Inheritance and genetic mapping of resistance to Asian soybean rust in cultivar TMG 803

This study analyzed the inheritance and identified microsatellite markers linked to the resistance gene to Phakopsora pachyrhizi in soybean cultivar TMG 803. Hybridization between the cultivars TMG 803 and BRS Valiosa RR was performed to obtain F1 progenies and the F2 population. The response of the parents ‘TMG 803’ and ‘BRS Valiosa RR’ to P. pachyrhizi was, respectively, resistant and susceptible, and among the 116 F2 plants, 93 were resistant and 23 susceptible, under natural infection and field conditions. It was found that the resistance of cultivar TMG 803 is controlled by one gene with complete dominance, mapped as resistance locus Rpp4 of linkage group G. Of the 16 tested, one microsatellite marker, sc21_3420, was completely linked to the resistance gene (distance 0.0cM) and the favorable allelic form was present in cultivar TMG 803, which may therefore be useful in assisted selection in segregating populations.


INTRODUCTION
Soybean (Glycine max (L.) Merr.) is the chief oilseed grown and consumed worldwide.Brazil, with an estimated production of 86,273,200 tons in the 2013/14 harvest, ranks as the second largest producer of the grain (CONAB 2014), but has the potential to become the largest, according to Sediyama et al. (2009).
However, a major negative influence on the national soybean production is soybean rust (SR), a disease caused by Phakopsora pachyrhizi Sydow & Sydow (Dhingra et al. 2009), first reported in Brazil in the 2000/01 growing season (Yorinori et al. 2005).According to Lima et al. (2012), the cost of the mainly fungicide-based control of this crop disease, must have exceeded US$ 16 billion by the 2010/11 growing season.Thus, Morales et al. (2012) emphasized that the best way to avoid significant production losses is disease prevention by planting resistant cultivars.This is directly related to knowledge about the pathogen and plant defense mechanisms as well as the development of resistant cultivars and effective control strategies.
With regard to SR, five loci (Rpp1, Rpp2, Rpp3, Rpp4, and Rpp5) were reported that allocate dominant alleles: Rpp1 identified in PI200492 (McLean and Byth 1980), Rpp2 of PI230970 (Bromfield and Hartwig 1980), Rpp3 in PI462312 (Bromfield and Melching 1982), Rpp4 in PI459025 (Hartwig 1986), and Rpp5 in PI 200526 (Garcia et al 2008); as well as two alleles of alternative resistance (Rpp1?And Rpp1-b) at locus Rpp1 (Chakraborty et al. 2009, Ray et al. 2009).Since then, the pyramiding of these genes has been studied with a view to soybean breeding.According to Yamanaka et al. (2013), lines containing the pyramided alleles Rpp2, Rpp4 and Rpp5 are promising for breeding for SR-resistance and may be effective against a population of several P. pachyrhizi isolates.
However, for the success of a breeding program resistant pathogen genotypes in segregating populations must be identified in the early and intermediate breeding stages.This can be done with DNA marker-assisted selection.In many crops, DNA markers have been successfully used, of which E Matsuo et al.
microsatellites and SNPs (Single Nucleotide Polymorphism) are the most commonly used for the selection and mapping of genomic regions associated with genes and QTLs, as in soybean (Fuganti et al. 2004, Morceli et al. 2008, Dias et al. 2009), rice (Pinheiro et al. 2012) and maize (Teixeira et al. 2013).Certainly, microsatellite markers have been used due to their high polymorphism, codominance, reproducibility and good distribution throughout the genome (Ferreira andGrattapaglia 1998, Song et al. 2004).
In this context, the segregation of an F 2 population originated from the cross 'TMG 803' x 'BRS Valiosa RR' was investigated to identify microsatellite markers linked to the resistance gene against Asian soybean rust in 'TMG 803'.

