Directionality of point mutation and 5-methylcytosine deamination rates in the chimpanzee genome

Abstract

Background: The pattern of point mutation is important for studying mutational mechanisms, genome evolution, and diseases. Previous studies of mutation direction were largely based on substitution data from a limited number of loci. To date, there is no genome-wide analysis of mutation direction or methylation-dependent transition rates in the chimpanzee or its categorized genomic regions.

Results: In this study, we performed a detailed examination of mutation direction in the chimpanzee genome and its categorized genomic regions using 588,918 SNPs whose ancestral alleles could be inferred by mapping them to human genome sequences. The C-->T (G-->A) changes occurred most frequently in the chimpanzee genome. Each type of transition occurred approximately four times more frequently than each type of transversion. Notably, the frequency of C-->T (G-->A) was the highest in exons among the genomic categories regardless of whether we calculated directly, normalized with the nucleotide content, or removed the SNPs involved in the CpG effect. Moreover, the directionality of the point mutation in exons and CpG islands were opposite relative to their corresponding intergenic regions, indicating that different forces govern the nucleotide changes. Our analysis suggests that the GC content is not in equilibrium in the chimpanzee genome. Further quantitative analysis revealed that the 5-methylcytosine deamination rates at CpG sites were highly dependent on the local GC content and the lengths of SNP flanking sequences and varied among categorized genomic regions.

Conclusion: We present the first mutational spectrum, estimated by three different approaches, in the chimpanzee genome. Our results provide detailed information on recent nucleotide changes and methylation-dependent transition rates in the chimpanzee genome after its split from the human. These results have important implications for understanding genome composition evolution, mechanisms of point mutation, and other genetic factors such as selection, biased codon usage, biased gene conversion, and recombination.

Figure 1

Rates of CpG transition, GpC transition, and 5^mC deamination varied with local GC content and SNP sequence length. (A) Rates of CpG→TpG/CpA per CpG dinucleotide (solid line) and GpC→GpT/ApC per GpC dinucleotide (dashed line) in the SNP flanking sequences. SNP GC content was calculated from the SNP flanking sequences. (B) 5^mC deamination rates, measured by the difference between the rates of CpG transition and GpC transition in A. (C) Plot of log₁₀(5^mC deamination rate) versus SNP GC content.

Figure 2

Rates of CpG transition, GpC transition, and 5^mC deamination varied with local GC content and among the genomic regions. (A) Rates of CpG→TpG/CpA per CpG dinucleotide (solid line) and GpC→GpT/ApC per GpC dinucleotide (dashed line) in the SNP flanking sequences (length category 101 nt). In CpG islands, the rates at GC content bin 0.325 were not calculated due to the insufficient number of SNPs. (B) 5^mC deamination rates, measured by the difference between the rates of CpG transition and GpC transition in A. (C) Plot of log₁₀(5^mC deamination rate) versus SNP GC content.

Figure 3

Slopes of linear regression lines. In CpG islands, the slopes for lengths 601 and 1001 nt were not included due to the insufficient number of SNPs in low GC-content bins.