E. Coli DNA Topoisomerase I
67K N-terminal fragment

David Kysela and David Marcey

Contents: I. Introduction

II. Structural Features

III. DNA Cleaving

IV. Online Links

V. References

I. Introduction

DNA topoisomerases are a class of enzymes involved in the regulation of DNA supercoiling. Type I topoisomerases change the degree of supercoiling of DNA by causing single-strand breaks and re-ligation, whereas type II topoisomerases (such as bacterial gyrase) cause double-strand breaks. The different roles of DNA topoisomerase I and II may indicate an opposing pair of roles in the regulation of DNA supercoiling. Both activities are especially crucial during DNA transcription and replication, when the DNA helix must be unwound to allow proper function of large enzymatic machinery, and topoisomerases have indeed been shown to maintain both transcription and replication.

Recently, new topoisomerases of both types I and II (classified as topoisomerase III and IV) have been discovered. These topoisomerases may indicate even more roles for topoisomerases, with some topoisomerase III enzymes implicated in regulation of recombination events, and topoisomerase IV implicated in the process of segregating newly replicated chromosomes.

The focus of this tutorial is E. coli DNA topoisomerase I, responsible primarily for the relaxation of negative supercoils. Topoisomerase I has also been implicated in knotting and unknotting DNA and in linking complementary rings of single-stranded DNA into double-stranded rings. The intact holoenzyme is a 97K protein with three Zn(II) atoms in tertacysteine motifs near its carboxy-terminus. Topoisomerase I appears to reverse supercoiling by transiently breaking a segment of single-stranded DNA, passing an intact single- or double-stranded strand of DNA through the gate, then rejoining the broken segment. To understand the function of this mechanism, a supercoil may be thought of as a knot on a string where the string is fixed at both ends, and so the knot may not simply be untied. Instead, the gating mechanism appears to open, untie, and reseal the "knotted" DNA precisely where it lies.

The 67K N-terminal fragment that is the subject of this tutorial is capable of this gating/passage activity, but not of directly relaxing negative supercoils. Other domains include a zinc-binding domain implicated in the process of relaxing DNA supercoils and a 14K C-terminal DNA-binding domain.

Topoisomerase I has several unusual features. Unlike type II topoisomerases, topoisomerase I does not require ATP hydrolysis to catalyze the complex topological rearrangements of DNA for which it is responsible. Whereas most enzymes involved in complex rearrangements of DNA are oligomeric, topoisomerase I appears to be a fully functional monomer.

II. Structural Features

The 67K N-terminal fragment of topoisomerase I is a single polypeptide with extensive secondary structure <>. There is a 27Å hole in the center of the protein large enough to comfortably encircle either a single- or double-stranded piece of DNA with no steric hindrance between the DNA sugar-phosphate backbone and protein side chains within the torus. This hole is lined with basic (positively charged) residues <> that provide strong electrostatic attraction to the negatively charged phosphate oxygens of the DNA backbone. Thus, both the size and electrostatic character of the torus make it a likely candidate for temporarily holding a DNA segment.

The fragment consists of four MAJOR DOMAINS. Domain I is made up of a beta sheet with four parallel beta strands and four alpha helices. This domain resembles the Rossman fold known to bind nucleotides in many other proteins, and so it is a likely candidate for binding a segment of DNA. Domain II consists of two beta sheets, each with three antiparallel strands, and one alpha helix. Domain III consists of five alpha helices, and this domain also contains the active site, Tyr 319 (see below). Domain IV has eight alpha helices and apparently provides structural support for the protein.

There are extensive interfaces between domain III and domain I and between domain IIIand domain IV. Seven hydrogen bonds <> exist between residues of domain III and domain IV. Eight hydrogen bonds <> and two salt bridges <> are present between domain III and domain I. Domains II and III appear to be linked as a rigid body <>, maintaining structure even after domains I and IV are cleaved off.

