Introduction
Chickpea is one of the most important grain legumes grown globally (FAO, 2017). Globally, 14.78 Mt chickpea is harvested annually from 14.56 Mha area (FAO, 2017). It plays a crucial role in providing protein-based nutrition security to the increasing global human population – especially in the developing countries (Graham and Vance, Reference Graham and Vance2003; Jukanti et al., Reference Jukanti, Gaur, Gowda and Chibbar2012; Bohra et al., Reference Bohra, Pandey, Jha, Singh, Singh, Datta, Chaturvedi, Nadarajan and Varshney2014). Being a leguminous crop, chickpea improves soil nitrogen content by incorporating atmospheric nitrogen in association with root inhabiting active rhizobia (Graham and Vance, Reference Graham and Vance2003).
Chickpea yield is seriously challenged by a range of biotic and abiotic stresses. Among the various biotic stresses, vascular wilt and Ascochyta blight caused by Fusarium oxysporum f. sp. ciceris and Ascochyta rabiei, respectively remain the most devastating diseases, causing up to 90% yield losses (Sabbavarapu et al., Reference Sabbavarapu, Sharma, Chamarthi, Swapna, Rathore, Thudi, Gaur, Pande, Singh, Kaur and Varshney2013; Jha et al., Reference Jha, Bohra, Pandey and Parida2020). Importantly, Fusarium wilt (FW) resistance is an essential prerequisite for variety identification and release system in major chickpea growing countries such as India. To date, a total of eight FW races (race 0, 1A, 1B/C, 2, 3, 4, 5 and 6) have been recognized in chickpea (Jiménez-Gasco and Jiménez-Diaz, Reference Jiménez-Gasco and Jiménez-Diaz2003). Initial studies based on Mendelian genetics in chickpea have elucidated the genetic makeup of plant resistance against FW races 1A, 2, 3, 4 and 5 (Sharma et al., Reference Sharma, Chen and Muehlbauer2005). However, genetic resistance of chickpea for FW races 1B, 1C and 6 remains to be studied.
Trait-mapping studies in chickpea have provided a set of DNA markers linked to genomic regions controlling resistance to multiple FW races. Examples include DNA markers OPJ20600 and TR59 for Foc0 race, TA110 and H3A12 for Foc1 race, H3A12 and TA96 for Foc2 race (Cobos et al., Reference Cobos, Fernández, Rubio, Kharrat, Moreno, Gil and Millán2005), TA196 and TR19 for Foc4 (Mannur et al., Reference Mannur, Babbar, Thudi, Sabbavarapu, Roorkiwal, Yeri, Bansal, Jayalakshmi, Singh Yadav, Rathore, Chamarthi, Mallikarjuna, Gaur and Varshney2019) and TA59 for Foc5 race (Castro et al., Reference Castro, Piston, Madrid, Millan, Gil and Rubio2010; Caballo et al., Reference Caballo, Madrid, Gil, Chen, Rubio and Millan2019). Similarly, DNA markers for selection against FW race 1 (Foc1) and race 2 (Foc2) were reported (Gowda et al., Reference Gowda, Radhika, Kadoo, Mhase and Gupta2009). Other examples of FW mapping in chickpea include against races Foc3 (Sharma et al., Reference Sharma, Winter, Kahl and Muehlbauer2004; Gowda et al., Reference Gowda, Radhika, Kadoo, Mhase and Gupta2009), Foc4 (Winter et al., Reference Winter, Benko-Iseppon and Hüttel2000; Sharma et al., Reference Sharma, Winter, Kahl and Muehlbauer2004, Reference Sharma, Chen and Muehlbauer2005; Mannur et al., Reference Mannur, Babbar, Thudi, Sabbavarapu, Roorkiwal, Yeri, Bansal, Jayalakshmi, Singh Yadav, Rathore, Chamarthi, Mallikarjuna, Gaur and Varshney2019) and Foc5 (Cobos et al., Reference Cobos, Winter, Kharrat, Cubero, Gil, Millan and Rubio2009; Caballo et al., Reference Caballo, Madrid, Gil, Chen, Rubio and Millan2019). However, these studies involved the analysis of mapping populations based on two FW-responsive parents with contrasting disease reactions (Varshney et al., Reference Varshney, Mohan, Gaur, Chamarthi, Singh, Srinivasan, Swapna, Sharma, Pande, Singh and Kaur2014; Mannur et al., Reference Mannur, Babbar, Thudi, Sabbavarapu, Roorkiwal, Yeri, Bansal, Jayalakshmi, Singh Yadav, Rathore, Chamarthi, Mallikarjuna, Gaur and Varshney2019).
