Thrombin is generated by the coagulation cascade, and it is the enzyme responsible for cleaving fibrinogen to make the fibrin clot. Thrombin has several binding sites: The active site is where proteolytic cleavage occurs; Exosite 1 is where fibrinogen, thrombomodulin, and hirudin (a leach inhibitor of thrombin) bind; Exosite 2 is where heparin binds to accentuate the inhibition by antithrombin III. Our lab has mainly been interested in allosteric communication between Exosite 1 and the active site. We have used amide H/D exchange to map changes in amide exchange that show possible “pathways” of allostery (Figure).
We have also used isothermal titration calorimetry to understand the thermodynamics of the allosteric process. Using NMR, we are currently measuring backbone dynamics to gain a deeper view of the allosteric process and to understand the effects of mutations in thrombin.
Thrombomodulin (TM) is a cell-surface receptor that binds thrombin and changes its catalytic activity from pro-coagulant to anticoagulant. Of the six epidermal growth factor-like (EGF) domains in TM, the EGF4-5-6 fragment has full activity. It binds thrombin in Exosite 1 (Figure) but somehow changes the way thrombin interacts with substrates and inhibitors. When TM is bound, thrombin no longer cleaves fibrinogen, instead it cleaves protein C to make activated protein C (aPC) and aPC then cleaves and inactivates Va and VIIIa the two limiting essential cofactors in blood coagulation. We solved the solution structure of the smallest active fragment of TM and using H/D exchange, we showed that TM allosterically regulates thrombin by changing the loops surrounding the active site. Recently, we have been working to make a soluble, fully-active smaller TM molecule and using NMR to understand how the dynamic loops of TM are involved in binding and allosteric affects in thrombin.
The NFκB family of transcription factors responds to inflammatory cytokines with rapid activation of transcription and subsequent signal repression. Much of the system control depends on the unique characteristics of the inhibitor, IκBα, which appears to have complex weakly-folded parts that are critical for function. We have shown that of the six ankyrin repeats (ARs) in IκBα, the fifth and sixth are only completely folded when IκBα is bound to NFκB. These repeats exchange all their amides within one minute and NMR signals are absent for most of AR5 and 6. Single molecule FRET studies have shown that in free IκBα, AR5 and AR6 fluctuate between a compact state and a more extended state but upon binding NFκB, the fluctuations cease. Combined solution biophysical measurements and quantitative protein half-life measurements inside cells have shown that free IκBα is degraded quickly by a ubiquitin-independent proteasome-mediated process Bound IκBα remains extremely stable, and requires phosphorylation and ubiquitinylation for targeted proteasome-mediated degradation. Through a collaboration with G. Ghosh, P. Wolynes, J. Dyson and A. Hoffmann, we are working towards integrating our biophysical understanding of the protein interactions with how signals are processed inside cells.
ASB-containing Ubiquitin E3 Ligase
K48 poly-ubiquitin (Ub) conjugation signals for proteolytic degradation of its substrate protein and is responsible for the large majority of regular protein degradation, which prevents the breakdown of cellular function, such as in neurodegenerative diseases. Ub is attached to its substrate as a monomer by an E3 ubiquitin Ligase as the third step of the ubiquitin activation cascade. The E3 complex contains a substrate-recognition subunit, and upon binding of the appropriate substrate, the E2 subunit catalyzes the transfer of ubiquitin from its cysteine onto a lysine residue on the substrate protein as an isopeptide bond. The E2 will subsequently ubiquitinate previously attached ubiquitin residues in order to form K48 polyubiquitin linkages.
The Ankyrin-repeat and SOCS-Box containing protein 9 (ASB9) substrate recognition subunit of one E3 ligase binds to and facilitates the ubiquitination of Creatine Kinase. So far we have used a combination of biophysical binding experiments and HDXMS to map the binding of ASB9 with creatine kinase.
Our work in mass spectrometry has several facets. We first demonstrated the possibility of measuring amide H/D exchange by MALDI mass spectrometry and we also showed that interaction surfaces on proteins could be mapped by measuring changes in solvent accessibility of amides on the surface of the protein. Lately, we have been working to improve proteomics work-flows through the characterization of additional proteases with alternative substrate specifities for increased proteome coverage and post-translational modification mapping
The urokinase plasminogen activator (uPA) is a serine protease that activates the “anti-coagulation” pathway. Once thrombin makes the fibrin clot, it is eventually necessary to digest the clot so that blood flow can resume. UPA activates plasminogen forming plasmin, the clot-digesting protease. Plasmin also can digest the extracellular matrix, and tumor cells display higher concentrations of the uPA-receptor (uPAR), which plays a critical role in tumor progression and metastasis. When uPA binds to uPAR, it becomes more active, but how this process occurs is not known. Few biophysical studies have been performed on this system. Therapeutic targeting of this important pathway has slowed due to a poor understanding of the mechanisms by which these proteins interact. The functional activation of uPA resulting from uPAR binding strongly suggests allostery, but no studies of dynamics to explicitly look for allosteric sites have been performed on this system. A structure of the uPAR with only the N-terminal EGF-kringle domains of uPA is available, but structures of uPAR with full-length uPA (EGF-kringle-protease) are not available. The structure of full-length uPA is also not known. Our recent measurements of amide hydrogen/ deuterium exchange (HDXMS) in the uPA protease domain revealed that the C-terminal b-barrel is very dynamic, and it adopts an alternate fold in some crystal structures. Large differences in H/D exchange were observed throughout the protease domain upon binding different ligands such as allosteric nanobodies. The results revealed the protease domain is highly dynamic and can adopt multiple conformational states. In this project, we will study the dynamics of uPA and how the dynamics change upon binding uPAR to determine if uPAR allosterically activates uPA.