Single Nucleotide Polymorphisms in Bmy1 Intron III Alleles Conferring the Genotypic Variation in β-amylase Activity under Drought Stress Between Tibetan Wild and Cultivated Barley

β-amylase activity is related to the polymorphism of Bmy1 intron III; however, no attention has been given to such relationship under environmental stresses like drought. In this study, 73 cultivated barley genotypes and 52 Tibetan wild barley accessions were used to test the association between Bmy1 gene intron III polymorphisms and β-amylase activity under drought stress. Our results showed that three alleles, Bmy1.a, Bmy1.b and Bmy1.c, existed in the examined barley genotypes. Tibetan wild barley had higher proportion of Bmy1.b, whereas cultivated barley showed higher proportion of Bmy1.a. Impressively, barley genotypes with Bmy1.b showed signicant increase in β-amylase activity under drought stress, compared with those with Bmy1.a or Bmy1.c, indicating that Bmy1.b allele might provide more chances for developing barley cultivars with higher β-amylase activity under water stress than both Bmy1.a and Bmy1.c alleles. Furthermore, the Tibetan wild barley XZ147, belonging to Bmy1.b allele type, showed signicant higher β-amylase activity than the cultivar Triumph under drought stress. This might result from the unique amino acid substitution M527 or the amino acid composition of R115, D165, A233, S347 and M527 of XZ147.

Genes controlling β-amylase synthesis may differ among plant tissues; however, their functional domain should be similar because all of them belong to the same gene family. In malting barley, β-amylase activity is an important quality trait, being closely associated with malt quality (Arends et al. 1995). There are two forms of β-amylase expressed in barley. The endosperm speci c form is a dominant one, which encoded by Bmy1 gene, about 5 kbp, located on the long arm telomere of 4H chromosome (Nielsen et al. 1983;Powling et al. 1981). It consists of 7 extrons and 6 introns, and encodes a polypeptide chain of 535 amino acids (Kreis et al. 1988;Yoshigi et al. 1994). The Bmy1 gene was closely correlated with DP (Hayes et al. 1993). The polymorphisms in the intron III of Bmy1 are extremely abundant. There are four different allele types according to the presence and absence of 4 insertion/deletions (INDELs) (126-, 38-, 11-and 21-bp), namely Bmy1.a, Bmy1.b, Bmy1.c and Bmy1.d . Erkkilä et al. (1998) identi ed two different indels (126-bp and 38-bp) through southern blot analysis using the rst 320bp located at the 5' region of Bmy1 intron III. Currently, all these four types of alleles derived from Bmy1 intron III INDELs have been identi ed, including the insertion of 126-bp and 38-bp in the cultivated barley Adorra, and the deletion of 126-bp and 38-bp INDELs in the wild barley PI 296897 (Erkkilä et al. 1998). Sjakste andZhuk (2006) also observed abundant polymorphisms of Bmy1 intron III, and found a potential binding site for a transcript factor. Numerous studies have con rmed that the polymorphism of Bmy1 intron III is correlated with enzyme activity, thermal stability and enzymatic kinetics of β-amylase (Erkkilä et al. 1998;Erkkilä and Ahokas 2001;Kaneko et al. 2000;Ma et al. 2001;Paris et al. 2002). Erkkilä and Ahokas (2001) and Gunkel et al. (2002) reported that the presence or deletion of a 126-bp INDEL in the 5' end of Bmy1 intron III was associated with low activity and high thermal stability of β-amylase, respectively. Coventry et al. (2003) determined the activity and thermal stability of β-amylase, and DP value and identi ed a primer pair which could discriminate the presence or absence of the 126-bp INDEL.
Meanwhile, the Single Nucleotide Polymorphisms (SNPs) in the coding region of Bmy1 also affected βamylase morphology, activity and thermal stability of the enzyme (Li et al. 2001). All these studies indicate that it is practicable to identify enzyme activity and thermal stability of β-amylase by only detecting the polymorphisms of either coding or noncoding regions of Bmy1 under normal environmental condition. However, little attention has been given to this issue under the environmental stresses like drought.
Tibetan wild barley is considered as one of ancestors of modern cultivated barley, rich in genetic diversity (Dai et al. 2012). Till the present, however, most studies about the wild barley are mainly concentrated on salinity tolerance (Qiu et al. 2011), aluminum tolerance ) and grain protein content (Wei et al. 2009), and no research has been done to investigate the genetic variation of malt quality under drought stress. In this study, we used 73 cultivated and 52 Tibetan wild barley genotypes to investigate the correlation between Bmy1 intron III polymorphism and grain β-amylase activity under drought stress, and to compare the difference between wild and cultivated barley.

