Analysis of time courses of metabolic precursors and products in heterogeneous rat brain tissue: limitations of kinetic modeling for predictions of intracompartmental concentrations from total tissue activity

Does shuttling of glycogen-derived lactate from astrocytes to neurons take place during neurotransmission and memory consolidation?

Glycogen levels in resting brain and its utilization rates during brain activation are high, but the functions fulfilled by glycogenolysis in living brain are poorly understood. Studies in cultured astrocytes have identified glycogen as the preferred fuel to provide ATP for Na⁺ ,K⁺ -ATPase for the uptake of extracellular K⁺ and for Ca²⁺ -ATPase to pump Ca²⁺ into the endoplasmic reticulum. Studies in astrocyte-neuron co-cultures led to the suggestion that glycogen-derived lactate is shuttled to neurons as oxidative fuel to support glutamatergic neurotransmission. Furthermore, both knockout of brain glycogen synthase and inhibition of glycogenolysis prior to a memory-evoking event impair memory consolidation, and shuttling of glycogen-derived lactate as neuronal fuel was postulated to be required for memory. However, lactate shuttling has not been measured in any of these studies, and procedures to inhibit glycogenolysis and neuronal lactate uptake are not specific. Testable alternative mechanisms to explain the observed findings are proposed: (i) disruption of K⁺ and Ca²⁺ homeostasis, (ii) release of gliotransmitters, (iii) imposition of an energy crisis on astrocytes and neurons by inhibition of mitochondrial pyruvate transport by compounds used to block neuronal monocarboxylic acid transporters, and (iv) inhibition of astrocytic filopodial movements that secondarily interfere with glutamate and K⁺ uptake from the synaptic cleft. Evidence that most pyruvate/lactate derived from glycogen is not oxidized and does not accumulate suggests predominant glycolytic metabolism of glycogen to support astrocytic energy demands. Sparing of blood-borne glucose for use by neurons is a reasonable explanation for the requirement for glycogenolysis in neurotransmission and memory processing.

MRS glucose mapping and PET joining forces: re-evaluation of the lumped constant in the rat brain under isoflurane anaesthesia

Although numerous positron emission tomography (PET) studies with (18) F-fluoro-deoxyglucose (FDG) have reported quantitative results on cerebral glucose kinetics and consumption, there is a large variation between the absolute values found in the literature. One of the underlying causes is the inconsistent use of the lumped constants (LCs), the derivation of which is often based on multiple assumptions that render absolute numbers imprecise and errors hard to quantify. We combined a kinetic FDG-PET study with magnetic resonance spectroscopic imaging (MRSI) of glucose dynamics in Sprague-Dawley rats to obtain a more comprehensive view of brain glucose kinetics and determine a reliable value for the LC under isoflurane anaesthesia. Maps of Tmax /CMRglc derived from MRSI data and Tmax determined from PET kinetic modelling allowed to obtain an LC-independent CMRglc . The LC was estimated to range from 0.33 ± 0.07 in retrosplenial cortex to 0.44 ± 0.05 in hippocampus, yielding CMRglc between 62 ± 14 and 54 ± 11 μmol/min/100 g, respectively. These newly determined LCs for four distinct areas in the rat brain under isoflurane anaesthesia provide means of comparing the growing amount of FDG-PET data available from translational studies.

Voxel-based estimation of kinetic model parameters of the L-[1-(11)C]leucine PET method for determination of regional rates of cerebral protein synthesis: validation and comparison with region-of-interest-based methods

We adapted and validated a basis function method (BFM) to estimate at the voxel level parameters of the kinetic model of the L-[1-(11)C]leucine positron emission tomography (PET) method and regional rates of cerebral protein synthesis (rCPS). In simulation at noise levels typical of voxel data, BFM yielded low-bias estimates of rCPS; in measured data, BFM and nonlinear least-squares parameter estimates were in good agreement. We also examined whether there are advantages to using voxel-level estimates averaged over regions of interest (ROIs) in place of estimates obtained by directly fitting ROI time-activity curves (TACs). In both simulated and measured data, fits of ROI TACs were poor, likely because of tissue heterogeneity not taken into account in the kinetic model. In simulation, rCPS determined from fitting ROI TACs was substantially overestimated and BFM-estimated rCPS averaged over all voxels in an ROI was slightly underestimated. In measured data, rCPS determined by regional averaging of voxel estimates was lower than rCPS determined from ROI TACs, consistent with simulation. In both simulated and measured data, intersubject variability of BFM-estimated rCPS averaged over all voxels in a ROI was low. We conclude that voxelwise estimation is preferable to fitting ROI TACs using a homogeneous tissue model.