In recent years, association genetics has been implemented in chickpea for the identification of significant gene–trait associations for a variety of important traits (Thudi et al., Reference Thudi, Upadhyaya, Rathore, Gaur, Krishnamurthy, Roorkiwal, Nayak, Chaturvedi, Basu, Gangarao, Fikre, Kimurto, Sharma, Sheshashayee, Tobita, Kashiwagi, Ito, Killian and Varshney2014; Upadhyaya et al., Reference Upadhyaya, Bajaj, Das, Kumar, Gowda, Sharma, Tyagi and Parida2016a, Reference Upadhyaya, Bajaj, Narnoliya, Das, Kumar, Gowda, Sharma, Tyagi and Paridab; Jha et al., Reference Jha, Jha, Bohra, Parida, Kole, Thakro, Singh and Singh2018). In this context, the current study identified significant marker-trait associations (MTAs) for FW (Foc2) using simple sequence repeat (SSR)-based association mapping of 75 chickpea genotypes. The significant MTAs will pave the way for rapid and targeted breeding of chickpea cultivars with enhanced resistance to Foc2 race.
Materials and methods
A panel of FW-responsive 75 chickpea genotypes along with WR 315 (FW-resistant check) and JG 62 (FW-susceptible check) (see online Supplementary Table S1) was screened for two consecutive years 2016 and 2017 in FW (Foc2)-sick nursery at Indian Institute of Pulses Research (IIPR), Kanpur, India. The genotypes were sown in the first week of November in both years (2016 and 2017) in augmented block design arranged in five blocks, with each block containing 15 genotypes along with the checks. The genotypes were sown in 60 cm × 10 cm spacing and the row length was kept as 3 m. The susceptible check (JG 62) was planted after every two rows. Wilt incidence was recorded at three different stages i.e. pre-flowering, post flowering and post podding (from November to February). The disease incidence was calculated according the following method suggested by Sharma et al. (Reference Sharma, Ghosh, Telangre, Rathore, Saifulla, Mahalinga, Saxena and Jain2016):
$$\% \;{\rm disease}\;{\rm incidence} = \displaystyle{{{\rm Number}\;{\rm of}\;{\rm FW}\hbox{-}{\rm infected}\;{\rm plants} \times 100} \over {{\rm Total}\;{\rm number}\;{\rm of}\;{\rm plants}}}$$Based on the % disease incidence, genotypes displaying less than 10.0% incidence were classified as resistant, whereas genotypes having incidence between 10.1 and 20.0% were classified as moderately tolerant. Genotypes with 20.1–50.0% incidence were considered as moderately susceptible and genotypes showing >50.0% incidence were considered as susceptible.
DNA extraction and SSR analysis
Genomic DNA was isolated from 3 weeks old chickpea seedlings following the CTAB method (Saghai-Maroof et al., Reference Saghai-Maroof, Soliman, Jorgensen and Allard1984). A total of 125 SSRs were screened on the given set of genotypes, of which 75 SSRs yielded polymorphic fragments. The SSRs used here are reported previously by different research groups (Winter et al., Reference Winter, Pfaff, Udupa, Huttel, Sharma, Sahi, ArreguinEspinoza, Weigand, Muehlbauer and Kahl1999, Reference Winter, Benko-Iseppon and Hüttel2000; Sethy et al., Reference Sethy, Shokeen and Bhatia2003, Reference Sethy, Shokeen, Edwards and Bhatia2006; Nayak et al., Reference Nayak, Zhu, Varghese, Datta, Choi, Horres, Jüngling, Singh, Kishor, Sivaramakrishnan, Hoisington, Kahl, Winter, Cook and Varshney2010; Gaur et al., Reference Gaur, Sethy, Choudhary, Shokeen, Gupta and Bhatia2011; Gujaria et al., Reference Gujaria, Kumar, Dauthal, Dubey, Hiremath, BhanuPrakash, Farmer, Bhide, Shah, Gaur, Upadhyaya, Bhatia, Cook, May and Varshney2011; Choudhary et al., Reference Choudhary, Gaur, Gupta and Bhatia2012) and the selected SSRs span entire eight linkage groups (LGs) of chickpea (online Supplementary Table S2).
