5-Formylcytosine mediated DNA-peptide cross-link induces predominantly semi-targeted mutations in both Escherichia coli and human cells

Abstract

Histone proteins can become trapped on DNA in the presence of 5-formylcytosine (5fC) to form toxic DNA-protein conjugates. Their repair may involve proteolytic digestion resulting in DNA-peptide cross-links (DpCs). Here, we have investigated replication of a model DpC comprised of an 11-mer peptide (NH₂-GGGKGLGK∗GGA) containing an oxy-lysine residue (K∗) conjugated to 5fC in DNA. Both CXG and CXT (where X = 5fC-DpC) sequence contexts were examined. Replication of both constructs gave low viability (<10%) in Escherichia coli, whereas TLS efficiency was high (72%) in HEK 293T cells. In E. coli, the DpC was bypassed largely error-free, inducing only 2 to 3% mutations, which increased to 4 to 5% with SOS. For both sequences, semi-targeted mutations were dominant, and for CXG, the predominant mutations were G→T and G→C at the 3'-base to the 5fC-DpC. In HEK 293T cells, 7 to 9% mutations occurred, and the dominant mutations were the semi-targeted G → T for CXG and T → G for CXT. These mutations were reduced drastically in cells deficient in hPol η, hPol ι or hPol ζ, suggesting a role of these TLS polymerases in mutagenic TLS. Steady-state kinetics studies using hPol η confirmed that this polymerase induces G → T and T → G transversions at the base immediately 3' to the DpC. This study reveals a unique replication pattern of 5fC-conjugated DpCs, which are bypassed largely error-free in both E. coli and human cells and induce mostly semi-targeted mutations at the 3' position to the lesion.

Keywords: 5-formylcytosine; DNA polymerase; DNA replication; DNA-protein crosslinks; mutagenesis; peptides; protein-DNA interaction; translesion synthesis.

Figure 1

Schematic diagram forconstruction of a single DpC-containing pMS2 vector anditsreplication in human cells.

Figure 2

TLS efficiency of 5fC-linked DpC CXG and CXT in E. coli AB 1157 cells with or without SOS relative to control. The data represent the mean and standard deviation from four independent replication experiments. Black dots represent the individual data points; error bars represent the standard deviation. The statistical significance of the difference in % viability between –SOS and +SOS cells was calculated using two-tailed, unpaired Student’s t test (∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001).

Figure 3

Types and frequencies of mutations induced by 5fC-conjugated DpCs in CXGand CXTsequences in E. coli cells without and with SOS.A and B, CXG. C and D, CXT.

Figure 4

TLS efficiencies of 5fC-linked DpCin HEK 293T cells with or without knockout of various TLS polymerases relative to control. The data in panel A (CXG) and panel B (CXT) represent the mean and standard deviation from three to four independent replication experiments. Black dots represent the individual data points; error bars represent the standard deviation. The statistical significance of the difference in % TLS efficiency between HEK 293T and polymerase-deficient cells was calculated using two-tailed, unpaired Student’s t test (∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001).

Figure 5

Mutation frequencies of 5fC-linked DpCin HEK 293T cells with or without knockout of various TLS polymerases. The data in panel A (CXG) and panel B (CXT) represent the mean and standard deviation from three to four independent replication experiments. Black dots represent the individual data points; error bars represent the standard deviation. The statistical significance of the difference in % mutation frequency between HEK 293T and polymerase(s)-deficient cells was calculated using two-tailed, unpaired Student’s t test (∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001).

Figure 6

Types and frequencies of mutations induced in the progeny from the 5fC-linked DpC in CXG sequence in HEK 293T cellsand various polymerase knockout HEK 293T cells. The data in panel A (CXG sequence in HEK 293T cells) and panels B–F (various polymerase knockout HEK 293T cells) represent the mean and the standard deviation from 3 to 6 independent experiments.

Figure 7

The types and frequencies of mutations induced in the progeny from the 5fC-linked DpC in CXT sequence in HEK 293T cellsand various polymerase knockout HEK 293T cells. The data in panel A (CXG sequence in HEK 293T cells) and panels B–F (various polymerase knockout HEK 293T cells) represent the mean and the standard deviation from 3 to 6 independent experiments.

Figure 8

Single-nucleotide insertion assays for replication across 3ʹ position ofthe 5fC-linked DpC. The assays in two DNA sequence contexts by hPol ƞ (A and C) and hPol ɩ (B and D) were performed with primer-template duplex containing unmodified dC or 5fC-conjugated 11-mer peptide cross-link. The reactions with hPol ƞ were quenched at 30 min and those with hPol ɩ were quenched at 60 min.

Figure 9

Schematic representation ofmutagenesis assay.