Marker-assisted introgression of multiple resistance genes confers broad spectrum resistance against bacterial leaf blight and blast diseases in Putra-1 rice variety

This experiment was conducted with the aim of introgressing multiple resistance genes against bacterial leaf blight (BLB) and blast diseases through marker-assisted backcross breeding. Two dominant (Xa4 and Xa21) and two recessive (xa5 and xa13) BLB resistance genes were introgressed into a Malaysian elite rice variety Putra-1 with genetic background of three blast resistance (Piz, Pi2 and Pi9) genes and high yielding. Eight polymorphic tightly linked functional and SSR markers were used for foreground selection of target genes. 79 polymorphic SSR markers were used in background selection. The plants were challenged at initial stage of breeding and challenged again at BC2F2 with the most virulent Malaysian pathotypes of Xoo (P7.7) and Magnaporthe oryzae (P7.2) to test their resistance. Results obtained from foreground marker analysis showed that the BC1F1 and BC2F1 both fitted into the Mendel’s single gene segregation ratio of 1:1 for both Xoo and blast resistance. At BC2F2, result obtained indicated that foreground marker segregation fitted into the expected Mendelian ratio of 1:2:1 for blast resistance only. Marker-assisted background selection revealed high percentage of recurrent parent genome recovery (95.9%). It was concluded that resistance to Xoo pathotype P7.7 in IRBB60 was neither due to two independent gene action nor epistasis but substantially due to single nuclear gene action. Also, the inheritance of blast resistance in the pyramided lines to pathotype P7.2 was also attributed to single gene action. The incorporation of four bacterial leaf blight and three blast resistance genes (Xa4+xa5+xa13+Xa21+Pi9+Pi2+Piz) in the newly developed lines provides for broad spectrum and durable resistance against the two major diseases studied.


Introduction
Rice production is a very important activity all around the world. Many African and Asian countries rely on rice as a source of income and daily calories [1,2,3]. Bacteria leaf blight (BLB) which is caused by the pathogen Xanthomonas oryzae pv. oryzae is one of the most serious diseases responsible for significant yield reduction in rice. The disease constraints the photosynthetic area of the crop thereby causing partial grain filling and this leads to poor yield [4]. Use of resistant varieties of rice is the best approach to guard against the disease as chemical control is not effective. To date, 42 Xoo resistance genes have been identified in rice and some of them have been incorporated into new varieties [5]. Generally, the disease is favoured by temperatures at 25 to 34°C, with relative humidity more than 70% and most rice growing regions in Malaysia are affected by the disease [6]. Host plant resistance offers the most viable, low cost effective and sustainable strategy to manage the disease than chemical control and biological method. Use of multiple R-genes is the most attractive approach to combat BLB [7].
Rice blast caused by the fungus Pyricularia oryzae [Magnaporthe oryzae (anamorph)] is one of the biotic agents responsible for severe yield loss in rice [8]. It is among the most destructive diseases affecting rice in Malaysia and other parts of the world's major rice growing regions. The disease causes considerably large grain yield losses in almost all major ecosystems where it is grown [9,10]. Blast infection could be noticed in every stage of rice growth and the symptoms occur in all above-ground parts. The early symptoms look like white/gray/green spots or necrosis surrounded by dark green colour. Molecular markers that co-segregate with the gene are a powerful method to develop a resistant cultivar. The availability of different molecular markers allows characterization of genes of interest [11].
Resistance genes for BLB and blast are usually obtained from rice varieties, land races, primitive cultivars and wild species. About 100 blast resistance genes have been identified, and many of them have been cloned and characterized [12]. Variability in pathogens leading to overcoming of resistance effect of major gene after 4-5 years, has forced the breeder for continuous search of novel R-genes. Widespread use of resistant cultivars leads to development of selection pressure in the pathogen that resulted into breaking of resistance barrier. Majority of R-genes are dominant in nature, but recessive gene has also been reported [6]. Dominant R-genes for BLB are xa-1, xa-3, xa-4, xa-7, xa-10, xa-14, xa-21 and xa- 22(t) and some major recessive genes are xa-5, xa-13. Most blast R-genes are dominant in nature.
Pyramiding of genes is an effective strategy for enhancing the durability of R-genes over time and space. Marker assisted selection is one of the most sustainable and attractive technology that facilitates incorporation of two or more major genes in different genetic background of rice. In the presence of a number of virulent pathogens, the genes may differ in their level of resistance to them. Thus, gene pyramiding is currently being pursued in an effort to develop more durable and comprehensive resistance rice varieties to combat the effect of BLB and blast pathogens. Therefore, this work was carried out to introgress resistance genes from two genetically diverse parents using marker-assisted selection for broad spectrum resistance against BLB and blast diseases.

