Psoralen-induced DNA interstrand cross-links block transcription and induce p53 in an ataxia-telangiectasia and rad3-related-dependent manner

Abstract

Psoralen plus UVA light (PUVA) is commonly used to treat psoriasis, a common skin disorder associated with rapid proliferation of cells. PUVA exerts its antiproliferative activity through formation of DNA monoadducts and interstrand cross-links (ICLs). However, this treatment may lead to skin malignancies as a direct result of inducing carcinogenic DNA damage. Inactivation of the p53 tumor suppressor gene is an important event in the development of skin cancer. p53 is rapidly phosphorylated and stabilized in response to DNA damage, and the induction of apoptosis by p53 is an important mechanism by which p53 exerts its tumor-suppressive activity. To better understand the mechanism by which PUVA treatment induces p53, we exposed human skin fibroblasts with PUVA under conditions that differentially produce monoadducts and ICLs and found that psoralen-induced ICLs induced phosphorylation of the Ser-15 site of p53 and apoptosis much more effectively than psoralen-induced monoadducts. The induction of p53 phosphorylation by psoralen ICLs did not require factors believed to be involved in the repair of psoralen ICLs [xeroderma pigmentosum (XP)-A, XP-C, XP-F, Cockayne's syndrome-B, Fanconi anemia] but did require the ataxia-telangiectasia and Rad3-related but not the ataxia-telangiectasia mutated kinase. Psoralen-induced ICLs blocked transcription and replication more efficiently than monoadducts, and induction of p53 and apoptosis correlated with doses causing interference with transcription rather than DNA replication. Our finding that cells underwent apoptosis preferentially during S-phase suggests that the combined blockade of transcription and DNA replication by psoralen ICLs during S-phase elicits a strong apoptotic response.

Fig. 1.

Induction of Ser-15 phosphorylation and p53 protein levels by PUVA. A, human diploid fibroblasts were treated with HMT (1 μg/ml = 3.8 μM) and irradiated with 30 kJ/m² with UVA (10 min). Cells were incubated in fresh media at 37°C and harvested at different times, and the levels of total p53 protein and Ser-15 phosphorylated p53 were determined by Western blotting. β-Actin was used as a loading control. B, human diploid fibroblasts were treated with different doses of HMT followed by UVA irradiation (30 kJ/m²). Cells were collected 6 h later, and the levels of total p53 protein and Ser-15 phosphorylated p53 were determined by Western blotting. After exposure to film, the membranes were stained with Coomassie blue stain to visualize total protein loading and transfer in the different lanes. C, human diploid fibroblasts were treated with HMT (1 μg/ml = 3.8 μM) and irradiated with different doses of UVA. Cells were harvested 6 h later, and the levels of total p53 protein and Ser-15 phosphorylated p53 were determined by Western blotting. Note that neither HMT nor UVA (45 kJ/m²) alone induced p53 or phosphorylation of p53, whereas HMT with the higher exposure of UVA light induced high levels of both. D, diploid fibroblasts derived from a patient with xeroderma pigmentosum complementation group A (XP-A) were treated as in B. E, diploid fibroblasts from various DNA repair-deficiency syndromes were treated as described in B and analyzed for Ser-15 phosphorylation. F, diploid normal, XP-F, and CS-B fibroblasts were treated with 100 ng/ml (380 nM) HMT and 30 kJ/m2 UVA and incubated for 24 h before the cellular levels of p53 and Ser-15 phosphorylated p53 were determined by Western blot. ^*, a nonspecific band was used to assess loading.

Fig. 2.

PUVA-induced ICLs induce a stronger phosphorylation of p53 and H2AX than psoralen monoadducts. A, human diploid fibroblasts were treated with HMT (0.5 or 1 μg/ml) and irradiated with 500 J/m² (low; lanes 2-4) or 30 kJ/m² UVA (high; lane 5). Two samples (lanes 6 and 7) were washed after exposure to the low 500 J/m² UVA dose to remove unbound HMT, and the cells were reirradiated with 30 kJ/m² UVA to convert monoadducts into ICLs. Cells in lane 8 were irradiated with 30 kJ/m² in the presence of 0.5 μg/ml HMT. After UVA irradiation, cells were incubated at 37°C for 6 h before cells were harvested and levels of Ser-15 phosphorylation (top) and γH2AX (bottom) were determined by Western blot. The top bands denoted with ^* are nonspecific bands that serve as surrogate loading control. B, the same as in A, but angelicin was used instead of HMT. 0.5 μg/ml HMT = 1.9 μM; 0.5 μg/ml angelicin = 2.7 μM.

