Computational systems biology approaches to investigate cellular senescence

Abstract

Cellular senescence is a complex stress response that results in an irreversible cell cycle arrest, accompanied by phenotypic, metabolic, and secretory changes termed as the senescenceassociated secretory phenotype (SASP). Senescent cells accumulate with age and are associated with several age-related pathologies, including fibrosis and chronic wounds. However, they also play important functional roles, such as in cancer suppression, embryogenesis, and wound healing. The inherent complexity of cellular senescence requires an approach where multiple levels of complexity can be addressed individually, followed by an investigation of the interactions between these levels, as well as the whole system. For this reason, a computational systems biology approach was adopted, which uses bioinformatics methods and computational/mathematical modelling to address biological processes. A tissue-level agentbased computational model was developed to investigate the dynamics of senescent cells in healthy wound healing. This model was then used to explore how perturbations in the system lead to pathological healing outcomes, such as fibrosis and chronic wounds. The model suggests that an appropriate number of senescent cells be present during the mid to late stages of wound healing to promote healthy healing, whereas higher background levels of senescence promote an inflammatory chronic wound. Furthermore, the model suggests that an abundance of pre-existing senescent cells, as during ageing, also promotes a chronic wound response. In contrast, an insufficient number of senescent cells was shown to promote a fibrotic response due to a lack of fibrolytic factors. Lastly according to the model, the delayed induction of senescent cells also results in fibrotic responses, due to a delay in the necessary fibrolytic activity. Overall, these results suggest that the functional versatility of senescent cells can be attributed to differences in SASP composition, duration of senescence and temporal induction of senescence relative to the healing phase. The range of outcomes demonstrated by the model strongly highlight the dynamic and heterogenous role of senescent cells in wound healing, as well as fibrotic and chronic wounds, and their necessary fine-tuned control. At the intracellular level, a bioinformatic workflow was developed to analyse a recently published multi-omic time-series dataset gathered during replicative senescence (RS). The primary output from the workflow were mechanistic networks driving RS, which confirmed the multi-step and gradual nature of senescence progression. Network analysis also revealed that senescence processes, especially involving cell cycle arrest iii and apoptosis evasion, begin early through to intermediate senescence, followed by effectors involved in SASP production during late senescence. Overall, multiple computational methods were used in this thesis to investigate the role of cellular senescence, at the tissue-level both in physiological and pathological settings, as well as to investigate intracellular mechanisms driving senescence. Furthermore, the wound healing model provides an extensible resource that can be used for further investigation with appropriate data. The bioinformatic workflow can be readily used to analyse other multi-omic datasets to generate mechanistic hypotheses.