Introduction

Wheat is an important staple food crop worldwide, providing 20.0% of calorie for humans [1, 2]. The quality of wheat-based products has gained increasing attention due to people’s growing demand for nutritious and healthy diets [3, 4]. The nutritional quality of bread can be improved through antioxidant enzyme and non-enzymatic antioxidants [5, 6]. Importantly, superoxide dismutase (SOD) plays a pivotal role in scavenging reactive oxygen species [7]. Superoxide dismutase in wheat grains can tremendously enhance dough rheological properties by reducing the production of free sulfhydryl groups [8]. Increased free sulfhydryl groups cause decreased strength of gluten network structure and flour quality [9]. Meanwhile, the elevated SOD activity contributes to improved wheat quality by increasing the content of vitamin C and vitamin E [10]. Therefore, breeding wheat varieties with high SOD activity represents a significant step forward for improvement of wheat quality.

SOD activity in wheat grain is a quantitative trait controlled by multiple genes and also influenced by environments [11]. To date, numerous QTL for SOD activity have been identified in wheat. For example, several SOD QTL were identified in wheat leaf SOD QTL on chromosomes 1B, 1D, 2 A and 6D [11]. Twenty-six SOD genes from the wheat genome were detected by bioinformatics, including Cu/Zn-SODs, Fe-SODs, and Mn-SODs [12]. These results indicate that the SOD activity of wheat is controlled by several major-effect QTL and numerous minor-effect QTL, distributing across wheat chromosomes.

SOD in wheat is associated with protecting plant cells from oxidative damage and maintaining metabolic equilibrium in organisms, but research on the correlation of SOD activity with quality traits has been rarely reported. In particular, there are few reports about SOD in wheat grains and knowledge of corresponding molecular markers, which is limiting breeding practices. There may be a risk to wheat production and resistance due to research limitations on wheat grain SOD activity. The development of functional molecular markers is important for efficient breeding. Therefore, the present study was to map the QTL for SOD activity, predict the potential candidate genes and develop molecular markers for precise marker-assisted selection, which would be served as a reference for future genetic improvement and quality breeding.

Materials and methods

Plant materials

Three hundred and nine F6 RILs developed from the cross of ‘Berkut’ × ‘Workatta’, kindly provided by Dr. Susanne Dreisigacker at the International Maize and Wheat Improvement Center (CIMMYT), were used for QTL mapping. The RILs and parental lines were grown at the Manas experimental Station of Xinjiang Academy of Agricultural Sciences in 2016, 2017, 2018 and 2019 (referred as E1, E2, E3, and E4, respectively). The experiments were conducted in randomized complete blocks with three replications, and the RILs were planted in single 2 m rows spacing in 25 cm between rows. The management of field trials was following local practices with drip irrigation. The grains of RILs were harvested at maturity, and used to extract the SOD enzyme.

Extraction of SOD from wheat grain

The extraction of the crude SOD enzyme was performed following Wang et al. [13]. For each line, 10 g grains were milled into whole wheat flour using a 0.8 mm cyclone mill (laboratory mill 120, Perten, Swedish). The 0.1 g of whole wheat flour was immersed in 1 mL of 50 mmol/L potassium phosphate (pH 7.8) in a 2 mL Eppendorf tube, and subsequently oscillated for 1 min on a mixer oscillator. The homogenized mixture was put into an ice bath shaker for 2 h, then centrifuged at 10,000 rpm for 15 min at 4℃ (Centrifuge 5425R, Eppendorf, USA). Finally, the supernatant was collected as crude SOD enzyme extract.

