The widespread assumption that COI and other mitochondrial genes will be ineffective DNA barcodes for anthozoan cnidarians has not been well tested for most anthozoans other than scleractinian corals. Here we examine the limitations of mitochondrial gene barcoding in the sub-class Octocorallia, a large, diverse, and ecologically important group of anthozoans. Pairwise genetic distance values (uncorrected p) were compared for three candidate barcoding regions: the Folmer region of COI; a fragment of the octocoral-specific mitochondrial protein-coding gene, msh1; and an extended barcode of msh1 plus COI with a short, adjacent intergenic region (igr1). Intraspecific variation was <0.5%, with most species exhibiting no variation in any of the three gene regions. Interspecific divergence was also low: 18.5% of congeneric morphospecies shared identical COI barcodes, and there was no discernible barcoding gap between intra- and interspecific p values. In a case study to assess regional octocoral biodiversity, COI and msh1 barcodes each identified 70% of morphospecies. In a second case study, a nucleotide character-based analysis correctly identified 70% of species in the temperate genus Alcyonium. Although interspecific genetic distances were 2× greater for msh1 than COI, each marker identified similar numbers of species in the two case studies, and the extended COI + igr1 + msh1 barcode more effectively discriminated sister taxa in Alcyonium. Although far from perfect for species identification, a COI + igr1 + msh1 barcode nonetheless represents a valuable addition to the depauperate set of characters available for octocoral taxonomy.

HEMERA (HMR) is a nuclear and plastidial dual-targeted protein. While it functions in the nucleus as a transcriptional coactivator in phytochrome signaling to regulate a distinct set of light-responsive, growth-relevant genes, in plastids it is known as pTAC12, which associates with the plastid-encoded RNA polymerase, and is essential for inducing the plastomic photosynthetic genes and initiating chloroplast biogenesis. However, the mechanism of targeting HMR to the nucleus and plastids is still poorly understood. Here, we show that HMR can be directly imported into chloroplasts through a transit peptide residing in the N-terminal 50 amino acids. Upon cleavage of the transit peptide and additional proteolytic processing, mature HMR, which begins from Lys-58, retains its biochemical properties in phytochrome signaling. Unexpectedly, expression of mature HMR failed to rescue not only the plastidial but also the nuclear defects of the hmr mutant. This is because the predicted nuclear localization signals of HMR are nonfunctional, and therefore mature HMR is unable to accumulate in either plastids or the nucleus. Surprisingly, fusing the transit peptide of the small subunit of Rubisco with mature HMR rescues both its plastidial and nuclear localization and functions. These results, combined with the observation that the nuclear form of HMR has the same reduced molecular mass as plastidial HMR, support a retrograde protein translocation mechanism in which HMR is targeted first to plastids, processed to the mature form, and then relocated to the nucleus.

BackgroundModels of sequence evolution typically assume that different nucleotide positions evolve independently. This assumption is widely appreciated to be an over-simplification. The best known violations involve biases due to adjacent nucleotides. There have also been suggestions that biases exist at larger scales, however this possibility has not been systematically explored.ResultsTo address this we have developed a method which identifies over- and under-represented substitution patterns and assesses their overall impact on the evolution of genome composition. Our method is designed to account for biases at smaller pattern sizes, removing their effects. We used this method to investigate context bias in the human lineage after the divergence from chimpanzee. We examined bias effects in substitution patterns between 2 and 5 bp long and found significant effects at all sizes. This included some individual three and four base pair patterns with relatively large biases. We also found that bias effects vary across the genome, differing between transposons and non-transposons, between different classes of transposons, and also near and far from genes.ConclusionsWe found that nucleotides beyond the immediately adjacent one are responsible for substantial context effects, and that these biases vary across the genome.