FISH-based mitotic and meiotic diakinesis karyotypes of Morus notabilis reveal a chromosomal fusion-fission cycle between mitotic and meiotic phases

Mulberry (Morus spp.), in family Moraceae, is a plant with important economic value. Many polyploid levels of mulberry have been determined. In the present study, the fluorescence in situ hybridization (FISH) technique was applied in Morus notabilis, using four single-copy sequences, telomere repeats, and 5S and 25S rDNAs as probes. All the mitotic chromosomes were clearly identified and grouped into seven pairs of homologous chromosomes. Three dot chromosome pairs were distinguished by the FISH patterns of the 25S rDNA probe and a simple sequence repeat (SSR2524). According to the FISH signals, chromosome length and morphology, detailed meiotic diakinesis karyotype was constructed. Interestingly, only six bivalent chromosomes were observed in diakinesis cells. The 25S rDNA probe was used to illustrate chromosome alterations. The results indicated that mitotic chromosomes 5 and 7 fused into diakinesis chromosome 5 during the meiotic phase. In mitotic cells, the fused chromosome 5 broke into chromosomes 5 and 7. A chromosomal fusion-fission cycle between the meiotic and mitotic phases in the same individual is reported here for the first time. This finding will contribute to the understanding of karyotype evolution in plants.

sequence probe 20 . Mitotic metaphase chromosomes and pachytene chromosomes have been used widely for chromosome identification, physical mapping, and transgene detection 21,22 . Meiotic diakinesis chromosomes have been applied in the studies of polyploidy and particular aspects of chromosomes. However, diakinesis chromosomes have not been systemically applied in FISH-based karyotyping 23 .
In this study, FISH-based karyotypes on mitotic and meiotic diakinesis chromosomes of M. notabilis were determined. All the individual chromosomes were identified by four single-copy sequence probes, and 5S and 25S rDNA probes. The 25S rDNA probe was also used to trace the dynamics of mitotic chromosomes and the different stages of meiotic chromosomes. Here, we describe a chromosomal fusion-fission cycle between the mitotic and meiotic phases in the same individual M. notabilis for the first time.

FISH-based mitotic karyotype of M. notabilis. The mitotic cells of M. notabilis contain 14 chromo-
somes grouped into seven pairs 13 . We used the FISH technique to further examine the karyotype of M. notabilis, especially to identify the two pairs of middle length chromosomes and three pairs of dot chromosomes. All the chromosomes were located by at least one of the probes. Table 1 summarizes the information for four out of the six single-copy sequence probes, and the 5S and 25S rDNA probes used in this study. The longest chromosome, chromosome 1, was located by probe morus027496 (Fig. 1a). The two middle length chromosomes, chromosomes 2 and 3, were located by probes morus027717 and morus026579, respectively. The morus027717 mapped onto the distal part of chromosomes 2 (Fig. 1b,c). Chromosome 4 was recognized by the 5S rDNA probe (Fig. 1d). The 25S rDNA probe mapped onto two of the three dot chromosomes. Based on minor morphology differences and signal intensities, we named these two dot chromosomes as chromosomes 5 and 7 since chromosome 5 was longer and had a larger signal than chromosome 7, the smallest chromosome (Fig. 1e). The remaining dot chromosome, chromosome 6, was distinguished by the location of probe SSR2524, which was named after the simple sequence repeat (SSR) 2524 locus (Fig. 1f).
FISH-based diakinesis karyotype of M. notabilis. Surprisingly, only six fully paired bivalents in the diakinesis phase were found when we observed the meiotic chromosomes of M. notabilis (Fig. 2). To get details of the organization of the diakinesis chromosomes, we used all the probes applied above to identify the chromosomes by FISH. In contrast to the results for the mitotic chromosomes, probes morus027496 and 5S rDNA were located on chromosomes 1 and 4, respectively (Fig. 2a,d). Chromosomes 2 and 3 showed distinguishable differences in morphology and length (Table 2). We named the longer chromosome containing the morus027717 locus as chromosome 2 (Fig. 2b), and the other shorter chromosome containing the morus026579 locus and a distinct primary constriction as chromosome 3 (Fig. 2c). The 25S rDNA probe mapped to only one chromosome, which was similar in length to chromosome 4 and contained a distinct morphology, as shown in the inset in Fig. 2e. We named this chromosome containing the 25S rDNA locus as chromosome 5 because it was similar to the largest dot mitotic chromosome in the signal pattern. The SSR probe SSR2524 mapped onto the smallest chromosome, which we named chromosome 6 because of its similarity to mitotic chromosome 6 ( Fig. 2f).
