Identification of Sclerotinia stem rot resistance quantitative trait loci in a chickpea (Cicer arietinum) recombinant inbred line population

Keywords: chickpea, disease resistance, Fabaceae, legume, polygenic disease resistance, quantitative trait locus analysis, Sclerotinia stem rot, Sclerotinia white mold.

Chickpea (Cicer arietinum L.) is a self-pollinated diploid (2n = 2x = 16) legume crop with a genome size of approximately 738 Mb (Varshney et al. 2013). Chickpea is produced in over 50 countries, including India, Australia, USA, Canada, Turkey and Ethiopia, and is third in the world among pulse crops in production, behind dry bean (Phaseolus vulgaris L.) and field pea (Pisum sativum L.) (Merga and Haji 2019). Together with other grain legume crops such as soybean (Glycine max L.), lupin (Lupinus spp.), and lentil (Lens culinaris L.), chickpea contributes a significant amount of protein to a plant-based diet, particularly in low-income countries (Bohra et al. 2014). Chickpea is also a valuable source of micronutrients such as phosphorous, calcium, magnesium, iron and zinc (Jukanti et al. 2012). Additionally, through symbiosis with rhizobacteria, chickpea plants are able to fix around 80% of their nitrogen requirement (Gaur et al. 2012).

Chickpea is produced under diverse agro-ecological conditions, and achieving stable yields is constrained by various abiotic and biotic stresses (Abbo et al. 2003; Jha et al. 2014). Currently, chickpea is produced on approximately 14 million ha, with an average production of 760 kg/ha globally (FAOSTAT 2019). Research has shown that a chickpea crop that is free from biotic and abiotic stresses can produce up to 3500 kg/ha (Merga and Haji 2019). Among the biotic constraints affecting chickpea are diseases such as Fusarium wilt, Ascochyta blight, Botrytis grey mould and Sclerotinia stem rot (SSR), caused by Fusarium oxysporum f. sp. ciceris, Ascochyta rabiei, Botrytis cinerea, and Sclerotinia sclerotiorum, respectively (Knights and Hobson 2016). Increasing resistance to diseases would both enhance and stabilise yields in chickpea production throughout the world. In addition, the identification of molecular markers associated with resistance may help to speed up the breeding process.

In Australia, SSR is a destructive chickpea disease that can cause up to 100% yield loss under conducive conditions (Pulse Australia 2020). Fuhlbohm et al. (2003) reported the first incidence of SSR in chickpea in eastern Australia. Since then, the pathogen S. sclerotiorum has emerged as a threat in all Australian chickpea growing regions due to its remarkably diverse hosts that include cultivated crops and weeds (Boland and Hall 1994). The cultivation of chickpea as a break crop in rotation with canola (Brassica napus L.), an important S. sclerotiorum host, increases the incidence of disease outbreaks in chickpea crops. In Australia, SSR can be controlled in canola with timely fungicide applications, but there are no fungicides registered for controlling SSR in chickpea (Pulse Australia 2020). Further, the S. sclerotiorum resting structures (sclerotia) can survive in the soil for over 7 years (Brooks et al. 2018; Lane et al. 2019). Management of SSR in chickpea requires observing cultural practices such as crop rotation, which can sometimes be ineffective due to S. sclerotiorum’s broad host range (Boland and Hall 1994). Therefore, exploring the feasibility of breeding for disease resistance to S. sclerotiorum infection should be explored.

No sources of resistance to S. sclerotiorum have been identified in chickpea. However, lines with partial resistance to various chickpea diseases have been identified and were successfully used in chickpea breeding to develop new resistant cultivars. QTL mapping of an intra-specific recombinant inbred line (RIL) population identified major and minor QTLs responsible for both Fusarium wilt and Ascochyta blight resistance (Garg et al. 2018). Single dominant and recessive genes controlling Ascochyta blight have been reported in chickpea (Tewari and Pandey 1986; Dey and Singh 1993; Li et al. 2017). Another study involving intraspecific crosses of desi and kabuli chickpea cultivars identified three QTLs responsible for Botrytis grey mould (Anuradha et al. 2011) caused by B. cinerea, a closely related pathogen of S. sclerotiorum.

