Association of crossover points with topoisomerase I cleavage sites: a model for nonhomologous recombination.

Nonhomologous DNA recombination is frequently observed in somatic cells upon the introduction of DNA into cells or in chromosomal events involving sequences already stably carried by the genome. In this report, the DNA sequences at the crossover points for excision of SV40 from chromosomes were shown to be associated with eukaryotic topoisomerase I cleavage sites in vitro. The precise location of the cleavage sites relative to the crossover points has suggested a general model for nonhomologous recombination mediated by topoisomerase I.

Crystal structures of human topoisomerase I in covalent and noncovalent complexes with DNA.

Topoisomerases I promote the relaxation of DNA superhelical tension by introducing a transient single-stranded break in duplex DNA and are vital for the processes of replication, transcription, and recombination. The crystal structures at 2.1 and 2.5 angstrom resolution of reconstituted human topoisomerase I comprising the core and carboxyl-terminal domains in covalent and noncovalent complexes with 22-base pair DNA duplexes reveal an enzyme that "clamps" around essentially B-form DNA. The core domain and the first eight residues of the carboxyl-terminal domain of the enzyme, including the active-site nucleophile tyrosine-723, share significant structural similarity with the bacteriophage family of DNA integrases. A binding mode for the anticancer drug camptothecin is proposed on the basis of chemical and biochemical information combined with these three-dimensional structures of topoisomerase I-DNA complexes.

A model for the mechanism of human topoisomerase I.

The three-dimensional structure of a 70-kilodalton amino terminally truncated form of human topoisomerase I in complex with a 22-base pair duplex oligonucleotide, determined to a resolution of 2.8 angstroms, reveals all of the structural elements of the enzyme that contact DNA. The linker region that connects the central core of the enzyme to the carboxyl-terminal domain assumes a coiled-coil configuration and protrudes away from the remainder of the enzyme. The positively charged DNA-proximal surface of the linker makes only a few contacts with the DNA downstream of the cleavage site. In combination with the crystal structures of the reconstituted human topoisomerase I before and after DNA cleavage, this information suggests which amino acid residues are involved in catalyzing phosphodiester bond breakage and religation. The structures also lead to the proposal that the topoisomerization step occurs by a mechanism termed "controlled rotation."

Mapping in vivo topoisomerase I sites on simian virus 40 DNA: asymmetric distribution of sites on replicating molecules.

Complexes between simian virus 40 DNA and topoisomerase I (topo I) were isolated from infected cells treated with camptothecin. The topo I break sites were precisely mapped by primer extension from defined oligonucleotides. Of the 56 sites, 40 conform to the in vitro consensus sequence previously determined for topo I. The remaining 16 sites have an unknown origin and were detectable even in the absence of camptothecin. Only 11% of the potential break sites were actually broken in vivo. In the regions mapped, the pattern of break sites was asymmetric. Most notable are the clustering of sites near the terminus for DNA replication and the confinement of sites to the strand that is the template for discontinuous DNA synthesis. These asymmetries could reflect the role of topo I in simian virus 40 DNA replication and suggest that topo I action is coordinated spatially with that of the replication complex.

Structural insights into the function of type IB topoisomerases.

Topoisomerases relax the DNA superhelical tension that arises in cells as a result of several nuclear processes, including transcription, replication and recombination. Recently determined crystal structures of human topoisomerase I in complex with DNA and of the 27 kDa catalytic domain of the vaccinia virus topoisomerase have advanced our understanding of the eukaryotic type IB topoisomerases. These recent structural results provide insights into functional aspects of these topoisomerases, including their DNA binding, strand cleavage and religation activities, as well as the mechanism that these enzymes use to relax DNA superhelical tension. In addition, two proposed models of the anticancer drug camptothecin bound to a covalent complex of human topoisomerase I and DNA suggest a structural basis for the mode of action of the drug.

Domains of human topoisomerase I and associated functions.

