Cytological and molecular characterizations of a novel 2A nullisomic line derived from a widely-grown wheat cultivar Zhoumai 18 conferring male sterility

A dwarf, multi-pistil and male sterile dms mutant was previously reported by us. However, the genetic changes in this dms are unclear. To examine the genetic changes, single nucleotide polymorphism (SNP) association, chromosome counting, and high-resolution chromosome fluorescence in situ hybridization (FISH) techniques were employed. By comparing tall plants (T) with dwarf plants (D) in the offspring of dms mutant plants, SNP association analysis indicated that most SNPs were on chromosome 2A. There were three types in offspring of dms plants, with 42, 41 and 40 chromosomes respectively. High-resolution chromosome painting analysis demonstrated that T plants had all 42 wheat chromosomes; the medium plants (M) had 41 chromosomes, lacking one chromosome 2A; while D plants had 40 wheat chromosomes, and lacked both 2A chromosomes. These data demonstrated that dms resulted from a loss of chromosome 2A. We identified 23 genes on chromosome 2A which might be involved in the development of stamens or pollen grains. These results lay a solid foundation for further analysis of the molecular mechanisms of wheat male sterility. Because D plants can be used as a female parent to cross with other wheat genotypes, dms is a unique germplasm for any functional study of chromosome 2A and wheat breeding specifically targeting genes on 2A.


INTRODUCTION
Allohexaploid wheat (Triticum aestivum L., 2n = 6x = 42, genomic constitution AABBDD) has a genome from three diploid species: Triticum urartu Thum. (the source of the A genome), Aegilops speltoides (Tausch) Gren. or a closely related species (the source of the B genome), and Aegilops tauschii Coss. (the source of the D genome) (Huang et al., 2002). Because allohexaploid wheat has high level of functional redundancy, it can host a range of diverse whole-chromosome aneuploids (Zhang et al., 2013;Zhang et al., 2017). There are various types aneuploid variations available in wheat, such as nulli-tetrasomic lines, nullisomic lines, monosomic lines, ditelosomic lines, as well as chromosome fragmental deletion lines (Endo & Gill, 1996;Qi et al., 2003). The aneuploid stocks are useful in wheat gene mapping, and especially, genes can be located to a small segment using a series of chromosome fragmental deletion lines (Endo & Gill, 1996;Qi et al., 2003). A set of homozygous chromosome deletion lines were obtained in cv. Chinese Spring (Endo & Gill, 1996). Most of the homozygous chromosome 2A short arm deletion lines are sterile; genes involved in male development are located on the chromosome 2A short arm (Endo & Gill, 1996).
Previously, we reported a mutant of dwarf, multi-pistil and male sterile dms in wheat (Duan et al., 2015;Zhu et al., 2016). Pollen vigor and hybridization tests demonstrated that dms mutant was male sterile. Male sterility and male fertility followed a segregation ratio of 1:3 [D:(T +M) = 1:3] (Duan et al., 2015). However, the genetic basis of this mutant is unknown. The present study is to discover the genetic basis at the cytological and molecular levels. We also identified a set of genes playing potential key roles accounting for male sterility on chromosome 2A.

Plant materials
Mutant dms used in this study was obtained as previously described by us (Duan et al., 2015). All plant materials were planted in our experimental field at Houwang Village, Xingyang City, Henan, P. R. China (34 • 51 N, 113 • 35 E, 49 m above sea level). We have got permission prior to accessing the field site by Henan Seed Company, and the administrator's name is Haolong Yang.

Morphological and histochemical observations
Phenotype of plants, young spikes and stamens of dms were observed at different developmental stages (Zadoks, Chang & Konzak, 1974). Developmental stages of the young spike primordia were referenced as previously described (Vahamidis et al., 2014). The florets with anthers and pistils at different developmental stages were dissected from spikes using an anatomical needle. Anthers were observed using scanning electron microscope (SEM) (SU8010, Hitachi, Tokyo, Japan) as previously described (Jiao et al., 2019). Structural and histochemical observation on the anthers of T and D plants were carried out using a method previously described by Geng et al. (2018).