Cultivated chickpea has a high morphological diversity but narrow genetic variation for trait improvement (Udupa et al. 1993; Abbo et al. 2003); therefore, research on whether SSR resistance alleles exist in this narrow gene pool is needed. Recently, a RIL population derived from Australian cultivars PBA HatTrick and Kyabra was employed to map QTLs associated with resistance to the root-lesion nematode Pratylenchus thornei (Khoo et al. 2021). The parents of this population differ in their responses to S. sclerotiorum infection, with PBA HatTrick having partial stem resistance (Mwape et al. 2021a, 2021b). In the research reported here, the F₇ generation of that population was used to investigate the response of 200 F₇ RILs to S. sclerotiorum inoculation and to map QTLs associated with SSR resistance.

The Australian desi chickpea (Cicer arietinum L.) cultivars PBA HatTrick and Kyabra were obtained from the Australian Grains Genebank (AGG, Horsham, Victoria, Australia). An additional 18 Australian chickpea varieties and 29 breeding lines were obtained from Dr Kristy Hobson of the Chickpea Breeding Australia (CBA) programme at the New South Wales Department of Primary Industries (NSW-DPI, Tamworth, NSW, Australia). In addition, an F₇ recombinant inbred line (RIL) population (n = 200) derived from cultivars PBA HatTrick and Kyabra (Khoo et al. 2021) was used.

All seeds were sown in 4 L pots filled with all-purpose potting mix (UWA plant biology mix, Richgro, Perth, WA, Australia). Plants were watered regularly for optimal growth, and at 4 weeks, they were fertilised with Nitrophoska Perfect fertiliser (Incitec Pivot fertilisers, Victoria, Australia). The experiment to evaluate the 49 chickpea lines was conducted in a hoop house environment under natural light and an average temperature of 24°C day/18°C night at the field trial area, Curtin University, Bentley, WA, Australia (32°0′19.272″S, 115°53′38.144″E) between June 2019 and August 2019. The F₇ RIL population, along with their parental lines PBA HatTrick and Kyabra, were screened for SSR response under a controlled greenhouse environment, under natural light and an average temperature of 24°C day/18°C night, at the Shenton Park field station of the University of Western Australia (31°57′2.2″S, 115°47′52.3″E) from September to November 2019.

The isolate CU8.20 is a highly aggressive isolate of S. sclerotiorum collected from B. napus fields in Western Australia and has been tested and found to be pathogenic to B. napus and chickpea (Denton-Giles et al. 2018; Mwape et al. 2021a). Single sterile sclerotia were dissected and germinated on a 9 mm Petri dish containing 39 g/L potato dextrose agar (PDA) (Becton Dickinson, NJ, USA) and incubated at 20°C for 5–7 days in the dark to produce inoculum from actively growing culture. The cultures were further sub-cultured by using a sterile cork borer and forceps to transfer 5 mm agar plugs from the original plates onto fresh media for 2 days at 20°C to source actively growing mycelia for plant inoculation.

Stem inoculation was conducted on 8-week-old plants following a stem inoculation assay previously described in B. napus by Denton-Giles et al. (2018) and adapted for chickpeas as described by Mwape et al. (2021a). Briefly, a 5 mm PDA plug with S. sclerotiorum mycelium from the leading edge of the culture was cut using a sterile cork borer, and a sterile metal spatula was used to transfer the plug onto Parafilm. The plug was placed upright on a strip of Parafilm and wrapped around the middle of the main stem of an individual plant, with the mycelium making direct contact with the chickpea stem. Plants were phenotyped by measuring stem lesion length development over time. The stem lesion length was measured at 3, 7, 10, 14, 17, and 21 days after inoculation (dai) using a ruler (Supplementary Fig. S1). At the end of the assessments, the stem lesion length data obtained were used to calculate the area under the disease progress curve (AUDPC) (Jeger and Viljanen-Rollinson 2001) for each line evaluated.

Randomised complete block designs (RCBD) were applied for both the hoop house evaluation of the 49 chickpea lines and the glasshouse evaluation of the RIL population to assess the susceptibility of all the chickpea genotypes to S. sclerotiorum. Negative controls with PDA-only agar plugs were included for all experiments. The experiment assessing 49 chickpea lines for S. sclerotiorum resistance was designed with three replicates per line, and the experiment evaluating the RIL population with four replicates per line.

