Biochemical and Structural Studies

Valuable insights into how DNA polymerases process their substrates were obtained as a result of detailed kinetic studies of the enzymes. Benkovic and co-workers employed rapid quenching techniques to study the kinetics of transient intermediates in the reaction pathway of DNA polymerases [5]. Intensive studies revealed that E. coli DNA polymerase I follows an ordered sequential reaction pathway when promoting DNA synthesis. Important aspects of these results for DNA polymerase fidelity are conformational changes before and after the chemical step and the occurrence of different rate-limiting steps for insertion of canonical and non-canonical nucleotides. E. coli DNA polymerase I discriminates between canonical and non-canonical nucleotide insertion by formation of the chemical bond. Bond formation proceeds at a rate more than several thousand times slower when an incorrect dNTP is processed compared with canonical nucleotide insertion.

Fig. 4.1.2. A. Structure of Thermus aquaticus (Taq) DNA polymerase bound to the DNA primer-template complex. B. "Open" and "closed" conformation after dNTP binding of Taq DNA polymerase. Structures were built on PDB entries 3KTQ and 4KTQ.

Recently determined crystal structures of several DNA polymerases in complexes with their DNA and dNTP substrates have contributed significantly to our understanding of structure and substrate recognition by these complex enzymes [6]. Most DNA polymerases with known structures have a large cleft in which the primer template complex is embedded (Box 16). By analogy of this conformation with a half open right hand the enzyme domains are termed thumb, palm, and fingers (Figure 4.1.2A).

The palm domain harbors the catalytic center comprising the essential carboxy-lates involved in the phosphoryl transfer reaction. The high conservation of this domain throughout distinct DNA polymerase families, e.g. eukaryotic, prokaryotic, and viral DNA polymerases, is striking. In contrast, the finger and thumb domains, which make extensive contact with the primer template complex and the incoming dNTPs, differ significantly among DNA polymerases. Results from structural investigation of DNA polymerases strongly support the occurrence of large conformational changes from an "open" to a "closed" conformation before phosphodiester bond formation, triggered by dNTP binding (Figure 4.1.2B). Editing of nascent nucleotide base-pair geometry during these transitions is believed to be a crucial determinant of DNA polymerase selectivity.

Data available from crystal structures of DNA polymerases suggest the formation of nucleotide binding pockets which preferentially accommodate Watson-Crick base pairs. Nevertheless, DNA polymerase selectivity often varies significantly depending on the DNA polymerase [7]. By comparison of crystal structures derived from high- and low-fidelity enzymes valuable insights were gained into structural differences which might be the origin of the often considerably different selectivity of these enzymes. Error-prone Sulfolobus solfataricus P2 DNA polymerase IV (Dpo4) was found to adopt the familiar overall shape of a half-open right hand, found in several DNA polymerases, in which parts of the primer-template are embedded [6d]. The palm domain is structurally similar to those of other DNA polymerases and the essential carboxylates in Dpo4 are in identical positions as found in high-fidelity DNA polymerases and in HIV-1 reverse transcriptase. The finger and thumb domains of Dpo4 which surround the incoming triphosphate and template nucleotide are, however, unusually small. The "O-helix" (Figure 4.1.2B) that is believed to be involved in editing of correct nucleobase geometry and present in all high-fidelity DNA polymerases is absent in Dpo4. In total, the nascent base pair between the template and the incoming nucleotide in Dpo4 is less tightly surrounded in the vicinity of this enzyme in comparison to high-fidelity DNA polymerases. The open and solvent-accessible active site might be one structural reason for the error-prone replication of DNA by this kind of DNA polymerase. Another structure of this enzyme had a complexed non-canonical dNTP bound to the active site. From the structure it is apparent that conformations of the sugar phosphate moieties of the primer, template and nucleoside triphosphate in the active site differ significantly from that found when a canonical nucleotide is bound [6d]. Because of translocation of the template without replication of the first template base (G), the incoming ddGTP forms a canonical base pair with the next template base (C). Such an alignment might be the origin of the apparent faulty DNA synthesis of this kind of enzyme leading to frameshifts (Figure 4.1.3).

Taking together, DNA polymerase structural data indicate a high degree of shape complementary between the active sites of the enzymes and the nucleotide substrates, suggesting that geometrical constraints are at least one cause of DNA polymerase fidelity.

These assumptions are further supported by the finding that mutations that are believed to alter the geometry of the binding pocket or the conformational changes

Fig. 4.1.3. DNA bound in the active site of Dpo4. A. Primer-template and incoming canonical triphosphate ddATP. B. Primer-template and incoming non-canonical triphosphate (ddGTP). Structures were built on PDB entry code 1JX4 and 1JXL.

needed to trigger catalysis effect the fidelity of the DNA polymerase [3a,b]. One of the most striking examples of this is the Arg283Ala mutation in DNA polymerase b. Arg283 is part of the nucleotide-binding pocket and its substitution with a steri-cally less demanding alanine moiety results in a marked decrease in fidelity. Similar results were obtained by mutation of E. coli DNA polymerase I and HIV-1 reverse transcriptase, strongly suggesting the participation of steric constraints in DNA polymerase selectivity mechanisms.

0 0

Post a comment