Quantitative estimation of lactate in the brain by 1H NMR

Detection of mobile proteins by proton nuclear magnetic resonance spectroscopy in the guinea pig brain ex vivo and their partial purification

Proton nuclear magnetic resonance (1H NMR) spectroscopy was used to study metabolites of the brain cortex ex vivo. The superfused brain cortex preparation was judged to be metabolically viable on the basis of the 31P NMR spectrum (intracellular pH of 7.23 +/- 0.03 and phosphocreatine/ATP ratio of 1.21 +/- 0.09). Using 1H NMR a group of previously unidentified signals was detectable at 0.94, 1.22, and 1.40 ppm with a water-suppressed spin-echo sequence. These signals had shorter spin-spin relaxation times (51-54 ms) than N-acetylaspartate and lactate (84-93 ms) and also smaller saturation factors, an indication of shorter spin-lattice relaxation times than the latter two low-molecular-weight metabolites. The unidentified signals also displayed homonuclear coupling to other spins in the methine region of the spectrum. Acid extraction of the brain slices or cortex from animals that were killed yielded a mixture of proteins that exhibited NMR properties matching the 1H NMR signals in the brain cortex. The molecular mass of these thermoresistant, "mobile" proteins, which contained proline plus hydroxyproline (9-16% of all amino acids), ranged between 8 and 40 kDa. These "new" assignments of 1H NMR-detectable compounds may influence interpretation of NMR data of some metabolites, as their signals are in the vicinity of the -CH3 1H NMR peaks of lactate and alanine.

Localized magnetic resonance spectroscopy measurement of brain lactate during intravenous lactate infusion in healthy volunteers

Proton magnetic resonance spectroscopy (1H MRS) localized to the left temporal-parietal region in 8 healthy volunteers detected a 2.1-fold +/- 0.7-fold increase (all values +/-SD) in brain lactate during intravenous infusion of 0.5 molar (M) sodium lactate (5 meq/kg over 20 minutes). Significant increases in brain lactate occurred within 5-10 minutes after starting lactate infusion, progressively rose during the infusion, then decreased towards baseline levels during 30 minutes post-infusion. Venous lactate concentration increased from 0.8 +/- 0.2 mM to 10.9 +/- 4.1 mM or 13.6-fold during the infusion. Flow phantom findings in vitro suggest attenuation of 1H MRS blood lactate signal from arteries and veins as a result of flow velocity effects. Correlations between paired blood and brain lactate measurements at each sampling time indicate a non-linear relationship between compartments during lactate infusion.

Three-dimensional 1H spectroscopic imaging of cerebral metabolites in the rat using surface coils

Three dimensional metabolite maps of protonated metabolites were obtained using 1H magnetic resonance spectroscopic imaging at 7 T. Surface coils were used to increase sensitivity and spatial resolution significantly over a volume coil two-dimensional acquisition. Adiabatic pulses were employed to provide homogeneous B1 excitation and frequency selective refocusing over the volume of the rat brain. These techniques were employed to obtain three-dimensional spectroscopic imaging spectra from nominal voxel volumes of 9-30 microliters from rat brain. The improved spatial resolution and sensitivity are also demonstrated with studies of focal ischemia in the rat.

Nondestructive detection of glutamate by 1H nuclear magnetic resonance spectroscopy in cortical brain slices from the guinea pig: evidence for changes in detectability during severe anoxic insults

31P and 1H nuclear magnetic resonance spectroscopy (NMR) was used to study the metabolism of intact superfused cortical brain slices during normoxia and anoxia. Attention was focused on quantification of 1H NMR-detected glutamate by a water-suppressed spin-echo method, using N-acetyl aspartate as an internal concentration reference. To quantify the 1H NMR signals, the spin-spin relaxation times and saturation effects were estimated for given metabolites. In addition, absolute concentrations of metabolites were determined by biochemical methods from acid extracts of the preparations after NMR experiments. Under aerobic conditions, 1H NMR detected 79% of the glutamate determined biochemically from the brain slice extracts. During anoxia in the absence of glucose when a severe energetic failure was evident, both 1H NMR and biochemical assays gave closely matching levels for glutamate. We conclude that in the brain cortex 21% of glutamate is located in an intracellular compartment in which this amino acid does not contribute to the 1H NMR signal. However, during severe anoxia an intracellular reorganisation occurs increasing the detectability of this amino acid neurotransmitter by NMR.

