Background

Pre-harvest sprouting (PHS), a phenomenon that physiologically matured kernels germinate on their mother plants before harvesting due to high humidity [1], is a major constraint for grain yield and end-use quality of wheat (Triticum aestivum L.) [2]. Global estimate of the annual losses directly caused by PHS was about $1 billion [3]. In China, 83% of wheat planting areas suffer from PHS, mainly in the Middle and Lower Yangtze River Valleys Winter Wheat Zone (MLWZ), Southwestern Winter Wheat Zone (SWWZ) and Northeastern Spring Wheat Zone (NSWZ) [4]. In recent years, PHS frequently occurred in the Yellow and Huai River Valleys Winter Wheat Zone (YHWZ), resulting in serious reductions in grain yields in 2013, 2015, and 2016 [5]. Thus, resistance to PHS has become an important target trait in many wheat breeding programs.

Wheat grain color is associated with PHS, with better tolerance to PHS for the red-grained wheat cultivars than for the white-grained ones [6]. However, white-grained cultivars had higher flour yield than red-grained wheat cultivars at the same grade of whiteness of flour, which makes them more popular for millers [7, 8]. Most Chinese prefer the shine and white color of products, mainly including steamed bread, boiled dumpling and noodle. The production acreages of white-grained wheats increase due to the strong market demand. Among the 75 wheat cultivars with the annual planting area > 667,000 hectares during the last several decades in China, 61 are white grains [9]. Hence, improvement of white-grained wheat with strong resistance to PHS is important for expanding its production in order to meet the market demands.

Resistance to PHS is a complex quantitative trait that is controlled by both genetic factors and external environmental factors [10]. Marker-assisted selection (MAS) on PHS resistance can not only shorten breeding cycles but also enhance selection efficiency. To date, more than 42 quantitative trait loci (QTL) governing PHS resistance are cataloged in wheat, most of which are associated with grain color and seed dormancy [2, 11, 12]. The transcription factor Tamyb10 is a candidate gene of grain color-related gene R, and three homoeologous genes Tamyb10-A1, Tamyb10-B1, and Tamyb10-D1 were identified [9, 13,14,15]. Gene TaDFR-B affects grain color and PHS resistance by controlling anthocyanin synthesis, and the TaDFR-Bb allele was tightly associated with PHS resistance in TaDFR-B [16].

Seed dormancy is another genetic factor affecting PHS resistance [17, 18]. Gene TaMFT, a wheat homolog of TaPHS1 related to seed dormancy, was cloned from red-grained cultivar Zen [19], and white cultivar Rio Blanco [20]. Five SNP mutations or InDels associated with low germination index (GI) were characterized in the promoter and coding regions of TaMFT [19, 21,22,23]. TaMKK3-A, previously known as Phs1, on chromosome 4A is another major gene governing seed dormancy [18]. The PHS-resistant and -susceptible cultivars carry the TaMKK3-Aa and TaMKK3-Ab alleles in TaMKK3-A, respectively [24]. The seed dormancy-related genes TaSdr, TaVp-1, and TaGASR34 were cloned by means of the homology-based cloning approach [25,26,27,28,29], and the corresponding functional markers have been developed.

Resistance to PHS can be assessed under field and controlled environmental conditions [23]. Field evaluation of PHS resistance needs suitable weather conditions including humidity and temperature, hence, it is difficult to obtain consistent phenotypic data in different years or environments. By contrast, controlled environmental evaluation of PHS resistance is relatively easy, and phenotypic data of PHS can be repeated in other environments. Sprouting rate of whole spikes and seed germination test are two of the main methods under controlled environment [30, 31]. Calculation of the GI by testing seed germination is the most direct approach to detect seed dormancy, and was widely used to evaluate PHS resistance in previous studies [27,28,29]. However, improvement of PHS resistance based on phenotypic selection is time-consuming and labor-intensive.

