Type IB eukaryotic DNA topoisomerases modulate the topological state of DNA by cleaving and rejoining 1 strand of the DNA duplex. Cleavage is a transesterification reaction in which the scissile NpN phosphodiester is attacked by a tyrosine of the enzyme, resulting in the formation of a DNA-(3’-phosphotyrosyl)-enzyme intermediate and the expulsion of a 5’-OH polynucleotide. During strand rejoining, the 5’-OH end attacks the covalent intermediate to form a phosphodiester bond and expel the tyrosine. There is no involvement of a divalent cation in transesterification chemistry. The lambda Int family of site-specific DNA recombinases (referred to now as the tyrosine recombinases) employs an identical mechanism to catalyze the formation and resolution of Holliday junctions.
Insights into the common catalytic mechanism of the type IB topoisomerases and tyrosine recombinases have emerged from the crystal structures of Cre recombinase and human topoisomerase I bound to DNA, the unliganded structures of XerD, lambda Int, HP1 Int, and vaccinia topoisomerase, and the extensive mutational analysis of vaccinia topoisomerase. The catalytic domain of each of these proteins adopts a common fold composed of 8 alpha helices and a 3-strand antiparallel beta sheet.
Crystal Structure of the Topoisomerase Catalytic Domain
Overall fold of the topoisomerase catalytic domain
We crystallized the catalytic domain of vaccinia topoisomerase and, in collaboration with our SKI colleague Nikola Pavletich, determined its structure at 2.3 Å resolution. The catalytic domain is an oblong globular protein consisting of 10 alpha helices and a 3-strand antiparallel beta-sheet on one surface. An N-terminal lobe (colored purple) comprises helices 1 to 5 and the beta-sheet; a C-terminal lobe (colored cyan) consists of helices 6 to 10. The long N-terminal alpha helix is integrated into both lobes. There was no interpretable electron density from residues 129 to 137. This segment, which is depicted as a dashed purple line connecting helices 2 and 3, includes residues that are protease-sensitive in the free topoisomerase, but protected upon DNA binding. Thus, the proteolysis and crystallographic findings together indicate that this segment is flexible in the free enzyme.
Two sulfate ions (colored red) were bound to the enzyme, which had been crystallized in ammonium sulfate. The distance between the sulfate densities was equal to the distance between vicinal backbone phosphates in 1 strand of duplex DNA. One sulfate is within hydrogen bonding distance of essential residue His-265, and essential residue Arg-223 is close by. We posit that this sulfate is located at a site that would be occupied by the scissile phosphate of the DNA substrate. Remarkably, the active site nucleophile Tyr-274 (located in helix 8) is actually oriented away from the putative binding pocket for the scissile phosphate and therefore is not poised to engage in catalysis. This implies that Tyr-274 must be reoriented upon binding of the topoisomerase to the DNA target site.
A Model of the Topoisomerase-DNA Complex
Topo I forms a C-shaped clamp around DNA.
We built a model of the noncovalent topoisomerase-DNA complex by fitting the structure of the catalytic domain onto a B-form duplex, such that the sulfates are superimposed on the scissile phosphodiester and 5’-proximal phosphodiester, respectively, of the CCCTpTpA target sequence. Placement of the N-terminal domain in the major groove as dictated by the observed UV crosslinking contacts easily connected the 2 domains. The protein spans an ~18 bp DNA segment (from +10 to -9 relative to the scissile phosphodiester), which is consistent with the dimensions of the DNase I and exo III footprints. The model suggests that topoisomerase engages the DNA circumferentially, with the protein in the form of a C-shaped clamp, essentially as predicted by our footprinting studies of the topoisomerase-DNA interface. It implies conformation flexibility in the free topoisomerase, which must open up to accommodate DNA within the cleft of protein. This is most likely achieved through domain motion about the interdomain bridge segment. Also, the model places essential catalytic residues Arg-223 and His-265 in a position to interact directly with the scissile phosphodiester.
Catalytic Mechanism of Topoisomerase I
A constellation of 4 conserved amino acid side chains (e.g., Arg130, Lys167, Arg223, and His265 in the vaccinia topoisomerase) provides most of the catalytic power for the estimated 12 order-of-magnitude rate enhancement of transesterification by topoisomerase IB. The 2 arginines and the histidine interact directly with the nonbridging scissile phosphate oxygens in the DNA cocrystal structures of recombinase and human topoisomerase I. Mutational and structural data together suggest that such interactions activate the scissile phosphodiester for attack by the 5’-OH DNA or tyrosine nucleophiles; and that they enhance catalysis by stabilizing the developing negative charge on the presumptive pentacoordinate phosphorane transition state. The stereochemical outcome of the net cleavage-rejoining reaction of vaccinia topoisomerase on phosphorothioate-modified DNA is a retention of configuration at the scissile phosphodiester, as expected for 2 serial in-line SN2-type displacements.
A mechanism of general acid-base catalysis has also been proposed; whereby a general base on the enzyme (-B:) accepts a proton from the attacking tyrosine nucleophile during covalent adduct formation, and a separate enzymic general acid (-AH) donates a proton to expel the 5’-OH leaving group. In the religation step, -BH would donate a proton to the leaving tyrosine and -A: would accept a proton from the attacking 5’-OH. The amino acid side chain corresponding to -B: has not yet been identified, although many candidate residues, including all histidines, have been ruled out by mutational analysis of vaccinia topoisomerase. The available crystal structures of vaccinia and human type IB topoisomerases provide no clues to the identity of potential base catalysts that can be reconciled with mutational data; i.e., there is no essential residue in the immediate vicinity of the tyrosine nucleophile that could serve as a proton acceptor and no essential residue in direct contact with the 5’-bridging oxygen of the DNA leaving strand that could serve as the proton donor.
This situation raises the prospect that either: (i) transesterification by topoisomerase IB and tyrosine recombinases does not entail general acid-base catalysis, or (ii) the side chains on the enzyme that serve as general acid or general base are only transiently in contact with their atomic targets and their function is coordinated with conformational changes of the enzyme bound to DNA. The delineation of all relevant catalytic moieties and the conformational changes that drive transesterification is the principal challenge in the topoisomerase/tyrosine recombinase field.
Lys167 is the General Acid Catalyst of DNA Cleavage
The “general acid loop” of vaccinia topoisomerase is colored red and denoted by the arrow.
We used 5’-bridging phosphorothiolate-modified DNAs to identify Lys167 as the long-sought but elusive general acid catalyst of DNA cleavage. The lower pKa of the 5’-S leaving group versus 5’-O restored activity to the catalytically defective K167A mutant; whereas there was no positive 5’-bridging thio effect for alanine mutants of catalytic residues Arg223 and His265 that are postulated to stabilize the transition state. Lys167 is located in the loop connecting the second and third antiparallel beta strands of the catalytic domain of vaccinia topoisomerase. Lysines occupy equivalent positions in the structures of human topoisomerase I and the tyrosine recombinases. The hairpin loop, in which the lysine general acid is situated, displays considerable conformational flexibility in the various tyrosine recombinase and topoisomerase IB structures. The lysine shifts into the minor groove upon binding to the DNA target site, where it is poised to interact transiently with the 5’-O leaving group.