How does the molecular machinery choose where to cut a chromosome for recombination?

Question

I'm wondering about a few technicalities of crossover in meiosis. The point of crossover is to create new chromosomes that don't have the same allele combinations as the original two chromosomes. Usually, the chromosomes are cut at the same place on both chromosomes, and each piece is then stitched to that place on the other. This is to avoid unequal recombination, a scenario in which one chromosome has several instances of a gene and the other no instance at all. I'm wondering how the molecular machinery knows where to cut.

So here's my question: How does the molecular machinery choose where to cut a chromosome for recombination?

This question has two parts: At what type of place does it occur (does the machinery choose a completely random place, regardless of where genes start and end, does it just cut at the start of genes, or does it do something else)? Given that it happens at this type of place (e.g. start of a gene), how does it decide that it will cut here (the start of this gene) and not there (the start of that gene)?

Answer 1

In humans and mice anyway ,a lot of it boils down to the recognition of a specific sequence that marks recombination hotspots by PRDM9. http://www.sciencemag.org/content/327/5967/836

Edit - I'm expanding in response to the comment below...

Meiotic recombination occurs at vastly greater frequencies in some locations in the genome than others and these are called Recombination Hotspots. For example, the figure here, sourced from http://www.sciencemag.org/content/327/5967/876/F1.large.jpg

shows the recombination rate at the hotspot and further away from it for Chimps and Humans.

These hotspots are recognised by the cutting machinery thanks to the binding of PRDM9, a zinc finger protein, to a DNA sequence that it specifically recognises and is present at hotspots. The DNA sequence varies from species to species (as does the sequence and the function of PRDM9) but in humans is a well-characterised motif 13 base pairs long and of the sequence CCNCCNTNNCCNC (where N is any one of the 4 bases in DNA) , and accounts for the activity of nearly 40% of known hotspots.

The first link I posted has evidence to suggest that variation in composition of PRDM9 is a determinant of which hotspots get used. PRDM9, it appears, stimulates the formation of a specific histone modification - H3k4me3 (DNA is wrapped around histones, there are five histones - H1, H2A, H2B, H3, H4, and each of them has a tail that can be chemically modified, and eight of these (2 of H2A, 2 of H2B, 2 of H3 and 2 of H4) form a nucleosome - in this case the fourth lysine residue of the tail of Histone H3 is trimethylated as regulated by PRDM9, and facilitates crossing over and recombination initiation. Edit Summary

Answer 2

The question is very broad and complicated, since the situation may differ in prokaryotes and eukaryotes. Nevertheless, I'm citing a good paper that is closely related to your question:

Studies in yeast show that initiation of recombination, which occurs by the formation of DNA double-strand breaks, determines the distribution of gene conversion and crossover events that take place in nearby intervals. Recent data in humans and mice also indicate the presence of highly localized initiation sites that promote crossovers clustered around the region of initiation and seem to share common features with sites in yeast. On a larger scale, chromosomal domains with various recombination rates have been identified from yeast to mammals. This indicates a higher level of regulation of recombination in the genome with potential consequences on genome structure... ...DSBs (Double Strand Breaks) occur in highly localized regions and spread over 70–250 bp. DNA sequence analysis reveals no unique conserved consensus sequences, although a degenerate 50-bp motif partly correlates with DSB sites. However, one common feature is that DSBs are located in accessible regions of the chromatin next to either promoters or binding sites for transcription factors. Based on two studies, DSB activity does not correlate with local transcriptional activity, but depends on transcription-factor binding (HIS4 in S. cerevisiae and ade6-M26 in S. pombe).

Bernard de Massy, Distribution of meiotic recombination sites. TRENDS in Genetics Vol.19 No.9 September 2003

I'll sum that up into a more answer-like form. Seems like the process is not random, because the double strand breaking events are clearly not evenly distributed across a genome. As the paper says there are no specific consensus motifs found as of yet, though there clearly is something special before promoters and TF binding sites, which makes them more likely to be a breaking site. How the machinery choses the place? Once again, as the paper says the breaking event depends on TF binding. But that is for S. cerevisiae. There are 17 hot spots found in human and mice genomes some of which are intergenic (they occupy introns or 5'/ 3' flanking regions).

Here is the distribution of recombination frequencies across one chromosome (the figure is taken from the paper).

Here is a list of recombination sites in humans and mice

Answer 3

In meiosis, homologous recombination occurs. Two Spo11 proteins utilize tyrosines to induce a double-stranded break in the DNA. Spo11 has no specific cleavage site. However, cleavage by Spo11 lead to the discovery of Spo11-Oligonucleotide (Spo11 w/ attached oligonucleotide post-cleavage) complexes which could be mapped to 'hotspots' which the linked study finds a number of discriminating factors to Spo11 cleavage in meiosis:

From the precise locations of the 2.2 million Spo11-oligonucleotide sequences, Pan et al. [1] showed that local DNA composition also influences Spo11 cleavage sites. As expected from previous studies, Spo11 does not have a specific recognition or cleavage site. However, sequence biases were detected: the 10 to 12 bp surrounding the cleavage site and predicted to be bound directly by Spo11 is relatively AT rich, predicting relatively narrow and deep helix grooves facing the bound Spo11 dimer. Cleavage favors sites immediately 3' of a C and disfavors G in the same position. Also, within a 32-bp core surrounding the Spo11 cleavage sites, a twofold rotational symmetry for complementary dinucleotide composition can be discerned, suggesting separate contributions of the flanking 'half sites' to Spo11 binding and/or cleavage. In addition, cleavage sites are negatively correlated with positioned nucleosomes. Similarly, Spo11 is generally occluded from cleaving sites where transcription factors are bound, even though the binding sites of several different transcription factors positively correlate with hotspot sites.

Answer 4

While all the answers above are correct and insightful for the process of DSBs formation, recombination has another layer of complexity.

Mediating proper chromosomal segregation is more likely the real point of crossovers since crossovers still happen in inbred genomes, which do not result in novel haplotypes. CO create a heteroduplex structure between DNA strands, Holliday junctions. This structure creates tensions forces which are needed to pass the spindle assembly checkpoint before the meiotic cell can progress into anaphase (see papers by Nicklas work on grasshopper chromosomes and Hirose et al 2011). In most* organisms, at least one CO is needed per chromosome arm to pass the spindle assembly checkpoint and mediate proper disjunction. Lacking COs or misplaced COs can result in non-disjunction and increased risk for aneuploidy (Hassold, Hall, Hunt 2007). (*an exception are male Drosophila melanogaster, which have a CO independant chromosome segregation pathway).

Additionally it should be noted that DSB =/= COs that is, not all DSBs are converted into COs. In all organisms, DSB outnumber the resulting COs. The majority of DSBs are resolved into non-crossovers (NCOs). There is new evidence (in mice at least) that these numbers are not proportional. That is, if total DSBs in the genome are reduced, the total CO number is unaffected. (Cole et al 2012).

I think a more interesting phrasing of your question would be, How does the genome choose which DSBs to resolve into COs vs NCO?