Genetic Mapping of ms1s, a Recessive Gene for Male Sterility in Common Wheat

The utilization of heterosis is an important way to improve wheat yield, and the production of wheat hybrid seeds mainly relies on male-sterile lines. Male sterility in line 15 Fan 03 derived from a cross of 72,180 and Xiaoyan 6 is controlled by a single recessive gene. The gene was mapped to the distal region of chromosome 4BS in a genetic interval of 1.4 cM and physical distance of 6.57 Mb between SSR markers Ms4BS42 and Ms4BS199 using an F2 population with 1205 individuals. Sterile individuals had a deletion of 4.57 Mb in the region presumed to carry the Ms1 locus. The allele for sterility was therefore named ms1s. Three CAPS markers were developed and verified from the region upstream of the deleted fragment and can be used for ms1s marker-assisted selection in wheat hybrid breeding. This work will enrich the utilization of male sterility genetic resources.


Introduction
As a major food crop, common wheat (Triticum aestivum L., AABBDD) plays a vital role in ensuring the food security of human beings. Development of wheat semi-dwarf varieties during the 1950s and 1960s improved lodging resistance and permitted significant increases in yield that led to a stable global food supply and eliminated the threat of hunger. These events became known as the "Green Revolution". However, with arable land currently decreasing due to urbanization, environmental degradation, and rising population levels, there is an emerging need to further increase wheat yield. Heterosis, a common biological phenomenon widely used in maize (Zea mays L.) and rice (Oryza Sativa L.), offers a way to increase wheat yield [1,2]. Despite decades of research, the use of male sterility for the production of hybrid wheat has achieved only limited success. Wilson and Ross (1962) introduced the nucleus of common wheat into the cytoplasm of Triticum timopheevii to develop a common wheat T-type male-sterile line, and fertility restorer lines (T-type restorer lines) were obtained by breeding and selection [3]. Thus, a three-parent hybrid system was devised. Subsequently, new types of sterility systems, e.g., K-and V-type cytoplasmic male sterility systems, a photo-thermal sterility system, and chemical male sterilizing agents were described. However, critical biological and economic problems hindered each system, and hybrid wheat as a commercial enterprise never exceeded 0.2% of the total production [4,5].
In 1931, Bovicini first discovered the phenomenon of nuclear male sterile in wheat, and, to date, five nuclear male sterility loci have been identified, i.e., Ms1, Ms2, Ms3, Ms4, and Ms5 [6][7][8][9][10]. Sterility alleles ms1 and ms5 are recessive and located on chromosome 4BS and 3AL, respectively [11,12], whereas Ms2, Ms3, and Ms4 are dominant. Ms2 and Ms4 are located on chromosomes 4DS, whereas Ms3 is on chromosome 5AS [7,8,10,13,14]. Many sterile alleles of the Ms1 gene have been described. These presumably arose as natural mutants, such as Pugsley's male sterile (ms1a, [6,15]) and LZ (ms1g, [16]); as X-rayinduced mutants, such as Probus male sterile (ms1b, [17]) and Cornerstone (ms1c, [18]); or as EMS induced mutants, including FS2 (ms1d), FS3 (ms1e), and FS24 (ms1f ) [10,11]. Recently, 12 isomorphic variants of Ms1 gene, i.e., ms1d.1, ms1d.2, ms1h, ms1i, ms1j, ms1k, ms1l, ms1m, ms1n, ms1o, ms1p, and ms1q, from the spring wheat variety Ningchun 4 were identified following EMS treatment [19]. Many scientists believe that nuclear male-sterile lines are ideal lines for the production of hybrid varieties, but a lack of stable male-sterile maintainers prohibits their widespread application. Although a "two-line system" [20] and an "XYZ hybrid" system using the nuclear male-sterile mutant "Cornerstone" were proposed, neither proved successful [21,22]. Huang et al. (1988Huang et al. ( , 1991) discovered a nuclear male-sterile mutant in the offspring of a cross of wheat varieties 72,180 and Xiaoyan 6 and created a blue-aleurone nuclear male-sterile maintainer system (BM system) by introducing a 4E chromosome from Agropyron elongatum containing a dominant blue endosperm gene and a fertility restorer gene [23,24]. The seeds of the wheat 4E monomer alien addition line are light blue. This line can produce seeds normally by selfing, and the offspring can separate about 64% non-blue seeds (nuclear male sterile), 16% light blue seeds, 16% medium blue seeds, and 4% deep blue seeds (fertile). All plants grown from non-blue seeds were male sterile, whose fertility could be restored well in F 1 by almost any variety of bread wheat. The light blue seed is used as a maintainer for male sterility. Zhou et al. (2006) also introduced the same 4E chromosome into Lanzhou Mutant 257A (ms1g) and established a recessive nuclear male-sterile 4E-ms system [25]. This solved the difficulties in obtaining stable maintainer lines and producing homozygous male-sterile seeds in large quantities. It was a significant step forward in the applied research of wheat nuclear male sterility.
In this study, we mapped a nuclear male-sterile gene in line 15 Fan 03 (derived from the blue-aleurone nuclear male-sterile maintainer system) and found a deletion of chromosome fragment within the Ms1 locus. We developed molecular markers for MAS to speed up hybrid breeding if a suitable system were to be developed.