Polymerase chain reaction (PCR) analysis
The PCR protocol was followed as suggested by Jha et al. (Reference Jha, Jha, Bohra, Parida, Kole, Thakro, Singh and Singh2018) and Bohra et al. (Reference Bohra, Jha, Lamichaney, Singh, Jha, Naik, Datta, Maurya, Tiwari, Yadav and Singh2020). A reaction mixture of 10 μl volume containing 5.9 μl of sterilized distilled water, 1.00 μl template DNA (25 ng), 0.5 μl of forward and 0.5 μl of reverse primers (5 μM), 1.00 μl 10× PCR buffer (10 mM Tris-HCl, 50 mM KCl, pH 8.3), 1.00 μl dNTP mix (0.2 mM each of dATP, dGTP, dCTP and dTTP) and 0.1 μl Taq polymerase (5 U/μl) (Thermo Fisher Scientific Mumbai, India, Pvt. Ltd.) was prepared. This reaction mixture was polymerized by using G-40402 thermo cycler (G-STORM, Somerset, UK) using the touchdown PCR profile for amplification with initial denaturation at 94°C for 5 min followed by 10 cycles of touchdown 61–51°C, 30 s at 94°C, annealing for 30 s at 61°C (the annealing temperature for each cycle being reduced by 1°C per cycle) and extension for 30 s at 72°C. This was followed by 40 cycles of denaturation at 94°C for 30 s, annealing at 51°C for 30 s, elongation at 72°C for 45 s and 10 min of final extension at 72°C. Amplified fragments were resolved on 3% agarose gel using 0.5× TBE running buffer and images were analysed with Quantity one software (Bio-Rad, CA 94547, USA).
Genetic diversity and population structure analysis
Genetic diversity parameters such as number of alleles per locus (N a), gene diversity (H e) and polymorphism information content (PIC) were computed with PowerMarker v. 3.25 (Liu and Muse, Reference Liu and Muse2005). With 1000 bootstrap value, neighbourhood joining tree and factorial analysis was performed with DARwin v. 6.0.13 (Perrier and Jacquemoud-Collet, Reference Perrier and Jacquemoud-Collet2006). To determine population structure (Q) and the subpopulation (K) in the given set of genotypes, model-based analysis was conducted with STRUCTURE v. 2.3.4 (Pritchard et al., Reference Pritchard, Stephens and Donnelly2000). By applying admixture model, five independent runs were conducted with 200,000 Markov Chain Monte Carlo iterations for each K value ranging from 1 to 10 with a burn-in length of 200. In parallel, the best K value was obtained according to the ΔK method of Evanno et al. (Reference Evanno, Regnaut and Goudet2005) by processing the STRUCTURE results by STRUCTURE HARVESTER (Earl and von Holdt, Reference Earl and von Holdt2012) (http://taylor0.biology.ucla.edu).
Association mapping for the identification of significant MTAs and putative candidate genes
The wilt disease scores and the genotypic data were analysed to discover significant MTAs using generalized linear model (GLM) and mixed linear model (MLM) models in TASSEL v. 3.0 (Bradbury et al., Reference Bradbury, Zhang, Kroon, Casstevens, Ramdoss and Buckler2007; Zhang et al., Reference Zhang, Ersoz, Lai, Todhunter, Tiwari, Gore, Bradbury, Yu, Arnett and Ordovas2010). The MTAs were detected at the thresholds of P = 0.05 and at P = 0.01. Furthermore, to find the possible candidate genes corresponding to the reported MTAs and the putative proteins encoded by these, we performed BLASTn search for the associated SSRs against the reference genome sequence of CDC frontier (Varshney et al., Reference Varshney, Song, Saxena, Azam, Yu, Sharpe, Cannon, Baek, Rosen, Tar'an, Millan, Zhang, Ramsay, Iwata, Wang, Nelson, Farmer, Gaur, Soderlund, Penmetsa, Xu, Bharti, He, Winter, Zhao, Hane, Carrasquilla-Garcia, Condie, Upadhyaya, Luo, Thudi, Gowda, Singh, Lichtenzveig, Gali, Rubio, Nadarajan, Dolezel, Bansal, Xu, Edwards, Zhang, Kahl, Gil, Singh, Datta, Jackson, Wang and Cook2013). In parallel, the proteins were predicted for the corresponding sequences using InterPro (https://www.ebi.ac.uk/interpro/).