RNA extraction and mRNA sequencing
Young spikes of T, M and D plants at floret primordium visible stage (stage 7) (Vahamidis et al., 2014) and anthers of T and D plants at 3-nucleate pollen stage (stage 12) (Browne et al., 2018) were sampled for RNA extraction. All the five samples for mRNA sequencing had 3 biological replicates. Total RNAs of fifteen samples were extracted respectively with TRIzol R reagent (TransGen Biotech, Beijing, China). RNA sequencing and basic analysis were carried out in BioMarker Company (Beijing, China).

mRNA data analysis
The mRNA reads of the fifteen samples were aligned to the draft assembly (IWGSC v1.0) of the wheat genome survey sequence (http://www.wheatgenome.org/) using Tophat2 tool software (Kim et al., 2013). Reads distribution on 21 wheat chromosomes of T, M and D plants were analyzed and drawn using Circos tool software (http: //mkweb.bcgsc.ca/tableviewer/). Gene functional annotation was carried out as previously described by us (He et al., 2018). Gene expression levels were estimated by fragments per kilobase of transcript per million fragments mapped (FPKM) (Florea, Song & Salzberg, 2013). Differentially expressed genes (DEGs) between two sample groups were analyzed using DESeq R package (Wang et al., 2010). The FDR < 0.01 (false discovery rate) and FC ≥ 2 were set as the thresholds for significantly DEGs.
Principal components were calculated using the FPKM values (Schulze et al., 2012). For the single gene, heat maps were drawn using software Hem I v 1.0.3.7 according to the FPKM values. Tissue specific expression patterns of wheat genes were analyzed using k-means cluster on the BMKCloud platform (https://www.biocloud.net/). Clustering was performed in R using the k-means function, where k = 8 within the cluster package by Euclidean distance. Wheat tissue type, expression profile data sources and sample description in origin article (Garcia-Seco et al., 2017;Feng et al., 2017;Li et al., 2019) were listed in Table S1.

Single nucleotide polymorphism (SNP) association analysis
Clean reads data of the two bulks (young spikes of T and D plants) were selected for SNPs identification. After the reads were aligned to the Chinese Spring genome, the SNPs were called using module 'HaplotypeCaller' software GATK v3.6 (McKenna et al., 2010). Euclidean distance (ED) algorithm was used to calculate the grade of different SNPs between T and D plants (Hill et al., 2013).

Simple sequence repeat (SSR) analysis
Genomic DNAs of T, M and D plants were extracted from young leaves using cetyltrimethyl ammonium bromid (CTAB)-method (Clarke, 2009). DNA concentration was measured with a DU800 Nucleic Acid Protein Analyzer (Beckman coulter, Fullerton, California, USA). SSR markers distributed on wheat chromosome 2A were used to analyze the genetic polymorphisms among T, M and D plants. SSR primers employed in this study included those of GWM (Röder et al., 1998), WMC (Somers, Isaac & Edwards, 2004), BARC (Song et al., 2005) and GPW (Gupta et al., 2002). PCR amplification reactions were performed using a method previously described by Zhu et al. (2019). PCR products were separated on an 8% PAGE gel with a standard DNA molecular weight marker in the first lane of the gel. The products were run at 60 W for about 1.0 h. Then the gel was removed from the apparatus and stained using the silver staining method (Li et al., 2018b). Primers used for SSR analysis are list in Table S2.

Cytological analysis
Chromosome configurations and high-resolution chromosome painting analysis of plants derived from M plants at metaphase of mitosis. Chromosome samples were prepared as previously described (Du et al., 2017). For chromosome painting, eight single strand oligonucleotides were used to form a modified multiplex probes for karyotype analysis in wheat, which included TAMRA (6-carboxytetramethylrhodamine)-modified oligonucleotides pAs1-1, pAs1-3, pAs1-4, pAs1-6, AFA-3 and AFA-4, and two FAM (6carboxyfluorescein)-modified oligonucleotides pSc119.2-1 and (GAA) 10 . Oligonucleotide probes used for FISH are list in Table S3. The FISH procedure was tested as previously described (Du et al., 2017). Chromosomes were visualized with microscope Olympus BX51 and pictures were captured with SPOT CCD (SPOT Cooled Color Digital Camera). Image analysis was conducted using Photoshop v6.0.

