Introduction

Wheat male sterility has been widely used in traditional and hybrid breeding. Male sterility is common in angiosperms (Kaul et al. 1988), it refers to a condition in which flower anthers display partial or complete abortion without any diminution of pistil development. Two types of male sterility, sporophytic and gametophytic, have been documented (Yamagata et al. 2007). Sporophytic male sterility is caused by developmental defects in sporophytic tissues (Wilson et al. 2001), while gametophytic male sterility is caused by developmental defects in pollen grains (Durbarry et al. 2005). In wheat, eight genic male sterility (GMS) loci have been reported (McIntosh et al. 2013) and the genes associated with three of them have been cloned, i.e., ms1 (Tucker et al. 2017; Wang et al. 2017), Ms2 (Ni et al. 2017; Xia et al. 2017), and ms5 (Pallotta et al. 2019).

Ms2 is a dominant gene conferring a ‘non-pollen type’ and complete male sterility (Deng and Gao 1982; Ni et al. 2017). F1 seeds obtained from crossing will segregate in a 1:1 ratio between male-sterile and male-fertile progeny (Deng and Gao 1982) guaranteeing that Ms2 plants will always be heterozygous. In 1980s, Liu et al. developed a dwarf male-sterile wheat by linking the dominant dwarf gene Rht-D1c (Rht10) and Ms2 (collectively called RMs2) (Liu and Yang 1991). Both Ms2 and RMs2 have been widely used in wheat breeding programs in China (Yang et al. 2009). Until 2009, more than 30 wheat cultivars and 66 lines have been released by the Ms2 and RMs2 breeding systems in China (Zhai et al. 2009). In 2017, Ni et al. cloned the Ms2 gene using map-based cloning and this work was soon confirmed by an independent study (Xia et al. 2017). Ms2 is a gain-of-function allele due to a retrotransposon insertion in the promoter of the recessive ms2 gene (Ni et al. 2017). Ms2 represents a novel protein that interacts with eukaryotic initiation/elongation factors, GTP-binding proteins, and seven ribosomal proteins in yeast-two-hybrid assays (Ni et al. 2017). These interactions with the cellular translation machinery may be the cause of Ms2 male sterility.

In China, the Ms2-derived male-sterile and fertile plants were comparable in appearance and in anthesis (Deng and Gao 1980, 1982). Other than its exclusive use in China, there is no report of the Ms2 gene effect in non-Chinese wheat germplasm. To extend the utilization of Ms2 gene, its effect on agronomic traits of PNW wheat varieties/lines was investigated in this study.

Materials and methods

Plant materials and growth conditions

Two male-sterile wheat lines, ‘Xiaoyan 6 + Ms2’ (‘XY6Ms2’) and ‘LM15RMs2’ (Ni et al. 2017), were used as donors for the Ms2 gene introgressions. Four PNW wheat varieties/lines, including ‘01-10704A’, ‘08-00802B’, ‘LCS Artdeco’, and ‘Puma’, were used as recurrent parents to generate respective BC2F1 hybrids. Winter wheat ‘Brundage’ was used as explant for the genetic transformation of Ms2. The BC2F1 plants were grown on the Parker Farm (Moscow, Idaho, USA) in the 2019–2020 growing season. For each genetic background, the recurrent parent, Ms2 hybrids and RMs2 hybrids were each planted in neighboring rows that were 25 cm apart and 1.5 m long. Materials were planted in October 2019 and data were collected in July 2020. The transgenic plants and wild-type (WT) ‘Brundage’ were grown in greenhouse of the University of Idaho (UI) under a 16-h photoperiod with a day temperature of 22–25 ℃ and a night temperature of 15–20 ℃.

Genetic transformation with the Ms2 gene

The wheat Ms2 gene was excised from the plasmid PC976 (Ni et al. 2017) and ligated into a high copy plasmid, PC414C, using the restriction enzymes AscI and NotI. The resulting plasmid, PC414C-Ms2, was amplified in E. coli, and the Ms2 fragment was gel purified after cutting with AscI and NotI. The Ms2 fragment and the BAR (bialaphos resistance) gene on PC174 (Zhang et al. 2019) were co-transformed into the ‘Brundage’ using biolistic bombardment (Lv et al. 2014). The treated immature embryos were selected in media supplemented with 3 mg·L−1 bialaphos (Gold Biotechnology, St Louis, MO, USA). Putative transgenic plants were tested for their resistance to 340 ppm glufosinate-ammonium (Finale Herbicide; Bayer, Leverkusen, Germany).

