NMR Methodologies On Protein Backbone And Methyl Assignment

Solution NMR spectroscopy is one of the most powerful techniques for protein structure,interaction and dynamic studies at physiological condition.Assignment of protein backbone or methyl group signals is the fundamental,and often the toughest step for these studies.Backbone signals can provide numerous NMR probes,but these signals attenuate rapidly with the increment of protein size.Though the number of methyl signals is not so much as backbone,excellent transverse relaxation property makes them detectable by NMR even in super protein systems.Therefore,assignment of backbone and methyl signals can supply residue-level structure,interaction and dynamic information at different signal amount and protein size.Due to the fast transverse relaxation rate of 13Cp and high frequency degeneracy of 13Ca,it remains a difficulty for backbone assignment of proteins larger than 30kDa at high completeness and accuracy,even though with the help of perdeuteration and TROSY techniques aimed at reducing transverse relaxation rate.Although many methyl assignment approaches have been presented over the past 20 years,none of them is general and effective enough,especially for the assignment of proteins in the range of 30kDa-50kDa at high completeness and accuracy.Hence,we presented a covariance LCC assignment method,which can simultaneously assign the backbone and methyl of a protein at high completeness,accuracy and efficiency using only one sample.The covariance LCC method selects the most sensitive backbone spectra HN(CO)CA&HNCA and methyl spectra CA-TOCSY-(CM)HM&methyl-HMQC as original spectra,reliefing the problem of signal attenuation by increment of protein size.By constructing covariance spectra between HN(CO)CA and HNCA,HNCA and CA-TOCSY-(CM)HM,CA-TOCSY-(CM)HM and methyl-HMQC along the donor dimension 13Ca?13C?a?1HM,the problem of multiple assignment caused by chemical shift degeneracy of 13Ca?13Ca and 1HM met in traditional backbone and methyl assgiment method,is converted to the problem of identifying fake peaks in covariance spectra.We presented a filter,i.e.,the peak lineshape linear correlation coefficient(LCC),which can identifying nearly 90%fake peaks in the covariance spetra,meaning that nearly 90%problem of multiple assignments can be solved.We developed a platform for backbone and methyl assignment using the covariance LCC method on Matlab.This covariance LCC assignment method has been used to assign the backbone signals of a U-[2H,13C,15N]-MBP protein at 306.6K and a U-[1H,13C,15N]-TRIM66 protein at 302.6K,also the methyl signals of a U-[2H,13C,15N]-I(?1 only)]-MBP protein at 306.6K and a U-[2H,13C,15N]-[1(?1 only),LproR,VProR]-MBP protein at 283.6K,at high completeness and accuracy.