Genetic materials and method of plant breeding
The donor parent IRBB60 which contains four BLB R-genes Xa4, xa5, xa13 and Xa21 [6] was collected from the International Rice Research Institute (IRRI), the Philippines. Xa4 and Xa21 are the dominant genes while xa5 and xa13 are in the recessive condition. The recurrent parent was Putra-1, a high yielding rice variety that is resistant to blast disease with genetic background of three blast R-genes, namely; Piz, Pi2, Pi9. Putra-1 variety was developed at the Institute of Tropical Agriculture and Food Security (ITAFoS), Universiti Putra Malaysia, from a cross between MR219 × Pongsu Seribu 1 [13]. Putra-1 was used as the female parent to hybridize the male parent IRBB60 and F 1 plants were developed. The Putra-1 was subsequently used as the recurrent parent during backcrossing stages to backcross with F 1 leading to development of BC 1 and later backcrossed to BC 1 leading to development of BC 2 plants. Marker-assisted backcross breeding method was adopted up to BC 2 F 2 generation. At each stage of backcrossing, at least 200 lines were genotyped using polymorphic tightly linked foreground markers for selection of target genes. At BC 1 F 1 and BC 2 F 1 , only plants that showed heterozygous alleles for both bacterial leaf blight and blast resistance genes were selected while plants that showed homozygous alleles for both resistances genes at BC 2 F 2 were finally selected. Foreground selection for multiple target genes was done up till BC 2 F 2 where recombined and pure homozygous lines for all seven target genes were selected.
Background selection using 79 polymorphic SSR markers was also done up till BC 2 F 2 to ensure that only lines with recovered recurrent parent genome as well as pyramided seven target genes for BLB and blast were selected. Phenotypic selection was combined with foreground and background selections at every generation to reconfirm the results of molecular screening. Management practices such as planting spacing, weed control, fertilizer application, pest and disease control were done according to the Malaysian Agricultural Research and Development Institute (MARDI) recommended manual.

Molecular marker screening
Fifteen molecular markers including both SSR and functional markers found in literature that were linked to Xoo resistance genes were screened to confirm their polymorphism between the two parents; Putra-1 and IRBB60. Also, two SSR DNA-based markers found in literature that were linked to blast resistance genes were also tested for their polymorphism between the two parents. A total of 472 rice SSR markers were screened to select the polymorphic markers between the parents Putra-1 and IRBB60 which were subsequently used for background selection.

Genomic DNA extraction
To extract genomic DNA, young and fresh rice leaves were collected from 4-week-old individual plants using the CTAB (Cetyltrimethylammonium bromide) method modified from Ashkani et al. (2013).

Confirmation of DNA bands
Three μl of the extracted DNA were mixed with 2 μl of DNA loading dye on top of a parafilm wax and then loaded into the wells of a 3g metaphor agarose gel and subjected to electrophoresis. After running the electrophoresis for about 45 minutes at 90V, 2500 aM, the gel was viewed on image viewer machine for confirmation that the DNA bands were present.