Determination of SOD activity

SOD activity was measured using the nitro-blue tetrazolium (NBT) photoreduction method [14]. Assay of each sample includes a sample tube and two control tubes. The sample tube contained 10 µL of SOD enzyme extract homogenized in a substrate composed of 150 mL of 0.05 mol/L phosphate buffer solution, 20 µL of distilled water, 30 µL of 130 mmol/L Met solution, 30 mL of 750 µmol/L NBT solution, 30 mL of 100 µmol/L EDTA-Na2 solution and 30 mL of 20 µmol/L solution, whereas the control tube contained the buffer without enzyme extract. One control tube was put in the darkness, while the other was placed under a 4000 lx fluorescent light for 20 min. The control tube in the dark is blank, control tube in the light is absorbance of illumination. The OD (optical density) value was determined at 560 nm using microplate reader (Synergy H1, Biotek, USA). One unit of SOD was defined as the amount of hemoglobin that inhibits the rate of NBT reduction by 50.0%. The SOD activity was calculated using the following formula:

$${\rm{SOD}}\,{\rm{activity}}\,{\rm{U\cdot}}{{\rm{g}}^{{\rm{ - 1}}}}{\rm{ = (}}{{\rm{A}}_{{\rm{ck}}}}{\rm{ - }}{{\rm{A}}_{\rm{E}}}{\rm{) \times V/(}}{{\rm{A}}_{{\rm{ck}}}}{\rm{ \times 0}}{\rm{.5 \times W \times }}{{\rm{V}}_{\rm{t}}}{\rm{)}},$$

where ACK is absorbance of illumination for the control tube, AE is absorbance of illumination for the sample tube, V is total sample volume (mL), Vt is sample volume for determination (mL) and W is sample fresh weight (g). The SOD value of each sample was measured in duplicate to ensure the accuracy of the assay. If the difference between two repeats was more than 10.0%, the experiments were carried out again.

Statistical analysis

The data analysis was performed using Excel 2016 and SPSS 21.0. The broad-sense heritability and variance were analyzed using the QTL IciMapping V4.1 software [15]. The broad-sense heritability (H2) was estimated as \({H^2} = {\sigma _g}^2/({\sigma _g}^2 + {\sigma _{ge}}^2/e + {\sigma _ \in }^2/re)\), where \({{}_{g}}^{2}\) is genotypic variance, \({{}_{ge}}^{2}\) is the variance of genotype × environment interaction, \({\sigma _ \in }^2\) is residual errors, and e and rare the number of environments and replications, respectively.

Linkage map construction

The genomic DNA was extracted from young wheat leaves using a modified cetyltrimethylammonium bromide (CTAB) method. The RILs and parents were genotyped by the wheat 50 K SNP array (the Axiom® HD Wheat Genotyping Array, CapitalBio Technology Inc, Beijing). The polymorphism analysis was performed using the Genomestudiov1.0 software. A total of 11,375 polymorphic SNP markers were obtained after removing markers of distorted segregation (P < 0.001), monomorphic markers, and those with more than 20.0% missing data, and 1,604 bin markers were selected to construct linkage groups. The genetic linkage map was constructed with 1604 bin SNP markers, generating 28 linkage groups on 21 wheat chromosomes, spanning 2220.3 cM with an average genetic distance of 1.38 cM between two bin markers. The genetic mapping was carried out using CARTHAGENE [16].

QTL mapping

The ICIM-ADD function in the QTL IciMapping V4.1 software was used to map QTL for the SOD activity. The LOD (logarithm of the odds) threshold was determined by 1000 × permutations, at a significance level of P < 0.05. The locations of QTL on wheat chromosomes were compared with the Chinese Spring (CS) genome sequence (IWGSC RefSeq v2.1) to obtain the physical positions of QTL. MapChart was used to draw genetic linkage maps [17].

Candidate gene prediction

QTL mapping for the SOD activity was projected for further comparison. The candidate genes involved in the SOD metabolism and regulation pathways were predicted based on the physical positions of QTL flanking markers and the CS reference genome sequences. The flanking sequences of SNP markers were blasted on the NCBI website (http://www.ncbi.nlm.nih.gov/) and the CS reference genome database (http://www.wheatgome.org/).