Quantitation of metabolites by 1H NMR

In this report we describe the three factors that need to be measured when quantitating an edited resonance relative to an internal standard using a surface coil. These factors are necessary by virtue of the water, fat suppressing, and localization schemes used in studing a 1H metabolite. First, use of semi-selective pulses requires amplitude correction of the edited and reference resonances. Second, use of a single surface coil results in different sensitive volumes for different resonances due to the inhomogeneous B1 and therefore require separate acquisition of the resonances. Third, editing pulses alter the sensitive volumes and this correction must be made internally by applying the same editing pulse to the reference resonance. A rationalization of this correction is given in terms of rotation operators. We apply these corrections to quantitate edited lactate relative to total creatine in a MnCl2-doped phantom and find 91% rather than 145% of known concentration. In human skeletal muscle in vivo after exhaustive exercise, we measured the lactate after exercise and found it to be 27.2 mM in two experiments, in reasonable agreement with literature values for the given exercise protocol.

The visibility of the 1H NMR signal of ethanol in the dog brain

In vivo, high-resolution, volume-selected 1H NMR spectroscopy was used to monitor the concentration of ethanol in the dog brain following intravenous injection of ethanol. Equilibration of ethanol in the body water should result in approximately equivalent concentrations of ethanol in the blood and brain. However, the mean equilibrium brain ethanol concentration determined using N-acetylaspartate as an internal standard was only 23 +/- 5% of the blood ethanol concentration. The disparity between blood and brain ethanol concentrations was attributed to underestimation of the ethanol concentration due to overlapping resonances with NAA and to T2 attenuation or possible nondetection of the 1H signal from ethanol bound at the surface of cell membranes and partitioned into the hydrophobic core of membrane lipids.

ACE: a single-shot method for water-suppressed localization and editing of spectra, images, and spectroscopic images

A versatile method for localized (1H) NMR spectroscopy is presented. The method intrinsically combines B0-based spatial localization with the possibility of water suppression and spectral editing. With this sequence it is feasible to localize not only single spectra but also phase-encoded images and spectroscopic images. The technique essentially integrates the "Hahn spin-echo" with the "stimulated echo" sequence and is therefore called ACE (acquiring combined echoes). It realizes water-suppressed three-dimensional localization in a single shot and can be used for localized shimming. Studies in which the new method is applied to phantoms with metabolites diluted at low concentrations are presented. Discrimination between lactate and alanine, employing an adapted spectral editing method with complete inversion, combined with simultaneous water suppression and localization of a 0.06-cc volume is shown. The suppression of signals from outside the selected volume is greater than or equal to 24,000. Also, the method is demonstrated by in vivo experiments at 6.3 T. Localized water-suppressed 1H spectra are obtained completely noninvasively, leaving scalp and fur intact, from well-defined volumes of 0.15 cc in the brain of a living rat. Water-suppressed spectroscopic imaging over a localized volume with "body" coil excitation and noninvasive surface coil detection yielded spectra from voxels as small as 25 microliters in the in vivo rat brain.

Surface coil spectroscopic imaging: time and spatial evolution of lactate production following fluid percussion brain injury

Detailed temporal and spatial distributions of lactate production are presented for graded fluid-percussion brain injury in the rat. A one-dimensional proton spin-echo spectroscopic imaging (1D SESI) technique, performed with a surface coil, is presented and evaluated. This technique, which represents a practical compromise, provides spatially localized proton nuclear magnetic resonance (NMR) brain spectra from a series of small voxels (less than 0.15 cm3) in less than 10 min, thus enabling both spatial and temporal monitoring of lactate production. These high-resolution lactate maps are correlated with hyperintense regions observed in T2-weighted images taken 10 h after impact, which, in turn, correlate with histology. The data demonstrate that, following severe trauma there is delayed production and propagation of lactate to regions of the brain that are remote from the trauma site. The extent of lactate production depends on the severity of impact. More significantly, the data show that following severe trauma, local lactate concentrations exceed 15 mumol/g, the concentration that has been claimed as the threshold for brain injury. Therefore high lactate levels cannot be ruled out a priori as a possible factor in brain injury following severe head trauma.

