Introduction

Soil salinity is one of the most serious environmental factors that decrease crop performance and productivity, especially under arid and semi-arid conditions (Akladious and Hanafy 2018). It was found that 6% of the world’s soil and 20% of agricultural lands are already affected by salinity (Flowers et al. 2010). According to recent reports, salinity stress costs the agricultural sector over $27.3 billion in lost revenue annually, exacerbating the food security crisis. More than 0.8 billion hectares of arable land worldwide are salt-affected (Munns and Tester 2008). Parallelly, the world’s human population is expected to reach 9.3 billion by 2050 (Flowers et al. 2010). Based on that evidence, developing salinity-tolerant crop plants is an important challenge to meeting the expected food demand.

Soil salinization significantly decreases soil water retention and mineral absorption, resulting in oxidative damage, nutrient deficits, and water insufficiency. Further, high sodium ion concentrations may damage photosynthetic leaves, causing chlorosis and premature leaf senescence (Hanin et al. 2016). Panta et al. (2014) reported that glycophytic plant crops considerably reduce average yields by 50–80% under moderate salt stress (EC 4–8 dS m−1). Plant morphology, physiology, and metabolic activity are all altered due to salt stress (Shrivastava and Kumar 2014). Therefore, understanding plant physiological and biochemical pathways is critical in breeding new barley genotypes that can grow under salt stress. Due to low sodium ion (Na+) sequestration into vacuoles, salinity stress conditions cause severe cellular toxicity resulting in premature senescence of older leaves and toxicity symptoms (i.e., chlorosis and necrosis) in mature leaves (Deinlein et al. 2014; Munns and Tester 2008). Also, Na+ content can disturb the integral potassium ion (K+) content, interfering with enzyme activity, cellular protein, and transcription factors (Sarker and Oba 2020a; Tester and Davenport 2003). Hence, the tissue K+/Na+ content ratio is often considered a basis of differential salt tolerance in various plant species (Chen et al. 2007). Moreover, Na+ could be prevented from entering the cytosol by controlling Na+ influx and efflux from the cytoplasm into the vacuoles. Hence, plants’ ability to sustain a high K+/Na+ content ratio is among the critical factors in response to salt stress (Ishikawa and Shabala 2019). Soil salinity increases the generation of active oxygen radicles, including hydrogen peroxide (H2O2), which could damage or kill the plant (Ahanger et al. 2020).

In response to external stimuli, molecular oxygen (O2) acts as an electron acceptor, producing different types of reactive oxygen species (ROS) molecules. Singlet oxygen (1O2), hydroxyl radical (OH−), superoxide (O2−), and hydrogen peroxide (H2O2), all seem to be highly oxidizing molecules that damage chlorophyll, proteins, membrane lipids, and nucleic acids, resulting in toxicity. Therefore, plants synthesize several metabolites and proteins with a protective function to protect these components (Ahmad et al. 2008). Many ROS detoxifying proteins, such as superoxide dismutase (SOD), catalase (CAT), ascorbate peroxidase (APX), and glutathione reductase (GR), as well as antioxidants like glutathione (GSH), ascorbic acid (AsA), tocopherols, and phenolic compounds, present in almost all subcellular compartments, play a key role to vanquish the deleterious effects of ROS (Siringam et al. 2011).

Due to the high genetic diversity in stress tolerance, barley is considered one of the most saline-tolerant crops and it is frequently used as a model to understand salinity adaptive responses in cereals (Mwando et al. 2020; Thabet and Alqudah 2019). This adaptability to various environments and abiotic stress conditions differs significantly among barley genotypes and developmental growth stages (Xue et al. 2009).

Genome-wide association studies (GWAS) are intensely used to discover and elucidate the genetic basis of complex traits frequently controlled by many genes (Alqudah et al. 2020; Thabet et al. 2022; Yu et al. 2018). GWAS generally achieves higher mapping resolution than conventional QTL mapping due to the higher recombination levels among linked marker loci and the traits of interest (Alqudah et al. 2020). Moreover, GWAS could accurately locate polymorphisms and the underlying genetic loci responsible for phenotypic variations, allowing gene-targeted searches (Yu et al. 2018). GWAS has been widely applied in trait-associated genetic studies in barley; for instance, single nucleotide polymorphisms (SNPs) had a significant association with antioxidant components at the vegetative stage under stress conditions (Thabet et al. 2020, 2021a, b, c). Although GWAS has identified a set of allelic variations in controlling plant salt tolerance, there is still a lack of applicable and sufficient salt-tolerant loci or genes in crop breeding. This may be due to losing elite germplasm resources or optimal conditions.

Our study detected genomic regions with putative candidate genes for morphological, physio-biochemical, and yield-related traits in response to salt stress using a core collection of 138 worldwide spring barley accessions that were genotyped using genotyping-by-sequencing (GBS). Our results indicated the natural variation in response to salt stress in barley plants and the most promising genomic regions. The findings will provide an extensive insight into the potential salinity tolerance mechanisms that underlie salt tolerance in barley during the reproductive phase. They could be used in breeding programs to develop barley varieties with improved salinity tolerance.