Results
Genetic diversity for FW resistance
Analysis of the FW response of the 75 chickpea genotypes during years 2016 and 2017 suggested a broad range of genetic variability. Figure S1 (see online Supplementary material) depicts the frequency distribution of disease incidence in the tested chickpea genotypes for the two consecutive years. Based on the average of disease scoring data of the two years, a total of 30 resistant (R), 34 moderately resistant (MR) and 11 susceptible (S) genotypes were obtained (online Supplementary Table S1).
SSR-based molecular diversity analysis
Assaying 75 chickpea genotypes with 75 SSRs revealed a total of 267 alleles with an average of 3.56 alleles per marker (online Supplementary Table S2). The number of alleles ranged from two to eight, while the PIC values varied between 0.17 and 0.77. Similarly, gene diversity ranged from 0.19 to 0.8 with an average value of 0.59.
As shown in Fig. 1, entire 75 genotypes were clustered into two groups according to unweighted neighbour joining. Cluster I contained 42 genotypes, while cluster II had 33 genotypes. Similarly, factorial analysis placed all the genotypes into two coordinates (online Supplementary Fig. S3).
Neighbour-joining trees of the 75 chickpea genotypes for FW resistance phenotyping.
Structure analysis
Population structure of the 75 chickpea genotypes was investigated with Bayesian approach using the STRUCTURE program. The ln P(D) as well as Evanno's ΔK values identified three genetically distinct populations (i.e. K = 3) (Fig. 2 and online Supplementary Fig. S4).
Population structure analysis of 75 chickpea genotypes.
Identification of MTAs for FW resistance
We used both GLM and MLM approaches for the detection of significant MTAs for FW (Foc2) resistance. Following GLM analysis, seven markers (NCPGR40, NCPGR180, NCPGR249, NCPGR231, NCPGR149, CESSR433 and ICCM0284) explained 7.9–45.5% of the phenotypic variance (PV) in the year 2016 (Table 1). Although MLM analysis showed significant association of three SSR markers CESSR433, ICCM0284 and NCPGR231, with PVs found in the range of 21.7–35.6% in the same year (Table 1). Figure 3 depicts the Q–Q plot for FW resistance in 2016 and 2017 based on GLM analysis. In year 2017, GLM revealed significant association with R 2 (11.7–37.5%) for seven SSR markers (NCPGR40, NCPGR180, NCPGR249, NCPGR231, NCPGR149, CESSR433 and ICCM 0284) (Table 2). However, association of the three markers CESSR433, ICCM0284 and NCPGR231 was evident through MLM analysis, with PVs extending up to 33.4% in the year 2017 (Table 2). The Q–Q plots for years 2016 and 2017 based on MLM are shown in Fig. 3.
Q–Q plots of AM for FW resistance in chickpea evaluated in 2016 and 2017 using MLM and GLM analyses.
Significant MTA for FW resistance in chickpea obtained in the year 2016 by Q GLM and MLM approaches of association mapping
*Significant at 5% level.
**Significant at 1% level.
Significant MTAs for FW resistance in chickpea obtained in the year 2017 by Q GLM and MLM approaches of association mapping
*Significant at 5% level.
**Significant at 1% level.
Candidate gene identification
The candidate genomic regions showing significant association with FW resistance were BLASTed for gene prediction against the CDC frontier genome sequence (Varshney et al., Reference Varshney, Song, Saxena, Azam, Yu, Sharpe, Cannon, Baek, Rosen, Tar'an, Millan, Zhang, Ramsay, Iwata, Wang, Nelson, Farmer, Gaur, Soderlund, Penmetsa, Xu, Bharti, He, Winter, Zhao, Hane, Carrasquilla-Garcia, Condie, Upadhyaya, Luo, Thudi, Gowda, Singh, Lichtenzveig, Gali, Rubio, Nadarajan, Dolezel, Bansal, Xu, Edwards, Zhang, Kahl, Gil, Singh, Datta, Jackson, Wang and Cook2013). As a result, seven candidate genes with putative function (see Table 3) were predicted for significant MTAs for FW (Foc2) tolerance.
