Synthesis of a Wheat/Maize Hybrid CENH3 Gene, the Genetic Transformation of Wheat, Its Chromosomal Localization and Effects on Chromosome Behaviors in Wheat/Maize Somatic Hybrids

Centromere-specific histone H3 (CENH3) replaces the canonical histone H3 in nucleosomes of functional centromeres, and plays important roles in faithful chromosome segregation during cell division. CENH3 is also important in the recognition of alien centromeres and determines the accommodation or elimination of alien chromosomes in interspecific or intergenic hybridization. In this study, a maize full length CENH3 with a yellow fluorescent protein (YFP) tag at C-terminus (ZmCENH3-YFP) and a synthetic hybrid wmCENH3 with the N-terminus from wheat CENH3 and the histone fold domain (HFD) from maize tagged with a red fluorescent protein (RFP) at the C-terminus (wmCENH3-RFP) were transformed to wheat by biolistics transformation. Transgenic wheat plants with both ZmCNEH3-YFP and wmCENH3-RFP genes were identified by PCR. The expression of ZmCENH3-YFP was not observed, while the expression of wmCENH3-RFP could be detected by RT-PCR, direct fluorescence microscopy, and immunostaining with anti-RFP antibody. The expressed wmCENH3-RFP was localized to nuclei as dotted patterns, indicating its targeting to wheat centromeres. Somatic hybridization was performed between wmCENH3-RFP transgenic wheat and transgenic maize that expressed a ZmCENH3-YFP gene to investigate chromosome behaviors in somatic hybrids. Cytological and FISH analyses of somatic hybrid cells showed the formation of micronuclei and lagging chromatin in both somatic hybridizations with or without the wmCENH3-RFP transgene, indicating that ectopically expressed wmCENH3 could not overcome chromosome elimination in wheat/maize somatic hybrids. Immunostaining of wmCENH3-RFP and ZmCENH3-YFP in early stage somatic hybrid cells indicated that both wmCENH3-RFP and ZmCENH3-YFP proteins were expressed, but their binding patterns changed from the commonly observed dotted patterns to diffused ones, suggesting that the inactivation of CENH3 might be a factor for chromosome elimination in wheat/maize somatic hybridization.


Introduction
Centromeric histone H3 (CENH3) is a variant to replace canonical histone H3 nucleosome in functional centromere for kinetochore assembly. CENH3 is only 50% identical to canonical histone H3 in the histone fold domain (HFD) [1]. It plays an important role in chromosome behaviors during cell divisions, especially in chromosome elimination and adaptation. For example, haploid Arabidopsis plant could be produced by chromosome elimination when a wide type plant was crossed with a cenh3 mutant expressing an engineered CENH3 or naturally occurring CENH3 from distant related plant species [2] [3]. The chromosome elimination is always on chromosomes from the hybridization parent that expresses the divergent CENH3. Chromosome elimination has also been reported in the unstable interspecific barley hybrids caused by the loss of CENH3 protein and centromere inactivation [4]. In contrast, in the oat/maize addition lines, oat CENH3 can nucleate functional kinetochores on maize chromosomes while the maize CENH3 gene is silenced [5]. Thus, maize chromosomes can be retained in the oat genetic background.
Two regions of CENH3 protein are highly variable: N-terminal tail for kinetochore assembly and the loop 1 region which interacts with nucleosomal DNA and is crucial for centromere targeting and the recognition between CENH3 and centromeric DNA sequence [6] [7]. Domain swap experiments showed that chimeric CENH3 with the N-terminal tail from Lepidium oleraceum and the HFD from A. thaliana caused chromosome missegregation and genome elimination, while the reversely swapped chimeric CENH3 with the A. thaliana N-terminal tail and the L.oleraceum HFD acted normally as the Arabidopsis CENH3 [3].
Thus, the evolution of CENH3 could contribute to hybridization barrier that prevents genetic cross between distantly related plant species.
Hybridization barriers are generally divided into pre-zygotic and post-zygotic barriers [8]. Pre-zygotic barriers include spatial and temporal separation [9], morphologically and ethologically floral isolation [10], and failure in pollen-pistil interactions [11]. Technologies have been developed to overcome hybridization barriers to broaden genetic resources for plant breeding. For example, artificial hybridization could overcome some habitat, temporal and ethological barriers [8]; gametic barriers could be solved via in vitro fertilization technology, in which male and female gametes were isolate from reproductive organs of parental plants, and fused to produce zygotes through calcium-mediated, electrofusion or microinjection, and ion measurement methods [12]. Post-zygotic barriers include hybrid non-viability, weakness and sterility [13] [14], hybrid breakdown [15], and hybrid necrosis [16]. Post-zygotic barriers are mainly caused by aborted embryo or endosperm, abnormal chromosome behaviors, and selective chromosome elimination [17] [18]. Some techniques have been developed to tackle post-zygotic barriers. For example, chromosome doubling technique was developed to solve hybrid sterility [19], and embryo rescue technique was developed to recover unviable and weak hybrid embryos from abortion [20]. Somatic hybridization technique was also developed to generate somatic hybrids between sexually incompatible plants that could not be realized by conventional genetic cross.
Recent studies have revealed molecular mechanisms of chromosome elimination in hybridization barrier. The importance of CENH3 in chromosome eliminations hints that CENH3 might be manipulated to overcome chromosome eliminations. In wheat/maize hybrids from genetic cross and somatic hybridization, maize chromosomes are quickly eliminated during the first several cell cycles [21]- [26]. Chen et al. [27] reported an attempt to ectopically express a maize CENH3 gene (ZmCENH3) in transgenic wheat. However, although the ZmCENH3 could be transcribed at a low level, ZmCENH3 protein was not detected by both western blot and immunostaining. Chromosome elimination was not suppressed in genetic crosses between transgenic wheat and maize. To overcome the suppression of ZmCENH3 transgene expression in transgenic wheat, we synthesized a hybrid wmCENH3 gene, which has the N-terminus before loop 1 domain from wheat TaCENH3, the C-terminal HFD after loop 1 from maize ZmCENH3, and a red fluorescent protein (RFP) tag at the C-terminus. The synthesized wmCENH3 gene was cloned into a gene expression cassette under a strong maize ubiquitin promoter [28], and transformed into wheat by biolistic transformation. Transgenic wheat was generated and the influence of ectopically expressed wmCENH3 on chromosome behaviors in wheat/maize somatic hybrids was analyzed.

