UV-B radiation induces epithelial tumors in mice lacking DNA polymerase eta and mesenchymal tumors in mice deficient for DNA polymerase iota

Abstract

DNA polymerase eta (Pol eta) is the product of the Polh gene, which is responsible for the group variant of xeroderma pigmentosum, a rare inherited recessive disease which is characterized by susceptibility to sunlight-induced skin cancer. We recently reported in a study of Polh mutant mice that Pol eta is involved in the somatic hypermutation of immunoglobulin genes, but the cancer predisposition of Polh-/- mice has not been examined until very recently. Another translesion synthesis polymerase, Pol iota, a Pol eta paralog encoded by the Poli gene, is naturally deficient in the 129 mouse strain, and the function of Pol iota is enigmatic. Here, we generated Polh Poli double-deficient mice and compared the tumor susceptibility of them with Polh- or Poli-deficient animals under the same genetic background. While Pol iota deficiency does not influence the UV sensitivity of mouse fibroblasts irrespective of Polh genotype, Polh Poli double-deficient mice show slightly earlier onset of skin tumor formation. Intriguingly, histological diagnosis after chronic treatment with UV light reveals that Pol iota deficiency leads to the formation of mesenchymal tumors, such as sarcomas, that are not observed in Polh(-/-) mice. These results suggest the involvement of the Pol eta and Pol iota proteins in UV-induced skin carcinogenesis.

FIG. 1.

Targeted disruption of the mouse Polh gene. (A) Schematic representation of the targeting strategy for the mouse Polh locus. The coding exons are numbered and boxed. The positions of selected restriction sites are shown. (B) Southern blot analysis of EcoRI-digested (for the 3′ and neo probes) or BamHI-digested (for the 5′ probe) genomic DNA and Northern blot analysis of mRNA prepared from MEFs. The position of each probe is shown in panel A. For Northern blot analysis, hybridization with a glyceraldehyde-3-phosphate dehydrogenase (GAPDH)-specific probe was performed to normalize for RNA amounts and transfer efficiency.

FIG. 2.

Expression of the Pol η and Pol ι proteins in MEFs. (A) Expression of Pol η in MEFs was monitored with a polyclonal antibody (60 μg/lane). The positions of the intact Pol η protein expressed by the wild-type allele and the truncated protein derived from the mutated allele are indicated by the filled and open arrowheads, respectively. Purified six-His-tagged mouse Pol η (0.5 ng) was used as a size marker (lane 6). (B) Expression of the Pol ι protein in MEFs was monitored with a polyclonal antibody (60 μg/lane) after stripping anti-Pol η antibody from the membrane shown in panel A. The position of the intact Pol ι protein is indicated by the filled arrowhead. Purified six-His-tagged mouse Pol ι (0.5 ng) was used as a size marker (lane 7). (C) Schematic of the protein derived from the mutant Polh allele. Highly conserved regions shared by Y-family polymerases are indicated. Like the predicted Pol η protein in human XP2SA cells, the truncated Pol η protein from the 1.7-kb transcript (Fig. 1B) from the mutated allele lacks part of the highly conserved domain essential for TLS activity but has an extra 49 amino acids derived from the sequence of the neomycin resistance gene cassette on the carboxy terminus.

FIG. 3.

Effects of UV irradiation on cell proliferation. (A and B) MEFs were irradiated as indicated with UV-C. After irradiation, cells were cultured in the absence (A) or the presence (B) of 1 mM caffeine for 2 days. The number of proliferating cells was measured by [³H]thymidine incorporation. All experiments were performed in duplicate at least three times. Consistent results were obtained for different sets of experiments. Data are presented as mean survival rates ± standard deviations. wt, wild type.

FIG. 4.

Tumorigenesis induced by chronic treatment with UV-B. (A) Kaplan-Meier curves of mice free of skin tumors after chronic UV irradiation (2 kJ/m²/day). (B) Number of UV-induced skin tumors per animal. Bars represent the exposure period. wt, wild type.

FIG. 5.

Histopathological examination of UV-B-induced skin tumors from Polh Poli double knockout mice. (A) Left and right panels represent typical examples of normal skin histology and dysplasia observed in wild-type mice, respectively. (B and C) Typical examples of squamous cell carcinoma predominantly observed in Polh^−/− Poli^+/+ mice (B) and Polh^−/− Poli^−/− mice (C). (D) Soft tissue sarcoma found in a Polh^−/− Poli^−/− mouse. White arrows indicate squamous cell carcinomas adjacent to the sarcoma.

FIG. 6.

Incidence of skin tumors after chronic UV exposure in mice bearing various combinations of Polh and Poli alleles. Gray, dark gray, and white bars indicate Poli^+/+, Poli^+/−, and Poli^−/− mice, respectively. We have not examined Polh^+/+ Poli^+/− mice. (A) Incidence of epithelial skin tumors. *, values for Polh^−/− mice are statistically significant versus those of corresponding Polh^+/− mice (P = 0.0072 for Poli^+/+, P = 0.0013 for Poli^+/−, and P = 0.004 for Poli^−/−, respectively); **, value for Polh^+/− mice is statistically significant versus that of Polh^+/+ mice (P = 0.0204). (B) Incidence of mesenchymal skin tumors. *, value for Poli^−/− mice is statistically significant versus that of Poli^+/− mice (P = 0.0106); **, value for Poli^−/− mice is statistically significant versus that of Poli^+/+ mice (P = 0.0025).