Methods

Germplasm and genotyping

A core set collection consisting of 138 worldwide spring barley accessions was selected out of 22,621 wild and domesticated barley accessions hosted in the German ex-situ IPK-Gatersleben GeneBank as highly diverse accessions. The collection included 54 cultivars and 84 landraces from different geographical origins. The core set was split into 63 two-rowed and 75 six-rowed barley. All the accessions were genotyped using GBS that yielded 19,276 SNPs as reported in (Milner et al. 2019). The detailed information on barley accessions is provided in Table S1.

Experimental design

A pot experiment was conducted under field conditions during the two growing seasons (2020 and 2021) at the Experimental Station of Fayoum University (29° 11′ 20.36″ N latitude, 30° 10′ 06.45″ E longitude). A diverse set of 138 barley accessions was grown under both control and salt treatment. Four seeds per each accession were sown in plastic pots (22 cm × 60 cm × 40 cm) filled with sandy loam soil. Each accession was replicated three times (three pots) for each treatment per year using a randomized complete block design (RCBD). Each pot received 5 g (17:11:10/N:P:K) fertilizer and was manually irrigated as needed. At the beginning of the reproductive phase, around 25 days after sowing (DAS), 200 mM NaCl and distilled water were added to the salt treatment and control pots, respectively. Salinity treatments were continued up to 55 DAS, after which irrigation with distilled water was continued until harvest. At the end of the treatments, leaf samples from each accession (three replicates) were collected to measure the physiology indices, ionic contents, and enzymatic and non-enzymatic antioxidants. Five phenotypic traits, such as spike length (SL), the distance between the spike’s base and tip (excluding awns), the number of spikelets per spike (NSS), and the number of grains per spike (NGS), were measured. For the weight of grains per spike (WGS) and thousand kernel weight (TKW), 1000 random seeds from each plot were weighed. Additionally, salt tolerance indices (STI) for all traits were calculated using the following equation: STI = trait value under salinity/trait value under control (Fernandez 1992).

Determination of antioxidants

Enzyme extracts were prepared from barley leaf samples for all accessions (three replicates) according to the procedure reported by Aebi (1984) and Garratt et al. (2002). The photochemical reduction of nitro blue tetrazolium (NBT) through inhibition of enzyme activity was assessed to determine superoxide dismutase activity (SOD, EC 1.15.1.1) (Dhindsa and Matowe 1981). The reaction mixture was prepared by mixing 0.1 mL enzyme extract with methionine (13.33 mM), 50 mM buffer (phosphate, PH 7.8), sodium carbonate (50 mM), NBT (75 Μm), 0.1 mL riboflavin (2 mM), and EDTA (0.1 mM), and then exposed to light for 15 min. The absorbance was read spectrophotometrically at 560 nm.

Catalase activity (CAT, E.C. 1.11.1.6) was calculated using the decomposed H2O2 measurement (Aebi 1984). Enzyme extract (50 μL) was mixed with 0.5 mL H2O2 (75 mM), and 1.5 mL buffer (phosphate 0.1 M, pH 7). An absorbance decrease was observed at 240 nm for 1 min. Catalase activity was estimated from the amount of decomposed H2O2. CAT activity (1 unit) was described as the amount of enzyme needed to catalyse the oxidation of 1 µmol of H2O2 min−1.

The activity of ascorbate peroxidase (APX, E.C. 1.11.1.11) was determined by calculating the decrease in optical density caused by ascorbate at 290 nm (Nakano and Asada 1980). Enzyme extract (0.1 mL) was mixed with 50 mM buffer (potassium phosphate, pH 7.0), 0.1 mL H2O2 (0.1 mM), EDTA (0.1 mM), ascorbate (0.5 mM), and water. APX activity (1 unit) was described as the amount of enzyme required to convert 1 µmol of ascorbate min−1 at 25 °C.

Glutathione reductase activity (GR, E.C. 1.6.4.2) was performed by estimating NADPH oxidation at 340 nm (Carlberg and Mannervik 1985). Enzyme extract (37 μL) of the was mixed with 50 mM Tris–HCl buffer (pH 7.5), MgCl2 (3 mM), GSSG (0.5 mM), and NADPH (0.2 mM), and GSSG was added to start the reaction. GR activity (1 unit) was described as the amount of enzyme needed to catalyse the oxidation of 1 µmol of NADPH min−1. The contents of ascorbic acid and glutathione were extracted and determined using the methods described in De Kok et al. (1986).

Estimation of ionic Na+ and K+ content

Leaf samples were dried using an oven at 70 °C until reaching constant weight. The ionic content of Na+ and K+, and the K+/Na+ ratios were estimated from the dried leaves using the nitric-perchloric acid digestion method according to Westerman (1990). The absorbance was measured at 766.5 nm and 589.0 nm for K and Na, respectively.

Data analyses

Analysis of variance (ANOVA) and broad-sense heritability (H2) were performed to compare salt and control treatments at p < 0.05 for all studied traits using R software. R-studio was used to calculate a comparison between treatments at p < 0.05, including boxplots for all the studied traits (Allaire 2012). The mean value for the measured traits was computed as Best Linear Unbiased Estimates (BLUEs) using the lme4 package (Bates et al. 2015). Pearson correlation analysis at p-value < 0.05 was used to calculate the correlation matrix among the traits for each treatment (Wei and Simko 2017):

$${H}^{2}=\frac{{\sigma }^{2}G}{{\sigma }^{2}G+\frac{{\sigma }^{2}e}{r}}$$

where, σ2G is the variance of genotypes (accessions); σ2e is the error variance, and r is the number of replicates.

