Extended Data Fig. 5: Analysis of the SPO11 dimer interface and TOP6BL GHKL domain.
From: Reconstitution of SPO11-dependent double-strand break formation
a,b, SPO11 dimer interface. Panel a shows distances between conserved interfacial residues; panel b compares the SPO11 and Top6A dimer interfaces (M. mazei Topo VI; pdb: 2q2e)14. c, Agreement between AlphaFold3 model and crystal structure (pdb: 1z5a)32 for an ATP-mediated dimer of the S. shibatae Top6B GHKL domain. d, Failure of AlphaFold3 to predict correct ATP-dependent dimerization of the TOP6BL GHKL domain. The dimer interface is in the wrong location and the ATPs are not at the correct surface. e, Comparison of Top6B and TOP6BL GHKL domain topologies. β strands (arrows) and α helices (cylinders) are shown, along with conserved ATP-interacting elements highlighted in panel f. f, Structure-based sequence alignment between S. shibatae Top6B and mouse TOP6BL GHKL domains. A previous alignment suggested that TOP6BL has degenerate versions of all three G boxes9, but AlphaFold3 suggests that the G1 box is missing instead, with degenerate versions of the other two boxes plus the previously shown absence of the ATP lid61. g, Hypothetical conformation change between pre-DSB and post-DSB SPO11 complexes. A post-cleavage state was modeled by aligning the WH (hot pink) and Toprim (purple) domains separately to the orthologous domains in the yeast Spo11 cryo-EM structure24. The AlphaFold3 model (presumptive pre-cleavage state) is shown in gray. h, EMSA of E224A mutant SPO11 complexes binding to a 5′-labeled 25-bp hairpin substrate with a two-nucleotide 5′ overhang end. A representative gel is shown at left, quantification is at right (mean ± s.d. of n = 3 experiments; apparent Kd given as mean ± s.e.). Retention of DNA binding is unlike the equivalent yeast Spo11-E233A25. i, Supercoil relaxation by wild-type protein. Wild-type SPO11–TOP6BL complexes were incubated with supercoiled pUC19 and 20 mM CaCl2 then deproteinized and separated on agarose gels without (top) or with (bottom) 1 µg/ml ethidium bromide (EtBr). Red line, topoisomer ladder that disappears on EtBr-containing gels. Red asterisk, plasmids that were relaxed covalently closed circles (CCC) after the reaction and became positively supercoiled upon binding EtBr. Performed twice with similar results. j, Model for the mechanism of supercoil relaxation. The cartoon (not to scale) shows the arrangement of the WH and Toprim domains of each SPO11 monomer on nicked DNA, with the 5′ end covalently attached to Y138 of the darker pink monomer on the left, and the 3′ OH bound by the Toprim domain of the lighter colored monomer on the right. The red arrow signifies rotation of the right-hand SPO11 and the DNA arm it is bound to, relative to the left-hand monomer and its bound DNA arm. This rotation, which necessitates disruption of the SPO11 dimer interface, would allow the DNA to swivel around the uncut strand, thereby relaxing supercoils. See Supplementary Discussion 3 for more detail.