For both the evaluation of the 49 chickpea lines and the RIL population, linear mixed models (LMM) were fitted using ASReml-R (Butler et al. 2018) to examine spatial variations, including local autocorrelations, global trends and extraneous variations. The cultivar was fitted as fixed effects (Smith et al. 2005). The blocking structures of the experiments were fitted as random effects. Spatial trends and residual variances with auto-regressive correlations at first-order for rows and columns were examined and fitted when the global trends and autocorrelations were significant. Likelihood ratio tests were used for random effects, and conditional Wald tests (Kenward and Roger 1997) were used for fixed effects. Residual diagnostics were performed to examine the validity of the model assumptions of normality and homogeneity of variance. For each of the fitted models, the empirical best unbiased linear estimations (eBLUEs) were produced. The R package AsremlPlus (Brien 2021) was used to compute the least significant difference (l.s.d. with α = 0.05) values.

Genotypic data of the RIL population has been described by Khoo et al. (2021). Briefly, leaf tissue was sampled from the parents and individual F₆ plants. DNA was isolated and subjected to genotyping-by-sequencing analysis by Diversity Arrays Technology Pty. Ltd. (Bruce, ACT, Australia) using its chickpea DArTseq (1.0) GBS platform. A linkage map was constructed using the R package ASMap (Taylor and Butler 2017). Linkage groups were assigned to chromosomes and oriented based on BLASTn analysis (Altschul et al. 1990) of GBS sequence tags against the kabuli chickpea reference genome (Version 2.6.3) (Edwards 2016).

QTL analysis was conducted using the inclusive composite interval mapping (ICIM) method, which is implemented in the integrated software for QTL mapping (QTL IciMapping ver. 4.1) available at http://www.isbreeding.net/ (Meng et al. 2015). The QTL mapping was conducted using the functionality of inclusive composite interval mapping of additive and dominant QTL (ICIM-ADD) (Li et al. 2007, Zhang et al. 2008). The stepwise regression was performed to identify the most significant markers and marker-pairs at a significance level of 0.001 and a scanning step of 1 cM. The threshold LOD (logarithm of the odds) score to declare significant QTL at a chromosome-wise type I error rate of 0.05 (Churchill and Doerge 1994) were determined by performing 1000 permutations.

The CDC Frontier reference genome ver. 2.6.3; http://www.cicer.info/databases.php/downloads/kabuli2.6.3rawdata.zip (Edwards 2016) was interrogated by filtering out the genes between the flanking markers of each QTL to identify candidate genes within the same intervals as the estimated QTL positions. Subsequently, the coding sequences (CDSs) were translated into amino acid sequences, and the Pfam database (http://pfam.xfam.org/) queried for putative domains to infer function (Bateman et al. 2002; Finn et al. 2014).

Gene expression data from a previous study generated from the moderately resistant (MR: PBA HatTrick) and susceptible (S: Kyabra) parents (Mwape et al. 2021b) was queried to identify differentially expressed genes in the QTL regions. Briefly, the data was previously generated by stem inoculation of 6-week-old plants with S. sclerotiorum isolate CU8.20 at 0 (control), 6, 12, 24, 48, and 72 hours post-inoculation with three biological replicates for each of the treatments. RNA extracted from stem segments was sequenced by Novogene (Beijing, China) on an Illumina HiSeq 2500 platform. The sequenced reads were aligned to the reference chickpea genome (Edwards 2016) and gene expression analysis was conducted using the Limma package in R ver. 4.0.2. A false discovery rate (FDR) cut-off of 0.05 was applied, and a log₂ fold change cut-off of ≥1 to indicate upregulation and ≤−1 to indicate downregulation. The RNAseq data can be accessed as NCBI sequence read archive under BioProject ID: PRJNA687280.

A set of 20 varieties and 29 breeding lines were evaluated for their responses to inoculation with S. sclerotiorum isolate CU8.20 to determine if there is any resistance to S. sclerotiorum in the Australian chickpea breeding programme. Lesion length measurements over time were used to calculate the AUDPC for each cultivar/breeding line. The 49 lines showed significant differences (P < 0.05, l.s.d._0.05 = 445.3) in disease responses, with PBA HatTrick showing the lowest susceptibility with a mean AUDPC of 2169 and Moti showing the highest susceptibility with a mean AUDPC of 5221 (Fig. 1; Supplementary Table S1).