qRT-PCR for mRNAs
Anthers of T and D plants at 3-nucleate pollen stage were prepared for real-time PCR. The experimental samples were consistent with the samples of RNA-seq. The qRT-PCR was performed as previously described by us (He et al., 2018). Reverse transcription was performed with 1 µg RNA using Hifair R II 1st Strand cDNA Synthesis SuperMix (11123ES60, Yeasen, Shanghai, China). Real-time PCRs of mRNAs were performed using Hieff R qPCR SYBR Green Master Mix (1201ES08, Yeasen, Shanghai, China) and a CFX ConnectTM Real-Time System (Bio-Rad, Hercules, CA, USA) following the production instructions. The wheat actin gene was used as an internal control. All primer sequences are listed in Table S4. The gene expression levels were calculated according to the 2 − Ct method (Livak & Schmittgen, 2001). The SPSS version 17.0 software (SPSS Inc., Chicago, IL, USA) was applied for statistical analysis for the qRT-PCR.

Male abortion in D plants of dms
There were three typical phenotypes in the progeny of dms, tall (T), semi-dwarf (M) and dwarf (D) plants (Fig. S1). D plants were male sterile according to the result of I 2 -KI staining (Duan et al., 2015). To discover the morphological causes of the male abortion in D, scanning electron microscope was employed to observe the anther ultrastructure ( Fig. 1). Unlike the fertile T plants, the anthers of D plants were not dehiscent at the trinucleate stage, and no mature pollen grains were released (Figs. 1A and 1B). The outer anther epidermal cells of T plants were arranged neatly, whereas these of D plants were irregular (Figs. 1C and 1D). Moreover, compared with the anthers of the T plants, the inner epidermal of D plants showed aberrant sized Ubisch bodies, which suggested a structural abnormality (Figs. 1E-1H). The pollen grains of T plants were plump and rounded; by contrast, the pollen grains of D plants were extremely shriveled and atrophied (Figs. 1I and 1L). Furthermore, we performed a histological analysis of the anthers in the dms mutant and the wild type. In the initial stages of pollen development (Figs. S2A and S2B), there were no differences between the wild type and dms mutant. The young microspores were normal ( Fig. S2C and S2D). However, compared with the wild type, degenerated pollen grains were observed in the dms mutant at 3-nucleate pollen stage (Figs. S2E and S2F).
The female organs in flowers of D plants developed unwell, most of the pistils were normal, a few were multi-pistils (Fig. S3). D plants were male sterile but female fertile, it can be used as a female parent to cross with other wheat genotypes. We got F 1 seeds from dwarf plants (as female parent) crossed with Guomai 301 (as pollen parent) (Fig. S4).

The most SNPs were on chromosome 2A
In order to clarify the genetic changes in dms, SNP association analysis was carried out.  Table S5). The result implied that the mutation occurred on chromosome 2A. The results were consistent with the sequencing reads distribution on chromosome 2A of young spikes in T, M and D plants (Fig. 3). About 5.24% of the total reads was identified on chromosome 2A in T plants, which was about twice that of M plants (2.86%) (Table S6). and D genomes were used to detect the polymorphism between tall and dwarf plants of dms.
The SSR markers were used to quantitatively amplify the DNA templates of tall, semidwarf and dwarf plants. The amounts of the amplified products of the tall plants (Fig. 4;lane 1,3,5,8) were about double that of the semi-dwarf plants ( Fig. 4; lane 2, 6, 10), and dwarf plants without PCR products (Fig. 4). In another word, the SSR primers specifically amplifying fragments on chromosomes 2A could be used to distinguish the T, M and D plants in the progeny of dms.

Chromosome 2A was absent in D plants
To explore the chromosome number of dms, 42 plants derived from M plants were investigated. Among them, 17 plants had 42 chromosomes, 21 had 41 chromosomes and 4 had 40 chromosomes (Table 1). After chromosome analysis, the seedlings were planted in field. The 17 plants with 42 chromosomes showed regular plant height, pistil and fertility, the 21 plants with 41 chromosomes showed medium plant height and regular pistil and fertility, while all the 4 plants with 40 chromosomes showed dwarf status, multi-pistil and complete male sterility (Fig. S7).
To validate and identify the chromosome constitution of dms, high-resolution chromosome painting was applied using eight single strand oligonucelotide probes. The result showed that all 21 wheat homoeologous chromosome pairs can be reliably discriminated. Total 45 plants derived from M plants were clearly characterized, among