Fig. 3.

PUVA-induced ICLs inhibit transcription preferentially compared with psoralen monoadducts. A, diploid human fibroblasts prelabeled with [¹⁴C]thymidine were treated with different doses of HMT followed by UVA (30 kJ/m²). The cells were then incubated for 1 h in fresh media followed by incubation with for 1 h in the presence of [³H]uridine to label nascent RNA. The ratios of ³H/¹⁴Cin total RNA or isolated mRNA of the different samples are expressed relative to mock-treated control cells (=100%). B, same as in A, but cells were treated with 1 μg/ml (3.8 μM) HMT followed by irradiation with different doses of UVA. C, ICLs induce a stronger inhibition of RNA synthesis than monoadducts. Diploid human fibroblasts were treated with 1 μg/ml HMT or angelicin alone or with combinations of 500 J/m² UVA (low) and/or 30 kJ/m² UVA (high) with a washout in between irradiations. Total nascent RNA synthesis values were compared with untreated control cells as described in A. D, ICLs inhibit mRNA synthesis. Cells were treated with 1 μg/ml HMT and irradiated with 500 J/m² UVA and/or 30 kJ/m² UVA with a washout in between, and nascent mRNA synthesis was determined as in A. Error bars show S.D.; ^*, p < 0.05; ^**, p < 0.001.

Fig. 4.

The ATR kinase is required for Ser-15 phosphorylation after PUVA treatment. A, diploid human fibroblasts were pretreated with different concentrations of wortmannin for 30 min followed by treatment with HMT (1 μg/ml) and UVA (30 kJ/m²). Cells were incubated for 6 h in the presence of wortmannin before the cells were harvested, and Ser-15 phosphorylation of p53 was analyzed by Western blot. β-Actin was used as loading control. B, diploid AT fibroblasts were treated with different concentrations of HMT followed by UVA light (30 kJ/m²). Cells were harvested 6 h later, and the levels of total p53 protein and Ser-15 phosphorylated p53 were determined by Western blotting. C, human diploid fibroblasts were microinjected into the nucleus with either IgG or anti-ATR antibodies. After 1-h incubation, the cells were treated with HMT (1 μg/ml) and UVA (30 kJ/m²), and 2 h later, the cells were fixed and stained for Ser-15 phosphorylation of p53. Arrows indicate microinjected cells. D, quantification of the percentage of cells staining positive for phosphorylated Ser-15 of p53 treated as in C. The total numbers of injected cells were 100 and 102 for IgG and anti-ATR, respectively. The error bars represent the S.D. of results from at least 5 different experiments with approximately 20 microinjected cells for each.

Fig. 5.

PUVA-induced ICLs induce apoptosis preferentially in cells progressing through S-phase. A, diploid human fibroblasts were treated with different concentrations of HMT followed by UVA (30 kJ/m²). Cells were harvested 72 h later, and the percentages of cells undergoing apoptosis were determined by the sub-G₁ fraction of cells detected using flow cytometry. Note that after a dose of 10 ng/ml, cells are accumulating in the S- and G₂/M-phases of the cell cycle without inducing any significant amounts of apoptosis. B, diploid fibroblasts were treated as described in Fig. 2A and were harvested 72 h later, and the percentages of cells undergoing apoptosis were determined by the sub-G₁ fraction of cells detected using flow cytometry. Note that the cell population reirradiated with 30 kJ/m² UVA to convert monoadducts into ICLs induced 13% of apoptotic cells compared with 3% in the cell population that was only exposed to 0.5 kJ/m² of UVA. C, diploid human fibroblasts pretreated with BrdU for 15 min were treated with different concentrations of HMT and UVA (30 kJ/m²). Cells were then incubated in the presence of BrdU and harvested 72 h later, and the percentages of BrdU-containing cells undergoing apoptosis (sub-G₁ fraction) were assessed. It was found that 63 and 68% of the apoptotic cells detected after treatment with 100 and 1000 ng/ml, respectively, stained positive for BrdU.