Introduction

Cowpea (Vigna unguiculata (L.) Walp.) is one of the most important legume crops globally. It is widely grown in tropical and subtropical regions, especially in Africa and Asia. Four groups or subspecies of cultivated cowpea are generally recognized, including grain cowpea (ssp. unguiculata), yardlong bean (ssp. sesquipedalis (L.) Verdc.), biflora/catjang bean (ssp. cylindrica (L.) Verdc.) and textilis (ssp. textilis (L.) Verdc.) (Singh 2005). Among them, grain cowpea and yardlong bean are the most popular for cultivation. These crops have highly contrasting morphological and physiological characteristics. Grain cowpea (also known as black-eye pea) is usually cultivated for its dry seed, although its young pods are also consumed as a vegetable. Yardlong bean (also known as asparagus bean) is cultivated chiefly for its long-tender-young pods. Pods of grain cowpea are about 15–35 cm in length, while pods of yardlong bean are about 30–120 cm in length (Singh 2005). Grain cowpea is principally cultivated in Africa, whereas yardlong bean is mainly cultivated in Asia and is a chief source of protein for people in developing counties. Young pods of yardlong bean are consumed as a vegetable in several ways in both raw and cooked forms and thus provide dietary minerals, vitamins and fibers. The young pods can be harvested several times beginning at about 45–50 days after sowing (Kongjaimun et al. 2012). In general, compared with grain cowpea, yardlong bean needs wetter and cooler environments to thrive. However, grain cowpea appears to be more resistant to biotic and abiotic stresses than yardlong bean.

Fungal disease is one of the main biotic stresses affecting the growth, development and yielding of cowpea (Singh 2005). Cercospora leaf spot (CLS) disease caused by Cercospora canescens Ellis and Martin or Pseudocercospora cruenta (Sacc.) Deighton (Crous and Braun 2003; Deighton 1976) is one of the important fungal diseases of cowpea. CLS disease widely occurs in the tropical and subtropical regions, especially in the rainy season when high moisture levels and warm temperatures prevail. The leaf spot symptoms of C. canescens are circular, while those of P. cruenta are angular. C. canescens is comparatively a weaker parasite than P. cruenta, but it has a wider host range under tropical climates (Fery et al. 1976). The fungi start to affect cowpea plants from the beginning of the flowering stage onwards. The disease can cause yield loss from 35 to 40% in susceptible varieties (Schneider et al. 1976; Fery et al. 1976). In yardlong bean, the spot symptom also appears on young pods, resulting in them becoming unmarketable. Yardlong bean is more susceptible to CLS disease than grain cowpea because field cultivation of yardlong bean requires higher soil moisture content (more irrigation) and the large bushy canopy creates more shading (Duangsong et al. 2016). Thus, enhancement of resistance to CLS is a major objective in yardlong bean breeding programs.

To date there have been some reports published on the genetics of resistance to CLS disease in grain cowpea. Genetic studies revealed that the resistance to CLS disease caused by P. cruenta is controlled by a single dominant gene (Fery et al. 1976; Castro et al. 2003) or single recessive gene (Fery et al. 1976) or oligogenes or polygenes (Booker and Umaharan 2008) depending on resistance sources. The single-dominant gene and the single-recessive gene for the resistance are designated as Cls1 and cls2, respectively. These two genes are not linked (Fery and Dukes 1977).

Recently, we reported that resistance to CLS disease caused by C. canescens and P. cruenta in the grain cowpea breeding line IT90K-59-120 is controlled by a single gene based on segregation and quantitative analyses (Duangsong et al. 2016; 2018). In addition, a major QTL for the resistance, qCLS9.1, in IT90K-59-120 was identified (Duangsong et al. 2016). qCLS9.1 is located in a region of 3.5 cM on linkage group 9 (LG9; corresponding to chromosome 10 of the cowpea reference sequence) between simple sequence repeat (SSR) markers CEDG070 and CEDG304. It accounted for up to about 90% and 30% of the disease variation caused by C. canescens and P. cruenta, respectively (Duangsong et al. 2016). In addition, qCLS9.1 is possibly the same locus as ALS10.1, which is a major QTL conferring resistance to angular leaf spot disease caused by P. griseola (Sacc.) Crous and U. Braun in common bean [Phaseolus vulgaris (L.)] (Oblessuc et al. 2015). Therefore, qCLS9.1 for the resistance to CLS is interesting. In this study, we reported fine mapping and identification of candidate genes for qCLS9.1 conferring CLS resistance in IT90K-59-120.