Chromosome 2A carries key genes determining fertility
We assessed the effects of several different chromosome 2As by making six crosses (   Notes. a At F 1 generation, all six cross combinations segregated into two phenotypes, the total number of T plant was 37, the total number of M plant was 37. b All seeds harvested from M plants at F 2 generation will segregate into T, M and D plants at F 3 generation as dms. Zhoumai 18× M at F 1 generation were 2A zhoumai18 2A dms and 2A zhoumai18 , the corresponding phenotypes were T and M plants. This demonstrated that the transmitting frequency of null-2A pollens was about half of the 2A pollens (Table 2). All the T plants of the six crosses at F 1 generation didn't segregate at F 2 generation, but all M plants at F 1 generation segregated into T, M and D plants at F 2 generation. All M plants at F 2 also segregated into T, M and D plants at F 3 ( Table 2). The 2As played key roles during whole plant development in all the monosomic lines of 2As from Taishan 4429, Jing 08-426, Yangmai 11, Yuyou 1 and Guomai 301. These data demonstrated that the six different origin chromosomes 2As had the similar function as the chromosome 2A in dms. In addition, the heterozygous genetic backgrounds of the six crosses, including chromosomes 2Ds and 2Bs couldn't complement the effects of chromosome 2A. D plants were used as a female parent to cross with other wheat genotypes, so as to construct inter-cultivar chromosome 2A substitution lines (Fig. S8). At F 1 generation, all the lines had one chromosome 2A, they were male fertile, which indicated chromosome 2A carries key genes determining male fertility.

Stamen-specifically expressed genes on chromosome 2A
To identify the key genes determining fertility in wheat, 5939 genes located on chromosome 2A (TraesCS2A01G000100 -TraesCS2A01G593900) were analyzed. Total 4613 genes on chromosome 2A were identified and found to be homeologous to genes on other chromosomes (Table S7). Deletion of these genes on chromosome 2A might not affect the phenotype of D plants because of gene redundancy. Total 1326 2A-specific genes were identified (Table S8). Tissue specific expression analysis showed that 188 genes on chromosome 2A were stamen specifically expressed genes ( Fig. 6A; Table S9). Venn diagram analysis showed that 23 genes were chromosome 2A specific and stamen specifically expressed genes ( Fig. 6B; Table S9). Among them 6 genes were reported to be involved in pollen development related biological processes (Table 3), they may be the key genes determining male fertility.

DEGs involved in pollen development related signal transduction in dms
A total of 5199 genes were significantly differentially expressed (FC ≥ 4) between stamens of T and D plants. Among them 4761 DEGs expressed less, only 438 DEGs expressed highly in D-ST (Table S10). Obviously, the expressions of most DEGs in D plants of dms were lower because lacking of the chromosome 2A. We identified 229 putative TF DEGs between T-ST and D-ST. They belonged to 47 TF families. The top three with the most number of DEGs were MYB, C2H2 zinc finger protein (C2H2) and APETALA 2/ethylene-responsive element binding factor (AP2/ERF) transcription factor families and most of them expressed less in D-ST (Table S11). Totally 45 DEGs were associated with auxin homeostasis, such as biosynthesis, response, signaling and metabolism (Fig. S9). Among auxin signal transduction-related genes, homologs of auxin response factor (ARF) genes and several auxin biosynthesis-related genes expressed less in D-ST. Reduced expression of auxin biosynthesis and signal transduction related genes were closely related to male sterility (Su et al., 2019). Abnormality of auxin homeostasis might be a major factor leading to the phenotype of dms.

DEGs involved in pollen development related metabolism in dms
Enrichment analysis of the DEGs between T-ST and D-ST revealed that most of them were involved in pollen development related metabolism such as the GO terms of 'lipid metabolism', 'cell wall biogenesis' and 'pollen development' (Fig. 7). Among lipid metabolism related genes, fatty acid biosynthetic related genes expressed less and lipid Table 3 The key genes determining male fertility in wheat on chromosome 2A.  catabolic process related genes expressed highly in D-ST. Among cell wall biogenesis related genes, cell wall modification related genes expressed less and cellulose biosynthetic process related genes expressed highly in D-ST (Fig. 7). This indicated that fatty acid metabolic and cell wall assembly disorder might be the critical factors causing male sterility in dms.