Sequence Analysis
To analyze the similarity of three different CENH3 proteins, amino acid sequences of TaCENH3, ZmCENH3 and a synthetic wmCENH3 were aligned with the canonical histone H3 of rice using the CLC Sequence Viewer 7.6 software program (https://www.qiagenbioinformatics.com/products/clc-sequence-viewer/).

Constructs for Wheat Transformation
The maize CENH3 gene with a YFP tag was constructed in pTF101, and was kindly provided by Prof. James Birchler (University of Missouri, Columbia). The pTF101-ZmCENH3-YFP (Figure 1(A)) had streptomycin resistance in E. coli, and two expression cassettes: a maize full length CENH3 gene with a YFP tag at C-terminus driven by 2 × 35S promoter from cauliflower mosaic virus (CaMV), and a Bar gene driven by 2 × 35S promoter as the selectable marker. A wmCENH3-RFP gene was synthesized and cloned into the pCAMBIA3301 vector to construct pCAMBIA3301-wmCENH3-RFP (Figure 1(B)). This construct had kanamycin resistance in E. coli, and three expression cassettes on T-DNA region: a wmCENH3-RFP gene with wheat TaCENH3 N-terminus before Loop1 domain, the maize ZmCENH3 C-terminus after Loop1, and a RFP gene tag at C-terminus, driven by a maize ubiquitin promoter [28]; a Bar gene driven by 2 × 35S promoter from cauliflower mosaic virus as the selectable marker; and a β-glucuronidase (GUS) gene [30] driven by 35S promoter as the reporter gene.

Biolistic Transformation of Wheat and Plant Regeneration
Plasmid DNAs were extracted by alkaline extraction method [31], and purified

Genomic DNA Isolation from Transgenic Wheat
Genomic DNAs were isolated from wild type and transgenic wheat lines by SDS method [34] with modification.  PCR reactions were performed in a total volume of 20 μL containing 10 μL Premix Ex Taq Hot Start Version (Takara Code No. HRR030A), 50 ng wheat genomic DNA or 10 ng plasmid DNA, 0.5 μM each of forward and reverse primers with the following PCR program: an initial denaturalization at 95˚C for 3 min, 35 cycles of denaturalization at 98˚C for 10 s, annealing at 60˚C for 30 s and extension at 72˚C for 40 s, followed by a final extension at 72˚C for 5 min. PCR products were checked on 0.8% agarose gel after staining by ethidium bromide (EB).