MATERIAL AND METHODS
The cultivars TMG 803 and BRS Valiosa RR were hybridized to obtain F 1 progenies and the F 2 population, in a greenhouse of the soybean breeding program of the Crop Science Department, Universidade Federal de Viçosa (UFV).After harvesting the seeds from artificial pollination, the F 1 generation was planted and grown in the same environment where hybridizations had been performed, as recommended for the crop (EMBRAPA 2008).
The F 2 population and parents were grown on the experimental field Diogo Alves de Mello of the UFV (lat 20º 46' 03'' S, long 42º 52' 10'' W, alt 650m asl), according to the technical recommendations for the crop (EMBRAPA 2008).In this experiment, a F 2 population of 116 plants and 38 plants of each parent were used.The plants were grown in 2-m long rows, spaced 0.80m apart, with five plants per meter.Additionally, cultivar BRS Valiosa RR was planted around the borders and additional sprinkler irrigation was provided, according to crop requirements.For an undisturbed development of the pathogen (P.pachyrhizi) on the plant leaves, no disease control was applied.
After identifying the pathogen on the leaves, the plants were discriminated for the typical SR reactions (RB: Red-Brown lesion, indicating resistance and TAN: Tan lesion, indicating susceptibility).The data of the phenotypic evaluation were tested for fitting the segregation model 3: 1, in relation to the control provided by a dominant gene, and 13: 3, in relation to the control provided by two genes with epistatic interaction, by the chi-square test, at the 5% level of significance.
Molecular analyses were performed at the Laboratory of Genomics, Department of Plant Pathology/Bioagro, UFV.The DNA was extracted from the plant leaves by the CTAB method (Doyle and Doyle 1987), with modifications.
The DNA integrity and concentration were determined on ethidium bromide-stained agarose gel (1%).Based on the estimated concentration, the samples were diluted in autoclaved Milli-Q H 2 0 to a final concentration of 5 ng µL -1 , to be used in the work solution.Part of each sample was stored undiluted at -20 °C for later use.
After identifying polymorphisms, 10 resistant and 10 susceptible plants were analyzed by the BSA (Bulked Segregant Analysis) developed by Michelmore et al. (1991).It is worth emphasizing that the plants were analyzed separately without mixing DNA for bulk formation.The SSR markers with polymorphism between the parents and the bulks (resistant and susceptible) were considered candidates for the evaluation of F 2 plants.
The PCR product was first analyzed on (1%) agarose gel stained with ethidium bromide and visualized on an ultraviolet light transilluminator.Once successful, the PCR product was diluted in autoclaved Milli-Q H 2 0 and analyzed on an automated DNA sequencer model ABI3100 or, when possible, directly on the agarose gel.The size of the amplified base pairs analyzed by an automated sequencer was estimated with software GeneMapper, version 3.5.
The segregations of the amplified microsatellite markers in all F 2 plants were tested for fitting to a codominant gene model, by the chi-square test at the 5% level of significance.For linkage analysis and map construction software GQMOL (Schuster and Cruz 2008) was used with the Kosambi mapping function.The linkage criterion was a LOD score > 3.0 and the maximum distance was 30.0 cM.
The selection efficiency (SE%) of microsatellite markers, identified in this study, was calculated as described by Silva et al. (2007).Similar studies were performed by Morceli et al. (2008).