Candidate genes underlying the marker intervals with their putative functions (based on InterPro)
Discussion
Marker attributes such as genome abundance, reproducibility and cost-efficiency make SSR the preferred marker system for marker-assisted breeding programmes (Semagn et al., Reference Semagn, Bjornstad and Ndjiondjop2006; Kalia et al., Reference Kalia, Manoj, Sanjay, Rohtas and Dhawan2011). In the current study, 267 alleles with an average of 3.56 alleles per marker were obtained following analysis of 75 FW-responsive chickpea genotypes with 75 SSRs. The level of heterozygosity ranged from 0.3 (NCPGR171) to 0.80 (NCPGR255) with a mean of 0.59. These results were in close agreement with earlier findings in chickpea (Upadhyaya et al., Reference Upadhyaya, Dwivedi, Baum, Varshney, Udupa, Gowda, Hoisington and Singh2008; Ghaffari et al., Reference Ghaffari, Talebi and Keshavarz2014; Hajibarat et al., Reference Hajibarat, Saidi, Hajibarat and Talebi2015; Jha et al., Reference Jha, Jha, Bohra, Parida, Kole, Thakro, Singh and Singh2018).
Identification of MTAs for traits of breeding relevance has received greater attention for accelerating crop improvement. Some noteworthy examples of MTAs identification in chickpea include mapping of drought (Thudi et al., Reference Thudi, Upadhyaya, Rathore, Gaur, Krishnamurthy, Roorkiwal, Nayak, Chaturvedi, Basu, Gangarao, Fikre, Kimurto, Sharma, Sheshashayee, Tobita, Kashiwagi, Ito, Killian and Varshney2014), heat stress (Jha et al., Reference Jha, Jha, Bohra, Parida, Kole, Thakro, Singh and Singh2018), seed weight, seed protein content (Upadhyaya et al., Reference Upadhyaya, Bajaj, Das, Kumar, Gowda, Sharma, Tyagi and Parida2016a) and grain zinc content (Upadhyaya et al., Reference Upadhyaya, Bajaj, Narnoliya, Das, Kumar, Gowda, Sharma, Tyagi and Parida2016b).
The genetic underpinnings of FW resistance were elucidated earlier in chickpea (Winter et al., Reference Winter, Benko-Iseppon and Hüttel2000; Sharma et al., Reference Sharma, Winter, Kahl and Muehlbauer2004, Reference Sharma, Chen and Muehlbauer2005; Sabbavarapu et al., Reference Sabbavarapu, Sharma, Chamarthi, Swapna, Rathore, Thudi, Gaur, Pande, Singh, Kaur and Varshney2013; Caballo et al., Reference Caballo, Madrid, Gil, Chen, Rubio and Millan2019; Mannur et al., Reference Mannur, Babbar, Thudi, Sabbavarapu, Roorkiwal, Yeri, Bansal, Jayalakshmi, Singh Yadav, Rathore, Chamarthi, Mallikarjuna, Gaur and Varshney2019), and the MTAs in these studies were identified using biparental quantitative trait locus (QTL) mapping. However, the potential of association genetics had not previously been explored to identify genomic regions underlying FW resistance. In the current study, significant association of the marker ICCM0284 with FW was detected on LG06 by both GLM and MLM approaches in both years (2016 and 2017). Earlier, Sabbavarapu et al. (Reference Sabbavarapu, Sharma, Chamarthi, Swapna, Rathore, Thudi, Gaur, Pande, Singh, Kaur and Varshney2013) recorded two major QTLs on LG06 that explained up to 18% PV for FW resistance (race 1). In the current study, another significant MTA (NCPGR40) for wilt resistance was identified on LG02 by GLM analysis in two years. A similar QTL analysis reported QTLs for early and late wilt on LG02 (Patil et al., Reference Patil, Ravikumar, Bhat and Soregaon2014). Notably, chickpea LG02 harbours resistance gene(s)/QTLs for FW races 1 (Foc1), 3 (Foc3), 4 (Foc4) and 5 (Foc5) (Ratnaparkhe et al., Reference Ratnaparkhe, Santra, Tullu and Muehlbauer1998; Tullu et al., Reference Tullu, Muehlbauer, Simon, Mayer, Kumar, Kaiser and Kraft1998; Winter et al., Reference Winter, Benko-Iseppon and Hüttel2000; Sharma et al., Reference Sharma, Winter, Kahl and Muehlbauer2004; Varshney et al., Reference Varshney, Mohan, Gaur, Chamarthi, Singh, Srinivasan, Swapna, Sharma, Pande, Singh and Kaur2014; Caballo et al., Reference Caballo, Madrid, Gil, Chen, Rubio and Millan2019). The GLM and MLM analyses also detected the marker CESSR433 on LG01, showing linkage with wilt resistance in both years 2016 and 2017. Occurrence of QTLs for FW resistance in LG01 was reported earlier in chickpea (Jingade and Ravikumar, Reference Jingade and Ravikumar2015). It is interesting to note that three SSR markers NCPGR231, CESSR433 and ICCM0284 displayed significant MTAs for FW tolerance consistently for two years in both GLM and MLM analyses. Thus, these SSR markers could be used as proxies for selecting desirable level of resistance against FW in chickpea.
