REACTION OF SOURCES OF RESISTANCE TO WHITE MOLD TO MICROSATELLITE HAPLOTYPES OF SCLEROTINIA SCLEROTIORUM

1 Estudante de Graduação de Agronomia, Universidade Federal de Viçosa. Viçosa, MG, Brasil. E-mail: bruno_2102@hotmail.com; 2 Pósdoutorando, Fitopatologia, Universidade Federal de Viçosa. Viçosa, MG, Brasil; 3 Estudante de Graduação de Agronomia, Universidade Federal de Viçosa. Viçosa, MG, Brasil; 4 Pósdoutorando EPAMIG, UREZM. Viçosa, MG, Brasil; 5 Pesquisador da EPAMIG, UREZM. Viçosa, MG, Brasil.


Introduction
Resistance to white mold caused by Sclerotinia sclerotiorum (Lib.) de Bary is a challenge to plant breeders.This pathogen has a necrotrophic lifestyle and secretes oxalic acid and a wide array of cell-wall-degrading enzymes to colonize plants (Bolton et al., 2006).Additionally, the genetic control of the disease is complex, because resistance is conditioned by both avoidance and physiological mechanisms (Schwartz and Singh, 2013).
Complete resistance to white mold is found only in non-host crops.With regard to the common bean (Phaseolus vulgaris L.), breeding for white mold resistance has advanced recently, especially in the United States, where several studies on the genetic mechanisms of resistance and breeding strategies have been carried out (Terán and Singh, 2009;Terán and Singh, 2010;Singh et al., 2014;Viteri and Singh, 2015) and a number of partially resistant lines have been released (Griffiths, 2009;Kelly et al., 2012;Miklas et al., 2014).
Recent studies concerning white mold resistance have been conducted in Brazil to identify sources of resistance (Carvalho et al., 2013;Souza et al., 2014;Lehner et al., 2015b) or understand the genetic control of resistance (Antonio et al., 2008;Carneiro et al., 2011).In a previous study (Lehner et al., 2015b), the physiological resistance of 20 putative sources of white mold resis-tance from the Bean White Mold Nursery (University of Nebraska, USA) was assessed.The lines A195, Cornell 605 and G122 were considered as the best sources of resistance to white mold for common bean breeding programs in Brazil.However, the physiological resistance of such genotypes was assessed using only one S. sclerotiorum isolate collected in Oratórios, in the state of Minas Gerais.Thus, this study aimed to assess the physiological resistance of the lines A195, Cornell 605 and G122 to eight genetically distinct isolates of S. sclerotiorum collected from the main producing regions of common bean in Brazil.

Materials and Methods
The experiment was conducted in a greenhouse located in Viçosa, in the state of Minas Gerais, Brazil.Five common bean genotypes (A195, Cornell 605, G122, Beryl, and BRSMG Madrepérola) and eight S. sclerotiorum isolates  were used.Beryl was one of the most susceptible genotypes to white mold in our previous study (Lehner et al., 2015b) and was used in this study as a susceptible check.BRSMG Madrepérola is a Brazilian cultivar, susceptible to white mold in the field (Vieira et al., 2012) and with low physiological resistance (Lehner et al., 2015b).This cultivar was used as an intermedi-  1A).Together, these states from the southeast (Minas Gerais and São Paulo), south (Paraná and Santa Catarina), and mid-western (Goiás) regions were responsible for about 55 % of total bean production in Brazil in 2014-2015(CONAB, 2015)).
Seeds were rolled up in germination paper saturated with water and germinated in the dark at ambient temperature for 48 h.Afterwards, three seeds were selected and planted at 1-cm depth in 3-L pots containing the substrate Tropstrato HT Hortaliças (Vida Verde, Mogi-Mirim, in the state of São Paulo, Brazil).The temperature in the greenhouse ranged from 16 °C to 29 °C and relative humidity was maintained at over 80 % by a sprinkler irrigation system.Pathogen inoculation was performed using the straw test methodology (Petzoldt and Dickson, 1996) with minor modifications (Lehner et al., 2015b).Treatments were arranged in factorial combinations of five common bean genotypes (A195, Cornell 605, G122, Beryl, and BRSMG Madrepérola) and eight S. sclerotiorum isolates (Ss-2, Ss-44, .A completely random design with four replicates was used.Each replicate was a pot with three plants. The disease severity was evaluated in individual plants at 14 days after inoculation (DAI) using the 1-to-9 rating scale where 1 = no infection, 2 = infection occurred but invasion of the first internode < 2 cm, 3 = invasion progressed > 2 cm but did not reach the first node, 4 = invasion reached the first node but no further, 5 = invasion passed the first node but progressed < 2 cm on the second internode, 6 = invasion of the second internode > 2 cm but did not reach the second node, 7 = invasion reached the second node but no further, 8 = invasion passed the second node but progressed < 2 cm on the third internode and no plant death, and 9 = inva- based on Bruvo's genetic distance (Bruvo et al., 2004) of the microsatellite data according to the Ward criterion (Ward, 1963).
Data were analyzed for normality with Lilliefors's test using the R program.The disease severity estimated using the 1-to-9 rating scale did not meet the normality assumption.Thus, it was transformed using log10 (x + 1).Analysis of variance was conducted and treatment means were separated using an LSD test at p < 0.05.Both analyses were conducted using the R program.

