The human genome encodes eighty tyrosine kinases that phosphorylate a wide range of substrates. Tyrosine phosphorylation alters protein function either by modulating its structure/activity or by altering its ability to engage in protein-protein interactions. It is therefore not surprising that aberrant tyrosine phosphorylation, caused by disregulation of cellular tyrosine kinases, contributes to cancer, developmental disorders, and neurodegenerative diseases. The work described in this thesis is aimed at increasing our understanding of the molecular mechanisms that connect upstream tyrosine phosphorylation events with the signaling pathways they activate and the downstream phenotypic effects they control. To this end, we developed a general method to identify sites of tyrosine phosphorylation on a protein of interest using tandem mass spectrometry. We then synthesized phosphopeptides representing each site and used them to probe protein microarrays comprising virtually every Src Homology 2 (SH2) domain and Phosphotyrosine Binding (PTB) domain encoded in the human genome. Spanning several biological systems, we demonstrate that combining mass spectrometry with protein domain microarrays provides a broad and unbiased way to uncover novel biophysical interactions and to generate testable hypotheses regarding signal transduction.
We begin in Chapter 2 by presenting data demonstrating that the sequences surrounding physiological sites of tyrosine phosphorylation are fundamentally different in their recruitment properties from those surrounding tyrosines that are not phosphorylated, supporting the notion that SH2 domains and kinases that generate their docking sites co-evolve. Next, in Chapter 3, we describe general method that combines targeted and untargeted mass spectrometry with protein microarrays to systematically identify pTyr-SH2/PTB domain interactions for a Receptor Tyrosine Kinase (RTK). We then use this method to uncover signaling pathways activated by ErbB4. Although a great deal is known about other ErbB family members (EGFR, ErbB2 and ErbB3), much less is known about ErbB4. By using a broad and unbiased approach, we were able to identify 19 sites of tyrosine phosphorylation on ErbB4, uncover a novel interaction with the transcription factor STAT1, and show that ErbB4 is far more selective in its recruitment properties than any other ErbB family member.
In Chapter 4, we use this same approach to study a cytokine receptor that does not contain an tyrosine kinase domain, but instead recruits and actives a cytoplasmic kinase upon ligand binding. We found that this receptor, c-Mpl, binds the SH2 domain of Tensin2, and then showed that Tensin2 acts to recruit Phosphoinositol-3-Kinase and activate Akt signaling. Finally, in Chapter 5, we explore the signaling network activated by the oncogenic fusion protein Etv6-NTRK3, which results from the chromosomal translocation t(12;15)(p13;g25). We show that one site, Y314, which has been predicted to regulate MAPK signaling through its consensus Grb2 binding sequence, is not, in fact, responsible for MAPK signaling, but instead signals through a variety of cytoplasmic tyrosine kinases and adaptor proteins including Bone Marrow X Kinase (BMX). This study highlights how the combination of proteomic techniques and molecular biology provides insight into signal transduction that is not obtained by either approach used in isolation. Although sites of tyrosine phosphorylation are often described as recruiting a single signaling protein, this work reveals that a single site of tyrosine phosphorylation are often capable of recruiting a diverse set of SH2 or PTB domain containing proteins. The combination of proteomic and more traditional cellular biological approaches provides a more thorough understanding of the signaling events that originate from each pTyr site in the human proteome, as well as insights into how these signals interconnect.