Sterile Plants in Line 15 Fan 03 Cannot Form Mature Pollen
Compared to the fertile line Mian 07-374, the glumes of sterile plants in 15 Fan 03 had a larger opening angle opened for a longer time, were translucent, and had shorter filaments causing the anthers to remain in the glume. Staining with 1% I 2 -KI showed that pollen grains from sterile plants were yellowish-brown and malformed, demonstrating abnormal development and lack of fertility ( Figure 1). The anthers of sterile and fertile plants were observed anatomically at different stages, and no significant differences were observed in tapetum cell structure and microspore development during the mononucleosis stages; however, degradation became evident at the tri-nuclear stage ( Figure 2). Furthermore, at the tri-nuclear stage, the sterile line had less pollen grains compared to the fertile line; this was observed by scan electron microscopy (SEM) (Figure 3). There were no discernible differences in the surface pattern on the anther walls between the two lines; however, the inner anther walls of pollen grains in sterile plants were smooth and lacked grainy ubisch bodies, whereas granulated ubisch bodies separated by similar cellular structures were densely distributed on the inner walls of pollen from the fertile line. Pollen grains from the fertile line were plump and near-spherical with a bulged ring protruding from the germination hole. By contrast, the aborted pollen grains from sterile plants were collapsed, irregular, and deformed; the germination hole was caved-in. This implies that sterile pollens are not fully developed and cannot pollinate normally.
line. Pollen grains from the fertile line were plump and near-spherical with a bulged ring protruding from the germination hole. By contrast, the aborted pollen grains from sterile plants were collapsed, irregular, and deformed; the germination hole was caved-in. This implies that sterile pollens are not fully developed and cannot pollinate normally.   line. Pollen grains from the fertile line were plump and near-spherical with a bulged ring protruding from the germination hole. By contrast, the aborted pollen grains from sterile plants were collapsed, irregular, and deformed; the germination hole was caved-in. This implies that sterile pollens are not fully developed and cannot pollinate normally.

The Nuclear Male Sterility Gene is Located on Chromosome 4BS
The F1 generation of the cross between a male-sterile plant in 15 Fan 03 and Mian 07-374 showed 100% fertility. The F2 population derived from F1 plants segregated 892 fertile: 313 sterile, conforming with a 3:1 Mendelian ratio (χ 2 = 0.458, p = 0.498) and indicating that sterility was conferred by a recessive gene ( Figure 4).