Materials and methods

Plant materials and population development

As the qCLS9.1 was localized to the interval between SSR markers CEDG070 and CEDG304 on LG9 using an F2:3 population developed from the cross CSR12906 × IT90K-59-120 (Duangsong et al. 2016), an F2:3 family segregating for the QTL qCLS9.1 was selected based on the marker genotype of CEDG070 being 0.5 cM from the qCLS9.1 and disease scores of its F2 ancestor plant (#80). This F2 plant showed a heterozygous genotype at the marker CEDG070, but showed homozygous genotypes at all the other markers on the LG9. Twenty-three F3 plants derived from self-pollination of the #80 F2 plant were grown and genotyped with CEDG070 in which two plants showing heterozygosis at this marker were selected and self-pollinated to generate an F3:4 population segregating at qCLS9.1 as a mapping population. The original F2 population used to identify qCLS9.1 (Duangsong et al. 2016) was also used in this study to confirm the QTL(s) identified by fine mapping.

DNA extraction and valuation for disease resistance

The F3:4 population was used to evaluate resistance to CLS disease. In total, 699 F3:4 individuals and their parental lines (IT90K-59-120 and CSR1906) were grown during August to October 2017 (rainy season) under field conditions in Paktho District, Ratchaburi Province, Thailand, which is a hot spot for CLS disease in yardlong bean in Thailand (Duangsong et al. 2016). The space between plants and rows was 50 and 50 cm, respectively. Each row contained 80 individuals. The susceptible parent CSR1906 was grown around the experimental field as a natural source of CLS inoculums. Fifteen days after planting (DAP), young leaves from each plant were collected and extracted for genomic DNA following a CTAB method described by Lodhi et al. (1994).

At 40 and 50 DAP, the F3:4 and parental plants were inoculated with C. canescens as per the procedures described by Chankaew et al. (2011), which were the same as those used by Duangsong et al. (2016). Before inoculation, conidia taken from infected plants were examined using a light microscope to confirm that the disease symptoms were the result of infection by C. canescens. At 20 and 30 days after inoculation, CLS disease symptoms on leaves of each tested plant were scored. The scoring was carried out by three trained staff using the disease rating scale of 1–5 where 1 = no visual disease infection, 2 = 1–25% leaf area infected, 3 = 26–50% leaf area infected, 4 = 51–75% leaf area infected and 5 = 76–100% leaf area infected.

Development of new SSR markers for fine mapping qCLS9.1

The QTL qCLS9.1 region was delimited by SSR markers PvM13 and CEDG304 (Duangsong et al. 2016). The location of these markers on the cowpea reference genome was determined by conducting a BLASTN search of the sequences of these markers against the cowpea reference genome in the Phytozome database (https://phytozome.jgi.doe.gov/pz/portal.html#!info?alias=Org_Vunguiculata_er) (Muñoz-Amatriaín et al. 2017; Lonardi et al. 2019). Once the physical genome locations of the PvM13 and CEDG304 were identified, DNA sequences between these two locations were downloaded and then searched for SSR sequences using SSRIT software (Temnykh et al. 2001). Primers for the SSRs were designed using Primer3 software (Untergasser et al. 2012) (Supplementary Table S1) and used to screen for polymorphisms between CSR12906 and IT90K-59-120. SSR analysis was carried out as per Duangsong et al. (2016) with minor modifications. Briefly, PCR was carried out in a total volume of 10 µL containing 5 ng of DNA template, 1 × Taq buffer, 2 mM MgCl2, 0.2 mM dNTPs, 1 U Taq DNA polymerase (Thermo Fisher Scientific) and 2.5 µM each of forward and reverse primers. Amplification was performed in a GeneAmp PCR 9700 System thermocycler (Applied Biosystems) programmed as follows: 94 °C for 2 min followed by 35 cycles of 94 °C for 30 s, 55 °C for 30 s, 72 °C for 1 min, and 72 °C for 10 min. PCR products were electrophoresed on 5% polyacrylamide gel electrophoresis and visualized by silver staining. Nine markers showing polymorphism between the parents were selected and used to genotype the F3:4 population.

Linkage map construction and QTL analysis

Genotypes of nine polymorphic markers were used to construct a genetic linkage map using QTL IciMapping 4.1 software (Meng et al. 2015). The markers were grouped with a minimum logarithm of the odds (LOD) value of 3.0. Markers on the linkage group were ordered using the REcombination Counting and ORDering (RECORD) algorithm (van Os et al. 2005). The marker orders were rippled using the Sum of Adjacent Recombination Frequencies (SARF) function (Falk 1989). Recombination frequencies were converted into genetic distance using the Kosambi mapping function (Kosambi 1944).

