Selection by Genetic Expression Profiles of Desi and Kabuli Chickpea (Cicer arietinum L.) Genotypes Tolerant to High Temperature Stress

Chickpea (Cicer arietinum L.) is the third most important grain legume worldwide. Developing countries are responsible for 96 % of its production, being India the leading producer; Mexico ranks in ninth place and produces two types of chickpea: white or kabuli and forage or desi type. The first one is produced in the northwest region for export purposes. The desi type is used mainly in the central-western region (Guanajuato, Jalisco and Michoacán) as livestock feed (Merga and Haji, 2019). High temperature is one of the main abiotic factors that reduces grain yield. Studies have found that temperatures above 30°C induce the plants to mature rapidly. Plant adaptation strategies to heat stress depend on climatic and soil conditions, but mainly on the plant genetic makeup. The increase in production of thermal shock protein (HSP) occurs when plants experience an abrupt or gradual increase in temperature (Devasirvatham and Tan, 2018). High temperatures produce protein unfolding that leads to the formation of protein aggregates and precipitates. Several HSPs function as chaperones and participate in the correct folding of other proteins and keep their conformation stable during the stress (Parankusam et al., 2010). Heat Stress Transcription Factors (HSFs) regulate the expression of HSPs by recognizing the heat stress elements (HSE, 5’GAAnnTTC-3’). There are five major classes of HSPs: HSP100, HSP90, HSP70, HSP60 and small heat shock proteins (sHSPs) (Iba, 2002; Guo et al., 2016). HSP101, HSP70, HSP17.7 and HSP17.6 are now known to protect plant cells from programmed cell death induced by stress at high temperatures (Rikhvanov et al., 2007). HSP21 protects the transcription of plastid-encoded RNA polymerase-dependent in chloroplast under heat stress (Zhong L. et al., 2013). Recently, Lin et al., (2018) described that the expression of miR160 is induced by heat stress and its activity regulates the expression of HSPs.
       
Biotechnological tools represent an opportunity to help reduce the losses generated by high temperatures throughout the identification of genes transcribed as a response to this stress, as well as the determination of their function (Gaur et al., 2018). In this research, we studied genes related to the heat stress tolerance and measured their expression levels by RT-PCR, results allowed us to identify kabuli and desi chickpea genotypes that potentially can tolerate high temperatures.

Chickpea genotypes
 
Thirteen kabuli and sixteen desi genotypes were tested (Table 2). The desi type included ICC 4958, a drought-tolerant genotype and for which its complete genome has been sequenced (Parween et al., 2015). The ICC and ICCV materials come from the germplasm bank of ICRISAT and were included due to outstanding agronomic characteristics. The kabuli materials included fixed lines from the breeding Program of the National Research Institute for Forestry, Agricultural and Livestock (INIFAP).
 

 
Gene selection for expression studies
 
We searched the chickpea genome of genotype ICC 4958 (Jain et al., 2013) for sequences of HSPs. Sixty-five genes were identified for which we designed specific primers with the Primer-BLAST algorithm (Ye et al., 2012) and the company Sigma-Aldrich carried out their synthesis. With the aid of final point PCR we identified and used 35 sequences which produced only one amplicon (Table 1).
 

 
Treatments
 
Fifty seeds of each material were sown and ten morphologically uniform plantlets were selected and divided into two groups, five to be kept under non-stress conditions (26°C) and five under a high-temperature treatment during the vegetative stage, phase characterized by the appearance of secondary branches. When plants reached 32 days after sowing, the second group was kept for two hours in a MRC® growth chamber with the temperature set at 40°C.
 
Expression profiles
 
To determine the level in the expression of HSPs, a bulk of leaf tissue of each genotype under the tested conditions was used. From each bulk total RNA was extracted and purified following the protocol of Logemann et al., (1987), subsequently their concentration was standardized to 1,000 ng μL^-1. Single-stranded cDNAs were synthesized by reverse transcription using the Super SMART^™ PCR cDNA Synthesis Kit, following the conditions suggested by the manufacturer. To determine the gene expression in response to both temperature treatments, the baseline level in samples from each treatment was normalized by amplifying the 26S gene by RT-PCR according to the methodology of Montero-Tavera et al., (2017). The reactions, amplification program and documentation of results were carried out as Montero-Tavera et al., (2017).
 
Densitometric analysis of expression levels
 
The expression levels generated by RT-PCR were measured using the TotalLabQuant software. The relative expression of each gene was defined in relationship to the expression of the constitutive gene 26S given it a value of 1. The selection criterion of induced genes was based on a completely random ANOVA analysis by calculating the minimum significant difference, which results of 0.1.

Identification of putative heat-tolerant genotypes
 
Based on the level of expression and in order to discriminate the genotypes, cluster analyses were carried out in both chickpea types. The relative expression values from all genes were used in a heatmap analysis with the heatmap.2 algorithm of the R statistical package gplots v3.0.1.1 (Warnes et al., 2015); distances were calculated with the Ward and Shakespeare (1998) coefficient. Furthermore, we calculated a modified Tolerance Index (TI) of Rosielle and Hamblin (1981) using as input the expression levels. This index represents the relationship of the expression under stress and non-stress conditions, which was normalized in each test considering the expression of the constitutive 26S gene. TI was calculated as follows:

 
                                 (TI) = Σ(a₁-b₁/c)/n

 
Where
a₁ = expression of gene 1 under normal temperature.
b₁ = expression of gene 1 under heat stress.
c = constitutive gene expression.
n = number of genes.