Results

Phenotypic variation of SOD activities in wheat grains

The SOD activities of Berkut ranged from 1817.49 to 1843.44 U·g− 1 across four environments, whereas those of Worrakatta ranged from 1698.44 to 1711.11 U·g− 1 (Table 1). The mean SOD activities of RILs in four environments were between 1746.99 U·g− 1~1785.97 U·g− 1 (Table 1), exhibiting a normal distribution in the population (Fig. S1). The SOD activities showed significant correlations (P < 0.01) among different environments (Table 1). The coefficients of variation in the RIL population ranged from 3.9 to 7.1% across environments. The broad-sense heritability based on the mean SOD activity was 0.8.

Table 1 Statistical analysis of SOD activities in wheat grains in the RIL population of ‘Berkut’ \(\times\) ‘Worrakatta’
Full size table

QTL for SOD activity

Six QTL for SOD activity were detected on chromosomes 1B, 4D, 5 A (2), and 5D (2), designated QSOD.xjau-1BL, QSOD.xjau-4DS, QSOD.xjau-5 A.1, QSOD.xjau-5 A.2, QSOD.xjau-5DL.1, and QSOD.xjau-5DL.2 (Fig. 1), respectively, explaining 2.2 to 7.4% of the phenotypic variances (Table 2). The favorable alleles for increasing SOD activity at QSOD.xjau-4DS and QSOD.xjau-5DL.2 loci, were derived from the parental line Berkut, whereas those of QSOD.xjau-1BL, QSOD.xjau-5AL.1, QSOD.xjau-5AL.2 and QSOD.xjau-5DL.1 came from another parental line Worrakatta. QSOD.xjau-5AL.2 and QSOD.xjau-5DL.2 were identified in E2 environment and the averaged value of four environments in the RIL population. QSOD.xjau-5AL.2 was mapped on chromosome 5AL in the region of 577.69 ~ 579.98 Mb, flanked by markers AX-111566544 and AX-110511959, explaining 7.1 ~ 7.4% of the phenotypic variances. QSOD.xjau-5DL.2 was located between markers AX-94670615 and AX-111652649 in the 473.90 ~ 481.42 Mb region of chromosome 5DL, accounting for 2.6 ~ 4.0% of the phenotypic variances.

Fig. 1
figure 1

QTL for wheat grain SOD activity identified in the RIL population of ‘Berkut’ × ’Worrakatta’. Note: Markers and QTL are on the right side of the chromosomes, and the genetic positions of the QTL are on the left. The environments where the QTL were detected are shown in different colors (proximal region of linkage groups in red color, the distal region of linkage groups in blue color) on the right of the linkage groups. The flanking markers of QTL are highlighted in bold

Full size image
Table 2 Locations and effects of QTL for SOD activity detected in the RIL population of ‘Berkut’ \(\times\) ‘Worrakatta’
Full size table

Candidate genes

Four candidate genes for QSOD.xjau-4DS, QSOD.xjau-5AL.1, and QSOD.xjau-5DL.1 (2), were identified based on the significant SNPs associated with SOD activity (Table 3). These candidate genes are involved in the SOD metabolism or regulation pathways. TraesCS4D01G059500 for QSOD.xjau-4DS was predicted to be the third type of mammalian superoxide dismutase, an extracellular Cu and Zn superoxide dismutase, referred to as SOD3. TraesCS5A01G371600 and TraesCS5D01G343100LC in QSOD.xjau-5AL.1 and QSOD.xjau-5DL.1 loci, respectively, were found to be superoxide dismutase [Cu-Zn], an important enzyme in cellular oxygen metabolism. TraesCS5D01G299900 for QSOD.xjau-5DL.1, encodes an Ethylene-responsive transcription factor, related with SOD activity in plant salinity tolerance.

