In situ 13C NMR quantification of hepatic glycogen

Direct validation of in vivo localized 13C MRS measurements of brain glycogen

With the use of localized 13C MRS in conjunction with [1-(13)C]-D-glucose infusion, it is possible to study brain glycogen metabolism in vivo. The purpose of this study was to validate in vivo 13C MRS measurements by comparing them with results from a standard biochemical assay. To increase the [1-(13)C] glycogen concentration, 11 rats were subjected to an episode of acute hypoglycemia followed by a mild hyperglycemic recovery period during which [1-(13)C]-D-glucose was infused. The total brain [1-(13)C] glycogen of the same animal was determined from the enzymatically determined total brain glycogen content, which was fixed by focused microwave irradiation (4 kW in 1.4 s) immediately after the end of the in vivo NMR measurements. The corresponding isotopic enrichment (IE) of glycogen was measured by in vitro 1H MRS of protons bound to glucose C1-alpha. The in vivo [1-(13)C] glycogen concentration was strongly correlated to the in vitro [1-(13)C] glycogen content determined by biochemical measurement in a linear manner (R=0.79). The results are consistent with the notion that localized 13C MRS measurements closely reflect 13C glycogen content in the brain.

Evaluation of muscle glycogen content by 13C NMR spectroscopy in adult-onset acid maltase deficiency

Muscle glycogen storage was measured by in vivo, natural abundance 13C nuclear magnetic resonance spectroscopy in distal and proximal lower limb segments of patients suffering from adult-onset acid maltase deficiency. Interleaved T1-weighted acquisitions of glycogen and creatine served to quantify glycogen excess. For acid maltase deficient patients (n=11), glycogen:creatine was higher than controls (n=12), (1.20+/-0.39 vs. 0.83+/-0.18, P=0.0005). Glycogen storage was above the normal 95% confidence limits in at least one site for 7/11 patients. The intra-individual coefficient of reproducibility was 12%. This totally atraumatic measurement of glycogen allows repeated measurement at different muscle sites of acid maltase deficient patients, despite selective fatty replacement of tissue. This could provide an additional parameter to follow the development of disease in individual patients, including in the perspective of forthcoming therapeutic trials. It may also offer an appropriate tool to study the role of glycogen accumulation in progression of the pathology.

Quantification of the glycogen 13C-1 NMR signal during glycogen synthesis in perfused rat liver

We studied glycogen synthesis from glucose in perfused livers of fed (n = 4) and 24 h starved (n = 7) rats. Glycogenolysis was inhibited by BAY R3401 (150 microM) and proglycosyn (100 microM). After 60 min, we replaced 99% (13)C-1 glucose by natural abundance glucose. This pulse-chase design allowed us to recognize residual ongoing futile glycogen turnover from the release of initially deposited (13)C-label, into the (13)C-free chase medium. Net residual turnover was less than 2 +/- 0.7% and 0.6 +/- 0.2% of 1-(13)C glycogen deposition rates of 0.31 +/- 0.04 and 0.99 +/- 0.04 micromol glucose g(-1) min(-1), in starved and fed livers, respectively. The 1-(13)C glycogen signal was monitored throughout the experiment with proton-decoupled (13)C NMR spectroscopy and analyzed in the time domain using AMARES. We noticed progressive line-broadening in any single experiment in the chase phase. One or a sum of two to three overlapping Lorentzians, with different exponential damping factors, were fitted to the signal. When the S/N was better than 40, the fit always delivered a small and a broad component. In the chase phase, the fit with a single Lorentzian resulted in a decline of glycogen signal by about 15 +/- 4 and 12 +/- 2% in starved and fed rats, respectively. This apparent decline in 1-(13)C glycogen signal could not be accounted for by the appearance of equivalent amounts of (13)C-labeled metabolites in the perfusate. The fit with a sum of two Lorentzians resulted in a decline of glycogen signal intensity of 7 +/- 5 and 5 +/- 3% in starved and fed rats, respectively, which reduced the apparent turnover to 8 +/- 9% and 6 +/- 4%, respectively. Quantification of the growing (13)C-1 glycogen signal requires a model function that accommodates changes in line shape throughout the period under study.

On the inhibition of hepatic glycogenolysis by fructose. A 31P-NMR study in perfused rat liver using the fructose analogue 2,5-anhydro-D-mannitol

