Model for the catalytic domain of the proofreading epsilon subunit of Escherichia coli DNA polymerase III based on NMR structural data.

The DNA polymerase III holoenzyme (HE) is the primary replicative polymerase of Escherichia coli. The epsilon subunit of the HE complex provides the 3'-exonucleolytic proofreading activity for this enzyme complex. epsilon consists of two domains: an N-terminal domain containing the proofreading exonuclease activity (residues 1-186) and a C-terminal domain required for binding to the polymerase (alpha) subunit (residues 187-243). Multidimensional NMR studies of (2)H-, (13)C-, and (15)N-labeled N-terminal domains (epsilon186) were performed to assign the backbone resonances and measure H(N)-H(N) nuclear Overhauser effects (NOEs). NMR studies were also performed on triple-lableled [U-(2)H,(13)C,(15)N]epsilon186 containing Val, Leu, and Ile residues with protonated methyl groups, which allowed for the assignment of H(N)-CH(3) and CH(3)-CH(3) NOEs. Analysis of the (13)C(alpha), (13)C(beta), and (13)CO shifts, using chemical shift indexing and the TALOS program, allowed for the identification of regions of the secondary structure. H(N)-H(N) NOEs provided information on the assembly of the extended strands into a beta-sheet structure and confirmed the assignment of the alpha helices. Measurement of H(N)-CH(3) and CH(3)-CH(3) NOEs confirmed the beta-sheet structure and assisted in the positioning of the alpha helices. The resulting preliminary characterization of the three-dimensional structure of the protein indicated that significant structural homology exists with the active site of the Klenow proofreading exonuclease domain, despite the extremely limited sequence homology. On the basis of this analogy, molecular modeling studies of epsilon186 were performed using as templates the crystal structures of the exonuclease domains of the Klenow fragment and the T4 DNA polymerase and the recently determined structure of the E. coli Exonuclease I. A multiple sequence alignment was constructed, with the initial alignment taken from the previously published hidden Markov model and NMR constraints. Because several of the published structures included complexed ssDNA, we were also able to incorporate an A-C-G trinucleotide into the epsilon186 structure. Nearly all of the residues which have been identified as mutators are located in the portion of the molecule which binds the DNA, with most of these playing either a catalytic or structural role.