Validation of ¹³C NMR measurements of liver glycogen in vivo

Abstract

The natural abundance 13C NMR intensity of the glycogen C1 resonance was measured in the surgically exposed liver of rabbits in vivo (n = 17) by integration from 98 to 104 ppm and compared double blindedly to the subsequent biochemical measurement. Coil loading was measured each time from a reference sphere at the coil center and the NMR intensity was normalized accordingly. For quantification, the normalized NMR intensity was calibrated using aqueous glycogen solutions ranging from 110 to 1100 mumol glucosyl units/g (n = 14). An in vivo range from 110 to 800 mumol glucosyl units/g wet weight was measured with a highly linear correlation with concentration (r = 0.85, P < 0.001). The in vivo NMR concentration was 0.95 +/- 0.05 (mean +/- standard error, n = 17) of the concomitant enzymatic measurement of glycogen content. We conclude that the 13C NMR signal of liver glycogen C1 is essentially 100% visible in vivo and that natural abundance 13C NMR spectroscopy can provide reliable noninvasive estimates of in vivo glycogen content over the physiological range of liver glycogen concentrations when using adequate localization and integration procedures.

MRI detection of glycogen in vivo by using chemical exchange saturation transfer imaging (glycoCEST)

“…Because there are numerous practical disadvantages to measuring glycogen by tissue biopsy in humans, there has been widespread interest in detection of glycogen in vivo by 13 C magnetic resonance (MR) spectroscopy (MRS) in multiple organs, including the heart (2), liver (3), skeletal muscle (4), and brain (5). Although glycogen is a large molecule, 13 C NMR allows accurate measurements of glycogen content in liver (6) and in skeletal muscle (7), presumably because internal motions remain sufficiently unrestricted in vivo (8). Detection of glycogen using 1 H NMR also has been demonstrated (9)(10)(11).…”

Technical and experimental features of Magnetic Resonance Spectroscopy of brain glycogen metabolism

“…Differently from 1 H-MRS, the complete visibility of glycogen by 13 C-MRS has been demonstrated in vivo in the liver [71], skeletal muscle [72] and also brain [58]. In fact, because 13 C-MRS sequences applied in the aforementioned studies are essentially pulseacquire, this technique is suited for compounds with an ultrashort T 2 , such as glycogen.…”

“…However, these methods are not applicable to 13 C-MRS of cerebral glycogen due to the absence of a natural abundance 13 C signal from any highly concentrated compound in the brain. An alternative option is to reproduce 13 C-MRS measurements in vivo on a phantom containing an aqueous glycogen solution of known concentration, the socalled external reference method [71]. To accurately perform this method, a reference 13 C signal from a 13 C-enriched compound (e.g.…”

“…To accurately perform this method, a reference 13 C signal from a 13 C-enriched compound (e.g. formic acid in a closed sphere) should be provided at the 13 C coil for corrections related with coil performance, such as loading or adjustments of transmitting and receiving powers [71,97]. Other factors, including T 1 , T 2 and NOE, have to be taken in to account between the external reference phantom solution and in vivo scans.…”

Direct, noninvasive measurement of brain glycogen metabolism in humans

“…Casey et al, 2000;Kunnecke and Seelig, 1991;Shulman et al, 1990;Shulman and Rothman, 2001;Van Den Bergh et al, 2000 and references therein). 13 C NMR has been shown to provide accurate measurements of glycogen concentrations based on the C1 resonance at 100.5 ppm that is well resolved from the C1 resonances of ␣-and ␤-glucose (Gruetter et al, 1991(Gruetter et al, , 1994a. In contrast to other tissues, however, the glycogen concentrations reported in the mammalian brain range from 2 to 5 mol/g (Choi et al, 1999 and references therein), which precludes detection of natural abundance 13 C glycogen signals from individual subjects.…”