The reliable molecular markers are a prerequisite for MAS, but PHS resistance and effects of allelic variation and haplotypes in known genes affecting PHS resistance in winter wheat cultivars are less studied. The aims of this study were to 1) evaluate PHS resistance in a set of winter wheat cultivars in China; 2) identify the allelic variation and haplotypes of 10 PHS resistance genes and compare the effects of contrasting alleles at each gene, haplotypes of TaMFT, and allelic combinations on PHS resistance; and 3) validate the effects of PHS resistance allele and haplotype under the genetic background of a PHS-susceptible wheat cultivar.

Materials and methods

Plant materials

An array of 326 Chinese winter wheat cultivars released from 1964 to 2020, including 19 cultivars from the Northern Winter Wheat Zone (NWWZ), 270 from the YHWZ, 21 from the MLWZ, and 16 from the SWWZ, were used in this study. These cultivars originated from 13 provinces, i.e., Beijing (14), Ningxia (1), Shanxi (5), Tianjin (1), Xinjiang (2), Hebei (60), Shandong (50), Henan (128), Shaanxi (10), Anhui (5), Hubei (4), Jiangsu (30), and Sichuan (16). All cultivars were approved by the National Genebank (Beijing, China) and wheat breeders. Seeds of these cultivars were originally acquired from the National Genebank, Beijing, or breeders. The detailed information of each cultivar is listed in Table S1.

The F2 (206 plants) and derived F3 (206 lines) populations from the cross Lunxuan 13 (TaMFT-Hap4/TaSdr-B1b, susceptible to PHS) × Bainong 3217 (TaMFT-Hap1/TaSdr-B1a, resistant to PHS) were used to validate the PHS resistance of alleles in TaSdr-B1, haplotypes of TaMFT, and their combinations. Both Bainong 3217 [Funo/Neixiang 5//Xiannong 39/3/Xinong 64(4)3/Yanda 24] and Lunxuan 13 (Shimai 12/Zhoumai 16//Zhoumai 16) are white-grained cultivars but differ in the resistance to PHS.

Field trials

The whole set of wheat cultivars were planted at the experimental station of the Chinese Academy of Agricultural Sciences (CAAS) in Xinxiang (35°31′N, 113°85′E), Henan province, during the 2017–2018, 2018–2019, and 2019–2020 wheat cropping seasons. All cultivars were arranged in one-row plots of 2-m length and 0.25-m width with 40 seeds. The soil of field is a typical clay loam. Fertilization, irrigation, and other field managements were carried out as described previously [32]. The meteorological data of daily average temperature (°C), relative humidity (%) and precipitation (mm) of the three cropping seasons is showed in Fig. S1.

Germination index assay

Resistance of the wheat entries to PHS was assessed using the seed GI method for three years from 2018–2020. The GIs of 206 F2 plants and derived F3 lines from the Lunxuan 13 × Bainong 3217 population were assessed in 2019 (F2) and 2020 (F3), respectively. The F2 plants were individually harvested and separately evaluated of GI. About 30 spikes were harvested from each cultivar and F3 line at the physiological maturity stage (about 35 d after flowering), air dried at ambient temperature (~ 25 °C) for 3 d, hand-threshed, and stored in a refrigerator at -20 °C. Fifty seeds were sterilized with 5% NaClO, evenly embedded on two layers of filter paper in Petri dishes (15 cm in diameter), and incubated in a growth cabinet at 25 °C for 7 d with 50 mL of sterile water. Germinated seeds were counted daily and removed. Germination index was calculated according to the method described by Zhang et al. [27]. This experiment was carried out thrice.