Fig. 1.  The mean area under the disease progress curve (AUDPC) scores for Australian chickpea cultivars and breeding lines following inoculation with an aggressive S. sclerotiorum isolate. The vertical bar represents the least significant difference (l.s.d. with α = 0.05) value across all genotypes (y-axis). The light bars are Australian chickpea cultivars and breeding lines. Indicated in dark bar is the partially resistant variety PBA HatTrick and in the white bar, the susceptible variety Kyabra that are the parents of the recombinant inbred line population used in this study.

The cultivars PBA HatTrick and Kyabra differed in their response to S. sclerotiorum isolate CU8.20, with mean AUDPC values of 2169 and 4281, respectively (Fig. 1). Therefore, a RIL population derived from these parents was inoculated with the same isolate to investigate the underlying genetic control of the partial resistance observed in PBA HatTrick. The Shapiro–Wilk test of normality (Shapiro and Wilk 1965; Royston 1982) for the AUDPC of the RIL population indicated that the data were approximately normally distributed with the Shapiro–Wilk statistic of 0.9875 (P > 0.05) (Fig. 2). The results showed that the AUDPC values were different for PBA HatTrick (3329) and Kyabra (3724) and a broader range of AUDPC variation among the RILs: AUDPC range of 2489–4609 (Fig. 2; Table S1). The mean AUDPC values across each individual RIL and the parents showed a continuous trait distribution with significant differences (l.s.d._0.05 = 885.3) between the parents and among the RILs (Table S2).

Fig. 2.  Phenotypic distribution of the area under the disease progress curve (AUDPC) after S. sclerotiorum inoculation of 200 RIL population and the parents. The x-axis shows the mean AUDPC, and the y-axis shows the RIL population and their parents. The white and black arrowheads indicate AUDPC scores for the highly susceptible parent (Kyabra) and moderately resistant parent (PBA HatTrick), respectively.

QTL analysis to detect loci contributing to the observed variation in SSR resistance in the RIL population was conducted using the mean AUDPC values. In the present study, four QTLs, with phenotypic variation explained (PVE) ranging from 4.2–15.8%, were detected and mapped on three chromosomes (Table 1; Fig. 3): Ca4 (qSSR4-1, qSSR4-2), Ca6 (qSSR6-1) and Ca7 (qSSR7-1). The most significant of these was qSSR4.1 (PVE = 15.8%, LOD = 10.6) between markers 11 062 500|F|0–36:G > A-36:G > A and 8 822 765|F|0–8:C > T-8:C > T (Fig. 3).

Table 1.  List of quantitative trait loci (QTL) associated with resistance response to a highly aggressive S. sclerotiorum isolate in PBA HatTrick × Kyabra chickpea recombinant inbred line population (n = 200).

Fig. 3.  Genetic positions of the QTLs associated with Sclerotinia stem rot resistance on chromosomes Ca4, Ca6 and Ca7. Resistance is expressed using the area under the disease progress curve (AUDPC) derived from the stem lesion length measurements for 200 F₇ individuals of the PBA HatTrick × Kyabra population. On the right side of the chromosome are the markers and their logarithm of the odds (LOD); on the left side are their corresponding positions in centimorgans (cM). The names of the major QTLs, their flanking markers and confidence intervals are depicted in blue. The blue dashed line indicates the significance threshold with a LOD score of 2.5.

To determine which genes underlying the QTLs might be involved in the resistance response to S. sclerotiorum infection, the physical regions in the chickpea cultivar CDC Frontier reference genome were assessed. A total of 52, 102, 622 and, 351 genes in the intervals between the flanking markers for QTLs qSSR4-1, qSSR4-2, qSSR6-1, and qSSR7-1 were identified, respectively (Table S3).

Genes involved in the plant hormone signal transduction pathway, including auxin-induced protein (LOC101514996), three ethylene transcription factors (LOC101502435, LOC101502737 and LOC101504146), and signal recognition particle protein (LOC101512174), were identified in the region for qSSR4-1 (Tables S3, S4). Other genes located in this region related to disease resistance were a WAT1-related protein (LOC101504468), an F-box/LRR-repeat protein 17 (LOC101514669), a calmodulin-like protein (LOC101491221), which is involved in plant-pathogen interactions, a glutathione reductase (LOC101514119) involved in glutathione metabolism, and a beta-carotene isomerase (LOC101489176) involved in carotenoid biosynthesis (Table S3).

