Molecular Marker-Assisted Backcrossing of Anthracnose Resistance Genes into Common Beans ( Phaseolus vulgaris L.) Varieties

Anthracnose, caused by Colletotrichum lindemuthianum, is a major disease of common bean and results in high yield loss. Due to the high degree of pathogenic variability of the fungus and the continual emergence of new races, genetic resistance in the host is not durable. Gene pyramiding using Marker Assisted Selection (MAS) is proposed as a viable approach to improve the durability of major genes conditioning resistance to anthracnose. In this study a common bean line Urugezi x AND 1062 susceptible to anthracnose but already improved for Pythium root rot resistance was improved for anthracnose resistance through a backcross breeding program. Genotypic selection was done in Rubilizi laboratory in Kigali, Rwanada whereas phenotypic selection was conducted in an anthracnose hotspot at Rwerere, a research Centre of the Rwanda Agricultural and Animal Resources Development Board (RAB). Analysis of variance for effect of bean varieties and anthracnose isolates on disease expression showed significant differences (p < 0.001) among varieties and isolates and for the interaction between isolates and varieties. Developed BC 2 F 1 plants were 41% of them resistant and 59% susceptible


Introduction
Bean anthracnose causes a significant yield reduction thereby causing food shortages and loss of income to Rwandan farmers. Anthracnose in common bean is caused by Colletotrichum lindemuthianum, which is one of the most destructive diseases of common bean. The pathogen for anthracnose is greatly variable and it has co-evolved with Middle American and Andean gene races [1] [2]. The disease is reportedly one of the most important and widely distributed throughout the world.
Anthracnose causes particularly severe economic losses in Latin America, Central and Eastern Africa. Yield losses of 95% have been recorded in Colombia and of over 92% in Malawi. Bean Anthracnose can cause losses of up to 100% particularly on susceptible varieties under favourable conditions [3] [1], and also when infected seed is used. Production is reduced because of poor seed germination, poor seedling vigour, and low yields.
In Rwanda, the bean anthracnose can lead to total yield losses when susceptible varieties are grown under favourable pathogen environmental conditions [4]. Breeding for anthracnose resistance can be a positive answer for reducing losses caused by anthracnose disease in common bean.
More than 10 different anthracnose resistance genes have been identified in a number of bean varieties [5]. Cultivar G2333, a climber released in the great lakes region as Umubano and in Uganda as NABE 10C carries three complementary genes; Co-4 2 , Co-5 and Co-7 [5] and so far is the most anthracnose resistant variety. It is currently the best choice as a donor for anthracnose resistance genes. Fortunately, enough, this cultivar is widely adapted and yields heavily [6]. Co-4 2 confers the broadest resistance to most known C. lindemuthianum isolates. Introgression of genes for resistance to anthracnose into susceptible bean varieties could be a strong response to anthracnose. Therefore, the objective of this study was to introgress genes for anthracnose resistance in a common bean line with Pythium root rot resistance and to develop a common bean line resistant against anthracnose and Pythium root rot. Cultivar (G2333), is also a climbing bean type released in the great lakes region as Umubano and in Uganda as NABE 10C, which carries three complementary genes; Co-4 2 , Co-5 and Co-7 [6] and so far it is the most anthracnose resistant variety. Urugezi × AND 1062, is a red mottled bean line already improved for Pythium root rot resistance by Nzungize et al. [7] and is susceptible to anthracnose. These genotypes were obtained from the gene bank at Rwanda Agricultural and Animal Resources Development Board (RAB). Urugezi × AND 1062 used as recurrent parent and was crossed with G2333 which is used as donor parent and source of anthracnose resistance genes Co-4 2 and Co-5 in 2016 at RAB. Then F 1 was backcrossed to Urugezi × AND 1062 to generate BC 1 F 1 which was also backcrossed to Urugezi × AND 1062 to generate BC 2 F 1 segregating population with anthracnose resistance and possessing commercial quality (red mottled market class) using Marker Assisted Selection.

