ALBERT

Description: Abstract 19 Coagulation factor XI (FXI) is a uniquely dimeric coagulation protein, which in its activated form (FXIa) activates FIX to FIXa. We have previously shown that the dimeric structure of FXI is essential for normal autoactivation and activation by thrombin and FXIIa, but not for the expression of enzymatic activity against FIX (Wu W, et al J. Biol. Chem. 283:18655-18664, 2008). A comparison of three separate structures of FXI/XIa from our laboratory (i.e., the crystal structure of the catalytic domain of FXIa in complex with the kunitz protease inhibitor domain of protease nexin-2; the crystal structure of full-length, dimeric FXI; and the NMR structure of the FXI A4 domain) predicts a major conformational change accompanying the conversion of FXI to FXIa. We now show that when FXI binds to the negatively charged polymer, dextran sulfate and is autoactivated to generate FXIa, changes of intrinsic fluorescence are observed, i.e, a decrease in fluorescence intensity and a red shift of emission wavelength, which also suggests that a conformational change accompanies FXI activation. To investigate the mechanism of FXI zymogen activation and the allosteric transition accompanying the conversion of FXI to FXIa, which exposes binding sites for FXIa ligands, we have carried out fluorescence resonance energy transfer (FRET) studies to characterize the conformational changes accompanying zymogen activation. Using a sensitive free thiol quantitation assay, we confirmed the presence of a single free cysteine residue (Cys11) per subunit of recombinant FXI, which was quantitatively labeled with the thiol reactive fluorescence dye IAEDANS (5-({2-[(iodoacetyl)amino]ethyl}amino)naphthalene-1-sulfonic acid). Fluorescence excitation of AEDANS-labeled FXI at 280 nm shows a prominent dansyl emission peak (∼450 nm) in addition to the Trp emission peak (∼325 nm) indicative of efficient FRET from Trp donors to the AEDANS acceptor. Controls using a C11S mutant of FXI showed ∼10-fold lower levels of AEDANS labeling, confirming that Cys11 is the predominant labeling site. Autoactivation of FXI in the presence of dextran sulfate results in a major decrease in donor emission, but has little effect on acceptor emission. This indicates that, for wild-type FXI, FRET is dominated by transfer within the A1 domain originating from Trp55, which is located at a distance of 18 Å from Cys11, far closer than any other tryptophan. The changes in Trp emission, which are similar in the presence and abence of AEDANS, allow us to follow the kinetics of zymogen activation. The S557A active-site mutant of FXI, which cannot undergo autoactivation, showed no fluorescence changes upon addition of dextran sulfate, confirming that the observed decrease in Trp fluorescence is due to formation of active FXIa enzyme. In an effort to observe specific inter-domain FRET, we prepared an AEDANS labeled W55H mutant of FXI, which eliminates the Trp donor in the A1 domain that dominates energy transfer in wild-type FXI. Our data show that autoactivation of the W55H mutant is accompanied by a significant increase in AEDANS emission that can be attributed to the movement of the labeled Cys11 (in A1) relative to Trp228 in the A3 domain of the opposite dimer subunit. In the crystal structure of FXI, the distance for this donor-acceptor pair is 29 Å (compared to a distance of 40 Å for the second closest Trp, Trp407 in the catalytic domain), making it a sensitive and specific FRET probe for monitoring changes in domain arrangement associated with enzyme activation and ligand interactions. A comparison of the FXI crystal structure with our model of FXIa showed that the distance between the active site serines (Ser557) of each catalytic triad is shortened from ∼118 Å in the zymogen to 40–75 Å in the enzyme. Since the distance between the two scissile bonds of each subunit of FXI is also ∼75 Å, we propose that during autoactivation, either the active site of each catalytic domain of FXIa is positioned to cleave the Arg369-Ile370 bond of the opposite subunit (intersubunit transactivation) or a FXIa dimer positions its two active sites adjacent to the two scissile bonds of a separate FXI dimer (intermolecular activation). These studies support a model in which the autoactivating transition from zymogen to enzyme is accompanied by the movement of each catalytic domain of the dimer to facilitate efficient autoactivation of FXI. Disclosures: No relevant conflicts of interest to declare.