Genotyping of transgenic plants

Plant DNA was extracted from leaf tissue (Yuan et al. 2012), quantified on an ND-1000 spectrophotometer (ThermoFisher Scientific, Waltham, MA, USA), and diluted to 100 ng ul−1. PCR markers, HT5 and X26 (Ni et al. 2017), were used to genotype the Ms2 gene in transgenic plants. PCR primers were 5’-GGCTCTGATACCAAATGTTGTTG-3’ plus 5’-CGTAGATGCGGACCCAGGGGAT-3’ for HT5, and 5’-CAAATTCCATCTCACCGATCTCTT-3’ plus 5’- ATGGTGGTGTGCCCCTAAAAAG-3’ for X26. PCR conditions were 95 ℃ for 5 min, 38 cycles of 95 ℃ for 30 s, 65 ℃ for HT5 (or 60 ℃ for X26) for 30 s, and 72℃ for 60 s for HT5 (or 30 s for X26), and a final extension at 72 ℃ for 7 min. PCR products were separated on a 1% agarose gel and visualized by ethidium bromide staining.

Collection and analysis of agronomic traits

For the BC2F1 and transgenic T2 plants, we recorded plant height, spike length, spikelets per spike and fertility status at the late-anthesis stage. For spikelets, we only counted the well-developed ones, and excluded the basal and terminal spikelets that were poorly developed. The seed-setting rates per spike measured the fertility of the first two primary florets in the well-developed spikelets in an open pollination environment. Only the primary tiller, normally the tallest and strongest shoot of each plant, was used for measuring plant height and seed-setting rates. We used SAS 9.4 (SAS Institute Inc., Cary, NC, USA) to process the collected data, which involved the use of GLM and UNIVARIATE procedures. Analysis of Variance (ANOVA) was performed for significance test.

Results

Conventionally transferred Ms2/RMs2

In the BC2F1 generation, plants segregated into the male-sterile and male-fertile groups, as the Ms2 (or RMs2) and ms2 genetic groups. Plant height was compared among different groups. Under all four genetic backgrounds, the Ms2 (Ms2ms2) and ms2 (ms2ms2) genetic groups were not significantly different from the WT ms2 group (ms2ms2, recurrent parent) (P > 0.05) (Table 1)—apart from Ms2 of ‘Puma’ (4.2 cm difference). Except for the 8.1 cm difference of ‘08-00802B’ (P < 0.05), the Ms2 and ms2 genetic groups in BC2F1 were comparable to each other (P > 0.05) (Table 1). Therefore, Ms2 had no essential effect on plant height in the tested wheat. In contrast, due to the Rht-D1c gene, the RMs2 groups in BC2F1 were always dwarf, extremely shorter than any other genetic groups (P < 0.01) (Table 1). This is well characterized by the fact that while the average plant height of all four Ms2 genetic backgrounds was 82.4 cm that of the RMs2 was only 44.6 cm. So, the Rht-D1c gene itself caused a 45.9% reduction in plant height averagely.

Table 1 Plant height and the spike traits in BC2F1
Full size table

Other agronomic traits including spike length, spikelets per spike and seed-setting rate were further analyzed. For spike length, the Ms2 and ms2 as well as RMs2 genetic groups were not significantly different in each of the four wheat genetic backgrounds (P > 0.05) (Table 1). However, the spike length of the Ms2 and/or RMs2 groups in BC2F1 was somewhat more reduced than those in the WT ms2 groups of ‘01-10704A’, ‘08-00802B’, and ‘Puma’ lines (P < 0.05)—these differences ranged from 0.9 to 1.4 cm. It may require a few more backcrosses to compensate this spike length difference.

Spikelets per spike are an important yield component. In the BC2F1 plants, the Ms2 and ms2 genetic groups were not significantly different in each of the four wheat genetic backgrounds (P > 0.05); and the RMs2 linkage did not impact the spikelet numbers compared to the ms2 group (P > 0.05) (Table 1). In comparison to the WT ms2 group of the recurrent parent, the spikelet numbers per spike of the Ms2 and RMs2 groups in BC2F1 were only reduced in ‘01-10704A’ (P < 0.05)—15.5 and 13.4% differences, respectively. From this, it is concluded that the Ms2 gene does not essentially affect spikelet numbers per spike in common wheat.

