A novel discovery of a long terminal repeat retrotransposon-induced hybrid weakness in rice

Abstract

Hybrid weakness is a post-zygotic hybridization barrier frequently observed in plants, including rice. In this study, we describe the genomic variation among three temperate japonica rice (Oryza sativa ssp. japonica) varieties 'Aranghyangchalbyeo' ('CH7'), 'Sanghaehyangheolua' ('CH8') and 'Shinseonchalbyeo' ('CH9'), carrying different hybrid weakness genes. The reciprocal progeny obtained from crossing any two varieties displayed characteristic hybrid weakness traits. We mapped and cloned a new locus, Hwc3 (hybrid weakness 3), on chromosome 4. Sequence analysis identified that a long terminal repeat (LTR) retrotransposon was inserted into the promoter region of the Hwc3 gene in 'CH7'. A 4-kb DNA fragment from 'CH7' containing the Hwc3 gene with the inserted LTR retrotransposon was able to induce hybrid weakness in hybrids with 'CH8' plants carrying the Hwc1 gene by genetic complementation. We investigated the differential gene expression profile of F1 plants exhibiting hybrid weakness and detected that the genes associated with energy metabolism were significantly down-regulated compared with the parents. Based on our results, we propose that LTR retrotransposons could be a potential cause of hybrid weakness in intrasubspecific hybrids in japonica rice. Understanding the molecular mechanisms underlying intrasubspecific hybrid weakness is important for increasing our knowledge on reproductive isolation and could have significant implications for rice improvement and hybrid breeding.

Keywords: japonica; F1 hybrids; LTR retrotransposon; gene expression profiles; genome re-sequencing; hybrid weakness; polymorphism; rice (Oryza sativa).

Fig. 1.

Morphological and phenotypic characterization of hybrid weakness. (A–C) Morphological evaluation of hybrid weakness in ‘CH7’, ‘CH8’, and the reciprocal hybrid ‘CH7/8’ and ‘CH8/7’ F₁ hybrid plants (A), ‘CH7’, ‘CH9’, and the reciprocal hybrid ‘CH7/9’ and ‘CH9/7’ F₁ hybrid plants (B), ‘CH8’, ‘CH9’, and the reciprocal hybrid ‘CH8/9’ and ‘CH9/8’ F₁ hybrid plants (C) at 30 d after transplantation. (D) Plant height comparison of the parental plants and their hybrid weakness F₁ progeny at the seedling and tillering stages. (E) Root length comparison of the parental plants and their F₁ hybrids with hybrid weakness at the seedling stage. Data points in (D, E) represent the mean ±SE. (This figure is available in color at JXB online.)

Fig. 2.

Comparison of yield components in the parents and their F₁ progeny exhibiting hybrid weakness at the maturity stage. Comparison of tiller number (A), panicle length (B) and panicle number per plant (PNPP) (C) between parents and their F₁ exhibiting hybrid weakness. Significant difference determined by the one-way-ANOVA: **P<0.01, ***P<0.001; ns, P>0.05).

Fig. 3.

Genetic characterization of hybrid weakness. (A) Physical mapping of Hwc3 on rice chromosome 4. Approximate location of the RM3687 and RM5473 DNA markers used for recombinants around the Hwc2 locus. The square denotes the Hwc2 locus reported by Kuboyama et al. (2009). The circle represents the candidate Hwc3 gene Os.89494. (B) RT-PCR analysis of the candidate Hwc3 gene, where 7 represents the cDNA of ‘CH7’, 8 represents the cDNA of ‘CH8’, 7/8 represents the cDNA of ‘CH7/8’ F₁ hybrid and ck represents the genomic DNA of ‘CH7’. The Actin gene was used as the internal control to normalize gene expression. (C) Gene expression analysis at the Hwc3 locus, where 7 represents the cDNA of ‘CH7’, 8 represents the cDNA of ‘CH8’, 7/8 represents the cDNA of ‘CH7/8’ F₁ hybrid and ck represents genomic DNA of ‘CH7’. (D) Genomic organization at Hwc3 locus in ‘CH7’ and ‘CH8’. (This figure is available in color at JXB online.)

Fig. 4.

Genetic complementation and postulated model of hybrid weakness. (A) AT70 and AT71 segments derived from ‘CH7’ used for complementation tests. (B) Independent transformants carrying AT70 and AT71 fragments in ‘CH8’ induced the expression of hybrid weakness. ‘LiyuB’ was used as the wild-type. Each line was represented by five independent transgenic plants. (C) Postulated model for Hwc1–Hwc3 interaction leading to hybrid weakness in ‘CH7/8’ F₁ hybrid, where X represents no transcription of the Hwc3 gene and O represents transcription activation of the Hwc3 gene due to the inserted promoter and transcription activator Hwc1 in the ‘CH8’ genome. (This figure is available in color at JXB online.)