Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2019 Feb;31(2):368-383.
doi: 10.1105/tpc.18.00613. Epub 2019 Jan 16.

Genome-Scale Sequence Disruption Following Biolistic Transformation in Rice and Maize

Jianing Liu  1 , Natalie J Nannas  2 , Fang-Fang Fu  3 , Jinghua Shi  4 , Brooke Aspinwall  1 , Wayne A Parrott  5 , R Kelly Dawe  6   3
Affiliations
Free PMC article

Genome-Scale Sequence Disruption Following Biolistic Transformation in Rice and Maize

Jianing Liu et al. Plant Cell. .
Free PMC article
Favorites

Abstract

Biolistic transformation delivers nucleic acids into plant cells by bombarding the cells with microprojectiles, which are micron-scale, typically gold particles. Despite the wide use of this technique, little is known about its effect on the cell's genome. We biolistically transformed linear 48-kb phage lambda and two different circular plasmids into rice (Oryza sativa) and maize (Zea mays) and analyzed the results by whole genome sequencing and optical mapping. Although some transgenic events showed simple insertions, others showed extreme genome damage in the form of chromosome truncations, large deletions, partial trisomy, and evidence of chromothripsis and breakage-fusion bridge cycling. Several transgenic events contained megabase-scale arrays of introduced DNA mixed with genomic fragments assembled by nonhomologous or microhomology-mediated joining. Damaged regions of the genome, assayed by the presence of small fragments displaced elsewhere, were often repaired without a trace, presumably by homology-dependent repair (HDR). The results suggest a model whereby successful biolistic transformation relies on a combination of end joining to insert foreign DNA and HDR to repair collateral damage caused by the microprojectiles. The differing levels of genome damage observed among transgenic events may reflect the stage of the cell cycle and the availability of templates for HDR.