Southern Blot for T0 Transgenic Wheat Lines
Southern blot analysis was performed to confirm gene transformation in T0

Expression Analysis of wmCENH3-RFP in Transgenic Wheat
RT-PCR was performed in T0 transgenic wheat for the detection of wmCENH3-RFP expression. Total RNA isolation from leaves of wild type and wmCENH3-RFP transgenic wheat was performed by using an RNeasy Plant Mini Kit (QIAGEN) according to the manufacturer's instruction. The first strand cDNA synthesis was performed using PrimeScript RT-PCR Kit (Takara). PCR amplification was performed using primers of P5 (5'-AGAGCGCTATACCGCAGAAG-3') and P6 (5'-GGGTGCTTCACGTACACCTT-3') for wmCEHN3-RFP gene with the same program mentioned above. The PCR products were separated by 2% agarose gel electrophoresis and stained with EB for visualization.

Localization of Alien CENH3 Protein by Fluorescence Microscopy
Root tips of transgenic wheat were squashed on glass slides with cover glasses, and examined for red or yellow fluorescence under the 100 × oil immersion objective of Leica DM5500B fluorescence microscope (Leica Microsystems). Images were captured with a Leica DFC490 digital camera and processed by Adobe Photoshop CS software.

Immunostaining of wmCENH3-RFP in Transgenic Wheat
Root tips of wild type wheat and T0 transgenic wheat lines (11-1, C2901, C3057 and C3216) were checked for the expression of wmCENH3-RFP by immunostaining [5] with some modifications.

Establishment of Cell Suspension Culture for Wheat
Wild type or wmCENH3-RFP transgenic wheat calli were used to initiate cell  Table Shaker (CRYSTAL, Product number 1109012) at 28˚C in the dark with 130 rpm rotation.  The qualities of wheat and maize protoplasts were checked under bright field inverted microscope (Nikon, TE300) immediately after isolation and images were captured with a digital camera (V-Tphoto adapter Nikon MBB74700).

Symmetric Hybridization between Wheat and Maize
The protoplast fusion between wmCENH3-RFP transgenic wheat line C3216 and ZmCENH3-YFP transgenic maize was performed by PEG-mediated somatic hybridization [25] [37] [38]. After re-suspending wheat and maize pellets in P5 medium, wheat and maize protoplasts were mixed together at a ratio of 1:

Cytological Analysis of Cell or Cell Clusters from Protoplast Fusion
The cell cultures from the fusion protoplasts of wheat and maize were fixed at 7, 14 and 28 d after fusion with Carnoy's fixative solution [39].

FISH Analysis of Cell or Cell Clusters from Somatic Hybridization
The maize B repeat sequence [41] was found to specifically hybridize to the sub-

Immunostaining of CENH3 in Somatic Hybridized Cells
High-frequency chromosome elimination usually occurs during the 10 -14 d after protoplast fusion [46].

Sequence Analysis of Wheat and Maize CENH3 and the Design of a Synthetic wmCENH3
Sequence alignment of wheat, maize and rice CENH3 proteins were performed (Figure 2(A)). The result showed conserved and diverged regions among the CENH3 proteins. The N-terminal regions were highly diverged among the CENH3 proteins, whereas the HFDs including αN-helix, α1-helix, Loop 1, α2-helix, Loop2 and α3-helix were relatively conserved.
The large divergence between wheat and maize CENH3 proteins (74% similarity) could be a factor for the suppression of ZmCENH3 expression in transgenic wheat [27]. To reduce the difference between transgenic CENH3 and the endogenous TaCENH3, a synthetic wmCENH3 gene was designed to have the N-terminus of TaCENH3 before the loop 1 domain, and the C-terminus of ZmCENH3 after loop 1 (Figure 2(B)). The synthetic wmCENH3 protein had 95% similarity to the TaCENH3. An RFP tag was fused in frame to the C-terminus of the synthetic wmCENH3 gene to trace the behavior of the wmCENH3 in transgenic plants.

Biolistic Transformation of Wheat and Plant Regeneration
Wheat plants were grown in Plant Growth Chamber and 3900 immature embryos   events with wmCENH3-RFP (3 events with wmCEHN3-RFP only and 13 events with wmCENH3-RFP + Bar) and 2 events with ZmCENH3-YFP (1 event with ZmCENH3-YFP only and 1 event with ZmCENH3-YFP + Bar). Transformation efficiency was 2.69% for Bar gene, 0.65% for wmCENH3-RFP, and 0.13% for ZmCENH3-YFP. More Bar gene was integrated into wheat genome than both alien CNEH3 genes although the Bar gene was located on the same construct with either wmCENH3-RFP or ZmCENH3-YFP. Transgenic plants were grown in growth chamber and self-pollinated to produce seeds (Figure 3(E)).