Location of the QTL qCLS9.1 was determined by inclusive composite interval mapping (ICIM) (Li et al. 2007) using the same software as for the linkage analysis. ICIM was performed at every 0.1 cM. Significant LOD score threshold for the QTL was determined by running a 10,000-permutation test at p = 0.001. As the LOD graph indicated the possibility of the existence of two tightly linked QTLs for the resistance (see “Results”), a two-linked QTL model was tested using MultiQTL software (http://www.multiQTL.com) as described by Peng et al. (2003). A single QTL model and two-linked QTL model were tested by comparing hypotheses with a single QTL present (H1) with no QTL (H0) and two-linked QTLs (H2) to the H0, respectively. Then, the H2 was compared to the H1 using a 1000-permutation test to ensure whether the two-linked QTL model fit the data better than the single QTL model.

Confirming QTL(s) for the resistance

The QTLs identified by fine mapping in the F3:4 population were confirmed in the original F2 population of 190 individuals segregating for qCLS9.1 (Duangsong et al. 2016). The F2 population was genotyped with the nine polymorphic markers used in the F3:4 population. Marker analysis, linkage map construction and QTL detection were the same as described above.

Sequencing of candidate genes for qCLS9.1

After QTL analysis, the reference genome location of the QTL regions was inspected to find candidate genes for qCLS9.1. Annotated genes located between the markers flanking the QTL(s) were considered as candidate gene(s) for CLS resistance. Primers were designed to the downloaded reference sequence to amplify the full transcript sequence and upstream sequence of the candidate gene(s) using Primer3 (Supplementary Table S2). cDNA (see “Expression analysis of the candidate genes” section) and genomic DNA of CSR12906 and IT90K-59-120 were amplified using the primers designed for each candidate gene. PCR was carried out in a total volume of 10 µL containing 5 ng of DNA template, 1 × Taq buffer, 2 mM MgCl2, 0.2 mM dNTPs, 1 U Taq DNA polymerase (Thermo Fisher Scientific) and 0.5 µM each of forward and reward primers. Amplification was performed in a GeneAmp PCR 9700 System thermocycler (Applied Biosystems) programmed as follows: 94 °C for 2 min followed by 35 cycles of 94 °C for 30 s, 55 °C for 1 min, 72 °C for 1 min, and 72 °C for 10 min. PCR products were run on 1.5% agarose gel electrophoresis to confirm the single DNA fragment amplified. The PCR products were sequenced by Sanger sequencing using an ABI 3730xl DNA analyzer with Big Dye Terminator v3.1 kit (Applied Biosystems, USA). Sequences were edited and assembled using Sequencher v5.4.6 (Gene Codes Corporation). The sequences of the CSR12906, IT90K-59-120 and reference sequence were aligned to identify nucleotide polymorphism(s) that may cause amino acid change. The cDNA sequences of CSR12906, IT90K-59-120 were translated into protein sequences and aligned to find amino acid polymorphisms. The effect of amino acid change in the protein was predicted by using Protein Variation Effect Analyzer (PROVEAN) software (Choi et al. 2012). The significant threshold PROVEAN value was set at 2.5. If the PROVEAN value is above this threshold, it is predicted to have a deleterious effect on protein function.

Development of allele-specific markers for candidate genes

The SNPs between parental lines identified in the candidate gene(s) were used to develop cleaved amplified polymorphic sequence (CAPS) (Konieczny and Ausubel 1993) or derived cleaved amplified polymorphic sequence (dCAPS) markers (Neff et al. 1998). Primers for CAPS and dCAPS were designed to amplify SNP regions using Primer3 and dCAPS Finder 2.0 (Neff et al. 2002), respectively (Supplementary Table S4). PCR amplification was conducted in the parental lines using the same procedure as described above. PCR products were run on 1.5% agarose gel electrophoresis to confirm the single DNA fragment amplified. The PCR products were digested with FastDigest RsaI restriction enzyme (Thermo Scientific), which recognizes GT^AC, at 37 °C for 15 min. After cutting, the PCR products were run on 3.0% agarose gel electrophoresis and stained with ethidium bromide for visualization.

