Measurement of mobility distributions for vesicles by electrophoretic NMR

Abstract

High-resolution electrophoretic NMR (ENMR) resolves NMR spectra of mixtures on the basis of the electrophoretic mobilities of contributing ions ( 1-4). This method is particularly convenient for the study of supramolecular systems, e.g., microemulsions, micelles, and vesicles. These systems tend to give large signals because of the large number of nuclei involved, the relatively long nuclear relaxation times, and the small mass diffusion coefficients. They are also polydisperse, and one of the aims of current research is to determine size and charge distributions for such particles. In this Communication we report the direct determination of velocity distributions for phospholipid vesicles in electric fields by a type of ENMR experiment previously suggested but not demonstrated (2). These vesicles consist of a single phospholipid bilayer in the form of a spherical shell that encloses an aqueous region ( 5). The ENMR experiments monitor both phospholipid protons in the bilayer and protons in entrapped molecules. Signals from water and from small molecules external to the vesicles are easily quenched by magnetic field gradients because of the large diffusion coefficients. The preliminary experiments reported here make use of the STE-ENMR sequence shown in Fig. 1 rather than the standard ENMR spin-echo sequence because T, b T2 for protons in the bilayer and because signals from entrapped molecules are degraded by the effects of J modulation (3). The new feature in this experiment is that 2D ENMR spectra are produced by Fourier transformations with respect to t2 and Edc rather than t2 and t I as in previously reported 2D NMR (2). The matched gradient pulses g, and g3 have areas G6, where G is the amplitude of the magnetic field gradient and 6 is the duration; and the gradient pulse gz in the time interval tl may be introduced to eliminate interference from the primary echo ( PE) and the secondary echoes ( SE lSE3). In the experiments described here it was more convenient to filter out unwanted coherence pathways by phase-cycling methods. The single-quantum coherence-transfer pathways for the stimulated-echo experiment are displayed in Ref. (6) with the exception of SE2 where p = 0 + 0 + + 1 + 1. Many different phase-cycling schemes are satisfactory. For sample I (described below) we cycled all three RF pulses with N,,* = 2 and ZV3 = 4 in the notation of Ref. (6) to eliminate PE, SEl-SE3, and various multiquantum pathways, while for sample II we cycled only RF pulse 3 with i’V3 = 4 to eliminate SE 1 -SE3 and experienced no problems.