An efficient allotriploid-mediated system of generating genomic introgression from Brassica oleracea to B. rapa

Abstract

Germplasm resources with rich genetic diversities are indispensable, but often fail short to need the demand for crop breeding (Snowdon et al., 2015). Genomic introgression in hybrids of closely related species is considered to be an important source for genetic diversity enhancement, which is mainly through homoeologous exchanges (HEs) among homoeologous chromosomes (Zhou et al., 2021). However, the technical bottleneck in terms of the low introgression frequency of genome segments in interspecific hybrid offspring (Quezada-Martinez et al., 2021), has not been solved, even though the cross between diploid B. rapa (A_rA_r) and B. napus (A_nA_nC_nC_n) was reported to increase the recombination frequency between A_n and A_r (Boideau et al., 2021). On the other hand, how to elevate homoeologous chromosome recombination and genome introgression between two different species remains to be studied. Here we report an allotriploid-involved genetic system that can tremendously increase introgression frequency between two closely related species and thereby generate rich phenotypic variations, and a pipeline for genome-wide identification of introgressed segments.

The chromosomes of a diploid Brassica rapa L. ssp. pekinensis (Chinese cabbage) NDCCBr (AA, 2n = 20) were doubled by the colchicine treatment to create a tetraploid B. rapa 9403Br (AAAA, 2n = 40) that was then hybridized with a diploid B. oleracea L. var. capitata (cabbage) 9501Bo (CC, 2n = 18) to produce allotriploid hybrids (AAC, 2n = 29) via ovary and ovule culture. Three allotriploids were subsequently backcrossed with another genotype diploid B. rapa 01-4-11Br, then 12 resultant offsprings were randomly selected and selfed for two successive generations. Finally, microspores were isolated from these selfed progenies and cultured to generate homozygous B. rapa–B. oleracea introgression lines (Boideau et al., 2021) (Figure 1a). The introgression lines were diploid, showing normal meiotic behaviour (Figure S1a–h). The presence of B. oleracea introgressed segments was confirmed by B. oleracea genome-specific InDel markers compared to B. rapa (Figures S1i and S2).

Figure 1

Open in figure viewerPowerPoint

(a) The pipeline for the construction of introgression lines and identification of B. oleracea introgressed segments. For SNP tracking-based identification of introgressed segments see the detailed methods in Data S1. (b) Stacked bar graph showing the genomic proportions of 9403Br, 9501Bo and 01-4-11Br in each line. (c) Distribution of introgressed segments from 9501Bo (A) and homologous recombination segments between 9403Br and 01-4-11Br (B) along the 10 B. rapa chromosomes. (d) Frequency for different lengths of introgressed segments. (e) Comparison of length between the replaced B. rapa segments and introgressed B. oleracea segments in all accessions. ***P < 0.001, Student's t-test. (f) The representatives of abundant phenotypic variations. A, swollen roots. B, leaf colour. C, plant architecture. D, leafy head shape. E, primary colour inside cut. Scale bars: 5 cm in panel A, 10 cm in panels B–E. (g) The heatmap represents the normalized phenotypic variation data. From left to right: three parents, 9403Br, 9501Bo and 01-4-11Br; and introgression lines, GDH1–GDH46. (h) Presence of the introgressed B. oleracea segment on the A03 chromosome of GDH23. The purple line represents the position of molecular markers for the B. oleracea segment. (i) Comparison of flowering times in the progeny populations with (+) and without (−) the 4.1 Mb introgressed B. oleracea segment. *P < 0.05, Student's t -test.

To check the frequency of genome introgression, we resequenced introgressed lines, and performed genome analyses (Figure 1d). We surprisingly found that the allotriploid-involved genetic system was able to generate very high introgression events (Figure 1b,c; Table S1), likely via homoeologous exchanges/replacement mediated by allotriploidy. In total, 1301 introgressed B. oleracea segments (introgression from C genome to A genome) were detected in the 46 introgression lines with an average of 28.28 segments and a range of from 3 to 67 in individual lines, extremely higher than those reported in previous studies (Figure S3). The majority of B. oleracea segments introgressed into the B. rapa genome are those with sizes of 0.5–2.5 Mb (Figure 1d; Figure S3). A total of 3212 segments of 01-4-11Br (homologous recombination between the A genomes of two different genotypes) were detected in the 46 lines with an average of 69.83 segments and a range of from 18 to 110 in individual lines (Figure S3; Table S2). 0.71%–27.40% and 55.57%–97.69% of the genome of the recipient parent 9403Br were replaced by 9501Bo and 01-4-11Br in introgression lines, respectively (Figure 1b; Tables S2 and S3), indicating excessive variations in genomic extent. Interestingly, the sizes of the introgressed segments from B. oleracea were significantly longer than the replaced counterpart segments of the B. rapa genome in the introgression lines (Figure 1e; Figure S4). The introgression frequency was 0.010–0.226 segments per Mb in the B. rapa–B. oleracea lines, which was significantly higher than that in natural B. napus accessions (0.004–0.051) and synthetic B. napus accessions (0.008 to 0.086). Introgression ratio ranged from 0.71% to 27.40%, significantly higher than 0.01%–2.86% or 0.09%–11.39% in natural or synthetic B. napus accessions (Table S3). We speculate that there are multiple reasons for the occurrence of higher introgression, such as sequence collinearity between A and C genomes, or/and alterations in chromosome spatial structure. Furthermore, after evaluating the linkage disequilibrium (LD) of 199 natural B. rapa accessions, particularly those with strong LD regions (r² > 0.7). We found that a number of introgressed B. oleracea segments occurred within the regions with strong LD (Figures S5 and S6; Table S4), indicating that elevated recombination mediated by allotriploid can break highly linked regions.

Accordingly, we observed very wide variations of phenotypes in the introgression lines, which can be divided into three categories: (i) absence in both parental B. rapa and B. oleracea, including swollen roots (Figure 1fA); (ii) presence specifically in B. oleracea but not B. rapa, such as beneficial glucoraphanin (Figure S7); and (III) increased variation range of the B. rapa traits (Figure 1fB–E,g; Figures S7 and S8). To demonstrate how an introgressed B. oleracea segment affects phenotypic changes, we examined flowering time in the introgression line GDH23. A 4.1-Mb introgressed segment from B. oleracea chromosome C03 harbouring four flowering time genes, BoAGL2, BoFLC, BoFY and BoNF-YA1 exists in GDH23 (Figure 1h). We backcrossed GDH23 with 01-4-11Br to construct a pair of near isogenic lines, with and without the 4.1-Mb introgressed segment. We found that flowering times of the 4.1-Mb segment-contained plants were delayed by 6 days when compared to those free of the segment (Figure 1i).

The chromosome complement in allotriploid hybrid AAC contains a single C and a pair of As which can form trivalents during meiosis. Such synapsis is likely to increase introgression rate from C to A genomes. Moreover, unlike AC, AAC is fertile and introgressions are able to pass on to its offspring. In this work, we establish an effective genetic system in which allotriploid can mediate extreme elevation of the frequency of genomic introgression, and thereby enrich genetic diversities of the B. rapa genome and phenotypic variations in the introgression lines.