Genotyping

The alleles of 10 genes associated with PHS resistance were determined using the functional markers as described in Table S2. For the sequence tagged site (STS) and cleaved amplified polymorphic sequence (CAPS) markers, a 20 μL reaction mixture was prepared by mixing 1 μL of 50–100 ng μL−1 template DNA, 1 μL each of the forward and reverse primers (10 μM), 10 μL 2 × Taq PCR Master Mix (P111-03, Nanjing Vazyme Biotech Co. Ltd., Nanjing, China), and 7 μL of sterilize ddH2O. Amplification of DNA was performed in a SimpliAmp thermal cycler (Thermo Fisher Scientific (China) Co., Ltd., Shanghai, China). The Kompetitive Allele-Specific PCR (KASP) assays were carried out in a 5 μL reaction mixture including 2.5 μL PARMS SNP master mix (GTE001-2, Wuhan Genetides Biotech Co., Ltd., Wuhan, China), 0.056 μL primer mix, 0.04 μL Mg2+, 2.2 μL template DNA (20–50 ng μL−1) and 0.204 μL ddH2O using a BIO-RAD S1000 Thermal Cycler PCR System (Bio-Rad Laboratory Inc., Hercules, CA, USA). Genotyping of KASP markers was performed following the protocol as described by Rasheed et al. [33].

Data analysis

Analysis of variance (ANOVA) for the GI values of 326 winter wheat cultivars over three years was performed using the PROC GLM program in the Statistical Product and Service Solutions (SPSS) software 22.0 (International Business Machines Corporation, Armonk, New York, USA) [34]. Phenotypic comparison between white-grained and red-grained wheats, and differences of PHS resistance between contrasting alleles of each gene were determined by the t-test in SPSS software 22.0. Multiple comparisons (PROC GLM) for the GI values of wheat cultivars from different wheat zones or provinces and phenotypic differences among haplotypes or allelic combinations (ACs) were performed using Tukey–Kramer at P < 0.05 in SPSS software 22. The broad-sense heritability (h2) and correlation coefficients between years were estimated according to the method described by Li et al. [35].

Results

Phenotypic evaluation on PHS resistance

The mean squares of genotypes, years and genotype × year interaction were significant as shown by ANOVA (P < 0.01) (Table S3). The broad-sense heritability (h2) of GI was 0.96. There were significant differences of GI in different years, and the mean GI values of the 326 cultivars were 48.9%, 59.2%, and 34.5% in 2018, 2019 and 2020, respectively. The wide range of phenotypic variation in GI was observed in each of the three years (Fig. S2a). A total of 43, 10 and 86 cultivars showed the GI values lower than 25.0% in 2018, 2019 and 2020, respectively (Fig. S2a, Table S4), and 8 cultivars had stable PHS resistance across the three years.

The difference in GI was significant among the four wheat zones, in which MLWZ had the lowest GI value in each year (Fig. S2b). The cultivars from Jiangsu and Shaanxi provinces showed better PHS resistance than those from the other provinces (Fig. S2c). Grain color was associated with PHS resistance (Fig. S2d), and the red-grained cultivars had lower mean GI than the white-grained cultivars (P < 0.05). The GI values for the 326 cultivars measured in different years were significantly correlated with a range of correlation coefficients from 0.69 to 0.79 (Fig. S3).

Allelic and haplotypic frequencies

Allelic variation of 9 genes and haplotypes of TaMFT in the 326 cultivars were determined (Table S4). The mean frequencies of alleles conferring low GI and high GI were 36.4% and 63.6%, respectively (Fig. 1). The low-GI alleles had higher frequencies than the corresponding high-GI alleles in TaSdr-A1, TaVp-1B, TaDFR-B and TaMKK3-A. Among them, the frequency of TaSdr-A1a allele only was over 60.0%. Nine main TaMFT haplotypes (Hap1-Hap9) were identified based on the genotypes at the positions- 194,- 222, + 219, + 646, + 666 in the 326 winter wheat cultivars (Table 1). Haplotypes Hap4 (29.4%) and Hap6 (25.8%) had the higher frequencies than the other haplotypes, whereas the frequency of Hap1 with five PHS resistance allele was only 3.4% (Fig. 1).