RESULTS AND DISCUSSION
On the plants of the cultivars TMG 803 and BRS Valiosa RR, we observed RB lesions (resistant) and TAN lesions (susceptible), respectively, indicating that the quantity and quality of the P. pachyrizi inoculum in this area and the environmental conditions were favorable for the pathogen development on the plant leaves.
Among the 116 F 2 plants evaluated, 93 behaved as resistant and 23 as susceptible to SR. Segregation was tested for the agreement with the expected Mendelian ratio of 3:1 (3 resistant plants and 1 susceptible plant) and 13:3 (13 resistant and 3 susceptible plants).
Of the 16 SSR markers tested, six (Satt191, Sat_372, Satt460, Satt708, sc21_3420, and sc21_4808) indicated polymorphism between the cultivars TMG 803 and BRS Valiosa RR and after analyzing the resistant and susceptible bulks, two (Satt191 and sc21_3420) were used to analyze the adequacy of segregation, which satisfactorily fit the expected rate of codominant inheritance (1:2:1) by the chi-square test (Table 2).
Between the phenotypic evaluation of F 2 plants and the DNA amplification pattern with microsatellite marker sc21_3420 an agreement of 100% was also observed.That is, the phenotypically resistant plants (lesion RB) had a similar amplification pattern to 'TMG 803' or were heterozygous with marker sc21_3420.The susceptible plants had an amplification pattern similar to that of cultivar BRS Valiosa RR.Thus, it can be stated that the resistance of cultivar TMG 803, for lesion RB and TAN, is controlled by one gene with complete dominance.
It is noteworthy that if a greater number of F 2 plants of this cross is evaluated, the values obtained by the chi-square test and probabilities could be changed.In addition, the agreement between the phenotypic and molecular evaluation may be less than 100%.For the acceptance of the ratio of 13: 3 in this study, at least one phenotypically resistant (lesion RB) plant (1/13) had to have the amplification pattern of BRS Valiosa RR for marker sc21_3420.
The estimates of the distances between each marker and the SR-resistance gene and LOD score values are shown in Table 3.The distance between the marker sc21_3420 score was equal to 0.0cM and the LOD score was 25.4 at a ratio of 3:1 for the lesion type (RB and TAN) of the F 2 population.For 13:3, the LOD value was 22.1.The other estimates, of the distance and LOD, were also similar between the two ratios (3:1 and 13:3).
The LOD is a statistic associated with an acceptable test to evaluate the hypothesis of linkage or independent segregation between locus pairs, since it is directly related to the likelihood ratio considering the probability of two genes being linked or not.Loci are linked when the LOD score is greater than 3 (probability of 1:1000), i.e., LOD = 3 means that the likelihood considering the r value is 1000 Table 3. Estimates of the distances between each marker and the resistance gene (Gene R) and LOD scores from the analysis of 116 F 2 plants considering a Proportion 3:1 (3 resistant and 1 susceptible) and 13:3 (13 resistant and 3 susceptible)  Garcia et al. (2008), Map E: Linkage map presented by Silva et al. (2008).Distance between the identified markers (in cM) on the left side of each map.
times greater than the likelihood in the case of no linkage (Schuster and Cruz 2008).
Considering the 3:1 segregation (accepted in this study), the resistance gene of cultivar TMG 803 was mapped at a distance of 0.0cM from marker sc21_3420 and these markers (sc21_3420) were 20.4cM from marker Satt191 (Figure 1).The results indicate that the resistance gene in cultivar TMG 803 is intimately linked to the resistance locus Rpp4, in linkage group G. Marker Satt612, located at a distance of 3.61cM from Satt288 (Song et al. 2004), showed no polymorphism between the cultivars TMG 803 and BRS Valiosa RR.With this marker (Satt612), Silva et al. (2008) identified no polymorphism either between BRS 184 and PI 459025 (carrier of the dominant resistance allele -Rpp4).The microsatellite marker Satt288 was mapped at locus Rpp4 by Garcia et al. (2008), Silva et al. (2008) and Meyer et al. (2009).In this study, this marker was not used to identify polymorphism between the cultivars TMG 803 and BRS Valiosa RR because the genomic DNA could not be amplified.
The selection efficiency of the resistant plants with microsatellite markers Satt191 and sc21_3420 was 80.2% and 100.0%, respectively.
The phenotypic segregation of an F 2 population with two independent genes and with complete dominance each is 9:3:3:1 (Table 4).Considering that MFMF (plants selected correctly as resistant, based on marker analysis and phenotypic evaluation) = 9; mfmf (plants selected correctly as susceptible, based on markers and phenotypic evaluation) = 1; and TP (Total number of studied plants) = 16, the selection efficiency is 62.5%.The value of 62.5% would result from the use of microsatellite markers unlinked to the resistance gene.Thus, the SE % limits would be 62.5% for unlinked genes, or microsatellite markers unlinked to the resistance gene, and 100.0%for completely linked genes, i.e., for microsatellite markers completely linked to the resistance gene.It is worth emphasizing that the value of 62.5% is a result of the F 2 segregating population analysis with complete dominance of two independent genes.The marker combination Satt309 + Satt356 in the selection of soybean genotypes resistant to race 14 cyst nematode resulted in 100% selection efficiency (Silva et al. 2007).By combining the markers Sat_275 + Sat_280 flanking the resistance gene Rpp5, the resulting SE % was also 100% (Morceli et al. 2008).According to this author, the use of these markers for marker-assisted selection results in total selection efficiency and is valid, because it identifies homozygous genotypes and the fixation of the resistance gene can be completed in a few selection cycles.
Microsatellite marker sc21_3420 enabled the correct identification of the dominant homozygote and heterozygote genotypes, and is therefore suitable for marker-assisted selection for the development of SR resistant genotypes, from the resistance source TMG 803.Thus, the resistance to soybean rust in cultivar TMG 803 is controlled by one gene with complete dominance; the resistance gene in cultivar TMG 803 is at the Rpp4 resistance locus in linkage group G; and microsatellite sc21_3420 may be indicated for marker-assisted selection of soybean genotypes, to incorporate Phakopsora pachyrhizi resistance of TMG 803 in other genotypes.
Where MFMF = Plants selected correctly as resistant based on marker analysis and phenotypic evaluation; mfmf = Plants selected correctly as susceptible, based on markers and phenotypic evaluation; TP = Total plants analyzed.

Figure 1 .
Figure 1.Linkage map of locus Rpp4 of linkage group G, showing the position of the resistance gene of cultivar TMG 803 to Asian soybean rust.Map A: Consensus linkage map of soybean for the region between Satt191 and Satt288 in linkage group G with the distance (in cM) adapted from Song et al. (2004), Map B: Linkage map presented by Meyer et al. (2009), C Map: Linkage map obtained in this study from the analysis of 116 F 2 plants derived from a cross between TMG 803 x BRS Valiosa RR, D Map: Linkage map presented byGarcia et al. (2008), Map E: Linkage map presented bySilva et al. (2008).Distance between the identified markers (in cM) on the left side of each map.

Table 1 .
Annealing temperature of the markers used in the study

Table 4 .
Phenotypic segregation of an F 2 population with two independent genes and complete dominance