Plant cultivation and drought treatment
In the present study, 73 cultivated barley genotypes and 52 Tibetan wild barley accessions were used (Supplemental Table S1). All genotypes or accessions were sown in mid-November 2018, and grew in two rain shelters (60 m × 20 m) in Changxing experimental station of Zhejiang University (Huzhou, Zhejiang province, China) for control and drought stress respectively. The experiment was arranged in random block design, with three replicates for each treatment. In each replicate, every barley genotype was sown in three rows, and only the grains from the middle line were harvested for the following investigations.
After germination, all barley plants in two rain shelters were well-irrigated with a sprinkling irrigation system, to keep the soil water content around 35% (equaling to water potential of − 0.15 MPa, monitored by HH2 Moisture Meter, Delta-T Devices, Cambridge, UK). When approximately 85% of barley genotypes reached the heading stage, drought stress was started when approximately 85% of barley genotypes were at heading stage, by stopping water supply until the soil water content dropped to 14% (equaling to water potential of − 0.75 MPa). The drought stressed barley plants were thereafter kept this water level to the maturity stage, whereas the control plants were maintained normal water supply to keep soil water content around 35%. During drought treatment, soil water contents in the two rain shelters were monitored every three days by randomly measuring soil water content at 30 positions over the whole shelter.
To analyze the polymorphism of Bmy1 gene, seeds of each barley genotype were surface sterilized with 12.5% NaClO solution, thoroughly rinsed with tap water (for at least 30 min) and then grown using paper roll with 1/5 strength of Hoagland solution in the well-controlled growth chamber, with a day-length of 14 h; light/dark temperatures, 23/16 o C; and relative humidity, 65% at Zhejiang University, China. The rst fully expanded leaf was collected to extract DNA for further investigations.

Measurement of β-amylase activity
Grains of each barley genotype were harvested at its respective maturity, dried at 40°C, milled to pass through 0.5 mm sieve, and then store at -20℃ for further use.

DNA extraction
DNA extraction was conducted according to the CTAB (hexadecyltrimethyl ammonium bromide) protocol with the following procedures: approximately 0.5 g leaf tissues were ne-grounded in liquid nitrogen and transferred to a 2 mL centrifuge tube containing 1 mL CTAB buffer (2% w/v CTAB, 1.42 M NaCl, 20 mM EDTA, 100 mM Tris-HCl, and 0.2% β-mercaptoethanol, preheated to 60 o C). After incubated in the water bath (65 o C) for 30 min, the leaf extract was thoroughly mixed with 750 µL chloroform: isoamylol (24:1) solution by vortex shaking and then centrifuged at 12000 rpm for 10 min at 4 o C. The supernatant was transferred to a new tube and treated with chloroform: isoamylol solution to make the DNA extraction purer. Thereafter, the supernatant was then well mixed with 0.7 v/v isopropanol, placed at room temperature for 10 min and centrifuged again at 12000 rpm 10 min at 4 o C. The supernatant was discarded and the pellet was washed with 1000 µl of 70% ethanol and then centrifuged at 12000 rpm for 15 min at room temperature. The cleaned pellet was later dried in air, suspended with 15-20 µl of 0.1M TE, and then stored at -20 o C for the following detection.
Polymorphism analysis of Bmy1 gene, cloning and sequencing of Bmy1 intron III According to Hayden et al. (2008), 12 Simple Sequence Repeat (SSR) marker primers (Table 1) were used to screen Bmy1 diversity. PCR products were analyzed by scoring the presence of a band as 1 and the absence as 0. All the data was imported into SPSS using Hierarchical Cluster Analysis with default settings to conduct cluster analysis (Fig. 1).  In the previous study, we found that the Tibetan wild barley XZ147 was a drought tolerant genotype with the largest increase of β-amylase activity and smallest grain weight loss under water stress compared with other genotypes, and Triumph was a drought sensitive malting barley (Wu et al. 2015). In the present study, therefore, we sequenced the full Bmy1 gene of XZ147 and Triumph to analyze the molecular difference between the wild and cultivated barley (Gong et al. 2013), in order to further identify the mRNA differences in SNP and amino acid substitution based on the cDNA of Huruna Nijo through Clustal W (Table 3).

