Male-driven biased gene conversion governs the evolution of base composition in human alu repeats

Regional biases in substitution pattern are likely to be responsible for the large-scale variation in base composition observed in vertebrate genomes. However, the evolutionary forces responsible for these biases are still not clearly defined. In order to study the processes of mutation and fixation across the entire human genome, we analyzed patterns of substitution in Alu repeats since their insertion. We also studied patterns of human polymorphism within the repeats. There is a highly significant effect of recombination rate on the pattern of substitution, whereas no such effect is seen on the pattern of polymorphism. These results suggest that regional biases in substitution are caused by biased gene conversion, a process that increases the probability of fixation of mutations that increase GC content. Furthermore, the strongest correlate of substitution patterns is found to be male recombination rates rather than female or sex-averaged recombination rates. This indicates that in addition to sexual dimorphism in recombination rates, the sexes also differ in the relative rates of crossover and gene conversion.     

DNA strand specificity for UV-induced mutations in mammalian cells.

The influence of DNA repair on the molecular nature of mutations induced by UV light (254 nm) was investigated in UV-induced hprt mutants from UV-sensitive Chinese hamster cells (V-H1) and the parental line (V79). The nature of point mutations in hprt exon sequences was determined for 19 hprt mutants of V79 and for 17 hprt mutants of V-H1 cells by sequence analysis of in vitro-amplified hprt cDNA. The mutation spectrum in V79 cells consisted of single- and tandem double-base pair changes, while in V-H1 cells three frameshift mutations were also detected. All base pair changes in V-H1 mutants were due to GC----AT transitions. In contrast, in V79 all possible classes of base pair changes except the GC----CG transversion were present. In this group, 70% of the mutations were transversions. Since all mutations except one did occur at dipyrimidine sites, the assumption was made that they were caused by UV-induced photoproducts at these sites. In V79 cells, 11 out of 17 base pair changes were caused by photoproducts in the nontranscribed strand of the hprt gene. However, in V-H1 cells, which are completely deficient in the removal of pyrimidine dimers from the hprt gene and which show a UV-induced mutation frequency enhanced seven times, 10 out of 11 base pair changes were caused by photoproducts in the transcribed strand of the hprt gene. We hypothesize that this extreme strand specificity in V-H1 cells is due to differences in fidelity of DNA replication of the leading and the lagging strand. Furthermore, we propose that in normal V79 cells two processes determine the strand specificity of UV-induced mutations in the hprt gene, namely preferential repair of the transcribed strand of the hprt gene and a higher fidelity of DNA replication of the nontranscribed strand compared with the transcribed strand.     

SHORT INDELS ARE SUBJECT TO INSERTION‐BIASED GENE CONVERSION

Recombination between homologous loci is accompanied by formation of heteroduplexes. Repairing mismatches in heteroduplexes often leads to single nucleotide substitutions in a process known as gene conversion. Gene conversion was shown to be GC‐biased in different organisms; that is, a W(A or T)→S(G or C) substitution is more likely in this process than a S→W substitution. Here, we show that the insertion/deletion ratio for short noncoding indels that reach fixation between species is positively correlated with the recombination rate in Drosophila melanogaster, Homo sapiens, and Saccharomyces cerevisiae. This correlation is both due to an increase of the fixation rate of insertions and decrease of the fixation rate of deletions in regions of high recombination. Whole‐genome data on indel polymorphism and divergence in D. melanogaster rule out mutation biases and selection as the cause of this trend, pointing to insertion‐biased gene conversion as the most likely explanation. The bias toward insertions is the strongest for single‐nucleotide indels, and decreases with indel length. In regions of high recombination rate this bias leads to an up to ～5‐fold excess of fixed short insertions over deletions, and substantially affects the evolution of DNA segments.     

Differential repair of premutational UV-lesions at tRNA genes in E. coli.