Fig. 1
figure 1

Allelic and haplotypic frequencies of 10 genes associated with pre-harvest sprouting resistance in the 326 winter wheat cultivars. Purple and blue boxes indicate pre-harvest sprouting (PHS) resistance and susceptibility alleles/haplotypes, respectively. The allelic composition of TaMFT haplotypes refers to Table 1

Full size image
Table 1 Main TaMFT haplotypes identified in the 326 wheat cultivars
Full size table

Allelic effects and haplotype analysis

Cultivars carrying the low-GI allele showed better PHS resistance than those carrying the high-GI allele in each of the 9 genes identified (Table 2). Among them, cultivars with the low-GI alleles TaSdr-B1a and Tamyb10-D1b had significantly lower GI values than those with the contrasting high-GI alleles in TaSdr-B1 and Tamyb10-D1 in the three years, respectively (P < 0.05). The allele Tamyb10-D1b had the largest phenotypic effect on PHS resistance at the single gene level. Compared to Tamyb10-D1a, the allele Tamyb10-D1b decreased GIs by 10.4%, 14.9% and 10.6 in 2018, 2019 and 2020, respectively.

Table 2 Comparison of germination index (GI) between contrasting alleles at each gene in the 326 winter wheat cultivars for the three years
Full size table

Haplotypes Hap1-Hap8 with at least one PHS resistance allele at five positions exhibited lower GIs than the Hap9 with five PHS susceptibility alleles (Fig. 2). Among them, Hap1 showed stable and better PHS resistance than the other haplotypes in each individual year. Compared to haplotype Hap9, Hap1 had significantly lower GIs, and averagely decreased by 32.1%, 28.5%, and 23.3% in 2018, 2019, and 2020, respectively (P < 0.05).

Fig. 2
figure 2

Effects of TaMFT haplotypes on germination index (GI) in 2018 (a), 2019 (b), 2020 (c), and mean value (d). “ + ” and “–”, the PHS resistance allele and susceptibility allele, respectively. Different letters indicate significance of GI among haplotypes at P < 0.05

Full size image

Effects of allelic combinations

Among the 10 genes associated with PHS resistance analyzed, 3 genes had significant difference in GIs between the contrasting alleles or among haplotypes. Hence, these genes were used to analyze effects of the ACs. A total of 14 major ACs were detected in the 326 cultivars (Fig. 3). There was a significant difference in the GI values among the AC classes. Compared to the AC14 with the high-GI alleles at the three genes, four ACs (AC1, AC2, AC5, and AC8) had significantly lower GIs in at least two environments and the values decreased by the ranges of 17.0% (AC8)-38.2% (AC1) in 2018, 13.2% (AC8)-40.3% (AC1) in 2019, and 9.9% (AC8)-28.9% (AC1) in 2020.

Fig. 3
figure 3

Comparison of germination index (GI) among 14 main allelic combinations (ACs) in the three pre-harvest sprouting (PHS) genes in the three years. Purple and blue boxes indicate pre-harvest sprouting (PHS) resistance and susceptibility alleles, respectively. Different letters after GIs in each year indicate significant differences at P < 0.05

Full size image

Validation of effects of Hap1 and TaSdr-B1a

Based on the haplotypic analysis of TaMFT, 206 homozygous plants comprised of 113 Hap1 and 96 Hap4 were detected from the 495 plants in the Lunxuan 13 × Bainong 3217 F2 population. Fifty-six plants carried the homozygous TaSdr-B1a genotype and 53 plants carried the homozygous TaSdr-B1b genotype in TaSdr-B1. These F2 plants and derived F3 lines were evaluated for their GIs. To avoid confusion, the heterozygous plants or lines were excluded.

The GI values varied in 206 F2 plants (Fig. S4a) and F3 (Fig. S4b) lines. More than 90% F2 plants and F3 lines showed significantly lower GIs than the PHS-susceptible parent Lunxuan 13. Thirteen F2 plants and five F3 lines were not significantly different in GI values from the PHS-resistant parent Bainong 3217.

