The alleles bc‐u ^d and bc‐u ^r (previously bc‐4 gene), representing coding mutations within Vps4 AAA+ ATPase ESCRT protein, interact with other genes to condition resistance to BCMV and BCMNV in common bean

Abstract

Bean common mosaic virus (BCMV) and bean common mosaic necrosis virus (BCMNV) have a damaging impact on global common bean (Phaseolus vulgaris L.) cultivation, causing potential yield losses of over 80%. The primary strategy for controlling these viruses is through host plant resistance. This research aimed to identify and validate structural variations for the bc‐u ^d gene as revealed by long‐read sequencing, develop an efficient DNA marker to assist selection of bc‐u ^d in snap and dry beans, and examine the interactions between the bc‐u ^d allele and other BCMV resistance genes. A gene (Phvul.005G125100) model on chromosome Pv05, encoding a vacuolar protein‐sorting 4 (Vps4) AAA+ ATPase endosomal sorting complexes required for transport (ESCRT) protein, was identified as the best candidate gene for bc‐u ^d. An 84‐bp repetitive insertion variant within the gene, exhibited 100% co‐segregation with the bc‐u ^d resistance allele across 264 common bean accessions. The 84‐bp repetitive insertion was labeled with an indel marker IND_05_36225873 which was useful for tracking the bc‐u ^d allele across diverse germplasm. A different single nucleotide polymorphism variant within the same candidate gene was associated with the bc‐4 gene. Segregation in F₂ populations confirmed bc‐u ^d and bc‐4 were alleles, so bc‐4 was renamed bc‐u ^r to fit gene nomenclature guidelines. The interactions of bc‐u ^d and bc‐u ^r with other resistance genes, such as bc‐1 (receptor‐like kinase on Pv03) and bc‐2 (Vps4 AAA+ ATPase ESCRT protein on Pv11), validated gene combinations in the differential “host groups” effective against specific BCMV/BCMNV “pathogroups.” These findings increase our understanding of the Bc‐u locus, and enhance our ability to develop more resilient bean varieties through marker‐assisted selection, reducing the impact of BCMV and BCMNV.

Core Ideas

• The Bc‐u locus contains multiple recessive resistance alleles (bc‐u ^d and bc‐u ^r).

• A gene encoding vacuolar protein‐sorting 4 AAA+ ATPase endosomal sorting complexes required for transport protein on Pv05 is the candidate gene for Bc‐u.

• Different causal mutations within the candidate gene underlie bc‐u ^d and bc‐u ^r alleles.

Abbreviations

BCMNV

bean common mosaic necrosis virus

BCMV

bean common mosaic virus

DDP

Durango Diversity Panel

HG

host group

IND

indel

PG

pathogroup

RLK

receptor‐like kinase

SnAP

Snap Bean Association Panel

SNP

single nucleotide polymorphism

Vps4

vacuolar protein‐sorting 4

1. INTRODUCTION

Bean common mosaic virus (BCMV) and bean common mosaic necrosis virus (BCMNV) are closely related positive‐stranded RNA viruses in the Potyvirus genus that infect common bean (Phaseolus vulgaris L.) worldwide. These viruses are primarily transmitted by several aphid species in a nonpersistent manner and through infected seeds. Seed‐borne transmission is a major concern for subsistence farmers who plant their saved seeds from previous harvests. The yield loss caused by BCMV and BCMNV in common bean production fields can exceed 80% (Morales, 2003).

The main strategy for disease control is to deploy host plant resistance. Drijfhout (1978) conducted a comprehensive study that describes the interaction between common bean and BCMV. His model for host plant resistance included four loci containing six recessive resistance alleles (bc‐1, bc‐1 ², bc‐2, bc‐2 ², bc‐3, and bc‐u), along with the dominant “I gene.” The presence of the bc‐u gene was necessary for the expression of the other recessive resistance genes, except in the presence of the I gene. The model categorized isolate diversity into seven pathogroups (PG‐I to PG‐VII) based on interactions with 11 host groups (HG‐1 to HG‐11) with different resistance gene combinations. Some minor adjustments to the model include a new PG (PG‐VIII) with a BCMV isolate capable of overcoming the bc‐3 gene in the HG‐7 genotype IVT‐7214 (Feng et al., 2015) and a new HG (HG‐12) with the I and bc‐3 combination of resistance genes (Larsen et al., 2005).

Moreover, recent major adjustments described by Soler‐Garzón et al. (2021a) include that the Bc‐1 locus has only one resistance allele, “bc‐1,” which exhibits a differential resistance effect dependent upon presence versus absence of bc‐u. Because the bc‐u “helper” gene was absent in HG‐2, HG‐4, HG‐5, and HG‐7 differential cultivars, where it was expected, they designated the presence of bc‐u in HG‐3, HG‐6, HG‐10, and HG‐11 with the new symbol, “bc‐u ^d.” Additionally, Soler‐Garzón et al. (2021b) found that the Bc‐2 locus similarly possessed just one resistance allele, bc‐2, with a differential resistance reaction based on presence of bc‐u ^d (e.g., in HG‐6 and HG‐11) versus another tightly linked gene they named bc‐4 (e.g., in HG‐4 and HG‐5).

