3.1 Molecular analysis with SNP markers
Eight of the nine SNPs that were selected out of the HTPG project data and genotyped at Intertek (88.8%) were polymorphic and were considered suitable for genotyping analysis. The snpPV0071 marker, associated with the Phg-2 locus (Nay et al. 2019a), was the only one that displayed a monomorphic profile for the susceptible associated allele (G:G) and was discarded from the analysis. This marker is specific to tag the resistance Phg-2 allele at G10474 line, not genotyped in this study.
Table 2
Gene annotation performed on the Phytozome platform for the five SNPs associated with common bean disease resistance genes selected for analysis, validation and marker-assisted selection.
SNP
|
Allele
|
Primary transcript
|
Description
|
snpPV0025
|
Phg-2
|
Phvul.008G280700.2
|
PTHR11132:SF38 - GB
|
snpPV0046
|
Co-u
|
Phvul.002G328300.1
|
Mitogen-Activated Protein Kinase 16-Related
|
snpPV0068
|
Co-42
|
Phvul.008G028000.1
|
MYB Transcription Factor
|
snpPV0070
|
Co-42
|
Phvul.008G028400.1
|
Domínio Serine Threonine Kinase/homólogo ao gene COK-4
|
snpP8282v3-817
|
Co-42
|
Phvul.008G028200
|
Domínio Serine Threonine Kinase
|
3.1.1 SNP markers associated with the Phg loci
The snpPV0051 (Phg-1) marker is located in a region that contains repeats in the genome (identified via RepeatMasker). Of the seven markers that exhibit a techncally adequate genotyping profile, snpPV0033 did not amplify the target allele in the sources carrying the Phg-2 allele.
The target allele was amplified by the snpPV0051 (A:G, where the resistance associated allele is the first in bold) marker in only 103 (27.4%) of the 376 lines evaluated. This SNP is located in a repetitive region of the common bean genome, identified using the RepeatMasker tool (Table 3), which may explain why the fragment was not amplified well. Of the 103 genotyped lines, the “A” allele associated with Phg-1 was amplified in 11 lines (Table S4), including AND 277, the source of the Phg-1 allele (Souza et al. 2016) and resistant to the 11 most important P. griseola pathotypes in Brazil, such as 63-23, 63-31, 63-47 and 63-63 (Damasceno-Silva et al. 2015; Balbi et al. 2009). Pathotype 63-63 is considered the most aggressive, causing susceptibility symptoms in all differential cultivars from the Andean and Mesoamerican gene pool (Nay et al. 2019b). The angular spot-resistant CAL 143 line was extracted from the cross between G12229 and AND 277 (Nay et al. 2019b) and contains the "A" allele associated with Phg-1 resistance in its genome (Table S4). Furthermore, the Phg-1 and Co-14 alleles were reported to be strongly linked, at a distance of 0.0 cM (Gonçalves-Vidigal et al. 2011), jointly conferring resistance to the P. griseola pathotype 63-23 and C. lindemuthianum pathotypes 65, 73 and 2047 (Gonçalves-Vidigal et al. 2011). Due to the inconsistent amplification of this marker in the evaluated lines, it was not validated in a reduced set of contrasting lines for resistance to angular leaf spot. However, the snpPV0051 marker has potential for use in the assisted selection of populations from crosses with known sources of the Phg-1 allele.
Table 3
Genotyping summary of 10 SNP markers associated with common bean disease resistance genes.
SNP
|
Allele
|
Polymorphism
|
Allele 1
|
Allele 2
|
Amplified?
|
NA's
|
%NA
|
Monomorphic?
|
RepeatMasker?