Summary Ultraviolet radiation produces bacterial revertants that frequently are the result of suppressor mutation. When irradiated cells are incubated under conditions unfavorable for protein synthesis there may be a large decrease in the frequency of observed mutants (mutation frequency decline, or MFD). MFD occurs only in excision-proficient strains and is inhibited by inhibitors of pyrimidine dimer excision. It has therefore been interpreted as enhanced excision of some premutational lesions. Potential de novo UAG suppressor mutation is very susceptible to MFD. Potential conversion mutation, the conversion of a UAG to a UAA suppressor, is at least ten times less susceptible to MFD. A base pair transition at a GC target in a particular tRNA gene is suggested for both de novo suppressor mutation and for conversion mutation. We interpret these results as indicating differential repair of premutational UV photoproducts at two closely spaced sites in the same tRNA gene. The significant difference between these two types of mutation may be the orientation of this target base pair in double helical DNA. The C would be in the transcribed strand of DNA when a nucleic acid alteration produces de novo suppressor mutation. The C would be in the nontranscribed strand, two base pairs removed, when a mutagenic alteration produces suppressor conversion. A model involving facilitated incision by hybridization of the transcribed strand of DNA to its cognate tRNA, under conditions promoting MFD, is described to explain this differential repair.     

Can mutation or fixation biases explain the allele frequency distribution of human single nucleotide polymorphisms (SNPs)?

One of the most abiding controversies in evolutionary biology concerns the role of neutral processes in molecular evolution. A main focus of the debate has been the evolution of isochores, the strong and systematic variation of base composition in mammalian genomes. One set of hypotheses argue that regions of similar GC are owing to localised mutational biases coupled with neutral evolution. The alternatives point to either selection or biased gene conversion as mechanisms to preferentially remove A or T bases, favouring G and C instead. Using a novel method, we compare models including such fixation biases to models based on mutation bias alone, under the assumption that non-coding, non-repetitive human DNA is at compositional equilibrium. While failing to fully explain the allele frequency distributions of recent single nucleotide polymorphism data, we show that the data are best fitted if the mutation bias is assumed to be constant across the genome, while fixation bias varies with GC content. We also attempt to estimate the strength of fixation bias, which increases linearly with increasing GC. Our approximation suggests that this force exists within the necessary parameter range: it is not so weak as to be drowned by random drift, but not so strong as to lead to exclusive use of G and C alone. Together these results demonstrate that mutation bias fails to explain the evolution of isochores, and suggest that either selection or biased gene conversion are involved.     

UV-induced hprt mutations in a UV-sensitive hamster cell line from complementation group 3 are biased towards the transcribed strand.

The molecular nature of 254 nm ultraviolet light (UV)-induced mutations at the hypoxanthine-guanine phosphoribosyltransferase (hprt) locus in UV24 Chinese hamster ovary (CHO) cells, which are defective in nucleotide excision repair, was determined. Sequence analysis of 19 hprt mutants showed that single base substitutions (9 mutants) and tandem base changes (7 mutants) dominated the UV mutation spectrum in this cell line. Sixty-five percent of the base substitutions were GC greater than AT transitions, whereas the rest consisted of transitions and transversions at AT base pairs. Analysis of the distribution of dipyrimidine sites over the two DNA strands, where the photoproducts causing these mutations presumably were formed, showed that 12 out of 14 mutations were located in the transcribed strand of the hprt gene. A similar strand distribution of mutagenic photoproducts as in UV24 has previously been found in two other UV-sensitive Chinese hamster cell lines (V-H1 and UV5), indicating that under defective nucleotide excision repair conditions the induction of mutations is strongly biased towards lesions in the transcribed strand of the hprt gene. A plausible explanation for this phenomenon is that during DNA replication large differences exist in the error rate with which DNA polymerase(s) bypass lesions in the templates for the leading and lagging strand, respectively.     

Adaptation or biased gene conversion? Extending the null hypothesis of molecular evolution

The analysis of evolutionary rates is a popular approach to characterizing the effect of natural selection at the molecular level. Sequences contributing to species adaptation are expected to evolve faster than nonfunctional sequences because favourable mutations have a higher fixation probability than neutral ones. Such an accelerated rate of evolution might be due to factors other than natural selection, in particular GC-biased gene conversion. This is true of neutral sequences, but also of constrained sequences, which can be illustrated using the mouse Fxy gene. Several criteria can discriminate between the natural selection and biased gene conversion models. These criteria suggest that the recently reported human accelerated regions are most likely the result of biased gene conversion. We argue that these regions, far from contributing to human adaptation, might represent the Achilles' heel of our genome.     

