Prostaglandin E2 regulates senescence and post-senescence neoplastic escape in primary human keratinocytes

Abstract

Aging of the epidermis partially occurs as a consequence of epidermal cell senescence, a non-proliferative state in which cells remain metabolically active and acquire changes in their secretome. We previously reported that senescent normal human epidermal keratinocytes (NHEKs) have two opposite outcomes: either cell death by excess of autophagic activity or escape from senescence to give rise to post-senescence neoplastic emerging (PSNE) cells. In this study, we investigated the role of PTGS2, the inducible enzyme of the prostaglandin biosynthesis pathway, in the onset of NHEK senescence and in the switch from senescence to pre-transformation. We provide evidence that the PTGS2/PGE₂/EP4 pathway plays a critical role in NHEK senescence as well as in senescence escape. We show that treating proliferating NHEKs with prostaglandin E₂ (PGE₂) or with an agonist of one of its receptors, EP4, induced the establishment of the senescent phenotype, according to several markers including the senescence-associated β-galactosidase activity. Conversely, treating already senescent NHEKs with an antagonist of EP4, or knocking-down PTGS2 by siRNA resulted in the decrease of the percentage of senescence-associated β-galactosidase-positive cells. We also demonstrate that the PSNE frequency was significantly decreased upon PTGS2 silencing by siRNA, pharmacological PTGS2 inhibition, or treatment by an EP4 antagonist, while on the contrary treatments with PGE₂ or EP4 agonist increased the PSNE frequency. These results indicate that the PTGS2/PGE₂/EP4 pathway is required to induce and maintain the senescent phenotype of NHEKs, and that PGE₂ level is a potential determinant of the initial steps of the age-related oncogenic process.

Keywords: EP receptors; PTGS2; keratinocyte; neoplastic transformation; prostaglandins; senescence.

Figure 1

PTGS2 is up-regulated during NHEK in vitro senescence and skin aging. (A) The mRNA levels of PTGS2 were measured by RT-qPCR in extracts from exponentially growing, pre-senescent and senescent NHEKs (donor 4F0315). PTGS2 levels were normalized to EAR levels. The bars represent the mean ±SD (p < 0.05; ^**p < 0.01). (B) PTGS2, MnSOD, PCNA, and GPX4 protein levels were evaluated by Western Blot in extracts from exponentially growing, pre-senescent and senescent NHEKs (upper panel: donor 4F0315; lower panel: donor K40FH1). The gel was equicharged with extracts from an equal number of cells. The equicharge was verified a posteriori by detecting the levels of actin and GAPDH. (C) Exponentially growing and senescent NHEKs (donor K3MC1) were fixed and processed for immunofluorescence detection of endogenous proteins PTGS2 (red) and XRCC1 (green). Cell nuclei were detected by DAPI staining (blue). Upper panel: Representative confocal photomicrographs of PTGS2 inmmunostaining and XRCC1 foci (white arrows). Bars represent 20 µm. Lower panel: XRCC1 foci were quantified. Measures were done in five independent microscopic fields for a total of at least 100 cells for each condition. The histogram represents the average ± S.D. of five counts. Results are representative of at least two independent experiments. (D) Immunofluorescence detection of PTGS2 (red) performed in sections of skin samples from human young (n = 4) and old (n = 5) healthy subjects (see Material and Method) (^*p < 0.05). Cell nuclei were detected by DAPI staining (blue). Upper panel: representative confocal microscopy images for epidermis and dermis of a young (37 years old) and an elderly donor (85 years old). The squares delimit the below images at higher magnification. Bars represent 40 µm. Lower panel: scatter dot plots indicating the mean fluorescence intensity in cells of the basal layer. A minimum of 25 cells per sample were selected individually in order to obtain a mean fluorescence for each donor. The horizontal black lines denote median values and the boxes the interquartile ranges (IQR). Vertical lines extend from max value (upper quartile + 1.5*IQR) to min value (lower quartile – 1.5*IQR).

Figure 2

PTGS2 induces and maintains NHEK senescence. (A) NHEKs (donor 4F0315) were treated with NS398 at 5 or 10 µM or DMSO every 48 h. Left panel: During the treatment, cells were passaged when reaching 70% confluence, counted, and the number of population doublings was calculated (see Methods). The experiment was performed in triplicate, each point representing the mean of three counts. Significant differences between DMSO control and 5 or 10 µM NS398 treatment are indicated. Right panel: Senescent NHEKs were treated as in (A), then, a SA-β-Gal assay was performed seven days post-treatments. (B) Senescent NHEKs (donor K23FC1) were transfected with a pool of 4 siRNAs targeting PTGS2, or with non-target siRNAs. Left panel: Evaluation of the efficacy of the siRNAs by Western Blot. GAPDH was used as a loading control. Right panel: Four days after transfection, a SA-β-Gal assay was performed. The bars represent the mean ±SD of three counts of blue cells (^**p < 0.01). (C) Senescent NHEKs (donor K23FC1) were treated with NS398 or rofecoxib at the indicated concentrations for four days. Then, a SA-β-Gal assay was performed. The bars represent the mean ±SD of three counts of blue cells (^*p < 0.05; ^**p < 0.01). (D) ELISA assays for measuring the amounts of GM-CSF and G-CSF in the conditioned media (secreted) of exponentially growing, pre-senescent and senescent NHEKs (donor K40FH1) treated or not with NS398 (5 µM) for 16 hrs. Measures were performed in triplicate. The bars represent the mean ± SD of three counts. Significant differences are indicated with asterisks with ^*p < 0.05; ^**p < 0.01; ^***p < 0.001.

