Hoogsteen base pair formation promotes synthesis opposite the 1,N 6 -ethenodeoxyadenosine lesion by human DNA polymerase i Deepak T Nair 1,3 , Robert E Johnson 2,3 , Louise Prakash 2 , Satya Prakash 2 & Aneel K Aggarwal 1 The 1,N 6 -ethenodeoxyadenosine (edA) lesion is promutagenic and has been implicated in carcinogenesis. We show here that human Poli, a Y-family DNA polymerase, can promote replication through this lesion by proficiently incorporating a nucleotide opposite it. The structural basis of this action is rotation of the edA adduct to the syn conformation in the Poli active site and presentation of its ‘Hoogsteen edge’ for hydrogen-bonding with incoming dTTP or dCTP. We also show that Polf carries out the subsequent extension reaction and that efficiency of extension from edA  T is notably higher than from edA  C. Together, our studies reveal for the first time how the exocyclic edA adduct is accommodated in a DNA polymerase active site, and they show that the combined action of Poli and Polf provides for efficient and error-free synthesis through this potentially carcinogenic DNA lesion. The edA adduct is promutagenic and has been implicated in carci- nogenesis 1 . The adduct is formed in DNA as a result of exposure to chemical carcinogens such as vinyl chloride, and it is also formed in unexposed animals and humans through DNA interaction with aldehydes derived from lipid peroxidation 2 . Lipid peroxidation, a normal chain-reaction process, initiates from the oxidation of polyunsaturated fatty acids in cell membranes and results in the production of a variety of highly reactive aldehydes, including acrolein, malonaldehyde and trans-4-hydroxy-2-nonenal (HNE). Enals such as HNE are converted to epoxyaldehyde by further oxidation reactions, and the reaction of epoxyaldehyde with adenine in DNA generates the edA adduct 2,3 . This adduct is highly inhibitory to synthesis by the replicative DNA polymerases (Pols) and also to synthesis by Y-family Pols such as Z and k 4 . The inability of these polymerases to replicate the edA adduct reflects their depen- dence upon Watson-Crick (W-C) base-pairing, as the exocyclic ring between the N1 and N6 positions of edA impairs the ‘W-C edge’ of adenine (Fig. 1). Poli differs greatly in many respects from other DNA polymerases 5 . By contrast to almost all the DNA Pols, which form the four correct W-C base pairs with very similar catalytic efficiencies, Poli incorporates nucleotides with a much higher efficiency and fidelity opposite template purines than opposite template pyrimidines. Furthermore, among template purines, Poli has a higher efficiency and fidelity opposite template A than opposite template G. Poli is highly inefficient at incorporating the correct nucleotide opposite template C and T; opposite template T, it misincorporates a G about ten times more frequently than it incorporates the correct nucleotide, A 6–10 . The ternary crystal structures of Poli bound to template purines and the correct incoming dNTP have yielded major insights into the action mechanism of this polymerase. Specifically, in the structure of Poli bound to template A and an incoming dTTP, the template A adopts a syn conformation and forms a Hoogsteen base pair with the incoming T, which remains in the anti conformation 11 . Similarly, in the structure of Poli paired with a template G and an incoming dCTP, aG  C+ Hoogsteen base pair is formed in the polymerase active site 12 . Hoogsteen base-pairing provides a basis for the much higher efficiency and fidelity of nucleotide incorporation opposite template purines than opposite pyrimidines, because only the templates A and G have a ‘Hoogsteen edge’ by which they can establish two hydrogen bonds with the correct incoming pyrimidine nucleotide. Because of the lack of a Hoogsteen edge, there are few hydrogen-bonding possibilities opposite template pyrimidines. The ability of the Poli active site to push template purines into the syn conformation provides a mechanism by which this polymerase could promote replication through minor groove adducts, such as those derived from the reaction of acrolein (from lipid peroxidation) with the N2 of G. In particular, rotation from anti to syn would displace these adducts into the major groove, where there is much less steric interference. The dependence of Poli upon Hoogsteen base- pairing raises the further possibility that, in addition to tolerating lesions that impinge upon the minor groove, Poli could also tolerate lesions such as edA that disrupt the W-C edge but not the Hoogsteen Received 15 February; accepted 5 June; published online 2 July 2006; doi:10.1038/nsmb1118 1 Department of Molecular Physiology and Biophysics, Mount Sinai School of Medicine, Box 1677, 1425 Madison Avenue, New York, New York 10029, USA. 2 Sealy Center for Molecular Science, University of Texas Medical Branch, 6.014 Medical Research Building, 11 th & Mechanic Streets, Galveston, Texas 77755, USA. 3 These authors contributed equally to this work. Correspondence should be addressed to A.K.A. (aneel.aggarwal@mssm.edu) or S.P. (saprakas@utmb.edu). NATURE STRUCTURAL & MOLECULAR BIOLOGY VOLUME 13 NUMBER 7 JULY 2006 619 ARTICLES © 2006 Nature Publishing Group http://www.nature.com/nsmb