|
snpPV0051
|
Phg-1
|
A:G
|
A:A (R)
|
G:G
|
Yes
|
273/376
|
72.60%
|
|
Yes
|
snpPV0025
|
Phg-2
|
G:T
|
G:G (R)
|
T:T (S)
|
Yes
|
0/376
|
0.00%
|
|
No
|
snpPV0033
|
Phg-2
|
T:C
|
T:T (R)
|
C:C (S)
|
Yes
|
0/376
|
0.00%
|
|
No
|
snpPV0071
|
Phg-2
|
T:G
|
T:T (R)
|
G:G (S)
|
Yes
|
0/376
|
0.00%
|
Yes
|
No
|
snpPV0027
|
Phg-5
|
T:C
|
T:T (R)
|
C:C (S)
|
Yes
|
1/376
|
0.26%
|
|
No
|
snpPV0079
|
Phg-5
|
A:G
|
A:A (R)
|
G:G (S)
|
Yes
|
2/376
|
0.53%
|
|
No
|
snpPV0068
|
Co-42
|
G:C
|
G:G (R)
|
C:C (S)
|
Yes
|
0/376
|
0.00%
|
|
No
|
snpPV0070
|
Co-42
|
G:T
|
G:G (R)
|
T:T (S)
|
Yes
|
25/376
|
6.60%
|
|
No
|
snpP8282v3-817
|
Co-42
|
A:G
|
A:A (R)
|
G:G (S)
|
Yes
|
1/181
|
0.55%
|
|
No
|
snpPV0046
|
Co-u
|
A:G
|
A:A (R)
|
G:G (S)
|
Yes
|
2/376
|
0.53%
|
|
No
|
(R): Resistant; (S): Susceptible; NA: No Amplification; The alleles in bold are associated with resistance. |
The snpPV0025 marker (G:T) associated with the Phg-2 allele, located on the Pv-08 chromosome, exhibited the “G” allele in the Mexico 54 and Cornell 49-242 breeding lines (Table S5), which are known sources of Phg-2 (Souza et al. 2016; Nay et al. 2019b). This Phg-2 allele, which has a wide diversity of functional haplotypes for a resistance locus in common bean (Nay et al. 2019b), confers resistance to pathotypes 63-19 and 63-39, and was found to be the most relevant allele in Brazil (Bassi et al. 2017). It was also detected in the PT 65 line, which has shown a high level of resistance to angular leaf spot under controlled inoculation and in the field (Pereira et al. 2019; Pereira et al. 2016). The line MAIII 16.159 from the recurrent selection program for angular leaf spot resistance conducted at the Federal University of Lavras (Pereira et al. 2019) and 11 lines from Embrapa’s program also displayed the “G” allele of the snpPV0025 marker, suggesting the presence of the resistant Phg-2 allele (Table S5). In addition, studies showed that Phg-2 was responsible for the resistance found in breeding lines evaluated under field and greenhouse conditions, and effective against P. griseola isolates from Colombia, Uganda and Brazil (Sartorato et al. 2000; Nay et al. 2019b). Phg-2 different resistance haplotypes, which confers a broad spectrum of resistance to different P. griseola pathotypes from the Andean and Mesoamerican gene pool were described by Nay et al. (2019a).
Due to its importance for breeding programs, the use of Phg-2 allele sources has been frequently adopted in crossing blocks, in addition to studies that seek to develop markers strongly linked to Phg-2 (Gil et al. 2019; Miller et al. 2018; Sartorato et al. 2000). For example, the improved MAB 348, MAB 349, MAB 351, MAB 352, MAB 353, MAB 354 and MAB 484 lines, with high angular leaf spot resistance, contain the G10474 line, source of Phg-2, as one of their parents (Gil et al. 2019). The results of this study indicate that the Phg-2 allele is present in elite germplasms developed by Embrapa, the Federal University of Lavras, IAC, IAPAR, Agropecuária Terra Alta and CIAT, and have been detected in 72 lines/cultivars (Table S5). However, as pointed out by Gi et al. (2019), care should be taken when using the snpPV0025 (ALS_08_62193174) for MAS, since it has been used to tag and introgress the Phg-2 locus from Mesoamerican MAB sources into Andean breeding lines, and do not tag specific meso alleles. Nay et al. (2019a) shed greater light on this by identifying pathotype-specific haplotypes at the Phg-2 and offering new molecular markers to be tested and used in MAS.