The rate, not the spectrum, of base pair substitutions changes at a GC-content transition in the human NF1 gene region: implications for the evolution of the mammalian genome structure

The human genome is composed of long stretches of DNA with distinct GC contents, called isochores or GC-content domains. A boundary between two GC-content domains in the human NF1 gene region is also a boundary between domains of early- and late-replicating sequences and of regions with high and low recombination frequencies. The perfect conservation of the GC-content distribution in this region between human and mouse demonstrates that GC-content stabilizing forces must act regionally on a fine scale at this locus. To further elucidate the nature of these forces, we report here on the spectrum of human SNPs and base pair substitutions between human and chimpanzee. The results show that the mutation rate changes exactly at the GC-content transition zone from low values in the GC-poor sequences to high values in GC-rich ones. The GC content of the GC-poor sequences can be explained by a bias in favor of GC > AT mutations, whereas the GC content of the GC-rich segment may result from a fixation bias in favor of AT > GC substitutions. This fixation bias may be explained by direct selection by the GC content or by biased gene conversion.     

Fidelity of replication of the leading and the lagging DNA strands opposite N-methyl-N-nitrosourea-induced DNA damage in human cells.

Semi-conservative replication of double-stranded DNA in eukaryotic cells is an asymmetric process involving leading and lagging strand synthesis and different DNA polymerases. We report a study to analyze the effect of these asymmetries when the replication machinery encounters alkylation-induced DNA adducts. The model system is an EBV-derived shuttle vector which replicates in synchrony with the host human cells and carries as marker gene the bacterial gpt gene. A preferential distribution of N-methyl-N-nitrosourea (MNU)-induced mutations in the non transcribed DNA strand of the shuttle vector pF1-EBV was previously reported. The hypermutated strand was the leading strand. To test whether the different fidelity of DNA polymerases synthesizing the leading and the lagging strands might contribute to MNU-induced mutation distribution the mutagenesis study was repeated on the shuttle vector pTF-EBV which contains the gpt gene in the inverted orientation. We show that the base substitution error rates on an alkylated substrate are similar for the replication of the leading and lagging strands. Moreover, we present evidence that the fidelity of replication opposite O6-methylguanine adducts of both the leading and lagging strands is not affected by the 3' flanking base. The preferential targeting of mutations after replication of alkylated DNA is mainly driven by the base at the 5' side of the G residues.     

The Influence of Local DNA Sequence and DNA Repair Background on the Mutational Specificity of 1-Nitroso-8-Nitropyrene in Escherichia Coli: Inferences for Mutagenic Mechanisms

We have examined the mutational specificity of 1-nitroso-8-nitropyrene (1,8-NONP), an activated metabolite of the carcinogen 1,8-dinitropyrene, in the lacI gene of Escherichia coli strains which differ with respect to nucleotide excision repair (+/- delta uvrB) and MucA/B-mediated error-prone translesion synthesis (+/- pKM101). Several different classes of mutation were recovered, of which frameshifts, base substitutions, and deletions were clearly induced by 1,8-NONP treatment. The high proportion of point mutations (> 92%) which occurred at G.C sites correlates with the percentage of 1,8-NONP-DNA adducts which occur at the C(8) position of guanine. The most prominent frameshift mutations were -(G.C) events, which were induced by 1,8-NONP treatment in all strains, occurred preferentially in runs of guanine residues, and whose frequency increased markedly with the length of the reiterated sequence. Of the base substitution mutations G.C-->T.A transversions were induced to the greatest extent by 1,8-NONP. The distribution of the G.C-->T.A transversions was not influenced by the nature of flanking bases, nor was there a strand preference for these events. The presence of plasmid pKM101 specifically increased the frequency of G.C-->T.A transversions by a factor of 30-60. In contrast, the -(G.C) frameshift mutation frequency was increased only 2-4-fold in strains harboring pKM101 as compared to strains lacking this plasmid. There was, however, a marked influence of pKM101 on the strand specificity of frameshift mutation; a preference was observed for -G events on the transcribed strand. The ability of the bacteria to carry out nucleotide excision repair had a strong effect on the frequency of all classes of mutation but did not significantly influence either the overall distribution of mutational classes or the strand specificity of G.C-->T.A transversions and -(G.C) frameshifts. Deletion mutations were induced in the delta uvr, pKM101 strain. The endpoints of the majority of the deletion mutations were G.C rich and contained regions of considerable homology. The specificity of 1,8-NONP-induced mutation suggests that DNA containing 1,8-NONP adducts can be processed through different mutational pathways depending on the DNA sequence context of the adduct and the DNA repair background of the cell.     

