13C NMR tracers in neurochemistry: implications for molecular imaging

An overview of 13C nuclear magnetic resonance (NMR) spectroscopy methods and their applications in the study of the metabolism of brain cells in vitro and in the in vivo brain is presented as well as their implications for modern molecular imaging techniques. Various topics will be discussed, such as general properties of the 13C NMR spectrum, 13C NMR spectroscopy acquisition protocols, determination of fractional 13C enrichment, 13C(2H) NMR methodologies, and the use of 13C hyperpolarized substrates for NMR spectroscopy and imaging. Some illustrative applications are described, both in vitro and in vivo.

Sources of hepatic triglyceride accumulation during high-fat feeding in the healthy rat

Hepatic triglyceride (HTG) accumulation from peripheral dietary sources and from endogenous de novo lipogenesis (DNL) was quantified in adult Sprague-Dawley rats by combining in vivo localized (1)H MRS measurement of total hepatic lipid with a novel ex vivo (2)H NMR analysis of HTG (2)H enrichment from (2)H-enriched body water. The methodology for DNL determination needs further validation against standard methodologies. To examine the effect of a high-fat diet on HTG concentrations and sources, animals (n = 5) were given high-fat chow for 35 days. HTG accumulation, measured by in vivo (1)H MRS, increased significantly after 1 week (3.85 +/- 0.60% vs 2.13 +/- 0.34% for animals fed on a standard chow diet, P < 0.05) and was maintained until week 5 (3.30 +/- 0.60% vs 1.12 +/- 0.30%, P < 0.05). Animals fed on a high-fat diet were glucose intolerant (13.3 +/- 1.3 vs 9.4 +/- 0.8 mM in animals fed on a standard chow diet, for 60 min glycemia after glucose challenge, P < 0.05). In control animals, DNL accounted for 10.9 +/- 1.0% of HTG, whereas in animals given the high-fat diet, the DNL contribution was significantly reduced to 1.0 +/- 0.2% (P < 0.01 relative to controls). In a separate study to determine the response of HTG to weaning from a high-fat diet, animals with raised HTG (3.33 +/- 0.51%) after 7 days of a high-fat diet reverted to basal HTG concentrations (0.76 +/- 0.06%) after an additional 7 days of weaning on a standard chow diet. These studies show that, in healthy rats, HTG concentrations are acutely influenced by dietary lipid concentrations. Although the DNL contribution to HTG content is suppressed by a high-fat diet in adult Sprague-Dawley rats, this effect is insufficient to prevent overall increases in HTG concentrations.

Tricarboxylic acid cycle inhibition by Li+ in the human neuroblastoma SH-SY5Y cell line: a 13C NMR isotopomer analysis

Li+ effects on glucose metabolism and on the competitive metabolism of glucose and lactate were investigated in the human neuroblastoma SH-SY5Y cell line using 13C NMR spectroscopy. The metabolic model proposed for glucose and lactate metabolism in these cells, based on tcaCALC best fitting solutions, for both control and Li+ conditions, was consistent with: (i) a single pyruvate pool; (ii) anaplerotic flux from endogenous unlabelled substrates; (iii) no cycling between pyruvate and oxaloacetate. Li+ was shown to induce a 38 and 53% decrease, for 1 and 15 mM Li+, respectively, in the rate of glucose conversion into pyruvate, when [U-13C]glucose was present, while no effects on lactate production were observed. Pyruvate oxidation by the tricarboxylic acid cycle and citrate synthase flux were shown to be significantly reduced by 64 and 84% in the presence of 1 and 15 mM Li+, respectively, suggesting a direct inhibitory effect of Li+ on tricarboxylic acid cycle flux. This work also showed that when both glucose and lactate are present as energetic substrates, SH-SY5Y cells preferentially consumed exogenous lactate over glucose, as 62% of the acetyl-CoA was derived from [3-13C]lactate while only 26% was derived from [U-13C]glucose. Li+ did not significantly affect the relative utilisation of these two substrates by the cells or the residual contribution of unlabelled endogenous sources for the acetyl-CoA pool.

Intracellular lithium and cyclic AMP levels are mutually regulated in neuronal cells

In this work, we studied the effect of intracellular 3',5'-cyclic adenosine monophosphate (cAMP) on Li+ transport in SH-SY5Y cells. The cells were stimulated with forskolin, an adenylate cyclase activator, or with the cAMP analogue, dibutyryl-cAMP. It was observed that under forskolin stimulation both the Li+ influx rate constant and the Li+ accumulation in these cells were increased. Dibutyryl-cAMP also increased Li+ uptake and identical results were obtained with cortical and hippocampal neurons. The inhibitor of the Na+/Ca2+ exchanger, KB-R7943, reduced the influx of Li+ under resting conditions, and completely inhibited the effect of forskolin on the accumulation of the cation. Intracellular Ca2+ chelation, or inhibition of N-type voltage-sensitive Ca2+ channels, or inhibition of cAMP-dependent protein kinase (PKA) also abolished the effect of forskolin on Li+ uptake. The involvement of Ca2+ on forskolin-induced Li+ uptake was confirmed by intracellular free Ca2+ measurements using fluorescence spectroscopy. Exposure of SH-SY5Y cells to 1 mm Li+ for 24 h increased basal cAMP levels, but preincubation with Li+, at the same concentration, decreased cAMP production in response to forskolin. To summarize, these results demonstrate that intracellular cAMP levels regulate the uptake of Li+ in a Ca(2+)-dependent manner, and indicate that Li+ plays an important role in the homeostasis of this second messenger in neuronal cells.

Effects of Li+ transport and intracellular binding on Li+/Mg2+ competition in bovine chromaffin cells

Li(+) transport, intracellular immobilisation and Li(+)/Mg(2+) competition were studied in Li(+)-loaded bovine chromaffin cells. Li(+) influx rate constants, k(i), obtained by atomic absorption (AA) spectrophotometry, in control (without and with ouabain) and depolarising (without and with nitrendipine) conditions, showed that L-type voltage-sensitive Ca(2+) channels have an important role in Li(+) uptake under depolarising conditions. The Li(+) influx apparent rate constant, k(iapp), determined under control conditions by (7)Li NMR spectroscopy with the cells immobilised and perfused, was much lower than the AA-determined value for the cells in suspension. Loading of cell suspensions with 15 mmol l(-1) LiCl led, within 90 min, to a AA-measured total intracellular Li(+) concentration, [Li(+)](iT)=11.39+/-0.56 mmol (l cells)(-1), very close to the steady state value. The intracellular Li(+) T(1)/T(2) ratio of (7)Li NMR relaxation times of the Li(+)-loaded cells reflected a high degree of Li(+) immobilisation in bovine chromaffin cells, similar to neuroblastoma, but larger than for lymphoblastoma and erythrocyte cells. A 52% increase in the intracellular free Mg(2+) concentration, Delta[Mg(2+)](f)=0.27+/-0.05 mmol (l cells)(-1) was measured for chromaffin cells loaded with the Mg(2+)-specific fluorescent probe furaptra, after 90-min loading with 15 mmol l(-1) LiCl, using fluorescence spectroscopy, indicating significant displacement of Mg(2+) by Li(+) from its intracellular binding sites. Comparison with other cell types showed that the extent of intracellular Li(+)/Mg(2+) competition at the same Li(+) loading level depends on intracellular Li(+) transport and immobilisation in a cell-specific manner, being maximal for neuroblastoma cells.