economy, Khon Kaen University, Khon Kaen 40002. 4 Plant Breeding Research Center for Sustainable Agriculture, Khon Kaen University, Khon Kaen, 40002, Thailand.


INTRODUCTION
Blast caused by the fungus Magnaporthe oryzae (Hebert) Barr.and bacterial leaf blight (BLB) caused by Xanthomonas oryzae pv.oryzae (Xoo) are the most serious diseases of rice, that cause severe yield losses throughout the world (Ou, 1985).These two diseases occur in more than 80 rice growing countries resulting yield losses estimated at more than 50% (Ou, 1985;Mew, 1989).Both pathogens have tremendous genetic diversity under natural environment conditions, and a high genetic diversity of the pathogens has been observed in most planting areas.M. oryzae has a wide range of alternative hosts, especially grass species that persist throughout the year (Mackill et al., 1986).
The fragrant glutinous rice cultivar, RD6 is a high quality and popular rice cultivar among rice growers in North and Northeast Thailand.However, this cultivar is suscep-tible to both blast (Wongsaprom et al., 2010) and BLB diseases (June, 1994).
The improvement of rice varieties for resistance to the diseases that are prevalent and destructive is necessary for the sustainability of rice grain yields.Past attempts to achieve varietal resistance to blast and BLB disease have been disappointing, largely due to high levels of variability in the disease populations in growing areas (Sreewongchai et al., 2010).Breeding for broad spectrum resistance is necessary to improve blast resistance in rice.Resistance genes can be specific for different causal pathogens.Pyramiding disease resistant genes into a single genetic background might be expected to give more durable disease resistance, as more resistant genes are incorporated into single genotypes (Koide et al., 2010).Marker assisted backcrossing (MAB) is one of the most anticipated and frequently cited benefits of molecular markers as indirect selection tools in breeding programs (Semagn et al., 2006).The BLB resistance genes have as many as 24 major genes of host plant resistance which have been identified and used in rice breeding programs (Rao et al., 2002).In addition, Naveed et al. (2010) detected the BLB resistance gene xa5 in Pakistani rice germplasm by using linked markers.
Marker-assisted selection (MAS) allows the identification of multiple resistance genes in plants (Akhtar et al., 2010).The introgression of two quantitative trait loci (QTLs) conferring resistance to blast disease from Jao Hom Nin (JHN) into RD6 has been successful through MAS, with two introgression lines being released for cultivation in North and Northeast Thailand in 2008 (Wongsaprom et al., 2010).The improved cultivar of RD6 has become extensively grown because of its resistance to blast.The resistance to BLB disease is quantitative when using NILs with four resistance genes (R gene) (Xa4,xa5,xa13,and Xa21) that expressed a higher level and more durable resistance after pyramiding of the R gene (Li et al., 2001).Phuc et al. (2005) reported that marker assisted selection was accurate for improving the resistance of rice varieties to BLB.The resistant genes to BLB, Xa4, and xa13, links to microsatellites markers RM144 and RM122, respectively, and xa5 links to STS marker (RG136).The objective of this study was to pyramid blast and bacterial leaf blight resistance genes into RD6 using marker assisted backcrossing, in order to achieve durable disease resistance.

Population development
A near inbred line (NIL) derived from pyramiding between (1) the cross of RD6 (susceptible) × JHN (lowland indica cultivar with broad-spectrum resistance to blast disease) and (2) RD6 × P0489 (recombinant derived from Azucena × IR64 with blast resistance gene) was used as a recurrent parent, while IR62266 was used as a BLB donor parent.JHN is a blast resistant line carrying QTLs conferring blast resistance in chromosomes 1 and 11 (Noenplab et al., 2006), while P0489 is also a blast resistant line carrying resistant QTLs in chromosomes 2 and 12 (Suwannual et al., 2009).IR 62266 carries the gene xa5, conferring bacterial leaf blight resistance in chromosome 5 (Pattawatang, 2005).
A schematic diagram of breeding program is presented in Figure 1.Population development was started with the crossing between NIL and IR62266.F 1 plants were back crossed to the recurrent parent to achieve BC 1 F 1 .MAS were used to select individual BC 1 F 1 plant with resistance alleles.Resistant plants were identified and consequently backcrossed.BC 2 F 1 plants were derived by the same method.Resistant BC 2 F 1 plants were identified and subsequently allowed to self-pollinate to produce BC 2 F 2 .MAS were performed to classify the genotype group of the BC 2 F 2 plants.The homozygous BC 2 F 2 plants were identified and grown to produce BC 2 F 2:3 seed for validation.

