Histological Validation of a Predictive Cosmetogenomics Study Ex Vivo
Total Page:16
File Type:pdf, Size:1020Kb
Histological validation of a predictive cosmetogenomics study ex vivo Peno-Mazzarino L.1, Patatian A.2, Bader Th.2, Percoco G.1, Gasser Philippe1, Scalvino S.1 Reby D.1, Durand C.1, Lati E.1 and Benech P.2, 3 1- Laboratoire BIO-EC, 1 chemin de Saulxier, 91160 Longjumeau, France 2- Genex, 1 chemin de Saulxier, 91160 Longjumeau, France 3- Aix Marseille Université, CNRS, NICN UMR 7259, 13916 Marseilles, France Introduction The ex vivo human skin explants is a choice model system to objectivate the activity of cosmetic products. Close to in vivo conditions, it allows evaluating some final products by taking into count the penetration dimension. In addition, it allows the use of more invasive methods requiring biological samples (cells, tissues, supernatants). The genomic study which is a powerful tool that allows early to identify genes activated by cosmetic products, can be combined with the ex vivo studies. Thus, the laboratories BIO-EC and GENEX have developed a model system of cosmetogenomics using ex vivo human skin explants. In order to explore the capability of this association and to correlate the protein expression by histology to gene activation by transcriptomic analysis, we investigated the effects of a commercial cosmetic product. The use of topically applied vitamins has become an ubiquitous part of skin care. While a part of the skin's antioxidant system that assists in protecting skin from oxidative damage, vitamins A, C, and E have also proven their ability to treat photoaging, acne, cutaneous inflammation, and hyperpigmentation [1]. Retinoic acid, a metabolite of vitamin A, is known to be a key signaling molecule in regulating epithelial cell differentiation. Topical natural retinoic acid precursors such as retinaldehyde or retinol are less irritant than acidic retinoids and may be combined with other compounds with complementary actions against ageing. According to its recognized activity and its well known action process, we have chosen a cosmetic product containing retinol. In this study, we investigated on human skin explants, derived from two donors, the effects of a commercial product containing retinol. For an extended period of more than a week, skin explants were treated with the product and samples were taken at different time points and compared to untreated controls. Genes of interest, identified by microarray analysis, were followed-up by immunochemistry on thin sections of explants as well as by RT-qPCR. Materials and methods Tested products A commercial product, a cosmetic anti-ageing product containing retinol (R) has been applied to ex vivo human skin explants cultured for 6 days. Ex vivo human skin explants Human skin explants were obtained from two abdominoplasties from a 30 and a 49-year- old woman (plastic surgery). The hypodermis was removed from the skin and circular samples were prepared using a punch instrument. The diameter of each explant was ~10mm. The explants were maintained in survival in a liquid–air interface in BIOEC’s ® Explant Medium (BEM ) at 37°C in moist atmosphere containing 5% CO2 for 24h before the study began. Half of the medium (1 ml) was renewed every other day. Living skin explants were treated as shown in Table 1. Batch Treatments Explants Nb Sampling times D-1 none 3 D-1 U Untreated batch 21 D0/D0+3h/D0+9h/D1/D2/D3/D6 R Retinol 18 D0+3h/D0+9h/D1/D2/D3/D6 Table 1 Experimental conditions of living skin explants Product applications A commercial cosmetic anti-ageing product was applied to ex vivo human skin explants cultured for more than one week. The retinol-containing product was applied topically every other day at 1 mg/cm². Sampling Explants were collected at the following incubation times: 3 hours, 9 hours, 1 day, 2 days, 3 days and 6 days. The explants were then cut in three parts. One part was fixed in buffered formol solution, the second one frozen at -80° C and the third used for RNA extraction and analysis. Gene Expression Profile Total RNAs were extracted at each sampling time for each explant using the RNeasy Fibrous Tissue Kit-Qiagen after disruption and homogenization using the TissueLyser Kit. To evaluate the effects of a treatment, we focused our first investigation on the samples coming from a 49 years old donor. After 3h, 9h, 24h (D1), 48(D2), 72 (D3) of treatment, five hundred nanograms of RNA of each sample were processed further for retrotranscription, amplification and cyanine labeling, using the Illumina Whole Human Genome Microarray (HT12 v 4.0) which presents more than 47.000 probes, derived from the National Center for Biotechnology Information Reference Sequence (NCBI) RefSeq. In a second step with donor #2 the more prominent time point(D2) was validated by microarray starting from sixty nanograms of RNA and using the low input Quick Amp one- color labeling kit (Agilent technologies) which contains 60.000 probes. Subsequently, the expression of key genes the modulation of which showed an impact of the treatment at crucial