Multi-locus genome-wide association scan (GWAS)

Multi-locus GWAS analysis was performed using Genomic Association and Prediction Integrated Tool (GAPIT) in R software (Lipka et al. 2012) using 19,276 SNPs and BLUE values of studied traits of the accessions. The GAPIT statistical model used in GWAS analysis is known as the Fixed and random model Circulating Probability Unification (FarmCPU) model. The FarmCPU model was also used by combining multiple markers as covariates in a fixed-effect model and optimizing the associated covariate markers in a random effect model separately, enabling us to control false-positive associations by attempting to prevent model overfitting (Liu et al. 2016b). Moreover, a threshold P-value was at 0.00001 that equal to − log10 (P) ≥ 4. The phenotypic variance explained by significant makers (R2) ≥ 10% was used to determine the significance of marker-trait associations (MTAs) that were associated with more than two traits.

Linkage disequilibrium (LD) between pairs of polymorphic SNPs on each chromosome was calculated by genome-wide pairwise estimates as a squared correlation (r2). The LD decay estimated as r2 versus the distance between pairs of polymorphic SNPs was presented as Mbp.

The most significant SNPs were used to describe the relatively close potential candidate gene inside the LD interval based on their physical position using the barley database BARLEX (Mascher et al. 2021). BARLEX is a web-based platform that includes gene annotations, gene ontology (GO), and a summary of the barley genomic space and gene annotations. The annotation of the high-confidence (HC) candidate gene was detected using Morex v3 (Mascher et al. 2021). For gene expression and molecular validation, we used the OPEN-ACCESS version of Genevestigator as an expression and transcriptomes database (Hruz et al. 2008). The database includes a wide range of gene expression datasets across different experiments or biological contexts such as anatomy, development, stresses, diseases, and mutation for the latest version of barley genes Morex v3.

Results

Natural variation of agronomic trait performance

The phenotypic data for all barley accessions under control and salt stress are presented in Table S1. The phenotypic distribution and variation for all agronomic traits under both treatment conditions are shown in Figure S1. The minimum, maximum, and mean of each trait for all genotypes are presented in Table S2. As expected, all yield traits (SL, NSS, NGS, WGS, and TKW) had significantly lower mean performance under salinity stress than under control conditions (Fig. 1a–e and Table S2). On average, all yield traits had a reduction due to salt stress ranging from 15% (NGS) to 48% (WGS) (Fig. 1a–e). Based on salt-tolerant indices (STI), mean genotypic values for all yield traits ranged from 52% (WGS) to 85% (NGS) (Table S2). The broad-sense heritability (H2) values ranged from 96.32% for NSS_C to 99.38% for NGS_C. In response to salt stress, the highest H2 was detected for NGS (99.3%), followed by SL (97.91%), and the lowest was 93.82% for TKW. Moreover, H2 values ranged from 53.64% (TKW_STI) to 81.26% (SL_STI) in the case of STI (Table S2).

Fig. 1
figure 1

Natural phenotypic variation for all agronomic traits; a spike length (SL), b number of spikelet per spike (NSS), c number of grains per spike (NGS), d weight of grains per spike (WGS), and e thousand kernel weight (TKW) in the degree of significance is indicated as *p 0.05; **p 0.01; ***p 0.001; ****p 0.0001

Full size image

Natural variation of antioxidants

Significant natural variation was also detected for all antioxidants with normal distribution in response to salinity conditions (Table S3 and Figure S2). Under control conditions, the mean genotypic values for SOD, CAT, APX, and GR ranged from 2.29 μmol mg−1 protein (SOD) to 1.10 μmol mg−1 protein (APX) (Fig. 2a–d and Table S3). The mean values ranged from 5.60 μmol mg−1 protein (SOD) to 2.37 μmol mg−1 protein (APX) under salt treatment. A significant increment in SOD, CAT, APX, and GR was observed under salt stress (58%, 48%, 50%, and 65%, respectively) compared to the control condition (Fig. 2a–d and Table S3). For STI, mean genotypic values ranged from 252.18 μmol mg−1 protein (APX) to 1986.54 μmol mg−1 protein (CAT). The highest value was detected for CAT_STI by 53,666.67 μmol mg−1 protein. The heritability values ranged from 86–97% under control conditions and 84–99% under salt treatment. The highest value was for CAT_S (99.74%) and GR_C (97.93%). Heritability values for STI ranged from 77.96% (SOD_STI) to 99.74% (CAT_STI), as shown in Table S3.