The markers snpPV0027 (T:C) and snpPV0079 (A:G), both for the Phg-5 gene, contained the "T" and "A" alleles, respectively, only in the G5686 line identified as a Phg-5 source ( Keller et al. 2015). The Phg-5 allele, found in the line G5686, originates from the Andean gene pool and confers resistance to several P. griseola pathotypes from the Andean and Mesoamerican gene pools. Mahuku et al. (2009) studied the reaction of the G5686 line against 15 P. griseola pathotypes and reported resistance for 53 and 72% of the Andean and Mesoamerican isolates tested, respectively. However, for pathotypes of Mesoamerican origin, such as 63-23, 63-31, 63-47 and 63-63, which occur more frequently on Brazilian farms, the line G5686 was moderately susceptible or susceptible (Mahuku et al. 2009), justifying the limited use of Phg-5 allele sources in crossing blocks of the main common bean breeding program in Brazil.
In the common bean - P. griseola pathosystem, combining the Phg-1 (Andean) and Phg-2 genes (Mesoamerican) would be important to produce plants resistant to pathotypes 63-19, 63-23 and 63-39 (Bassi et al. 2017). The molecular characterization of Embrapa's elite germplasm allowed the identification of elite lines containing pyramided alleles, such as CNFC 16636 (Phg-1 + Phg-2) (Table S4) and the differential varieties of anthracnose pathotypes G2333, G2858 and PI 207262 (Co-u + Co-42) (Table S6). This demonstrates the need to obtain lines that simultaneously combine alleles that confer resistance to anthracnose and angular leaf spot, in addition to the other agronomic characteristics demanded by the market.
3.1.2 SNP markers associated with the Co-4 and Co-u loci
The Co-4 locus is located close to a telomeric region of the chromosome Pv-08, characterized by containing about 18 copies of the COK-4 gene and described as being associated with anthracnose resistance in the common bean (Oblessuc et al. 2015). In the present study, four SNP markers associated with the anthracnose-resistant locus were analyzed. In case of the markers snpPV0068 (G:C) and snpPV0070 (G:T), the "G" alleles (of both SNPs) were identified in the G-2333, SEL 1308, PI 207262, K-10, K-13, and CNFC 5547 lines (Table S6), which are known sources of the resistant allele of Co-42 (Kelly and Vallejo 2004; Vieira et al. 2018). Another 22 black grain lines also displayed this same allele (Table S6). Some of these lines, already characterized in terms of resistance/susceptibility, are part of the differential varieties of anthracnose pathotypes or originated in the anthracnose resistance breeding program (Table S6). The presence of “G” alleles in the two SNPs also coincided with the polymorphisms identified in the snpP8282v3-817 marker (GRAF1, A:G) (Cieslak et al. 2015) in detecting the Co-42 allele of the Co-4 locus in lines K-10, K-13 and CNFC 5547 (Table S6). However, there was no “A” allele amplification of the snpP8282v3-817 marker in the 22 black bean lines amplified by snpPV0070 (Table S6). This result indicates that snpPV0070 is not specific for the Co-42 allele, but capable of detecting resistance at the Co-4 locus. An alternative to snpPV0070 are the markers snpPV0068 and snpP8282v3-817, both amplified in lines that are known to carry the Co-42 allele (Table S6). Portilla et al (2021) when evaluating the Co-4 marker snpPV0069 (closely linked to the snpPV0068 and snpPV0070), developed from G2333 genotype (Lobaton et al. 2018) and available at HTPG portfolio, did not identify any significant effect of association, which was suggestive of a race specific resistance gene interaction. The Co-42 is a dominant allele of the Co-4 locus and has been used by breeding programs worldwide (Kelly and Vallejo 2004; Vieira et al. 2018) due to its broad spectrum of resistance to several C. lindemuthianum pathotypes (Balardin and Kelly 1998; Silvério et al. 2002).