DNA identification and quantification
One μl of each DNA sample was put on NanoDrop spectrophotometry (ND-1000, NanoDrop Technologies Inc., Wilmington, DE, USA) and relative purity with concentration of the extracted DNA was estimated from the computer display data value. The final concentration of each DNA sample was diluted with TE buffer (10 mMTris-HCl, pH 8.0, 1 mM EDTA, pH 8.0) to get the required concentration and maintained in a refrigerator of -20 °C for PCR analysis.

Polymerase chain reaction (PCR)
After confirming the DNA quantity and quality, the selected suitable samples were taken for PCR analysis using 15 μl PCR mixture. Thereafter, the PCR tubes were loaded in the PCR machine (Product specification: Thermocycler, T100TM, Bio-Rad, UK) and the PCR was run following three hours plus three minutes touchdown PCR programme. After initial denaturation for 5 min at 94 °C, each cycle comprised 1 min denaturation at 94 °C, 1 min annealing at 55 °C, and 2 min extension at 72 °C with a final extension for 5 min at 72 °C at the end of 35 cycles.

Gel Electrophoresis
A 3% Metaphor TM agarose (Lonza) gel was prepared using 3 g Metaphor TM agarose powder dissolved in 100 ml 1×TBE buffer with the aid of the microwave machine. After proper dissolution, 10 ul of Gel view was added and then shaked to mix very well. The gel was run at a steady voltage of 90 set at maximum 2500A for 45 min and the resultant bands were visualized using Molecular Imager® (Product specification: GelDocTM XR, Bio-Rad Laboratories Inc., USA).

Disease challenging for bacterial leaf blight resistance
Diseased samples of rice were collected from 15 different locations used for isolation of Xoo.
Infected leaf pieces 28 × 7 mm of rice were excised with sterile scalpel. The leaf surface was sterilized with 1% Clorox (Sodium hypochloride) for three minutes and then washed with sterile distilled water (SDW) twice of one to two minutes each. The cut leaf pieces (6-7) were transferred to nutrient agar (NA) medium after drying on sterile blotting paper and incubated at room temperature (25-28 °C) for 72 hrs [14]. The single bacterial colonies were subcultured onto NA plates and to get pure culture, cultures were preserved for longer duration (at 4 °C) on NA slants. The isolates were subjected to the pathogeniecity test to confirm the pathogenicity of the isolated bacteria [15]. Pure cultures of bacterial colonies were suspended into 5 ml distilled water and the concentration of inoculums was adjusted to 10 8 cfu/ml [9] and homogenised. The plants were sprayed with water to create wet conditions favourable for disease development. Inoculation was done by cutting five leaves, approximately 5 cm from the tips of each line with sterilized scissors dipped in the prepared inoculum. After 14 days, disease data were assessed to identify the degree of pathogenicity on 0 to 9 disease rating scale using Standard Evaluation System of IRRI [16]. The disease incidence and severity were determined. At BC 2 F 2 , pure cultures of the most virulent Xoo pathotype P7.7 obtained from MARDI was used for inoculation.

Disease challenging for blast resistance
On a special request to the Malaysia Agriculture and Research Development Institute (MARDI), Seberang Perai, the most virulent pathotype P7.2 of blast pathogen Magnaporthe oryza was provided. Potato dextrose agar (PDA) was prepared in the Physiology laboratory, ITAFoS and used for sub-culturing the collected pathotype P7.2. Plating was done under a sterile condition in the laminar flow hood and the petri dish was allowed to solidify and sealed with parafilm to avoid any form of contamination. The plate cultures were streaked from the master culture using a sterilized needle and sealed with tape to avoid contamination [17]. The plates streaked with pathogen were kept under a light and dark cycle of 16 and eight hours, respectively, under room temperature for four days. Then the cultures were incubated for another 10 days at 28 o C for maximum conidial production. A sterile slide was used to harvest the conidia by scraping the fungal mycelia on the culture plates. Then, the conidia suspension was filtered using the cheese cloth to remove the mycelia and agar pieces [15,18]. In order to measure the concentration of the initial conidial suspension, a conidial suspension was delivered to both chambers of haemocytometer by capillary action [19,20].
The spore concentration was adjusted to 1.5 x 10 5 spores/ml standard by haemocytometer.
Prior to inoculation, 0.05% Tween 20 was added to the suspension to increase the spores adhesion on leaves of plants. The donor parent (IRBB60), recurrent parent (Putra-1) and the newly developed lines were inoculated with the pathogen. In order to maintain an ideal environment for the disease development, the plants were covered with plastic bags, maintaining about 90% humidity. After nine days of inoculation, the challenged plants were phenotypically screened for blast disease symptoms. The disease incidence and severity were determined [16].