In the region of qSSR4-2, genes involved in the biosynthesis of secondary metabolites, purine metabolism, metabolic pathways, biosynthesis of antibiotics, phenylpropanoid biosynthesis, oxidoreductase and plant-pathogen interaction were identified (Table S4). Genes related to disease-resistance pathways identified in this region included ethylene-responsive transcription factor (LOC101505675), calcium-dependent protein kinase 4-like (LOC101502238), cellulose synthase-like protein (LOC101514030, LOC101514359, and LOC101506417) peroxidase 5-like (LOC101492647), aminoacylase-1 (LOC101489624), uricase-2 isozyme 1-like (LOC101500535), WAT1-related protein (LOC101497530), and serine/threonine-protein kinase (LOC101499994) (Table S3).

The highest number of genes was identified in the qSSR6-1 region. These genes are involved in pathways including plant hormone signal transduction, biosynthesis of antibiotics, biosynthesis of secondary metabolites, phenylalanine metabolism, ABC transporters, and plant–pathogen interaction pathways (Table S4). Genes involved in plant hormone signal transduction included serine/threonine-protein kinase (LOC101489210, LOC101489210, LOC101489533, and LOC101494601), signal recognition particle subunit (LOC101498515, and LOC101504832), two-component response regulator (LOC101509325 and LOC101497765), putative ETHYLENE INSENSITIVE 3-like 4 protein (LOC101500668) and histidine kinase 3-like (LOC101506212). Genes involved in plant-pathogen interaction included calcium-dependent protein kinase (LOC101493107) and respiratory burst oxidase (LOC101491892) (Table S3).

Genes involved in plant resistance, including plant hormone signal transduction, biosynthesis of antibiotics, plant–pathogen interaction, phenylpropanoid biosynthesis and metabolism and oxidative phosphorylation, were identified in the region of qSSR7-1 (Table S4). Five genes that play a role in plant hormone transduction included two-component response regulators ARR2 (LOC101510188) and protein TIFY 3B (LOC101502388), which were located in the qSSR7-1 region. Plant–pathogen interaction pathway-related genes including pto-interacting protein (LOC101511806) and squidulin (LOC101501000), defence response pathway (ABA-responsive protein ABR18-like: LOC101511589 and LOC101511270) and those involved in the biosynthesis of antibiotics (dihydrolipoyl dehydrogenase, alpha-aminoadipic semi-aldehyde synthase, ATP-citrate synthase beta chain protein 2-like and 1-deoxy-d-xylulose-5-phosphate synthase) were identified in the region of qSSR7-1.

A previously generated transcriptome dataset (Mwape et al. 2021b) was used to investigate the differential expression of candidate genes underlying the QTLs in the parents of the RIL population. A time course 0–72 h post-inoculation (hpi) with S. sclerotiorum isolate CU8.20 was generated for both the moderately resistant (MR) variety PBA HatTrick and the susceptible (S) variety Kyabra. Of the 1127 genes across the four QTLs, 120 genes showed differential expression patterns between non-inoculated (time 0) and inoculated samples (time points 6–72 hpi) for both parents (Figs 4, 5, 6; Table S5). There were 10, 8, 47 and 55 differentially expressed genes in the qSSR4-1, qSSR4-2, qSSR6-1 and qSSR7-1 regions, respectively.

Fig. 4.  Heat maps showing the patterns of expression of all the differentially expressed genes identified in the regions of QTLs, qSSR4-1 and qSSR4-2. Positive LogFC values (shown in green) represent upregulation relative to expression at the time of inoculation, LogFC values of 0 (shown in black) represent no significant change in expression and negative LogFC values (shown in red) represent downregulation of expression at 6, 12, 24, 48 and 72 hpi for moderately resistant (MR) and susceptible (S) chickpea lines. The genes with different patterns of expression between parents are marked with asterisks. The vertical axis represents the genes, and the horizontal axis represents the chickpea line and time points.

Of the genes located in qSSR6-1, sucrose synthase (LOC101494314), serine/threonine-protein kinase (LOC101511372 and LOC101503660), monothiol glutaredoxin-S2-like (LOC101503006 and LOC101503330), mitochondrial thiamine pyrophosphate carrier-like (LOC101495703), F-box protein (LOC101497457), chitotriosidase-1-like (LOC101501002), basic 7S globulin-like (LOC101500045), alpha-aminoadipic semialdehyde synthase (LOC101500267), alkaline/neutral invertase A, mitochondrial (LOC101491074) and ABC transporter (LOC101511590) were differentially expressed in the MR line only. A pathogenicity related protein (LOC101512575) located in qSSR6-1 was upregulated earlier (24 hpi) in the MR line compared to the S line (72 hpi) (Fig. 5).

