O6-2'-Deoxyguanosine-butylene-O6-2'-deoxyguanosine DNA Interstrand Cross-Links Are Replication-Blocking and Mutagenic DNA Lesions

Abstract

DNA interstrand cross-links (ICLs) are cytotoxic DNA lesions derived from reactions of DNA with a number of anti-cancer reagents as well as endogenous bifunctional electrophiles. Deciphering the DNA repair mechanisms of ICLs is important for understanding the toxicity of DNA cross-linking agents and for developing effective chemotherapies. Previous research has focused on ICLs cross-linked with the N7 and N2 atoms of guanine as well as those formed at the N6 atom of adenine; however, little is known about the mutagenicity of O⁶-dG-derived ICLs. Although less abundant, O⁶-alkylated guanine DNA lesions are chemically stable and highly mutagenic. Here, O⁶-2'-deoxyguanosine-butylene-O⁶-2'-deoxyguanosine (O⁶-dG-C4-O⁶-dG) is designed as a chemically stable ICL, which can be induced by the action of bifunctional alkylating agents. We investigate the DNA replication-blocking and mutagenic properties of O⁶-dG-C4-O⁶-dG ICLs during an important step in ICL repair, translesion DNA synthesis (TLS). The model replicative DNA polymerase (pol) Sulfolobus solfataricus P2 DNA polymerase B1 (Dpo1) is able to incorporate a correct nucleotide opposite the cross-linked template guanine of ICLs with low efficiency and fidelity but cannot extend beyond the ICLs. Translesion synthesis by human pol κ is completely inhibited by O⁶-dG-C4-O⁶-dG ICLs. Moderate bypass activities are observed for human pol η and S. solfataricus P2 DNA polymerase IV (Dpo4). Among the pols tested, pol η exhibits the highest bypass activity; however, 70% of the bypass products are mutagenic containing substitutions or deletions. The increase in the size of unhooked repair intermediates elevates the frequency of deletion mutation. Lastly, the importance of pol η in O⁶-dG-derived ICL bypass is demonstrated using whole cell extracts of Xeroderma pigmentosum variant patient cells and those complemented with pol η. Together, this study provides the first set of biochemical evidence for the mutagenicity of O⁶-dG-derived ICLs.

Figure 1

Replication-dependent repair of DNA interstrand cross-links.

Figure 2

(A) Chemical structure of O⁶-dG-C4-O⁶-dG cross-link. (B) Primers and templates used in this study. T^dd and C^dd indicate dideoxynucleotides to prevent polymerization.

Figure 3

Primer-extension reactions catalyzed by Dpo1, Dpo4, pol κ, and pol η. (A) Extension reactions with a 13/27-mer duplex with or without ICL (running start assays). Arrows indicate full-length products. Major products terminated at the nucleotide prior to the lesion, at the lesion, or after the lesion are labeled as −1, 0, and +1. Assays were performed at 37 °C with 100 nM DNA duplex, dNTP mixtures at 100 μM each, 80 nM polymerase, 4% (v/v) glycerol, 5 mM DTT, 50 mM NaCl, 5 mM MgCl₂, and 100 μg mL⁻¹ bovine serum albumin in 50 mM Tris-HCl pH 7.4. (B) Quantification of extended products from 10 min reactions shown in (A). (C) Extension reactions with a 20/27-mer duplex with or without ICL (standing start assays). Reaction conditions were the same as described in (A).

Figure 4

Normalized catalytic efficiencies of nucleotide incorporations at the −1 position with an unmodified template, ICL-S, or ICL-L for Dpo1 (A), Dpo4 (B), pol κ (C), and pol η (D). Results were derived from steady-state kinetic data shown in Table S1, Supporting Information. The catalytic efficiency of the native base pair with an unmodified duplex is considered as 1 for each enzyme. Normalized catalytic efficiencies were obtained using respective catalytic efficiency (k_cat/K_m) divided by the catalytic efficiency of the correct nucleotide with an unmodified duplex. Assays were performed at 37 °C with 80–100 nM duplex DNA (with or without ICL), 0.5–20 nM DNA pol, varying concentrations of a single dNTP, 4% (v/v) glycerol, 5 mM DTT, 50 mM NaCl, 5 mM MgCl₂, and 100 μg mL⁻¹ BSA in 50 mM Tris-HCl (pH 7.4 at 37 °C). Changes in catalytic efficiency relative to a native base pair were calculated from (k_cat/K_m,dNTP)_{unmodified, dTTP}/(k_cat/K_m,dNTP)_{ICL, dTTP}, and are indicated as x-fold decrease. Changes less than 2-fold are considered as no change. Complete data are in Table S1.

Figure 5

Normalized catalytic efficiencies of nucleotide incorporations at the 0 position with an unmodified template, ICL-S, or ICL-L for Dpo1 (A), Dpo4 (B), pol κ (C), and pol η (D). Results were derived from steady-state kinetic data shown in Table S2, Supporting Information. Experimental conditions were the same as those indicated in Figure 4. Normalized catalytic efficiencies and x-fold decrease were calculated in the same way as described in Figure 4. Changes in catalytic efficiency relative to a native base pair were indicated as x-fold decrease. Complete data are in Table S2.

Figure 6

LC-MS sequencing of pol η-catalyzed ICL bypass products. (A) Extracted ion chromatograms of products with different nucleotides incorporated opposite the lesion (underlined), i.e., m/z 1395.9 with C, m/z 1407.9 with A, m/z 1403.2 with T, and m/z 1415.2 with G. Product m/z 1231.3 is with a one-nucleotide deletion. (B) Fragmentation mass spectra of the different species in (A). The observed product ions that match theoretical fragmented ions are labeled. Reactions contained 1 μM pol η, 2 μM primer–template complex; 2% (v/v) glycerol, physiological concentrations of dNTPs in the nucleus (10 μM dGTP, 40 μM dATP, 40 μM dCTP, and 40 μM dTTP), 5 mM DTT, 50 mM NaCl, 5 mM MgCl₂, and 50 μg mL⁻¹ BSA in a total volume of 50 μL.

Figure 7

(A) Western blot of whole cell extracts of XP-V cells complemented with empty vector or pol η-expressing vector. (B) Primer-extension reactions with whole cell extracts of XP-V cells complemented with empty vector or pol η-expressing vector. The substrate is an ICL-S-containing duplex. Reactions were performed with 40 nM DNA duplex, dNTPs at 100 μM each, 2 μg proteins from whole cell extract (2 μL), 5 mM DTT, 50 mM NaCl, 5 mM MgCl₂, and 100 μg mL⁻¹ BSA in 50 mM Tris-HCl pH 7.4.