In order to be able to study larger complexes by NMR we developed and adapted new technology in the laboratory. We adopted new NMR experiments to study larger proteins as well as labeling procedures (deuteration) that was not previously available for our group. In addition we showed by careful experiments and controls that one can use scalar coupling in addition to chemical shift to map out molecular interactions. This enables one to probe allosteric process that was previously difficult to distinguish from chemical shift information alone. At the same time we also showed that thru the use of paramagnetic spin label one can probe weak and transient interaction. We illustrated a protocol where one can determine a bound conformation of an amino acid binding protein, glutamine binding protein (GlnBP), from a known free or apo conformation using paramagnetic label alone. We further showed in the case of glutamine free form of GlnBP, using extensive paramagnetic relaxation enhancement data, does not sample the close conformation in solution. This is in contrast to Maltose binding protein that seems to transiently sample its close conformation, with 5% population, in solution. Based on what we learned here we can extend this protocols to look at various weak interactions in proteins involved in cell signaling cascades. In parallel we also observed a unique change of scalar coupling due to the presence of paramagnetic label. We believe that one contribution is due to the polarization of the spin orbital due to the electron field at the nucleus. In addition a contribution thru relaxation mechanism due to the interference between the electron dipole and nuclear dipole can not be ruled out. This is currently an on going project to test whether one can take advantage of this for structural information. We have adapted the paramagnetic spin labeling technology that we learned to study protein complexes important for the regulation of actin polymerization. The first such complex structure that we solved was Capping protein and V-1 (Myotrophin). We showed that the V-1 binding site overlaps that of the actin barbed end on capping protein. This explains the regulation of actin polymerization by capping protein thru its binding site sequestration in the cell by V-1. We confirmed out finding by mutation studies on V-1 as well as capping protein that modulate their affinity. The second complex that we solved was between capping protein and a peptide corresponding to the CAH3a and CAH3b regions of CARMIL. We showed that the binding site of CARMIL on capping protein is extensive and there are no overlap between CARMIl and actin binding sites on capping protein. Interestingly CARMIL peptide doesn't adopt any secondary structure in the free as well as the bound forms. Furthermore we showed that in either of these two complexes the terminal tentacles of the capping protein are not involved. They have been implicated in the past as important for actin regulation. We further showed a more realistic model, based on our structure of the complex, where tight binding regions on CARMIL orient its polypeptide chain such that a interference loop between the CAH3a and CAH3b domains is positioned close to the basic patch where actin supposed to bind. This close proximity can explain the reduction of capping protein affinity to actin, but not a complete inhibition. In the future we will investigate the ternary complex of capping protein, V-1, and CARMIL. We developed a new method to characterize inter-domain motion. We applied this new approach to study the functional flexibility of a three domain modules of factor-H, which is a protein involved in immune signaling against host pathogens. We used residual dipolar coupling (rdc) measured by NMR. The rdc is an average quantity which reflects the ensemble population of structures in solution. We showed that there is a maximum of 20 degrees cone angle between these domains. We also used a shape empirical potential in the calculation to test our finding. The agreement to the rdc was worse when a shape potential is used to limit the amplitude of motion. This amplitude of motion can explain the conformation observed for this protein when it binds the target protein C3b. We are currently carrying out simulation to test stochastic diffusion under various interaction potential that can reproduce the observed amplitude of motion in factor-H. Our results indicated that the use of Model-Free approach to analyze NMR relaxation data for multi-domain proteins is still valid as long as the inter-domain motion amplitude is less than 60 degrees. We concluded a study in which we probed, by NMR, the temperature dependent of amplitude of motion of protein backbone which is related to heat capacity. We used GlnBP in the free and substrate bound form which we have studied previously. We showed that certain sites in the protein backbone showed decrease flexibility as the temperature increased. These residues include those involved in substrate binding as well as those making up the hinge region of the protein that allow domain closure upon substrate binding. This behaviour could be correlated to our earlier finding, where this protein doesn't sample close conformation in the absence of substrate, unlike other members of the family. In addition we also showed that hydrophobic residues forming ring stacking and salt bridge surrounding them also decreased their flexibility as a function of increased temperature. As a progression in developing new technology to characterize dynamic molecular events which regulate important biological function, we chose to look at retroviral capsid assembly. This protein is a part of the Gag-poly protein which is processed as part of the maturation of the virus. The assembly and disassembly of the capsid particle is crucial for viral budding from and entry into the host cell, respectively. We showed that capsid assembly occurs due to two types of distinct molecular interactions. The N-terminal beta hairpin promotes the elongation of helix 1 which forms the oligomerization interface of the capsid particle. This event occurs at a slower timescale than the dimerization that involves the C-terminal domain of the capsid. We could only established the above observations by using a barrage of NMR experiments. This is largely due to the dynamic nature of the molecular interactions. We continued to develop an improved method to do NMR resonance assignment using pattern recognition algorithm. We showed that our method can accomplish this computational task in under a second. We employed fast Fourier transform method to correct for any shifts in the resonance positions between multiple experiments. We showed that there are multiple ways to present the data so that it would be suitable for pattern recognition. We also synthesized a compound (methylated-DOTA) that can coordinate lanthanide ligand with reduced flexibility. This was done in collaborating with the Imaging Probe Development Group. The goal was to achieve a substantial increase in observable Pseudo Contact Shift (PCS) and use the information for structure determination. In addition we also showed that the methylated-DOTA-lanthanide adopts two isomers. The populations of these isomers depend on the size of the lanthanide metal being coordinated. The population ratios that we measured by observing PCS on a protein matched those obtained from HPLC on the methylated-DOTA-lanthanide. Furthermore, we are now developing a new method to extract the PCS data automatically from the NMR spectra. Plus to to search for structures in the database that match our PCS data using pattern recognition algorithm.
Showing the most recent 10 out of 37 publications