Fig. 5.  Heat maps showing the patterns of expression of all the differentially expressed genes identified in the region of QTL qSSR6-1. Positive LogFC values (shown in green) represent upregulation relative to expression at the time of inoculation, LogFC values of 0 (shown in black) represent no significant change in expression and negative LogFC values (shown in red) represent downregulation of expression at 6, 12, 24, 48 and 72 hpi for moderately resistant (MR) and susceptible (S) chickpea lines. The genes with different patterns of expression between parents are marked with asterisks. The vertical axis represents the genes, and the horizontal axis represents the chickpea line and time points.

Fig. 6.  Heat maps showing the patterns of expression of all the differentially expressed genes identified in the region of QTL qSSR7-1. Positive LogFC values (shown in green) represent upregulation relative to expression at the time of inoculation, LogFC values of 0 (shown in black) represent no significant change in expression and negative LogFC values (shown in red) represent downregulation of expression at 6, 12, 24, 48 and 72 hpi for moderately resistant (MR) and susceptible (S) chickpea lines. The genes with different patterns of expression between parents are marked with asterisks. The vertical axis represents the genes, and the horizontal axis represents the chickpea line and time points.

The ethylene-responsive transcription factor (LOC101502737) was identified in the qSSR4-1 region and was expressed earlier in the MR parent than in the S parent (Fig. 4). The peroxidase 5-like (LOC101492647), Armadillo repeat-containing protein 6 (LOC101513382) and WAT1-related protein (LOC101497530) were upregulated while glycerol-3-phosphate dehydrogenase (LOC101495780) downregulated in the MR line only (Fig. 4).

Among the genes differentially expressed in qSSR7-1 were a leucine-rich repeat receptor-like kinase (LOC101503887), probable WRKY transcription factor 11 (LOC101513839), pathogen-associated molecular pattern-induced protein A70 (LOC101492689), alpha-trehalose-phosphate synthase (LOC101507748), calcium-binding protein CML45 (LOC101505709), homeobox-leucine zipper protein (LOC101489128), probable serine/threonine-protein kinase (LOC101514185), a probable WRKY transcription factor 11 (LOC101513839) and a caffeoyl shikimate esterase (LOC101510635) (Fig. 6). A probable F-box protein (LOC101501001) located in the qSSR7-1 region was differentially expressed at 24–72 hpi in the MR line and 24 hpi in the S line (Fig. 6).

This study investigated the inheritance of resistance to S. sclerotiorum in a bi-parental mapping population derived from Australian varieties PBA HatTrick and Kyabra, following the evaluation of 49 varieties and 200 breeding lines. Evaluation of the 49 chickpea lines identified variety PBA HatTrick as having the lowest mean AUDPC values, whereas variety Kyabra was one of the more susceptible varieties and had a significantly different AUDPC value from PBA HatTrick. As such, the bi-parental RIL population was employed to dissect the genetic inheritance of the partial resistance identified in PBA HatTrick. Our findings indicate that resistance to S. sclerotiorum in chickpea is a complex quantitative trait and is affected by several genes with small effects, which is consistent with previous studies in B. napus (Yin et al. 2010), soybean (Arahana et al. 2001), common bean (Ender and Kelly 2005), and sunflower (Helianthus spp.) (Yue et al. 2008). Breeding varieties that are resistant to SSR is important for crops such as chickpea, which is a highly valued source of protein (Kottapalli et al. 2009). Thus, it is important to identify SSR resistance-related genes through QTL mapping from the current chickpea germplasm. This finding provides direct genetic resources for genetic improvement of SSR resistance and the knowledge required for developing effective strategies for SSR resistance breeding.

To date, complete resistance to S. sclerotiorum has not been identified in crop hosts. However, partial resistance has been reported in B. napus (Denton-Giles et al. 2018), soybean (Kim et al. 1999), sunflower, dry bean (Vuong et al. 2004) and chickpea (Mwape et al. 2021a). Partial resistance to S. sclerotiorum was found in chickpea line PBA HatTrick, while the Kyabra line showed high susceptibility to SSR compared to other Australian cultivars and breeding lines (Fig. 1). Similar findings of partial resistance to S. sclerotiorum in the PBA HatTrick line and high susceptibility in the Kyabra line compared to a subset of wild chickpea germplasm were reported previously by Mwape et al. (2021a). The ability of PBA HatTrick to maintain durable partial resistance in the field remains to be tested.

