Synapsis — the process by which paired homologous chromosomes are brought into close alignment — is a general feature of meiosis. But whereas Saccharomyces cerevisiae requires a protein called Spo11p to initiate synapsis, knockout studies in Caenorhabditis elegans and Drosophila melanogaster have shown that these organisms do not. Because Spo11p and its homologues generate double-stranded DNA breaks (DSBs), which initiate meiotic recombination, these experiments have led to a model for synapsis — that, in worms and flies at least, it occurs in specialized 'pairing centres', without the need for meiotic recombination.

Could a dependence on pairing centres have emerged with increased genome complexity? And, if so, might mammals rely on such centres? Two papers in Molecular Cell report that, surprisingly, they don't.

Both groups have knocked out the mouse Spo11 gene, and they find that both male and female mice have severe gonadal abnormalities from defective meiosis. There is little or no synapsis between chromosomes and, where it does occur, it's mostly between non-homologous partners. Moreover, two 'markers' for meiotic recombination — Rad51 and Dmc1, which load onto single-stranded DNA ends at the sites of DSBs — do not form characteristic foci in the Spo11−/− mice. In other words, synapsis in mice does seem to depend on meiotic recombination.

Romanienko and Camerini-Otero also propose a second function for Spo11 in mice. They localized the Spo11 protein in meiotic chromosome spreads, and saw that it decorated the lengths of synapsed homologues during pachytene. This was unexpected — Spo11 is thought to catalyse the formation of DSBs at an earlier stage, leptotene. But, because pachytene is the stage at which homologous pairs become fully synapsed, the authors suggest that Spo11 could also have a structural function in stabilizing synapsis.