Table 3 Candidate genes predicted in the QTL regions
Full size table

Discussion

Determination of SOD activity

Superoxide dismutase is a key enzyme involved in cellular defense against reactive oxygen species in living organisms, and is widely distributed in microorgnisms, plant and animals [18]. Wang et al. [13] reported SOD activity of 1316.83 ~ 2025.14 U·g− 1 in wheat grains, and the coefficient of variations in different environments ranged from 4.3 to 5.2%, indicating that different materials exhibit rich variations of SOD activity among environments. Therefore, accurate determination of SOD activity in wheat grains is important for discovering genes or genetic loci in the wheat genome. The methods for determining SOD activities were described previously, and one method based on SOD matrix -O2- needs an expensive equipment and is not stable [19]. In contrast, indirect assays of SOD activities are frequently used, e.g., photoreduction method based on the NBT reduction [20], pyrogallol autoxidation [21], Cytochrome C reduction method [22] and xanthene / xanthene oxidase (XOD) system [23]. The NBT method is widely used for SOD determination due to its low cost, small sample volume, high degree of automation, and convenient procedure [24]. Particularly, the determination of superoxide anion radical (o2-) and SOD activity using NBT photoreduction are performed mostly in both plants and animals [25, 26]. Whereas, the NBT method has drawback of a relatively poor soluble formazan, and its accuracy is also affected by the protein content in the reaction system, however, the catalase, glutathione peroxidase and co-enzyme Q10 did not interfere with SOD activity determination in NBT assay [24]. In low temperature and low pH, SOD activity was stable, the crude enzyme solution has higher activity [27].

QTL associated with SOD activity in wheat

The SOD activities in crops are conditioned by polygenes [11], exhibiting significant variations among environments [28, 29], and the high variations of SOD activities in the stress due to different sensitivities of lines to stress [30], which is in agreement with the present study. And SOD activities are widely employed to indicate plant seed germination [31, 32], premature senescence [33], water stress [34], and micronutrient [35]. Previously, several seed SOD QTL were identified in wheat on chromosomes 1 A, 1B, 1D, 2D, 3 A, 4 A, 7 A [36]. Here, QTL QSOD.xjau-1BL, QSOD.xjau-4DS, QSOD.xjau-5AL.1, QSOD.xjau-5AL.2, QSOD.xjau-5DL.1 and QSOD.xjau-5DL.2 for SOD activities in wheat grain were identified on chromosomes 1B, 4D, 5 A (2) and 5D (2), respectively. Among these, QTL QSOD.xjau-1BL was identified on chromosome 1BL at region from 665.80 to 668.96 Mb. An unconditional QTL for SOD activity were detected on chromosome 1B (swes119a-ubc857a) in wheat grain during germination [36]. The QTL qSOD.1B.SH for SOD activity was detected on chromosome 1B (xcfd15-xgwm264) in wheat leaves [36]. The QSOD.xjau-5DL.1 was located on the interval 393.06 ~ 400.88 Mb of chromosome 5DL, in agreement with Wang et al. [13] who identified a QTL linked with the marker RAC875_c49940_385 at 399.30 Mb on chromosome 5D in wheat grain by GWAS panel, whereas QSOD.xjau-5DL.1 was at 6.24 Mb to 1.58 Mb that is much distant from marker RAC875_c49940_385. Therefore, QSOD.xjau-1BL, QSOD.xjau-4DS, QSOD.xjau-5 A.1, QSOD.xjau-5 A.2, and QSOD.xjau-5DL.2 identified in the present study are likely to be new loci for SOD activity.

Candidate genes of QTL

The four candidate genes identified in the present study are involved in regulation of SOD dismutase, including three SOD genes and a gene related with SOD activity. TraesCS4D01G059500, TraesCS5A01G371600 and TraesCS5D01G343100LC were predicted to be SOD genes, located on chromosomes 4DS, 5AL and 5DL, respectively. In addition, a Cu / Zn-SOD gene was found to be present in chromosome 4DL (at chromosome region of 405.45 Mb) [12]. TraesCS5D01G299900 found in QSOD.xjau-5DL.1 locus encodes Ethylene-responsive transcription factor. In the MAPK signaling pathway, ERF1b interacts with MYC2, indicating that ERF1b may be a key integration factor in ETH/JA dependent defense regulation of ETH/JA signaling [37, 38]. ERF transcription factors and ethylene along with oxygen dependent signal transduction, play an important role in hypoxic responses [39]. Several QTL on chromosomes 5 A and 5D are associated with SOD, and these QTL are stable across environments. Future studies can focus on these two chromosomes to further dissect the genetic mechanism underlying the activity of SOD in wheat grain.