Characterization for agro-morphological traits
The agro-morphological performance was evaluated using lines that already recovered their recurrent parent genome with high yielding genetic background and all seven target genes pyramided. Agronomic traits associated with yield and yield components were measured and recorded. These included number of days to 50% flowering and maturity, plant height, number panicles per plants, effective tillers per plants, panicle length, number of filled grain per plant, 1000-grain weight, total yield per plant, seed length, seed width, flag leaf length, flag leaf width, leaf area, leaf area index, and other traits of economic importance. These traits were recorded from all of the best selected lines of BC 2 F 2 along with the recurrent parent. The morphological and agronomical characteristics were measured following standard procedures.

Statistical analysis
The genotypic data obtained from the DNA scoring were subjected to Chi-square analysis using the formular: 2 E Where: O and E are the observed and expected values. The χ 2 calculated was compared with the χ 2 tabulated at P=0.05 in order to determine whether the markers showed significant differences from the Mendelian segregation ratio of 1:1 between the observed and expected values obtained from the backcross population. The DNA scoring data obtained from background selection with the aid of the 79 polymorphic markers were further subjected to analysis using the genetic software Graphical Genotyper (GGT 2.0) [21]. This analysis was done to determine the genotypes with maximum recurrent parent genome recovery. With respect to selection for maximum recovery, the sorted SSR data were analysed with the Graphical Genotyper (GGT 2.0) software.

F 1 confirmation using polymorphic tightly linked foreground markers
A cross was made between Putra-1 and IRBB60 which produced the F 1 seeds. A total of 72 F 1 plants were grown from the F 1 seeds and tested for heterozygosity using the foreground polymorphic markers. Four BLB (Xa21FR, Xa13prom, pTA248 and RM13) and two blast (RM8225 and RM6865) polymorphic foreground markers were used for the foreground selection. Out of the 72 grown F 1 plants, a total of 46 plants were confirmed heterozygous true F 1 plants and scored as 'H' using the Xa21FR polymorphic tightly-linked foreground marker. In order to determine the plants that carry all the target genes, the 46 heterozygous plants were screened with the other foreground markers and five plants were confirmed to carry both BLB (Xa4, xa5, xa13 and Xa21) and blast (Piz, Pi2 and Pi9) R-genes. These five progenies were selected for use in crossing to produce BC 1 F 1 because the Xoo and blast resistance genes have been confirmed to be pyramided into them. Figure 2 shows banding patterns obtained from genotyping of F 1 plants.

A H H A H A H H H A H H A H H H A H H H H A F1
RM13 -xa5 gene F1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21      1:1 Mendelian expected test cross ratio for a single gene model at BC 2 F 1 . This is an indication of close association with the BLB and blast genes for resistance to these particular bacterial and fungal infections. The result confirms the successful pyramiding of the target genes into the BC 2 F 1 populations using marker-assisted foreground selection. Foreground markers are closely associated or linked to the gene of interest on the same chromosome and as such are carried from one generation to another [22,23].     Table 3 shows the genetic composition of the 16 newly developed lines with their respective gene pyramids.

Marker-assisted background selection of selected pyramided lines
In order to confirm that the pyramided lines maintained high yielding characteristic from the genetic background of Putra-1, the improved selected lines were molecularly screened using    Table 2 shows the percentage RPGR and heterozygous segment of selected BC 2 F 2 lines. The recurrent parent genome recovery obtained ranged from 93.2% in BC 2 F 2 -109 to 98.5% in BC 2 F 2 -4. The mean percentage recurrent parent genome recovery recorded in all the 16 selected lines across the 12 chromosomes was 95.9. The Chromosome-wise recurrent parent genome recovery (RPGR) of the 16 improved selected BC 2 F 2 lines are shown in Figure 9.