Next‐generation sequencing (NGS) technologies have played a crucial role in identifying candidate genes and DNA markers associated with the BCMV resistance loci. For the bc‐1 gene, two receptor‐like kinases (RLKs) on chromosome 3 (Pv03) were identified as candidate genes, and the bc‐u ^d gene was associated with a basic leucine zipper transcription factor protein on Pv05 (Soler‐Garzón et al., 2021a). The bc‐2 and bc‐4 loci were associated with two homologous genes on Pv05 and Pv11, respectively, which encode a vacuolar protein‐sorting 4 (Vps4) AAA+ ATPase endosomal sorting complexes required for transport (ESCRT) protein (Soler‐Garzón et al., 2021b). For the bc‐3 gene, a eukaryotic translation initiation factor 4E protein was identified as the candidate gene located on Pv06 (Hart & Griffiths, 2013; Naderpour et al., 2010). Although the I gene was associated with a cluster of genes encoding nucleotide‐binding site‐leucine‐rich repeat (NBS‐LRR) proteins (Vallejos et al., 2006), a specific candidate gene has been difficult to find due to the repetitive sequences across the region. To track the I gene, markers next to the NBS‐LRR gene cluster have been used, including one single nucleotide polymorphism (SNP) marker, S02_48908259, located downstream from the NBS‐LRR gene cluster and tightly linked with the I gene (Soler‐Garzón et al., 2021a).

Core Ideas

• The Bc‐u locus contains multiple recessive resistance alleles (bc‐u ^d and bc‐u ^r).

• A gene encoding vacuolar protein‐sorting 4 AAA+ ATPase endosomal sorting complexes required for transport protein on Pv05 is the candidate gene for Bc‐u.

• Different causal mutations within the candidate gene underlie bc‐u ^d and bc‐u ^r alleles.

The linked candidate gene markers for the BCMV resistance loci have been useful for tracking and deploying the resistance genes in common bean germplasm. However, we have recently observed numerous recombination events between the Pvbzip1_A_C SNP marker associated with the favorable bc‐u ^d allele in snap bean germplasm. This observation necessitates reexamination of the bZIP candidate gene model for bc‐u ^d in both snap and dry beans. The specific objectives of this study were to (i) identify structural variations in the bc‐u ^d region using long‐read whole‐genome sequencing, (ii) develop a DNA marker to enhance marker‐assisted selection efficiency for bc‐u ^d in snap and dry bean germplasm, and (iii) investigate the interactions of bc‐u ^d with other BCMV resistance genes in segregating populations inoculated with different BCMV and BCMNV strains.

2. MATERIALS AND METHODS

2.1. GN 1140 genome assembly and variant calling

To further examine the genomic region for bc‐u ^d, we used GN 1140 (syn US‐1140), a great northern dry bean cultivar developed and released by the United States Department of Agriculture‐Agricultural Research Service (USDA‐ARS), the Montana Agricultural Experiment Station, and the Idaho Agricultural Experiment Station in 1960. GN 1140 is a member of the Durango Diversity Panel (DDP) with accession number DDP075, which contributed to the discovery of bc‐u ^d (Soler‐Garzón et al., 2021a). GN 1140 carries the bc‐1 and bc‐u ^d genes, exhibiting a mild mosaic (mM) reaction to BCMNV NL‐3 strain. A draft genome assembly of GN 1140 was obtained by creating 10X linked read libraries (https://www.10xgenomics.com/) and sequencing and assembling the reads into scaffolds (McClean et al., 2022).

A total of 93,329 scaffolds of GN 1140 were mapped to the P. vulgaris Andean G19833 v2.1 reference genome using Minimap2 (Li, 2018). Executing the next‐generation sequencing experience platform pipeline (Duitama et al., 2014; Lobaton et al., 2018), a total of 5,354,277 variants, including single nucleotide variants, insertions or deletions (indels), and short tandem repeats (STRs), were mapped to the 11 chromosomes and scaffolds of the Chaucha Chuga (G19833) v2.1 P. vulgaris reference genome downloaded from Phytozome v13 (Schmutz et al., 2014). To confirm the accuracy of the variant calling in the bc‐u ^d region, a visual examination of the variants was performed using Geneious v9.1.8 (Kearse et al., 2012).

2.2. Plant material

The Snap Bean Association Panel (SnAP) and DDP, comprising 378 and 180 accessions, respectively, were genotyped using the BCMV SNP markers developed by Soler‐Garzón et al. (2021a, 2021b) for detection of I (S02_48908259), bc‐1 (S03_4203361), bc‐u ^d (Pvbzip1_A_C), and bc‐4 (Pvmit‐1_T_G) genes (Table 1). These SNP markers were amplified by polymerase chain reaction (PCR) using the protocol conditions standardized by Soler‐Garzón et al. (2021a, 2021b) and described below in the “DNA extraction and genotyping” section.