Description: Activation of the tyrosine kinase c-Src is a fundamental element of integrin αIIbβ3-mediated outside-in signaling in platelets. Presently, it is thought that c-Src is bound via its SH3 domain to the C-terminal Arg-Gly-Thr (RGT) sequence of the β3 cytoplasmic tail (CT). Although a recent NMR study supports this contention, it is likely that such binding would be precluded in the inactive conformation of c-Src because the linker connecting the SH2 and kinase domains of c-Src physically occludes the putative β3 CT binding site in the SH3 domain. Accordingly, we have re-examined the interaction between c-Src and β3 by immunoprecipitating β3 bound c-Src from resting and agonist-stimulated platelets and then applied the biophysical techniques of surface plasmon resonance (SPR) and NMR spectroscopy to directly characterize the interaction of the SH3 domain with the β3 CT. In unstimulated platelets, there was little or no association between c-Src and β3. However, following platelet stimulation with 1 unit/ml thrombin, c-Src binding to β3 is detectable within 10 sec of stimulation and binding persists for at least the next 90 sec. Phosphorylation of c-Src residue Tyr419 located in its activation loop is an indicator of c-Src activation. We detected the phosphorylation of β3-bound c-Src on Tyr419 at 20 sec following thrombin stimulation. Tyr419 phosphorylation then persisted for the next 40 sec after which it rapidly declined. Thus, platelet stimulation induces the rapid interaction of c-Src with the β3 subunit of αIIbβ3 and this interaction is accompanied by the induction of c-Src kinase activity. Since it has been reported that c-Src binds to β3 via its SH3 domain, we used SPR spectroscopy to measure the strength of this interaction. In initial experiments, we measured binding of the C-terminal β3 heptapeptide NITYRGT to the immobilized SH3 domain and detected NITYRGT binding with a dissociation constant (Kd) of approximately 700 mM. We then repeated the measurements by flowing the SH3 domain over the entire 43 residue β3 CT appended to the SPR chip. Under these conditions, we detected a Kd for SH3 domain binding to the β3 CT of 7.21 ± 2.42 µM. Previously, we reported similar behavior for talin-1 FERM domain binding to the β3 CT. Thus, these experiments suggest that like binding of the talin-1 FERM domain to the β3 CT, the interaction of c-Src with β3 is a two-dimensional ternary interaction in which the membrane surface makes an important contribution. Next, to identify the sites in the SH3 domain that interact with β3, we used NMR spectroscopy. In these experiments, we measured the interaction of the β3 peptide NITYRGT with the 15N-labeled SH3 domain. Because NITYRGT binding to the SH3 domain is weak, we used chemical shift perturbations (CSP) to identify the SH3 domain residues interacting with NITYRGT. We found that NITYRGT interacted with the SH3 domain RT-loop and surrounding residues. A control peptide whose last three residues where replaced with those of the β1 CT (NITYEGK) induced only small CSP on the opposite face of the SH3 domain. Next, to mimic inactive c-Src, we found that the canonical polyproline peptide RPLPPLP prevented binding of the β3 peptide to the RT-loop. Under these conditions, the β3 peptide induced CSP similar to the negative control. Thus, these studies indicate that the primary interaction of c-Src with the β3 CT occurs in its activated state at a site that overlaps with the polyproline binding site in its SH3 domain and suggest that protein-membrane interactions make an important contribution to the strength of binding. By contrast, interactions of inactive c-Src with β3 are weak and insensitive to β3 CT mutations. Disclosures No relevant conflicts of interest to declare.

Description: Abstract 383 An essential component of αIIbβ3-mediated outside-in signaling is activation of the tyrosine kinase c-Src, some of which is constitutively bound via its SH3 domain to the C-terminal Arg759-Gly760-Thr761 (RGT) sequence of the β3 cytoplasmic tail. RGT is quite different from the canonical polyproline sequence recognized by SH3 domains in which a polyproline helix packs against a shallow groove composed of aromatic residues (Tyr93, Tyr95, Tyr139 in c-Src). A specificity pocket located at the end of the groove and composed of residues from the n-Src- and RT-loops affects substrate specificity. Because of the obvious difference between RGT and polyproline sequences, we asked how RGT binds to the c-Src SH3 domain and what implications this has for c-Src regulation by αIIbβ3. Initially, we employed CD spectroscopy and tryptophan (Trp) fluorescence because these techniques are sensitive to changes in the local environment surrounding aromatic residues. However, there were no differences in the CD spectrum of the SH3 domain in absence or presence of the β3 peptide NITYRGT, whereas there was a clear shift in the presence of the core polyproline peptide RPLPPLP. Polyproline binding to Trp in the SH3 specificity pocket also results in a blue shift in Trp fluorescence from 355 nm to 347 nm; however, the fluorescence spectrum was essentially unchanged in the presence of NITYRGT. These experiments suggest that either the interaction of NITYRGT with SH3 is extremely weak and not observed at the concentrations used or occurs outside of the aromatic groove and the specificity pocket. Accordingly, we turned to NMR, a method able to detect weak protein-protein interactions. Two dimensional 1H-15N HSQC spectra of the SH3 domain in the presence of NITYRGT exhibited a number of changes in chemical shift compared to the spectrum in the absence of ligand. Sixteen residues located in the n-Src and RT-loops, grouped around the specificity pocket, had chemical shift changes 〉 0.05 ppm. The largest changes occurred in residues in or adjacent to the RT-loop, especially residues Arg98, Glu100, and Asp102. Of the resides forming the aromatic groove, only Tyr95 which is adjacent to the specificity pocket was perturbed by NITYRGT. Plots of the chemical shift changes for NH groups in SH3 vs. NITYRGT concentration were linear, indicating that the majority of SH3 domain was unbound. Further, a Kd for NITYRGT binding to SH3, estimated from these experiments, was between 175–350 mM. Next, we obtained HSQC spectra for SH3 in the presence of either RPLPPLP or a negative control peptide NITYEGK. Major perturbations due to RPLPPLP occurred in three regions: residues 98–103 (RT-loop), 116–122 (n-Src loop and specificity pocket), and residues 134–138; residues in the aromatic cluster were unaffected by the ligand. By contrast, only a handful of residues showed small perturbations in the presence of NITYEGK and there was no overlap between the affected residues and those affected by RPLPPLP. In conclusion, our results indicate that compared to polyproline sequences, the C-terminus of the β3 cytoplasmic tail binds to the c-Src SH3 domain in the region of the SH3 specificity pocket. Because chemical shifts for acidic residues located in the RT-loop were particularly sensitive to the presence of NITYRGT, it is likely that Arg759 in β3 makes an important contribution to the interaction. Moreover, we found that the interaction between NITYRGT and the c-Src SH3 domain is substantially weaker than was previously reported for the interaction of β3 with c-Src. This suggest the possibility that a third component is required for this interaction to occur under biological conditions. Recently we found that the β3 cytoplasmic tail in solution has weak affinity for the talin-1 FERM domain, but appending the tail to acidic phospholipids increased its affinity by three orders of magnitude. Since the c-Src SH3 domain contains a conserved patch of basic residues that are necessary for binding to acidic phospholipids, it is possible that the interaction of c-Src with β3 is also a ternary interaction in which protein-lipid interactions play an important role. Disclosures: No relevant conflicts of interest to declare.