B A A A A A A B A A A A A A H A A B B
This result is an indication that the improved lines not only have been pyramided with Xoo and blast resistance genes, thereby incorporating a broad spectrum resistance against bacterial leaf blight and blast diseases but also maintained high yielding trait.

Phenotypic screening of selected improved BC 2 F 2 lines and their parents for bacterial leaf blight resistance
No symptom (lesion) of bacterial leaf blight was observed on the leaves of IRBB60

Phenotypic screening of BC 2 F 2 lines and their parents for blast resistance
The donor parent for blast resistance genes Putra-1 which carries the Pi genes, displayed high level of resistance to rice blast with a disease score ranging from 0 to 1 while the mean score obtained was 0 indicating high resistance to blast infection.
Incidentally, the other parent IRBB60 showed a moderate resistance to blast infection with a mean disease score of 3. The BC 2 F 2 selected improved lines displayed resistance to blast disease with disease score ranging from 0 to 1 and a mean score of 1. The distribution of lesion length after challenging with Magnaporthe oryzae pathotype P7.2 in the parents and BC 2 F 2 lines are presented in  Table 6 shows there was a significant difference (P<0.05) in the observed vs expected ratio for the two independent gene model and epistasis effects both in bacterial leaf blight and blast. Therefore, the segregation ratio did not fit into the 9:3:3:1 and 15:1 ratio of two independent gene model and epistasis effects respectively. This is attributed to maker segregation distortion phenomenon. mm mm mm  , χ 2 (0.05,1)=3.84 Table 6: Chi-square test for independent genes (9:3:3:1) and epistasis effects (15:1) for BLB and blast resistance in BC 2 F 2 population inoculated with the most virulent pathotype P7.7 and P7.2 of Xoo and Magnaporthe oryzae respectively.

Agro-morphological performance of selected improved lines compared to the recurrent parent
The selected improved lines were evaluated based on twenty agro-morphological characteristics and the result was compared to the performance of their recurrent parent. The result showed that the newly improved lines differed significantly (P<0.05) from their recurrent parents in all the twenty agro-morphological traits evaluated except leaf area, leaf area index and number of unfilled grains which were statistically the same (Table 7) (Table 7).