Candidate genes underlying the marker intervals with their putative functions are presented in Table 3. The major genes underlying candidate genomic regions could be predicted as Ca_23618, Ca_14730, Ca_13845, Ca_24529, Ca_03351, Ca_17303 and Ca_20129. The probable function of Ca_23618 candidate gene is to code for a RING type zinc finger protein. Similarly, the candidate gene Ca_14730 encodes a putative uncharacterized protein, while the possible protein coded by Ca_20129 is a MYB-like protein. Among these, the role of the zinc finger protein domain in the plant disease response has been reported in barley (Shirasu et al., Reference Shirasu, Lahaye, Tan, Zhou, Azevedo and Schulze-Lefert1999), rice (Xu and He, Reference Xu and He2007) and Arabidopsis (Shi et al., Reference Shi, Wang, Ye, Chen, Deng, Yang, Zhang and Chan2014). Similarly, MYB protein participating in disease resistance has been demonstrated in Arabidopsis (Mengiste et al., Reference Mengiste, Chen, Salmeron and Dietrich2003), potato (Tai et al., Reference Tai, Goyer and Murphy2013) and wheat (Al-Attala et al., Reference Al-Attala, Wang, Abou-Attia, Duan and Kang2014; Shan et al., Reference Shan, Rong, Xu, Du, Liu and Zhang2016). Previously, a set of candidate genes controlling FW resistance had been identified in chickpea that included Ca_14301 encoding NB-ARC domain disease resistance protein, controlling tolerance against race Foc4 (Mannur et al., Reference Mannur, Babbar, Thudi, Sabbavarapu, Roorkiwal, Yeri, Bansal, Jayalakshmi, Singh Yadav, Rathore, Chamarthi, Mallikarjuna, Gaur and Varshney2019), and LOC101511605 encoding CBL-interacting serine/threonine-protein kinase 8-like for Foc5 (Caballo et al., Reference Caballo, Madrid, Gil, Chen, Rubio and Millan2019). Functional validation of these candidate genes and cloning of the causative gene could greatly assist precise incorporation of FW resistance into chickpea cultivars.
Conclusion
We investigated the disease response of 75 chickpea genotypes for FW using the association mapping approach. To the best of our knowledge, this is the first MTA analysis for FW (Foc2) in chickpea that analyses an FW-responsive chickpea collection with genome-wide SSRs. Based on the results of both GLM and MLM, three SSR markers NCPGR231, CESSR433 and ICCM 0284 consistently displayed significant associations with FW (Foc2) resistance across both years. However, fine mapping will be needed to precisely delineate the causative gene(s) underlying the candidate genomic regions for future research and breeding for FW resistance in chickpea.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S1479262121000228.
Acknowledgements
The authors acknowledge support from Indian Council of Agricultural Research (ICAR), India.
Author contributions
UCJ conceived the idea and wrote the MS with AB, SP and NPS. SKC provided the material; ML and PRS conducted the phenotyping; RJ and VT performed the genotyping and analysis. AB edited the manuscript. All authors have read and approved the manuscript.
Conflict of interest
The authors declare that they have no conflict of interest.