Quantification of cerebral glucose metabolic rate in mice using 18F-FDG and small-animal PET

The aim of this study was to evaluate various methods for estimating the metabolic rate of glucose utilization in the mouse brain (cMR(glc)) using small-animal PET and reliable blood curves derived by a microfluidic blood sampler. Typical values of (18)F-FDG rate constants of normal mouse cerebral cortex were estimated and used for cMR(glc) calculations. The feasibility of using the image-derived liver time-activity curve as a surrogate input function in various quantification methods was also evaluated.

METHODS

Thirteen normoglycemic C57BL/6 mice were studied. Eighteen blood samples were taken from the femoral artery by the microfluidic blood sampler. Tissue time-activity curves were derived from PET images. cMR(glc) values were calculated using 2 different input functions (one derived from the blood samples [IF(blood)] and the other from the liver time-activity curve [IF(liver)]) in various quantification methods, which included the 3-compartment (18)F-FDG model (from which the (18)F-FDG rate constants were derived), the Patlak analysis, and operational equations. The estimated cMR(glc) value based on IF(blood) and the 3-compartment model served as a standard for comparisons with the cMR(glc) values calculated by the other methods.

RESULTS

The values of K(1), k(2), k(3), k(4), and K(FDG) estimated by IF(blood) and the 3-compartment model were 0.22 +/- 0.05 mL/min/g, 0.48 +/- 0.09 min(-1), 0.06 +/- 0.02 min(-1), 0.025 +/- 0.010 min(-1), and 0.024 +/- 0.007 mL/min/g, respectively. The standard cMR(glc) value was, therefore, 40.6 +/- 13.3 micromol/100 g/min (lumped constant = 0.6). No significant difference between the standard cMR(glc) and the cMR(glc) estimated by the operational equation that includes k(4) was observed. The standard cMR(glc) was also found to have strong correlations (r > 0.8) with the cMR(glc) value estimated by the use of IF(liver) in the 3-compartment model and with those estimated by the Patlak analysis (using either IF(blood) or IF(liver)).

CONCLUSION

The (18)F-FDG rate constants of normal mouse cerebral cortex were determined. These values can be used in the k(4)-included operational equation to calculate cMR(glc). IF(liver) can be used to estimate cMR(glc) in most methods included in this study, with proper linear corrections applied. The validity of using the Patlak analysis for estimating cMR(glc) in mouse PET studies was also confirmed.

Regional rates of cerebral protein synthesis measured with L-[1-11C]leucine and PET in conscious, young adult men: normal values, variability, and reproducibility

We report regional rates of cerebral protein synthesis (rCPS) measured with the fully quantitative L-[1-(11)C]leucine positron emission tomography (PET) method. The method accounts for the fraction (lambda) of unlabeled amino acids in the precursor pool for protein synthesis derived from arterial plasma; the remainder (1-lambda) comes from tissue proteolysis. We determined rCPS and lambda in 18 regions and whole brain in 10 healthy men (21 to 24 years). Subjects underwent two 90-min dynamic PET studies with arterial blood sampling at least 2 weeks apart. Rates of cerebral protein synthesis varied regionally and ranged from 0.97+/-0.70 to 2.25+/-0.20 nmol/g per min. Values of rCPS were in good agreement between the two PET studies. Mean differences in rCPS between studies ranged from 9% in cortical regions to 15% in white matter. The lambda value was comparatively more uniform across regions, ranging from 0.63+/-0.03 to 0.79+/-0.02. Mean differences in lambda between studies were 2% to 8%. Intersubject variability in rCPS was on average 6% in cortical areas, 9% in subcortical regions, and 12% in white matter; intersubject variability in lambda was 2% to 8%. Our data indicate that in human subjects low variance and highly reproducible measures of rCPS can be made with the L-[1-(11)C]leucine PET method.

Functional imaging of focal brain activation in conscious rats: impact of [(14)C]glucose metabolite spreading and release