A stem inoculation assay measures the host–pathogen interaction under a consistent favourable environment but does not measure escape mechanisms such as early flowering, early maturity, and canopy size. Here, we used a highly reproducible stem inoculation technique that has previously been used for SSR resistance screening of canola and chickpea under controlled environment conditions (Denton-Giles et al. 2018; Mwape et al. 2021a). The frequency distribution of RILs for SSR on length (calculated as AUDPC) depicted a distribution for a continuous trait indicating that genetic control of resistance to S. sclerotiorum may be complex. There were significant differences in AUDPC among RILs (l.s.d._0.05 = 885.3), with 32% of phenotyped lines showing higher resistance to S. sclerotiorum than the moderately resistant parent PBA HatTrick (Fig. 2).

In recent years, extensive efforts have been made using identified partial resistance to detect the loci controlling resistance to S. sclerotiorum in soybean (Arahana et al. 2001; McCaghey et al. 2017) canola (Qasim et al. 2020), pea (Ashtari Mahini et al. 2020) peanut (Liang et al. 2021) and dry bean (Miklas 2007), an indication of the importance of understanding the genetic basis of SSR resistance. In the present study, we mapped four QTLs: two on chromosome Ca4 (qSSR4-1 and qSSR4-2), one on chromosome Ca6 (qSSR6-1) and one on chromosome Ca7 (qSSR7-1). Each of these explained between 3.5 and 14.2% of the phenotypic variation, with the favourable alleles contributed by both parents (Fig. 3; Table 1). The detection of some QTLs with favourable alleles from PBA HatTrick and some with favourable alleles from Kyabra is consistent with the observation that some lines had more extreme phenotypes than the parents (Fig. 2). This is the first QTL report for resistance to S. sclerotiorum in chickpea to the best of our knowledge.

The earlier mapped resistance loci in chickpea include those for resistance to Ascochyta blight, Fusarium wilt, Botrytis grey mould and root lesion nematode (Anuradha et al. 2011; Sabbavarapu et al. 2013; Deokar et al. 2019; Garg et al. 2018; Khoo et al. 2021). A chickpea genomic region on Ca4 has consistently been reported to contain QTLs for Ascochyta blight resistance (Sharma and Ghosh 2016) and Fusarium wilt (Garg et al. 2018). In the present study, QTLs with major effects (qSSR4-1 and qSSR4-2), explaining the phenotypic variance of 8.8–14.2%, were located on chromosome Ca4. The markers associated with root lesion nematode and Fusarium wilt QTLs on Ca4 are located approximately 2 Mb distal of the region of interest for qSSR4-2 (Garg et al. 2018; Khoo et al. 2021) when using BLASTn to anchor the markers on the chickpea reference genome. For those molecular markers associated with Ascochyta blight resistance with sequence information available two markers (H1G20 and TS54) were identified that were located approximately 2 and 4 Mb upstream of the qSSR4-2 region, but none of the markers linked to Ascochyta blight, Fusarium and root lesion nematodes thus reside in the Sclerotinia stem rot resistance regions (data not shown). Genes involved in pathways associated with plant disease responses such as plant-pathogen interaction, phenylpropanoid pathways and plant hormone signalling are located in these regions of Ca4.

The identified SNP markers associated with the S. sclerotiorum resistance loci could be adopted for marker-assisted selection by chickpea breeding programmes, which will allow them to retain the partial resistance that exists in the breeding programme whilst also incorporating novel sources of partial stem resistance such as those identified in a collection of wild Cicer germplasm (Mwape et al. 2021a). Investigation into the genomic regions of QTLs identified several candidate genes that were differentially expressed in response to S. sclerotiorum. With the help of the RNA sequencing data, we were able to identify some genes that may play important roles in resistance to S. sclerotiorum. For instance, a gene involved in the thiamine biosynthesis pathway (LOC101495703) was identified in the region of qSSR6-1. Thiamine is known to play a key role in enhancing anti-oxidative capacity in the plant; thus, an increase in plant thiamine increases resistance to biotic stresses (Zhao et al. 2011; Zhou et al. 2013). Thiamine metabolism was associated with modulation of the redox environment, reducing the disease progress during S. sclerotiorum infection in Arabidopsis thaliana L. (Zhou et al. 2013).