TABLE 1.

T _m‐shift single nucleotide polymorphism (SNP) and indel markers linked to genes for resistance to bean common mosaic virus (BCMV) assayed in F₂ individual plants and parental lines.

BCMV resistance alleles	ID marker	Chr.	Position (bp) a	Allele resistant	Allele susceptible	Reference
bc‐u d	IND_05_36225873	Pv05	36,225,873	Insertion 84 bp	–	This study
bc‐u r (formerly bc‐4)	Pvmit‐1_T_G	Pv05	36,225,550	G	T	Soler‐Garzón et al. (2021b)
Linked to bZIP gene	Pvbzip1_A_C	Pv05	36,114,516	A	C	Soler‐Garzón et al. (2021a)
I	S02_48908259	Pv02	48,908,259	T	A	Soler‐Garzón et al. (2021a)
bc‐1	S03_4203361	Pv03	4,203,361	T	A	Soler‐Garzón et al. (2021a)

Abbreviation: Chr., chromosome.

^a Based on the Andean reference genome G19833 v2.1, available in Phytozome 13 (https://phytozome‐next.jgi.doe.gov/info/Pvulgaris_v2_1).

A subset of 188 accessions with the I, bc‐1, and bc‐u ^d combination of genes, but variable for phenotypic reaction (lethal systemic top necrosis or restricted vein necrosis) to the NL‐3 strain of BCMNV in PG‐VI was selected from the SnAP (Table S1). The SnAP consists of snap bean cultivars, including public and privately released varieties primarily from the United States, some landrace cultivars, and a few breeding lines. Recently, the SnAP was used for genome‐wide association studies (GWAS) for resistance to white mold (Sclerotinia sclerotiorum Lib. de Bary) (Arkwazee et al., 2022) and to beet curly top virus (Soler‐Garzón et al., 2023). A subset of 59 accessions from the DDP were selected for presence of either I/bc‐1/bc‐u ^d (nine accessions), bc‐1/bc‐u ^d (47 accessions), or bc‐1/bc‐4 (three accessions) genotypes (Table S2).

Two F₂ populations segregating for bc‐u ^d based on linked markers or phenotypic reactions were generated by crossing Beryl (DDP055) great northern bean with the SnAP accessions FR 266 (SnAP136) and Ulysses (SnAP359). Beryl possesses the markers linked to the I + bc‐1 + bc‐u ^d genes supported by a restricted vein necrosis phenotypic reaction to NL‐3. FR 266 has the markers linked to the I + bc‐1 + bc‐u ^d + bc‐4 genes representing the rare recombinant for bc‐u ^d + bc‐4 genes in coupling phase linkage; however, a top necrosis phenotypic reaction to NL‐3 strain indicates that bc‐u ^d gene is absent. Ulysses possesses the I + bc‐1 + bc‐4 gene combination based on markers and, like FR 266, exhibits a top necrosis reaction to NL‐3. From these two populations, there were 121 individual F₂ plants genotyped with BCMV linked markers bc‐u ^d (Pvbzip1_A_C) and bc‐4 (Pvmit‐1_T_G) developed by Soler‐Garzón et al. (2021a, 2021b) and evaluated for reaction to NL‐3 strain as described below (Table S3).

2.3. DNA extraction and genotyping

Genomic DNA from a subset of the DDP (59 accessions), SnAP (188 accessions), and 17 additional lines were isolated for SNP marker testing. Leaf tissue weighing 20 mg was collected from individual plants grown in the USDA‐ARS greenhouses in Prosser, WA. The Qiagen DNeasy 96 Plant Kit (Hilden) was used for DNA extraction. For individual F₂ plants, total DNA was extracted from four‐leaf disks (approximately 30 mm²) using an alkaline extraction method with modifications based on Xin et al. (2003). The method involved using Buffer A (composed of 50 mM NaOH and 1% Tween 20) to degrade cell walls and Buffer B (consisting of 100 mM Tris‐HCl and 1.7 mM EDTA pH 8) for DNA neutralization and protein binding. The extracted DNA was diluted 1:7, and a 5 µL DNA template dilution was used for PCR in a final volume of 100 µL in a 96‐well plate.

All SNP and indel markers used in this study were amplified by PCR using an Eppendorf Mastercycler (Eppendorf AG). The PCR volume was 20 µL, which included 1X Taq buffer, 1.5 mM of MgCl₂, 0.2 mM of dNTPs mix (Promega), 0.15 µM of each primer, 1X EvaGreen dye, 0.1 µL of Taq polymerase (Promega), and 20 ng of genomic DNA. The PCR was performed with the following thermal profile: initial denaturation at 94°C for 2 min, followed by 38 cycles of denaturation at 92°C for 20 s, annealing for 20 s (i.e., the annealing temperature was specific to each primer assay), and extension at 72°C for 20 s. A final extension step was performed at 72°C for 5 min. DNA melting analysis for allele determination of the PCR products was conducted using a fluorescence‐detecting QuantStudio 5 Real‐Time PCR (Thermo Fisher Scientific). EvaGreen fluorescent detection was performed for 15 s at 95°C, and the melting curve data were collected with a ramp rate of 0.04°C/s from 70°C to 98°C.