Human topoisomerase I can be divided into four domains based on homology alignments, physical properties, sensitivity to limited proteolysis, and fragment complementation studies. Roughly the first 197 amino acids represent the N-terminal domain that appears to be devoid of secondary structure and is likely important for targeting the enzyme to its sites of action within the nucleus of the cell. The core domain encompasses residues approximately 198 to approximately 651, is involved in catalysis, and is important for the preferential binding of the enzyme to supercoiled DNA. The C-terminal domain extends from residue approximately 697 to the end of the protein at residue 765 and contains the catalytically important active site tyrosine at position 723. The core and C-terminal domains are connected by a poorly conserved, protease-sensitive linker domain (residues approximately 652 to approximately 696) that has been implicated in DNA binding and may influence how long the enzyme remains in the nicked stated. Mutations that confer resistance to the topoisomerase I poison camptothecin are located in the core and C-terminal domains and presumably identify residues important for drug binding.

Overexpression of human topoisomerase I in baby hamster kidney cells: hypersensitivity of clonal isolates to camptothecin.

The 3645-base pair human topoisomerase I complementary DNA (cDNA) clone isolated by D'Arpa et al. (Proc. Natl. Acad. Sci. USA, 85:2543-2547, 1988) and a mutated version of the cDNA encoding a protein with phenylalanine instead of tyrosine at position 723 have been overexpressed 2- to 5-fold in stably transfected baby hamster kidney cells. The overexpressed proteins are the same size as the topoisomerase I present in Hela cells, indicating that the cDNA clone contains the complete topoisomerase I coding sequence. Some human colon carcinoma cells have increased levels of topoisomerase I and are hypersensitive to the drug camptothecin. The overexpressed wild-type topoisomerase I does not affect the cell growth or morphology of the baby hamster kidney cells, suggesting that elevated levels of topoisomerase I alone are not sufficient to cause cell transformation. However, the overexpressed wild-type protein is active, as shown by the hypersensitivity of clonal cell lines to camptothecin. The mutant form of topoisomerase I is enzymatically inactive by two criteria. First, extracts of Escherichia coli cells carrying the mutant cDNA contain no activity capable of relaxing superhelical DNA under conditions where activity is easily detectable in extracts from cells containing the wild-type cDNA. Second, baby hamster kidney cells stably transfected by the mutant cDNA are no more sensitive to camptothecin than control untransfected cells. These results indicate that tyrosine 723 is essential for enzyme activity and are consistent with predictions based on homology comparisons with the yeast enzymes, that this is the active-site tyrosine in the human topoisomerase I.

The effect of salt on the binding of the eucaryotic DNA nicking-closing enzyme to DNA and chromatin.

The optimum monovalent cation concentration (Na+ or K+) for the relaxation of superhelical DNA by the rat liver nicking-closing enzyme under conditions of DNA excess was found to be 150-200 mM. The detection of a nicked DNA species after stopping a reaction with alkali depends on having a high molar ratio of enzyme to DNA and is maximal between 50 and 100 mM monovalent cation. Varying the salt concentration from 15 to 200 mM appears to have no effect on the catalysis of the nicking -closing reaction by the enzyme. Instead different salt optima in these two assays can be explained by the observation that the nicking-closing enzyme acts by a processive mechanism below 100 mM salt and becomes nonprocessive above 150 mM. The salt elution of the nicking-closing enzyme from resting cell chromatin appears to be similar to that which one would expect for the elution of the enzyme from naked DNA. However, greater than 70% of the chromatin associated enzyme activity remained bound to chromatin from growing cells at 300 mM salt, a concentration at which there is no significant binding to naked DNA in vitro.

DNA strand breakage by wheat germ type 1 topoisomerase.

Properties of strand breakage in duplex and single-stranded DNA by the wheat germ type 1 DNA topoisomerase were investigated. Strand breakage in duplex DNA is dependent upon the use of denaturing conditions to inactivate the enzyme and terminate the reaction, whereas breakage of single-stranded DNA occurs under the normal reaction conditions and is not dependent upon denaturation. Breakage generates a free 5' hydroxyl group and enzyme bound to the 3' side of the break, presumably via the 3' phosphate group. The location of sites of breakage with both duplex and single-stranded DNA is not random. In all these respects the wheat germ enzyme closely resembles the rat liver type 1 topoisomerase. A comparison of the locations of the sites of breakage in duplex DNA generated by the topoisomerases from wheat germ and rat liver indicates a number of common sites, although the patterns of breakage are not identical.