Significant phenotypic difference in GIs was found between the progenies with different TaMFT haplotypes and TaSdr-B1 alleles in the Lunxuan 13 × Bainong 3217 population (Fig. 4). The Hap1 progenies reduced the GI values by 13.5% and 14.4% compared to those with the haplotype Hap4 in the F2 and F3 populations (Fig. 4a), respectively (P < 0.05). Consistently, the progenies with the low-GI allele TaSdr-B1a showed significantly lower GI values than those with the high-GI allele TaSdr-B1b (P < 0.05) (Fig. 4b). Compared to the PHS-susceptible parent Lunxuan 13, the progenies with the Hap1 and TaSdr-B1a reduced the GI values by 29.9% and 27.1% in the F2 population, respectively, and 40.8% and 40.0% in the F3 population, respectively (P < 0.05). The AC TaMFT-Hap1/TaSdr-B1a had the smallest GIs among the four ACs in the F2 (Fig. 5a) and F3 (Fig. 5b) populations. In comparison with Lunxuan 13, the progenies with the TaMFT-Hap1/TaSdr-B1a genotype reduced the GI values by 32.7% and 44.3% in the F2 and F3 populations, respectively (P < 0.05).

Fig.4
figure 4

Comparison of germination index (GI) between two haplotypes of TaMFT (a) and contrasting alleles of TaSdr-B1 (b) in the Lunxuan 13 × Bainong 3217 F2 and F3 populations. Different letters in F2 and F3 indicate significance of GI at P < 0.05, respectively

Full size image
Fig.5
figure 5

Comparison of germination index (GI) among allelic combinations of TaMFT and TaSdr-B1 in the Lunxuan 13 × Bainong 3217 F2 (a) and F3 (b) populations. Different letters indicate significance of GI at P < 0.05

Full size image

Discussion

Performance of PHS resistance in winter wheat cultivars

ANOVA and broad-sense heritability analysis suggested that the genetic variation of PHS resistance was mainly controlled by genotype, but year and genotype × year interaction also affected PHS resistance. The consistent finding was reported in the previous study [23]. Temperature is a major external determinant on seed dormancy during seed development in wheat [36]. High temperatures during seed development reduces the level of seed dormancy [19]. Late flowering delayed sampling and continuous high temperature (Figure S1) occurred at late maturity stage in 2019, which might result in reducing the depth of seed dormancy and making all wheat cultivars with higher GIs than other years. In all 10 genes identified, the Low-GI alleles/haplotype had better PHS resistance than the corresponding high-GI alleles/haplotype in each individual year (Table 2), suggesting that these Low-GI alleles/haplotype had stable effects on PHS resistance in different environment, and they can be used to improve wheat PHS in different breeding programs.

A wide variation in the GI values was observed in the winter wheat cultivars examined, but most of them were susceptible to PHS (Fig. S2a, Table S4). This is attributed to the fact that modern wheat cultivars have been domesticated by human in order to improve their adaptation and productivity, so they have relative uniform and rapid germination ability [37]. The PHS-resistant cultivars identified will be useful as donors for PHS improvement.

The red-grained cultivars showed usually more resistant to PHS than the white-grained ones [6, 15]. All the red-grained cultivars examined proved to carry at least one resistance allele from the Tamyb10-A1, Tamyb10-B1, and Tamyb-D1 genes. Furthermore, these cultivars mainly adapted to the MLWZ and SWWZ where PHS occurs frequently during harvesting seasons due to wet weather conditions [4]. This might make wheat cultivars more tolerant to PHS in those wheat areas by natural and artificial selections.

Comparison of PHS resistance of resistance allele, haplotype and allelic combination

Cultivars with the low-GI allele showed higher PHS resistance than those with the contrasting high-GI allele for each gene examined, which is in agreement with the previous studies [16, 18,19,20,21,22, 26,27,28,29]. However, the level of PHS resistance between the contrasting alleles at certain genes varied in different genetic backgrounds [38]. For example, a main-effect locus TaMKK3-A associated with PHS resistance was identified in different populations and explained 30–38% phenotypic variations [10]. The difference of GI between alleles TaMKK3-Aa and TaMKK3-Ab was 15.5% in the Tutoumai A/NW97S186//NW97S186 BC2 population [10], but the corresponding value was only 2.3%, 1.8% and 1.7% in 2018, 2019, and 2020, respectively (Table 2) in this study. This might be attributed to the impacts of genetic backgrounds.

