NRF2 Shortage in Human Skin Fibroblasts Dysregulates Matrisome Gene Expression and Affects Collagen Fibrillogenesis.

NRF2 is a master regulator of the antioxidative response that was recently proposed as a potential regulator of extracellular matrix (ECM) gene expression. Fibroblasts are major ECM producers in all connective tissues, including the dermis. A better understanding of NRF2-mediated ECM regulation in skin fibroblasts is thus of great interest for skin homeostasis maintenance and aging protection. In this study, we investigate the impact of NRF2 downregulation on matrisome gene expression and ECM deposits in human primary dermal fibroblasts. RNA-sequencing‒based transcriptome analysis of NRF2 silenced dermal fibroblasts shows that ECM genes are the most regulated gene sets, highlighting the relevance of the NRF2-mediated matrisome program in these cells. Using complementary light and electron microscopy methods, we show that NRF2 deprivation in dermal fibroblasts results in reduced collagen I biosynthesis and impacts collagen fibril deposition. Moreover, we identify ZNF469, a putative transcriptional regulator of collagen biosynthesis, as a target of NRF2. Both ZNF469 silenced fibroblasts and fibroblasts derived from Brittle Corneal Syndrome patients carrying variants in ZNF469 gene show reduced collagen I gene expression. Our study shows that NRF2 orchestrates matrisome expression in human skin fibroblasts through direct or indirect transcriptional mechanisms that could be prioritized to target dermal ECM homeostasis in health and disease.

In vitro tools for photobiological testing: molecular responses to simulated solar UV of keratinocytes growing as monolayers or as part of reconstructed skin.

Epidermal keratinocytes are critical targets for UV-induced genotoxicity as their transformation by sunlight overexposure can lead to skin cancer such as basal cell carcinoma (BCC) and squamous cell carcinoma (SCC). Therefore, assessment of photoprotection should involve early markers associated with DNA photodamage. Here, the same normal human keratinocytes either in monoculture (KC) or in full thickness reconstructed skin (RS) were compared with respect to their response to simulated solar UV (SSUV) exposure. Irradiation conditions (spectral power distribution and doses) were designed to mimic environmental zenithal UV from sunlight. At doses where survival was higher than 80%, comet assay showed more single strand breaks (SSB) and cyclobutane pyrimidine dimers (CPD) in keratinocytes in RS than in KC one hour post-exposure. The transcription factor p53 was activated in both models. While in KC p53 accumulation displayed a linear dose-dependency up to 24 h post-exposure, in RS it followed a bell-shaped profile and reverted to its basal rate. QRT-PCR demonstrated that among genes controlled by p53, P21 and MDM2 were clearly induced by SSUV in KC, whereas GADD45 expression was strongly and almost exclusively up-regulated in RS. Nrf2-dependent antioxidant genes (Ferritin light chain, NQO1) were only induced in RS, yet at low doses for NQO1. In vitro models such as KC or RS allowing the development of quantitative methodologies should be used as surrogates for in vivo tests assessing photogenotoxicity.

Photo-pollution stress in skin: Traces of pollutants (PAH and particulate matter) impair redox homeostasis in keratinocytes exposed to UVA1.

BACKGROUND: It is likely that skin is exposed to low concentrations of pollutants such as Polycyclic Aromatic Hydrocarbons (PAH) either through topical penetration by ultrafine particles or by systemic distribution. No precise estimation of pollutants in living skin is available, but literature has reported contamination of blood by PAH at concentrations in the nanomolar range. Some pollutants (PAH for example) are photo-reactive and phototoxic: sunlight and pollution might thus synergistically compromise skin health. OBJECTIVE: Here, the biological effects of particulate matter, PM extract and various PAH were compared in normal human epidermal keratinocytes (NHEK) and reconstructed skin model exposed to either daily UV (d-UV 300-400nm) or UVA1 (350-400nm). Impact of pollutants (PM, PAH or PM extract) combined to UV was studied on NHEK by measuring toxicity, redox homeostasis and GSH metabolism in NHEK. METHODS: NHEK were exposed to UV from solar simulator (either d-UV or UVA1) combined with pollutants. Viability, clonogenic efficiency, redox homeostasis and GSH metabolism were assessed. RESULTS: Pollutants (PAH, PM or PM extract) ±UVA1 irradiation was associated with a significant phototoxic effect that was equal to or greater than that produced by d-UV. This result is interesting considering that UVA1 represents around 80% of daily UV and reaches the dermal-epidermal junction with ease. Moreover, among PAH studied, benzo[a]pyrene and indeno[1,2,3-cd]pyrene were phototoxic at very low concentrations (nanomolar range) on cultured cells or in reconstructed epidermis and also impaired keratinocyte clonogenic potential at sub-toxic doses. ROS generation within cells and in the inner mitochondrial compartment, mitochondrial membrane depolarization and/or reduced ATP production were also noted. Meanwhile, intracellular glutathione concentrations transiently decreased several hours post-treatment and reduction of its synthesis by buthionine sulfoximine potentiated PAH phototoxicity. Consequently, expression of GSH neo-synthesis genes such as SLC7A11 or GCLc was upregulated several hours post-treatment. CONCLUSION: These results obtained using PAH concentrations in the range of those reported in blood of pollution-exposed people suggest that exposure to such a photo-pollution stress, particularly if chronic, may impair cutaneous homeostasis and aggravate sunlight-induced skin damage.