Intracellular Selection, Conversion Bias, and the Expected Substitution Rate of Organelle Genes

A key step in the substitution of a new organelle mutant throughout a population is the generation of germ-line cells homoplasmic for that mutant. Given that each cell typically contains multiple copies of organelles, each of which in turn contains multiple copies of the organelle genome, processes akin to drift and selection in a population are responsible for producing homoplasmic cells. This paper examines the expected substitution rate of new mutants by obtaining the probability that a new mutant is fixed throughout a cell, allowing for arbitrary rates of genome turnover within an organelle and organelle turnover within the cell, as well as (possibly biased) gene conversion and genetic differences in genome and/or organelle replication rates. Analysis is based on a variation of Moran's model for drift in a haploid population. One interesting result is that if the rate of unbiased conversion is sufficiently strong, it creates enough intracellular drift to overcome even strong differences in the replication rates of wild-type and mutant genomes. Thus, organelles with very high conversion rates are more resistant to intracellular selection based on differences in genome replication and/or degradation rates. It is found that the amount of genetic exchange between organelles within the cell greatly influences the probability of fixation. In the absence of exchange, biased gene conversion and/or differences in genome replication rates do not influence the probability of fixation beyond the initial fixation within a single organelle. With exchange, both these processes influence the probability of fixation throughout the entire cell. Generally speaking, exchange between organelles accentuates the effects of directional intracellular forces.(ABSTRACT TRUNCATED AT 250 WORDS)     

Evidence of selection on silent site base composition in mammals: potential implications for the evolution of isochores and junk DNA.

It has been suggested that mutation bias is the major determinant of base composition bias at synonymous, intron, and flanking DNA sites in mammals. Here I test this hypothesis using population genetic data from the major histocompatibility genes of several mammalian species. The results of two tests are inconsistent with the mutation hypothesis in coding, noncoding, CpG-island, and non-CpG-island DNA, but are consistent with selection or biased gene conversion. It is argued that biased gene conversion is unlikely to affect silent site base composition in mammals. The results therefore suggest that selection is acting upon silent site G + C content. This may have broad implications, since silent site base composition reflects large-scale variation in G + C content along mammalian chromosomes. The results therefore suggest that selection may be acting upon the base composition of isochores and large sections of junk DNA.     

Evolution of the isochore structure in the scale of chromosome: insight from the mutation bias and fixation bias

Following the development of reliable methods for inferring the direction of mutations of the single nucleotide polymorphism (SNP), and the revealing of the human isochore map, it has become possible to investigate the evolution of the isochore structure in a continuous region. In this study, the recent evolution of the isochore structure on human chromosome 18, as inferred from the SNP, was examined. A remarkable mutation bias was found, which was destroying the present isochore structure. However, a fixation bias contributed by the biased gene conversion (BGC) effect and a rising fixation probability of derived alleles with increasing GC content was extending the present isochore structure. Combining the two opposing processes, the old isochore structure was declining and a more homogenous isochore structure with higher GC content was being formed on the chromosome. During this process, both the CpG and genic sites, which were present in the isochore but were paid little attention to before, played an important role. In addition, the recombination was confirmed to promote the GC alleles fixed in the genome because of the BGC effect. For the first time, it was observed that with the occurrence of little recombination, AT alleles had the identical fixation probability with GC alleles in the recombination cold spots.     

Context-dependent mutation rates may cause spurious signatures of a fixation bias favoring higher GC-content in humans

Understanding the proximate and ultimate causes underlying the evolution of nucleotide composition in mammalian genomes is of fundamental interest to the study of molecular evolution. Comparative genomics studies have revealed that many more substitutions occur from G and C nucleotides to A and T nucleotides than the reverse, suggesting that mammalian genomes are not at equilibrium for base composition. Analysis of human polymorphism data suggests that mutations that increase GC-content tend to be at much higher frequencies than those that decrease or preserve GC-content when the ancestral allele is inferred via parsimony using the chimpanzee genome. These observations have been interpreted as evidence for a fixation bias in favor of G and C alleles due to either positive natural selection or biased gene conversion. Here, we test the robustness of this interpretation to violations of the parsimony assumption using a data set of 21,488 noncoding single nucleotide polymorphisms (SNPs) discovered by the National Institute of Environmental Health Sciences (NIEHS) SNPs project via direct resequencing of n = 95 individuals. Applying standard nonparametric and parametric population genetic approaches, we replicate the signatures of a fixation bias in favor of G and C alleles when the ancestral base is assumed to be the base found in the chimpanzee outgroup. However, upon taking into account the probability of misidentifying the ancestral state of each SNP using a context-dependent mutation model, the corrected distribution of SNP frequencies for GC-content increasing SNPs are nearly indistinguishable from the patterns observed for other types of mutations, suggesting that the signature of fixation bias is a spurious artifact of the parsimony assumption.     

