Development of an Ultrasensitive Phosphoproteomic Workflow to Enable Primary Cell Phosphoproteomics

© 2021 Rune Hertz LarsenPhosphorylation is a reversible post-translational modification which participates in nearly all human signalling events. The role of phosphorylation in biology has been studied for more than 70 years, yet only a tiny fraction of substrates and their corresponding function have been characterised. Over the last 20 years, phosphoproteomics has developed and now established itself as the primary tool to investigate biological signalling events. Phosphoproteomics enables the global unbiased quantitative detection of more than 10,000 phosphorylated sites on a routine basis. However, despite this sensitivity, phosphoproteomic experiments have primarily been performed in immortalized cell lines and tissue samples containing multiple cell types. Moreover, complex biological information is often distilled and collated into single pathways that attempt to capture general signalling cascades. These generalizations through simplified pathway analysis risks losing valuable insights into biological system and does not make full use of the exquisite sensitivity that phosphoproteomic workflows provide. Additionally, cellular responses to ligands can be radically different between different cell types and cellular states, thus, the generalization of pathways combined with most experiments being performed in immortalized cell lines significantly reduces the power of phosphorylation analysis in human systems. The use of cell lines and mixed tissues has been necessary due to the large protein input requirements in phosphoproteomic experiments. Historically, the protein input requirements have been more than 1mg of protein. However, recent developments decreased the input amount required to 0.2mg of protein per sample. Despite this large improvement, the protein input is still highly prohibitive for most primary cell experiments from mice and men. Primary cells are derived from limited sources and expansion ex vivo is highly limited for most primary cell populations. Based on the need for primary cell compatible phosphoproteomic methods, I present the development of an ultrasensitive phosphoproteomic method which is 4-fold more sensitive than current published methods. Moreover, I present an experimentally derived framework to assist in the design of limited input phosphoproteomic experiments. The framework assists in the selection of timepoints, through comprehensive mapping of TNF induced signalling. Furthermore, the choice of protein input is recommended through an investigation of stimulation induced signalling at different inputs. The experiment is based off three donors and investigates how much material is required to identify donor specific differences in signalling. Lastly, we explore the recovery of phosphorylated peptides from formaldehyde crosslinked cells, which will enable the sorting of cell populations post stimulation. This improved workflow is demonstrated on a murine model of TNF induced T cell proliferation using T cells derived from the spleen that are stimulated ex vivo. However, as ex vivo expansion is not possible in this model, we performed the experiment using 12ug of protein input per sample. Even at these incredibly low levels of input, I determine that TNFR2 stimulation may regulate the metabolism of T cells towards a more sustainable energetic environment favouring memory and regulatory T cells rather than an effector phenotype