Alternatively, the resulting PCR products from the indel marker were evaluated for polymorphisms on a 2% agarose gel in 1 × Tris‐acetate‐EDTA buffer on a Bio‐Rad (Sub‐cell Model 192) gel electrophoresis system for 1 h at 100 V. GelRed (Biotium) was added to the gel to visualize the DNA bands on a UVP GelDoc IT 310 Imaging system.

2.4. Evaluation of BCMV and BCMNV reactions

All F₂ populations, accessions, and additional lines were evaluated for reaction to the NL‐3 (D) strain of BCMNV, which traces back to the Drijfhout collection. NL‐3 was maintained in the universal susceptible HG‐1 differential cultivar Sutter Pink, and the youngest leaves with mosaic symptoms were used as inoculum. Inoculations were conducted in the USDA‐ARS greenhouses in Prosser, WA, under controlled conditions (22°C–28°C temperature range and 14‐h photoperiod, supplemented with artificial lighting as needed). Test plants were sown in 9‐cm³ pots with a maximum of three plants per pot, using a commercial potting mix (Sun Gro Horticulture). About 10 days after planting, the two primary leaves of each plant were mechanically inoculated with the same strain using the procedure described by Drijfhout (1978).

Disease reactions were observed once a week for 5 weeks post‐inoculation (wpi). The symptoms for each plant were recorded as follows: VN, restricted vein necrosis on inoculated leaves, no systemic symptoms; VN⁺, restricted vein necrosis on inoculated leaves, with some small patches (10 mm²) of systemic restricted vein necrosis on upper trifoliolate leaves observed from 3–5 wpi; TN, lethal systemic top necrosis by 7–10 days post‐inoculation (dpi), resulting in plant death; and dTN, delayed top necrosis beginning >11 dpi, most often resulting in plant death. HG‐9 genotype, Topcrop (I + bc‐1), which exhibited TN to NL‐3, was included as a control.

3. RESULTS

3.1. Examination of the bc‐u ^d region in GN 1140

A comparative analysis was performed between the genotyping data obtained from GN 1140 using short‐read Illumina technology by Soler‐Garzón et al. (2021a) and the draft genome scaffolds of GN 1140 (McClean et al., 2022). This analysis focused on the genomic interval (Pv05: 36,001,025–36,249,945 bases), which includes both bc‐u ^d and bc‐4, that was previously identified by GWAS in the DDP genotyped with SNP markers and evaluated for BCMV reaction by Soler‐Garzón et al. (2021a, 2021b). Six new variants, undetected by short‐read sequencing across a narrowed coding region, were observed in two GN 1140 scaffolds, one of 1.3 Mb and other of 55.6 kb in length. Remarkably, among these new variants, a repetitive in‐frame insertion of 84 bp was observed at Pv05: 36,225,873 bases, located within the first exon of the Phvul.005G125100 gene model from the G19833v2.1 reference genome (Figure 1A). The Phvul.005G125100 gene possesses eight exons, encoding a Vps4 AAA+ ATPase ESCRT protein. It spans 434 amino acids (aa) and contains three domains as follows: (i) microtubule interacting and transport (MIT) domain, (ii) AAA‐type ATPase (AAA+) domain, and (iii) Vps4 domain.

FIGURE 1.

Variation in (A) first nucleotide exon and (B) complete amino acid sequences of Phvul.005G125100, corresponding with bc‐u alleles. (A) Indel of 84‐bp insertion region in a first exon (yellow bar) of Phvul.005G125100 in G19833 (Bc‐u) and GN 1140 (bc‐u ^d). Bar in dark blue represents repetitive insertion. (B) Amino acid alignment of proteins of representative G19833 (wild‐type Bc‐u), Robust (bc‐u ^r, formerly bc‐4), and GN 1140 (bc‐u ^d), genotypes for the vacuolar protein‐sorting 4 (Vps4) AAA+ ATPase endosomal sorting complexes required for transport (ESCRT) protein.

Computational analysis of the in silico translated Vps4 AAA+ ATPase ESCRT protein derived from the open reading frame for Phvul.005G125100 with the 84‐bp frame insertion revealed no disruptions in the protein sequence. However, the 84 bp frame insertion was found to cause the incorporation of 28 aa at position 140, located between the MIT and AAA+ domains in the GN 1140 genotype, which possesses bc‐u ^d (Figure 1B). Henceforth, the 84‐bp frame insertion in candidate gene Phvul.005G125100 is postulated as the causal mutation for bc‐u ^d. Phvul. 005G125100 is also the candidate gene for bc‐4. An SNP missense variant, S05_36225550, in the MIT domain of Phvul.005G125100, resulting in the change of a methionine to an arginine aa in the MIT domain of the Vps4 AAA+ ATPase ESCRT protein, was postulated by Soler‐Garzón et al. (2021b) as the causal mutation for bc‐4.