Southern Blot and RT-PCR Analysis for Transgenic Wheat
Southern blot with the DIG-labeled Bar gene was performed in T0 transgenic plants. The results showed wmCENH3-RFP construct was successfully integrated into the genome of T0 transgenic wheat lines of 11-1, C2901, C3057 and C3216. The copy numbers of the transgene were one for 11-1 and C3057 events, two for C2901 event, and four for C3216 event (Figure 4).
To analyze wmCNEH3-RFP gene expression in transgenic wheat, RT-PCR was performed with wild type wheat and transgenic events 11-1, C2901, C3057 and C3216. PCR primers were designed to detect a 399 bp product from the wmCENH3-RFP gene if it was expressed by RT-PCR. A 399 bp wmCENH3-RFP gene product was detected in the four transgenic wheat lines of 11-1, C2901, C3057 and C3216, but no product was detected in the WT wheat ( Figure 5(A)).

Localization of Ectopically Expressed CENH3
The expression of fluorescence protein tagged CENH3 proteins were analyzed under a fluorescence microscope in root tips of regenerated transgenic plants.
Dotted RFP signals were observed in the nuclei of root tip cells of wmCENH3-RFP transgenic plants (Supplementary Figure S3), suggesting that the wmCENH3-RFP was targeted to functional wheat centromeres. In contrast, fluorescence protein was not observed in the two transgenic plants with the ZmCENH3-YFP gene, being consistent with the previous report by Chen et al. [27]. This evidence indicated the suppression of ZmCENH3 gene expression in transgenic wheat.
The wmCENH3-RFP fusion protein was further detected by immunostaining with an antibody to the RFP tag at the C-terminus ( Figure 5(B)). The expressed wmCENH3-RFP protein was detected in the nucleus of transgenic wheat in a

Symmetric Hybridization of Transgenic Wheat and Maize
Cell suspension cultures for WT wheat and wmCENH3-RFP transgenic wheat lines, 11-1, C2901, C3057, and C3216 were established. There was no significant difference in growth performance between WT and four wmCENH3-RFP transgenic wheat cell suspension cultures (Supplementary Figure S4 and Supplementary Figure S5). WT and one of the transgenic line C3216 cultured cells were used for protoplast isolation and somatic hybridization to maize protoplasts.
Spherical protoplasts were isolated from WT and C3216 wheat suspension culture and maize young leaves (Supplementary Figure S6(A) and Figure   S6(B)). About equal amount of isolated wheat and maize protoplasts were mixed and cell fusions were induced by PEG. Fused protoplasts were cultured in microplates with P5 medium at a density of 1 -5 × 10 6 protoplasts per ml. The occurrence of hybridization was confirmed by checking the hybridization products with inverted microscope at 1 -2 d after fusion (Supplementary Figure S6(C)).
The growth of fused cells at early stage after fusion was observed by an inverted microscope (Supplementary Figure S6(D)).

Cytological Examinations of Somatic Hybrids
Aceto-orcein staining analysis was performed to investigate chromosome behaviors during mitosis in somatic hybrids. Abnormal chromosomes behaviors such as mi-

CENH3 Behavior in Somatic Hybrids
The

Fragmentation of Transformed DNAs
Genetic transformation by particle bombardment usually produces complex transgene loci with multiple copies, truncated, rearranged sequences interspersed with genomic DNA fragments [47] [48]. In this research, we observed only 14 out of 109 transgenic wheat plants (12.84%) that had full length DNA fragments, and most of the transgenic events had fragmented transgenes, indicating the   [47]. Southern blot and FISH analysis of transgene loci in transgenic hexaploid oat revealed that transgene integration was associated with chromosomal breakage and rearrangements [50], and more complex transgene integration loci and configuration could be induced through the minor rearrangements and undesirable plasmids backbone vector.

