No abstract

The exogenous application of small peptides can beneficially affect clinical skin appearance (wrinkles) and architecture (collagen and elastic fibre deposition and epidermal thickness). However, the discovery of new bioactive peptides has not been underpinned by any guiding hypothesis. As endogenous extracellular matrix (ECM)-derived peptides produced during tissue remodelling can act as molecular signals influencing cell metabolism, we hypothesised that protease cleavage site prediction could identify putative novel matrikines with beneficial activities. Here, we present an in silico to in vivo discovery pipeline, which enables the prediction and characterisation of peptide matrikines which differentially influence cellular metabolism in vitro. We use this pipeline to further characterise a combination of two novel ECM peptide mimics (GPKG and LSVD) which act in vitro to enhance the transcription of ECM organisation and cell proliferation genes and in vivo to promote epithelial and dermal remodelling. This pipeline approach can both identify new matrikines and provide insights into the mechanisms underpinning tissue homeostasis and repair.

The extracellular matrices (ECMs) of mammalian tissues play important roles in mediating and maintaining tissue function. However, aberrant and progressive remodelling of ECM components is a key feature in the pathology and ageing of many organs, including skin. Crucially, these degradative processes not only impair function, but may also release peptide fragments, known as matrikines, with cell signalling capabilities. In addition to endogenously produced matrikines, many exogenously applied bioactive peptides, with proven activity in skin, are homologous to naturally occurring amino acid sequences within ECM proteins. Despite the ability of some peptides to induce clinical benefits in aged skin, there has been no published conceptual framework to guide the prediction of new therapeutic matrikines. Here, we tested the hypothesis that small bioactive peptides (matrikines) can be predicted by the in silico digestion of dermal proteins by ECM proteases. We first identified a target cohort of 27 ECM proteins that were abundant in the dermis and/or were reported to undergo age-related remodelling. Utilizing an established machine-learning tool (PROSPER) we developed a PYTHON algorithm that could predict peptides liberated by in silico cleavage with eight skin-active enzymes [matrix metalloproteinase (MMP)-2, MMP-3, MMP-7 and MMP-9, cathepsin G and cathepsin K, granzyme B and elastase-2]. For the 27 target proteins this approach predicted the identity of 453 tetrapeptides. These peptides were predicted to be liberated predominantly from collagens (I, III, IV, VI and VII), some elastic fibre-associated proteins (EMILIN1 and fibulin-1) and the adhesive glycoprotein fibronectin. Eight peptides were selected for synthesis and biological activity testing based on their predicted protein sources, high solubility and likelihood of successful manufacture. The ability of these peptides to promote ECM synthesis (procollagen I, fibronectin, decorin, collagen IV, hyaluronic acid and fibrillin-1) in cultured human dermal fibroblasts (HDFs) was assessed by enzyme-linked immunosorbent assay or immunofluorescence (fibrillin-1 only). With the exception of hyaluronic acid, all of the peptides enhanced the synthesis of at least some ECM markers, including procollagen-I and decorin. Deposition of fibrillin-1, which is a sensitive marker of both skin ageing and repair, was significantly enhanced by three peptides. The ability of the peptides to influence the wider HDF secretome was assessed by liquid chromatography tandem mass spectrometry proteomics. Four of the peptides modulated synthesis of proteins in a wide range of functional classes, including enzyme inhibitors, ECM-affiliated proteins, cell adhesion molecules and basement membrane components, while the remaining four exhibited more targeted activities. We conclude that bioinformatic prediction of cleaved peptide fragments from ECM proteins can identify multiple small peptides with potentially beneficial activities for skin. This approach can both identify new matrikines and provide insights into the mechanisms underpinning tissue repair. Funding sources: this study was funded by the No. 7 Beauty Company.

No abstract