DEGs involved in abnormal pollen development of dms
The DEGs associated with various aspects of pollen development, such as pollen tube growth, pollen germination, megagametogenesis, anther dehiscence and anther wall tapetum formation, expressed less in D-ST (Fig. 7). A series of pollen development related genes were identified (Table 4). The homologs of AtSAC1B and OsGAMYB involved in exine formation and AtAGC1.5, OsSUT1, AtRopGEF8, AtAGC1.7, AtFIM5 and etc. involved in pollen germination and pollen tube growth expressed less in D-ST (Table 4). The qRT-PCR results showed that the expression patterns of the representative genes were well consistent with that of the sequencing results (Fig. 8).   TraesCS7A01G281300  232  53  TraesCS7B01G179800  277  66   TraesCS7D01G280000  471  109 Os08g0560700/Phl p 7 Involved in pollen tube growth Verdino et al. (2002) Notes. *The asterisks indicate the genes verified by real-time qRT-PCR.

A hypothesis of the molecular regulatory network in dms wheat lines
In summary, we put forward a hypothesis on the molecular regulatory network in dms (Fig. 9). Several evidences support this hypothesis: (1) Chromosome 2A is absent in D plants (Fig. 4); (2) key genes involved in pollen development are identified on chromosome 2A in dms (Table 3); and (3) fatty acid biosynthetic related genes expressed less and catabolic related genes expressed highly in dms (Fig. 7).