Statistical analysis
Statistical analysis was performed using SPSS Statistics 20 (IBM, New York, NY, USA). Hierarchical cluster analysis of 125 genotypes based on SSR markers was conducted with the method of averagelinkage-between-groups. Two-way variance analysis (ANOVA) was carried out to evaluate the signi cance among barley genotypes (G), drought treatments (E) and the interaction between genotype and drought treatment (G × E).

Results
Polymorphism of Bmy1 DNA SSR markers based on Bmy1 intron III DNA sequence divided the cultivated barley into numerous small groups (Fig. 1), and the wild accessions were scattered over these groups. Unfortunately, there was no outstanding cluster found in the present study. Thus, based on our previous ndings, XZ147 and Triumph were selected from the 125 barley genotypes for further analysis (Wu et al. 2015).

Bmy1 gene intron III alleles
Four INDELs of Bmy1 intron III alleles, 126-bp 38-bp 11-bp and 21-bp were identi ed in this study (Table 2). According to different combinations of the INDELs, the 125 barley accessions used in this study were classi ed into three allele-types: Bmy1.a, Bmy1.b and Bmy1.c, with no Bmy1.d allele being detected. Among these three allele-types, Bmy1.c group occupied the largest proportion of the barley accessions (consisting of 36 cultivated and 19 wild barley accessions), followed by Bmy1.b (consisting of 10 cultivated and 26 wild barley accessions) and Bmy1.a (consisting of 27 cultivated and 7 wild barley accessions). It was surprising that half of the wild barley were classi ed into group Bmy1.b, but most of the cultivated barley were classi ed into group Bmy1.a and c.

Polymorphism of Bmy1 cDNA and amino acid composition
In this study, six SNPs and corresponding amino acid substitution were identi ed based on the alignment of cDNA between wild barley XZ147 and cultivated barley Triumph (Table 3). Wild barley XZ147 showed a great difference from Triumph and Huruna Nijo mRNA and amino acid composition, especially in D165E (495C→D) and V430A (1289T→C).

