FF483-484 motif of human Polη mediates its interaction with the POLD2 subunit of Polδ and contributes to DNA damage tolerance

Abstract

Switching between replicative and translesion synthesis (TLS) DNA polymerases are crucial events for the completion of genomic DNA synthesis when the replication machinery encounters lesions in the DNA template. In eukaryotes, the translesional DNA polymerase η (Polη) plays a central role for accurate bypass of cyclobutane pyrimidine dimers, the predominant DNA lesions induced by ultraviolet irradiation. Polη deficiency is responsible for a variant form of the Xeroderma pigmentosum (XPV) syndrome, characterized by a predisposition to skin cancer. Here, we show that the FF483-484 amino acids in the human Polη (designated F1 motif) are necessary for the interaction of this TLS polymerase with POLD2, the B subunit of the replicative DNA polymerase δ, both in vitro and in vivo. Mutating this motif impairs Polη function in the bypass of both an N-2-acetylaminofluorene adduct and a TT-CPD lesion in cellular extracts. By complementing XPV cells with different forms of Polη, we show that the F1 motif contributes to the progression of DNA synthesis and to the cell survival after UV irradiation. We propose that the integrity of the F1 motif of Polη, necessary for the Polη/POLD2 interaction, is required for the establishment of an efficient TLS complex.

Figure 1.

Polη interacts with the POLD2 subunit of Polδ. (A) Polη full length or truncation mutants and POLD2 proteins were expressed in the yeast strain AH109 as a transcription activation domain fusion protein (in pACT2) and a DNA binding domain fusion protein (in pGBKT7), respectively. Yeast transformants expressing both Polη and POLD2 fusion proteins are selected on double drop out medium (-W-L). Positive interactions are indicated by growth on quadruple drop out medium (-W-L-A-H). (B) Minimum amino acid sequence of the Polη region that interacts with POLD2. Residues that were mutated to alanine in the full-length Polη coding sequence and tested for their interaction with POLD2 are boxed. (C) Wild type and FF_483–484AA (Polη F1*) of full-length Polη were examined for the interaction with POLD2. (D) Association of POLD2 with Polη in vitro. Left panel: physical interaction between the purified human DNA polymerase η and the POLD2 subunit of DNA polymerase δ. GST pull-down experiment was carried out using Flag-POLD2 and GST-Polη or GST followed by immobilization on GTH beads. The bound proteins were analyzed by immunoblotting or Coomassie blue staining. Right panel: selective binding of POLD2 with His-GST-Polη_393–511. Pull-down experiments were carried out using in vitro translated POLD2, His-GST, His-GST-Polη_393–511 or His-GST-Polη_393–511 (F1*) followed by immobilization on IMAC magnetic beads. Input and bound proteins were analyzed by Coomassie blue staining or immunoblotting. Inputs correspond to 1/10th the protein amount used for pull-down.

Figure 2.

Wild-type or mutant Polη displays the same catalytic activities. Time course of DNA synthesis catalyzed by wild-type (WT) or mutant Polη using a primed single-stranded template (pUC118.ss). Upper panel: DNA products were subjected to electrophoresis on a 20% polyacrylamide–7-M urea denaturing gel. Lower panel: immunoblot of the different forms of Polη produced in rabbit reticulocytes lysates, using an anti-myc antibody.

Figure 3.

The F1 motif of Polη contributes to TLS in vitro through an AAF adduct and a TT-CPD lesion. (A) Outline of the experiment and diagram of the modified plasmids. The length of the strand produced upon elongation of the [32P]-labeled primer, up to the lesion site, is indicated. nts : nucleotides. (B) TLS efficiency through a G-AAF adduct located on the third guanine of a run of 3G. Monomodified DNA substrates (10 fmol) were incubated 20 min at 37°C in the presence of XPV cell-free extracts (20 μg) complemented with an equal amount (0.2 μl) of different forms of Polη produced in vitro. Samples were analyzed by electrophoresis through an 8% denaturing polyacrylamide gel. Product L-1 is generated when synthesis is blocked one nucleotide before the lesion. TLS0 and TLS-1 are TLS products through an AAF adduct. Below the gel: immunoblot of the different forms of Polη produced in rabbit reticulocytes lysates added to the reactions. (C) TLS efficiency through a TT-CPD lesion. Upper panel: monomodified DNA substrates (10 fmol) were incubated 10 min at 37°C in the presence of various amounts (0.33, 0.25 and 0.125 μl) of either Polη wild-type (WT) or Polη mutant (Polη F1*) mixed with XPV cell-free extracts (20 μg). Below the gel: immunoblot of the different forms of Polη produced in rabbit reticulocytes lysates added to the reactions. Lower panel: monomodified DNA substrates (10 fmol) were incubated 10 min at 37°C in the presence of XPV cell-free extracts (20 μg) complemented with the same amount (0.2 μl) of different forms of Polη produced in vitro. Below the gel: immunoblot of the different forms of Polη produced in rabbit reticulocytes lysates and present in the reactions. Right panel: quantitative analysis of TLS efficiency with the different versions of Polη. Error bars denote the standard deviation (SD) of at least two experiments performed with independent Polη samples (0.2 μl).

Figure 4.

The F1 and the PIP motifs of Polη both contribute to cell survival and cell cycle progression following UV irradiation. (A) Immunoblots showing the expression of ectopic Polη versions in the different XPV-complemented cell lines. POLD1 was used as a loading control. The values under the immunoblot represent the ratio of the intensity of Polη to POLD1 bands, relative to MRC5. (B) Clonogenic survival assay with XPV-derived cell lines expressing Polη wild type or mutant following UV irradiation (8 J/m²) and growth in 0.375-mM caffeine-containing medium (EV: empty vector). Error bars represent SD from at least three independent experiments. (C) Cell cycle profiles of XPV cells complemented with the indicated wild-type and mutant forms of Polη determined by flow cytometry after DNA staining with propidium iodide. (D) Representative images of CSK-extracted nuclei from MRC5 cells cotransfected with the indicated plasmids and UV-irradiated (8 J/m²). Scale bar = 10 μm. Representative images of one of the three independent PLA assays for Polη/POLD2 (E) or Polη/PCNA (F) interactions. XPV cells complemented with the indicated wild-type and mutant forms of Polη were either non-treated or UV-irradiated (20 J/m²) and processed after 6 h. CSK-extracted nuclei were labeled with DAPI (blue). PLA signals were observed as white dots. The percentage of positive cells (containing at least four fluorescent spots) was scored (more than 450 cells were examined for each condition in panel (E) and more than 300 cells in panel (F)). The mean values of two (panel (F)) or three independent experiments (panel (E)) ±SD are indicated.