Isolates Collection and Inoculum Preparation
Fifty isolates used in this study were collected in the northern and southern agricultural zones of Rwanda, from the most anthracnose hotspot areas. Potato Dextrose Agar (PDA) and Marthur media were prepared following the International Center for Tropical Agriculture (CIAT) protocol. Small portions of bean pods and leaves with anthracnose symptoms were plated on both media and incubated upside down to grow at the temperature of 24˚C, then incubated for 5 days. Then after single spore culture was done after one month, from old isolates which had shown anthracnose spores features under the microscope. Twelve anthracnose isolates were identified (named as, a, b, c, d, e, f, g, h, i, j, k, l, and m used control-which is distilled water) and inoculum per isolate was prepared at a concentration of 1.2 × 10 6 spores per millilitre. In order to confirm the most virulent among the twelve, three common bean varieties known to be high and less susceptible to anthracnose were screened by the twelve isolates. Inoculation of bean seedlings was done by applying the spore suspension with paint brush starting from the abaxial to the adaxial sides of primary leaves and the first trifoliate leaf until run off was evident. Inoculated plants were covered with a transparent plastic bag for 48 hours in order to maintain a relative humidity of approximately > 92%. Two days later, the transparent plastic bag was removed. Watering was done twice a day then [8] scale was used to evaluate common bean with anthracnose symptoms. Plants with scores of 1 to 3 were considered as resistant, whereas those with scores of 4 to 9 were rated as mildly to severely susceptible.

Evaluation of Segregating BC2F1 Plants
The experiment was carried out in Rwerere located at Latitude 1˚36'27"S and Longitude 29˚19'52"E. Rwerere is one of seven sectors of Burera District. Its average altitude reaches 2, 100 meter above sea level. District's climate is humid, with two dry seasons and two rainy seasons: a short dry season taking place in mid-December till February; a long rainy season going from March till mid-June; a long dry season from June till September; and a short rainy season going American Journal of Plant Sciences from mid-September till December. Yearly rains reach 1, 400 mm, whilst the temperature can go from 9˚C to 29˚C [9].
A split plot experimental design with three replications was used in this study.
Twelve isolates and one control (sterile distilled water free of inoculum) were used to screen three common bean susceptible varieties (RWV 3006, RWR 2355 and Gasilida) in order to identify the most virulent isolate which would then be used to screen developed sixty three BC 2 F 1 plants with resistance to anthracnose.
In a screen house, ten plants were planted per row with a spacing of 25 cm between rows and 20 cm between plants. Inoculation of bean seedlings was performed in a screen house by applying the spore suspension with paintbrush starting from the abaxial to the adaxial sides of primary leaves and the first trifoliate leaf until run off was evident. Inoculated plants were covered with a transparent plastic bag for 48 hours in order to maintain a relative humidity of approximately > 92%. Two days later, the transparent plastic bag was removed.
Watering was done once a day within a period of 7 -10 days then disease reaction was evaluated 10 days after inoculation using 1 -9 CIAT scale.

DNA Extraction and Amplification
Whatman FTA protocol [10]

Data Analysis
A split plot experimental design was used in field experiment. Molecular and phenotypic data were analyzed using a statistical package SAS 9.2 for laboratory and field experiments. Analysis of variance was done and means separation was performed using Duncan Multiple Range Test (DMRT).

Evaluation of Anthracnose Symptoms under Field Conditions
Analysis of variance revealed that the highly significant was found for the beans varieties and the anthracnose isolates at p < 0.01 (Table 1). Significant interaction was observed between bean varieties and isolates (Table 1). The mean separation was performed using Duncan Multiple Range Test (DMRT). Results showed isolate l and e to be the most virulent in inducing more anthracnose symptoms on the three tested common bean varieties, followed by isolates d, k, a, g, j, c and h, while isolates b, f, i and m being the isolates to induce very few symptoms on the three varieties as it is indicated in Table 2. The low means of isolates shown in Table 2 might be due to the very few symptoms observed on many plants and significant symptoms on some plants. Also, attributed disease score was between 1 -4 than 5 -7 according to the recorded scores. And many of the isolates were less virulent at the time of data collection; probably, at the time of inoculation, conditions for disease development were not too suitable for good disease expression as after some days after the scoring, there was an observation in disease severity increase according to the observation made.
Mean separation of bean varieties and bean isolates from ANOVA has shown RWR 2355 to have been attacked by all the twelve isolates followed by RWV 3006, while Gasilida was the less attacked variety. Also, isolate l has shown a high mean comparatively to other isolates.