Analysis of the biased distribution of topoisomerase I break sites on replicating simian virus 40 DNA.

We previously mapped the locations of breaks introduced by eukaryotic topoisomerase I (topo I) in replicating simian virus 40 (SV40) DNA and observed an approximate 3:1 bias in the distribution of the break sites for the template strand for discontinuous DNA synthesis. In the present study, this bias has been confirmed by the mapping of additional sites utilizing a standard primer extension assay and a sensitive repetitive primer extension (RPE) method. No new sites could be detected on either strand of SV40 by the RPE method, despite the 10 to 20 fold greater sensitivity of the technique. To investigate the nature of the bias, a detailed analysis of the SV40 DNA sequence was undertaken. A set of 17 pentanucleotide sequences derived from those sites observed to be broken in the viral DNA extracted from SV40-infected cells define an in vivo consensus sequence. We show that the observed strand bias is likely due to the intrinsic asymmetric distribution of these consensus sequences on the two strands of SV40 DNA. To confirm these observations, double-stranded oligonucleotides containing previously identified in vivo topo I break sites were introduced in both orientations into SV40 to generate insertion mutants. Mapping experiments utilizing these mutants revealed that the inserted topo I break sites were broken in vivo regardless of their orientation, confirming that the SV40 sequence is the major, if not the sole determinant, of the observed strand bias. The possible origins of the strand bias are discussed in relation to the evolution of the virus.

Structural alterations in the DNA ahead of the primer terminus during displacement synthesis by reverse transcriptases.

Unlike most DNA polymerases, reverse transcriptases can initiate DNA synthesis at a single-strand break and displace the downstream non- template strand simultaneously with extension of the primer. This reaction is important for generation of the long terminal repeat sequences in the duplex DNA product of retroviral reverse transcription. Oligonucleotide-based model displacement constructs were used to study the interaction of human immunodeficiency virus type 1 and Moloney murine leukemia virus reverse transcriptases with the DNA. Under conditions where the DNA is saturated with enzyme, there is no protection against DNase I cleavage of the 5' single-stranded extension that would correspond to the already-displaced strand. However, the DNase I footprint on the non-template strand extends from the +1 to the +9 position for the human immunodeficiency virus type 1 enzyme and from +1 to +7 or +8 for the Moloney enzyme. This extent of protection on the non-template strand is similar to what was observed previously for the template strand downstream from the primer terminus. Use of potassium permanganate as a probe for unpaired bases in the region ahead of the primer terminus reveals that the two base-pairs immediately in front of the enzyme are melted by the bound enzyme. These findings are consistent with a displacement mechanism in which the reverse transcriptase plays an active role in unpairing the DNA ahead of the translocating polymerase. The results are interpreted in light of a recent crystal structure showing the nature of the protein-DNA contacts with the template strand ahead of the primer terminus.

Plus-strand priming by Moloney murine leukemia virus. The sequence features important for cleavage by RNase H.

The reverse transcriptase-associated RNase H activity is responsible for producing the plus-strand RNA primer during reverse transcription. The major plus-strand initiation site is located within a highly conserved polypurine tract (PPT), and initiation of DNA replication at this site is necessary for proper formation of the viral long terminal repeats (LTRs). We present here a compilation of PPT sequences from an evolutionarily diverse group of retroviruses and retrotransposons, which reveals that there is a high degree of sequence conservation at this site. Furthermore, we found previously that secondary plus-strand origins, identified in vitro, also show strong similarity to the PPT. Taken together, these data suggest that RNase H recognizes a specific sequence at the PPT as a signal to cleave the RNA at a precise location, producing a primer for the initiation of plus-DNA strands. We have analyzed the RNase H recognition sequence by producing a large number of single and double mutations within the PPT. Our findings suggest that no single residue in the +5 to -6 region (where the cleavage occurs between -1 and +1) is essential; mutations at these positions introduced heterogeneity at the cleavage site, but cleavage is still predominantly at the correct location. Furthermore, base-pairing is not required at the +1 position of the RNase H cleavage site, but a mismatched base-pair at the -1 position causes imprecision in the cleavage reaction. Interestingly, the A residue at position -7 seems to be critical in positioning the RNase H enzyme for correct cleavage. The preference of the enzyme for cleaving between G and A residues may play a minor role in determining the specificity.

