Comparison of intracellular sugar-phosphate levels from 31P nmr spectroscopy of intact cells and cell-free extracts.

Estimation of intracellular intermediary metabolite levels is of fundamental importance to characterize cell metabolic processes and their regulation. Usually, intracellular intermediates are determined by stopping the cell metabolism, e.g., by immersing the cell sample in liquid nitrogen, and performing percNoric acid extracts of the cells. The metabolite levels are then obtained either by standard analytical methods'-3 or by using nuclear magnetic resonance (NMR) ~pectroscopy.~*~ Using this technique, it is possible to obtain only one point per sample because the method is destructive. Thus, transient studies are not possible on the same cell sample, and a series of aliquots are required with the assumption that all have the same dynamic behavior and are exposed to the same initial conditions. Also, information on compartmentalization within the cell is lost when extracts are prepared. For example, this prevents differentiation of compounds in the cytoplasm from those in the vacuole in yeast. In vivo determinations of cell metabolite content are possible using NMR, a technique that allows noninvasive measurement. Phosphorus-3 1 (3'P) NMR resolves small phosphate-containing metabolites in different compartments within the cell, as well as extracellular compound^.^^' Using in vivo 3'P NMR, the intracellular sugar phosphates in yeast utilizing glucose appear as a single broad peak which consists mainly of overlapping peaks of glucose-6phosphate, fructose-6-phosphate, fructose- 1,6-diphosphate, and 3-phosphoglycerate. This peak can be deconvoluted into its individual components by spectral analysis methods.' The deconvolution method consists in optimizing the correspondence between a linear combination of the individual peaks and the experimental sugar phosphate peak through a modified Levenberg-Marquardt algorithm. After