Expression analysis of the candidate genes

CSR12906 and IT90K-59-120 were grown in field conditions at Kasetsart University, Kamphaeng Saen Campus, Nakhon Pathom, Thailand, during September to October 2018 (rainy season). Forty days after planting, the plants were inoculated with spore suspensions of Cercospora canescens. At 0, 2, 4, 6, 12, 24 and 48 h post-inoculation (hpi), leaves were collected and extracted for total RNA following the protocol described by Laksana and Chanprame (2015). The RNA was treated with DNA-free™ DNaseI (Ambion) and converted to cDNA using a RevertAid H Minus First Strand cDNA Synthesis Kit (Thermo Scientific) according to the manufacturer’s instructions. cDNA concentration was quantified using a ND-1000 Spectrophotometer (NanoDrop Technologies Inc.). The cDNA was subjected to gene expression analysis by quantitative real-time polymerase chain reaction (qRT-PCR). Primers for qRT-PCR of candidate genes and ACTIN (Vigun07g025200) were designed using Primer3 (Supplementary Table S3). qRT-PCR was performed using ViiA 7 Real-Time PCR System (Applied Biosystems). Three biological and technical replicates were conducted for each sample. Reaction mixtures contained water, 1 × Master mix of Fast SYBR™ Green Master Mix (Thermo Fisher Scientific), 5 µM of forward primer, 5 µM of reverse primer and 50 ng cDNA. Thermocycler conditions included initial denaturation at 95 °C for 20 s, followed by 40 cycles at 95 °C for 3 s and 60 °C for 30 s. After 40 cycles, a melting curve was generated by slowly increasing (0.5 °C per 1 s) the temperature from 60 to 95 °C while the fluorescence was measured. Fluorescent data were acquired during each extension phase. Expression levels of Vigun10g019300 and Vigun10g019400 were calculated based on the ΔCT method by using reference gene (ACTIN) (Livak and Schmittgen 2001).

Results

Response to CLS disease in the F3:4 population

Parental line and F3:4 plant responses to CLS disease were scored at 70 DAP. The yardlong bean CSR12906 was highly susceptible to the disease with a disease score of 4.50, while cowpea IT90K-59-120 was highly resistant with a disease score of 1.20. The disease scores of the F3:4 plants showed continuous distribution and ranged from 1.00 to 5.00 with a mean of 3.72 (Fig. 1).

Fig. 1
figure 1

Frequency distribution of disease score for Cercospora leaf spot disease caused by Cercospora canescens in a cowpea F3:4 population of 699 individuals from the cross CSR12906 × IT90K-59-120

Full size image

Fine mapping for qCLS9.1 and confirmation of the QTLs for CLS resistance

The locations of PvM13 and CEDG304, in which SSR markers delimited the QTL qCLS9.1 region in the previous study (Duangsong et al. 2016), were 34.03 and 22.51 Mbp (11.52 Mbp) on chromosome 10 of the cowpea reference genome, respectively. Among 70 newly developed SSR markers residing in this region, 47 were able to amplify DNA of CSR12906 and IT90K-59-120 and 17 were polymorphic markers between these two cowpea lines. A linkage map constructed based on the eight new polymorphic markers and CEDG070 for the F3:4 population spanned 25.84 cM in length. All the marker orders completely agreed with their position on the cowpea reference sequence.

QTL analysis in the F3:4 population revealed that two tightly linked QTLs, named qCLS9.1A and qCLS9.1B, are involved in CLS resistance (Table 1 and Fig. 2). qCLS9.1A was located at 15.30 cM between markers VU10g3-11 and VU10g3-12, while qCLS9.1B was located at 20.10 cM between markers CEDG070 and VU10g3-14. qCLS9.1A showed a LOD score of 44.76 and accounted for 23.45% of the disease score variation in the F3:4 population. It possessed an additive effect of 0.53 and a dominant effect of 0.01. qCLS9.1B showed a LOD score of 60.76 and explained 33.72% of the disease score variation in the F3:4 population. It had additive and dominant effects of 0.65 and 0.03, respectively. At both qCLS9.1A and qCLS9.1B, allele(s) from IT90K-59-120 decreased the disease score (elevated the resistance). The existence of the tightly linked QTLs qCLS9.1A and qCLS9.1B was also supported by the two-linked QTL model created using MultiQTL (Supplementary Table S5). Genotypic analysis revealed that plants possessing IT90K-59-120 homozygous alleles at both qCLS9.1A and qCLS9.1B are highly resistant with disease scores < 2.0, those possessing IT90K-59-120 homozygous alleles at qCLS9.1A or qCLS9.1B are moderately resistant with disease scores of about 2.5–3.0, and those possessing CSR12906 homozygous alleles at both qCLS9.1A and qCLS9.1B are highly susceptible with disease scores > 3.50 (Fig. 2b).

