Improvement of spectral resolution in shift-reagent-aided 23Na NMR spectroscopy in the isolated perfused rat heart system.

The level of intracellular sodium (Nai) is maintained at approximately 14 mM in healthy myocytes. When myocytes are damaged, Nai increases and therefore the level of Nai may be a means of evaluating myocardial cell integrity. A particularly useful method to monitor Nai levels is 23Na NMR spectroscopy. However, because of the isochronous nature of the extracellular sodium (Nao) and Nai NMR signals, paramagnetic lanthanide shift reagents (LSR), such as dysprosium triphosphate, Dy(PPP)7-(2), have been used to shift the Nao signal. This reveals the unshifted Nai signal and allows the NMR monitoring of Nai in isolated perfused hearts and other systems. A major shortcoming of this method (the "shift-only" method) is in the need to minimize the Nao signal by not submerging the perfused hearts in Na(+)-containing buffer. An equally undesirable alternative is the utilization of relatively high concentrations of LSR to shift a large Nao signal sufficiently to enable reasonable resolution and quantitation of Nai. We present here a method, the "shift-relaxation" method, which is a combination of using a mixture of Dy(PPP)7-(2), a shift reagent, and gadolinium triphosphate, Gd(PPP)7-(2), a relaxation agent, with data acquisition using an inversion-recovery (IR) pulse sequence. This combination allows differentiation between Nao and Nai by the difference in their respective T1 values in addition to the shift between them. With this technique we can selectively minimize the extracellular signal and therefore minimize the need for a large Dy-induced shift, as well as allow data acquisition on a heart submerged in Na(+)-containing perfusate. The resulting improved discrimination between Nai and Nao at relatively low levels of LSR should be helpful for ultimate in vivo applications and potential clinical applications, where a lower dose of LSR also means a decreased possibility of physiologically deleterious effects. Also included in this paper is a method for the quick determination of an accurate 180 degrees pulse which is required for the optimization of the IR method.