For snpPV0046 (A:G, Co-u locus), located in the exon of the Phvul.002G328300 gene, the “A” allele is associated with the Co-u resistant allele, which also confers resistance to anthracnose (Zuiderveen et al. 2016; Oblessuc et al. 2014; Geffroy et al. 2008). In this study, the “A” allele was identified in 51 of the 376 lines genetically characterized with snpPV0046, including the BAT 93 breeding line (Table S7), which contains the parental genotype PI 207262 (Geffroy et al. 2008) and is a source of the Co-u allele (Geffroy et al. 2008). In the present study, the “A” allele was amplified in PI 207262, suggesting that Co-u may come from this parent or from a mutation that occurred during the evolutionary process of BAT 93. Of the 51 lines that contain the "A" allele, 13 are from the black bean group, 4 from the Carioca group and the remainder belong to different commercial classes, such as brindle, jalo, white, calima (Table S7), which are detected in Andean and Mesoamerican gene pools. Of these, 12 are resistant cultivars that belong to differentiating varieties of the anthracnose pathotypes (Table S7). The Co-u allele is located in the Pv-02 linkage group, very close to I locus, conferring resistance to important common bean viruses, such as the bean common mosaic virus, potyviruses and comovirus (Meziadi et al. 2016; Geffroy et al. 2008). At the molecular level, these two loci may have a common origin, since most resistance genes in plants encode nucleotide-binding site leucine-rich repeat (NBS-LRR) proteins, responsible for recognizing several pathogens (Geffroy et al. 2009). Recent information confirms this hypothesis, since the functional annotation of resistance genes has revealed common protein motifs, such as: Leucine Zippers-LZ, Leucine Rich Repeat-LRR and protein kinase domains (Protein Kinase - PK) (Fritsche-Neto et al. 2019; Nogueira et al. 2019; Queiroz et al. 2019; Nay et al. 2019a; Banoo et al. 2020; Gonçalves-Vidigal et al. 2020; Costa et al. 2021). Thus, validating the snpPV0046 marker for assisted selection is particularly important for programs aimed at incorporating multiple resistance into elite germplasms.
In addition, the introgression of resistance alleles from different gene pools (Andean and Mesoamerican) in the same line through assisted selection by molecular markers is an important strategy for developing bean cultivars with broad and long-lasting resistance (Miller et al. 2018; Vieira et al. 2018; Miklas et al. 2006; Kelly and Miklas 1998; Guzmán et al. 1995). In Brazil, an example of an important combination of anthracnose resistance alleles is that of the allele of Mesoamerican origin Co-42 ('SEL1308' and 'K13') and its Andean counterpart Co-1 (Co-14) (Vieira et al. 2018; Souza et al. 2014).
3.1.3 Simple linear regression analysis of marker effects
Of the eight SNP markers analyzed, five (snpPV0046, snpPV0068, snpPV0070, snpP8282v3-817 and snpPV0025) showed potential for indirect selection of common bean genotypes containing the Co-42 and Phg-2 alleles and were therefore submitted to simple linear regression analysis. Only for the marker snpPV0046 the regression model was significant (Table S8), but they do not explain much of the variability (low R-square of 3%). Although marker effects were not significant in the sample set of lines containing information regarding the reaction to anthracnose and angular leaf spot (snpPV0046 = 133; snpPV0068 = 134; snpPV0070 = 115; snpP8282v3-817 = 46 and snpPV0025 = 78), all lines carrying the target SNPs showed resistance, suggesting that these alleles can be monitored from these markers in the lines of interest. The absence of significance in the regression test is certainly due to the fact that many lines, even though resistant, did not exhibit the target alleles of this study, suggesting the presence of other anthracnose and angular leaf spot-resistant alleles. When the regression analysis was applied in a subset of genotypes contrasting for the resistance/susceptibility to the anthracnosis and ALS diseases (Table S9), the Co-42 and Phg-2 marker effects were significant (Table 4) and the slope values were negative for all markers (Table 4), revealing an association between SNPs and resistance to anthracnose and ALS. The introgression of resistance alleles into elite bean germplasms assisted by molecular markers is a promising strategy in breeding programs, given that it reduces time and costs in the initial selection stages (Sakiyama et al. 2014). Fig. 1 illustrates the importance of resistance alleles Co-u ("A"), Co-42 ("G") and Phg-2 ("G") in reducing mean phenotypic values in the lines that contain them.