Table 1 Positions and effects of QTLs controlling Cercospora leaf spot disease caused by Cercospora canescens confirmed in a cowpea F2 population of 190 individuals from the cross CSR12906 × IT90K-59-120
Full size table
Fig. 2
figure 2

Fine mapping of qCLS9.1A and qCLS9.1B for resistance to Cercospora leaf spot (CLS) disease caused by C. canescens using the F3:4 population of the cross CSR12906 × IT90K-59-120. The LOD graph of QTLs (a). Dotted line parallel to the X-axis represents the LOD threshold for the QTLs. Marker genotypes and disease scores of CSR12906, IT90K-59-120 and 16 selected F3:4 individuals carrying crossovers around qCLS9.1A and qCLS9.1B (b). The individuals carrying genotype B on both markers VU10g3-11 and CEDG070 are classified as highly resistant to CLS disease. Those individuals carrying genotype B on marker VU10g3-11 or CEDG070 are classified as moderately resistant to CLS disease. Any individuals carrying genotype A or H on both markers VU10g3-11 and CEDG070 are classified as susceptible to CLS disease. Physical locations of candidate genes and markers at QTLs qCLS9.1A and qCLS9.1B on the cowpea reference genome (c)

Full size image

The QTLs identified by fine mapping using the F3:4 population were re-analyzed using a linkage map constructed with the same markers for the original F2 population reported by Duangsong et al. (2016) in which disease scoring was conducted at 60 and 70 DAP. In the present study, QTL analysis consistently identified one QTL, qCLS9.1A, for the resistance at both 60 and 70 DAP. qCLS9.1A was located between markers PvM127 and VU10g3-9 (Table 2 and Supplementary Fig. S7). qCLS9.1A accounted for more than 80% of the disease score variation in the F2:3 population. The QTL possessed an additive effect of 0.94–1.44 and a dominant effect of nearly 0.

Table 2 Positions and effects of QTLs controlling Cercospora leaf spot disease caused by Cercospora canescens in a cowpea F3:4 population of 699 individuals from the cross CSR12906 × IT90K-59-120
Full size table

Identification of candidate genes for qCLS9.1A and qCLS9.1B

Based on the cowpea reference genome version 1.1, the physical region of marker interval VU10g3-9–VU10g3-14 covering qCLS9.1A and qCLS9.1B in the fine mapping population is only 60.6 Kbp (Fig. 2c). There are two annotated genes in this region: Vigun10g019300 and Vigun10g019400. Vigun10g019300 is located on the interval between markers VU10g3-11 and VU10g3-12, while Vigun10g019400 is between markers VU10g3-12 and CEDG070. In fact, VU10g3-11 is a part of Vigun10g019300, while CEDG070 is a part of Vigun10g019400. The distance between Vigun10g019300 and Vigun10g019400 is only 10.3 Kbp. Vigun10g019300 encodes NAD-dependent malic enzyme 1 (NAD-ME1), while Vigun10g019400 encodes for dynamin-related protein 1C (DRP1C). Vigun10g019400 is a homolog to AtDRP1E [known as enhanced disease resistance 3 (EDR3)]. These genes are considered as candidate genes for qCLS9.1A and qCLS9.1B, respectively, explaining the CLS resistance in IT90K-59-120.

Nucleotide and amino acid variations in Vigun10g019300 and Vigun10g019400

Transcript and upstream sequences of Vigun10g019300 and Vigun10g019400 were determined for CSR12906 and IT90K-59-120. The sequence alignment of Vigun10g019300 transcripts revealed 12 SNPs between CSR12906 and IT90K-59-120. Upstream sequence alignment of Vigun10g019300 showed one SNP in the 5′ untranslated region (UTR) (Supplementary Fig. S1). SNPs at positions 896, 1464 and 1631 in the CDS, corresponding to positions 18,642, 30,562 and 35,574 of the open reading frame (ORF), respectively, (Fig. 3a) cause the amino acid change at positions 299 aa (N → S), 488 aa (S → N) and 544 aa (S → N) of the CSR12906 NAD-ME1 protein, respectively (Fig. 4). cDNA sequence alignment of Vigun10g019400 revealed one SNP between CSR12906 and IT90K-59-120 (Fig. 3b). This SNP was located at position 1608 (corresponding to position 6146 of the ORF) but did not cause amino acid change (Supplementary Fig. S5). However, upstream sequence alignment of this gene revealed one SNP in the 5′ UTR and two SNPs in the TATA boxes (Fig. 5 and Supplementary Fig. S3).