Fig. 2
figure 2

Natural phenotypic variation for all physiological traits. a Superoxide dismutase (SOD), b catalase (CAT), c ascorbate peroxidase (APX), and d glutathione reductase (GR), e ascorbic acid (AsA), f glutathione (GSH), g sodium (Na+), h potassium (K+) content and i sodium/potassium ratio (K+/Na+) ratio in barley accessions under control and salt stress. The degree of significance is indicated as **p 0.01

Full size image

Natural variation of ionic Na+ and K+ contents

Natural phenotypic distribution was significantly detected for Na+ and K+ ions under both treatments (Figure S3). K+ content was decreased by 67% under salt treatment compared to the control across all accessions (Fig. 2g and Table S3). In contrast, Na+ content significantly increased by 87% under salt treatment compared to the control treatment (Fig. 2h and Table S3). Moreover, the K+/Na+ content ratio was significantly increased under control conditions by 27% compared to the salt treatment (Fig. 2i and Table S3). Their heritability values were 91% under control, while they ranged from 84.09–91.39% under salt stress conditions. For STI, mean values were 768.51, 32.64, and 4.25 for Na+, K+, and the K+/Na+ content ratio, respectively. Additionally, high H2 values were detected, ranging from 84.06% (K+/Na+ content) to 90.87% (Na+ content) (Table S3).

Correlations analyses

Pearson’s correlations among traits based on the mean of all accessions under each treatment condition revealed a significant correlation amongst all studied traits under both treatments. Under control conditions, SOD showed significant positive correlations at p-value 0.0001*** with CAT, APX, and GR (r = 0.83***, 1.0***, and 0.96***, respectively). Similarly, a significant positive correlation was noted between SOD and all agronomic traits except NGS** and WGS***. Moreover, APX showed positive correlations with GR and TKW (r = 0.95*** and 0.54*, respectively). Interestingly, Na+, K+, and the K+/Na+ content ratio negatively correlated with all physiological and agronomical parameters except NSS and TKW (Fig. 3a). Under salinity conditions, SOD activity showed high positive associations with CAT, APX, GR, AsA, and GSH at 0.75**, 0.95***, 0.94**, and 0.8***, respectively. Contrarily, SOD and APX were negatively correlated with all agronomic traits except for NGS and WGS. Consistent with salinity, a negative correlation was detected between Na+, K+, and the K+/Na+ content ratio with all physiological and agronomical parameters except NSS and TKW (Fig. 3b), suggesting the presence of common genetic determinants affecting all mineral accumulation (Na+, K+, and K+/Na+ content ratio) and grain filling in response to salt stress.

Fig. 3
figure 3

Correlations of the studied traits in barley accessions under both a control and b salinity treatment. The degree of significance is indicated as *p 0.05; **p 0.01; ***p 0.001

Full size image

Genetic associations and candidate genes

GWAS identified a total of 263 SNPs as significantly associated with 14 traits and passing thresholds of − Log10 (p) ≥ 4.0 and with R2 ≥ 10% under both control and salinity conditions (Table S4). Out of the 263 SNPs, we observed 16, 15, 14, 11, and 10 SNPs associated respectively with APX_STI, CAT_STI, AsA_C, CAT_S, and SOD_S, whereas less than 10 SNPs were associated with each of the remaining traits (Table S4 and Figure S4). Moreover, 37 SNP markers were significantly associated with the Na+ and K+ content and K+/Na+ ratio, where the highest number was detected for K+/Na+_S (12 SNPs), followed by K+/Na+_C (9 SNPs). Less than nine SNPs were associated with each of the remaining traits under both control and salinity treatments (Table S4 and Figure S5).

For all agronomic traits, 86 SNPs (− log10 p-value ≥ 4) were significantly detected on all chromosomes (Table S4). The highest number of markers was detected for SL_S (23 SNPs), followed by NGS_C and SL_C (11 SNPs each), whereas 10 SNPs were associated with NGS_S and 8 SNPs for TKW_C (Table S4 and Figure S6).

Interestingly, highly significant associations (− log10 (p-value) > 10) were discovered on chromosomes that were found to be highly associated with CAT_STI (Figure S7), CAT_S, NGS_C, NGS_S, and TKW_C (Table S4).

This study revealed five genomic regions based on chromosome number. Each genomic region was classified into sub-regions based on marker-trait associations (MTAs) and LD, which harbors 193 potential candidate genes distributed on chromosomes 1H, 2H, 4H, 6H, and 7H (Table S5 and Table S6). Of these candidate genes, 23 potential candidate genes were found to control all enzymatic and non-enzymatic antioxidant components under salt stress (Table 1 and Table S6).

Table 1 List of high confidence candidate genes associated with the variation of multi-traits in the barley collection under control salt stress treatment
Full size table

On chromosome 1H, two important candidate genes were discovered in the genomic region (1.1 and 1.2). The first one is the pentatricopeptide repeat (PPR) protein superfamily that controls the variation of Na+_C, K+_C, and Na+_STI contents at position (6,697,338–6,697,349 bp). The second is the FBD-associated F-box protein associated with APX_STI and SOD_STI at position 50,002,764–51,210,527 bp. Chromosome 2H harbours several candidate genes associated with enzymatic and non-enzymatic antioxidant components such as SOD, APX, GR, AsA, and GSH under salt stress. The most prominent candidate gene is HORVU.MOREX.r3.2HG0181470 at position 557,725,076–557,730,772 bp in genomic region number 2.2 on 2H that annotated as Beta-glucosidase (Fig. 4). The closest linked marker to this gene is chr2H:557720032:G:T at 557,720,032 bp, which strongly correlates with the abovementioned traits (Table S5). This SNP marker had negative influences, approximately − 0.33, − 0.1, and − 0.5 μmol mg−1 protein of APX, GR, and SOD, respectively (Table S5).

