Topoisomerase IIIalpha and Bloom's helicase can resolve a mobile double Holliday junction substrate through convergent branch migration

Abstract

It has long been suspected that a double Holliday junction (dHJ) could be resolved by a topoisomerase partnered with a helicase by convergent branch migration of the HJs. Genetic analysis of yeast TOP3 and SGS1 has lent considerable evidence to the notion that the protein products of these genes are involved in just such a process, although biochemical analysis of the metabolism of a dHJ has been hindered by the lack of a substrate that adequately replicates the endogenous structure. We have synthesized a dHJ substrate that recapitulates many of the features of an endogenous dHJ and represents a much earlier intermediate in the resolution pathway. Here, we show that Drosophila topoisomerase IIIalpha (Topo IIIalpha) and Blm (a homolog of Sgs1) are capable of resolving this substrate to non-cross-over products and that this activity is stimulated by replication protein A (RPA). We investigated the ability of other Drosophila topoisomerases to perform this reaction in concert with Blm and RPA and discovered that this resolution activity is unique to Topo IIIalpha. Examination of the mechanism of resolution reveals that Topo IIIalpha, Blm, and RPA resolve this substrate by convergent migration of the two HJs toward each other, collapsing the dHJ. This mechanism stands in contrast to classic resolvase activities that use a structure-specific endonuclease to cleave the HJs.

Fig. 1.

Pathways of resolution of the DHJS. In the upper pathway, the linking number between circles A and B is continuously reduced by the convergent branch migration of the HJs by a topoisomerase/helicase complex until the two circles are completely unlinked to yield NCO products. In the lower pathway, a structure-specific endonuclease can cut the two HJs, with the orientation of the two cuts relative to each other determining whether CO or NCO products will be produced. The arrows indicate the location of BamHI restriction sites on the molecules. Digestion of either the CO or NCO products by BamHI will result in two linear molecules that are 416 and 465 bp in length.

Fig. 2.

Topo IIIα and Blm can resolve the DHJS. There was no detectable activity on the substrate with either Topo IIIα alone (lane 2) or Blm alone (lane 3). When both proteins are added to the reaction, resolved DNA circles can be detected (lane 4). This activity depends on the activity of the Topo IIIα, because Topo IIIα with a mutated active-site tyrosine (mut) cannot support resolution (lane 5); it also depends on the helicase activity of Blm, because omission of ATP from the reaction buffer also abolishes activity (lane 6). Markers A and B are for the resolved products (lanes 7 and 8).

Fig. 3.

The activity of Topo IIIα and Blm is specifically stimulated by RPA. The amount of resolution of the DHJS increases with increasing concentrations of hRPA in the reaction (compare lane 2 with lanes 3–5), whereas hRPA alone has no activity on the substrate (lane 6). This stimulation was specific to RPA; including other ssDNA-binding proteins, E. coli SSB (lanes 7–9) or T4 gp32 (lanes 11–13), showed no stimulation of the resolution of DHJS by Topo IIIα and Blm. The supplemented proteins are at 44, 88, and 175 nM concentrations, respectively.

Fig. 4.

Of all of the Drosophila topoisomerases, only Topo IIIα supports resolution with Blm and RPA. Blm and RPA were incubated with 36 nM Topo I (lane 2), Topo II (lane 3), Topo IIIα (lane 4), or Topo IIIβ (lane 5) in resolution reactions. Only Topo IIIα produced products that comigrated with markers A and B (lanes 6 and 7, respectively).

Fig. 5.

Psoralen cross-linking inhibits DHJS resolution by Topo IIIα/Blm/RPA. After the DHJS, with or without psoralen cross-links, was reacted in the presence of the indicated proteins, the products of the reactions were digested with BamHI. Lane 1 contains a DNA size marker, with the size of the bands denoted to the left of the gel (in base pairs). There is no resolution of the substrate in the mock reactions (lanes 2 and 3). T7 endo I can resolve the cross-linked DHJS with little inhibition of activity (lanes 4 and 5), whereas resolution by Topo IIIα/Blm/RPA is almost completely inhibited by the psoralen cross-links (lanes 6 and 7). Lanes 8 and 9 contain BamHI-digested markers A and B, respectively.

Fig. 6.

DHJS resolution by Topo IIIα, Blm, and RPA occurs through migration of the HJs rather than by a resolvase mechanism. (A) A DHJS containing mismatches allows differentiation between resolvase and migration mechanisms of resolution. We created a DHJS by replacing the NheI restriction sites in the “B” vectors with a SphI site, pDHJS BS^+/−. The DHJS that we made with these vectors, MM-DHJS, contains a pair of two nucleotide mismatches, which render the substrate refractory to digestion by NheI and SphI. If the substrate is resolved by a classic resolvase activity (left pathway), the mismatches will be retained in the products. If the substrate is resolved through migration of the HJs (right pathway), then the mismatches will be re-paired with their complementary sequence, and the products of the reaction will now be sensitive to digestion by NheI and SphI. (B) The products of the Topo IIIα/Blm/RPA reactions are generated by the migration of the dHJ. MM-DHJS is refractory to digestion by NheI (lane 2) or SphI (lane 3). After incubation with Topo IIIα/Blm/RPA, resolution products are detected (lane 4), and these products are sensitive to NheI (lane 5) and SphI (lane 6) digestion. In contrast, resolution of MM-DHJS by T7 endo I (lane 7) yields products that are largely refractory to digestion by NheI (lane 8) and SphI (lane 9). Lanes 10–13 contain the digested marker molecules, and lane 14 contains a DNA size marker with the size of the bands denoted to the right of the gel in base pairs.

Fig. 7.

A model for Topo IIIα/Blm/RPA-mediated resolution of a dHJ. In this illustration, the ends of the DNA are attached to a hypothetically immobile structure to simulate topological constraints imposed on an endogenous dHJ. (A) If the HJs in a dHJ were to undergo convergent migration in the absence of a topoisomerase, negative writhe would be induced behind the migrating HJ, and positive writhe would accumulate in front of the HJ. (B) In the presence of Topo IIIα/Blm/RPA (represented as a green complex), the HJs undergo convergent migration until the linking number between the two DNAs is reduced to zero. Through established protein–protein interactions, it seems likely that these proteins function as a complex, and within this complex, Topo IIIα would be able to perform the strand passage events to unlink the two strands. The association between Topo IIIα and Blm could provide the directionality for the Topo IIIα strand passage events, although it is unclear at this time how directionality for Blm is established.