3.2. Mutation‐specific marker design

The 84‐bp repetitive insertion identified in the GN 1140 long‐read sequences within Phvul.005G125100 was targeted with a specific set of primers pairs (forward: GACCCCGAGCAAGCCAAG; reverse: TCACAGGCAAAATCACAGCC) (Figure 1A; Supporting Information S1). The resulting indel marker was named IND_05_36225873 and utilized to exploit this insertion as a putative gene/mutation‐specific marker for bc‐u ^d. To validate the presence of the insertion, a melting analysis was conducted. Lines carrying the insertion exhibited a higher temperature peak (T _m ∼90°C) compared to lines lacking the insertion (T _m ∼88°C) (Figure 2A). Alternatively, the same PCR product was visualized on a 2% agarose gel. Lines with the insertion displayed a fragment of 220 bp, while lines without the insertion showed a fragment of 136 bp (Figure 2B).

FIGURE 2.

The IND_05_36225873 indel marker developed to detect the bc‐u ^d allele. (A) DNA melting analysis, showing derivative melt curve plot. (B) Polymerase chain reaction (PCR) products visualized by 2% agarose gel electrophoresis set to 100 V for 1 h.

3.3. DDP, SnAP, and host group assays

The DDP and SnAP accessions were genotyped with the new IND_05_36225873 indel marker and tested with the NL‐3 strain. For the DDP, complete linkage disequilibrium (R‐squared = 1) was observed between indel marker IND_05_36225873 and the previous Pvbzip1_A_C SNP marker that detected a missense mutation in the bZIP protein gene Phvul.005G124100, which was identified by Soler‐Garzón et al. (2021a) as the candidate gene for bc‐u^d . Among the DDP, 48 accessions exhibited mM symptoms to NL‐3 and possessed bc‐1 and both candidate gene markers, IND_05_36225873 and Pvbzip1_A_C, for bc‐u ^d. Nine DDP accessions with a VN reaction to NL‐3 possessed I + bc‐1 and bc‐u ^d as detected by both candidate gene markers, IND_05_36225873 and Pvbzip1_A_C. There were three DDP accessions with mosaic symptoms that possessed bc‐1 and bc‐4, as detected by the ‘Pvmit‐1_T_G’ SNP marker, in the same Phvul.005G125100 gene model (Table 2).

TABLE 2.

Phenotypic reaction observed in plants inoculated with NL‐3 strain in two panels genotyped by the previous Pvbzip1_A_C single nucleotide polymorphism (SNP) marker with the bZIP candidate gene for bc‐u ^d, the indel marker IND_05_36225873 within new vacuolar protein‐sorting 4 (Vps4) candidate gene for bc‐u ^d, and the Pvmit‐1_T_G SNP marker with the same Vps4 gene for bc‐u ^r (formerly bc‐4).

		Phenotypes
Lines	Genotypes (markers)	M	mM	dTN	TN	VN	Total
DDP	bc‐1/bc‐u d (bZIP‐SNP and Vps4‐indel)	–	47	–	–	–	47
	I/bc‐1/bc‐u d (bZIP and Vps4‐indel)	–	–	–	–	9	9
	bc‐1/bc‐u r (Vps4‐SNP)	3	–	–	–	–	3
	TOTAL	3	47	–	–	9	59
SnAP	I/bc‐1 (bZIP‐SNP)	–	–	16	157	–	173
	I/bc‐1/bc‐u d (bZIP‐SNP and Vps4‐indel)	–	–	–	–	14	14
	I/bc‐1/bc‐u r (bZIP‐SNP and Vps4‐SNP)	–	–	–	1	–	1
	Total	–	–	16	158	14	188

Abbreviations: DDP, Durango Diversity Panel; dTN, delayed top necrosis; M, mosaic; mM, mild mosaic; SnAP; Snap Bean Association Panel; TN, top necrosis; VN, vein necrosis.

Conversely, the SnAP accessions exhibited low linkage disequilibrium (R‐squared = 0.07) between the IND_05_36225873 and the Pvbzip1_A_C SNP marker. Fourteen SnAP accessions with I + bc‐1, and bc‐u ^d, as detected by both IND_05_36225873 and Pvbzip1_A_C markers, exhibited VN reaction against NL‐3 strain. Notably, one accession, FR 266, which exhibited TN against NL‐3, displayed cis‐linkage between the Pvbzip1_A_C SNP (detects bZIP bc‐u ^d allele) and Pvmit‐1_T_G SNP (detects bc‐4 allele) (Table 2). Tellingly, 173 accessions carrying I + bc‐1 and the favorable allele for the Pvbzip1_A_C SNP marker exhibited dTN or TN reaction to NL‐3. This evidence underscores the inadequacy of the Pvbzip1_A_C SNP marker and rejects the bZIP protein encoded by Phvul.005G124100 as the candidate gene for bc‐u ^d. Moreover, these results for the 84‐bp insertion, detected by the IND_05_36225873 marker, support Phvul.005G125100 (Vps4 AAA+ ATPase ESCRT protein) as the most likely candidate gene for the bc‐u ^d resistance gene.

