The Effect Of Meiotic Recombination On Dinucleotide Bias And Processed Pseudogene Distribution

Genetic variation is the material basis of evolution that underlies the whole biology. There are two sources of genetic variation, mutation and recombination. Without recombination, genomes would undergo very little change merely by mutations, and the evolution of organisms would be severely restricted. Meiotic recombination occurs as a consequence of the exchange of genetic information between homologous chromosomes during meiosis. Recombination plays a crucial role in disjunction of the homologous chromosomes through generating chiasma. Besides, it plays various important roles in genome evolution either selectively or neutrally. With the increasing success in genome sequencing and construction of genetic maps, much progress has been made in the science of the mechanisms via which the recombination and genome interacts. Due to the complicity of the interaction between the recombination and genomic sequences in the whole genome background, however, many open questions still requires further investigation.The profile of the dinucleotide relative abundances is a "genomic signature" that reflects the overall selective pressures or mutational biases in the genome, and plays a unique and important role in the study of genome evolution and phylogeny. Therefore, revealing the pressures or factors that contribute to the formation and evolution of the genomic signature has been an important issue in the study of genome evolution. Pseudogene is a disabled copy of gene that lost the ability to encode protein. It records the evolution footprints at the molecular level and thus provides an ideal material for the study of the genome dynamics and evolution. Processed pseudogene has been a subject applied more in the study of genome evolution due to its reverse transcribed origin. It is very important to explore the evolutionary pressures that impose on the distribution of the processed pseudogenes. The present paper focus on the effect of the meiotic recombination on dinucletide bias and processed pseudogene distribution, and the underlying mechanisms are discussed. The main contributions are summarized as follows: 1. The recombination data of the Drosophila melanogaster is obtained first and then the correlation between dinucleotide bias and recombination is investigated. The results show that the overall dinucleotide bias is positively correlated with recombination rate for both non-coding and coding sequences. The correlation patterns of individual dinucleotide biases with recombination rate are presented and possible mechanisms of interaction between recombination and dinucleotide bias are discussed. We propose that recombination might influence the dinucleotide bias through a genome-wide universal mechanism, which is likely to be neighbor-dependent biased gene conversion.2. Based on the recombination data obtained from high resolution genetic map of human genome, the correlation between meiotic recombination rate and dinucleotide bias is analyzed. Our results show that the overall dinucleotide bias is positively correlated with recombination rate for coding sequences while negatively correlated for introns and intergenic sequences that both are non-coding sequences. The correlation patterns of individual dinucleotide biases with recombination rate are presented and possible mechanisms of interaction between recombination and dinucleotide bias are discussed comparing human with Drosophila melanogaster. Our results indicate that recombination has been playing an important role in the formation and evolution of the genomic signature.3. It is conventionally thought that the integration of the processed pseudogenes into genome is random. Our analysis, however, shows that the density of the processed pseudogene is correlated negatively with recombination rate and positively with gene density in human genome. Several possible models that could be invoked to explain the phenomena are proposed. The suppressor model describes an effect that processed pseudogene may act as a sppressor of recombination by reducing homology of the homologous chromosmomes. The deleterious insertion model refers to selection against deleterious insertion of processed pseudogene into the genome, which expects an accumulation of processed pseudogenes in regions of reduced recombination where selection efficiency decreased due to Hill-Robertson interference. The ectopic recombination model states processed pseudogenes will accumulate in regions of reduced recombination as a passive consequence of their elimination from high recombination rate regions, where ectopic recombination, supposed to be intense, would have more deleterious effects. The weak selection model, however, assumes that the insertion of processed pseudogenes into regions of low recombination rates might be favored by selection due to its effect of decreasing the Hill-Robertson interference between weakly selected mutations, which allows adjacent genes or exons evolve more efficiently. The positive correlation between processed pseudogene density and gene density has two possible interpretations. First, the processed pseudogene insertion into gene-dense regions might be selectively beneficial, because such insertion, as the effect descrbed in the weak selection model, could increase recombination between weakly selected mutations and facilitate adjacent genes or exons evolve more efficiently. Second, processed pseudogenes tend to be retained more in gene-dense regions than gene-rare regions probably because ectopic recombination less occurs in gene-dense regions.4. Several models including the suppressor model, deleterious insertion model, ectopic recombination model, and weak selection model might explain the negative correlation between the processed pseudogene density and recombination rate. It is of great importance to distinguish the various models. Our analysis shows that the homologous adjacent processed pseudogenes which have the potential of ectopic recombination more preferentially accumulate in the regions of low recombination rate (0.0-0.4 cM/Mb) than other processed pseudogenes, and the recombination rates of the regions where the homologous adjacent processed pseudogenes distribute are significantly lower (P<0.0001) than that corresponds to other processed pseudogenes. The results definitely indicate that there exists selective effect associated with ectopic recombination in the distribution of the homologous adjacent processed pseudogenes. Although not strong as the homologous adjacent processed pseudogenes, the processed pseudogenes that do not have the potential of ectopic recombination also exhibit a biased distribution in regions of low recombination, suggesting the existence of the effects that are not associated with ectopic recombination. Additionally, we found a length effect that long processed pseudogenes more preferentially accumulate in regions of low recombination rates.