Breakage of single-stranded DNA by eukaryotic type 1 topoisomerase occurs only at regions with the potential for base-pairing.

Eukaryotic type 1 DNA topoisomerases break single-stranded DNA at specific sites. A preferred site for rat liver topoisomerase breakage in single-stranded phi X174 DNA was located within a region of the DNA with the potential for duplex formation. To investigate the relationship between sites of breakage in duplex and single-stranded DNA, a restriction fragment containing sequences from the transcriptional regulatory and enhancer region of the simian virus 40 genome was used as a substrate for topoisomerase. While different patterns of breakage in the native and denatured forms of the DNA were observed, some sites of breakage were common to both forms. The break sites in the denatured DNA were a subset of the break sites in the duplex DNA and were located in regions which had the potential for intrastrand base-pairing due to distal complementary sequences. A series of single-stranded fragments were generated with the distal complementary sequences deleted and these fragments were used as substrates for topoisomerase breakage. The lack of detectable breakage at a site when the complementary sequence was deleted, suggests that topoisomerase acts at duplex regions in the single-stranded DNA and that it is not active on regions of single-stranded DNA that are not base-paired.

Properties of strand displacement synthesis by Moloney murine leukemia virus reverse transcriptase: mechanistic implications.

Previous results indicated that Moloney murine leukemia virus reverse transcriptase is capable of extensive synthesis under conditions where it must simultaneously displace a downstream non-template DNA strand. To investigate more fully the mechanistic basis for displacement synthesis and to characterize the activity with natural viral templates, displacement and non-displacement synthesis were compared under a variety of conditions using the viral long terminal repeat plus strand as the template. Although the rates of both displacement and non-displacement synthesis varied regionally over the template, on the average, displacement synthesis was slower by a factor of approximately 3 to 4. Surprisingly, with one particular primer situated downstream of the tRNA primer binding site, displacement synthesis was found to be at least tenfold more processive than non-displacement synthesis, approaching a value of 500 nucleotides. The sequence features associated with pausing during the two modes of synthesis are different in both nucleotide preference and position relative to the enzyme, suggesting that the enzyme contacts the DNA differently under the two modes of synthesis. It was found that pausing during displacement synthesis did not reflect those local regions of DNA with a predicted high degree of thermal stability. Moreover, the very similar effects of temperature on the rates of displacement and non-displacement synthesis make unlikely a strictly passive mechanism of displacement synthesis whereby breathing of the downstream duplex is sufficient for advancement of the polymerase. Together, these results suggest a mechanism of displacement synthesis in which reverse transcriptase actively participates in the process of strand separation in front of the translocating polymerase.

Reconstitution of human topoisomerase I by fragment complementation.

Human topoisomerase I (topo I, 91 kDa) is composed of four major domains; the unconserved and highly charged "N-terminal" domain (24 kDa), the conserved "core" domain (54 kDa), a poorly conserved and positively charged "linker" region (5 kDa), and the highly conserved "C-terminal" domain (8 kDa) which contains the active site tyrosine at position 723. Here we demonstrate that human topo I activity can be reconstituted by mixing a 58 kDa recombinant core domain (residues Lys175 to Ala659) with any one of a series of recombinant C-terminal fragments that range in size from 12 kDa (linker and C-terminal domains, residues Leu658 to Phe765) to 6.3 kDa (C-terminal domain residues Gln713 to Phe765). The C-terminal fragments bind tightly to the core domain, forming a 1:1 complex that is stable irrespective of ionic strength (0.01 to 1 M). The reconstituted enzymes are active only over a relatively narrow range of salt concentrations (25 to 200 mM KCl) as compared to the intact topo70 enzyme (missing the N-terminal domain). Under physiological conditions (150 mM KCl and 10 mM Mg2+) they are much more distributive in their mode of action than topo70. The reconstituted enzyme binds DNA with an affinity that is approximately 20-fold lower than that of the intact topo70. In addition, the cleavage/religation equilibrium of the reconstituted enzyme appears to be biased towards religation relative to that of the intact enzyme. Despite differences in the cleavage/religation equilibrium and affinity for DNA, the reconstituted and intact enzymes have identical sequence specificities for the cleavage of duplex DNA or suicide cleavage of oligonucleotide substrates.