To further analyze the diakinesis karyotype, the relative lengths of each chromosome were measured in 30 diakinesis cells in which the chromosomes had been identified (Table 2). Chromosome 1 made up 31.85 ± 4.26% of the total length of all the chromosomes. Chromosomes 2 and 3 had distinguishably different lengths of 20.11 ± 2.04% and 18.60 ± 1.83%, respectively. Chromosomes 4 and 5 had similar lengths (12.25 ± 1.56% and 11.05 ± 1.59%, respectively). Chromosome 6 was the shortest, making up only 6.14 ± 1.73% of the total length. We constructed the final karyotype of M. notabilis by combining the morphology and length of diakinesis chromosome and the FISH signal patterns (Fig. 3). The numbering of the chromosomes conformed to the descending order of chromosome length.
Location patterns of 25S rDNA on meiotic and mitotic chromosomes. To reveal the mechanism associated with the reduction of the mitotic chromosome complement of seven chromosomes to the complement of six meiotic chromosomes, we traced the pattern of the 25S rDNA probe signals on the meiotic chromosomes. Interestingly, we detected four loci on the leptotene chromosome (Fig. 4a). Because homologous chromosomes were paired, we observed two loci on the zygotene chromosome ( Fig. 4b) and, strangely, one large locus on the pachytene chromosome (Fig. 4c). One bright locus was resolved on the diplotene chromosome when the homologous chromosome arms were separated, as shown in the inset. In the pre-diakinesis stage, most chromosomes showed one signal, but some had two clearly separated signals (Fig. 4e)   one locus (Fig. 4f). Finally, one locus was observed on the metaphase I chromosome (Fig. 4g), and two loci were observed on the anaphase I homologous chromosomes and metaphase II sister chromosomes separated from each other (Fig. 4h,i).
To determine whether the mitotic chromosome number of 2n = 12 existed, for which the meiotic chromosome number would be doubled, we counted chromosome numbers in 100 metaphase cells of one female and two male wild adult M. notabilis trees for three replicates, and took into account only chromosome numbers between 11 and 15. The results are shown in Fig. 5a. We found that the highest chromosome number was 14 and 11-20% of the cells had 2n = 12 metaphase chromosomes. The cells that had 2n = 12 chromosomes were mapped by FISH using the 25S rDNA probe, and two signal loci were found on a pair of chromosomes (Fig. 5b).

Tel-FISH of M. notabilis.
Tel-FISH was performed on mitotic and meiosis chromosomes to track the dynamics of chromosomes. An Arabidopsis-like telomere sequence with (TTTAGGG) 42 was cloned using the genomic DNA of M. notabilis as template. 25S rDNA probe labelled with digoxigenin-11-dUTP was used as landmark of chromosomes 5 and 7 (Fig. 6a,b. red arrows for chromosome 5 and white arrows for chromosome 7). On mitotic chromosomes, the telomere signals were detected on five pairs of chromosomes at both ends. In the case of dot chromosomes 5 and 7, the telomere signals were only detected at one end and another end was located by 25S rDNA (Fig. 6b, red and white arrows). In addition, interstitial telomere repeat sequences (ITRs) existed on mitotic chromosomes 1-4. In contrast, telomere repeats were only observed at the ends of all diakinesis chromosomes.