It is noteworthy that the resistance allele TaPHS1-222C always present in the TaMFT haplotype Hap1, but absent in the other haplotypes. Wang et al. [39] also found that the TaPHS1-222C allele was consistently present with TaPHS1-646G and TaPHS1-666A alleles in haplotype GCA. Furthermore, there was no phenotypic difference in GI between the allele TaPHS1-222C and haplotype Hap1 in this study (data not shown). This suggests that the TaPHS1-222C marker is effective to select Hap1 genotypes in the process of MAS.

Among the 14 ACs, AC1 carrying Hap1 and TaSdr-B1a showed smaller GI than the other ACs (Fig. 3), and the pyramiding effect was further verified in the Lunxuan 13 × Bainong 3217 population (Fig. 5). Even though the Hap1 and Tamyb10-D1b alleles had the largest phenotypic effects on GIs at a single locus level, the effects of pyramiding two alleles are not clear because no cultivar carries both alleles in this study.

The known PHS resistance genes were not detected in the low-GI cultivars such as Yangmai 20 (11.6%), Lunan 11 (13.8%), and Luomai 4 (15.5%), demonstrating that these cultivars may carry new genes associated with the low GI. It warrants genetic analysis to dissect the QTL for their low GI performance.

Distribution of the low-GI alleles and allelic combinations

The haplotype Hap1 with the largest effect on PHS resistance was detected only in 11 cultivars (Table S4), suggesting that it is a rare haplotype in modern wheat cultivars. The low frequencies of the Hap 1 were also reported in Chinese accessions (2.79%) [39] and landraces (2.0%) from the Fertile Crescent and surrounding areas [40], and wheat accessions from the USA (24.4%) [40]. Hence, it is necessary to introgress the Hap1 into PHS-susceptible cultivars by molecular marker of TaPHS1-222C allele due to tightly association between the TaPHS1-222C allele with the other low-GI alleles [39]. Another allele Tamyb10-D1b also showed low frequency (8.6%), and mainly present in the red-grained cultivars (Fig. 2). In China, white-grained cultivars have been preferentially selected than red-grained cultivars by wheat breeders [6], which might result in low frequency of this allele.

Validation of effects of low-GI allele and haplotype on resistance to PHS

Lunxuan 13 is an elite high-yielding wheat cultivar, but susceptible to PHS [41]. We tried to introgress the low-GI Hap1 and allele TaSdr-B1a from the PHS-resistant parent Bainong 3217 to Lunxuan 13. The PHS resistance of the progenies were significantly enhanced. This indicates that they can efficiently improve PHS resistance of a susceptible cultivar. Pyramiding of Hap1 and TaSdr-B1a showed lower GIs than those carrying single haplotype or allele, suggesting that they have additive effects. Similar result was reported by analyzing the combining effects of TaPHS1 and TaMKK3-A [42]. Even if Hap1 and TaSdr-B1a showed additive effects, there was significant difference in mean GI between the pyramiding progenies and PHS resistance parent Bainong 3217, suggesting that Bainong 3217 might carry other unknown PHS-resistant loci.

Conclusions

The comparison of effects between contrasting alleles on GI in single gene combining haplotype analysis showed that the haplotype Hap1 of TaMFT gene had the best PHS resistance. This haplotype can be preferentially used to enhance PHS resistance due to its high effectiveness and low distribution frequency. Combining haplotype Hap1 and the TaSdr-B1a allele in AC1 exhibited additive effects on GIs in winter wheat cultivars and validated in genetic population. This study will facilitate the parental selection and MAS for wheat PHS resistance, and provide important materials for identifying new PHS-resistant genes.