Distance Measurement Between Selected Sites In Protein Molecules By EPR

The study of structures and functions of biomacromolecules (e.g. proteins) is animportant topic in modern biology research. Site directed spin labeling electronparamagnetic resonance (SDSL-EPR) technology is currently one of the principalmethods to study structures of proteins in physiological environments. SDSL-EPRuses SDSL technology to label the target sites selectively in biomolecules withnitroxides, measures the spectra of the spins on nitroxides by using EPR technology,and eventually via spectrum analysis, obtains the dynamics of the target sites and thedistance and its variation between the two spins. Its chief distinguishing features arewide ranges, anti-interference of other factors, and more effective analysis of proteinstructures and conformational changes. SDSL-EPR is capable of studying proteins insolution under the condition similar to physiological environment, and is particularlyapplicable to study the dynamic process of the conformational changes. Currently, thistechnology has solved many difficulties in biology research, such as the structuralanalysis of proteins that are difficult to crystallize, the conformational analysis ofmembrane proteins and macromolecular protein complexes, the dynamic changes ofproteins in solution, and interactions between proteins or between proteins andbiomembranes.However, there are still a few unsolved problems with SDSL-EPR distancemeasurement. Due to the diversity of biology samples, in practical studies, Continueswave EPR (CW-EPR) dipolar broadening can be caused by many factors, such asincomplete spin labeling reaction, global rotations of the biomolecules and localrotational motions of the nitroxides, relaxation effects of spins, measuring temperature,and parameters of spectral measurement, all of which may introduce errors in thedistance measurement. However, all current CW-EPR distance calculation approacheshave difficulties in completely avoiding the effects of the above factors on accuracy ofmeasurement. Therefore, how to effectively distinguish and abstract the spectralbroadening caused by dipolar-dipolar interactions between spins from an EPRspectrum, is the essential problem with the CW-EPR method that need to be solved.Considering the actuality of current SDSL-EPR method used in research ofbiomolecular structures and functions, the purpose of this work is to establish anexperimental method to measure distance between target sites in biomolecules and a new method for distance correction and calibration, and to improve the accuracy andprecision of CW-EPR distance measurement, providing a new technology to study therelationship between protein structures and functions in solution or physiologicalenvironment.Methods:(1) Based on Fourier deconvolution distance measurement (FDDM)method, a new distance calculation software that combines spectrumsimulation-Fourier deconvolution distance measurement (SS-FDDM) and Gaussfitting of the broadening function was written. In the process of optimizing spectrumsimulation, we added Gauss fitting program and diradical spectrum inverse fittingprogram, which improved the results of spectrum simulation. By adding thecorrection program of the single/double spin spectrum and broadening function in theprocess of calculating distances, the accuracy of the distance calculation wasimproved. This new distance calculation software was tested and modified byapplying it to calculating known distances of diradical compounds.(2) Experimentalmethod of measuring the spectral broadening caused by interactions between spins inminute solutio samples within the range of CW-EPR distance measurement underdifferent conditions (different solvent, concentration, temperature and power) wasestablished. A series of single/double spin compounds with rigid structure, definitemolecular dimension, and stable structure were synthesized as a distance ruler. Inaddition, distance correction and calibration method was developed under differentconditions.(3) Biological experiment models for distance measurement wereconstructed of protein molecules: the carbohydrate-recognition domains (CRD) of thehuman liver and lymph node sinusoidal endothelial cell C-type lectin (hLSECtin) wasselected as a template, in which single/double cysteines (Cys) mutants was designedin its Ca2+-dependent binding region. Single/double spins were introduced into themodel through SDSL. The model was expressed by pET28b-Tat prokaryoticexpression vector. This lays the foundation for further measurement of distancevariations and the application and verification of the established method.Results:(1) The original distance calculation method (SS-FDDM) was modifiedthrough the integrated application of EPR spectrum simulation, Gauss fitting of thebroadening functions and inverse fitting of diradical spectrum. Comparisons betweenthe results of C60biradical molecules under different conditions proved that the newcalculation method was more flexible in addressing the spectra, which extends theapplication of the spectrum simulation software, especially for the effective fitting of low temperature powder spectrum. This new method can effectively eliminate thehigh-frequency noise of the experiment spectrum, and can save the manual work ofdetermining cutoff points of the broadening function, which overcome the contrivedfactors to some extent, improving the accuracy and objectivity of the distancecalculation. The new Spectrum Simulation-Distance Measurement software cangenerally be applied to most interacting EPR spectrum analyses and deconvolutiondistance measurements.(2) Three diradical isomers based on C60structure withdifferent distances between the two spins were constructed as the distance ruler. Theresults showed that C60diradical isomer had definite molecular dimension, stableproperty, and rigid structure, which meets the requirement of the distance ruler.Moreover, it is structurally better than distance rulers with double helix nucleic acidstructure or alkyne-benzene structure used in similar works. According to thedistance measuring results under different conditions, the experimental conditions ofbroadening effects caused by dipolar-dipoalr interaction in minute solutio sampleswere optimized. Distance calibration method and correction standard were establishedunder different conditions (different solvent, concentration, temperature and power).The current range of distance calibration is8~14, and the precision of the distancemeasurement, the biggest standard deviation under all kinds of conditions, is about0.84.(3) The hLSECtin-CRD domain of single/double Cys mutation protein modelswere successfully constructed, and expressed at the high-yield and soluble expressionlevel by pET28b-Tat prokaryotic expression vector. Primary purified samples wereacquired, offering conditions for further study of the conformational changes ofhLSECtin-CRD.Conclusion: This work proposed the idea of using double spins compoundmolecules to correct and calibrate EPR distance measurement, and establishedcorresponding experimental method. New CW-EPR distance calculation software waswritten and modified. Stable double spins distance ruler was established within therange of8-14, as well as its experimental method of distance correction andcalibration. The protein model for distance measurement of single/double Cysmutants was constructed. The results of this work generally achieved the goal ofdeveloping a method for EPR distance measurement between target sites inbiomolecules.The deficiencies of this work are: current spectrum fitting program can be furtherimproved and optimized, especially the feasibility of applying the Gauss fitting method to more complicated broadening functions. Moreover, double spinscompounds with longer distance can be synthesized, further extending the scale of thedistance ruler. In addition, the practical application of the distance measurementmethod should be verified by protein experiments.