The Problematic Determination of Proton Magnetic Relaxation Rates of Protein-Containing Solutions

Abstract

16 data points were collected, with the times, , spanning the interval between 1.67 msec and a time at least nine mean half-lives longer. Times between consecutive 180° pulses were at least five mean half-lives. For transverse relaxation, 169 equally spaced data points were collected spanning the interval between time zero and at least 1.75 mean half-lives later. Because the built-in software of the relaxometer used was found to be unsatisfactory for fitting biexponential functions, data were analyzed with more modern commercial software (Statistica, StatSoft, Tulsa, Okla; Excel, Microsoft, Redmond, Wash). RESULTS Water, saline, and solutions of typical paramagnetic chelates of the kind used as extracellular MRI contrast agents showed strictly monoexponential decays of longitudinal and transverse magnetization, justifying the characterization of these decays by single relaxation rates, R1 and R2, respectively. In contrast, the presence of proteins in the same liquids caused biexponential decays. The chemical nature of the sample cell was without influence, arguing against a role of protein absorption to the walls. Here, we illustrate the situation only with protein solutions without contrast agents. The Table shows results for two pure proteins. Monoexponential fits were progressively unsatisfactory with increasing concentrations, although standard errors for the relaxation rates remained below 5%. Biexponential fits revealed a large-amplitude