Differential cultivars within HG‐3, HG‐6, HG‐10, and HG‐11, with expected phenotypic reactions to NL‐3 strain, all possessed the indel marker IND_05_36225873 indicating the presence of bc‐u ^d (Table 3). For instance, UI‐123 (HG‐3) with bc‐1 and bc‐u ^d, exhibited mM symptoms against NL‐3, whereas Othello (HG‐6) with bc‐1, bc‐u ^d, and bc‐2 alleles, exhibited no symptoms. Conversely, Sanilac (HG‐4) with bc‐2 and bc‐4, was susceptible to NL‐3. HG‐10 cultivars harboring the I gene and bc‐1 + bc‐u ^d combination displayed VN reactions, and HG‐11 cultivars with I + bc‐u ^d + bc‐2 exhibited local necrotic lesions reactions to NL‐3 strain.

TABLE 3.

Updated resistance genes for the differential host groups (HG) based candidate gene and linked markers.

		BCMV	BCMNV		S02_ 48908259 a	S03_ 4203361 a	Pvvps4_del a	Pvmit‐2_C_del a	PveIF4E1,3,4_PveIF4E 2a	Pvmit1_T_G a	IND_05 b _36225873
		PG‐VII	PG‐VI	Proposed resistance genotype	Pv02: 48,908,259	Pv03: 4,203,361	Pv11: 9,272,542 ‐ 9,262,459	Pv11: 9,278,765	Pv06: 27,204,768	Pv05: 36,225,550	Pv05: 36,225,873
	Host group	US‐6	NL‐3		I gene	bc‐1	bc‐2 [UI 111]	bc‐2 [Robust]	bc‐3	bc‐u r	bc‐u d
Line					NBS‐LRR c	RLK	Vps4	Vps4	eIF4E	Vps‐4	Vps‐4
Dubbele Witte	1	M	M	none	–	–	–	–	–	–	–
Immuna	2	M	M	bc‐1	–	+	–	–	–	–	–
UI‐123	3	M	mM	bc‐1/bc‐ud	–	+	–	–	–	–	+
Sanilac	4	NS	M	bc‐2 [Robust] /bc‐ur	–	–	–	+	–	+	–
UI‐111 (DDP077)	4	NS	M	bc‐2 [UI 111] /bc‐ur	–	–	+	–	–	+	–
UI‐114‐8	5	NS	M	bc‐1/bc‐2 [UI 111 ]/bc‐ur	–	+	+	–	–	+	–
Monroe	6	M	NS	bc‐1/bc‐ud/bc‐2 [Robust]	–	+	–	+	–	–	+
Othello (DDP109)	6	M	NS	bc‐1/bc‐ud/bc‐2 [UI 111]	–	+	+	–	–	–	+
IVT‐7214	7	NS	NS	bc‐2?/bc‐3/bc‐ur	–	–	–	–	+	+	–
Widusa	8	NS	TN	I	+	–	–	–	–	–	–
Jubila	9a	NS	VN+	I/bc‐1	+	+	–	–	–	–	–
Topcrop	9b	NS	TN	I/bc‐1	+	+	–	–	–	–	–
Beryl (DDP055)	10	NS	VN	I/bc‐1/bc‐ud	+	+	–	–	–	–	+
IVT‐7223	11	NS	NLL	I/bc‐2 [UI 111]/bc‐ud	+	–	+	–	–	–	+
Quincy (DDP103)	11	NS	NLL	I/bc‐1/ bc‐2 [UI 111] /bc‐ud	+	+	+	–	–	–	+
Raven	12	NS	NS	I/bc‐3	+	–	–	–	+	–	–
TARS‐VR‐8S	12	NS	NS	I/bc‐3/bc‐ud	+	–	–	–	+	–	+

Abbreviations: BCMV, bean common mosaic virus; BCMNV; bean common mosaic necrosis virus; dM, delayed mosaic; dTN, delayed top necrosis; eIF4E, eukaryotic translation initiation factor 4E protein; M, mosaic; mM, mild mosaic; NBS‐LRR, nucleotide‐binding site‐leucine‐rich repeat; NLL, local necrotic lesions; NS, no symptoms; SNP, single nucleotide polymorphism; TN, top necrosis; VN, vein necrosis; VN⁺, vein necrosis on upper trifoliolate leaves.

^a SNP markers designed by Soler‐Garzón et al. (2021a, 2021b).

