A systems biology investigation into age-related changes in the maintenance of collagen and extracellular matrix in human skin

Abstract

As skin ages, its capacity for providing an effective barrier between an organism and the outside world diminishes. The dermal extracellular matrix (ECM) is a critical component of skin that provides structural integrity and nourishment to the avascular epidermis. The ability for the dermis to perform its function is dependent on its molecular composition, which is mostly a fibrous mesh of type 1 collagens, elastins, proteoglycans and glycoproteins. Fibroblasts are the resident caretakers of the dermis as they are responsible for its constant remodelling and turnover. With age the composition of the dermis changes. Collagen and elastin fibres gradually become fewer in number and disorganised while matrix metalloproteinases (MMPs) that degrade the ECM become more prevalent. A consequence of these changes is that the aged dermis has reduced mechanical and tensile strength compared to the young dermis. Transforming growth factor β (TGF-β) is a pleotropic cytokine that induces the synthesis and deposition of dermal ECM, amongst other functions. The molecular processes that lead the demise of skin integrity and the contribution of TGF-β to these processes are not well understood. The goal of this thesis is to investigate skin ageing using a systems biology approach to gain insight into the differences between the regulation of young and old fibroblasts by TGF-β. This work describes the results of two high-throughput time series experiments. The first is an Affymetrix microarray experiment that was designed to understand how young fibroblasts transcriptionally respond to short term TGF-β treatment, with an emphasis on identifying transcriptional modes of cross-talk in the control of ECM synthesis. The second is a high-throughput quantitative polymerase chain reaction iii iv (qPCR) experiment that measured 68 hand-selected genes involved in the production of ECM components or TGF-β signalling proteins. By comparing temporal data from neonatal, adult and senescent fibroblasts we provide insight into the transcriptional differences that exist between young and old fibroblasts. The data collected using these high throughput methodologies were used in a combinatorial ordinary differential equation model selection problem that was designed to test the feasibility of four hypotheses concerning the fibroblast response to TGF-β in age and the role of connective tissue growth factor (CTGF) in TGF-β mediated collagen production. This work made use of two custom Python packages that are both available on Python package index (PyPI), PyCoTools (Welsh et., al. 2018) which is a toolbox for automating aspects of COPASI, a commonly used software package in systems modelling, and pytseries which provides a set of useful objects and methods for handling and manipulating time series data.