Discussion
Various studies have shown that mulberry contains abundant polyploid levels and similar chromosome morphologies 4,8 . These studies were focused on chromosome counting and chromosome behaviors during the meiotic phase. However, these works were carried out using traditional methods and most of the chromosomes were not traceable or distinguishable. In M. notabilis, which has 14 chromosomes of different lengths, two middle pairs of chromosomes and three pairs of dot chromosomes have still not been discriminated. Therefore, new techniques with higher resolution are needed for chromosome studies in these species. FISH has proved to be an important molecular cytogenetic technique and FISH-based chromosome identification can provide accurate karyotypes that can form the basis for phylogenetic analysis 24 .
The availability of the genome sequence of M. notabilis provided an opportunity to analyze the genome organization 13 . Single-copy sequences and repeat sequences can be identified and used as probes for FISH. Here, chromosome identification of M. notabilis was carried out by FISH using single-copy sequence probes and rDNA probes. All chromosomes were successfully identified. Prominent among these were three dot chromosomes easily ordered based on the signal patterns.
Meiotic karyotypes have been widely studied in plants 24 , especially pachytene karyotype. The pachytene chromosome contained four chromatids and was 10-40 fold longer than the metaphase chromosome 16 . The pachytene chromosome had a higher axial resolution and sensitivity in FISH and more morphology landmarks, which made it a better choice for cytogenetic studies. However, the well-spread and separated pachytene chromosome was difficult to prepare in most of the plants 25 , which we also found in M. notabilis.
Unlike the pachytene chromosome, the diakinesis chromosome contains homologous chromosomes that further condensed into bivalent or multivalent chromosome. The diakinesis chromosome has usually been used in polyploidy studies 23   number of Brachiaria humidicola. In this study, we obtained a systemic diakinesis karyotype of M. notabilis by analyzing chromosome morphology, relative chromosome length, and FISH signals (Figs 2 and 3). All the chromosomes were fully paired into bivalents with good sharps and more morphology landmarks, especially in diakinesis chromosomes 3 and 5. Chromosome 3 was clearly distinguished from chromosome 2 based on the FISH signal patterns and diakinesis chromosome morphology, which was impossible in mitotic chromosomes. Diakinesis chromosome 5 was easy to identify because it showed a unique morphology in which the long arm was thicker than the other arm (Fig. 2e). According to the mitotic and diakinesis karyotype of M. notabilis described above, the 14 mitotic metaphase chromosomes showed seven distinct pairs and the diakinesis chromosomes showed six fully paired bivalents (discussed below). Recently, wild mulberry resources with chromosome numbers of 2n = 35 and 2n = 49 have been found by our team. These chromosome numbers are multiples of seven, not 14. Thus, the basic chromosome number of Morus is convincingly considered to be seven.
Polyploidy, dysploidy, and aneuploidy are thought to have played important roles in the karyotype evolution of eukaryotes 28,29 . Dysploidy can cause increases or decreases in basic chromosome numbers through chromosome fission or fusion, respectively. Nested chromosome fusion that can cause basic chromosome number decrease have been reported recently 30 . However, all the plants used in these studies had already formed different basic chromosome numbers in their genus or species. Based on the synteny of chromosome segments, the pathway of karyotype evolution was reconstructed 31 . Here, the chromosome complement of M. notabilis showed only six chromosomes after the diakinesis stage, which was one less than the mitotic complement. To date, this has not been reported in any other plant species. In detail, chromosomes 5 and 7 maintained their separation from the leptotene stage to the zygotene stage (Fig. 4). These two chromosomes began to associate together from the diplotene stage to the pre-diakinesis stage, and two 25S rDNA signals existed for some time. During the diakinesis stage and the two meiosis stages, chromosomes 5 and 7 were fully associated, because only one 25S rDNA signal existed in the meiotic chromosome complement. In the somatic cells, some of the meiotic chromosome 5 s broke into mitotic chromosomes 5 and 7, and the mitotic chromosome number recovered to 2n = 14, although some of the somatic chromosome number 2n = 12 remained. Based on these signal patterns of the 25S rDNA probe, we proposed that mitotic chromosomes 5 and 7 fused or recombined into a slightly unstable diakinesis chromosome 5, and then the fused chromosome broke into chromosomes 5 and 7 in somatic cells. In short, this process suggested a chromosomal fusion-fission cycle in M. notabilis that might be mediated by rDNA.
