Single-Cell Proteomics Reveals Heterogeneous and Bimodal Proteome Responses to DNA Damage

DNA damage response (DDR) involves coordinated activation of repair, checkpoint, and stress adaptive signalling pathways following genomic toxicity. Bulk cell proteomics, however, averages divergent responses across neighbouring cells, obscuring the characterisation of biologically meaningful proteome remodelling during genotoxic stress. This is particularly relevant to DDR, where a subset of cells undergo apoptosis, resulting in the sampling of dead cells that confound downstream biological interpretation. Here, we applied single-cell proteomics (SCP) to investigate the proteome response of U2 OS cells to 6 thioguanine (6TG) induced DNA damage. By filtering individual cells according to predefined morphological, sampling and cell cycle state criteria, to ensure a matched comparison between control and DNA-damaged populations; 6TG treatment was found to activate canonical DDR, including mismatch repair and base excision repair, together with secondary stress responses linked to oxidative stress, mitochondrial dysfunction, inflammatory signalling and proteostasis. Underscoring this, DNA mismatch repair protein MSH3 and mitochondrial energy metabolism regulator PKRAB1, underwent the 4th and 3rd greatest increased protein abundances following DNA damage, respectively. We then exploited the individual cell resolution of SCP to interrogate neighbouring cell protein abundance heterogeneity, revealing significant globally increased protein abundance variability following DNA damage. In particular, proteins involved in mitochondrial oxidative phosphorylation and the tricarboxylic acid cycle exhibited a striking increase in cell to cell abundance variability, likely reflecting heterogeneity in the secondary stress responses elicited across U2 OS cells following 6TG treatment. Furthermore, analysis of protein abundance heterogeneity revealed proteins that exhibit bifurcated abundance states across neighbouring cells, with DNA damage inducing either the emergence or abolishment of these bimodal high low abundance distributions. Indeed, 6TG treatment was associated with remodelling of high low abundance subpopulation structure across several DNA damage response (PMS1, ORC6), inflammatory (LGALS3, IKKalpha), metabolic (IDH3A) and proteostasis related (STAMBP, NAA30, KLHDC4) proteins, indicating that DNA damage can reorganise cell to cell protein abundance distributions towards either distinct or more homogeneous proteomic states. Collectively, these findings demonstrate how SCP reveals cell to cell variation in the proteomic response to DNA damage that is obscured by population-level averaging.