Molecular responses to stress induced in normal human caucasian melanocytes in culture by exposure to simulated solar UV.

Melanocytes play a central role in the response of skin to sunlight exposure. They are directly involved in UV-induced pigmentation as a defense mechanism. However, their alteration can lead to melanoma, a process where the role of sun overexposure is highly probable. The transformation process whereby UV damage may result in melanoma initiation is poorly understood, especially in terms of UV-induced genotoxicity in pigmented cells, where melanin can act either as a sunscreen or as a photosensitizer. The aim of this study was to analyze the behavior of melanocytes from fair skin under irradiation mimicking environmental sunlight in terms of spectral power distribution. To do this, normal human Caucasian melanocytes in culture were exposed to simulated solar UV (SSUV, 300-400 nm). Even at relatively high doses (until 20 min exposure, corresponding to 12 kJ/m2 UV-B and 110 kJ/m2 UV-A), cell death was limited, as shown by cell viability and low occurrence of apoptosis (caspase-3 activation). Moreover, p53 accumulation was three times lower in melanocytes than in unpigmented cells such as fibroblasts after SSUV exposure. However, an important fraction of melanocyte population was arrested in G2-M phase, and this correlated well with a high induction level of the gene GADD45, 4 h after exposure. Among the genes involved in DNA repair, gene XPC was the most inducible because its expression increased more than two-fold 15 h after a 20 min exposure, whereas expression of P48 was only slightly increased. In addition, an early induction of Heme Oxygenase 1 (HO1) gene, a typical response to oxidative stress, was also observed for the first time in melanocytes. Interestingly, this induction remained significant when melanocytes were exposed to UV-A radiation only (320-400 nm), and stimulation of melanogenesis before irradiation further increased HO1 induction. These results were obtained with normal human cells after exposure to SSUV radiation, which mimicked natural sunlight. They provide new data related to gene expression and suggest that melanin in light skin could contribute to sunlight-induced genotoxicity and maybe to melanocyte transformation.

Molecular responses to photogenotoxic stress induced by the antibiotic lomefloxacin in human skin cells: from DNA damage to apoptosis.

Photo-unstable chemicals sometimes behave as phototoxins in skin, inducing untoward clinical side-effects when exposed to sunlight. Some drugs, such as psoralens or fluoroquinolones, can damage genomic DNA, thus increasing the risk of photocarcinogenesis. Here, lomefloxacin, an antibiotic from the fluoroquinolone family known to be involved in skin tumor development in photoexposed mice, was studied using normal human skin cells in culture: fibroblasts, keratinocytes, and Caucasian melanocytes. When treated cells were exposed to simulated solar ultraviolet A (320-400 nm), lomefloxacin induced damage such as strand breaks and pyrimidine dimers in genomic DNA. Lomefloxacin also triggered various stress responses: heme-oxygenase-1 expression in fibroblasts, changes in p53 status as shown by the accumulation of p53 and p21 proteins or the induction of MDM2 and GADD45 genes, and stimulation of melanogenesis by increasing the tyrosinase activity in melanocytes. Lomefloxacin could also lead to apoptosis in keratinocytes exposed to ultraviolet A: caspase-3 was activated and FAS-L gene was induced. Moreover, keratinocytes were shown to be the most sensitive cell type to lomefloxacin phototoxic effects, in spite of the well-established effectiveness of their antioxidant equipment. These data show that the phototoxicity of a given drug can be driven by different mechanisms and that its biologic impact varies according to cell type.

The significance of Nrf2 pathway in (photo)-oxidative stress response in melanocytes and keratinocytes of the human epidermis.

The expression of genes encoding antioxidant and/or phase 2 detoxifying enzymes can be enhanced in response to various environmental stresses. The main transcription factor involved in this response is nuclear factor erythroid 2-related factor 2 (Nrf2). Nrf2 activity is negatively regulated by the protein Kelch-like-Ech-associated-protein 1 (Keap1). While the roles of Nrf2 and phase 2 genes in chemoprevention of carcinogenesis have been well described; only few studies have dealt with their role in skin cancer. Normal human keratinocytes (NHK) and melanocytes (NHM) were treated by chemical inducers of the Nrf2 pathway or by small interfering RNAs (siRNA) used to knock down Keap1 mRNA. The above treatments resulted in significant stimulation of NQO-1 (NADPH-Quinone-Oxidoreductase 1) gene expression. GCL (gamma-Glutamyl-cysteinyl-ligase) gene was also induced but interestingly increased mRNA encoding the catalytic, heavy subunit GCLC was mainly stimulated in NHK, whereas the mRNA encoding the modifier, light subunit GCLM was mostly induced in NHM. HO-1 (Heme Oxygenase 1) gene induction was relatively strong in NHM, but generally absent in NHK, except when the cells were subjected to cytotoxic doses of the above chemicals. Exposure to solar UV (UVB + UVA, 300-400 nm) or to UVA alone (320-400 nm) confirmed this trend, but interestingly, at doses where cell growth reduction was comparable, UVA was generally more efficient than solar UV in inducing phase 2 genes. When siRNAs directed against Nrf2 were used, a strong down-regulation of NQO-1 expression was observed in both, NHM and NHK, whereas reduction of HO-1 expression was mainly detected in NHM. To our knowledge, this is the first study comparing phase 2 gene modulation in NHK and NHM. The results hereby presented should contribute to a better understanding of the molecular mechanisms involved in skin adaptation to environmental stress.