A putative ethylene-responsive factor (ERF) gene present in the qSSR4-1 region showed an early upregulation in the MR compared to the S parent. ERF transcription factors play important roles in plant development and response to biotic and abiotic stresses (Licausi et al. 2013). Overexpression of ERF genes enhanced resistance to S. sclerotiorum in broccoli (Jiang et al. 2019).

Members of the plant WRKY transcription factor family are implicated in regulating defence-related genes in response to fungal pathogens (Yang et al. 2009). A WRKY transcription factor identified in the QTL regions qSSR6-1 showed upregulation during infection at 24 hpi (Fig. 5). Other candidate genes identified in qSSR7-1 regions encode auxin response factors (ARFs) known for their signalling role during plant growth and development and have been linked to disease resistance (Yamada 1993; Li et al. 2016). In Arabidopsis, auxin signalling mutants that had defects in response to auxin showed increased susceptibility to B. cinerea and Plectosphaerella cucumerina, indicating that auxin signalling is important resistance to these necrotrophs (Llorente et al. 2008).

Plants have receptors that harbour a C-terminal LRR domain and can directly or indirectly perceive pathogen effectors to activate multiple defence signal transduction pathways that may result in a hypersensitive response to limit pathogen growth (Sagi et al. 2017). Genes with LRR domains were identified in the QTL regions of Ca4 and Ca7. Another domain is known to play a role in pathogen recognition and downstream signalling is the zinc finger protein (Carpita and Gibeaut 1993). Zinc finger protein-encoding genes were identified in the region of QTL qSSR7-1.

Peroxidases are known for their role in the resistance-related oxidative burst response in plants (Dmochowska-Boguta et al. 2013). Four peroxidase proteins were identified in the qSSR7-1 region; however, they did not show differential expression during infection at 6–72 hpi. The plant hormone abscisic acid (ABA) is known to promote resistance in some plant–pathogen interactions and susceptibility in others and is linked with SA, JA and ET signalling to affect pathogen resistance (Wang et al. 2012). Two ABA response proteins were identified in the regions of qSSR7-1 (Table S3) and were shown to be upregulated in both lines at 12–72 hpi. Three Arabidopsis mutants with defects in ABA signalling showed a complete loss of resistance to S. sclerotiorum (Perchepied et al. 2010). This suggests that ABA signalling may be involved in partial chickpea defence against S. sclerotiorum.

Putative candidate genes with defence functional categories related to pathogenesis-related (PR) genes were identified at qSSR7-1. Pathogen-associated molecular pattern-induced proteins A70 and pathogenesis-related proteins (PR1 and PR2) were identified in the qSSR7-1 region (Table S2). PR proteins produce glycosidic fragments, which weaken and decompose fungal cell walls containing glucans, chitin and proteins (Ali et al. 2018). PR-1 genes involved in signalling and plant defences were identified within regions of QTLs qSSR6.1 (Table S2).

The markers closely linked with SSR resistance QTLs identified in this study could facilitate identifying the genes that contribute to the partial resistance phenotype and may be used to retain these through pyramiding in one genotype. The QTLs identified in this study can facilitate marker-assisted breeding for SSR resistance.

Supplementary material is available online.

The raw data used in this study for RNA sequence analysis was obtained from NCBI sequence read archive deposit under BioProject ID: PRJNA687280.

The authors declare that they have no competing interests.

The Australian Grains Research and Development Corporation (GRDC) supported the development of the population and linkage map under grant UA00157 and the research on Sclerotinia stem rot under grants CUR00023 and CUR00024. A Curtin University RTP Scholarship and CSIRO Research Plus Top-up scholarship were awarded to VWM.

VWM, LGK, TEN, MCD and KC designed the project. KK and DEM provided the recombinant inbred lines, genotypic data, linkage map and guidance on QTL mapping and interpretation of QTL results. VWM, YK, LGK, TEN, MCD phenotyped the F₇ population, chickpea varieties and breeding lines. VWM and KC performed the phenotype and QTL data analysis. VWM performed the RNA-seq data analysis and wrote the first manuscript draft. All authors read and approved the final manuscript.