Inhibition of hormone-stimulated hepatic glycogenolysis by fructose (Fru) has been attributed to accumulation of the competitive inhibitor Fru1P and/or to the associated depletion of the substrate phosphate (Pi). To evaluate the relative importance of either factor, we used the Fru analogue 2,5-anhydro-D-mannitol (aHMol). This analogue is avidly phosphorylated, traps Pi, and inhibits hormone-stimulated glycogenolysis, but it is not a gluconeogenic substrate, and hence does not confound glycogenolytic glucose production. Livers were continuously perfused with dibutyryl-cAMP (100 microM) to clamp phosphorylase in its fully activated a form. We administered aHMol (3.8 mM), and studied changes in glycogenolysis (glucose, lactate and pyruvate output) and in cytosolic Pi and phosphomonoester (PME), using in situ 31P-NMR spectroscopy (n = 4). Lobes of seven livers perfused outside the magnet were extracted for evaluation, by high-resolution 31P-NMR, of the evolution of aHMol1P and of aHMol(1,6)P2. After addition of aHMol, both glycogenolysis and the NMR Pi signal dropped precipitously, while the PME signal rose continuously and was almost entirely composed of aHMol1P. Inhibition of glycogenolysis in excess of the drop in Pi could be explained by continuing accumulation of aHMol1P. A subsequent block of mitochondrial ATP synthesis by KCN (1 mM) caused a rapid increase of Pi. Despite recovery of Pi to values exceeding control levels, glycogenolysis only recovered partially, attesting to the Pi-dependence of glycogenolysis, but also to inhibition by aHMol phosphorylation products. However, KCN resulted in conversion of the major part of aHMol1P into aHMol(1,6)P2. Residual inhibition of glycogenolysis was due to aHMol1P. Indeed, the subsequent withdrawal of aHMol caused a further gradual decrease in the proportion of aHMol1P (being converted into aHMol(1,6)P2, in the absence of de novo aHMol1P synthesis), and this resulted in a gradual de-inhibition of glycogenolysis, in the absence of marked changes in Pi. Glycogenolytic rates were consistently predicted by a model assuming non-saturated Pi kinetics and competition by aHMol1P exclusively: In conclusion, limited Pi availability and the presence of competitive inhibitors are decisive factors in the control of the in situ catalytic potential of phosphorylase a.

1H NMR spectroscopy study of the dynamic properties of glycogen in solution by steady-state magnetisation measurement with off-resonance irradiation

The dynamics of size-selected fractions of glycogen in solution have been investigated by proton NMR spectroscopy, using a recently described relaxation study method which relies on strong offresonance irradiation. The dependence of the steady-state magnetisation on angle and intensity of the effective radio-frequency field was measured and compared to theoretical curves derived from different models of motion. Absence or presence of contributions to relaxation from molecular motions on the microsecond time scale can be tested with this method, without having to resort to models. We found that glycogen dipolar relaxation did not result from isotropic Brownian rotation, and despite some contribution from slow motion (> 1 microsecond) to relaxation in glycogen alpha-particles extracted from rat liver, bulk movement of the molecules did not appear to participate in averaging the dipolar term to zero. Whereas hepatic glycogen rat beta-particles and commercial oyster glycogen displayed very similar relaxation properties, alpha-particles showed significantly different behaviour. However, all results were compatible with a diversity of movements within the molecule, ranging from freely rotating pyranoside rings through collective chain motion and possibly to bulk movement of the beta sub-units within the alpha-particle.

Evaluation of signal processing methods for the quantification of a multi-exponential signal: the glycogen 13C-1 NMR signal

The 13C-1 NMR peak in proton-decoupled spectra of liver glycogen solution was quantitatively analyzed by three types of model-function fitting algorithms: iterative line-fitting in the frequency domain (MDCON); iterative least-squares fitting (VARPRO) in the time domain; and noniterative singular value decomposition-based analysis (HTLS), also in the time domain. Quantification results were compared with manual integration values. Performance of the algorithms was tested at different signal-to-noise ratios (S/N) of the glycogen C-1 peak. This was achieved by varying the number of scans summed prior to analysis. Since T2 relaxation in glycogen has been shown to be multiexponential [Overloop, K. et al. Magn. Reson. Med. 36, 45-51 (1996], the exact quantification of the C-1 glycogen signal requires a model function comprising a sum of Lorentzian components, each with a different broadening at the glycogen frequency. This paper focuses on the performances of the above methods to fit such a multicomponent resonance line. In the frequency domain, line fitting with two Lorentz lines gives good results at sufficiently high S/N. In the time domain, VARPRO performs better than HTLS because fixed values can be imposed to the linewidth of the components at the common C-1 frequency, thereby reducing convergence problems at low S/N.

13C-NMR relaxation in glycogen

This study is the first report on the multiexponential T2 relaxation of the 13C-1 carbon of glycogen. In contrast to T1 relaxation, which does not display observable multiexponential decay behavior, T2 relaxation is described by a continuous distribution of T2 times. Changes in molecular weight and sample viscosity, which affect the overall mobility of the glycogen particle have little influence on T1 and T2 relaxation times. This is in contradiction with earlier results that T2 is dominated by the overall motion of the glycogen particles [L.-H. Zang Biochemistry 29, 6815-6820 (1990)]. T1 depends strongly on the external field Bo and is almost temperature independent in the range 23-37 degrees C whereas T2 is field independent and varies appreciably with temperature. The experimental T1 and T2 relaxation data are shown to be consistent with existing theoretical models for relaxation, suitably modified to include a distribution of correlation times for the internal motions. The presence of fast decaying components (short T2) in the FID implies broad line components in the frequency spectrum and the corresponding need to appropriately set the integration limits for the quantification of the glycogen peak.