The Nuclear Male Sterility Gene Is Located on Chromosome 4BS
The F 1 generation of the cross between a male-sterile plant in 15 Fan 03 and Mian 07-374 showed 100% fertility. The F 2 population derived from F 1 plants segregated 892 fertile: 313 sterile, conforming with a 3:1 Mendelian ratio (χ 2 = 0.458, p = 0.498) and indicating that sterility was conferred by a recessive gene ( Figure 4).   DNA from the pooled samples of 15 Fan 03 (male-sterile line), Mian 07-374 (fertile line), F 2 individuals with 0% seed set (homozygous recessive, msms), and F 2 individuals with more than 99% seed set (heterozygous and homozygous dominant, Msms and MsMs) was used as a template for whole-genome SNP genotyping using the Wheat660 SNP chip. Of a total of 9042 SNPs detected homozygous in both 15 Fan 03 and 0% seed set pool, 10.05% (909) were located on chromosome 4B, whereas 1118 (12.36%) were located on chromosome 5B. Less than 7% were located on each of the remaining chromosomes, and the locations of 115 (1.27%) were not predictable ( Figure 5a). For chromosome 4B, 22.86% of the SNPs were located in the 0-150 Mb region, and 9.15% were in the 500-600 Mb region; for chromosome 5B 10.72% of the SNPs were located in the 640-730 Mb region (Figure 5b,c). These results predicted that the gene responsible for male sterility could be on chromosome 4B or 5B.
To further map the gene-regulating male sterile of 15 Fan 03 line, we used 364 individuals from the F 2 population and developed SSR markers in the predicted physical regions of chromosomes 4B and 5B. SSR markers Ms4BS42, Ms4BS54, Ms4BS390, and Ms4B118 located at the terminal region of chromosome 4BS were significantly associated with sterility, whereas the markers for chromosome 5B were not ( Figure 6). The genetic map positioned the sterility locus between SSR markers Ms4BS42 and Ms4BS54 in a genetic interval and physical distance of 4.27 cM and 14.23 Mb, respectively. Based on the Chinese Spring Reference Genome (IWGSCWGA v1.0), these markers were physically located on chromosomes 4BS at 10.61 and 24.84 Mb, respectively.
To narrow down the interval of the sterility locus, 1205 F 2 individuals were screened by SSR markers Ms4BS42 and Ms4BS54, and 184 recombinants were identified. New SSR markers developed from the target region allowed the locus to be fine-mapped between Ms4BS42 and Ms4BS199 at a genetic distance of 1.4 cM and a physical distance of 6.57 Mb. Based on IWGSCWGA v1.0, markers Ms4BS42 and Ms4BS199 were located on chromosomes 4BS at 10.61 and 17.18 Mb, respectively. Using the remaining polymorphic markers Ms4BS132, Ms4BS199, Ms4BS234, and Ms4BS236, the genetic distance of target region reduced. However, the physical distance based on IWGSCWGA v1.0 increased, indicating that the region between Ms4BS132 and Ms4BS236 of 15 Fan 03 line is significantly different from the Chinese Spring reference genome sequence.

Sterility in Line 15 Fan 03 is Caused by a Chromosome Deletion
Many newly developed chromosome-specific 4BS markers in the mapped interval were dominant makers. Therefore, Wheat660K SNP chip results of the fertile pool and the sterile pool of F2 generation were used, and 874 SNPs were detected in the 4B chromosome of the fertile plant pool but not in the sterile plant pool. Of these, 134 SNPs were clustered in the 10-15 Mb region of 4BS (Figure 7a). We speculated that a chromosome segment might be lost in this region.
Specific primers for 16 genes located close to the sterility gene were designed to verify the hypothesized deletion ( Table 1). The first 10 sequentially distributed genes were not amplified in the sterile pool and sterile F2 individuals. According to the physical  Specific primers for 16 genes located close to the sterility gene were designed to verify the hypothesized deletion (Table 1 and Table S1). The first 10 sequentially distributed genes were not amplified in the sterile pool and sterile F 2 individuals. According to the physical locations of the genes, it was predicted that a fragment of 5 Mb would be missing from the target region. The upstream break was between TraesCS4B02G015400LC and TraesCS4B02G015200 (physical locations 10.79 and 11.43 Mb, respectively), and the downstream break was between TraesCS4B02G021300 and TraesCS4B02G021600 (physical locations 15.41 and 15.60 Mb, respectively). Fluorescence in situ hybridization (FISH) was carried out to further verify the results. TraceCS4B017400 was used as a probe to hybridize with the sterile line 15 Fan 03, fertile line Mian 07-374, and homozygous recessive sterile and dominant fertile F 2 individuals (with homozygous and heterozygous individuals confirmed by SSR markers Ms4BS42 and Ms4BS54). Mian 07-374 and homozygous fertile F 2 individuals had two hybridization signals at the end of chromosome 4BS, heterozygous F 2 individuals had one signal, and sterile plants from 15 Fan 03 and sterile F 2 individuals had no signals (Figure 7c). We concluded that sterility was caused by deletion of a substantial fragment from chromosome 4BS.
In order to narrow down the location and size of the deleted fragment, a number of SSR markers were developed using IWGSCWGA v1.0 as reference. Primers MS13, MS19, MS571, MS572, and MS577 were amplified in all four pooled samples, whereas MS41, MS42, MS43, MS366, MS367 showed amplification only in Mian 07-374 and homozygous dominant fertile F 2 individuals. Based on the physical location of the SSR markers on chromosome 4BS and the deletion interval, the upstream boundary of the deleted fragment was between TraesCS4B025400LC and the marker MS41, physically located at 10.789 and 10.838 Mb, respectively. The downstream boundary was predicted between the gene TraesCS4B021300 and TraesCS4B02G021600, which are physically located at 15.409 and 15.596 Mb, respectively.
In previous studies [19,26], the dominant fertile allele at locus MS1 was mapped to position 13.126 Mb on chromosome 4BS. This position falls within the deleted fragment found in sterile individuals of 15 Fan 03. Continuing the naming of mutant alleles at this locus causing sterility, the present allele is named ms1s.

