Whole-gene positive selection, elevated synonymous substitution rates, duplication, and indel evolution of the chloroplast clpP1 gene

Abstract

Background: Synonymous DNA substitution rates in the plant chloroplast genome are generally relatively slow and lineage dependent. Non-synonymous rates are usually even slower due to purifying selection acting on the genes. Positive selection is expected to speed up non-synonymous substitution rates, whereas synonymous rates are expected to be unaffected. Until recently, positive selection has seldom been observed in chloroplast genes, and large-scale structural rearrangements leading to gene duplications are hitherto supposed to be rare.

Methodology/principle findings: We found high substitution rates in the exons of the plastid clpP1 gene in Oenothera (the Evening Primrose family) and three separate lineages in the tribe Sileneae (Caryophyllaceae, the Carnation family). Introns have been lost in some of the lineages, but where present, the intron sequences have substitution rates similar to those found in other introns of their genomes. The elevated substitution rates of clpP1 are associated with statistically significant whole-gene positive selection in three branches of the phylogeny. In two of the lineages we found multiple copies of the gene. Neighboring genes present in the duplicated fragments do not show signs of elevated substitution rates or positive selection. Although non-synonymous substitutions account for most of the increase in substitution rates, synonymous rates are also markedly elevated in some lineages. Whereas plant clpP1 genes experiencing negative (purifying) selection are characterized by having very conserved lengths, genes under positive selection often have large insertions of more or less repetitive amino acid sequence motifs.

Conclusions/significance: We found positive selection of the clpP1 gene in various plant lineages to correlated with repeated duplication of the clpP1 gene and surrounding regions, repetitive amino acid sequences, and increase in synonymous substitution rates. The present study sheds light on the controversial issue of whether negative or positive selection is to be expected after gene duplications by providing evidence for the latter alternative. The observed increase in synonymous substitution rates in some of the lineages indicates that the detection of positive selection may be obscured under such circumstances. Future studies are required to explore the functional significance of the large inserted repeated amino acid motifs, as well as the possibility that synonymous substitution rates may be affected by positive selection.

Figure 1. Multiple Copies of the clpP1 Gene Region Found in Two Species.

DNA fragments sequenced from the clpP1 region (thick black bars) for Lychnis chalcedonica (Lc) and Silene fruticosa (Sf). Shown in gray is the corresponding region, with the genes (boxes) and introns (dotted lines), in Spinacia. Thick white bars indicate non-homologous flanking regions (Lc2 is located differently in the chloroplast genome). Fragments with arrows on the left side consist of a continuous sequence from rbcL to petB (c. 18 kb). Thin lines mark lack of introns. Ψ indicates stop codon. ? indicates unknown flaking regions.

Figure 2. Positive Selection in the Chloroplast clpP Gene.

A) dN/dS ratios calculated on the Eudicot tree topology. Numbers on nodes indicate classification as follows: 1 Myrtales, 2 Eurosids II, 5 Eurosids I, 6 Rosids, 13 Caryophyllales, 17 Euasterids I, 19 Euasterids II, 20 Asterids ; 3 Hologalegina, 4 Fabaceae ; 7, 8, 9 subgenus Behen, 10 subgenus Silene, 11, 12 Sileneae ; 14 Solaneae, 15 Solanoideae, 16 Solanaceae . Values on branches are the dN/dS ratios, S: dN = 0, N: dS = 0. Ratios significantly above one (Bonferroni-corrected) are indicated by asterisks: ***P<0.001, *P<0.05. B) dN- and dS-branch (in grey and black, respectively) lengths imposed on topology from Bayesian analysis of third codon positions from the clpP1 exons Numbers on nodes are Bayesian posterior probabilities (Bpp). Only Bpp values>0.50 are shown.

Figure 3. dS and dN Values of Four Chloroplast Genes.

Diagram shows the pairwise dS values (top) and dN values (bottom) between eleven species of Sileneae and the outgroup (Heliosperma alpestre) for four chloroplast genes (clpP1; 591 bp, psbB; 1527 bp, cemA; 639 bp, and petA; 963 bp). To the right of the vertical gray line is the pairwise comparison between Oenothera elata and Eucalyptus globulus, for the same genes.

Figure 4. Bayesian Consensus Phylograms of the Tribe Sileneae.

A) The clpP1 exon sequences, and B) the clpP1 intron sequences. Branches in bold have Bayesian posterior probabilities (Bpp) of 1.00. Only Bpp values above 0.90 are shown at nodes. Note that A and B are drawn at different scales. ** indicates taxa that lack introns in clpP1.

Figure 5. The clpP1 Gene and Flanking Regions with Primer Sites.

Approximate position of general primers for Sileneae with reference to Spinacia. Boxes correspond to protein coding genes with their name and position in Spinacia. Boxes above the line denote genes that are transcribed from left to right, and those below are transcribed from right to left. Dotted lines represent introns.