Figures

Figure 1.
Spectrum of Genomic Outcomes Following Transformation with Lambda and Plasmid In Rice. All Circos plots are annotated as follows. The twelve rice chromosomes are shown along with λ and plasmid pPvUbi2H magnified at 1,000× and 5,000×. The outer track shows sequence coverage over each molecule or chromosome as histograms. The inner ring demonstrates DNA copy number profiles derived from read depth, with gray shown as 1 copy, orange as 3 copies, dark red as 4 copies, and black as more than 4 copies. The inner arcs designate inter- and intra-chromosomal rearrangements. Breakpoints within the genome are colored gray, whereas the breakpoints between λ or plasmid and the genome are colored to match the respective chromosomes. (A) Rice event λ-4, which contains a long transgene array in chromosome 2. The coverage values in histogram tracks of λ and plasmid are divided by 15 and 1.5, respectively. (B) Rice event λ-7, illustrating a complex event with severe genome damage. (C) A 26 Mb region on chromosome 3 (highlighted in cyan in Figure 1B) at high resolution. The horizontal lines show copy number states and vertical bars represent inter-chromosomal breakpoints (gray) and breakpoints involving λ (plum). The arcing links show local rearrangements of the deletion-type (gray), duplication-type (red), and intra-chromosomal translocation-type (blue). For a visual depiction of how local rearrangements are defined using paired end reads, see Supplemental Figure 5. (D) A region from 25.1 Mb to 25.2 Mb on chromosome 5 (highlighted in cyan in Figure 1B) as visualized with IGV. Deleted regions are shown in red and retained regions in white (top), as indicated by the alignment of discordant reads (middle) and read depth (bottom). (E) Swarm and violin-plots showing the distribution of the size and number of deletions, duplications, and triplications in all rice events transformed with λ. Each dot in the swarm plots represents a different SV. Violin plots represent the statistical distribution, where the width shows the probability of given SV lengths.
Figure 2.
Characteristics of the Long Transgene Array in Rice Event λ-4. (A) Bionano assembly depicting the 1.6-Mb insertion in chromosome 2. The middle panel represents the reference genome, and the top and bottom panels depict the assembled transgenic and wild type chromosomes in this heterozygous line. The blue bars indicate matching restrictions sites between the reference and assembled contigs, and red bars denote restriction sites within the insertion. The nucleotide sequences above the top panel show the breakpoint sequences, with chromosome sequences highlighted in blue, λ sequences highlighted in red, and new sequences in black. (B) A 1.1-kb region assembled from Illumina data showing five λ pieces and a single fragment of chromosome 9 in rice event λ-4. The direction of the arrows indicates the 3′ ends (Tails) of λ and chromosomal genomic fragments. Four different relative orientations between intra- and inter-chromosomal pieces can be found in this sequence: Tail (3′)-Head (5′), Tail-Tail, Head-Tail, Head-Head. (C) Size distribution of λ fragments in the array as determined by PacBio sequencing.
Figure 3.
Evidence of HDR in Rice Transgenic Events. (A) Circos plot of rice transgenic event λ-5 annotated as in Figure 1. The λ coverage is divided by 15. Region 2,138,442 - 2,139,257 on chromosome 1 and region 11,041,419 - 11,041,484 on chromosome 9 are displayed in IGV windows, where displaced fragments (110 bp and 66 bp) are highlighted in red. The top panels show only discordant reads (where one end maps to the fragment and the other maps to another chromosome). The bottom panels show all reads, illustrating the ∼50% increase in read depth indicative of an HDR event. (B) Complex rearrangements observed in rice event λ-8. Regions from chromosome 2 were assembled into an array with other broken fragments at an unknown location in the genome. The damaged regions of chromosome 2 were subsequently repaired as demonstrated by the ∼50% increase in read depth.
Figure 4.
Chromothripsis-Like Outcomes and BFB-like Genomic Rearrangements in Rice and Maize Transgenic Events. (A) Circos plot of rice transgenic event λ-8 annotated as in Figure 1. The coverage of λ in the histogram track is divided by 4. (B) Copy number states of region 29.7 - 43.7 Mb on chromosome 1 (highlighted in cyan in Figure 4A) annotated as in Figure 1C. (C) Circos plot of maize transgenic event λ-3, with coverage of λ in the histogram track divided by 5. Note the region of increased copy number states on chromosome 9 indicative of BFB. (D) Circos plot of maize transgenic event λ-4, with the coverage of λ and plasmid in the histogram track divided by 15 and 10, respectively. Note the regions of increased copy number states on chromosome 1 and 6 indicative of BFB.
Figure 5.
Similar Genomic Disturbances Following Single Plasmid Transformations. Circos plots of rice lines transformed with plasmid pANIC10A-OsFPGS1 (A) and (C) and pANIC12A-OsFPGS1 (B) and (D). (A) Simple insertion. (B) Complex insertion showing a network of interlinked genomic regions. (C) Extensive damage with a deletion on chromosome 7 and apparent chromothripsis on chromosome 1 (coverage of the 10A plasmid is divided by 2). See Supplemental Figure 6B for a detailed view of the chromothripsis region on chromosome 1 (highlighted in cyan). (D) Chromosome-scale disruption with a partially trisomic chromosome 4. (E) Relationship between transgene copy number and genome breakage at sites not involving the transgene (intra- and inter-chromosomal translocations). Blue triangles and orange circles show lambda and co-bombarded plasmid from the lambda transformation events. Gray squares show data from single plasmid transformations. There are no significant correlations. Pearson correlation coefficient (R) and p-value are indicated.
Figure 6.
Models for Genomic Outcomes After Biolistic Transformation. The stage of cell cycle may influence the outcome of biolistic transformation. The models are based on the fact that in animals and presumably plants, NHEJ is the most likely repair pathway in G1 and homology directed repair (HDR) is more likely in S and G2. (A) Simple insertion. Fragments of introduced molecules (yellow) are ligated with broken ends of native chromosomes by NHEJ (nonhomologous end joining). (B) Chromothripsis-like genome rearrangements. Localized regions from native genome are shattered, resulting in many double stranded breaks. Fragments of chromosomes and introduced molecules are stitched together through NHEJ, creating complex patterns that involve the loss of genomic DNA and changes in copy number state (lost regions are circled). (C) Breakage and joining of two different chromosomes and breakage-fusion-bridge (BFB)-like genome rearrangements. When two chromosomes are broken, they can be ligated together through NHEJ. The resulting dicentric chromosome is expected to undergo BFB, which can result in stable terminal deletions. (D) DNA damage repaired by HDR. Double stranded breaks in S or G2 phase may be repaired by HDR through recombination with an intact sister chromatid.
See this image and copyright information in PMC

Comment in

Similar articles

See all similar articles

Cited by 9 articles

See all "Cited by" articles

Publication types

LinkOut - more resources

Feedback