The non-covalent interface between domains I, III, and IV and the rigid domain II and III substructure indicate that domains II and III may be able to swing away from domains I and IV as if on a hinge. The enzyme appears, then, two have two conformations: one closed, with non-covalent bonding along the interface between domain III and domains I and IV; and a second open conformation in which the aforementioned non-covalent bonds are broken and domains II and III swing away from the rest of the protein as a single body.

A residue at the catalytic site, Tyr 319, is located in domain III at the interface between domains I, III, and IV. Several residues comprise the complete active site, interacting through at least five hydrogen bonds <> and one salt bridge <> . Many of these residues are also involved in maintaining the non-covalent interface between domains I, III, and IV <> .

III. DNA Cleaving

As mentioned earlier, the manipulation of DNA by topoisomerase involves transiently gating a segment of single-stranded DNA and passing either single- or double-stranded DNA through the gate. This section of the tutorial explores a proposed strand passage mechanism. The images presented in this section are models generated by manipulation of existing atomic coordinate files of crystallographic data. These models were produced with Hyperchem molecular modelling software, and are based on published models for the strand passing mechanism (see references).

One single-stranded and one double-stranded segment of DNA, both B-form, are external to the protein. The single-stranded segment serves as the gating element and the double-stranded segment as the passing element. A key catalytic residue, Tyr 319, is shown.

A transesterfication occurs between Tyr 319 and a phospho-diester bond in the backbone of the single-stranded DNA segment. The 3' oxygen in the backbone is displaced by a tyrosine oxygen, covalently linking the tyrosine oxygen to phosphorus at the 5' backbone nick. Thus, the 5' nick has been covalently linked to topoisomerase I in domain III.

This covalent linkage causes a structural deformation in the protein which breaks the non-covalent interface between domain III and domains I and IV <> (recall that many of the active site residues are also implicated in this interface), and the rigid body formed by domains II and III <> swings away from the remainder of the molecule, forming the "open" conformation of the enzyme. The 3' nick remains non-covalently associated with domains I and IV (recall the DNA-binding character of domain I), while the 5' nick swings up with domains II and III, thus creating a gate in the single-stranded element. (For another online explanation of the transesterfication mechanism, visit Bear's topoisomerase site.)

The passing element travels through the transient gate and into the torus.

Domains II and III swing back in as the enzyme adopts a closed conformation, but one which is slightly gapped along the interface between domains I, III, and IV in order to accommodate the single-stranded DNA fragment. The single-stranded DNA segment now rejoins itself through nucleophilic attack of the 5' nick phosphorus by the 3' nick OH group.

The enzyme again adopts its open conformation. The single-stranded DNA segment remains non-covalently bound to domains I and IV.

The passing DNA element is released from the torus, completing the strand passage event. The single-stranded DNA element may remain non-covalently associated with the enzyme for repetition of the mechanism.

Since topoisomerase operates in the absence of ATP, this process must be driven solely by interactions within and between the enzyme and DNA. There is no defined directionality of this mechanism, so the overall process must involve the release of free energy. It seems likely that the supercoiled and deformed DNA conformations which topoisomerase manipulates are highly energetically unfavorable, and thus the lower energy state of the manipulated conformation may drive the mechanism.

Although the model presented above involves a double-stranded passing element, it is important to recall that either single- or double-stranded DNA may serve as the passing element. Gating, however, has only been observed with single-stranded DNA.

The 14K C-terminal fragment of E. coli topoisomerase I enhances DNA binding activity and improves relaxation of negative supercoiling by the enzyme.

IV. Links to Other Topoisomerase Sites

Soaring Bear's Topoisomerase Site at University of Arizona's College of Pharmacy

V. References

Lima, C.D., J.C. Wang, and A. Mondragon. 1994. Three-dimensional structure of the 67K N-terminal fragment of E. coli DNA topoisomerase I. Nature 367:138-145.

Wang, J.C. 1991. DNA topoisomerases: why so many? Journal of Biological Chemistry 266:6659-6662.

Yu, L., C.X. Zhu, Y.C. Tse-Dinh, and S.W. Fesik. 1995. Solution structure of the C-terminal single-stranded DNA-binding domain of E. coli topoisomerase I. Biochemistry 34:7622-7628