To utilize Ms2 GMS system in a large-scale hybrid seed production, it is important to know how it affects the seed-setting rate in an open pollination environment. In this study, the seed-setting rate was only based on the first two florets of the well-developed spikelets in a spike. All spike heads of the BC2F1 and the recurrent parents were left open for possible cross-pollination in the field. In BC2F1, the Ms2 gene caused male-sterility, but when the Ms2-containing plants were allowed to cross pollinate, they displayed an average seed-setting rate of 79.4% in the Ms2 group and 60.2% in the RMs2 group as compared to the average (95.0%) of the ms2 group (Table 1). Apparently, Ms2-based male-sterile plants are able to produce seeds in an open-cross environment on a relatively high percentage. However, the seed-setting rates of the Ms2 and RMs2 groups were significantly reduced when compared to the ms2 group in BC2F1 and in the recurrent parents (P < 0.05) (Table 1). Due to the Rht-D1c gene a further significant reduction was also observed (19.2% difference between the averages of the Ms2 and RMs2 groups).

Genetically transformed Ms2

To validate the effect of Ms2, we transformed the dominant Ms2 gene into the PNW winter wheat ‘Brundage’. In total, 2190 immature embryos of ‘Brundage’ were treated, and 32 embryos survived in the selection media supplemented with bialaphos. From 11 immature embryos, we obtained 11 putative T0 plants, of which four plants, PT1 to PT4 each from an independent embryo, exhibited the classic male sterility of the Ms2 genotype and were positive for the dominant Ms2 gene (Fig. 1). Three transgenic T0 plants, PT2–PT4, were also resistant to the Finale herbicide (Fig. 1).

Fig. 1
figure 1

Characteristics of the T0 transgenic ‘Brundage’. a Florets from putative transgenic T0 plants displayed male-sterile (left) and male-fertile (right) phenotypes. Scale bar = 1 mm. b Putative transgenic T0 plants were resistant (R) or susceptible (S) to Finale herbicide as it shown in leaf segments. Scale bar = 1 cm. c PCR test of three putative transgenic T0 plants. PT2, PT3, and PT4 were positive for the dominant Ms2 gene. Black arrows mark specific band of a dominant Ms2 gene. WT represents the wild-type ‘Brundage’

Full size image

Pollen grains of the WT ‘Brundage’ were used to pollinate the male-sterile transgenic lines (PT2–PT4). As expected, nearly half male-fertile and half male-sterile plants were obtained in both T1 and T2 generations, and the male-sterility trait was inherited from T0 to T2 generations. The transgenic plants and the WT ‘Brundage’ were genotyped using the Ms2-specific markers: X26 and HT5 (Fig. 1). This clearly distinguished the dominant Ms2 group in the male-sterile T2 segregants from the recessive ms2 group. At the T2 generation, we compared the two genetic groups of each transgenic line. For plant height, spike length, and spikelets per spike, although there were slight differences between the Ms2 and ms2 groups in each line, these differences were insignificant (P > 0.05) (Table 2).

Table 2 Plant height and spike traits of the T2 transgenic ‘Brundage’
Full size table

To compare with the WT ‘Brundage’ (the WT ms2 group), we pooled the Ms2 group data and the ms2 group data of PT2, PT3, and PT4. All data sets for plant height, spike length, and spikelets per spike data met the ANOVA assumptions (P > 0.05). Again, no significant differences were present for plant height, spike length, and spikelets per spike among the three genetic groups: WT ms2, T2 Ms2, and T2 ms2 (P > 0.05) (Fig. 2a–c). When developing heads in greenhouse-grown plants were bagged, the T2 Ms2 group had zero seed set (Fig. 2d). In contrast, the WT ms2 group had a 96% seed set, and the T2 ms2 group had an 81% seed set (Fig. 2d, Supplemental Fig. 1), very similar to that of Ms2 group in BC2F1. Taken together, the dominant Ms2 gene conferred complete male sterility in transgenic wheat, while having insignificant effects on traits of height, spike length, and spikelet numbers per spike.