Breeding Molecular Marker Development for ms1s
The The ms1sCAPS1 cleaved an amplified 361 bp region into 232 bp and 129 bp fragments; ms1sCAPS2 cleaved an amplified 347 bp product into 232 bp and 115 bp fragments; and ms1sCAPS3 cut a 395 bp fragment into 232 bp and 163 bp fragments (Figure 8).

Discussion
Application of nuclear male sterility is one way to study heterosis in wheat. Although there are many reports on nuclear male sterility in wheat, only 5 loci Ms1-Ms5 have been identified and physically mapped. Sterility-causing alleles of Ms1 and Ms5 are recessive, and those of Ms2, Ms3, and Ms4 are dominant. Three of them have been isolated and cloned. The dominant Ms2 sterility allele in Taigu contains an originally unexpressed "orphan" gene in its promoter region, which is activated and specifically expressed in anthers causing male sterility [13]. Tucker et al. (2017) and Wang et al. (2017) cloned the wildtype allele of Ms1 and found that it encodes a lipid transport protein expressed only in microspores [19,26]. Pallotta et al. (2019) demonstrated that Ms5 encodes a lipid transfer protein anchored by glycosylphosphatidylinositol, which is required for normal development of the pollen outer wall [12].
The flowering phenotype of sterile plants in 15 Fan 03 differs from that in the fertile plants of Mian 07-374, as the glume remains open for a longer time at a wider angle, and at the tri-nuclear stage, the pollen grains degrade and the pollen activity is lost. However, genetically, the sterile line 15 Fan 03 was similar to ms1b and ms1g mutants in possessing Pooled DNA from random F 2 individuals and heterozygous F 2 individuals produced three fragments after the restriction enzyme Bgl II treatment, of which ms1sCAPS1 produced 361 bp, 232 bp, and 129 bp; ms1sCAPS2 produced 347 bp, 232 bp, and 115 bp; and ms1sCAPS3 produced 395 bp, 232 bp, and 163 bp fragments. These results confirmed the functionality of the three CAPS markers for the selection of male sterility allele ms1s.

Discussion
Application of nuclear male sterility is one way to study heterosis in wheat. Although there are many reports on nuclear male sterility in wheat, only 5 loci Ms1-Ms5 have been identified and physically mapped. Sterility-causing alleles of Ms1 and Ms5 are recessive, and those of Ms2, Ms3, and Ms4 are dominant. Three of them have been isolated and cloned. The dominant Ms2 sterility allele in Taigu contains an originally unexpressed "orphan" gene in its promoter region, which is activated and specifically expressed in anthers causing male sterility [13]. Tucker et al. (2017) and Wang et al. (2017) cloned the wildtype allele of Ms1 and found that it encodes a lipid transport protein expressed only in microspores [19,26]. Pallotta et al. (2019) demonstrated that Ms5 encodes a lipid transfer protein anchored by glycosylphosphatidylinositol, which is required for normal development of the pollen outer wall [12].
The flowering phenotype of sterile plants in 15 Fan 03 differs from that in the fertile plants of Mian 07-374, as the glume remains open for a longer time at a wider angle, and at the tri-nuclear stage, the pollen grains degrade and the pollen activity is lost. However, genetically, the sterile line 15 Fan 03 was similar to ms1b and ms1g mutants in possessing interstitial deletions of 4BS chromosome segments bearing the Ms1 locus.
Wheat nuclear male-sterile lines have many advantages. For example, they are easily restored by most common wheat varieties, with significant heterosis, and without the adverse effect of exogenous cytoplasm. Furthermore, it is easy to select excellent hybrid combinations. The main problem with this type of male-sterile line is that it is difficult to obtain a stable male-sterile maintainer line; that is, it is difficult to produce a large number of homozygous male-sterile seeds. However, using the blue-aleurone nuclear male-sterile maintainer system (BM system), we can easily obtain wheat male-sterile lines (white grains, such as 15 Fan 03) and maintainer lines (light blue grains). We obtained 11 combinations of BM systems with modern main varieties. Moreover, white and blue seeds can be separated at the seed stage, so it is easier to obtain pure male-sterile lines for the production of hybrid wheat [23,24]. This solves the difficulties in obtaining stable male sterile maintainer lines and producing a large number of homozygous male sterile seeds. Moreover, the easy restoration of these lines by common wheat varieties has opened up a new direction for the application of and research into wheat nuclear male sterility.
Because the genome of common wheat is large, complex, and heterogeneous, many chromosomal variations are preserved in the species. These chromosomal abnormalities may also alter the plant phenotype. For example, the absence of the end of chromosome 5AL causes reduced plant height [28]. In this study, a deleted chromosomal fragment was found in sterile plants, and the positions of some markers were in opposite orientations in genetic and physical maps (Figure 6), indicating that chromosomal inversion might have occurred in the parental material. The upstream position of the deletion was between TraesCS4B02G015400LC and marker MS41, whereas the downstream position was between TraesCS4B021300 and TraesCS4B02G021600. Using this information and the wheat660K SNP chip data, three CAPS markers were developed and validated on the F 2 pooled DNA and heterozygous F 2 individuals, as well as 12 widely grown modern wheat varieties. These CAPS markers can be used to genotype individuals in segregating populations at an early stage of growth. Since an SNP marker is usually developed from a sequence polymorphism between different cultivars, it is background genotype specific. We are currently exploring other types of molecular markers that might be more suitable for marker-assisted selection of ms1s.

