NMR imaging of 15N at natural abundance

Abstract

*‘N observation at natural abundance (0.37% of total nitrogen) has become conventional in NMR spectroscopic studies on liquids since the first report in 197 1 ( 1) . This technique has also been useful for solving problems in solids (2). Experiments involving enriched material have proved to be important in labeling studies in chemical applications (3)) in determination of protein structure (4)) and in improving the understanding of plant metabolism (5,6). In the case of plants, crude spatial selection has previously been achieved by physical sectioning of plants which had been fed with “N-labeled nutrients ( 5 ) . No 15N NMR images from samples at either natural abundance or enriched concentration have previously been reported, and only one 14N image (of liquid nitrogen) had been published ( 7) until recently (8). Given the ubiquity of nitrogen in metabolic processes and the interest in the processes of nitrogen fixation and nitrogen turnover in the environment, it seemed sensible to tackle the particular problems associated with ‘*N NMR imaging. These, as in 15N NMR spectroscopic studies, arise from the low isotopic abundance and the low magnetogyric ratio, which give the 15N isotope a receptivity of only 3.85 X 10e6 compared to ‘H. An additional drawback is the long relaxation times frequently encountered ( 9). This renders accumulation difficult but does at least ensure that “N lines are normally narrow. A faster accumulation rate may, however, become beneficial if relaxation times are shortened by the use of paramagnetic reagents or if steady-state techniques are employed. The maximum concentration of 15N spins at natural abundance occurs in liquid nitrogen, for which the molarity of i5N nuclei is 0.2 1 M. Additionally, low temperature (77 K) gives a favorable Boltzmann gain relative to 298 K of about 4 as well as reduced thermal noise to give an overall gain of approximately 7.5. The signal-to-noise ratio was measured to be approximately 10: 1 following a single 90” pulse and the linewidth was less than 70 Hz. The relaxation times are presumably shortened by contamination with oxygen, the amount of which varies from sample to sample. The sample was contained in a two-section, unsilvered Dewar and was bubbling while under investigation so that the signal-to-noise ratio was reduced by removal of magnetized material from the active volume of the RF coil. The upper section functioned as a reservoir and was 8 cm high (od, 6.5 cm) and ended in a finger (id, 1.7 cm and height, 7.5 cm) which was inserted into a lo-turn surface coil (id, 3.0 cm) made of 0.8 mm insulated copper wire. There was sufficient volume of liquid nitrogen in the reservoir to keep the finger full for 3 1 hours.