Discussion
In this study, eight polymorphic tightly linked foreground markers were used to confirm the pyramiding of seven Xoo and blast resistance genes in the newly improved lines. The four Xoo resistance genes (Xa4 + xa5 + xa13 + Xa21) were introgressed from IRBB60 variety into the Malaysia elite variety Putra-1 with genetic background of three Pi genes, namely; Pi9, Pi2 and Piz leading to development of lines with seven gene pyramid (Xa4+xa5+xa13+Xa21+Pi9+Pi2+Piz) for broad spectrum resistance. Pyramiding may involve incorporating genes from two or more parents. For example, Pradhan et al. [4] pyramided three bacterial leaf blight (xa5 + xa13 + Xa21) resistance genes for broadspectrum resistance in deep water rice variety, Jalmagna. Three major genes (Pi1, Piz-5 and Pita) were pyramided by Hittalmani et al. [24] using RFLP markers. Hittalmani et al. [24] and Castro et al. [25] combined genes originating from three parents for rice blast and stripe rust in barley, respectively. Gene pyramiding is difficult to achieve using conventional breeding alone because of linkage with some undesirable traits that is very difficult to break even after repeated backcrossings [26]. When two or more genes are introgressed, phenotypic evaluation is unable to distinguish the effect of individual gene precisely since each gene confers resistance to and combats multiple races of the pathogen. Moreover, in the presence of a dominant and a recessive allele, the effect of the recessive gene is masked. The advent and easy availability of molecular markers closely associated with each of the resistance genes makes identification of plants with multiple genes possible. Earlier efforts to generate three-gene pyramids in the backgrounds of rice cultivars PR106 [27], Pusa Basmati [28] and Samba Mahsuri by the collaborative work of the Directorate of Rice Research, Hyderabad, India (DRR) and the Centre for Cellular and Molecular Biology, Hyderabad, India (CCMB) have been successful [29].
Studies conducted to identify the best gene combinations conferring broad-spectrum resistance showed that the four-gene combination was the most effective and did not show any sign of breakdown of resistance to various strains of the BLB pathogen from different parts of the country [30]. Shanty et al. [31] worked on the pyramiding of four BLB resistance genes Xa4, xa5, xa13 and Xa21 through marker-assisted selection (MAS) in the BLB susceptible high yielding rice cv. Mahsuri and parental lines of hybrid rice, maintainers IR58025B and Pusa 6B and the restorer lines KMR3 and PRR78 simultaneously. The results showed increased and broad spectrum resistance to the pathogen populations from different parts of India. The work was the first successful example of the use of molecular markers in foreground selection in conjunction with conventional breeding for the simultaneous introgression of genes of interest into multiple backgrounds [31]. Popular plant varieties, adopted in large-scale production, may have some undesirable traits. These cultivars are generally high-yielding but may be highly susceptible to the attack of a certain disease or insect. In some cases, these are of poor grain quality, and thus command lower prices in the market. Also, long-term cultivation of single variety with wide adoption can lead to a shift in race or biotype frequency and ultimately cause resistance breakdown [32].
One way to delay loss of resistance is called gene pyramiding or gene stacking [33,27].
Thus, to further extend the use of these popular varieties, new traits or genes are being introduced through a series of backcrossing [34]. However, pyramiding is difficult or impossible using the conventional approach due to masking and/or epistatic effects of the genes being combined [35]. Moreover, some genes have similar reactions to two or more races or biotypes; thus they prove rather difficult to identify and transfer through conventional approaches [27]. Most of the rice varieties in the Philippines have Xa4 and xa5 genes for bacterial leaf blight resistance and some have no resistance at all [36]. With the identification of new genes like Xa21, Xa22 and Xa23 or the Om gene from Oryza minuta, varieties developed earlier can be improved through the incorporation of these new important genes [37]. Several genes for BLB have been mapped and tagged using DNA markers and these can be explored in improving resistance [38]. Among the more than two dozen genes reported [39,40,41,42,43], eight genes (Xa3, Xa4, xa5, Xa7, xa8, Xa10, xa13 and Xa21) have been transferred or pyramided in common genetic background using marker-aided selection [44,33,27,45,35,46]. Lines with single or combinations of genes have been identified using different DNA markers closely linked to the genes. Classical transfer of new genes is being done through a series of backcrossing and selection, and this procedure needs a longer time and requires large resources. The use of DNA as markers for selection has streamlined and facilitated the whole process even without inoculation [37].
In this study, marker segregation distortion was observed in Xoo resistance genes at BC 2 F 2 .
The markers at those loci failed to show a goodness of fit to the expected Mendelian ratios. A study conducted by Jin et al. [47] showed that in 300 F 2 populations evaluated, the foreground marker segregation for the BADH2 gene did not correspond with the expected 1:2:1 Mendelian segregation ratio. Also, Amarawathi et al. [48] reported non-fitness to the 1:2:1 ratio of the nksbad2 marker, hence segregation distortion in their RIL individuals. Lau [49] also observed a segregation ratio of 69:46:44 which failed to fit into the 1:2:1 ratio and the result was attributed to segregation distortion of foreground markers. In the same vein, Muthusamy et al. [50] reported distortion of foreground marker segregation which was observed in all the backcross stages performed in maize and this was without regard to the size of population evaluated. Therefore, population size of BC 2 F 2 progenies could not be taken as a factor responsible for the foreground markers segregation distortion observed in this study. The BC 1 F 1 and BC 2 F 1 both fitted into the Mendel's single gene segregation ratio despite having varying population sizes. The distortion of foreground markers segregation observed at Xoo resistance locus could be attributed to variation in environmental factors.
Bing et al. [51] and Xu et al. [52] all reported that distortion of marker segregation could be as a result of environmental issues. Similarly, studies conducted by Muthusamy et al. [50] showed that the distortion in foreground marker segregation in maize backcrossings was due to backcrossings performed during winter season. Wang et al. [53] discovered distortion of marker segregation particularly in F 2 population which were grown at a particular altitude, but when grown at other altitudes, marker segregation distortion was not detected. There is, therefore, every likelihood that the markers segregation distortion observed in this study was due to environmental effects as the BC 2 F 2 progenies were planted at varied weather conditions compared to BC 1 F 1 and BC 2  performance is a good indicator of the genotype provided that genes have major effect on phenotypic expression and the phenotyping error is less [55]. Phenotypic selection for desirable agro-morphological traits has always been done along with backcross selection [23]. On the other hand, Young and Tanksley [56] proposed that the genotypic selection which was later referred to as background selection [57] involved screening the allelic pedigree throughout the genome with the aid of markers. Singh et al. [58] reported that foreground selection for target genes followed by stringent phenotyping gives superior yield.
The high recurrent parent genome recovery in this study could be as a result of the stringent phenotypic selection done at every backcross generation and selfing. This strategy reduced both time and economic resources input. Singh et al. [59] also reported that phenotypic and marker-assisted selections increase the RPGR. The result obtained in this present study was in contrast with the findings made by Sabu et al. [60] and Martinez et al. [61] who found no significant difference in grain yield between the parental cultivar and its progenies. Hossain et al. [62] noted that number of panicles is the result of productive tillers which survived to produce panicles. The significant increase in yield contributing traits such as number tillers, number of panicles, number of leaves, etc., recorded in this present study could have resulted to the significant increase in grain yield per hectare obtained. Dutta et al. [63] and Kusutani et al. [64] both reported that higher grain yield in rice genotypes is associated with number of productive tillers per hill and number of grains per panicle. Similarly, seed length and width are essential quantifiable characteristics that are closely related to the external physical quality of rice [65,66]. Juliano [67] reported that seed length and width determines the grain/seed shape of rice. There was a significant difference in the seed length and width of the newly improved selected rice lines compared to their original recurrent parent. However, both lines have slender seed shape (seed length:width ratio >3.0). Shi and Zhu [68](1996) reported that seed shape is concurrently controlled by cytoplasmic and maternal genes as well as triploid endosperm