Fig. 4
figure 4

Manhattan plot for all physiological traits under salinity in barley with LD heat map. The x-axis shows the chromosomes and the SNP order. The y-axis shows the − Log10 (P-value) for each SNP marker

Full size image

The third genomic region on chromosome 4H (3.1) was strongly associated with the SNP chr4H:13826536:A:G detected inside the HORVU.MOREX.r3.4HG0336230 gene at position 13,823,277–13,824,982 bp and annotated as Glycosyltransferase. This SNP showed a tight association with SL_STI that negatively affected SL_STI by approximately − 4.12 cm (Table S5). Another genomic region located on chromosome 4H (3.2) harbours the variation of all ionic content parameters (K+/Na+_S, K+/Na+_STI, and K+_STI) under salt stress (Table 1, Table S5). The chr4H:569747682:G:A SNP found inside the gene HORVU.MOREX.r3.4HG0405320 (569,744,191–569,748,338 bp), is annotated as 3-ketoacyl-CoA synthase (Tables S5 and S6). Interestingly, this SNP marker positively influenced the ionic content, which was approximately 0.16, 0.45, and 3.74 μg g−1 for K+/Na+_S, K+/Na+_STI, and K+_STI, respectively (Table S5).

Interestingly, the fourth genomic region (4.3) was detected on 6H, chr6H:451682318:A:T SNP near the gene HORVU.MOREX.r3.6HG0605480 (451,676,227–451,677,811 bp) which annotated as the PLATZ transcription factor, regulating the variation of CAT_S, APX_S, GR_S, AsA_S, and GSH_S (Fig. 4). Markedly, an A:T SNP showed the highest − log10 (P-value) and had a positive effect on CAT_S, APX_S, GR_S, AsA_S, and GSH_S (approximately 0.60, 15.12, 0.17, 0.18, 12.9, 140.8 μmol mg−1 protein, respectively) (Table S5).

Ultimately, the genomic region (5.2) on 7H was associated with SOD_S and APX_S at 153,772,300–153,774,057 bp on chr7H:153773211:A:C SNP inside the gene HORVU.MOREX.r3.7HG0676830 that encodes L-gulonolactone oxidase (Tables S5 and S6). This SNP (A:C) showed a negative influence on SOD_S and APX_S (approximately—0.23 and 0.35 μmol mg−1 protein, respectively) (Table S5). The allelic variation at chr7H:153773211:A:C SNP inside the l-gulonolactone oxidase gene demonstrates a negative selection of accessions carrying the C allele. This allele appears in a cultivar with lower activity of enzymatic antioxidants, e.g., superoxide dismutase and ascorbate peroxidase, under salinity conditions (Fig. 5). Furthermore, high expression profiles for our candidate gene were detected during the vegetative and reproductive phases under optimal conditions (Figure S8a). The HORVU.MOREX.r3.7HG0676830 gene was upregulated during microspore development in the reproductive phase under osmotic stress conditions, showing that the candidate gene plays an essential role in enhancing salt stress-responsive in barley under salt stress conditions (Figure S8b).

Fig. 5
figure 5

a Locus zoom in on regional plot colocalization of highly significant SNPs with SOD and APX under salinity in barley. The x-axis shows the chromosomes and the SNP order. The y-axis shows the − Log10 (P-value) for each SNP marker. b The physical position and structure of the candidate gene at 7H, HORVU.MOREX.r3.7HG0676830 with allelic variation in the exon, Arrow indicated the SNP located physically inside the HORVU.MOREX.r3.7HG0676830 with LD heat map. c Allele effect on the accessions at the natural genetic variant (A/C), d natural variation of accessions based on the alleles (A/C)

Full size image

Discussion

Salinity, in the form of NaCl, is considered the most serious abiotic stress worldwide as it severely constrains cereal crops’ yield and production. Nevertheless, the response of different plant species and accessions to salt stress may vary significantly because it is genotype- and concentration-dependent. Therefore, we screened new diverse barley accessions for morphological, physiological, and antioxidant responses to salinity to detect new physiological, biochemical, and antioxidative genetic factors and mechanisms to develop new barley varieties that can tolerate salt stress (Fig. 6). In our study, significant variations were observed in a core collection of barley accessions for all the studied traits under salinity conditions. The high variation found in this study could be used to enhance barley salt tolerance in breeding programs. Moreover, higher heritability values were detected for antioxidants and ionic content of Na+ and K+ parameters under salinity conditions, indicating that the major part of the variation is genetically controlled. Similarly, Thabet et al. (2021a) detected higher heritability values for antioxidant components under salinity stress during the vegetative developmental phase. Under salinity treatment, all morphological traits showed a significant reduction when compared to the control, demonstrating that salt stress might cause a decline in the plant growth and performance of the whole studied panel. Thabet et al. (2021a) also detected a significant reduction of all agronomic traits, including SL, NSS, NGS, WGS, and TKW, under salt stress conditions.

