Systems modelling age-related changes in the maintenance of dermal extra-cellular matrix: mechanisms and interventions.

Age-related changes to the structure of the skin such as thinning, wrinkling and loss in flexibility are the result of changes in the regulation and composition of the dermal extra-cellular matrix (ECM). With age processes that degrade the ECM tend to dominate over regenerative processes. The consequences are the visible changes we all know together with increased incidence of conditions such as psoriasis, fibrosis, melanoma and impaired wound healing. Advanced anti-ageing skin-care products that target the causal mechanisms underlying these changes are a multi-billion pound industry which is forecast to increase over the coming decades. However much remains unknown about these mechanisms, and an in depth understanding of the processes involved in the maintenance of the ECM will provide better quality products and - perhaps more importantly - further our understanding of the skin ageing process itself.
The use of animals in skincare product development is widespread but is not necessarily very informative due to fundamental differences in biology with humans. Animal testing in cosmetics sold in the UK was first banned for tests carried out in the UK in 1998, EU in 2009 and extended to all countries in 2013. However, many animals, laboratory mice in particular, are routinely used in the basic research leading to identification of novel products for skin healthcare. Here, we aim to use a systems biology approach integrating in-silico discovery and in-vitro validation to offer a powerful alternative to animal experimentation. We will use this approach to generate computer models of age-related changes in the maintenance of ECM in the human dermis and use them to identify intervention strategies to counter undesirable changes.
Our computational models will be informed with data generated from human dermal cells, thereby avoiding focus on processes that may only be relevant in animal models. Once established the computer models will be used to explore treatment strategies and in more complex combinations that can be carried out solely in laboratory experiments. Those treatments that look promising will be tested in our in-vitro system which we have developed to behave very much like human skin. Our in-vitro dermal model allow us close control over which cells are grown within the tissue and how we can study them. We will use our system to help streamline product development for industry. An additional outcome will be a central computational resource where our data and models will be kept together and a software interface to allow others to interact with the models.

Overall, we will measure in-vitro and model in-silico the short-term biochemical network dynamics of extra cellular matrix maintenance (ECM) in populations of young, old and senescent dermal fibroblasts. We have shown in previous work that differences in network dynamics are highly informative and provide a means to identify parts of the network that could be targeted to restore healthy function. The work will be first carried out in 2D culture of human dermal fibroblasts then extended to a novel 3D in-vitro system which we have established and validation as a reliable surrogate for the human dermis. Intervention strategies will be explored in the computational model and subsequently tested in-vitro. The work will involve: generation of high throughput time series data; development and application of bioinformatics workflows to analyse the data and establish networks of molecular interactions governing the maintenance of ECM; building, calibration and validation of dynamic computational model(s) in vitro; use of the model(s) to identify intervention strategies to modify network behaviour; testing of molecules in vitro that could be used in potential treatments to improve skin healthcare.

Our project addresses the NC3Rs principles with obvious implications for both Reduction and Replacement for completely animal free research. Importantly this is also in line with the policy of our industrial sponsor, P&G. For Replacement, we use only 2D cell culture and 3D in-vitro tissue models of human derived dermal cells. No exogenous animal derived collagens are used in our completely humanised novel 3D dermal construct. The use of computational models expands the value of the laboratory experiments. In our case this reduces the need for further animal cell experiments but in a broader perspective the use of such models provides a powerful means to Reduce animal cell work in other experimental settings.

At the core of a systems modelling study is the computational model capable of simulating the biology of interest. Model development requires large time course datasets and if animals were sacrificed to acquire data at each time point then numbers would be substantial. The model calibration outlined in this proposal would require at least 1296 animals which have been directly Replaced with work in human derived cells. Other modelling studies have sacrificed animals to generate the necessary time course data and even with strict restrictions to only essential data the numbers of animals, most commonly mice, have been in the 100s-1000s. Once the model is calibrated, the numbers that are potentially replaced by its use to explore treatment strategies is more difficult to estimate but may in principle be very significant (1000s). This Reduction in numbers is likely to be large as they relate to the key advantage to using a computational model in testing the impact of interventions over a far larger range in terms of dose, timing and in combinations. Clearly animal experimentation on this scale is prohibitive.

As outlined in the proposal we will develop and implement a number of approaches with broad application to the NC3Rs including data-driven computational modelling, novel bioinformatic workflows for the analysis of time course data and the use of a novel in-vitro 3D cell culture system representative of human tissue. An advantage of the overall strategy is that once established it can be transferable to many other areas of study. This is in part enabled by our adoption of standards such as use of Systems Biology Markup Language (SBML) for model representation established by the Systems Biology research community. Many animals each year are sacrificed for basic research into ageing with questionable relevance to humans. Much of this work could be carried out in human cell lines and the value greatly expanded with generating computational models as we propose here. A google scholar literature search of ageing mechanisms using mice returned over 13000 publications in 2017. With a conservative estimate of 2 treatments, a group size of 6 in both young and old mice implies that over 300,000 (2x6x2x13000) animals were used that year. Our systems modelling approach could make a real impact and Reduce these numbers. Cell and tissue engineering are being increasingly used in the study of age-related conditions such as most cancers, neurodegenerative diseases, and cardiovascular disease. Our combined approach offers a proof of concept for the value of underpinning such work with a bespoke computational model.