Conclusion
The four Xoo R-gene from IRBB60 were successfully introgressed into the backcross progenies and fixed in plants that had homozygous Xa-genes similar to IRBB60 and homozygous Pi-genes similar to Putra-1 in the BC 2 F 2 generation. Based on the MABB method and combined phenotypic selection strategy adopted in this breeding programme, 16 progenies from the BC 2 F 2 generation were selected as advanced backcross rice lines with characteristic high yielding, resistance to bacterial leaf blight and blast diseases. The result obtained from this study reveals that resistance to Xoo pathotype P7.7 in IRBB60 was neither due to two independent gene action nor epistasis but most likely due to single nuclear gene action. The result also clearly showed that the inheritance of blast resistance to pathotype P7.2 was mainly due to single gene action. More so, the pyramiding of seven resistance genes  Note: PH=plant height, FLL=flag leaf length, FLW=flag leaf width, LA=leaf area, LAI=leaf area index, FLWR=flag-leaf length:width ratio, NP/H=number of panicles per hill, DF=days to 50% flowering, DM=days to maturity, NL=number of leaves, NT=number of tillers, PL=panicle length, TNG/P=total number of grains per panicle, NUFG=number of unfilled grain per panicle, 1000GW=one thousand grain weight, TGW/H=total grain weight per hill, SL=seed length, SW=seed width, SLWR=seed length:width ratio, Y/HA=yield per hectare, CV=co-efficient of variation.