Unrepaired or inaccurately repaired DNA damage can lead to a range of cell fates, such as apoptosis, cellular senescence or cancer, depending on the efficiency and accuracy of DNA damage repair and on the downstream DNA damage signalling. DNA damage repair and signalling have been studied and modelled in detail separately, but it is not yet clear how they integrate with one another to control cell fate. In this study, we have created an integrated stochastic model of DNA damage repair by non-homologous end joining and of gamma irradiation-induced cellular senescence in human cells that are not apoptosis-prone. The integrated model successfully explains the changes that occur in the dynamics of DNA damage repair after irradiation. Simulations of p53/p21 dynamics after irradiation agree well with previously published experimental studies, further validating the model. Additionally, the model predicts, and we offer some experimental support, that low-dose fractionated irradiation of cells leads to temporal patterns in p53/p21 that lead to significant cellular senescence. The integrated model is valuable for studying the processes of DNA damage induced cell fate and predicting the effectiveness of DNA damage related medical interventions at the cellular level.

Cellular senescence, a state of irreversible cell cycle arrest, is thought to help protect an organism from cancer, yet also contributes to ageing. The changes which occur in senescence are controlled by networks of multiple signalling and feedback pathways at the cellular level, and the interplay between these is difficult to predict and understand. To unravel the intrinsic challenges of understanding such a highly networked system, we have taken a systems biology approach to cellular senescence. We report a detailed analysis of senescence signalling via DNA damage, insulin-TOR, FoxO3a transcription factors, oxidative stress response, mitochondrial regulation and mitophagy. We show in silico and in vitro that inhibition of reactive oxygen species can prevent loss of mitochondrial membrane potential, whilst inhibition of mTOR shows a partial rescue of mitochondrial mass changes during establishment of senescence. Dual inhibition of ROS and mTOR in vitro confirmed computational model predictions that it was possible to further reduce senescence-induced mitochondrial dysfunction and DNA double-strand breaks. However, these interventions were unable to abrogate the senescence-induced mitochondrial dysfunction completely, and we identified decreased mitochondrial fission as the potential driving force for increased mitochondrial mass via prevention of mitophagy. Dynamic sensitivity analysis of the model showed the network stabilised at a new late state of cellular senescence. This was characterised by poor network sensitivity, high signalling noise, low cellular energy, high inflammation and permanent cell cycle arrest suggesting an unsatisfactory outcome for treatments aiming to delay or reverse cellular senescence at late time points. Combinatorial targeted interventions are therefore possible for intervening in the cellular pathway to senescence, but in the cases identified here, are only capable of delaying senescence onset.

MOTIVATION: SBML is quickly becoming the standard format to exchange biochemical models. The tools presented in this paper are loosely-coupled, and are intended to be incorporated into SBML aware applications. The rationale for this is to reduce the amount of repeated work carried out within the community and to create tools that offer a greater number of features to the end-user. AVAILABILITY: All tools described are available from http://www.basis.ncl.ac.uk/software and are licensed under GNU General Public License.

We describe a program STATSEARCH which implements the method of Mott et al. (1989) for searching DNA and protein sequence databanks for statistically significant similarities to a given query sequence. STATSEARCH is written to run in conjunction with the GCG sequence analysis package.

A method is developed, based on word-searching, which provides a rapid test for the statistical significance of DNA sequence similarities for use in databank searching. The method makes allowance for the lengths and dinucleotide compositions of the sequences being compared. A way is also described to calculate the power of the test, i.e. the probability of detecting a given similarity as being statistically significant. The effects on the power of the test of the scoring method, word length, sequence length, and sequence composition are examined. A novel scoring method is shown to be superior to the method currently used in most word-searching algorithms.