VIAF is a member of a conserved protein family (initially identified by our collaborators in Colin Duckett's lab in NCI), that associate with animal IAPs (inhibitor of apoptosis proteins). VIAF itself substantially protects cells from Fas- and Bax-induced apoptosis, while coexpression of VIAF with suboptimal quantities of XIAP confers almost complete protection from these inducers. VIAF and XIAP activate JNK in a synergistic manner. Hence, VIAF is a novel cofactor which modulates the anti-apoptotic and signaling properties of the IAP family. In order to obtain a basis for understanding the function of VIAF at the molecular level, we have initiated a determination of the three-dimensional structure of VIAF using NMR spectroscopy. Full length VIAF contains 239 residues, but two-hybrid-screening studies have shown that the C-terminal 128 residue region of the protein is necessary and sufficient for interaction with Op-IAP. Hence we have expressed the C-terminal domain of VIAF, henceforth referred to as VIAFC in E..coli using the pET11 vector. Cultures were grown in minimal media containing either 15N or 15N plus 13C labeled ammonium chloride and glucose, respectively. The protein was purified from inclusion bodies and successfully refolded. NMR spectra were recorded on the protein dissolved in a 10 mM phosphate buffer, pH 7.4, at 35 ?C. The NMR resonances of backbone heavy atoms were assigned using the usual suite of 3D-heteronuclear experiments. Specifically, CBCA(CO)NH, HNCA and HNCACB were used to assign the amide-N, -HN, alpha-C? and beta-C signals. Next, the carbonyl carbons were assigned using the 3D-HNCO experiment. The HCACO experiment was then used to confirm these assignments. The alpha-H? and beta-H signals were then assigned using HNHA and HBHA(CO)NH experiments. These signal assignments were used to predict the secondary structure of VIAF using the chemical shift index (CSI). Surprisingly the VIAF secondary structure was highly homologous to that of phosducin, in spite of the limited sequence homology shared by the two proteins. The predicted secondary structure of VIAFC and the existing crystal structure of phosducin were then used to build an initial model for VIAFC. We next weakly oriented VIAFC in a Pf1 phage liquid crystal medium. Using a 15N-13C double-labeled sample we measured five sets of different heteronuclear residual dipolar couplings. Homology searches using these dipolar couplings were conducted against seven residue fragments generated from a PDB structure file and the best fits for these dipolar couplings yielded a model of the protein structure that had the correct secondary structure. However, additional data, in the form of NOE distance restraints are needed to determine the three dimensional structure of VIAFC. In order to obtain the needed NOE restraints, proton resonance assignments have been extended to the amino acid sidechains using a combination of C(CO)NH, HC(CO)NH, and 3D-HCCH-TOCSY experiments. Currently we have assigned more than 90 % of the protons in the structured region of the protein. We have also acquired 15N and 13C three- and four-dimensional NOESY spectra and are currently analyzing these data sets to obtain proton-proton distance restraints. We have been encouraged by the fact that the VIAFC model based upon the phosducin structure is consistent with the long range NOES we have observed so far. Upon solving the three dimensional solution structure of VIAFC we plan to study the interaction of VIAFC with XIAP. Specifically, we will map out the surface of VIAFC that interacts with XIAP.