Discussion
Marker-assisted selection (MAS) has already been widely used in breeding, phyletic evolution, comparative genetics and gene mapping (Prasad et al. 2000;Stein and Graner 2004;Varshney et al. 2005). Su cient evidences have demonstrated the usefulness of MAS in accelerating breeding e ciency (Crepieux et al. 2005;Koebner and Summers 2003). In this study, SSR assay of Bmy1 DNA sequences showed abundant polymorphisms of the gene (Fig. 1). Genetic cluster and the distribution of β-amylase activity (  . The allele-type with the presence of 11-and 21-bp, and the absence of 126-and 38-bp INDELs was de ned as Bmy1.d, which was once identi ed in a wild barley PI 2976897 (Erkkilä et al. 1998). In this study, no Bmy1.d was detected in all barley genotypes. The impact of drought on β-amylase activity in grains varied greatly between genotypes. All genotypes belonging to Bmy1.b, and the wild barley accessions belonging to Bmy1.a and Bmy1.c showed an increase in β-amylase activity, while the cultivated barley genotypes belonging to Bmy1.a and Bmy1.c showed a decrease in β-amylase activity. Bmy1 intron III has been reported to be a useful marker in barley breeding for selecting high malt quality (Kaneko and Kihara 2000) and the 126-bp INDEL is closely correlated to β-amylase activity and thermo-stability (Coventry et al. 2003;Erkkilä et al. 1998;Erkkilä 1999;Kaneko et al. 2000;Gunkel et al. 2002). The 126-bp fragment in Bmy1.a allele-type may be a site of negatively regulated transcription factor, and could be linked with the low β-amylase activity (Erkkilä and Ahokas 2001). The current result also con rmed that determining Bmy1 intron III allele-type could help predict β-amylase activity under drought stress. In this study, all genotypes of Bmy1.b allele-type showed an increase in grain β-amylase activity, while those of Bmy1.a and Bmy1.c allele-types showed less changes, indicating that the genotypes belonging to Bmy1.b allele-type could be more useful in developing the barley cultivars with drought tolerance and high malt quality.
It has been documented that some amino acid substitutions derived from speci c SNPs of mRNA in Bmy1 gene were highly correlated with β-amylase activity in barley grains (Chiapparion et al. 2006;Filichkin et al. 2010;Ma et al. 2001;Zhang et al. 2007). Zhang et al. (2007) and Ma et al. (2001) reported that 115 R→C amino acid substitution was the main reason for high β-amylase activity of W127 and Ashqeleon. Chiapparino et al. (2006) found that the genotypes containing amino acid composition of C115, E165 and V233 had higher β-amylase activity than the ones with the composition of R115, D165 and A233. Based on the alignment of Bmy1 gene DNA, Filichkin et al. (2010) identi ed two genotypes only differing in V233A, with the genotypes A233 having higher β-amylase activity than the genotype V233. R115 or C115 alone had no effect on β-amylase activity, but their co-existence increased β-amylase activity and thermo-stability distinctly (Clark et al. 2003;Ma et al. 2001). In this study, β-amylase activity was higher in cultivated barley Triumph than wild barley XZ147 under normal condition, which may be attributed to the amino acid composition of C115, D165 and V233. Surprisingly, β-amylase activity was higher in XZ147 than Triumph under drought stress, which might result from its unique M527 or the composition of R115, D165, A233, S347 and M527. In addition, drought-induced increase in the abundance of β-amylase might also cause such increase in its activity. Therefore, the association of the amino acid M527 or the composition of R115, A233, S347 and M527 allele with the increase in β-amylase activity under drought stress, and their potentiality A233 in breeding barley cultivars with both drought tolerance and high β-amylase activity in grain need to be further studied.

Conclusion
Novel genetic variation is essential for successful breeding. In this study, based on the comparison of genetic variation of Bmy1 gene from both cultivated and wild barley, we evaluated the potential value of wild barley in developing cultivars with drought tolerant and high β-amylase activity. Introns play important roles in the functioning of genes or genomes, participate in the regulation of auto-catalytic activity of IREs (Intronic Regulatory Elements), and the mobility of endonucleases and reverse transcriptases (Lewin 2004); however, this still needs further study (Yutzey et al. 1989). In plants, introns regulate the transcription of genes, such as, in rice (Fiume et al. 2004) and Arabidopsis (Gazzani et al. 2003;Sheldon et al. 2002). In this study, β-amylase activity of Bmy1.b allele-type accessions was increased by drought treatment, indicating its higher implication value in malt barley breeding than the other two allele-types. Meanwhile, wild barley belonging to Bmy1.b allele type displayed higher β-amylase activity and minimum yield lose under drought stress (Wu et al. 2015). This might result from its unique amino acid substitution M527 or the amino acid composition of R115, D165, A233, S347 and M527. We conclude that the gene pool of Tibetan wild barley germplasm may provide a unique resource for improving drought tolerance and grain β-amylase activity of malt barley.

Declarations
Con ict of Interest Figure 1 Genetic cluster of 125 barley accessions by SSR markers.