Topoisomerase I
Created by Elizabeth Leeper
Topoisomerase I (PDB ID = 1A36) is an enzyme that alleviates topological problems inherent to DNA’s double-helical structure by regulating the underwinding or overwinding of DNA. These problems are associated with several nuclear processes including DNA replication, transcription, and recombination (Redinbo, Champoux, & Hol, 1999a; Redinbo, Stewart, Kuhn, Champoux, & Hol, 1998; Stewart, Redinbo, Qiu, Hol, & Champoux, 1998). Topoisomerases relax superhelical tension created during nuclear processes by forming transient breaks in the DNA helix. Hence, these enzymes are essential for providing DNA that is structurally able to participate in important nuclear activities. Additionally, some chemotherapy drugs work through the inhibition of eukaryotic topoisomerases to induce apoptosis in cancer cells. These drugs trap DNA to the topoisomerase enzyme, impeding the cell’s ability to complete transcription and replication, thereby increasing levels of DNA breaks that lead to programmed cell death (Redinbo et al., 1999a; Redinbo et al., 1998).
Topoisomerases are divided into two classes (Redinbo et al., 1999a). Type I topoisomerases are monomeric, require no energy cofactor, and cleave one strand of the DNA double-helix. Conversely, type II topoisomerases function as dimers, are ATP-dependent, and cleave both strands of DNA. Both classes have a conserved tyrosine residue despite their structural and mechanical differences (Redinbo et al, 1998). Furthermore, type I topoisomerases occur as either type IA or type IB, which are unrelated enzymes used to execute related functions (Redinbo et al., 1999a). Type IB topoisomerases are almost exclusively found in eukaryotic cells and differ from type IA in that they can relax positive and negative supercoils, require no metal ion or single-stranded region of DNA, and become attached to the 3´ end of DNA. The type IB topoisomerase found in humans is the focus of the remaining discussion. The molecular weight of human topoisomerase I (topo I) is 91 kDa (Redinbo et al., 1998), and its isoelectric point is 9.37 (Artimo et al., 2012).
Topo I functions as a monomeric protein of 765 amino acids consisting of four major regions: the NH2-terminal (not shown), core, linker, and COOH-terminal domains (Redinbo et al., 1999a; Redinbo et al., 1998; Stewart et al., 1998). The NH2-terminal, comprised of the first 210 residues, exhibits very limited spatial organization, is highly charged, and contains 90% polar residues (Redinbo et al., 1999a). This domain shows no catalytic function and is not required for the relaxation of DNA supercoils. However, the NH2 terminal experiences protein-protein interactions within the nucleus. The linker region, comprised of residues 636 to 712, connects the central core of the enzyme to the COOH-terminal. The linker region assumes a coiled-coil configuration with an asymmetric distribution of charges: the DNA proximal (top) side of the coiled-coil shows a large net positive charge while the bottom side is only slightly positively charged. This region is also not involved in DNA relaxation or catalytic functions, but it does affect the enzyme’s processivity (Redinbo et al., 1999a). The COOH-terminal, residues 713 to 765, contains the critical catalytic residue Tyr-732.The core region, residues 211 to 635, contains the protein’s active site and is divided into three conserved subdomains. The COOH-terminal and the core region create the catalytic domain of topo I and are the most integral to its activity.
Topo I acts as a bi-lobed clamp that temporarily binds to DNA by creating a DNA-binding pore. One lobe, core subdomains I and II, sits on top of the DNA helix in a tightly-folded complex and is termed the cap. Subdomain I contains two α helices and nine β sheets while subdomain II consists of five α helices and two β sheets (Redinbo et al., 1998). These subdomains also contain two unique ‘nose cone’ helices that extend from the body of the enzyme coming together in a ‘V.’ These ‘nose cone’ helices are positioned above the DNA major groove, and while highly positively charged, only Arg-316 makes direct contact with a phosphate on the cleaved strand. (Redinbo et al., 1999a; Stewart et al., 1998). The second lobe, core subdomain III and the COOH-terminal domain, sits below the DNA. Five α helices comprise the COOH-terminal, which also contains the active site Tyr-732. Subdomain III is a complex arrangement of ten α helices and three β strands, and it includes most of the active site residues, except for Tyr-732 (Redinbo et al., 1998). Hence, the bottom lobe is comprised of all the catalytic residues used for strand cleavage and religation. The two lobes are covalently linked by means of a continuous chain on one side of the DNA molecule. Their only point of direct contact is a region composed of 6 amino acids and one salt bridge on the other side of the DNA molecule. Subdomains I and III undergo extensive interactions with DNA substrate and are the main components of topo I’s molecular clamp.
The linker region protrudes from the second, bottom lobe, extending a positively charged face toward the DNA (Stewart et al., 1998). However, like the ‘nose cone’ helices, the linker region fails to make significant interactions with DNA. The positively charged top surface of the linker region contacts DNA only at two residues: Lys-650 and Arg-708. These residues are conserved as lysine and arginine in all type I topoisomerases. Thus, topo I includes two positively charged helical surfaces (‘nose cone’ helices and the coiled-coil linker) facing but not directly interacting with the DNA substrate. These helices are believed to “play an important role in the mechanism of topoisomerization” (Stewart et al., 1998, 1538).
