G-matrix Fourier Transform (GFT) Projection
NMR Spectroscopy

On GFT Projection NMR Spectroscopy


Defining features of G-matrix Fourier Transform (GFT) projection NMR spectroscopy (Kim and Szyperski, 2003):

(i) joint sampling of K+1 chemical shifts W0, W1, …, WK encoded in K+1 indirect evolution periods of an ND FT NMR experiment,
(ii) phase-sensitive detection of 2K linear combinations of chemical shifts W0±W1±…..±WK,
(iii) editing of the 2K linear combinations of shifts into 2K (N-K)D ‘basic’ projected sub-spectra, so that each contains one of the linear combinations.


Key technical features of GFT projection NMR:

G- or F-matrices perform step (iii) alluded to above in the time- or frequency-domain, respectively (Kim and Szyperski, 2003).

The increments for the joint sampling of shift evolution periods can be scaled (Kim and Szyperski, 2003).

Successive omission of jointly sampled shifts by setting scaling factors to zero yields the acquisition of ‘central peak sub-spectra’ of increasing order (Kim and Szyperski, 2003).

A strategy focusing on central peak sub-spectra allows one to retain the full correlation information of the parent ND experiments with a minimal number of 2K +1-1 projected sub-spectra (Kim and Szyperski, 2003).

Alternatively, a larger number of sub-spectra with differently scaled evolution periods can provide the same information. For example, one can record the basic spectra twice with different scaling factors, so that the information of the parent experiment is encoded in 2K +1 projected sub-spectra (for details, see Atreya and Szyperski, 2005).


GFT NMR experiments were implemented by our group for:

complete protein resonance assignment (Kim and Szyperski, 2004; Atreya and Szyperski, 2004),

studying proteins with extreme shift degeneracy (e.g., unfolded or membrane proteins) (Atreya et al., 2005),

using simultaneous heteronuclear NOESY (Shen et al., 2005).

A high-throughput protein structure determination protocol using GFT projection NMR was established, and its feasibility for proteins with molecular masses up to at least 20 kDa was demonstrated (Liu et al., 2005).


Support of the National Science Foundation and the
National Institutes of Health is acknowledged

Copyright © 2007 University at Buffalo. All rights reserved.