Biased Transformation and Recovery of Transgenic Wheat with CENH3 Transgenes
In this study we observed significantly more transgenic events with Bar gene only (83.5%) than those with either ZmCENH3 or wmCENH3 genes (16.5%), although the Bar gene was on the same constructs. The bias against CENH3 transgene indicates that the transformation might be not random. Two reasons wheat callus by PCR in T0 generation. But no YFP signals were observed in the root tips of these two events under fluorescence microscope. Similar results have been reported by Chen et al. [27], who observed the competition and suppression of ZmCENH3 transcription by endogenous wheat TaCENH3. The transgenic ZmCENH3 can be weakly transcribed and the ZmCENH3-YFP protein is not observed by both western and immunostaining analyses, whereas the expression of endogenous TaCENH3 is up-regulated in transgenic wheat [27]. In contrast, we observed higher transformation efficiency with the synthetic wmCENH3 gene (0.65%) than the ZmCENH3 construct (0.13%), and the expression of wmCENH3 was successful in transgenic wheat.
The difference between the two transgenes is the overall sequence similarity to wheat TaCENH3 and the N-terminal part of CENH3 proteins that is highly variable and critical for kinetichore assembly. In Arabidopsis, the N-terminal part of AtCENH3 is required for its loading to centromeres during meiosis and mitosis [51], and the divergence of the N-terminal tail could cause chromosome missegregation and genome elimination in hybrids between wild type plants and engineered Arabidopsis that expressed a chimeric CENH3 [3]. In addition, high sequence similarity to endogenous CENH3 protein has been reported to be required for heterologeous CENH3 recognition in A. thaliana centromeres [51], although there are other reports wherein the highly diverged maize ZmCENH3 localized in A. thaliana [3] [52]. The missegregation and chromosome elimination are positively related to the degree of CENH3 divergence. Crosses between wild type and Arabidopsis cenh3 mutants that express the maize ZmCENH3 can produce more missegregation and haploids than those crosses with Arabidopsis cenh3 mutants that express CENH3 from a closely related L. oleraceum species [52]. Alignment of maize and wheat CENH3s demonstrates that the full length proteins are 74% similarity. The similarities are 60% in the N-terminal part before Loop 1 and 92% in the C-terminal part after Loop 1, respectively. The domain swap between wheat TaCENH3 and maize ZmCENH3 resulted in a fusion protein with the N-terminal tail from TaCENH3, and the overall protein similarity between the synthetic wmCENH3 and wheat TaCENH3 was increased to 95% (Figure 2(A)). The relatively low frequency in obtaining transgenic wheat with the full length maize ZmCENH3 gene may suggest that transgenic plant with this transgene might have detrimental effect on the growth and development of transgenic events. We argue that the suppression of ZmCENH3 gene expression observed by us and Chen et al. [27] could in part reduce the detrimental effects on transgenic wheat, which can be ameliorated by the chimeric CENH3 with a wheat N-terminal tail, and the increased sequence similarity to wheat CENH3. Thus, the higher transformation efficiency with the synthetic wmCENH3 gene than the ZmCENH3 construct, and the successful expression and localization of wmCENH3 might be attributed to less competition between endogenous CENH3 and the ectopically expressed ones in transgenic wheat, although the endogenous TaCENH3 expression was not determined in this re-

Ectopically Expressed wmCENH3 in Wheat/Maize Somatic Hybridization
Abnormal chromosome behaviors were observed frequently in the somatic hybrids, indicating the occurrence of chromosome elimination. Cytological studies of these somatic hybrids revealed various abnormal chromosome behaviors such as micronuclei, asynchronous cell cycle, and chromosome lagging ( Figure 6).
Previous studies demonstrate that maize chromosomes without spindle microtubule attachment are stayed or delayed in the metaphase plate during first mitosis of zygotic hybrids of wheat/maize [26], and maize chromosomes could be gradually eliminated from the hybrids in several cell cycles [24].
Many evidences show that chromosome elimination is related to the loss or malfunction of CENH3 [2] [3] [4] [5] [52]. By analyzing CENH3 behavior in early somatic hybrids between transgenic wheat with a synthetic wmCENH3-RFP and transgenic maize with the ZmCENH3-YFP, we could detect the change of the CENH3 localization patterns from the normally dotted centromere localization to a diffused pattern in the whole nuclear chromatins (Figure 8).
Eukaryotic chromosomes can have three types of centromeres including point centromeres, monocentromeres and holocentromeres [53]. As a centromere marker, CNEH3 protein is mainly deposited to the active centromeres, and its distribution can vary depending on the type of centromeres. CENH3 on monocentric or point centromeres is usually observed as dotted patterns. In contrast, CENH3 is almost evenly distributed on the whole holocentric chromosomes.
Wheat and maize have monocentric chromosomes and thus their CENH3 localization usually displays a dotted pattern in the nuclei. The heterogeneous expression of wmCENH3-RFP in wheat also displayed a dotted pattern ( Figure   5(B)), indicating that this heterogeneous CENH3 might be incorporated into wheat centromeres. Similar results have also been reported in heterogeneous expressions of CENH3s from A. lyrata, A. arenosa, Capsella bursa-pastoris, Zea mays, Nicotiana tabacum, and L. oleraceum in A. thaliana [3] [51] [52] [54], as well as A. thaliana AtCENH3 in tobacco BY-2 cells [54]. Our observation of mislocalization of both ZmCENH3-YFP and wmCENH3-RFP in wheat/maize somatic hybrids may suggest a mechanism of centromere inactivation in somatic hybrids, which could be responsible for chromosome elimination partially. Agricultural Sciences