Fig. 6
figure 6

Schematic diagram illustrates the biosynthesis of a cascade of antioxidants in barley worldwide population during salt stress. ROS reactive oxygen species, O2· superoxide anions, H2O2 hydrogen peroxide, SOD superoxide dismutase, CAT catalase, APX ascorbate peroxidase, GR glutathione reductase, GSH reduced glutathione, GSSH oxidized glutathione, GPX glutathione peroxidase, DHAR dehydroascorbic acid reductase, AsA ascorbic acid, DHA dehydroascorbate, MDHAR monodehydroascorbate reductase, NADPH nicotinamide adenine dinucleotide phosphate

Full size image

Non-enzymatic antioxidant components such as AsA and GSH are generated to reduce oxidative damage and scavenge the ROS molecules in different plant species (Ashraf 2009) to maintain cellular ROS homeostasis under salt stress. Our investigation detected that the contents of AsA and GSH were higher under salt treatment than under the control treatment. It is indicated that those non-enzymatic antioxidants played critical roles in reducing oxidative damage and regulating the cellular ROS homeostasis under salinity conditions. Moreover, a significant increase in ascorbic acid contents signified the major contribution of the ascorbate–glutathione cycle in quenching the ROS molecules (Munns and Tester 2008). Taken together, Hernández et al. (2001) reported that the ascorbate–glutathione cycle activity may be higher under salt stress in the studied panel. Under salinity stress, tolerant plants have developed an enhanced enzymatic antioxidant defence system to combat the outcomes of ROS molecules (Sarker and Oba 2020a). All barley accessions detected significantly increased SOD, CAT, APX, and GR activities under salt treatment. Our results are in agreement with Thabet et al. (2021a), who observed high enzymatic antioxidant activities in response to salinity conditions. SOD contributed to the dismutation of O2− to O2 and finally to H2O2; after that, a variety of peroxidases used various reducing agents to convert H2O2 into H2O (Hwang et al. 2012). In parallel, CAT decomposes H2O2 produced in the peroxisome due to photorespiration (Sarker and Oba 2020b). Altogether, it is evident that the higher activities of POD, CAT, and APX cope with the activity of SOD to counteract the negative effects of O2− and H2O2 molecules, indicating that the activities of these enzymes showed a strong correlation with salinity stress tolerance in wheat and barley (Munns et al. 2006; Temel and Gozukirmizi 2015).

Interestingly, the higher GR activity might be increased by the concentrations of NADP+ to gain electrons from the photosynthetic electron transport chain via the reduction of ROS (Ramachandra Reddy et al. 2004). Altogether, barley genotypes may exhibit a more active AsA-GSH cycle under salinity conditions. Therefore, crop plants may induce high levels of enzymatic and non-enzymatic antioxidants that increase tolerance to salt-induced oxidative damage (Siringam et al. 2011).

In this study, all barley accessions showed a significant increase in Na+ content and a significant decrease in K+ content under salinity. High Na+ ion concentrations may inhibit nutrient uptake because it interferes with K+ selective ion channels in the root plasma membrane (Tester and Davenport 2003). Due to the K+ requirement for tRNA binding to ribosomes, osmotic damage and a decrease in K+ ion intake can disrupt protein synthesis (Blaha et al. 2000). It was postulated that a higher ionic K+/Na+ content ratio plays a key role in plant tolerance to salinity stress (Chen et al. 2007). Furthermore, Na+ should be prevented from reaching the cytosol by restricting Na+ influx and efflux from the cytoplasm into the vacuoles. Hence, sustaining a high K+/Na+ ratio is one of the critical factors in plant responses to salinity stress (Ishikawa and Shabala 2019).

Our genetic analyses identified five genomic regions harboring SNPs significantly associated with 14 traits at different developmental stages. These genomic regions are linked with 193 candidate genes and are distributed on chromosomes 1H, 2H, 4H, 6H, and 7H. Markedly, 23 candidate genes were potentially involved in controlling all enzymatic and non-enzymatic antioxidant components and ionic Na+ and K+ parameters under salt stress.

To our knowledge, these are newly detected loci and candidate genes based on the latest physical positions using a high-quality reference sequence of the cultivar Morex (Mascher et al. 2021). There could be other loci reported previously in the same genomic region; however, we cannot compare them with our loci and candidate genes because we used the latest physical map of barley, while previous studies mostly used the genetic map (cM) or old physical map (Mwando et al. 2021; Saade et al. 2020). Moreover, most of what we know about QTLs in response to salt stress tolerance from previous studies is based on mapping biparental populations (Mwando et al. 2021; Sayed et al. 2021; Xue et al. 2017) or diverse collections (Saade et al. 2020; Thabet et al. 2021b). We used a new diverse collection for the first time to detect new loci. Therefore, comparisons between our GWAS findings and other studies using biparental or diverse populations are tricky, as the association markers and positions we identified were based on a worldwide barley collection not previously investigated for salt stress conditions.