Figure 3

PGE₂ contributes to the establishment and maintenance of senescence in NHEKs. (A) PTGS2, PTGES1, PTDGS, and PGFS mRNA levels were measured by RT-qPCR in NHEKs (donor K67FA1) at the exponential growth phase or at the senescence plateau and were normalized to EAR levels. Results are presented in log 10 scale relative to PTGS2 expression in exponentially growing NHEKs (^*p < 0.05; ^***p < 0.001). (B) The amount of PGE₂ and PGF₂ in the culture media (secreted) of exponentially growing or senescent NHEKs (donor K67FA1) were measured by a competitive assay (see Material and Methods). Measures were performed in triplicate. The bars represent means ± SD. (C) Senescent NHEKs (donor K40FH1) were transfected with a pool of 4 siRNAs targeting PTGES1 or PGFS, or with non-target siRNAs (siCtrl). Four days after transfection, a SA-β-Gal assay was performed. The bars represent the mean ± SD of three counts (^*p < 0.05; ^***p < 0.001). (D) The amounts of IL-10, GM-CSF and G-CSF in the conditioned media (secreted) were measured by ELISA assays in NHEKs (donor K40FH1) treated as in (C). Measures were performed in triplicate. The bars represent the mean ± SD. Significant differences are indicated with asterisks with ^*p < 0.05. (E) Pre-senescent NHEKs (donor K23FC1) were treated or not with 10 nM to 10 µM PGE₂ 4-times a day. The percentage of SA-β-Gal positive cells 4 day after the beginning of the treatment was determined. The bars represent the mean ± SD of three counts (^*p < 0.05; ^**p < 0.01). (F) Pre-senescent NHEKs (donor K23FC1) were treated as in (E), the number of cells was then determined. The bars represent the mean ± SD of three counts (^**p < 0.01).

Figure 4

EP4 receptors mediate senescence of epidermal keratinocytes. (A) The mRNA levels of EPs were detected by RT-PCR in extracts from exponentially growing and senescent NHEKs (donor 4F0315) and in NHDFs at the replicative senescence plateau. Actin levels were used as control. The number of population doublings for NHDFs and NHEKs are indicated. (B) Immunofluorescence detection of EP1 and EP4 (green) performed in sections of skin samples from human young and old healthy subjects (see Material and Method). Cell nuclei were detected by DAPI staining (blue). Representative confocal microscopy images of the epidermis of a young (37 years old) and an elderly donor (83 years old). Positive staining of EP1 and EP4 (white arrows) are shown. Bars represent 20 µm. (C) Scheme of PGE₂ receptors and their agonists/antagonists used in the following experiments. (D) NHEKs (donor K40FH1) at the exponential growth phase were treated with the EP1 or EP4 agonists (Iloprost, L-902,688 and Rivenprost, respectively at 100 ng/mL, 1 µM and 100 nM) for 4 days. The percentage of SA-β-Gal-positive cells was determined 4 days after the beginning of treatment. SA-β-Gal-positive cells were counted in at least 3 different microscopic fields. The bars represent the mean ± SD of at least 3 counts (^**p < 0.01). (E) Senescent NHEKs (donor K23FC1) were treated with EP1 and EP4 antagonists (AH6809 and L-161,982, respectively at 10 µM and 1 µM). The percentage of SA-β-Gal-positive cells was determined 4 days after the beginning of treatment. SA-β-Gal-positive cells were counted in at least 3 different microscopic fields. The bars represent the mean ± SD of at least 3 counts (^*p < 0.05). (F) NHEKs (donor K40FH1) at the exponential growth phase were treated as in (D), the amounts of GM-CSF in the conditioned media (secreted) were measured by an ELISA assay. Measures were performed in triplicate. The bars represent the mean ± SD. Significant differences are indicated with asterisks with ^*p < 0.05 and ^***p < 0.001. (G) NHEKs (donor K40FH1) at the exponential growth phase were treated as in (E), and the amount of GM-CSF in the conditioned media (secreted) was measured by an ELISA assay. Measures were performed in triplicate. The bars represent the mean ± SD. Significant differences are indicated with asterisks with ^*p < 0.05.

Figure 5

Loss of PTGS2 activity reduces preneoplastic senescence escape of NHEKs. (A) Left panel: Scheme depicting the experimental process. Right panel: representative phase contrast microscopy image of what was counted as a clone in the following experiments. (B) Senescent NHEKs (donor K67JA1) were subjected to PTGS2 siRNAs vs. non-target control siRNAs. After four- or six-days post-transfection, emerging clones were manually counted under microscopic examination after fixation and coloration with crystal violet. (C) Senescent NHEKs (donor K23C1) were subjected to PTGS2 pharmacological inhibition by NS398 or Rofecoxib at the indicated concentrations. Four- or six-days post-treatment, emerging clones were manually counted as in B. (D) Senescent NHEKs (donor KNBMC1) were treated or not with 10 or 100 nM PGE₂ 4-times a day. After four days of treatment, the number of emerging clones was counted as in B. (E) Senescent NHEKs (donor K40FH1) were treated with the EP1 or EP4 agonists (Iloprost and Rivenprost at the indicated concentrations). After four days of treatment, the number of emerging clones was counted as in B. (F) Senescent NHEKs (donor K40FH1) were treated with EP1 and EP4 antagonists (AH6809 and L-161,982, respectively at 10 µM and 1 µM). After four days of treatment, the number of emerging clones was counted as in B. In (B–F), the bar chart represents the mean ± SD of the 3 independent measures (^*p < 0.05; ^**p < 0.01). In (B–F), each result is representative from 3 independent experiments, with each experiment corresponding of three or four technical replicates.