Comparative Genomic Analysis of Adaptive and Economic Traits Related Genes in Southern Pines

Abstract

Four major Southern pines, Pinus echinata Mill., P. elliottii Engelm., P. palustris Mill. and P. taeda L. are evolutionarily young and closely related. They have not been intensely researched except P. taeda. In this study we addressed the questions of exonintron structure, nucleotide variation and neutrality in adaptive and economic traits related genes, and phylogenetic relationships between these pines. Using publically available data in the NCBI databases, we first developed a series of statistical regression models. We defined functional relationships between the parameters that can be easily estimated from a small data sample (e.g. mean exon length and exon/gene ratio), and parameters that are difficult to assess (e.g. number of genes and exons). Second, we examined the effects of selection upon the set of studied genes in the four pines. We collected data from individuals representing all four Southern pines and merged them with previously published data, and applied four neutrality tests: Tajima’s D, McDonald- Kreitman (MK), Hudson-Kreitman-Aguade (HKA), and synonymous-nonsynonymous nucleotide substitutions ratio. Finally, we analyzed phylogenetic relationships between the four Southern pines, and with respect to other selected pine species (P. radiata, P. pinaster and P. sylvestris), for which the nucleotide sequence data orthologous to the sequences newly generated in this study were available in the NCBI GenBank. We applied Maximum Parsimony, Maximum Likelihood and Bayesian Inference approaches. Based on the statistical models we expect about 13-14 thousand genes in an organism with the mean exon length of 334.8 bp (like P. taeda). This number could be higher in plants (20-21 thousand). Furthermore, we identified signatures of selection in some of the studied genes, and demonstrated that different parts of a gene could be under different forms of selection. Therefore, the results of the neutrality tests performed at the entire gene level could be misleading. Finally, using twelve nuclear loci we confirmed very tight phylogenetic relationships within the subsection Australes, but the conclusions were not robust. Using two exclusive sets of three genes led to robust but conflicting results. Therefore, we demonstrated that conclusions about “species” trees based on “gene” trees may be misleading, especially for closely related species.