RT-PCR expression profiles
 
Fig 1 shows the expression profiles of all genotypes exposed to both temperature treatments. The tolerant desi induced on average 30.5 genes, the moderately tolerant 23.8 and the susceptible 11.5; whereas in the kabuli type induced genes were 28.4, 20.7 and 10.5, respectively (Table 2). Ten genes were identified whose expression profile is useful for the selection of putative heat-tolerant genotypes in both types of chickpea (CaHS3, CaHS5, CaHS19, CaHS22, CaHS26, CaHS37, CaHS38, CaHS41, CaHS56 and CaHS58); in addition to these, 8 exclusive genes were identified for the selection of kabuli genotypes (CaHS1, CaHS10, CaHS12, CaHS15, CaHS21, CaHS24, CaHS46 and CaHS53) and 6 exclusive for the selection of desi genotypes (CaHS14, CaHS16, CaHS43, CaHS48, CaHS59 and CaHS63).
 

 
Among the genes with a higher level of expression under stress in both types of chickpea, the ones that code for 70 kDa (HSP70) and 22 kDa (HSP22) proteins stand out. The expression of HSPs under normal conditions account for 5% of the total intracellular proteins; however, under stress conditions, they can increase up to 15% (Fragkostefanakis et al., 2015). This agrees with our results and confirms that it is possible to determine expression profiles associated with heat stress response, which could be used in a selection program for the development of genotypes tolerant to this stress.
 
Desi genotypes induced a higher number of genes. We observed in all genotypes the induction of CaHS22 and CaHS37 genes, corresponding to a non-yet described and a 22.7 kDa HSP, respectively. Genes CaHS3, CaHS4 and CaHS5 showed similar transcriptional behavior in all the desi genotypes, probably because they share the same domain: a HSP of 20 kDa. Genes CaHS2 and CaHS18 were induced in most of the genotypes under the stress condition. The induced expression of the CaHS41 gene, which contains a HSP20 domain, was observed in all kabuli genotypes. The response of the rest of the genes was similar to that shown by kabuli; however, the expression of some genes was different between chickpea types: CaHS9 had a higher induction in the desi and CaHS53 in the kabuli.

Overall, the heat stress condition induced the transcription of the studied genes; desi genotypes showed higher levels of expression (Fig 1 and 3), as well as a higher number of induced genes compared to those in the kabuli genotypes. The latter could be explained by the fact that desi materials, being more resilient and less domesticated than kabuli types, have maintained greater genetic diversity and diverse heat tolerance mechanisms. Thus, although the differences observed in the expression profiles are clear and allow us to classify each chickpea genotype into tolerance classes by plotting the first versus second discriminant and on the same manner classifies desi and kabuli types (Fig 2), it is also important to establish that these differences, although significant, do not imply the absence of tolerance mechanisms in the kabuli type, since we also identified five heat stress-tolerant genotypes in this type (Fig 3).
 

 

 
Selection of desi and kabuli chickpea genotypes tolerant to heat stress
 
Previous research suggests that genotypes that induce the expression of a higher number of minor effect genes are those with a greater potential for tolerance to heat stress (Barnabás et al., 2008; Porch and Hall, 2013). The cluster analysis (Fig 3) grouped the materials according to the total number of induced genes and this coincides with the tolerance reported in the field under heat conditions during the seed filling stage (>32°C); for desi and kabuli chickpea three groups were obtained: 1, genotypes with high potential tolerance to heat stress; 2, moderately tolerant genotypes; and 3, susceptible genotypes (Table 2, Fig 3). Regarding the desi genotypes, these outstanding were ICC 10259, ICC 13020, ICC 4958 and Annigeri; moderately tolerant genotypes were ICC 96029, ICC 5780, ICC 6671, ICC 3287, ICC 1282 and El Patrón; susceptible genotypes were ICC 2173, line 123 and ICC 1882. Some of these materials have been reported as tolerant to heat stress due to their physiological and phenological responses; Krishnamurthy et al., (2011) reported ICC 4958 as tolerant, its response mechanism is early flowering and high seed yield. Cultivar Annigeri has also been reported as tolerant due to its high cell membrane stability against heat stress (Srinivasan et al., 1996). The kabuli genotypes were grouped as follows: group 1, Mazocahui, ICCV 2, Blanco Sinaloa 92, Tequi Blanco 95, Combo-743 and CUGA 08-1210; group 2, HOGA 021, JAMU 96, HOGA 067; group 3, HOGA 2004-20-6, CUGA 08-751, Costa 2004 and Blancoson (Fig 3). ICCV 2 has also been reported as tolerant because of its physiological characteristics, showing early flowering as the main response against heat stress (Kumar and Abbo, 2001).
       
The Tolerance Index (TI) based on the adding of individual gene expression levels (Table 2) describes the chickpea genotypes in a similar way as the profiles of expression, since their values almost coincided with the classification of tolerant, moderately tolerant and susceptible. Therefore, it is proposed that both strategies could be used as complementary criteria for the identification of chickpea genotypes tolerant to high temperatures.

The methodology used in this research allowed for the classification of kabuli and desi chickpea genotypes into tolerant, moderately tolerant and susceptible. Twenty-four genes were identified that induced their transcription under heat stress, ten of which can be useful to study both types of chickpea, eight were exclusive of kabuli and six of desi type.
       
Desi genotypes with potential heat tolerance are ICC 10259, ICC 13020, ICC 4958 and Annigeri; and the kabuli Mazocahui, ICCV 2, Blanco Sinaloa 92, Tequi Blanco 95, Combo-743 and CUGA 08-1210. Results suggests that desi chickpea has higher heat tolerance potential than kabuli; however, the methodology also allowed the identification of tolerant kabuli genotypes. We will integrate this information into the local chickpea improvement program for the development of new varieties tolerant to heat stress.