Marker assisted selection
To select desirable BC 1 F 1 , BC 2 F 1 and BC 2 F 2 lines, 8 SSR flanking markers (RM319/RM212, RM48/RM207, RM224/RM144 and RM313/RM277) associated with blast resistance and 2 markers associated to BLB (RM122/RM159) (Table 1), were used for marker assisted selection (MAS).The DNA of individual plants was extracted using the method (with slight modifications) of Dellaporta et al. (1983).Aliquots of the extracted DNA were run on 1% agarose gel electrophoresis to check the quality and quantity when compared to λ-DNA standard.
Polymerase chain reaction (PCR) was carried out in 10 µl reaction containing 50 ng of DNA template, 1 X PCR buffer, 2 mM MgCl 2 , 0.2 mM dNTP, 0.2 mM reverse and forward primer, and 0.5 unit of Taq DNA polymerase.The standardized amplification was initial denaturation at 94°C for 5 min, followed by 35 cycles of denaturation at 94°C for 30 s; primer annealing at 55°C for 30 s; primer extension at 72°C for 2 min and final extension at 72°C for 7 min.The PCR products were separated on 4.5% polyacrylamide denaturing gel (SequiGen, BioRad Laboratory) at 70 watt for 1 h 30 min.DNA bands were resolved using silver staining.
In addition, three markers BADH, SNP3 and GLU23 associated with quality traits for aroma, gelatinization temperature and glutinous, respectively were used to select homozygous genotypes.

Evaluation for resistance to blast and bacterial leaf blight
Twelve (12) lines of selected BC 2 F 2:3 and the parental lines were grown in plastic trays (seed seeds per hill, two replications) at the Rice Gene Discovery Unit (RGDU), Kasetsart University, Thailand, in 2011.The susceptible check cultivar KDML105, resistant check varieties IR64, P0489 and Jao Hom Nin, and the two donors, were used as the controls for blast resistance.For BLB validation, 13 of BC 2 F 2:3 (12 lines with blast and BLB resistance and one line with blast resistance but not having BLB resistance), were grown together with KDML105 and IRBB5, as susceptible and resistance checks, respectively.Urea fertilizer (46-0-0) at the rate of 312.5 kg/ha was applied three days before inoculation.

Blast inoculation
A factorial experiment in Completely Randomized Design (CRD) was laid out.Eight isolates of M. oryzae (Hebert) Barr.(THL185, THL653, THL658, THL142, THL119, THL191, THL949 and B1-2) which represented geographical locations in rice growing areas in North and Northeast Thailand, were used for blast evaluation.The  Mackill and Bonman (1986) was used for the preparation of fungus conidia.Inoculation was done with an airbrush spray, using 21 days old seedlings, followed by incubation at 24 to 28°C.Disease scoring was done 7 days after inoculation, using the standard of Roumen et al. (1997).

Data analysis
Data recorded for blast and BLB severity scored were converted to a percentage severity index (SI) using the following formula: Where, N i is the number of plant in each level; V i is the disease score with differences among individual for plant number; V is the maximum disease score and N is the total plant number.Disease reaction for blast was classified by SI using the method described by Sirithanya (1998) whereas for BLB was done by SI, based on the method of IRRI (1996).Broad spectrum resistance (BSR) for both diseases was calculated using the method reported of Ahn (1994).Analysis of variance and mean comparisons were undertaken (Gomez and Gomez, 1984).
We measure photoperiod sensitivity of promising lines by growing the line in late of February (Start point of long day season in Thailand) and then records the days to flowering.

