Genomic structural variation in an alpha/beta hydrolase triggers hybrid necrosis in wheat

Abstract

Hybrid necrosis, a century-old mystery in wheat, is caused by complementary genes Ne1 and Ne2. Ne2, encoding a nucleotide-binding leucine-rich repeat (NLR) immune receptor, has been cloned, yet Ne1 remains elusive. Here, we report that Ne1, which encodes an alpha/beta hydrolase (ABH) protein generated by structural variation, triggers hybrid necrosis with Ne2 by activating autoimmune responses. We further verify that not only allelic variation but also copy number variation (CNV) of Ne1 are pivotal for hybrid necrosis diversity in wheat. Ne1 likely originates from wild emmer wheat, potentially through duplication and ectopic recombination events. Unlike Ne2, which is frequently selected for rust resistance in wheat breeding, the lower prevalence of Ne1 in modern wheat cultivars is attributed to its association with hybrid necrosis. Altogether, these findings illuminate the co-evolution of the NLR/ABH gene pair in plant development and innate immunity, offering potential benefits for wheat breeding.

Fig. 1. Map-based cloning of Ne1.

a Phenotypes of necrotic line M114 (Ne1Ne1Ne2Ne2), Zhoumai22 (ZM, ne1ne1Ne2Ne2), and F₁ plant derived from M114 × ZM at the grain filling stage in the field. Scale bar in plant, 8 cm; Scale bar in leaf, 2 cm. b Fine genetic map of Ne1 constructed using the F₂ populations derived from M114 × ZM. c Physical map of Ne1 region according to Chinese Spring (CS) reference genome sequence IWGSC RefSeq v2.1. The blue oval indicates Ne1. d Fine genetic map of Ne1 constructed using segregation populations derived from Zhengnong17 (ZN17, ne1ne1Ne2Ne2) and Yangbaimai (YBM, Ne1Ne1ne2ne2). Ne1 is highlighted in blue. e Gene structure and sequence variation of ABH.1 gene. Black boxes and gray lines indicate exons and introns, respectively. The variation of coding sequence (c) and amino acid (p) sequences between CS, YBM, and M114 are indicated. f Protein structure prediction of ABH.1. Hydrolase domain, AAA ATPase with vWA domain, and membrane protease YdiL domain are represented in orange, green, and purple, respectively. The signal peptide is indicated in red. Source data are provided as a Source Data file.

Fig. 2. Functional validation of ABH.1 by mutagenesis.

a, c Phenotypes of non-necrotic mutants generated from the necrotic NIL-Ne1 (Ne1Ne1Ne2Ne2) and M114 (Ne1Ne1Ne2Ne2) background by ethyl methanesulfonate (EMS) treatment. Gene structure and EMS mutant analysis of ABH.1 (b) and Ne2 (d). The positions of the EMS-derived loss-of-function mutations are indicated by black lines. Black boxes and gray straight lines represent exons and introns, respectively. The coding sequence (c) changes and their predicted effects on the amino acid (p) are indicated. Scale bar in (a and c), 1 cm. Mutation names in red indicate nonsense mutations. Source data are provided as a Source Data file.

Fig. 3. Transgenic validation of ABH.1.

a Structure of proUbi::ABH.1^YBM used for transformation of wheat cultivar Fielder. The construct contains the ABH.1^YBM coding sequence (CDS) and maize ubiquitin promoter region (proUbi). LB left border; RB, right border; Nos Ter, termination sequence of the Agrobacterium tumefaciens nopaline synthase gene. b Genetic complementation of ABH.1 and Ne2 was performed by crossing independent transformants (OEABH.1^YBM#1-#4) carrying proUbi::ABH.1^YBM and wheat cultivar ZZ6903 with the ne1ne1Ne2Ne2 genotype, which induced the expression of hybrid necrosis. Representative leaves of four F₁ hybrids and their corresponding parental lines are presented. Scale bar, 1 cm. c Structure of proABH.1^M114::ABH.1^M114 used for transformation of the Ne2 overexpression transgenic line OE-T₁−1-1 (ne1ne1Ne2Ne2, OENe2) in the Fielder genetic background and the F₁ plants (ne1ne1Ne2ne2) of ZM × Fielder, respectively, by Agrobacterium-mediated transformation. The construct proABH.1^M114::ABH.1^M114 consisted of a 13,118 bp genomic fragment of ABH^M114 from M114, comprising 8157 bp the entire gene body, 2406 bp upstream native promoter and 2,556 bp downstream regulatory sequences, respectively. d Phenotypes of OENe2, positive transgenic lines (OENe2 + COMABH.1^M114#1-#3) in OENe2 background, ZM, positive transgenic lines (Ne2Ne2 + COMABH.1^M114#1-#3, and ne2ne2 + COMABH.1^M114#1) in ZM × Fielder background. Scale bar, 1 cm. Source data are provided as a Source Data file.