Integrating genomics,bioinformatics, and classical genetics to study the effects of recombination on genome evolution

This study presents compelling evidence that recombination significantly increases the silent GC content of a genome in a selectively neutral manner, resulting in a highly significant positive correlation between recombination and "GC3s" in the yeast Saccharomyces cerevisiae. Neither selection nor mutation can explain this relationship. A highly significant GC-biased mismatch repair system is documented for the first time in any member of the Kingdom Fungi. Much of the variation in the GC3s within yeast appears to result from GC-biased gene conversion. Evidence suggests that GC-biased mismatch repair exists in numerous organisms spanning six kingdoms. This transkingdom GC mismatch repair bias may have evolved in response to a ubiquitous AT mutational bias. A significant positive correlation between recombination and GC content is found in many of these same organisms, suggesting that the processes influencing the evolution of the yeast genome may be a general phenomenon. Nonrecombining regions of the genome and nonrecombining genomes would not be subject to this type of molecular drive. It is suggested that the low GC content characteristic of many nonrecombining genomes may be the result of three processes (1) a prevailing AT mutational bias, (2) random fixation of the most common types of mutation, and (3) the absence of the GC-biased gene conversion which, in recombining organisms, permits the reversal of the most common types of mutation. A model is proposed to explain the observation that introns, intergenic regions, and pseudogenes typically have lower GC content than the silent sites of corresponding open reading frames. This model is based on the observation that the greater the heterology between two sequences, the less likely it is that recombination will occur between them. According to this "Constraint" hypothesis, the formation and propagation of heteroduplex DNA is expected to occur, on average, more frequently within conserved coding and regulatory regions of the genome. In organisms possessing GC-biased mismatch repair, this would enhance the GC content of these regions through biased gene conversion. These findings have a number of important implications for the way we view genome evolution and suggest a new model for the evolution of sex.     

Strand specificity for UV-induced DNA repair and mutations in the Chinese hamster HPRT gene.

DNA excision repair modulates the mutagenic effect of many genotoxic agents. The recently observed strand specificity for removal of UV-induced cyclobutane dimers from actively transcribed genes in mammalian cells could influence the nature and distribution of mutations in a particular gene. To investigate this, we have analyzed UV-induced DNA repair and mutagenesis in the same gene, i.e. the hypoxanthine phosphoribosyl-transferase (hprt) gene. In 23 hprt mutants from V79 Chinese hamster cells induced by 2 J/m2 UV we found a strong strand bias for mutation induction: assuming that pre-mutagenic lesions occur at dipyrimidine sequences, 85% of the mutations could be attributed to lesions in the nontranscribed strand. Analysis of DNA repair in the hprt gene revealed that more than 90% of the cyclobutane dimers were removed from the transcribed strand within 8 hours after irradiation with 10 J/m2 UV, whereas virtually no dimer removal could be detected from the nontranscribed strand even up to 24 hr after UV. These data present the first proof that strand specific repair of DNA lesions in an expressed mammalian gene is associated with a strand specificity for mutation induction.     

The evolutionary rate of duplicated genes under concerted evolution

The effect of directional selection on the fixation process of a single mutation that spreads in a multigene family by gene conversion is investigated. A simple two-locus model with two alleles, A and a, is first considered in a random-mating diploid population with size N. There are four haplotypes, AA, Aa, aA, and aa, and selection works on the number of alleles A in a diplod (i = 0, 1, 2, 3, 4). Because gene conversion is allowed between the two loci, when the mutation rate is very low, either AA or aa will fix in the population eventually. We consider a situation where a single mutant, A, arises in one locus when a is fixed in both loci. Then, we derive the fixation probability analytically, and the fixation time is investigated by simulations. It is found that gene conversion has an effect to increase the "effective" population size, so that weak selection works more efficiently in a multigene family. With these results, we discuss the effect of gene conversion on the rate of molecular evolution in a multigene family undergoing concerted evolution. We also argue about the applicability of the theoretical results to models of multigene families with more than two loci.