Targeted exome sequencing of unselected heavy-ion beam-irradiated populations reveals less-biased mutation characteristics in the rice genome

Abstract

Heavy-ion beams have been widely utilized as a novel and effective mutagen for mutation breeding in diverse plant species, but the induced mutation spectrum is not fully understood at the genome scale. We describe the development of a multiplexed and cost-efficient whole-exome sequencing procedure in rice, and its application to characterize an unselected population of heavy-ion beam-induced mutations. The bioinformatics pipeline identified single-nucleotide mutations as well as small and large (>63 kb) insertions and deletions, and showed good agreement with the results obtained with conventional polymerase chain reaction (PCR) and sequencing analyses. We applied the procedure to analyze the mutation spectrum induced by heavy-ion beams at the population level. In total, 165 individual M₂ lines derived from six irradiation conditions as well as eight pools from non-irradiated 'Nipponbare' controls were sequenced using the newly established target exome sequencing procedure. The characteristics and distribution of carbon-ion beam-induced mutations were analyzed in the absence of bias introduced by visual mutant selections. The average (±SE) number of mutations within the target exon regions was 9.06 ± 0.37 induced by 150 Gy irradiation of dry seeds. The mutation frequency changed in parallel to the irradiation dose when dry seeds were irradiated. The total number of mutations detected by sequencing unselected M₂ lines was correlated with the conventional mutation frequency determined by the occurrence of morphological mutants. Therefore, mutation frequency may be a good indicator for sequencing-based determination of the optimal irradiation condition for induction of mutations.

Keywords: heavy-ion beam; mutagenesis; mutation rate; rice (Oryza sativa); whole-exome sequencing.

Figure 1

Mutation types detected in 13 heavy‐ion beam‐induced mutants of rice The proportion of each type of mutation (SNV, single‐nucleotide variants; DEL, deletions; INS, insertions; RPL, replacements; INV, inversions) is shown. Targeted exome sequences were obtained for 11 mutants, induced by carbon‐ion and neon‐ion irradiation, followed by variant calling with the Genome Analysis Toolkit and Pindel programs integrated into the bioinformatics pipeline. Note that the 14‐45 mutant arose by seed contamination or outcrossing.

Figure 2

Phenotypes of the 5‐12 mutant and its transgenic derivatives Leaf blades were sampled from advanced‐stage plants and imaged under white light‐emitting diode (LED) illumination. wild‐type (WT): Nipponbare (wild‐type), 5‐12: a striated‐leaf mutant isolated from a population irradiated with carbon‐ion beams, Os01g0109300 complementation +: genomic Os01g0109300 fragment introgressed by Agrobacterium‐mediated transformation, Os01g0109300 complementation −: a segregate progeny of a transformed 5‐12 plant lacking the introgressed Os01g0109300 fragment. Bar: 2 mm.

Figure 3

Distribution and Gaussian fit of the number of mutations per line The total number of mutations detected for each of 110 M₂ lines derived from 150 Gy irradiation of dry seeds. On average 9.06 ± 0.37 (average ± SE) mutations were detected in an unselected M₂ line. The solid line indicates goodness of fit to the Gaussian distribution (μ = 8.66, σ = 4.37).

Figure 4

Nucleotide preference in single‐nucleotide mutations The number of transition and transversion events detected from the 110 M₂ lines derived from 150 Gy irradiation of dry seeds for each combination of original and alternate bases. The total number of transition and transversion events was 271 and 302, respectively (ratio 1:1.07).

Figure 5

Chromosomal distribution and frequency of mutations The chromosomal location of all mutation events detected within the target exon regions for each M₂ line derived from 150 Gy irradiation of dry seeds. The relative mutation frequency (per Mbp) is plotted above. The mutations were distributed on all chromosomes. Note that fewer target exon regions were subjected to target‐capture sequencing in the centromeric regions.

Figure 6

Number of line‐specific mutations in heavy‐ion‐irradiated, unselected Nipponbare populations The mean number of line‐specific mutations (bars) for populations subjected to different irradiation doses (15, 75, 100 and 150 Gy), linear energy transfer (30 and 50 keV μm⁻¹), and seed conditions (dry and imbibed seeds). The total number of mutations per irradiation dose is indicated by solid circles. Error bars represent the standard error. NIC, non‐irradiated control.