bioRxiv2020
Computational modelling is an important research tool, helping predict the outcome of proposed treatment plans or to illuminate the mechanics of tumour growth. In silico modelling has been used in every aspect of cancer research from DNA damage and repair, tumour growth, drug/tumour interactions, and mutational status. Indeed, modelling even holds potential in understanding the interactions between individual proteins on a single cell basis. Here, we present a computational model of the cell cycle network of the cyclin family of proteins (cyclin A, B, D and E). This model has been quantised using western blot and flow cytometry data from a synchronised HUVEC line to enable the determination of the absolute number of cyclin protein molecules per cell. This quantification allows the model to have stringent controls over the thresholds between transitions. The results show that the peak values obtained for the four cyclins are similar with cyclin B having a peak values of 5x106 to 9x106 molecules per cell. Comparing this value with the number of actin proteins, 5E8, shows that despite their importance, the level of cyclin family proteins are approximately 2 orders of magnitude lower. The efficiency of the model presented would also allow for its use as an internal component in more complex models such as a tumour growth model, in which each individual cell would have its own cell cycle calculated independently from neighbouring cells. Additionally, the model can also be used to help understand the impact of novel therapeutic interventions on cell cycle progression. Author SummaryProtein and gene networks control every physiological behaviour of cells, with the cell cycle being controlled by the network of genes that promote the cyclin family of proteins. These networks hold the key to creating accurate and relevant biological models. Normally these models are presented with relative protein concentrations without any real world counterpart to their outputs. The model presented within shows and advancement of this approach by calculating the absolute concentration of each cyclin protein in one cell as it progresses through the cell cycle. This model employs Boolean variables to represent the genetic network, either the gene is active or not, and continuous variables to represent the concentrations of the proteins. This hybridised approach allows for rapid calculations of the protein concentrations and of the cell cycle progression allowing for a model that could be easily incorporated into larger tumour models, allowing for the tracking of discrete cells within the tumour.