Wheat lines with high SOD activity

The SOD gene TaSOD showed up-regulation or significantly increased expression levels in wheat after drought and salt stresses [12]. The drought-tolerant genotype had significantly higher level of expression for Mn-SOD and Cu/Zn-SOD than susceptible genotype [40]. In the present study, some F6 RILs developed from the cross of ‘Berkut’ × ‘Workatta’ had high SOD activities in the population. The lines B104, B113, B109, B223, B224, B230, and B106 have higher level of SOD activity (2126.42 ~ 1911.47 U·g− 1) than other RILs. Favorable alleles for six QTL QSOD.xjau-1BL, QSOD.xjau-4DS, QSOD.xjau-5AL.1, QSOD.xjau-5AL.2, QSOD.xjau-5DL.1, and QSOD.xjau-5DL.2 contributed to the high SOD activities in these lines (Table S1). The mean SOD activities of RILs harboring six, five, four, three, two and one favorable alleles are 1954.80 U·g− 1, 1796.92 U·g− 1, 1773.88 U·g− 1, 1770.21 U·g− 1, 1757.59 U·g− 1 and 1737.42 U·g− 1 in four environments of SOD activities. The line B104 had the highest SOD activity of 2126.42 U·g− 1 possessing six favorable alleles from QTL QSOD.xjau-1BL, QSOD.xjau-4DS, QSOD.xjau-5AL.1, QSOD.xjau-5AL.2, QSOD.xjau-5DL.1 and QSOD.xjau-5DL.2; the line B109 containing five favorable alleles from five loci (QSOD.xjau-1BL, QSOD.xjau-5AL.1, QSOD.xjau-5AL.2, QSOD.xjau-5DL.1 and QSOD.xjau-5DL.2) had 1943.16 U·g− 1 of SOD activity; and the line B106 with five favorable alleles at loci of QSOD.xjau-1BL, QSOD.xjau-5AL.1, QSOD.xjau-5AL.2, QSOD.xjau-5DL.1, and QSOD.xjau-5DL.2 in had 1911.47 U·g− 1 of SOD activity; and four favorable alleles at loci of QSOD.xjau-1BL, QSOD.xjau-5AL.1, QSOD.xjau-5DL.1 and QSOD.xjau-5DL.2 were found in the line B222 that had 1898.94 U·g− 1 of SOD activity; and the other lines B104, B109, and B106 also had higher level of SOD activities among the population. The favorable alleles at these loci could be benefited for improvement of SOD activity in wheat breeding, and those lines with favorable alleles of QTL, and high and stable SOD activity can be integrated into wheat germplasm, and be used to improve the abiotic stress tolerance and quality in wheat.

Conclusion

Six QTL for SOD activity were detected on chromosomes 1BL, 4DS, 5AL (2), and 5DL (2), respectively, in the RIL population of ‘Berkut’ × ‘Worrakatta’, explaining 2.2 ~ 7.4% of the phenotypic variances. Among four candidate genes predicted in the region of QTL, TraesCS4D01G059500, TraesCS5A01G371600, TraesCS5D01G343100LC and TraesCS5D01G299900 are annotated as superoxide dismutase [Cu-Zn] 3, superoxide dismutase [Cu-Zn], superoxide dismutase [Cu-Zn] and Ethylene-responsive transcription factor, respectively. Seven wheat lines with high SOD activities can be used as germplasms to improve abiotic stress tolerance and quality in wheat breeding. In the future, fine mapping of SOD candidate genes will further elucidate the genetic dissection of SOD activity in wheat grains.