In vivo volume-selective metabolite editing via correlated z-order

Volume-selected 1H NMR spectroscopy was combined with spectral editing to selectively detect brain metabolites. The SPACE localization sequence was used to create a voxel of zeta-magnetization which could then be edited for any scalar coupled metabolite by the use of selective excitation in the ECZOTIC sequence to generate longitudinal spin order. The sequence returns an edited signal with no intrinsic loss of magnetization. The method was applied to observe approximately 10 mM ethanol and 17 mM lactate in the brain of a dog.

Changes in brain metabolism during hyperammonemia and acute liver failure: results of a comparative 1H-NMR spectroscopy and biochemical investigation

The effects of hyperammonemia on brain function have been studied in three different experimental models in the rat: acute liver ischemia, urease-treated animals and methionine sulfoximine-treated animals. To quantify the development of encephalopathy, clinical grading and electroencephalographic spectral analysis were used as indicators. In all three experimental models brain ammonia concentrations increased remarkably associated with comparable increases in severity of encephalopathy. Furthermore, in vivo 1H-nuclear magnetic resonance spectroscopy of a localized cerebral cortex region showed a decrease in glutamate concentration in each of the aforementioned experimental models. This decreased cerebral cortex glutamate concentration was confirmed by biochemical analysis of cerebral cortex tissue post mortem. Furthermore, an increase in cerebral cortex glutamine and lactate concentration was observed in urease-treated rats and acute liver ischemia rats. As expected, no increase in cerebral cortex glutamine was observed in methionine sulfoximine-treated rats. These data support the hypothesis that ammonia is of key importance in the pathogenesis of acute hepatic encephalopathy. Decreased availability of cerebral cortex glutamate for neurotransmission might be a contributing factor to the pathogenesis of hyperammonemic encephalopathy. A surprising new finding revealed by 1H-nuclear magnetic resonance spectroscopy was a decrease of cerebral cortex phosphocholine compounds in all three experimental models. The significance of this finding, however, remains speculative.

Improved quantification of in vivo 1H NMR spectra by optimization of signal acquisition and processing and by incorporation of prior knowledge into the spectral fitting

Quantification of localized in vivo brain 1H spectra is in general very difficult due to excessive spectral overlap. In addition, intensity distortions may result from the effects of the NMR pulse sequence on the spins. This paper describes an approach to solving these problems. It comprises optimization of the pulse sequence; correction of the experimental lineshape; determination of intensity distortions, of relative line positions, and of linewidths using model solutions; and incorporation of the thus obtained prior knowledge into a nonlinear least-squares spectral fitting procedure. This approach resulted in greatly improved accuracy, precision, and reliability of the quantitation of our in vivo spectra of rat brain, and enabled us to estimate absolute metabolite concentrations.

Cerebral energy metabolism and intracellular pH during severe hypoxia and recovery: a study using 1H, 31P, and 1H [13C] nuclear magnetic resonance spectroscopy in the guinea pig cerebral cortex in vitro

1H and 31P nuclear magnetic resonance spectroscopy was used to study intracellular pH (pHi), high-energy phosphates, lactate, and amino acids in cortical brain slices superfused in Krebs-Henseleit bicarbonate buffer during and after severe hypoxia at 0, 10, and 50 mM glucose. An extensive drop in phosphocreatine (PCr) and a rapid build-up of intracellular lactate and H+ were the first signs of hypoxia. Adenosine triphosphate (ATP) was significantly more resistant to hypoxia provided that glucose was present. In the preparations that had been exposed to hypoxia in the presence of glucose, PCr became detectable within 2 min of reoxygenation, and both PCr and ATP concentrations were restored to 72-80% of normoxic levels within 30 min. Lactate was washed out, and pHi returned to normal within 4-8 min. Using 1-[13C]glucose as a tracer, we demonstrated that the rate of lactate production in the immediate posthypoxic period was at the prehypoxic level, indicating that the elevated lactate during this period was due solely to that produced during hypoxia. During reoxygenation of the preparations that were denied glucose during hypoxia, only 30% of total creatine + PCr and 18% of PCr were restored, and ATP was not detectable. The lactate concentration rose twofold in this period, and pHi became significantly more alkaline than before the hypoxic insult. Thus acute metabolic damage was considerably greater if glucose was absent during the insult, suggesting that either anaerobic ATP production or low pH may exert some protective effect against acute cell damage.