RESULTS
Population development was started with the crossing between NIL and IR62266.F 1 plants were back crossed to the recurrent parent, resulting in 156 BC 1 F 1 .MAS were used to select individual plant with resistance alleles.Two plants were identified and consequently backcrossed.BC 2 F 1 plants were selected for the same purpose and 5 BC 2 F 1 plants were selected and subsequently allowed to self-pollinate to produce BC 2 F 2 .MAS were performed to classify the genotype group of the BC 2 F 2 plants (Figure 2).Twelve homozygous BC 2 F 2 plants were identified and grown to produce BC 3 F 2:3 seed for validation.
The markers related to aroma, high grain quality, were also used in selection program.Moreover, visualized and sensory test were used as routine for typical RD6 types.

DISCUSSION
Selection for quantitative trait loci (QTL) of disease resistant lines, in which numerous loci are accumulated, is referred to as gene pyramiding.This is difficult to achieve using conventional breeding approaches due to a low accuracy in the identification of desirable genotypes, and because of the laborious and time consuming process.This study demonstrated that MAS can deliver genotypes of interest in a limited time and a small population (Qi-ming et al., 2006).BC 2 F 2:3 pyramided lines were achieved in 2.5 years, while gene validation and the objectives of the breeding program were accomplished in three years.
The success of the breeding program was defined by BSR to blast and BLB.The recurrent parent obtained blast resistance on chromosomes 2 and 12 from P0489, the recombinant line of Azucena × IR64 (the donor, containing strongly resistant genes) (Sallaud et al., 2003) as well as the resistant genes on chromosomes 1 and 11 from RD6 × Jao Hom Nin (Wongsaprom et al., 2010).This study demonstrated the availability of recombinants for gene introgression.
For blast validation, P0489 and IR64 had the same level of resistance.However, the two varieties were not resistant to B1-2 and THL 185, respectively.Nonetheless, most of the resistant lines were moderately resis-tant to the isolate B1-2, the isolate with the most virulence.Due to only eight markers associated with two major and two minor QTLs from resistant lines were used in the program.The possibility that change the high resistant to moderate resistant performance might be due to: 1) Several minors QTL have not been selected, therefore incomplete additive gene effects from resistant line could not be selected, 2) there are interaction between genes from different parental lines (non-additive gene effects).Further pyramiding might be needed to incorporate more resistance genes, to overcome B1-2.The BSR of the two lines BC 2 F 2:3 -2-8-2-24 and BC 2 F 2:3 -2-8-2-25 was greater than that for P0489 and IR64.This may reflect the non-additive effects of other resistant loci contributed from susceptible parent.Korinsak et al. (2009) reported the blast resistance locus of KDML105 located on chromosome 8.IR62266 (the donor parent) and IRBB5 (the BLB resistant check) had the same reaction to X. oryzae with BSR of 1.00, reflecting the fact that two rice varieties possess the gene xa05 for BLB resistance (Blair and McCouch, 1997).Despite BC 2 F 2:3 introgression lines being selected on markers, crossing over due to the genetic distance of 1.8 cM between the flanking markers RM122 and RM159, might have occurred.Consequently, none of the BC 2 F 2:3 lines had the same BSR as donor.However, the line BC 2 F 2:3 -2-8-2-36 was most likely resistant to the 10 isolates of X. oryzae.Disease reaction also showed that the 758 was the most virulent isolate.We selected introgression line by only marker associated with major and some minor QTL effects.Therefore, it could not get the complete resistance.The photoperiod non-sensitivity occurred by donor transmission, which actually was an advantage in introgression line.This makes this line grown all years round.In the same year, field experiment was also conducted to assess resistance as well as grain yield.Unfortunately, the field experiment was flooded, resulting to failure for data collection.
Despite the pyramiding of lines containing four QTLs for blast and one gene for BLB, none of them were resistant to all isolates of the two disease pathogens.This indicates that further backcrossing and pyramiding are required to broaden the spectrum of resistance.Farmers in the North and Northeast regions of Thailand will benefit from growing the resistance RD6 cultivar.The promising lines derived from the blast and BLB evaluation studies carry the trait of non-sensitivity to photoperiod.This characteristic is an additional benefit in that it allows rice farmer to grow them at any time during the year.

Table 1 .
SSR markers used for MAS in F 1 , BC 1 F 1 , BC 2 F 1 and BC 2 F 2 populations.

Table 2 .
Severity index (SI) and blast disease reactions.

Table 3 .
Analysis of variance of BC 2 F 2:3 selected lines and check varieties against 8 isolates of P. oryzae and 10 isolates of X. oryzae.