Plant Material
An F 2 population of 1205 individuals derived from a cross between 15 Fan 03 (white grain, a nuclear male-sterile line, derived from the mutant in the offspring of a cross of wheat varieties 72,180 and Xiaoyan 6, which introduced a 4E chromosome from Agropyron elongatum containing a dominant blue endosperm gene and a fertility restorer gene) [24] and Mian 07-374 was used for genetic analysis, linkage mapping, and gene localization. At least three spikes from each plant at heading were bagged, and the seed-setting rate was calculated as follows: Seed setting rate = (seeds from the two florets of the spikelet base)(no. of spiklets) × 2. The seed setting rates lower than 10% represent sterile individuals; the seed setting rates higher than 10% correspond to fertile ones.

Morphological Observations
15 Fan 03, Mian 07-374, and the F 2 population were planted in 1.5 m rows, with plantto-plant and row-to-row spacings of 10 cm and 25 cm, respectively, at the experimental farm of Institute of Genetics and Developmental Biology in 2017. Anthers at the mature tri-nuclear stage were sampled and stained with 1% I 2 -KI solution for 10 min to determine the pollen fertility.
Anthers for histological studies were excised at different developmental stages, fixed in FAA solution for 24 h, sliced into 4 mm sections, stained with 1% aniline saffron and 0.5% aniline fast green, and observed under bright field using a Nikon Eclipse Ci-S microscope. For scanning electron microscopy (SEM), anthers at the tri-nuclear pollen stage were dried using a CO 2 dryer (Hitachi Critical Point Dryer HCP-2) for 30 mins, and the golden surface of the dried anthers was studied using a Hitachi S-3000N electron microscope.

Wheat660K SNP Chip Analysis
Genomic DNA from 15 Fan 03 (male-sterile line), Mian 07-374 (fertile line), and the F 2 population was extracted using a plant genomic DNA extraction kit (DP305, Tiangen Biochemical Technology Co., Ltd., Beijing, China). Four DNA pools, separately derived from 10 sterile plants of 15 Fan 03, the fertile line, F 2 individuals with more than 99% seed set, and F 2 individuals with 0% seed set, were pooled and scanned by the Wheat660 SNP chip for genome-wide SNP analysis [29]. Sequences flanking SNPs were blasted against IWGSCWGA v1.0 to predict physical locations on the Chinese Spring Reference Genome. The SNPs identified in the four DNA pools were statistically analyzed using Microsoft Excel.

SSR Marker Development and Genetic Map Construction
Genomic DNA of the F 2 individuals was extracted using the CTAB method [30]. SSRLocator software [31] was used to develop SSR markers based on IWGSCWGA v1.0. The parameters of primers were set at 18-23 bp, annealing temperature (Tm) 55-63 • C, GC content 50-60%, and amplified DNA sequence length 100-280 bp ( Table 2). All the markers developed within the mapping interval were used for genotyping. The linkage map was constructed using JoinMap 4.0 software [32] and MapChart 2.30 software [33].

Fluorescence In Situ Hybridization
Root tips were prepared for fluorescent in situ hybridization (FISH) as described by Han et al. [34]. The genomic DNA of TraesCS4B01G017400 was amplified using the following primer pair: forward, 5 -GGGCTGGGAAAGAGCA-3 and reverse, 5 -CCCCAAGATACGTGTTCC-3 . It was labeled with Alexa Fluor-488-dUTP (green) as needed. Data Availability Statement: The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.