Email updates

Keep up to date with the latest news and content from BMC Bioinformatics and BioMed Central.

This article is part of the supplement: Italian Society of Bioinformatics (BITS): Annual Meeting 2007

Open Access Research

Is plant mitochondrial RNA editing a source of phylogenetic incongruence? An answer from in silico and in vivo data sets

Ernesto Picardi* and Carla Quagliariello

Author Affiliations

Dipartimento di Biologia Cellulare, Università della Calabria, Arcavacata di Rende, (87036), Italy

For all author emails, please log on.

BMC Bioinformatics 2008, 9(Suppl 2):S14  doi:10.1186/1471-2105-9-S2-S14

Published: 26 March 2008

Abstract

Background

In plant mitochondria, the post-transcriptional RNA editing process converts C to U at a number of specific sites of the mRNA sequence and usually restores phylogenetically conserved codons and the encoded amino acid residues. Sites undergoing RNA editing evolve at a higher rate than sites not modified by the process. As a result, editing sites strongly affect the evolution of plant mitochondrial genomes, representing an important source of sequence variability and potentially informative characters.

To date no clear and convincing evidence has established whether or not editing sites really affect the topology of reconstructed phylogenetic trees. For this reason, we investigated here the effect of RNA editing on the tree building process of twenty different plant mitochondrial gene sequences and by means of computer simulations.

Results

Based on our simulation study we suggest that the editing ‘noise’ in tree topology inference is mainly manifested at the cDNA level. In particular, editing sites tend to confuse tree topologies when artificial genomic and cDNA sequences are generated shorter than 500 bp and with an editing percentage higher than 5.0%. Similar results have been also obtained with genuine plant mitochondrial genes. In this latter instance, indeed, the topology incongruence increases when the editing percentage goes up from about 3.0 to 14.0%. However, when the average gene length is higher than 1,000 bp (rps3, matR and atp1) no differences in the comparison between inferred genomic and cDNA topologies could be detected.

Conclusions

Our findings by the here reported in silico and in vivo computer simulation system seem to strongly suggest that editing sites contribute in the generation of misleading phylogenetic trees if the analyzed mitochondrial gene sequence is highly edited (higher than 3.0%) and reduced in length (shorter than 500 bp).

In the current lack of direct experimental evidence the results presented here encourage, thus, the use of genomic mitochondrial rather than cDNA sequences for reconstructing phylogenetic events in land plants.