Labeled glucose and its analogs are widely used in imaging and metabolic studies of brain function, astrocyte-neuron interactions, and neurotransmission. Metabolite shuttling among astrocytes and neurons is essential for cell-cell transfer of neurotransmitter precursors and supply and elimination of energy metabolites, but dispersion and release of labeled compounds from activated tissue would reduce signal registration in metabolic labeling studies, causing underestimation of focal functional activation. Processes and pathways involved in metabolite trafficking and release were therefore assessed in the auditory pathway of conscious rats. Unilateral monotonic stimulation increased glucose utilization (CMR(glc)) in tonotopic bands in the activated inferior colliculus by 35-85% compared with contralateral tissue when assayed with [(14)C]deoxyglucose (DG), whereas only 20-30% increases were registered with [1- or 6-(14)C]glucose. Tonotopic bands were not evident with [1-(14)C]glucose unless assayed during halothane anesthesia or pretreatment with probenecid but were detectable with [6-(14)C]glucose. Extracellular lactate levels transiently doubled during acoustic stimulation, so metabolite spreading was assessed by microinfusion of [(14)C]tracers into the inferior colliculus. The volume of tissue labeled by [1-(14)C]glucose exceeded that by [(14)C]DG by 3.2- and 1.4-fold during rest and acoustic activation, respectively. During activation, the tissue volume labeled by U-(14)C-labeled glutamine and lactate rose, whereas that by glucose fell 50% and that by DG was unchanged. Dispersion of [1-(14)C]glucose and its metabolites during rest was also reduced 50% by preinfusion of gap junction blockers. To summarize, during brain activation focal CMR(glc) is underestimated with labeled glucose because of decarboxylation reactions, spreading within tissue and via the astrocyte syncytium, and release from activated tissue. These findings help explain the fall in CMR(O2)/CMR(glc) during brain activation and suggest that lactate and other nonoxidized metabolites of glucose are quickly shuttled away from sites of functional activation.

Astrocyte activation in vivo during graded photic stimulation

Astrocytes have important roles in control of extracellular environment, de novo synthesis of neurotransmitters, and regulation of neurotransmission and blood flow. All of these functions require energy, suggesting that astrocytic metabolism should rise and fall with changes in neuronal activity and that brain imaging can be used to visualize and quantify astrocytic activation in vivo. A unilateral photic stimulation paradigm was used to test the hypothesis that graded sensory stimuli cause progressive increases in the uptake coefficient of [2-(14)C]acetate, a substrate preferentially oxidized by astrocytes. The acetate uptake coefficient fell in deafferented visual structures and it rose in intact tissue during photic stimulation of conscious rats; the increase was highest in structures with monosynaptic input from the eye and was much smaller in magnitude than the change in glucose utilization (CMR(glc)) by all cells. The acetate uptake coefficient was not proportional to stimulus rate and did not correlate with CMR(glc) in resting or activated structures. Simulation studies support the conclusions that acetate uptake coefficients represent mainly metabolism and respond to changes in metabolism rate, with a lower response at high rates. A model portraying regulation of acetate oxidation illustrates complex relationships among functional activation, cation levels, and astrocytic metabolism.

Glycolysis versus TCA cycle in the primate brain as measured by combining 18F-FDG PET and 13C-NMR

The glycolytic flux (cerebral metabolic rate of glucose CMRglc) and the TCA cycle flux (VTCA) were measured in the same monkeys by 18F-fluorodeoxyglucose (18F-FDG) positron emission tomography (PET) and 13C NMR spectroscopy, respectively. Registration of nuclear magnetic resonance (NMR) and PET data were used for comparison of CMRglc and VTCA in the exact same area of the brain. Both fluxes were in good agreement with literature values (CMRglc=0.23+/-0.03 micromol/g min, VTCA=0.53+/-0.13 micromol/g min). The resulting [CMRglc/VTCA] ratio was 0.46+/-0.12 (n=5, mean+/-s.d.), not significantly different from the 0.5 expected when glucose is the sole fuel that is completely oxidized. Our results provide a cross-validation of both techniques. Comparison of CMRglc with VTCA is in agreement with a metabolic coupling between the TCA cycle and glycolysis under normal physiologic conditions.

Measurement of regional rates of cerebral protein synthesis with L-[1-11C]leucine and PET with correction for recycling of tissue amino acids: I. Kinetic modeling approach

Measurements of regional rates of cerebral protein synthesis (rCPS) require correction for the effect of recycling of tissue amino acids back into the precursor pool for protein synthesis. The fraction of the precursor pool derived from arterial plasma, lambda, can be evaluated as the steady-state ratio of the specific activity of leucine in the tissue tRNA-bound fraction to that in arterial plasma. While lambda can be directly measured in terminal experiments in animals, an alternative method is required for use with PET. We report a method to estimate lambda based on a kinetic model of labeled and unlabeled leucine and labeled CO2 in the tissue. The kinetic model is also used to estimate the amount of labeled protein and rCPS. We measured time courses of [14C]leucine, [14C]protein, and 14CO2 in the blood and brain of anesthetized rats and estimated parameters of the kinetic model from these data. Simulation studies based on the kinetic parameters were then performed to examine the feasibility of this approach for use with L-[1-11C]leucine and PET. Lambda and rCPS were estimated with low bias, which suggests that PET can be used for quantitative measurement of rCPS with L-[1-11C]leucine and a kinetic modeling approach for correction for recycling of tissue amino acids.