Genotyping of the Segregating Population (BC1F1 and BC2F1)
A total number of 63 BC 2 F 1 plants were genetically screened for the presence or absence of anthracnose resistance genes Co-5 and Co-4 2 , using SBB14 and SAB 3 molecular markers (Figure 1 and     Genetic analysis of BC 2 F 1 progenies shown that observed frequencies for resistant and susceptible progeny from the cross (Urugezi × AND1062) × G2333 that they were significantly different from the expected frequencies of 75 resistant and 25 susceptible (Table 3). American Journal of Plant Sciences Chi-square test of BC 2 F 1 progeny for the presence/absence of Co-4 2 and Co-5 genes shown that observed proportion of resistant to susceptible progeny from the cross (Urugezi × AND1062) × G2333 was significantly different (p < 0.05) from the expected proportion of 75 resistant to 25 susceptible as shown in Table   3. Means that the observed ratio didn't fit to the goodness of fit of 3:1 which was expected to fit in. BC 2 F 1 progeny was screened genetically for the presence or the absence of the genes and phenotypically for disease reaction using 1 -9 scores. Combined observed genotypic and phenotypic data have shown that the proportion observed was highly statistically different (smaller to the expected 3:1 ratio) (p < 0.0001) ( Table 4).

Discussion
The results obtained from the screening experiment of three common bean varieties by twelve anthracnose isolates and one control revealed that; Mean separation from analysis of variance was showed RWR 2355 bean variety to be the most susceptible to the used anthracnose isolates followed by RWV 3006, while Gasilida was the most resistant variety. In addition, mean separation from analysis of variance also shown that isolate l and e to be the most virulent isolates, followed by isolates d, k, a, g, j, c and h, while isolates f and i and m (distilled water) used as a control were the least virulent. Obtained results confirmed the presence of susceptible varieties and virulent isolates in the study region.
Therefore, the combination of major genes if properly deployed after studying the population structure of the pathogen and its resistance spectrum should be of immense importance [11]. Balardin and Kelly [12] proposed pyramiding ge- used by many other researchers in common bean and found to be the most suitable one [13].
In this study, the fact that the obtained result from genotypic and phenotypic data of BC 1 F 1 and BC 2 F 1 plants was smaller to the expected ratio might be due to many of the improved plants which inherited either Co-4 2 or Co-5 anthracnose resistance genes when the expectation was to get more plants with the two targeted genes. Plants with one or another resistance gene were resistant others not resistant to the artificial inoculation. Also, in most cases, however, bean cultivars will be resistant to some races but not to others [14] leading to easy breakdown in its resistance. Although, some few plants which inherited the two genes were totally resistant. This is similar to what Muhalet et al. [15] and Peloso et al. [16] also reported that duplicate dominant genes were responsible for anthracnose resistance in some bean crosses.
Molecular markers have been developed for many diseases resistance genes and these markers have been successfully used to develop improved common bean cultivars and germplasm [17]. About 10 different anthracnose resistance genes (designated with Co symbol) have been characterized and markers linked to most of these resistance loci have been identified [18] [19] [20] [21].
In this study, two SCAR markers SAB 3 and SBB14 linked to anthracnose resistance genes Co-4 2 and Co-5 were screened in a recurrent and donor parents as well as in BC 2 F 1 population developed for anthracnose resistance. Only twenty six BC 2 F 1 plants inherited the target genes; Nine plants had the two resistance genes Co-4 2 and Co-5 and 17 plants inherited either Co-4 2 or Co-5 genes. The observed ratio of resistant to susceptible plants was significantly different from the expected ratio. This might be due to the screening of non-homozygous BC 2 F 1 plants as all the plants were heterozygous for Co-4 2 gene. Servin et al. [22] recommended crossing the root genotype with a blank parent to obtain a genotype carrying all favourable alleles in coupling phase then self-pollinate the F 1 to produce the ideotype with fixed target genes in one generation.
Based on the results obtained from the present study, the findings promise that it is possible to improve common bean for anthracnose and Pythium root rot diseases resistance in one cultivar. Presence of SCAR-markers, SAB3 and SBB14, in the developed resistant population revealed a successful resistance transfer of anthracnose resistance genes using a MAS backcrossing program.