Fig. 3
figure 3

Single nucleotide polymorphisms (SNPs) among CSR12906, IT90K-59-120 and cowpea reference genome sequence in Vigun10g019300 (a) and Vigun10g019400 (b). Number in parentheses under each SNP indicates its position in the open reading frame sequence. SNPs causing amino acid changes are in bold

Full size image
Fig. 4
figure 4

Alignment of amino acid sequences of NAD-dependent malic enzyme 1 encoded by Vigun10g019300 in CSR12906, IT90K-59-120 and the cowpea reference genome. Asterisk (*) indicates fully conserved residue. Period (.) indicates conservation between amino acids of weakly similar properties

Full size image
Fig. 5
figure 5

Single nucleotide polymorphisms (SNPs) in the upstream region of Vigun10g019400 in CSR12906, IT90K-59-120 and the cowpea reference genome. Portions and letters in red indicate the location of SNPs between CSR12906, IT90K-59-120 and the cowpea reference sequence. Letters and positions of TATA boxes are in bold and underlined

Full size image

The effect of amino acid changes on the function of NAD-ME1 encoded by Vigun10g019300 was predicted using PROVEAN software. Among the three amino acid changes in NAD-ME1, only the S488N change in CSR12906 possibly has a deleterious effect on the protein function. This amino acid change had a PROVEAN score of − 2.896, which is predicted to have a deleterious effect on protein function.

Differences in gene expression patterns of the candidate genes in parent lines

Gene expression of Vigun10g019300 and Vigun10g019400 in IT90K-59-120 and CSR12906 was measured by qRT-PCR at 0, 2, 4, 6, 12, 24 and 48 hpi. The expression level of Vigun10g019300 between IT90K-59-120 and CSR12906 was significantly different at 4, 6 and 12 hpi (Fig. 6a). IT90K-59-120 showed higher expression than CSR12906 at 4 and 6 hpi, and the difference was pronounced at 6 hpi, whereas CSR12906 showed higher expression than IT90K-59-120 at 2, 12 and 24 hpi. The expression level of Vigun10g019400 of IT90K-59-120 was significantly higher than that of CSR12906 at 4, 6, 12 and 48 hpi. The highest expression was at 6 hpi, followed by 4 and 12 hpi, respectively (Fig. 6b).

Fig. 6
figure 6

Relative gene expression level of Vigun10g019300 (a) and Vigun10g019400 (b) in the leaves of CSR12906 and IT90K-59-120 at 0, 2, 4, 6, 12, 24 and 48 h post-inoculation (hpi) with Cercospora canescens. *, ** and *** indicate significant differences of gene expression level between CSR12906 and IT90K-59-120 by t test at p = 0.05, p = 0.01 and p = 0.001, respectively

Full size image

Development of CAPS and dCAPS markers

CAPS marker cVU19300 and dCAPS marker dVU19400 were developed to detect the CDS SNP(s) at position 1078 (corresponding to position 25,662 of the ORF (position 2,208,171 on chromosome 10)) in Vigun10g019300 and the CDS SNP at position 1608 (corresponding to position 6146 of the ORF (position 2,236,679 on chromosome 10)) in Vigun10g019400, respectively (Fig. 3). These SNPs are detected by the restriction enzyme RsaI. dVU19300 primers amplified a fragment of 255, while dVU19400 primers amplified a region of 282. Analysis of CAPS marker cVU19300 revealed polymorphism between CSR12906 and IT90K-59-120. CSR12906 had three DNA bands of sizes 153, 70 and 32 bp, while IT90K-59-120 had two DNA bands of sizes 153 and 102 bp (Fig. 7a). F3:4 individuals possessing heterozygous SNP at this marker had four DNA bands (Fig. 7a). Analysis of the dCAPS marker dVU19400 showed polymorphism between CSR12906 and IT90K-59-120. SR12906 and IT90K-59-120 showed a DNA band size of 255 bp and 282 bp, respectively (Fig. 7b). F3:4 individuals possessing a heterozygous SNP at the marker dVU19400 showed two DNA bands (Fig. 7b).

Fig. 7
figure 7

DNA banding patterns of CAPS marker dVU19300 (a) and dCAPS marker dVU19400 (b). Amplified and digested DNA fragments with the restriction enzyme RsaI were resolved by 3% agarose gel. M, 100 bp DNA ladder; P1, CSR12906; P2, IT90K-59-120; 19, F3:4 individual number 19; 21, F3:4 individual number 21