Additionally, the linker region demonstrates limited interactions with the rest of the enzyme. The linker coiled-coil only contacts the last helix of core subdomain III with the COOH-terminal end of the second linker helix. Hydrophobic interactions between the two helices ensue, in addition to the formation of two salt bridges. These interactions create a three-helix sheet and are the only interactions observed between the linker helices and the rest topo I molecule. Stewart et al. (1998) speculates that these minimal interactions lead to a flexible linker domain that shifts by moving hydrophobic surfaces along each other.
Topo I’s active site is found in the central DNA-binding pore of the enzyme. This channel is highly positively charged and coordinates well the negative charges found on the DNA sugar-phosphate backbone (Stewart et al., 1998). The active site is composed of four conserved residues: Arg-488, Arg-590, His-632, and Tyr-723. These residues are assembled around the scissile phosphate, and the tyrosine residue is positioned for nucleophilic attack and subsequent covalent attachment to the 3´ end of the broken DNA strand (Stewart et al., 1998). The active site residues incur a small shift upon strand cleavage and covalent attachment (Redinbo et al., 1999a). However, Redinbo et al. (1998) demonstrate that there are few structural differences between the covalently attached protein-DNA complex and the noncovalent complex. The changes are reserved for the catalytic Tyr-723 residue and the scissile phosphate group.
Stewart et al. (1998) proposes a ‘controlled rotation’ mechanism for the relaxation of superhelical tension by topo I. This mechanism suggests that instead of rotating freely, the DNA substrate downstream of the cleavage site rotates under the influence of the surrounding protein. The DNA is thought to rotate about several bonds in the intact strand within a positively charged cavity formed by the ‘nose cone’ helices and the linker domain of the enzyme. The orthogonal arrangement and the empty triangular-shaped space between the two 'nose cone' helices are thought to guide the rotating DNA. Similarly, the linker domain was proposed to guide the relaxation event. The ‘nose cone’ helices and linker domain show high degrees of conformational flexibility consistent with the structural flexibility necessitated by the ‘controlled rotation’ mechanism (Redinbo, Stewart, Champoux, & Hol, 1999b).
The topoisomerization reaction begins with the binding of topo I to the DNA substrate (Stewart et al., 1998). The enzyme must initially exist in an ‘open’ conformation, which is likely achieved by a hinge-bending motion at Pro-431 and Lys-452. The binding event is sensitive to surface and charge complementarity of the enzyme and DNA. However, there is little evidence for the protein recognizing specific DNA sequences (Redinbo et al., 1998). Ultimately, binding results in the complete embrace of the DNA substrate by topo I, placing the substrate within the positively charged active site channel. The active site residues are in position for the attack and cleavage of the scissile strand and covalent attachment of Tyr-723 to the 3´ end of the DNA. This covalent intermediate is able to undergo one more or cycles of controlled rotation to release superhelical tension. The intermediate is religated upon release of the Tyr-723 residue from the end of the DNA, producing a DNA molecule with reduced superhelicity (Stewart et al., 1998).
Topo I is the sole target of camptothecin (CPT) and its derivatives. These compounds are emerging as potent anticancer drugs due to their ability to stabilize topo I-DNA complexes and increase the yield of covalent intermediates thereby inducing cell apoptosis. Specific structural components of CPTs have been elucidated as critical to their drug activity and inhibition of topo I (Redinbo et al., 1998). These components include the lactone, pyridone, and 20(S)-hydroxyl moieties. Additionally, mutagenesis studies have shown that when the Asn-722, Arg-364, and Asp-533 residues of topo I are mutated, the enzyme is rendered insensitive to CPT (Redinbo et al., 1999a). However, the structure of CPT bound to the covalent complex of topo I and DNA has yet to be determined.
Lambda integrase (PDB ID = 1Z1B) is found in enterobacteria phage lambda, a virus that infects Escherichia coli. This enzyme has conserved topo I’s catalytic pentad in its tertiary structure despite limited sequence similarity. The PSI-BLAST query showed that lambda integrase (λ-int) has very little sequence homology to topo I, providing an E score of 0.85 (Altschul et al., 1997). However, the Dali Server demonstrated that the fold of topo I and λ-int is quite similar, resulting in a Z score of 10.0 (Holm & Rosenström, 2010). Accordingly, topo I and λ-int maintain similar functions with similar tertiary structures.
λ-int is a member of a large family of site-specific recombinases that catalyze rearrangements between DNA sequences (Biswas et al., 2005). Like topo I, λ-int does not rely on high-energy cofactors and is able to cleave double-stranded DNA. A nucleophilic tyrosine residue is conserved between the two proteins, which leads to the formation of a covalent DNA-protein complex. Additionally, the other three catalytic residues of topo I, Arg-His-Arg, are conserved in λ-int. Hence, the catalytic domain of topo I is preserved in λ-int (Chen, Kussie, Pavletich, & Shuman, 1998). Additionally, λ-int contains flexible, positively charged regions similar to the ‘nose cone’ helices and linker region of topo I. However, λ-int functions as tetramer, each with four domains, and its functional form includes two active sites for the binding and recombination of two DNA strands (Biswas et al., 2005). The amino acid sequences of λ-int and topo I are very dissimilar and contain many deletions and insertions between the two proteins (Chen et al., 1998).