Table 4
Summary of the regression analysis framework between SNP markers and resistance reaction to anthracnose or angular leaf spot of common bean genotypes.
snpPV46 (Co-u) - Anthracnose
|
Source of variation
|
Df
|
SS
|
MS
|
F-value
|
p-value
|
R2
|
Inclination (2)
|
A vs G (1)
|
1
|
283.9
|
283.9
|
255.7
|
2.2E-16
|
0.86
|
-2.54
|
Residual
|
42
|
46.6
|
1.1
|
|
|
|
|
snpPV0068 (Co-42) - Anthracnose
|
Source of variation
|
Df
|
SS
|
MS
|
F-value
|
p-value
|
R2
|
Inclination
|
G vs C
|
1
|
158.5
|
158.5
|
175.7
|
2.97E-12
|
0.88
|
-2.95
|
Residual
|
23
|
20.7
|
0.9
|
|
|
|
|
snpPV0070 (Co-42) - Anthracnose
|
Source of variation
|
Df
|
SS
|
MS
|
F-value
|
p-value
|
R2
|
Inclination
|
G vs T
|
1
|
167.9
|
167.9
|
93.0
|
2.14E-10
|
0.77
|
-2.41
|
Residual
|
28
|
50.6
|
1.8
|
|
|
|
|
snpP8282v3-817 (Co-42) - Anthracnose
|
Source of variation
|
Df
|
SS
|
MS
|
F-value
|
p-value
|
R2
|
Inclination
|
A vs G
|
1
|
112.2
|
112.2
|
78.2
|
2.45E-07
|
0.83
|
-3.03
|
Residual
|
15
|
21.5
|
1.4
|
|
|
|
|
snpPV0025 (Phg-2) – Angular Leaf Spot
|
Source of variation
|
Df
|
SS
|
MS
|
F-value
|
p-value
|
R2
|
Inclination
|
G vs T
|
1
|
66.2
|
66.2
|
72.2
|
4.52E-08
|
0.78
|
-1.76
|
Residual
|
20
|
18.3
|
0.9
|
|
|
|
|
The alleles in bold are associated with resistance; (1) Contrast considered in the regression analysis between marker locus alleles and the severity of anthracnose or angular leaf spot infection; (2) Angular coefficient of the linear regression equation. The negative sign on the slope indicates that the allele is associated with resistance. |
Additionally, gene annotation revealed that snpPV0046 (Co-u), snpPV0068 (Co-42), snpPV0070 (Co-42) and snpP8282v3-817 (Co-42) are located in gene regions that encode defense proteins in plants (Table 2). The snpPV0046 marker is found in the gene that encodes the Mitogen-Activated Protein Kinase (MAPK) protein, which interacts with salicylic acid, a plant hormone known to play a role in plant-acquired resistance against pathogen infection (Jagodzik et al. 2018). The snpPV0068 marker is located in the MYB transcription factor (MYB Transcription Factor) coding region, playing an essential role in the control of cellular processes in response to biotic and abiotic stresses (Ambawat et al. 2013), such as resistance to Pseudomonas syringae pv. tomato (Pst DC3000) in Arabidopsis (Zhang et al. 2019). The snpPV0070 and snpP8282v3-817 markers are located in the Phvul.008G028400 and Phvul.008G028200 transcripts, respectively, homologous to the COK-4 locus (STK domain) (Table 2), previously described as an anthracnose resistance source in common bean (Melotto and Kelly 2001).
3.2 Validation of the snpPV0070, snpP8282v3-817 and snpPV0025 markers
TaqMan® hydrolysis probes were developed from the snpPV0070 and snpP8282v3-817 markers, targeting the Co-42 allele. 149 F2 plants from the cross between BRS Cometa and SEL1308 were genotyped with two probes and phenotyped for reaction to C. lindemuthianum pathotype 73 (Table S2). Of these, 110 were characterized as resistant and 39 as susceptible, in line with the expected ratio of 3R:1S (χ2 = 0.11; p = 0.74) (Table 5). Previous studies support the hypothesis that resistance to anthracnose contained in SEL 1308 is controlled by a single locus, with complete dominance (Young et al. 1998; Oblessuc et al. 2015). Markers snpPV0070 and snpP8282v3-817 maintained 1RR:2Rr:1rr ratio (Table 5), segregating as expected for codominant markers. Linkage analysis revealed that the snpPV0070 and snpP8282v3-817 markers used in F2 population genotyping (BRS Cometa x SEL 1308) are strongly linked to the Co-4 locus, with a recombination frequency of 0.026 (2.6 cM) and 0.019 (1.9 cM), respectively (Table 5). Markers snpPV0070 and snpP8282v3-817 showed selection efficiency (SE) of 99.7% and 99.8% respectively, indicating the high potential value of molecular markers in strong linkage disequilibrium in MAS programs.