Results and Discussion
Each S. sclerotiorum isolate was a distinct haplotype.Bruvo's genetic distance between haplotypes ranged from 0.087 to 0.615 forming a genetic tree with two clusters (Figure 1B).One cluster included the isolates Ss-44, Ss-115, and Ss-136 from the southeastern states (Minas Gerais and São Paulo).The second cluster included three isolates  from the southern states (Paraná and Santa Catarina), one from the state of Goiás (Ss-347), and one from the state of Minas Gerais (Ss-2).
The interaction between S. sclerotiorum isolates and common bean genotypes was significant as regards white mold reaction score (p < 0.001, Figure 2A) and AUDPC (p = 0.011, Figure 2B).These results suggest that resistance of common bean genotypes to white mold may vary according to the S. sclerotiorum isolate.Pascual et al., (2010) also found significant common bean genotype × S. sclerotiorum isolates interaction, but Otto-Hanson et al., (2011) did not.The different results may be due to the variability of isolates and/or cultivars, inoculation method, and environmental conditions under which the experiments were carried out.
The average white mold reaction score varied from 5.0 (Cornell 605) to 8.3 (Beryl) and the average AUDPC varied from 59 (Cornell 605) to 141 (Beryl) across the isolates (data not shown).In general, Beryl was the most susceptible genotype to white mold (Figure 2).On average, the white mold reaction scores of Cornell 605, A195 and G122 across the eight isolates were 40 %, 34 % and 22 % lower than those for the susceptible check Beryl; for AUDPC, they were 58 %, 43 % and 46 % lower, respectively.
In our previous study (Lehner et al., 2015b), Cornell 605, A195 and G122 showed high-yielding potential in the field and intermediate resistance or resistance to anthracnose, angular leaf spot, rust, and fusarium wilt.Furthermore, they exhibited a high level of physiological resistance to a specific isolate of S. sclero- This characteristic facilitates mechanical harvesting.Furthermore, it is an important avoidance mechanism, since it minimizes contact between plants within and between rows and propitiates sunlight penetration and air circulation in the plant canopy, disfavoring white mold development (Miklas et al., 2013).The results of this study indicate that the physiological resistance of Cornell 605, A195 and G122 is effective against distinct genotypes of S. sclerotiorum collected from the main producing regions of common bean in Brazil.
Although plants of Cornell 605, A195 and G122 have some desirable agronomic attributes, the grain characteristics are not attractive to the Brazilian market.Cornell 605 is a light red kidney common bean (Griffiths, 2009), A195 has large opaque beige seeds (Singh et al., 2007) and G 122 belongs to the cranberry common bean class (Viteri and Sigh, 2015).In Brazil the "carioca" grain type (medium-sized cream-colored grains with brown stripes) is the most consumed common bean class (about 70 %).Therefore, the greatest challenge from now on is to incorporate the white mold resistance found in Cornell 605, A195 or G 122 in the elite cultivars of "carioca" grain type.
G122 is a well-known source of resistance to white mold abroad (Griffiths, 2009;Miklas et al., 2014;Singh et al., 2014;Viteri and Singh, 2015) and recently has been used in breeding programs in Brazil (Carneiro et al., 2011;Carvalho et al., 2013).The physiological resistance present in G122 is controlled by one gene with a predominance of additive effects, which facilitates the use of this line in breeding programs (Carneiro et al., 2011).Common bean lines exhibiting good levels of physiological resistance to white mold and commercially favorable grain type were recently obtained from backcrossing between G122 and an elite cultivar of "carioca" grain type (Carvalho et al., 2013).The performance of these lines in the field, however, was not assessed.
A195 has been used as a donor parent in breeding programs aimed at white mold resistance mainly in the USA (Terán and Singh, 2009;Terán and Singh, 2010).The white mold resistance present in this line is monogenic dominant in the field and recessive in the greenhouse (Genchev and Kiryakov, 2002).The use of Cornell 605 for improving white mold resistance is still incipient.
In this study, only a few isolates from each region were assessed.Thus, caution must be applied, because the populations of S. sclerotiorum are variable, although at low levels under Brazilian conditions (Lehner et al., 2015a).In general, our results suggest that Cornell 605 is the most appropriate source of physiological resistance to white mold for the southern region, whereas Cornell 605 and A195 are somewhat superior to G122 for the mid-western and southern regions.

Figure 1 −
Figure 1 − (A) Dots in the Brazilian states shown in gray are the sites where eight Sclerotinia sclerotiorum isolates were collected.(B) Dendrogrambased on Bruvo's genetic distance(Bruvo et al., 2004) of the microsatellite data according to the Ward criterion(Ward, 1963).

Figure 2 −
Figure 2 − Mean white mold reaction scores evaluated 14 d after inoculation (A) and mean area under the disease progress curve (AUDPC) from 2 to 14 d after inoculation (B) of eight Sclerotinia sclerotiorum isolates onto five common bean genotypes using the straw test.These isolates were collected from five Brazilian states: MG = Minas Gerais; SP = São Paulo; PR = Paraná; SC = Santa Catarina; and GO = Goiás.A195, Cornell 605 and G122 are white mold resistance sources; Beryl is a susceptible check; and BRSMG Madrepérola is a Brazilian cultivar susceptible to white mold in field tests and with low physiological resistance.In panel A data were transformed using log10 (x + 1) before analysis, but untransformed means are presented.The bars labeled with different letters within each isolate are significantly different (p ≤ 0.05, LSD).