Fig. 4. Allelic and copy number variation of ABH.1.

a Haplotype analysis of ABH.1 using whole-genome re-sequencing data from 540 hexaploid and 157 tetraploid wheat accessions. Variants at the coding sequence of ABH.1 were clustered into seven major haplotypes (Hap1-Hap7), while 366 accessions lacked ABH.1 (Hap0). Introns and exons are represented by lines and black boxes, respectively. The coding nucleotide and amino acid sequences of Chinese Spring (CS) are shown in black font, and the bracketed numbers represent the relative positions of nucleotide or amino acid sequences concerning ATG or Met, respectively. Deletions are indicated by the symbol ‘-’. YM14, Yumai14; YBM, Yangbaimai. b Four major representative structural variations of the NE1 locus. Reads from these wheat samples were aligned to the Chinese Spring reference genome sequence IWGSC RefSeq v2.1, and coverage depth for each 5 kb window was computed. The coverage depth for each 5 kb window was then normalized based on the total coverage depth. Finally, the normalized depth for the target region (Chr5B:385.5-390.5 Mb) was presented in the form of a scatter plot, with each data point representing a 5 kb window. c Coverage depth of the ABH.1 gene in the published re-sequence dataset for 540 hexaploidy wheat accessions. The ABH.2 (TraesCS5B03G0565000) gene is situated downstream of ABH.1 (TraesCS5B03G0561800), with the sequencing coverage depth of ABH.2 used as an indicator of normal resequencing outcomes within this genomic region. Reads from wheat samples were aligned to the Chinese Spring RefSeq v2.1, and coverage depth was determined for each gene. The coverage depth for each gene was subsequently normalized by the total coverage depth, and the resulting normalized depth for the target gene was depicted in a scatter plot, with each data point representing a specific wheat accession. The n number is the sample size used to derive statistics. d The expression of ABH.1 was compared among the four CNV (copy number variation) groups (two-sided Student’s t test; centerline, median). RNA-seq data from 305 common wheat genomes (one biological replicate) were used to quantify ABH.1 expression for the four CNV groups. P-values are shown at the top of data points, and n, the sample size used to derive statistics, is listed at the bottom of the figure. TPM, transcripts per million. Source data are provided as a Source Data file.

Fig. 5. Effects of allelic and copy number variations at ABH.1 locus on hybrid necrosis.

a Phenotypes of a series of materials carrying different ABH.1 alleles. Scale bar, 10 cm. b Structural variations of the ABH.1 locus from materials carrying different ABH.1 alleles. c–e Phenotypes of ZZ6903, and the representative F₁ plants derived from crosses between wheat accessions (carrying ne2ne2 and different ABH.1 haplotypes) × ZZ6903 (ne1ne1Ne2Ne2 genotype). ZZ6903, Zhengzhou 6903; YM14, Yumai 14; YBM, Yangbaimai; CS, Chinese Spring. Scale bar in plant, 10 cm. Source data are provided as a Source Data file.

Fig. 6. Evolutionary trajectory and distribution of ABH.1.

a A proposed model for the evolutionary trajectory of ABH.1. Following a segmental duplication event encompassing the genomic loci of ABH.2, SP.4, and SP.5, designated as CNV-0 type, there was a subsequent deletional event within the duplicated segment. This resulted in the loss of the entire SP.4 gene sequence, as well as the promoter region and the initial 315 base pairs of the first exon of the ABH.2 gene. Subsequently, transposable elements (TEs) were inserted into the 3’ downstream region of the SP.5 gene, situated between Fragment A and Fragment B. The integration of these TEs with a segment of the downstream sequence of the SP.5 gene, in conjunction with the truncated remnant of the ABH.2 gene, catalyzed the genesis of a novel functional gene entity, characterized as the ABH.1 gene (CNV-1 type). As a hotspot for genomic recombination, this region has also undergone events resulting from one (CNV-2 type) and two (CNV-3 type) duplications. b Global distribution of Ne1/ABH.1 and ne1 alleles in a worldwide wheat collection (n = 853). Samples were categorized into Ne1/ABH.1 allele type and ne1 allele type based on their genetic type in the NE1 locus, and the proportions of samples in different regions were represented using pie charts. Initially, the geographic coordinates of the samples were imported as point layers and overlaid with corresponding national and continental boundary data. Subsequently, pie charts were added at the sampling locations in each region, with different colors representing distinct haplotype categories and the size of the pie charts reflecting the total number of samples in that region. c Ne1/ABH.1 allele was negatively selected in Chinese wheat breeding. A total of 89 landrace wheat and 449 cultivar wheat accessions were used to investigate the distribution of Ne1/ABH.1 and ne1 alleles. Source data are provided as a Source Data file.