Microbiology

Accelerated laboratory evolution reveals the influence of replication on the GC skew in Escherichia coli

Kono et al., Genome Biology and Evolution, 10(11):3110-3117, 2018

Most bacterial genomes display contrasting strand asymmetry in a variety of features, such as nucleotide composition and gene orientation, of the two replichores separated by the replication origin and terminus. The cause for the polarization is often attributed to mutations arising from the asymmetric replication machinery. Notably, a base compositional bias known as a GC skew is focused on as a footprint of the bacterial genome evolution driven by DNA replication. Previously, although a replication driven mutation pattern responsible for the GC skew formation or the related mathematical models have been well reported, an exact impact of the replication-related elements on the genomic structure is yet actively debated, and not confirmed experimentally. However, the GC skew formation is very time consuming and challenging in the laboratory. We, therefore, used cytosine deaminase as a DNA mutator, and by monitoring the mutations during an accelerated laboratory evolution procedure with Illumina sequencing, we enabled the trial and error of the GC skew formation in high resolution. Using this technology, we succeeded in reconfirming the influence of bacterial replication machinery on the genomic structure at high resolution.

Rif1 binds to G quadruplexes and suppresses replication over long distances

Kanoh et al., Nature Structural & Molecular Biology, 22(11), 889-897, 2015

Rif1 regulates replication timing and repair of double-strand DNA breaks. Using a chromatin immunoprecipitation–sequencing method, we identified 35 high-affinity Rif1-binding sites in fission yeast chromosomes. Binding sites tended to be located near dormant origins and to contain at least two copies of a conserved motif, CNWWGTGGGGG. Base substitution within these motifs resulted in complete loss of Rif1 binding and in activation of late-firing or dormant origins located up to 50 kb away. We show that Rif1-binding sites adopt G quadruplex–like structures in vitro, in a manner dependent on the conserved sequence and on other G tracts, and that purified Rif1 preferentially binds to this structure. These results suggest that Rif1 recognizes and binds G quadruplex–like structures at selected intergenic regions, thus generating local chromatin structures that may exert long-range suppressive effects on origin firing.

Undesigned selection for replication termination of bacterial chromosomes

Kono et al., Journal of Molecular Biology, 426(16), 2918-2927, 2014

The oriC DNA replication origin in bacterial chromosomes, the location of which appears to be physically identified, is genetically regulated by relevant molecular machinery. In contrast, the location of the terminus remains obscure for many bacterial replicons, except for terC, the proposed and well-studied chromosome termination site in certain bacteria. The terC locus, which is composed of specific sequences for its binding protein, is located at a site opposite from oriC, exhibiting a symmetric structure around the oriC–terC axis. Here, we investigated Bacillus subtilis 168 strains whose axes were hindered and found that the native terC function was robust. However, eradication of terminus region specific binding protein resulted in the natural terC sites not being used for termination; instead, new termini were selected at a location exactly opposite to oriC. We concluded that replication generally terminates at the loci where the two approaching replisomes meet. This site was automatically selected, and two replisomes moving at the same rate supported symmetrical chromosome structures relative to oriC. The rule, which was even validated by artificial chromosomes irrespective of oriC, should be general for replicons administered by two replisomes.

Validation of Bacterial Replication Termination Models Using Simulation of Genomic Mutations

Kono et al., PLoS One, 7(4), e34526, 2012

In bacterial circular chromosomes and most plasmids, the replication is known to be terminated when either of the following occurs: the forks progressing in opposite directions meet at the distal end of the chromosome or the replication forks become trapped by Tus proteins bound to Ter sites. Most bacterial genomes have various polarities in their genomic structures. The most notable feature is polar genomic compositional asymmetry of the bases G and C in the leading and lagging strands, called GC skew. This asymmetry is caused by replication-associated mutation bias, and this “footprint" of the replication machinery suggests that, in contrast to the two known mechanisms, replication termination occurs near the chromosome dimer resolution site dif. To understand this difference between the known replication machinery and genomic compositional bias, we undertook a simulation study of genomic mutations, and we report here how different replication termination models contribute to the generation of replication-related genomic compositional asymmetry. Contrary to naive expectations, our results show that a single finite termination site at dif or at the GC skew shift point is not sufficient to reconstruct the genomic compositional bias as observed in published sequences. The results also show that the known replication mechanisms are sufficient to explain the position of the GC skew shift point.