Full size image

Discussion

Previous study using qualitative and quantitative genetic analyses demonstrated that the CLS resistance in IT90K-59-120 is controlled by a single gene (Duangsong et al. 2018). Duangsong et al. (2016) showed that qCLS9.1 is a major QTL controlling CLS resistance in this cowpea. Based on the cowpea reference sequence (Muñoz-Amatriaín et al. 2017; Lonardi et al. 2019), the markers CEDG070 and CEDG304 flanking qCLS9.1 are 20.28 Mbp apart and contain 613 genes. In this study, by analyzing newly developed SSRs in a population segregating at only qCLS9.1 originating from an F2 plant in the mapping population (CSR12906 × IT90K-59-120) used by Duangsong et al. (2016), we successfully narrowed down the qCLS9.1 to a 60.6-Kbp region on chromosome 10 of cowpea (Fig. 2). In the present study, the previous qCLS9.1 locus was found to be composed of two tightly linked major QTLs, qCLS9.1A and qCLS9.1B (Fig. 2a). However, the LOD graph of the 60.6-Kbp region did not clearly indicate the existence of the two QTLs and it is also possible that only one strong QTL exists. Evaluation of many recombinants from a larger population size may be able to unambiguously resolve if one or two QTLs exist in this narrow genome region. In the 60.6-Kbp region, however, there was no sequence gap in the cowpea reference sequence and the region contains only two annotated genes: Vigun10g019300 coding for NAD-ME1 and Vigun10g019400 coding for DRP1E (Fig. 2). As both qCLS9.1A and qCLS9.1B had comparable genetic effects that contribute to the CLS resistance (Table 1), the two genes were further characterized.

As compared with IT90K-59-120, NAD-ME1 in CSR12906 possesses three amino acid changes: N299S, S488N and S544N (Figs. 3a and 4 and Supplementary Fig. S2). S488N was predicted to cause a deleterious effect on the NAD-ME1 function in CSR12906. Based on the Phytozome database, S488 locates in the NAD binding domain of NAD-ME1. Sequence alignment of NAD-ME1 proteins from cowpea, azuki bean (Vigna angularis (Ohwi) Ohwi and Ohashi), mungbean (Vigna radiata (L.) Wilczek), common bean (Phaseolus vulgaris L.), soybean (Glycine max (L.) Merr.), pigeon pea (Cajanus cajan (L.) Millsp.) and barrel medick (Medicago truncatula Gaertner) showed perfect conservation of S488 and S544 in these legumes (Supplementary Fig. S6), suggesting that S at these positions is important in the function of NAD-ME1.

In plants, NAD-MEs are found in mitochondria (Winning et al. 1994). NAD-MEs are a primary regulatory enzyme for the metabolism of malate in plant mitochondria. NAD-MEs catalyze the oxidative decarboxylation of L-malate using NAD+ as a coenzyme to generate pyruvate, CO2 and NADH. In plants, NADP-dependent malic enzymes (NADP-ME), a very similar decarboxylating enzymes to NAD-MEs, play an important role in disease resistance through their involvement in production of NADPH by NADPH oxidases for synthesis of reactive oxygen species (ROS). Apoplastic ROS bursts generated in elicited plant cells are sufficiently cytotoxic to kill invading pathogens (Legendre et al. 1993; Chi et al. 2009; Park et al. 2013). In addition, ROS act as signaling molecules that activate plant defenses against pathogen invasion (Tenhaken et al. 1995; Jabs 1999; Torres 2010). In Arabidopsis thaliana, a nadp-me2 mutant showed enhanced susceptibility toward pathogens Colletotrichum higginsianum (Voll et al. 2012), Botrytis cinerea and Pseudomonas syringae pv. tomato (Pto) DC3000 (Mhamdi and Noctor 2016). Singh et al. (2018) showed that rice blast fungus Magnaporthe oryzae secrete effector AVR-Pii to inhibit NADP-ME2-3 function in rice (Oryza sativa L.) and disrupt host immunity. Malate has been reported to participate in plant disease resistance (Bolwell et al. 2002; Finkemeier et al. 2013), although its specific function in disease resistance is not known. To date there have been no reports on the direct involvement of NAD-ME in plant disease resistance, despite it having been shown that transgenic arabidopsis overexpressing OsNAD-ME1 exhibit germination and growth advantages under salt, alkali, drought, and oxidative stress conditions, suggesting that NAD-ME1 plays an important role in plant responses to abiotic stresses (Zhou et al. 2012). However, as NAD-ME and NADP-ME function in a very similar fashion, NAD-ME may provide the plant resistance to pathogen invasion in the same way as NADP-ME. In addition, as the generation of O ·−2 , the proximal mitochondrial ROS, within mitochondria depends critically on proton motive force, NADH/NAD+ and CoQH2/CoQ (reduced CoQ/coenzyme Q) ratios and the local O2 concentration (Murphy 2009) and the NADH and NAD+ pools depend partially on NAD-ME (Schertl and Braun 2014); therefore, production of mitochondrial ROS partially depends on NAD-ME. It is worth mentioning that the rate of ROS production increases at high NADH/NAD+ ratios (Abdelwahid et al. 2007; Adam-Vizi and Chinopoulos 2006). Taken together, these results indicate that Vigun10g019300 encoding functional NAD-ME1 increases resistance to CLS in grain cowpea IT90K-59-120. Moreover, expression of Vigun10g019300 in IT90K-59-120 was higher than that in CSR12906 (Fig. 6a). Taken together, our results are the first line of evidence that show the involvement of NAD-ME in plant disease resistance.