A double quantum coherence transfer proton NMR spectroscopy technique for monitoring steady-state tumor lactic acid levels in vivo

If proton nuclear magnetic resonance (1H NMR) spectroscopy is to provide a clinically useful modality for monitoring tumor growth and treatment, the technique must be able to unambiguously detect steady-state metabolite concentrations in human tumors and differentiate these from normal tissue levels. To address this problem, a two-dimensional double quantum coherence transfer spectroscopy (2DDQCT) method was developed and tested in a series of tumor cell lines implanted in mice. Lactate-edited proton NMR spectra were determined from a roughly 1-cm3 region of interest in EMT6, RIF-1, and fibroma. In two-dimensional data matrix representations of the 2DDQCT experiments (double quantum frequency on the vertical axis and chemical shift on the horizontal axis) the lactate signal (330 Hz with the transmitter set at the water resonance) was well-resolved from lipid (480 Hz, 600 Hz). The resolution in the double quantum dimension was also sufficient to conclude that a detectable level of alanine, which would reside at 358 Hz, was not present in the three tumor types. Following the NMR experiment, tumors were chemically assayed for lactate giving 8.17, 9.1, and 6.73 mumols/g wet wt for RIF-1, EMT6, and fibroma, respectively. This technique is likely to provide a noninvasive method for monitoring the steady-state lactic acid levels in small tumors before and after therapy, as well as in tissues with impaired oxygen delivery using clinical and research NMR systems.

Proton magnetic resonance spectroscopy of human brain in vivo in the evaluation of multiple sclerosis: assessment of the load of disease

Image localized, water-suppressed proton magnetic resonance spectra were obtained from affected brain in patients with multiple sclerosis. In patients with moderate to severe chronic disease, spectra revealed a decreased ratio of N-acetylaspartate to creatine resonance intensities. A normal ratio was obtained from a large recently symptomatic MRI plaque that resolved without sequelae. We propose that the observed metabolite changes can be useful as an index of irreversible CNS injury.

In vivo measurements of ethanol concentration in rabbit brain by 1H magnetic resonance spectroscopy

In vivo 1H magnetic resonance spectroscopy was used to measure the cerebral ethanol concentration in the rabbit after both intraarterial and intragastric administration. There was good agreement between cerebral and blood ethanol concentrations at all times after administration by either route. Cerebral ethanol levels, measured using in vivo 1H spectroscopy, agreed well with those measured in perchloric acid extracts of brain, analyzed by both high-resolution 1H spectroscopy and gas chromatography. Ethanol may be useful as an indicator to measure cerebral blood flow by 1H spectroscopy and chemical shift-selective magnetic resonance imaging.

QUALITY: quantification improvement by converting lineshapes to the Lorentzian type

A method of time domain deconvolution, QUALITY, is described for general use in quantitative NMR spectroscopy. By division of the experimental NMR time domain signal by a reference signal, which may be obtained either in a separate experiment or via inverse Fourier transformation of an isolated single resonance in the experimental spectrum, perfect Lorentzian lineshapes can be obtained regardless of the magnetic field inhomogeneity. Experiments, both in vivo using a surface coil and in vitro using a surface coil and a HR NMR probe, show excellent performance of the method.

Observation of cerebral metabolites in an animal model of acute liver failure in vivo: a 1H and 31P nuclear magnetic resonance study

Acute liver failure was induced in rats by a single intragastric dose of carbon tetrachloride. This causes hepatic centrilobular necrosis, as indicated by histological examinations, and produces a large increase in the activity of serum alanine aminotransferase. The plasma NH4+ level (mean +/- SEM) was 123 +/- 10 microM in the control group and 564 +/- 41 microM in animals with acute liver failure (each n = 5). 31P nuclear magnetic resonance (NMR) was used to monitor brain cortical high-energy phosphate compounds, Pi, and intracellular pH. 1H NMR spectroscopy was utilised to detect additional metabolites, including glutamate, glutamine, and lactate. The results show that the forebrain is capable of maintaining normal phosphorus energy metabolite ratios and intracellular pH despite the metabolic challenge by an elevated blood NH4+ level. There was a significant increase in the brain glutamine level and a concomitant decrease in the glutamate level during hyperammonaemia. The brain lactate level increased twofold in rats with acute liver failure. The results indicate that 1H NMR can be used to detect cerebral metabolic changes in this model of hyperammonaemia, and our observations are discussed in relation to compartmentation of NH4+ metabolism.