Table 5
Genotypic and phenotypic segregation of individuals from the F2 population, originated from the crossing between BRS Cometa x SEL 1308, evaluated for reaction to Coletotrichum lindemuthianum pathotype 73, the causal agent of anthracnose in the common bean.
Genotype/Phenotype
|
Observed
|
Expected
|
Hypothesis
|
ꭕ2
|
P-value
|
rf
|
Distance1
|
SE
|
snpPV0070
|
T:T
|
40
|
37
|
1:2:1
|
0.72
|
0.70
|
0.026
|
2.6 cM
|
99.7%
|
G:T
|
76
|
75
|
G:G
|
33
|
37
|
P8282v3-817
|
G:G
|
39
|
37
|
1:2:1
|
0.65
|
0.72
|
0.019
|
1.9 cM
|
99.8%
|
A:G
|
77
|
75
|
A:A
|
33
|
37
|
Class
|
Resistant (1-3)
|
110
|
112
|
3:1
|
0.11
|
0.74
|
-
|
-
|
-
|
Susceptible (4-9)
|
39
|
37
|
Total
|
149
|
149
|
-
|
-
|
-
|
-
|
-
|
-
|
The alleles in bold are associated with resistance; rf Recombination fraction; SE Selection efficiency; 1Distance in centiMorgans; P-value associated to the null hypothesis not rejected (1:2:1 for molecular markers data and 3:1 for phenotype data); snpPV0070: "G:G" dominant homozygous for the Co-42 allele, "G:T" heterozygous for the Co-42 allele and "T:T" recessive homozygous for the Co-42 allele; snpP8282v3-817: “A:A” dominant homozygous for the Co-42 allele, “A:G” heterozygous for the Co-42 allele and “G:G” recessive homozygous for the Co-42 allele.
The TaqMan® probe derived from snpPV0025 (G:T) associated with the Phg-2 allele was evaluated in a set of contrasting elite lines for resistance to angular leaf spot (Table S3). This target SNP was able to detect the Phg-2 locus, despite its amplification in varieties that are susceptible to angular leaf spot (Table S3). This is in accordance with previously reported that snpPV0025 would only effectively tag Phg-2 in Andean backgrounds. Additional SNP markers have been identified as associates with Phg-2 and should be tested (Gil et al. 2019; Nay et al. 2019a). On the other hand, the complementary analysis performed with RAPD_SEO4 (Sartorato et al. 1999) and STS_g796 (Miller et al. 2018) markers resulted in specific markers for the Phg-2 locus, amplifying in resistance sources Mexico 54 and MAR-2 (Table S3). These two markers have been frequently used in studies aimed at Phg-2 allele introgression (Sanglard et al. 2016; Miller et al. 2018). In addition, SEO4 also amplified in lines DM 108, CNFC 17142, CNFC 17395 and CNFC 18710; while g796 amplified in lines CNFC 15086 and CNFC 17153 (Table S3). Among the parental lines used by the breeding program in crossing blocks, DM108 stands out, originating from the cross between BRS Rudá and MAR-2 (Sanglard et al. 2016). The MAR-2 line contains the resistant Phg-2 locus in its genome; however, the allelic variation of this locus in this variety is still not clear (Nay et al. 2019a). Thus, the presence of the Phg-2 locus in the elite lines of Embrapa’s common bean breeding program may be related to the previous use of the DM108 line as parental line in the crossing blocks. The Phg-2 locus is particularly important in angular leaf spot resistance, since it confers resistance to the most prevalent pathotypes in Brazilian common bean crops (Bassi et al. 2017).