The chromosomal fusion-fission cycle was first reported in field bean, Vicia faba (2n = 14), where one individual had one pair of metacentric chromosomes that broke into two pairs of telocentric chromosomes 32,33 . This fusion-fission cycle was observed in only one seeding among 2096 progenies, where half the somatic cells had a chromosome number of 2n = 14, and the other half had a chromosome number of 13. Together with the species that contained a chromosome number of 2n = 12 and the morphology of the involved chromosomes, Schubert et al. 33 proposed the fusion-fission cycle was reversible. However, in the present study, the fusion-fission cycle was observed between mitotic and meiotic chromosomes that existed in the same individual. This cycle is remarkably different from the cycle reported in V. faba. Therefore, the fusion-fission cycle is a novel one that has not been observed till now.
In this study, telomere signals were detected only on one end of the metaphase chromosomes 5 and 7 and no ITRs were observed on diakinesis chromosome 5, implying that the chromosomes 5 and 7 were fused at the ends without telomere signals. In combination with the signals of 25S rDNA on mitotic chromosomes 5, 7 and diakinesis chromosome 5, it seems reasonable to suggest that chromosome fusion occurred at the chromosome ends harbored 25S rDNA loci in M. notabilis.
In conclusion, we applied the molecular cytogenetic technique FISH in M. notabilis. The mitotic karyotype was constructed accurately and can act as the bases for future cytogenetic research in other mulberry species. The comprehensive FISH-based diakinesis karyotype was also built, and the chromosome information that was obtained will have a wide usage in other plants. The fusion-fission cycle of chromosomes 5 and 7 between the meiotic and mitotic phases in the same individual is reported here for the first time. This process suggests the natural basic chromosome number can be altered, and that the M. notabilis karyotype goes through this process.

Materials and Methods
Material and DNA preparation. The wild mulberry germplasm resource M. notabilis (2n = 14) used for all the FISH experiments was collected from a pristine forest in Ya'an, Sichuan Province, Southwest China (29°45.278′N, 102°53.878′E). The genomic DNA was extracted from M. notabilis leaves using the CTAB method 34 .
Chromosome preparation. The mitotic chromosomes were prepared as described previously with minor modifications 13 . In brief, young leaves were pretreated with 2 mM 8-hydroxyquinoline at room temperature for 3 h, then fixed in 3:1 methanol/glacial acetic acid at 4 °C for 24 h. Fixed leaves were incubated with 0.067 M KCl solution for 1 h, then digested by 2.5% (W/V) cellulase Onozuka R-10 (Biosharp, Hefei, China) and 2.5% (W/V) pectolyase Y-23 (Biosharp, Hefei, China) at 30 °C for 3 h. Digested leaves were rinsed with ddH 2 O at room temperature for 1 h. For meiotic chromosome preparation, young inflorescences were fixed in 3:1 methanol/glacial acetic acid at room temperature for at least 24 h and stored at −20 °C before use. After removal of calyx, anthers from an inflorescence were incubated with 0.067 M KCl solution for 1 h, and then digested by 2.5% (W/V) Probe preparation. The Morus Genome Database (MorusDB) (http://morus.swu.edu.cn/morusdb) was used to screen the single-copy sequences 35 . Based on BLASTn searches against MorusDB, the repeat-free sequences of single-copy genes and the flanking regions >10,000 bp were selected. The repeat-free sequences (morus027496, morus027717, morus026579 and SSR2524) were amplified into lengths of 1,154~3,100 bp with primers designed using the Primer5 software 36 . The amplified sequences were cloned into a pMD19-T vector, and verified by both end sequencing. The purified PCR products of the clones were mixed equally, and then labeled with digoxigenin-11-dUTP using a DIG-Nick Translation Kit (Roche, Mannheim, Germany) according to the product manual. These probes were used directly for FISH. The primers for the 5S and 25S rDNAs were designed from the conserved sequences according to the multiple sequence alignment results obtained using the BioEdit software 37 . The primers for cloning Arabidopsis-like telomere were designed according to Ling et al. 38 . The 5S and 25S rDNA sequences were cloned into a pMD19-T vector, and then labelled with biotin-16-dUTP and digoxigenin-11-dUTP, respectively, using a PCR DIG Probe Synthesis Kit (Roche) with labeling conditions based on the product manual. The telomere sequences were cloned into pMD19-T vector and labelled with biotin-16-dUTP. The PCR cycles consisted of an initial denaturation of 95 °C for 5 min; followed by 32 cycles of 95 °C for 30 s, 58 °C for 30 s, 72 °C for 1 min; then 7 min for the final extension. The primers used in this study are listed in Supplemental Table S1.