DISCUSSION
Aneuploids are large scale mutations greatly affect cellular physiology and have significant phenotypic consequences (Zhang et al., 2017). The typical phenotype of dms was significantly different from its parent Zhoumai 18 at three traits, plant height, pistil number and male fertility. Preliminary, this mutant was considered as a SNP mutation. Till the SNP association analysis showed that a large amount of SNPs between T and D plants, we thought their chromosomes should be clarified. High resolution chromosome painting is a new and efficient method for distinguishing chromosomes, which has many advantages including high sensitivity and resolution (Du et al., 2017). Using this method, we successfully distinguished the karyotypes of D, M and T plants derived from M plants of mutant dms, and their phenotypes were corresponded with their chromosome constitutions. Now it was clear that the mutant dms was resulted from the absence of chromosome 2A. Our data showed that there were 23 2A-specific genes were highly expressed in stamen. However, it can't exclude other genes on 2A involved in pollen development also. In the case of wheat Ms1 gene on 4BS, it has homoeologs on 4A and 4D. Mutant ms1 is responsible for male sterile phenotype due to that homoeologous Ms1 on 4A and 4D were not expressed (due to methylation) (Wang et al., 2017). Some genes in the list of 23 2A-specific genes could be candidates of key genes responsible for the male sterility, but not exclude other possibilities. The molecular regulatory network can be elucidated well till the key gene/genes have been investigated. Endo and Gill reported a set of nullisomic, monosomic, trisomic and tetrasomic lines from Chinese Spring (CS) (Endo & Gill, 1996). Among them, the 2A nullisomic line could not be maintained for they were sterile in both sexes. However, a stable self-fertile 2A nullisomic line was obtained from common wheat 'Abbondanza'. Although the pistils and stamens of the 2A nullisomic line were fertile, its female flower organ developed unwell (Xue et al., 1991). Different genetic backgrounds lead to nullisomic lines of 'Abbondanza' are greater vigor and fertility than those of Chinese spring (Ji, Xue & Wang, 1992). Our 2A nullisomic line dms is multi-pistil and male sterile, that is different from the other 2A nullisomic lines from 'Abbondanza ' and CS (Endo & Gill, 1996). A set of chromosome deletion stocks from CS are reported, five out of the nine chromosome 2AS deletion lines have irregular meioses with many univalents at metaphase I, and they are highly sterile in both sexes, the seed settings of the four chromosome 2AL deletion lines are reduced after selfing. In our study, the meiosis of the pollen development is normal in dms mutant, which is different from the 2A nullisomic lines from CS (Endo & Gill, 1996).
MYB TFs play pivotal roles in plant development and stress response (Verma, 2019;Zheng et al., 2018). Many MYB TFs have been functionally characterized in pollen development of Arabidopsis and rice, such as AtMYB32 (Preston et al., 2004), OsGAMYB (Aya et al., 2009) andOsTDF1 (Cai et al., 2015). In dms, three homologs of OsGAMYB, TraesCS7A01G458700, TraesCS7B01G357900 and TraesCS7D01G446700, expressed less. In rice, GAMYB is essential for pollen development, and it directly binds to promoter of β-KETOACYL REDUCTASE (KAR), a key enzyme essential for fatty acid synthesis. GAMYB activates the expression of KAR and other genes involved in the synthesis of sporopollenin, and are involved in the formation of exine and Ubisch bodies (Aya et al., 2009). Reduced expression of GAMYB might lead to the shriveled and atrophied pollen grains of D plants. Change in the levels of AtMYB32 expression influence pollen development by affecting the composition of the pollen wall in Arabidopsis (Preston et al., 2004). Knocking out OsTDF1 impaired tapetum development, leading to male sterility in rice (Cai et al., 2015). Similarly, some MYB TF genes expressed less in dms, their functions involved in exine formation in wheat needs further research.
Pollen germination is critical for double fertilization in angiosperms (Zhang, He & McCormick, 2009). The polarity of tip-growing pollen tubes is maintained through dynamic association of active Rho GTPases in plants (ROP-GTP) (Li et al., 2018a). Guanine nucleotide exchange factors for ROPs (RopGEFs) catalyze the activation of ROPs and thereby affect spatiotemporal ROP signaling (Gu et al., 2006). Deletion RopGEF mutant has the defects in pollen tube polar growth (Gu et al., 2006). AGC1.5 and AGC1.7 kinases phosphorylate RopGEFs to control pollen tube growth. Loss functions of AGC1.5 and AGC1.7 in pollen tubes results in meandering and depolarized growth morphology (Zhang, He & McCormick, 2009). In summary, the AGC1.5/1.7-RopGEFs-ROPs signaling pathway is involved in pollen germination and tip growth in Arabodopsis (Li et al., 2018a;Huang et al., 2019). In our research, the homologs of AtAGC1.5, AtAGC1.7 and AtRopGEF8 involved in ROP signaling expressed less in dms. Similarly, all the DEGs associated with various aspects of pollen germination related biological processes, such as pollen tube growth, regulation of pollen tube growth, pollen tube development and pollination, expressed less in dms. These data demonstrated that pollen germination and pollen tube growth might be suppressed in dms. Further experiments should be carried out to test the hypothesis.
In our study, all the six chromosome 2As from different cultivars and the genetic backgrounds of the six heterozygotes of the crosses had similar functions. These indicated that the interactions among 2A, 2B and 2D were similar to that in dms. Wheat chromosome 2A has many important agronomy trait genes such as powdery mildew resistance gene PmLK906 (Niu et al., 2008), photoperiod response locus Ppd-A1 (Beales et al., 2007), reduced height (Rht ) genes Rht7 (Worland, Law & Shakoor, 1980). Some quantitative trait loci (QTL) for thousand grain weight (Quan, Sean & Sparkes, 2015), floret primordia development (Guo et al., 2017) and grain protein-content (Groos et al., 2003) are also mapped on chromosome 2A. Because the 2A nullisomic line dms derived from Zhoumai 18 is male sterile but female fertile, it can be used as a female parent to cross with other wheat genotypes, so as to construct inter-cultivar chromosome 2A substitution lines. Backcross to dms can construct series lines with highly similar genetic backgrounds but different 2As (Fig. 10), which can be used to evaluate the functions of different 2As and wheat design breeding targeting 2As.

CONCLUSIONS
We characterized a dwarf, multi-pistil and male sterile mutant dms derived from a widelygrown wheat cultivar Zhoumai 18. Cytological and molecular analyses demonstrated that mutant dms was a novel wheat 2A nullisomic line. Twenty-three stamen and pollen development related genes are identified specifically on chromosome 2A. We put forward a hypothesis on the molecular regulatory network of the sterility trait in dms. dms is a unique germplasm for gene functional study about chromosome 2A and wheat design breeding targeting 2A.
• Zengjun Qi and Jishan Niu conceived and designed the experiments, authored or reviewed drafts of the paper, and approved the final draft.

Field Study Permissions
The following information was supplied relating to field study approvals (i.e., approving body and any reference numbers): Permission prior to accessing the field site was provided by Haolong Yang, administrator at the Henan Seed Company.

Data Availability
The following information was supplied regarding data availability: