Insulin-like Growth Factor-1 Impacts p53 Target Gene Induction in UVB-irradiated Keratinocytes and Human Skin

Abstract

The tumor suppressor protein p53 limits mutagenesis in response to ultraviolet-B (UVB) light exposure by activating the transcription of genes that mitigate the damaging effects of UVB radiation on DNA. Because most nonmelanoma skin cancers (NMSCs) occur in older individuals, it is important to understand the process of mutagenesis in the geriatric skin microenvironment. Based on previous studies demonstrating that geriatric skin expresses lower levels of the growth factor insulin-like growth factor-1 (IGF-1) than young adult skin, a role for IGF-1 in the regulation of p53 target genes was investigated in both human keratinocytes in vitro and human skin explants ex vivo. The products of the p53 target genes p21 and DNA polymerase eta (pol η) were found to be increased by UVB exposure in both experimental systems, and this induction was observed to be partially abrogated by depriving keratinocytes of IGF-1 in vitro or by the treatment of keratinocytes in vitro and human skin explants with an IGF-1 receptor antagonist. Because p21 and pol η function to limit mutagenic DNA replication following UVB exposure, these results suggest that NMSC risk in geriatric populations may be due to age-dependent decreases in IGF-1 signaling that disrupt p53 function in the skin.

Figure 1.

IGF-1 is required for the induction of p53 target genes in UVB-irradiated human keratinocytes in vitro. (a) In response to UVB radiation, p53 transactivates the expression of genes involved in slowing cell cycle progression, nucleotide excision repair and translesion synthesis across UVB-induced thymine dimers. (b) p53 phosphorylation on Ser15 was monitored by protein immunoblotting at the indicated time points after exposure to 100 J m⁻² in N-TERT cells grown in the presence (+ IGF-1) or absence (− IGF-1) of IGF-1. (c) Cells were treated as in (b) except that cells were harvested 24 h after UVB to examine protein induction. (d) Quantitation (average and SEM) of relative protein levels (+UVB/−UVB) from several independent experiments (n = 4–11, as indicated) performed as in (b). (e) Quantitative reverse transcription PCR analysis of pol η and p21 RNA levels 12 h after UVB exposure in cells treated as in (b) (n = 3). Unpaired, two-tailed t-tests were performed to determine whether the difference in relative gene product levels (+UVB/−UVB) was significant between the two treatment groups, and the p-values are indicated (n.s., not significant).

Figure 2.

Characterization of IGF-1 function in pol η and p21 induction in UVB-irradiated keratinocytes. (a) N-TERTs were treated as in Fig. 1 except that cells were exposed to the indicated doses of UVB radiation and harvested 24 h later. (b) Quantitation of protein levels from at least three independent experiments as shown in (a). A multiple t-test analysis was carried out to determine whether the difference between treatment groups was significant at each UVB dose. Asterisks indicate a significant difference (P < 0.05). (c) N-TERTs were cultured in the presence or absence of IGF-1, exposed to 100 J m⁻² UVB and then harvested at the indicated time points for protein immunoblotting. (d) N-TERTs were exposed twice (x2) to 100 J m⁻² UVB 24 h and 48 h after beginning incubation in medium containing or lacking IGF-1. Cells were harvested 6 h after the second UVB exposure to examine p53 phosphorylation on Ser15 and 24 h after the second UVB exposure to examine pol η and p21 induction. The graphs show the relative level of protein induction from two independent experiments. One-tailed t-tests were used to determine whether the relative protein induction in cells grown with IGF-1 was higher than in the absence of IGF-1.

Figure 3.

Inhibition of IGF-1R and p53 abrogates the UVB-dependent induction of pol η and p21 in keratinocytes in vitro. (a) N-TERT keratinocytes were treated with vehicle (DMSO) or IGF-1R inhibitor (IGF-1Ri, 10 μm AG538) for 30 min before exposure to 100 J m⁻² UVB. Cells were harvested 24 h later for protein immunoblot analysis. (b) Cells were treated as in (a) except that the p53 inhibitor (p53i; 20 μm pifithrin-α) was used. The relative level of protein induction was determined from 4 to 6 independent experiments, as indicated. Unpaired, two-tailed t-tests were performed to determine whether the difference in relative gene product levels (+UVB/−UVB) was significant between the two treatment groups.

Figure 4.

The loss of IGF-1 signaling impacts clonogenic growth potential in response to a low dose of UVB radiation. (a) Cells were treated as described in Fig. 1 except that cells were fixed in ethanol, stained with propidium iodide and analyzed by flow cytometry for DNA content. (b) Cells were treated as in (a) but were pulsed with EdU and processed with a Click-iT EdU staining kit to monitor S phase cells by fluorescence microscopy (n = 4–5 images per condition). (c) Cells were treated as described in (a) except that MTT assays were performed 24 h after UVB exposure. The results show the relative level of survival from 3 independent experiments. (d) Clonogenic survival assays were carried out with N-TERTs grown in the presence or absence of IGF-1 and exposed or not to 100 J m⁻² UVB. The results show the fraction of cells that formed colonies in 2–3 independent experiments. One-tailed t-tests were used to determine whether the percentage of EdU-positive cells or surviving cells was higher after UVB exposure or in the absence of IGF-1 (n.s., not significant).

Figure 5.

IGF-1 signaling is required for p21 and pol η induction in UVB-irradiated human skin epidermis ex vivo. (a) RT-qPCR was used to determine IGF-1 mRNA expression in skin samples (n = 39) discarded during routine surgeries. IGF-1 was normalized to B2M and then compared to a set of B2M standards of known copy number. A linear regression was performed, and the line of best-fit is plotted along with the 95% confidence bands. (b) Human skin explants were treated topically with either DMSO or 20 μm AG538 (IGF-1Ri) for 30 min before exposure to 800 J m⁻² UVB. Epidermis was isolated 24 h later for RNA analysis by RT-qPCR. The graphs show the induction of pol η and p21 in UVB-irradiated skin relative to nonirradiated skin from 4 different skin donors (36–62 years of age). Unpaired, two-tailed t-tests were performed to determine whether the difference in relative gene product levels (+UVB/−UVB) was significant. (c) A representative experiment in which skin was treated as in (a) except that the epidermis was analyzed by protein immunoblotting. (d) Quantitation of 7 independent experiments performed as in (c). For each experiment, the level of p21 protein (normalized to actin) in the DMSO + UVB sample was set to an arbitrary value of 100. All other signals were then compared to this value. A one-sample t-test was used to compare the treatment values to the DMSO + UVB control mean set to 100.