Vigun10g019400 coding for DRP1E, a dynamin-related protein (DRP), was found to associate with the CLS resistance in IT90K-59-120. DRP1E localizes in mitochondria (Tang et al. 2006; Li et al. 2017). Dynamins and DRPs are members of a protein superfamily of GTPases that are involved in diverse membrane-related processes in prokaryotic and eukaryotic cells, including membrane fusion (such as mitochondrial fission), membrane scission, membrane protection and/or membrane stabilization (Jilly et al. 2018). Although the specific function of DRPs in plant disease resistance remains poorly understood, studies have revealed that DRPs are implicated in plant disease resistance (Tang et al. 2006; Smith et al. 2014; Leslie et al. 2016; Li et al. 2017; Wu et al. 2018). For example, induction of cell death (Tang et al. 2006; Li et al. 2017) and vesicular trafficking through the perception of pathogen-associated molecular pattern (PAMP)-triggered immunity signaling (Smith et al. 2014) and regulation of the trafficking of proteins involved in callose synthase and/or callose degradation (Leslie et al. 2016). In Arabidopsis thaliana, Tang et al. (2006) demonstrated that dynamin-related protein 1e (drp1e) enhances susceptibility to B. cinerea. Similarly, Smith et al. (2014) showed that loss of Dynamin-related Protein 2B (DRP2B) results in decreased PATHOGENESIS-RELATED 1 (PR1) mRNA levels in response to bacterial infection and thus increased susceptibility to virulent and avirulent Pto DC3000 strains. In rice, Li et al. (2017) showed that DRP1E regulates programmed cell death (PCD) via the control of cytochrome c release from mitochondria. In our case, although the predicted DRP1E proteins encoded by Vigun10g019400 in IT90K-59-120 and CSR12906 were the same (Supplementary Fig. S5), qRT-PCR analysis showed that Vigun10g019400 was expressed statistically higher in IT90K-59-120 than CSR12906 during C. canescent ingress (Fig. 6b). SNPs in TATA boxes of this gene are likely the cause of the differential expression (Fig. 5). Taken together, the above data suggest that the high expression of Vigun10g019400 appears to enhance resistance in grain cowpea IT90K-59-120, possibly by increasing production of mitochondrial ROS.

Interestingly, both NAD-ME1 and DRP1E, which were associated with CLS resistance in this study, locate in mitochondria and their functions are involved in ROS production. DRP1E controls cytochrome c release from mitochondria (Li et al. 2017). Cytochrome c is a heme-containing redox protein active in electron transfer pathways, including the respiratory electron transport chain (ETC) (Alvarez-Paggi et al. 2017; Supplementary Fig. S8). The plant mitochondrial ETC core includes classical oxidoreductase complexes (complex I to IV), cytochrome c, ubiquinone (coenzyme Q) and alternative oxidoreductases (Schertl and Braun 2014). NAD-MEs are among the dehydrogenase enzymes in the ETC complex II (succinate dehydrogenase) in the mitochondrial matrix that transfer electrons in the form of NADH to the ETC in which the NADH is re-oxidized by complex I or the internal alternative NAD(P)H dehydrogenases (Schertl and Braun 2014). Malate is the prime precursor in complex II and is oxidized by NAD-MEs to produce pyruvate that is fed into the tricarboxylic acid cycle (Krebs cycle) (Schertl and Braun 2014). It has been proposed that complex II may function as a general sensor for PCD (Grimm 2013; Lemarie et al. 2011). In the respiratory ETC, the cytochrome c shuttles electrons from complex III (cytochrome c reductase) to complex IV (cytochrome c oxidase) (Alvarez-Paggi et al. 2017). Based on our findings and this information, NAD-ME1 and DRP1E appear to play an important role in the electron transfer pathways in plant mitochondria that regulate ROS production and PCD (Supplementary Fig. S8) and contribute to CLS resistance in grain cowpea IT90K-59-120.

Because both Vigun10g019300 and Vigun10g019400, which are associated with CLS resistance, are linked in the coupling phase, selection for a highly resistant genotype would not be difficult in breeding materials. The SSR markers VU10g3-9 and VU10g3-11 located in Vigun10g019300 and CEDG070 located in Vigun10g019300 can be used with a very high degree of precision for MAS for both genes to reduce time and resources for the development of CLS-resistant cowpea cultivar(s). Apart from these SSR markers, CAPS marker dVU19300 and dCAPS marker dVU19400 developed in the present study would be useful for MAS because both markers are allele-specific, segregate in a co-dominant manner and can be detected with simple molecular laboratory facilities.