Two candidate genes, namely pentatricopeptide repeat (PPR) superfamily protein superfamily that controls the variation of ionic content of Na+_C, K+_C, and Na+_STI, and FBD-associated F-box protein that controls the variation of APX_STI and SOD_STI were identified on chromosome 1H at position 6,787,041–6,788,808 bp and position 50,002,764–51,210,527 bp, respectively. The first gene, HORVU.MOREX. r3.1HG0003310, encodes the PPR superfamily protein involved in various physiological and developmental processes, such as photosynthesis, respiration, seed development, and embryogenesis (Liu et al. 2016a; Robles and Quesada 2019). In Arabidopsis, PPR40 had an important role in salinity tolerance (Robles and Quesada 2019). Mitochondrial mutants (ppr40) showed hypersensitivity towards these adverse conditions during early growth stages (Zsigmond et al. 2012). Also, the authors reported that upregulated expression of the mitochondrial PPR40 gene conferred salinity tolerance in Arabidopsis by enhancing the mitochondrial electron transport’s stability and reducing oxidative damage, which could also be the role of our candidate gene.

The second gene, HORVU.MOREX.r3.1HG0016850, encodes the F-box family protein, which is recognized by a conserved 40–50-amino acid F-box motif. This gene plays a critical role in abiotic stress tolerance through the ubiquitin pathway (Maldonado-Calderon et al. 2012; Zhang et al. 2008). Zhou et al. (2015) suggested that the TaFBA1 gene is involved in oxidative stress tolerance in wheat. In transgenic plants, enhanced oxidative stress responses were attributed to higher enzymatic antioxidant activities such as SOD, CAT, APX, and POD compared to the wild type, which may be related to the overexpression of the TaFBA1 gene (Zhou et al. 2015). Together with our candidate gene, it is critical to comprehend the roles of F-box proteins in barley tolerance to various abiotic stress conditions.

The QTL analysis for salt-stress at seedlings in a barley mapping population of Nure × Tremois detected a QTL between 35–40 cM for shoot-related traits on 1H (Xue et al. 2017), and the HvNax4 locus was mapped on 1HL using Clipper × Sahara mapping population (Rivandi et al. 2011). The association study conducted by Mwando et al. (2020) detected associated markers with two candidate genes (Piriformospora indica-insensitive protein 2 and protein kinase superfamily protein) on 1H enhancing salinity tolerance during germination in 350 worldwide barley collections. Neither these candidate genes nor QTL overlap with the loci/candidate genes detected in our study.

Chromosome 2H harbours several potential candidate genes involved in most of the enzymatic and non-enzymatic antioxidant components in response to salt stress. The most prominent candidate gene is HORVU.MOREX.r3.2HG0181470 at position 557,725,076–557,730,772 bp, annotated as Beta-glucosidase. Mwando et al. (2021). A QTL underlying salt stress tolerance at germination was mapped on the long arm of 2H using a doubled haploid (DH) barley population (Angessa et al. 2017). Based on the association mapping, Saade et al. (2020) detected an SNP at 141.87 cM associated with yield components in response to salinity stress, while Thabet et al. (2021a) identified associations with spikelet and grain weight under salt stress during the vegetative phase with an SNP at 577,284,290 bp. These studies demonstrated that 2H harbours important QTL/loci underlying salt stress tolerance in barley, but it is difficult to compare them as they are based on different genetic backgrounds.

Xu et al. (2012) reported that several β-glucosidase proteins hydrolyse abscisic acid (ABA) from glucose-conjugated abscisic acid (ABA-GE) and multiple ABA production pathways work together to enhance ABA levels. In Arabidopsis, a vacuole-glucosidase homolog1 (BG1) with glycosylated abscisic acid (Glc-ABA) hydrolysing activity is important in osmotic stress responses. Therefore, we suggest that β-glucosidase proteins play a key role in altering the antioxidant pool by inducing high levels of enzymatic and non-enzymatic antioxidant components that increase tolerance to salt-induced oxidative damage.

The third genomic region (3.1) is located on chromosome 4H at position 13,823,277–13,824,982 bp with two significant SNPs. The first one is A:G SNP (13,826,536 bp) inside the HORVU.MOREX.r3.4HG0336230 gene annotated as Glycosyltransferase, belonging to the family 1 UDP glycosyltransferases (UGTs). Lairson et al. (2008) detected that endogenous cytokinin levels are linked to small molecular modification by the UGTs family, which transports sugar substituents to small acceptor molecules. Hou et al. (2004) identified the Arabidopsis UGT76C1 and UGT76C2, which finely tune the glycosylation of cytokinins and regulate the cytokinin homeostasis in several plant species. Furthermore, Li et al. (2015) demonstrated that AtUGT76C2, Arabidopsis cytokinin glycosyltransferase, is implicated in the adaptation to osmotic and drought stress. The second SNP (G:A) is found inside the HORVU.MOREX.r3.4HG0405320 gene at position 569,744,191–569,748,338 bp, annotated as 3-ketoacyl-CoA synthase and controls the variation of K+/Na+_S, K+/Na+_STI, and K+_STI contents. 3-ketoacyl-CoA synthase is the key enzyme in the fatty acid elongation process (Yang et al. 2020). In Arabidopsis, upregulation of Ketoacyl-CoA Synthase from Vitis vinifera L. enhances salinity stress responses. The enhanced salt stress resulted from the cooperation of various mechanisms and the regulatory oversight of ion transporters, the accumulation of osmolytes, and the maintenance of membrane integrity (Yang et al. 2020). Altogether, these genomic regions might be a target in future studies to improve salt-tolerant barley varieties. Association analysis of the USDA barley core collection identified a major locus on 4H named HKT1;5 controlling Na+ content in the flag leaf under saline conditions (Hazzouri et al. 2018). Based on the literature, no known QTL or genes are located close to our associated genomic regions or candidate genes on 4H.

