Mosaic origins of a complex chimeric mitochondrial gene in Silene vulgaris

Abstract

Chimeric genes are significant sources of evolutionary innovation that are normally created when portions of two or more protein coding regions fuse to form a new open reading frame. In plant mitochondria astonishingly high numbers of different novel chimeric genes have been reported, where they are generated through processes of rearrangement and recombination. Nonetheless, because most studies do not find or report nucleotide variation within the same chimeric gene, evolution after the origination of these chimeric genes remains unstudied. Here we identify two alleles of a complex chimera in Silene vulgaris that are divergent in nucleotide sequence, genomic position relative to other mitochondrial genes, and expression patterns. Structural patterns suggest a history partially influenced by gene conversion between the chimeric gene and functional copies of subunit 1 of the mitochondrial ATP synthase gene (atp1). We identified small repeat structures within the chimeras that are likely recombination sites allowing generation of the chimera. These results establish the potential for chimeric gene divergence in different plant mitochondrial lineages within the same species. This result contrasts with the absence of diversity within mitochondrial chimeras found in crop species.

Figure 1. Structure and genomic context of the two variants of bobt, bobt_KR (A) and bobt_MV (B).

bobt_KR and bobt_MV differ in 43 nucleotide sites, but share the same general atp1-cox2-unknown ORF structure between the start and stop sites. They share a segment of atp1 at the 3′ end downstream of the stop codon, a segment of cox2, and a segment of unknown origin. The fourth segment, located just downstream of stop codon corresponds to atp1. Its size differs between both copies, because bobt_KR is adjacent to cob, whereas bobt_MV is not. Horizontal arrows designate the locations of the primers (atp1 lo, atp1 up) originally used to discover bobt via PCR co-amplification with the functional copy of atp1. A full-length atp1 gene is shown below the bobt variants, the homologous regions are indicated by dotted lines. A scale in bp is given above.

Figure 2. Southern hybridization results with bobt-specific and atp1-specific probes.

Genomic DNA of Silene from various sites was digested with EcoRI and hybridized with probes derived from atp1 and bobt_MV. The same membrane was hybridized with both (A) an bobt_MV probe and (B) an atp1 probe. Lanes 1–5 S. vulgaris Mt. View; 6–8 S. vulgaris Krasnoyarsk; 9–11 S. vulgaris Beagle; 12 S. vulgaris Krasnoyarsk; 13 S. vulgaris Beagle; 14–16 S. latifolia Prague. Individuals from the Beagle population differ in Southern-RFLP pattern of atp1 region, which is very common in S. vulgaris populations. Faint, but visible, bands visible in Beagle and Krasnoyarsk DNA hybridized with a bobt_MV probe may correspond to weakly homologous regions in nuclear or mt DNA. The bands corresponding to bobt_KR are a bit weaker than bobt_MV bands due to nucleotide divergence between the two variants. One band detected by bobt probes suggests the existence of only one bobt copy in the mt genome, whereas two or three atp1-specific bands reflect the existence of two or three atp1 copies. Marker sizes are shown along the right hand side of each blot. The results of PCR with bobt specific primers are shown below (+,−).

Figure 3. Maximum likelihood phylogenies of atp1 and homologous regions of bobt from Silene vulgaris accessions.

(A) Phylogeny using the complete sequence of atp1. The same topology was recovered when only the regions with homology to bobt were used. Numbers above lineages indicate bootstrap support for nodes in the phylogeny with the complete sequence, whereas numbers below the line represent support for node using only regions with homology to bobt. (B) Phylogeny including the shared regions of bobt, the atp1 genes found in the same individuals as bobt (KR & MV), and two additional atp1 copies from individuals from the Kovary CR population. Silene latifolia, Vitis vinifera, and Beta vulgaris atp1 sequences were used as outgroups. (C) Phylogeny of first 192 bases of the shared regions of bobt and atp1 homologous region found at the 5′ end of bobt. (D) Phylogeny of the region 3′ to the stop codon of bobt and the homologous region in atp1. This phylogeny includes 78 bases than are not found in bobt_KR, but are fond in the other accessions.

Figure 4. Transcript levels of bobt_MV and bobt_KR in flower buds estimated by qRT PCR.

Expression is presented as a ratio of the accumulation of PCR product for bobt and the mt 18S rRNA reference gene. Ratios for bobt_MV are multiplied by 100 to allow visualization of difference on the same scale as bobt_KR. The standard deviation was calculated on the basis of 3–5 sibling pairs F and H. The differences in bobt expression between genders were significant in both MV and KR plants (t-test P<0.05).

Figure 5. Relative transcript levels in flower buds assessed at different regions of the bobt_KR-cob co-transcript.

(Relative) transcript level at the atp1-cob junction and cob is more than twice that in the unknown region of bobt_KR. No signal was detected when RNA instead of cDNA was added to qPCR reaction mixture, which excluded contamination of RNA samples with genomic DNA. (A) Expression is presented as the ratio of the accumulation of PCR product for the specific region and the mt 18S rRNA reference gene and (B) copy numbers of different regions of bobt_KR-cob DNA as a ratio of the accumulation of PCR product for the specific region and the mt 18S rRNA reference gene. Whereas target gene and reference copy numbers are approximately equal, atp1-cob cotranscript and cob transcript levels are more than one order of magnitude lower than mt 18S rRNA. The black lines below the gene indicate the positions of PCR products used for quantitation. The standard deviation was calculated on the basis of 5 sibling pairs F and H. The significant differences (t-test P<0.05) between F and H are marked by **.