Fig. 2
figure 2

Plant height and spike development in the T2 transgenic ‘Brundage’. Boxplots of the plant height (a), spike length (b), spikelets per spike (c), and the seed-setting rate (d). Transgenic plants segregated in the T2 generation into the dominant Ms2 group and the recessive ms2 group. ‘Brundage’ (or BD) was included as the non-transgenic WT control, which contains only the recessive ms2 gene. For each genetic group, the box plot depicts the range of values shown by the vertical lines, the interquartile range (25–75 percentile) shown by the shaded box, the group means shown by the diamond symbol, and the group medians shown by the horizontal lines within each box. Outliers were plotted separately as cycles on the chart

Full size image

Discussion

The male sterility gene Ms2 is a valuable tool in wheat breeding. Although the Ms2 gene has been extensively studied in China, how well this gene might function in non-Chinese wheat germplasm remained unknown. In this study, the Ms2 gene and/or RMs2 linkage were backcrossed into four PNW SWW wheat lines and transferred into an additional one by biolistic bombardment. BC2F1 generation and T2 transgenic plants were investigated to illustrate the effect of Ms2 on agronomic traits.

Plant height, spike length and spikelet numbers per spike are important traits that directly or indirectly contribute to yield potential. The yield components of wheat can be multifaceted (Slafer et al. 2014); however, grain number per spike is always an important parameter. In winter wheat, elite varieties showed on average 38% more yield compared to genetic resources, and this yield increase is mainly contributed by a 19% increase in grains per spike, but with limited gain (4%) in thousand grain weight (Philipp et al. 2018). Wheat yield largely depends on the final numbers of well-developed spikelets and grains per spike which are established prior to flowering (Würschum et al. 2018). In this study, both BC2F1 plants and transgenic plants together proved that the Ms2 locus has no essential effect on plant height and spike-related traits, which provides a powerful breeding tool without masking the yield potentials in wheat breeding. Due to seed or space limitations, we did not include replication for each genotype; however, multiple backcross and transgenic lines unanimously agreed with each other, which served as a broad-sense replication within and between treatment groups.

The linkage of the dominant dwarfing gene Rht-D1c and the GMS gene Ms2 (RMs2) (Liu and Yang 1991) is very useful for large-scale hybrid seed production, as this dwarfing gene helps the identification of tall male-fertile plants from dwarf male-sterile ones. However, our observations showed that the Rht-D1c gene had an adverse effect on natural seed-setting rate of the male-sterile spikes in an open pollination environment, as the Ms2-positive BC2F1 plants had a 79% natural seed-setting rate, but the RMs2-positive ones had only a 60%. This difference probably due to the delayed anthesis (3–5 days) in RMs2 lines. It has been reported that Rht-D1cgenes have effects on wheat heading and flowering time under different backgrounds (Sun et al 2011). It might also come from the extreme dwarfing nature of the Rht-D1c; the dwarf plants may have less chance to capture pollen grains due to little air flow around the ground. Besides, the adverse effect of this dwarfing gene was also observed in non-male sterile background, as it significantly reduced grain yield (Alghabari et al. 2015) and root dry mass (Wojciechowski et al. 2009) as compared to various near-isogenic lines of winter wheat carrying various Rht alleles. Although the dwarf male-sterile wheat RMs2 system has been widely used in wheat breeding programs in China, however, due to the above adverse effects of Rht-D1c dwarfing gene the Ms2 system might be more practical for cross-pollination than the RMs2 one.

In addition to its use as a recurrent selection tool in conventional breeding, the Ms2 gene can also be designed for hybrid wheat breeding and hybrid seed production (Ni et al. 2017). To date, third-generation hybrid seed systems have been developed by integrating the use of male fertility/sterility genes and fluorescence marker genes in cereal crops (Chang et al. 2016; Zhang et al. 2018; Qi et al. 2020). A similar wheat hybrid system could also be developed by integrating the dominant Ms2 gene and the natural blue aleurone gene (Li et al. 2017). There were breeders trying to create “Blue-AiBai” lines to stack RMs2 and the blue aleurone gene via chromosome engineering. In 2004, Pu et al. reported a blue RMs2 (collectively called BRMs2) carrying the blue aleurone gene and RMs2 linkage on an additional chromosome. This monosomic wheat could be utilized in hybrid wheat breeding and hybrid seed production, but it has a main disadvantage that the inheritance of monosomic chromosome carrying the BRMs2 is only around 22% (Pu et al. 2004). More efficient or high-throughput system could be developed by using cisgenics or transgenics methods to create BRMs2 or the blue aleurone gene and Ms2 linkage (BMs2). We think that last version, i.e., without the use of Rht-D1c gene, would be more economical solution for large-scale hybrid production due to the higher seed-setting rate on the male-sterile mother line.