Sequence of the supernumerary B chromosome of maize provides insight into its drive mechanism and evolution

Significance B chromosomes are nonvital chromosomes found in thousands of plants and animals that persist through various drive mechanisms. The drive mechanism of the maize B chromosome consists of mitotic nondisjunction at the second pollen division to produce two unequal sperm and then the sperm with the B chromosomes preferentially fertilizes the egg in double fertilization. A high-quality sequence of the maize B chromosome together with genetic analysis reveals the cis factor for nondisjunction is a B chromosome-specific repeat interspersed in and around the centromere. The gene and transposable element content of the B chromosome and relaxed purifying selection of transposed protein-encoding genes suggest that the chromosome has been present in the evolutionary lineage for millions of years.

and a 22 mm × 50 mm glass coverslip was placed on the slide. After at least 20 min, spreads were screened using a fluorescence Olympus microscope. Images were processed with Photoshop.

B chromosome size estimation
The size of the B chromosome was estimated using flow cytometry in plants with

B chromosome sorting and survey sequencing
Flow cytometry was used for chromosome isolation (6) of maize B chromosomes.
Roots of young seedlings were treated with 3 mM hydroxyurea for 18 hours in order to synchronize the cell cycle. Accumulation of cells in metaphase was achieved using nitrous oxide at 5 atmospheres for 2 hours. To prepare the chromosome suspension, roots were cut and transferred into 3% formaldehyde fixative for 25 min at 4°C. After washing in Tris buffer (3x5 min), root-tip meristems were excised and collected in 1 ml of LB01 lysis buffer and homogenized. The homogenate was filtered through 25 μm nylon mesh into a sample tube. Chromosomal DNA was stained with 2 μg/ml of DAPI and chromosome analysis and sorting were achieved using FACSAria II SORP flow cytometer (Becton Dickinson, Franklin Lakes, NJ, USA). Purity check was performed via FISH using the B-specific probe (7). Total of 235,000 flow-sorted chromosomes were treated by proteinase K and column-purified as described previously (8), yielding 24 ng DNA, which was then amplified in three independent reactions using Illustra GenomiPhi V2 DNA Amplification kit (GE Healthcare, Chalfont St. Giles, United Kingdom). Resulting 2.5 micrograms amplified DNA were pooled and used to prepare two paired-end libraries: the first one with ≈500 bp insert using DNA Library Preparation Kit v2 (Illumina, San Diego, CA, USA) and of the second one with ≈800 bp insert using Nextera XT DNA Library Preparation Kit (Illumina). The sequencing of both libraries was then performed on an Illumina HiSeq2000 sequencer (Illumina) at the Institute of Applied Genomics (Udine, Italy).
The identification of LTR retrotransposons (LTR-TE), LINEs, SINEs, TIRs and helitrons was performed as described in Jiao et al. (5) using provided scripts with minor modifications. In brief, ZmBs, 180-bp knob repeat and CentC elements were masked prior analysis and LTR-TEs were identified using LTRharvest v1.5.10 (12) and LTRdigest v1.5.10 (13). Seventy-six iterations were performed to discover the nested elements. Identified elements longer than 100 kb were excluded from further analyses because they are likely to be false positives. Retrieved LTR-TEs were clustered into families by the 5' LTR similarity based on comparison with vsearch v2.13.6 (14) and SiLiX v1.2.9 (15). Superfamily membership was assigned to each element based on the order of the genes for integrase, reverse transcriptase (RT) and RNaseH (if found in a particular element). Identification of those genes was performed using homology (23) to recover LINE and TIR elements. Helitron candidates were identified using HelitronScanner v1.0 (24) and assigned to families with SiLiX and vsearch in a similar manner as performed for LTR-TEs. Fisher's exact test. The sum of members in A chromosomal complement served as a reference set for testing. The p-value was adjusted via FDR correction (26).

Identification of insertions of organellar DNA
The maize chloroplast genome (NC_001666.2) and 10 mitochondrial genomes FISH analysis of the B chromosome using probes to mitochondrial and chloroplastic DNA sequences was performed as described previously (29).

Expression analysis of B-linked genes
RNA-seq reads generated from leaves of maize B73 containing no, one and six B chromosomes (30)  used to run a differential expression analysis with DESeq2 (34) and with p-value set to < 0.05.

Test of homology of the B chromosome to the A chromosomes
The copy number variation analysis was performed as described previously (35).
Generally, single-end reads after filtering low quality reads were aligned to the maize W22 reference genome (36) using Bowtie2 (37). Only uniquely mapped reads with no more than 2 mismatches were kept. To examine the copy number change with a single gene window, the reads from per gene windows were counted along each chromosome and the results were plotted by ggplot2 in R (38  The delineation of the cis region for nondisjunction is described in the text. The region for preferential fertilization resides between the breakpoints of TB-10L18 and TB-8Lc. The region for trans-acting factor #1 resides distal to the breakpoint in TB-3Sb.    GO term enrichment of the B chromosome genes. GO terms enriched among genes encoded in the B chromosome sequence (red) compared to frequency of genes in the same GO term in A chromosomal complement (blue). The significance was tested using one-sided Fisher's exact test with correction via FDR and p-value threshold of 0.05.

Fig. S7.
Alignment of final scaffolds (Y-axis) to pseudomolecules of ten chromosomes of B73 genome (X-axis) (5). Note nearly perfect collinearity and high similarity of the sequence.     The estimates are based on the assumption that the size of the haploid reference of B73 inbred line is 2,135 Mb (5).