^b Indel marker developed in this study.

^c SNP marker linked to a cluster of NBS‐LRR genes.

The results above suggest that bc‐u ^d and bc‐4 may be different alleles in the same Phvul.005G125100 gene (Vps4 AAA+ ATPase ESCRT protein) conditioned by different mutations, an 84‐bp frame insertion and SNP, respectively. Because the bc‐u gene symbol predates the bc‐4 gene symbol by 45 years, we propose maintaining the gene nomenclature suggested by Drijfhout (1978) and renaming bc‐4 as bc‐u ^r. The superscript “r” signifies that the bc‐u ^r allele was initially identified in Robust navy bean of race Mesoamerican origin. Although the candidate gene Phvul.005G125100 for bc‐u ^d is new, the gene symbol remains unchanged.

3.4. Allelism test in F₂ populations

Two segregating F₂ populations were scored with the IND_05_36225873 marker (Table 4). Notably, the GN parent Beryl, with the I and bc‐1 markers, possessed the new indel marker for bc‐u ^d. Based on indel allele marker data, the snap bean FR266 parent lacked the 84‐bp insertion. This observation and an unprotected I gene TN reaction to NL‐3 strain support the absence of the bc‐u ^d gene in FR266. FR 266 possessed the Pvmit‐1_T_G SNP marker indicating presence of bc‐u ^r (formerly bc‐4), whereas Beryl lacked this marker. The phenotypic segregation of the FR 266/Beryl F₂ population in response to the NL‐3 strain has a ratio of 1 (VN):2 (VN⁺ or dTN):1 (dTN or TN). Among the genotypes observed, 28 individual plants with II/bc‐1bc‐1/bc‐u ^d bc‐u ^d exhibited VN, 31 individuals in a heterozygous state for the bc‐u ^d indel and the bc‐u ^r linked marker (e.g., bc‐u ^d bc‐u ^r) showed dTN or VN in primary leaves and systemic VN in upper (VN⁺) trifoliolate leaves which were observed from 3 to 5 wpi, and 19 individuals with II/ bc‐1bc‐1/bc‐u ^r bc‐u ^r had a TN or dTN reaction. These results further support the 84‐bp insertion as a causal mutation in the Vps4 candidate gene for bc‐u ^d. The intermediate dTN/VN+ phenotype for the bc‐u ^d bc‐u ^r heterozygotes indicates incomplete dominance.

TABLE 4.

Co‐segregation of bc‐u ^d (IND_05_36225873) and bc‐u ^r (Pvmit‐1_T_G) alleles in F₂ populations evaluated with bean common mosaic necrosis virus (BCMNV) NL‐3 strain.

Genotype	Genes	Phenotype	No. of plants	${\mathrm{\chi }}^2$ p = 0.05 (df = 2) a
Beryl	II/bc‐1bc‐1/bc‐u d bc‐u d	VN	3	–
FR 266	II/bc‐1bc‐1/bc‐u r bc‐u r	TN	3	–
Ulysses	II/bc‐1bc‐1/bc‐u r bc‐u r	TN	3	–
F2: FR266 × Beryl	II/bc‐1bc‐1/bc‐u d bc‐u d	VN	28	4.09
	II/bc‐1bc‐1/bc‐u d bc‐u r	VN+ or dTN	34
	II/bc‐1bc‐1/bc‐u r bc‐u r	TN or dTN	19
F2: Ulysses × Beryl	II/bc‐1bc‐1/bc‐u d bc‐u d	VN	9	1.56
	II/bc‐1bc‐1/bc‐u d bc‐u r	VN+ or dTN	23
	II/bc‐1bc‐1/bc‐u r bc‐u r	TN or dTN	8

Abbreviations: dTN, delayed top necrosis; TN; top necrosis; VN, vein necrosis; VN⁺, vein necrosis on upper trifoliolate leaves.

^a Expected 1:2:1 segregation ratio.

Ulysses snap bean with the same I + bc‐1 + bc‐u ^r genotype as FR266 exhibited a similar 1:2:1 genotypic/phenotypic ratio in the F₂ generation when crossed with Beryl (Ulysses/Beryl). Nine plants with II/bc‐1bc‐1/bc‐u ^d bc‐u ^d genotype exhibited a VN phenotype, 23 plants with heterozygous II/bc‐1bc‐1/bc‐u ^d bc‐u ^r genotype displayed a VN⁺ or dTN phenotype, and eight plants with II/bc‐1bc‐1/bc‐u ^r bc‐u ^r genotype exhibited TN or dTN reactions to NL‐3 strain.