points of different signaling pathways involved in skin biology was evaluated by RT-qPCR at D2; D3 and D6 of the kinetics. After qualitative validation and normalization of the microarray data, gene expression intensities of the treated samples were compared with untreated samples. Only those genes that were modulated at least in 2 consecutive time points with a fold-change (FC) of at least ≥1.45 and ≤0.5 were considered for evaluation by PredictSearch™ analysis. PredictSearch™ is a proprietary data and text mining software that searches and retrieves correlations between genes and biological processes or diseases through millions of scientific publications [2-3]. The functional correlation based on the Fisher test allows statistical co-citation analysis of annotated keywords in order to define relationships between genes, biological processes and concepts, metabolites, diseases, and tissue/cells/organs with a direction of effect rather than mere associations. Histological process After 24 hours of fixation in formol solution, the samples were dehydrated and paraffin impregnated with a Leica 1020 dehydrator automat. After preparation, explants were embedded at a Leica EG 1160 coating station and 5 μm sections were cut with a Minot-type microtome, (Leica 2125) and mounted on superfrosted silanized glass histology slides. Microscopical observations were performed by optical microscopy with a 40x objective. Photo-micrographs were performed with a DP72 Olympus camera and archived with Olympus CellD data storing software. Staining and immunostaining The observations of general morphology were carried out on formalin fixed paraffin embedded skin sections stained according to Masson’s trichrome Goldner variant, using a ST 4040 Leica staining automat. Cellular or tissular modifications were noted in the stratum corneum, living epidermis, papillary and reticular dermis. The general morphology resulted in observational and descriptive data and without statistical analysis. Formalin-fixed paraffin embedded skin sections were subjected to immunostaining according to the following process: deparaffinizing and rehydration (Clearen, alcohol), quenching of endogenous peroxidase with 1% H2O2 in phosphate-buffered saline (PBS) for 10 min; washing 3 times in PBS; blocking with normal horse serum (Vector Laboratories) for 30 min at RT; incubation for 1 hour at RT with the primary antibody (list below), washing 3 times in 0.05% Tween-20 in PBS; incubation for 30 min at RT with R.T.U. Biotinylated Universal Antibody (Vectastain Universal Elite ABC Kit, Vector Laboratories); washing 3 times in 0.05% Tween 20 in PBS; incubation for 30 min at RT with R.T.U. Elite ABC Reagent (Vectastain Universal Elite ABC Kit, Vector Laboratories); washing 3 times in 0.05% Tween-20 in PBS, and revelation with VIP substrate (Vector Laboratories) for a period specific for each primary antibody. Negative controls were performed by replacing the primary antibody with PBS. Primary antibodies: anti-Ki67 (7B11 mouse monoclonal, Zymed), anti-involucrine (SY5 mouse monoclonal, Novus Biologicals), anti-FABP5 (rabbit polyclonal, Santa Cruz Biotechnology), anti-filaggrin (AKH1 mouse monoclonal, Santa Cruz Biotechnology), and anti-CRABP2 (rabbit polyclonal, Sigma-Aldrich). Results & Discussion The number of modulated gene was evaluated for at least 2 consecutive time points (FC≥1.45 or FC≤0.5). Based on these criteria, 18 genes were significantly up-regulated at 3 and 9 hours; 14 genes at 9-24h; 51 genes at 24-48h and 75 genes at 48-72h. Similarly, 35 genes were repressed at 3-9h; 12 genes at 9-24h; 13 genes at 24-48h, and 92 genes at 48-72h. The number of modulated genes increased after 48 and 72h of treatment. Thus, our next transcriptomic approach was centered on the validation of the results after 48h treatment of skin explants from the second donor. As expected, the expression of a set of genes (DHRS3, DHRS9, CRABP2, FABP5 and CYP26B1) involved in retinol metabolism was induced in both donors (Table 2a). Among all modulated sequences, genes, which were induced at early times of treatment (3h and 9h), were associated with cell proliferation. In contrast, genes, which were induced at 24-48-72h were associated with differentiation (Table 2b). a) b) Table 2: a) Expression of genes involved in retinol metabolism. Fold-changes at 3h, 9h, 24h, 48h, 72h were calculated for donor 1. Only the 48h treatment was performed for donor 2 (indicated by an asterisk). b) Expression of genes involved in differentiation. It is important to underline that similar results were obtained with both donors although RNAs were processed differently and microarrays were performed on distinct platforms. (Illumina for donor 1 and Agilent for donor 2). Using PredictSearch™, we were able to insert these genes within biological networks highlighting the role of retinol in the induction of transcriptional activities that lead to dual functions (Figure 1). Indeed, the metabolism of retinol is regulated by groups of enzymes that control conversion of retinol into active retinoid aldehyde and retinoic acid.