Interestingly, the fourth genomic region (4.3) on 6H (chr6H:451682318:A:T SNP near the gene HORVU.MOREX.r3.6HG0605480 at position 451,676,227–451,677,811 bp) was annotated as PLATZ (plant AT-rich protein and zinc-binding protein) transcription factor, regulating the variation of CAT_S, APX_S, GR_S, AsA_S, and GSH_S related to antioxidant components. The PLATZ transcription factor family is a class of plant-specific zinc-dependent DNA-binding proteins. PLATZ plays a critical role in seed endosperm development during the early developmental crop stages. PLATZ proteins are involved in the response to environmental stress. GmPLATZ1 (Glycine max PLATZ1) transcription levels in soybean were increased by ABA and drought stress (Fu et al. 2020). Upregulation of GmPLATZ1 in transgenic Arabidopsis exhibited lower germination rates under osmotic and ABA stress (So et al. 2015). Altogether, this gene was linked to enzymatic antioxidants, suggesting that it participates in the salt tolerance pathway via the antioxidant defence system, which may enhance plant growth and performance. No known mapped QTL or candidate genes are co-located close to associated genomic regions or candidate genes detected on 6H in our study, confirming the novelty of our candidate gene.

The last genomic region (5.2) on 7H at 153,773,211 (bp chr7H:153773211:A:C SNP) inside the HORVU.MOREX.r3.7HG0676830 gene (153,772,300–153,774,057 bp) that encodes l-gulonolactone oxidase was associated with SOD_S and APX_S. Many QTL or candidate genes important in response to salt stress tolerance were reported on 7H based on a bi-parental mapping population or association population. For example, the QTL on 7H reported by Shavrukov et al. (2010); Xue et al. (2017) had a candidate gene named HvNax3 that can reduce Na+ content in shoots. Many associated genomic regions with candidate genes were identified in the association-based studies (Saade et al. 2020; Thabet et al. 2021a, b). However, none are genetically close to the genomic region from this study, and the importance of the l-gulonolactone oxidase gene in salt stress tolerance was shown for the first time.

The allelic variation of the chr7H:153773211:A:C SNP inside the l-gulonolactone oxidase gene reveals a negative selection of accessions carrying the C allele. This allele is found in landraces from the Far and Near East that have lower enzymatic antioxidant activity, e.g., SOD and APX, in response to salinity. l-gulonolactone oxidase (GLOase) is one of the key enzymes that catalyse the transformation of l-Gulonolactone to AsA (Hemavathi et al. 2010). Relative expression of GLOase showed a significant increase in treated Aeluropus littoralis shoots in response to salt stress (Malekzadeh et al. 2015). This enhancement may be correlated with the reduction in K+ and water content and the increase in Na+ concentration that induced oxidative stress, resulting in increasing ROS molecules. Therefore, Na+ should be prevented from reaching the cytosol by restricting Na+ influx and efflux from the cytoplasm into the vacuoles. Hence, one of the critical factors in plant tolerance to salt stress is maintaining a high K+/Na+ ratio. Altogether, GLOase actively participated in multiple stress responses by influencing the activity of ROS enzymes. The allelic variation at chr7H:153773211:A:C SNP inside the l-gulonolactone oxidase gene demonstrates a negative selection of accessions carrying the C allele. Markedly, A:C SNP was associated with important enzymatic antioxidant components, such as SOD_S and APX_S, and thus, could be a candidate genomic region for helping to improve barley tolerance to salt stress. Interestingly, the candidate gene HORVU.MOREX.r3.7HG0676830 is upregulated in microspore development under salt stress, indicating a potential involvement in salt tolerance responsive mechanisms in barley. It was demonstrated that reproductive developmental events, including microspore development, are extremely sensitive to environmental stress conditions such as temperature, drought, and salt/osmotic stress, resulting in high levels of microspore abortion (Alqudah et al. 2011; De Storme and Geelen 2014). Therefore, it is also plausible to suggest that our candidate is a key regulator for enzymatic and non-enzymatic antioxidants and ionic balance through scavenging the deleterious effects of ROS and oxidative damage under salt stress conditions.

Conclusions

In conclusion, our study provided detailed mechanistic insights into the genetic basis of salinity tolerance-related traits in barley under salt stress conditions. Highly reliable genomic regions were detected for multiple salinity tolerance-related traits, such as antioxidant components (SOD, APX, GR, AsA, and GSH), K+ content, and K+/ Na+ content ratio. Potential candidate genes, such as HORVU.MOREX. r3.2HG0181470 and HORVU.MOREX. r3.7HG0676830 that were annotated as Beta-glucosidase and l-gulonolactone oxidase, respectively, were also identified. Novel genomic regions and the linked genes discovered in this study could be potential candidates for developing barley salt-tolerant breeding lines. These findings maximize our understanding of the genetic factors underlying morpho-physiological and grain yield traits in barley in response to salinity that can be implemented through breeding for crop improvement.