Mitotic chromosome FISH. Slides with well-spread chromosome preparations were baked at 37 °C for 2 h.
To reduce the signal background, the slides were subsequently incubated in 100 µg/ml RNase at 37 °C for 1 h and in 1 µg/ml protease K at 37 °C for 15 min. The slides were washed in 2× standard sodium citrate (SSC), once at 37 °C for 5 min and twice at room temperature for 5 min. The chromosomes were denatured with 70% formamide in 2× SSC at 72 °C for 10 min, immediately dehydrated in 70%, 90%, and 100% ethanol for 5 min in each, and dried in air. The hybridization solution (80 µl of 50% formamide in 2× SSC, 10% dextran sulfate, 0.25% SDS, 125 ng/ml salmon sperm DNA, and 5 ng/µl probes) was denatured at 95 °C for 6 min, then chilled on ice for at least 10 min. Next, the chromosomes and hybridization solution were denatured together at 80 °C for 5 min. After overnight hybridization at 37 °C, the slides were washed with 10% formamide in 2× SSC at 37 °C for 10 s, twice in 2× SSC at 37 °C for 3 min, then in 0.2% Tween 20 in 4× SSC at room temperature for 3 min. The slides were blocked with 1× blocking solution (Roche) in 4× SSC at 37 °C for 20 min. The digoxigenin-and biotin-labelled probes were detected with 1ng/µl anti-digoxigenin-fluorescein, Fab fragments from sheep (Roche, Mannheim, Germany) and 1ng/µl streptavidin-Cy3 (Zymax Grade) (Invitrogen, CA, USA), respectively. After incubation at 37 °C for 1 h, the slides were washed with 0.2% Tween 20 in 4× SSC once at 37 °C and twice at room temperature for 3 min. Chromosomes were counterstained with 1 ng/µl of DAPI in darkness for 8 min, then washed with 2× SSC at room temperature for 3 min. Finally, the slides were covered with cover slips and sealed using nail polish.
Meiosis chromosome FISH. The meiosis chromosome FISH was carried out using the mitotic chromosome FISH protocol described above with the following modifications. The digestion times of RNase and protease K were lower, namely 30 min and 5 min, respectively. To fix the chromosome, two steps were added before the chromosomes were denatured with 70% formamide; i.e., the chromosomes were fixed with 4% formaldehyde at room temperature for 10 min and then dehydrated in 70%, 90%, and 100% ethanol for 5 min in each. with DAPI counterstaining in gray color, the metaphase chromosomes 5 and 7 were indicated by red and white arrows, respectively. (b) Red and white arrows indicated the FISH signals of the 25S rDNA probe on metaphase chromosomes 5 and 7, respectively, and (c) green arrow indicated the FISH signal on the diakinesis chromosome 5. Scale bars represent 5 μm.
using the cellSens Standard 1.13 software and a CCD camera DP80 in grayscale channels. The images were then pseudocoloured with the cellSens Standard 1.13 software. Finally, the images were adjusted using Adobe Photoshop CS6 and merged by ImageJ software 39 . To construct the karyotypes of M. notabilis, each chromosome pair was separated from the images with chromosomes identified. The relative lengths of the diakinesis chromosomes were measured using ImageJ software in 30 cells 39 .