4. DISCUSSION

A new candidate gene for bc‐u ^d resistance allele was identified by genotyping and phenotyping BCMNV reactions in two different common bean panels, the SnAP and the DDP. Initially, an SNP mutation (Pv05: 36,114,516 bases) in a bZIP protein encoded by gene model Phvul.005G124100 (G19833 v2.1) was proposed as a candidate gene for bc‐u ^d, based on 100% co‐segregation between the SNP marker and the bc‐u ^d gene in the DDP (Soler‐Garzón, 2021a). However, this candidate gene failed to show the same association with bc‐u ^d resistance in the SnAP panel examined in this study. To address this discrepancy for the marker‐gene association between panels, we reexamined the genomic sequence for the bc‐u region (Pv05: 36,001,025–36,249,945 bases) in DDP accession GN 1140, using updated NGS technologies.

By employing long‐read (PacBio) sequencing, we obtained detailed genomic information for the GN 1140 genotype, which revealed an 84‐bp repetitive in‐frame insertion within Phvul.005G125100, which is also the candidate gene for bc‐u ^r (formerly bc‐4) (Soler‐Garzón et al., 2021b). NGS long‐read technologies have proven helpful in characterizing genetic variations, even within repetitive regions, despite the challenges posed by high indel error rates in repeated regions (Narzisi & Schatz, 2015). The Phvul.005G125100 gene encodes a Vps4 AAA+ ATPase ESCRT protein or Vps4 gene for short.

The Vps4 gene is associated with virus resistance in other crops. Previous studies in Cucumis sativus and Cucumis melo infected with potyviruses led to the discovery of a resistance hotspot associated with the Vps4 gene, indicating its potential as a candidate gene for resistance to zucchini yellow mosaic virus and papaya ringspot virus (Monnot et al., 2022). Additionally, in melon, a Vps4 protein was identified as a susceptibility factor to watermelon mosaic virus, underscoring the importance of the Vps4 gene in potyvirus interaction (Agaoua et al., 2022). Barajas et al. (2014) propose that Vps4, along with the ESCRT machinery, plays a supportive role in building nuclease‐insensitive viral replicase complexes (VRCs), facilitating the evasion of host antiviral defenses and the protection of viral RNA. Moreover, the role of Vps4 and the ESCRT machinery might extend to other positive‐sense RNA viruses in plants and animals, which rely on membrane deformation and spherule formation for VRC assembly. In addition to bc‐u ^r, a second Vps4 gene Phvul.011G092700 on Pv11 in common bean, underlies another gene bc‐2, which confers resistance to BCMV and BCMNV.

An indel marker named IND_05_36225873 was developed to detect the repetitive 84‐bp in‐frame insertion. The IND_05_36225873 marker co‐segregated entirely (100%) with the presence of bc‐u ^d as determined by phenotypic reaction to NL‐3 across both DDP and SnAP diversity panels. Results of the allelism tests in two F₂ populations (FR 266/Beryl and Ulysses/Beryl), phenotyped with NL‐3 and genotyped with IND_05_36225873 and Pvmit‐1_T_G SNP markers, also support that bc‐u ^d and bc‐u ^r are alleles conditioned by different mutations in the same Vps4 gene Phvul.005G125100. Presence or absence of the bc‐u ^d and bc‐u ^r alleles, as detected by respective indel and SNP markers, was as expected across the 12 HG differential cultivars.

The role of the Vps4 protein as a candidate gene for resistance to different potyviruses in crops is now expanded by the bc‐u ^d and bc‐u ^r alleles, which interact with other genes like bc‐1 (RLKs) and bc‐2 (Vps4) to condition resistance to BCMV and BCMNV. For instance, bc‐u ^d interacts with bc‐1 (e.g., HG‐3) to provide resistance against all BCMV and BCMNV PGs except PG‐IV, PG‐VI, and PG‐VII, whereas the bc‐u ^d + bc‐2 combination (e.g., HG‐6) is only susceptible to PG‐VII. The bc‐u ^r allele interacts with bc‐2 (e.g., HG 4) to provide resistance against all BCMV PGs except PG‐V, but it does not interact with bc‐1. These findings provide important insights into the bc‐u gene and its association with resistance to BCMV/BCMNV, highlighting the significance of the identified insertion mutation and its potential as a reliable marker for detecting the bc‐u ^d allele.

AUTHOR CONTRIBUTIONS

Alvaro Soler‐Garzón: Conceptualization; formal analysis; investigation; methodology; writing—original draft; writing—review and editing. Phillip E. McClean: Data curation; formal analysis; methodology; writing—review and editing. Phillip N. Miklas: Conceptualization; formal analysis; investigation; writing—original draft; writing—review and editing.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

Supporting information

Tables S1, S2, and S3 with genotype and phenotype data.

File S1 with sequences from the IND_05_36225873 amplification product.

Soler‐Garzón, A. , McClean, P. E. , & Miklas, P. N. (2024). The alleles bc‐u ^d and bc‐u ^r (previously bc‐4 gene), representing coding mutations within Vps4 AAA+ ATPase ESCRT protein, interact with other genes to condition resistance to BCMV and BCMNV in common bean. The Plant Genome, 17, e20421. 10.1002/tpg2.20421

DATA AVAILABILITY STATEMENT

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request. Some of the data are presented in the Supporting Information.