Structural flexibility in human topoisomerase I revealed in multiple non-isomorphous crystal structures.

Human topoisomerase I plays a critical role in chromosomal stability by relaxing the DNA superhelical tension that arises from a variety of nuclear processes, including replication, transcription, and chromatin remodeling. Human topoisomerase I is a approximately 91 kDa enzyme composed of four major domains: a 24 kDa N-terminal domain, a56 kDa core domain, a7 kDa linker domain, and a6 kDa C-terminal domain containing the active-site Tyr723 residue. A monoclinic crystal structure of a 70 kDa N-terminally truncated form of human topoisomerase I in non-covalent complex with a 22 bp DNA duplex exhibited remarkable crystal-to-crystal non-isomorphism; shifts in cell constants of up to 9 A in the b -axis length and up to 8.5 degrees in the beta-angle were observed. Eight crystal structures of human topoisomerase I - DNA complexes from this crystal form were determined to between 2.8 and 3.25 A resolution. These structures revealed both dramatic shifts in crystal packing and functionally suggestive regions of conformational flexibility in the structure of the enzyme. Crystal packing shifts of up to 20.5 A combined with rotations of up to 11.5 degrees were observed, helping to explain the variability in cell constants. When the core subdomain III regions of the eight structures are superimposed, the "cap" (core subdomains I and II) of the molecule is observed to rotate by up to 4.6 degrees and to shift by up to 3.6 A. The linker domain shows the greatest degree of conformational flexibility, rotating and shifting by up to 2.5 degrees and 4.6 A, respectively, in five of eight structures, and becoming disordered altogether in the remaining three. These observed regions of conformational flexibility in the cap and the linker domain are consistent with the structural flexibility invoked in the "controled rotation" mechanism proposed for the relaxation of DNA superhelical tension by human topoisomerase I.

Structure-based analysis of the effects of camptothecin on the activities of human topoisomerase I.

The sole target for the anticancer drug camptothecin (CPT) is the type I topoisomerase. The drug poisons the topoisomerase by slowing the religation step of the reaction, thereby trapping the enzyme in a covalent complex on the DNA. In addition, CPT has been shown to inhibit plasmid DNA relaxation in vitro. The structural bases for these two activities of CPT are explored in relation to the recently published crystal structure of the enzyme with bound DNA.

Mechanism of catalysis by eukaryotic DNA topoisomerase I.

The elucidation of the chemistry of the topo I reaction has provided the first example of how a phosphodiester bond in DNA can be temporarily broken and the energy for reclosure stored in a covalent linkage between the end of the broken strand and the enzyme (Champoux, 1977a, 1981). This type of reaction offers several advantages to the cell. First, unnecessary exposure of DNA ends to nucleolytic attack is prevented. Second, breakage and reclosure of DNA strands can occur without an expenditure of ATP energy. Third, the combined breakage and rejoining reactions can be both spatially and temporally coordinated with other cellular activities by regulating the activity of a single protein molecule. This general mechanism has not only been extended to type II topoisomerases (see Chapters 3 and 5), but also to the specialized single-stranded phage replication proteins (e.g., phi X174 gene A protein) (Ikeda et al., 1976; Eisenberg et al., 1977) and to site-specific recombinases such as the bacteriophage lambda integrase (Craig and Nash, 1983), the delta gamma and Tn3 resolvases (Reed, 1981; Reed and Grindley, 1981; Krasnow and Cozzarelli, 1983; Hatfull and Grindley, 1986), and the yeast 2-microns circle FLP recombinase (Andrews et al., 1985; Gronostajski and Sadowski, 1985). Since the site-specific recombinases attach the broken strand to a different terminus rather than simply restoring the original phosphodiester bond as conventional topoisomerases do, they have been referred to as DNA strand transferases. It is conceivable that a similar mechanism applies to the rearrangement of immunoglobulin genes (Schatz et al., 1990) and to other specific genomic rearrangements that might occur during development (Matsuoka et al., 1991).