Structures of DNA polymerase beta with active-site mismatches suggest a transient abasic site intermediate during misincorporation

Abstract

We report the crystallographic structures of DNA polymerase beta with dG-dAMPCPP and dC-dAMPCPP mismatches in the active site. These premutagenic structures were obtained with a nonhydrolyzable incoming nucleotide analog, dAMPCPP, and Mn(2+). Substituting Mn(2+) for Mg(2+) significantly decreases the fidelity of DNA synthesis. The structures reveal that the enzyme is in a closed conformation like that observed with a matched Watson-Crick base pair. The incorrect dAMPCPP binds in a conformation identical to that observed with the correct nucleotide. To accommodate the incorrect nucleotide and closed protein conformation, the template strand in the vicinity of the active site has shifted upstream over 3 A, removing the coding base from the active site and generating an abasic templating pocket. The primer terminus rotates as its complementary template base is repositioned. This rotation moves O3' of the primer terminus away from the alpha-phosphate of the incoming nucleotide, thereby deterring misincorporation.

Figure 1. Closed Conformation of the Ternary Substrate Complex with an Active Site Mismatched Nascent Base Pair

(A) Licorice representation of the Pol β backbone of the binary DNA complex (PDB ID 1BPX; orange) and ternary substrate complex (green) with an incorrect incoming nucleotide (dAMPCPP; yellow carbons) with a templating guanine. The catalytic and DNA-binding subdomains superimposed (gray backbone) with an rmsd of 0.56 Å (177 Cα). The DNA is omitted for clarity, but the 5′—3′ direction of the primer entry into the active site is indicated with a solid arrow. The open and closed position of α-helix N is shown. The amino-terminal lyase domain and carboxyl-terminal N—subdomain (colored) move in response to binding an incorrect dNTP. The amino– (N) and carboxyl-terminal (C) ends are labeled. (B) Ribbon representation of the Pol β backbone of the ternary substrate complex with correct (gray) or incorrect (green) incoming nucleotides. The polymerase domain with a correct (dA—dUMPNPP; PDB ID 2FMS) and incorrect (dG—dAMPCPP) nascent base pair were superimposed with an rmsd of 0.62 Å (314 Cα). The superimposed structures indicate that α-helix N is in a ‘closed’ position like that observed with a Watson-Crick nascent base pair. The incoming dAMPCPP of the mismatch structure is shown (yellow carbons), but the DNA is omitted for clarity. The amino-terminus (N) of the lyase domain is also indicated.

Figure 2. Nucleotides of the Mismatch are in a Staggered Conformation

(A) The carbons of the mismatch (dG—dAMPCPP) are colored yellow and those of the primer terminus and α-helix N are green. The coding template base (n) is shifted upstream 3.2 Å while the incoming nucleotide is positioned appropriately in the dNTP binding pocket as illustrated by comparing the relative positions of the respective nucleotides in the ternary complex structure with a matched base pair (gray, PDB ID 2FMS). The superimposed structures indicate that α-helix N is in a ‘closed’ position. The primer terminus and its templating base (n-1) of the mismatched structure are shown as well as the backbone of the templating strand of both structures. A hydrogen bond (dashed green line) may occur between N6(dAMPCPP) and O6(dG). The atoms that directly participate in catalysis are labeled (O3’, Pα, and Mn²⁺). (B) The carbons of the mismatch (dC—dAMPCPP) are colored yellow and those of the primer terminus and α-helix N are light blue. In this case, a possible hydrogen bond (dashed blue line) can form between the templating cytosine(N4) and N3 of the primer terminal cytosine.

Figure 3. Primer Terminus is Re-positioned as a Consequence of Template Slippage

(A) The primer terminus (n-1) phosphates of the matched (gray) and dG—dAMPCPP mismatched (green) ternary complexes are in similar positions as illustrated by the ribbon representation, but the primer base and dexoxyribose of the mismatched complex are displaced from that observed in the matched ternary complex. This separates O3’ in these structures 2.7 Å, so that O3’ in the mismatch structure is now 3.9 Å from Pα of the incoming nucleotide. More importantly however, is that O3’ no longer participates in coordination of the catalytic metal and is no longer in-line with the bridging oxygen of Pα and Pβ of the incoming nucleotide. The superimposed structures also reveal that the divalent metals (Mn²⁺ or Mg²⁺, orange and grayspheres respectively) occupy identical positions in both structures. The carbons of the mismatch (dG—dAMPCPP) are colored yellow. (B) The catalytic Mn²⁺ is octahedrally coordinated with oxygens from three active site aspartates (D190, D192, D256), two waters, and a non-bridging oxygen (pro-R_P) on Pα of the incoming incorrect nucleotide. In contrast to when a correct nucleotide binds, the primer terminus O3’ is too far from the catalytic metal to participate in the inner coordination sphere. Instead, a water molecule (indicated) now occupies the position where O3’ (semi-transparent) is observed with a correct incoming nucleotide. The sugar of the primer terminus in the ternary complex structure with a correct incoming nucleotide is shown (gray). (C) A similar view as that illustrated in panel B for the dC—dAMPCPP active site. In this case, the carbons of the active site carboxylates and primer terminus are light blue. A similarly positioned water is also observed in the mismatch structure where O3’ of a correctly positioned primer terminus is observed (gray carbons).

Figure 4. Position of Key Protein Residues in Closed Polymerase Complexes with Correct or Incorrect Incoming Nucleotides

(A) Two views (major groove view, top; –90° rotation of the top view, bottom) of the nascent base pair of the mismatch structure (dG—dAMPCPP; green carbons) superimposed (see Figure 1) with that for a correctly matched base pair (gray carbons). The templating base is omitted from the correctly matched overlay for clarity. The structure illustrates the position of key protein side chains that can influence catalytic behavior. For catalytic activation with the correct incoming nucleotide, N–subdomain closing is associated with the loss of a salt bridge between Arg258 (R258) and Asp192 (D192), that coordinate both active site metals (M²⁺), and the formation of hydrogen bonds (black dashed lines) with Glu295 (E295) and Tyr296 (Y296). Phe272 (F272) is repositioned in the closed complex to insulate Asp192 from Arg258. Arg283 (R283) that is situated in the N–subdomain interacts with the minor groove edge of the templating strand (not shown). With an incorrect incoming nucleotide, R258 and R283 are in conformations that preclude catalytic activation. Arg283 is observed to hydrogen bond with the minor groove edge and phosphate backbone of the templating base. Other key residues (Asp192, Asn279, Phe272) are observed in similar positions as that found with a correct incoming nucleotide. (B) Major groove view of the nascent base pair of the dG/dC—dAMPCPP superimposed mismatch structures (dC template, light blue carbons; dG template, green).

Figure 5. Comparison of the Position of the Coding Templating Nucleotide in Various Polymerase Liganded States

Ternary complex structures of Pol β with a Watson-Crick base pair (gray, 2FMS) and dG— dAMPCPP mismatch (yellow) where superimposed with single-nucleotide gapped binary complexes of Pol β (orange) and Pol λ (green). The templating base (n) in each case is shown in a thick stick representation and the primer terminal base pair (n–1) is illustrated as thin sticks. The arrow indicates the upstream duplex. The correct incoming nucleotide is also shown base pairing (black dashed lines) with the templating adenine. Arg283 of Pol β (R283) fills the templating base binding pocket in the mismatch structure where the templating base has moved upstream producing an apparent abasic site. A similar position of the templating nucleotide and the conserved arginine (R517) in a Pol λ binary DNA complex is observed indicating that template slippage may be a general mechanism for active site de-activation.