Dynamic changes in glucose metabolism by lactate loading as revealed by a positron autoradiography technique using rat living brain slices

Neuroscience Letters 249 (1998) 155–158

Dynamic changes in glucose metabolism by lactate loading as revealed by a positron autoradiography technique using rat living brain slices Tetsuhito Murata a , b ,*, Atsuo Waki a, Naoto Omata a, Yasuhisa Fujibayashi a, Norihiro Sadato a, Ryoichi Yano a, Mitsuyoshi Yoshimoto a, Kiminori Isaki b, Yoshiharu Yonekura a a

Biomedical Imaging Research Center, Fukui Medical University, Matsuoka-cho, Fukui 910-1193, Japan b Department of Neuropsychiatry, Fukui Medical University, Matsuoka-cho, Fukui 910-1193, Japan Received 11 May 1998; accepted 19 May 1998

Abstract To demonstrate the preference of lactate over glucose as an energy substrate in normal brain tissue under normoxic condition, the dynamic changes in glucose uptake by lactate loading were investigated in living rat brain slices using a positron autoradiography technique. Fresh rat brain slices were incubated with [18F]2-fluoro-2-deoxy-D-glucose ([18F]FDG) in oxygenated Krebs– Ringer solution containing 10 mM glucose at 36°C. During incubation, serial two-dimensional imaging of [18F]FDG uptake in the slices was constructed on the imaging plates. Lactate loading (20 mM) reversibly suppressed the [18F]FDG accumulation up to 80 min. Compared with the pre-loading and the unloaded control values, [18F]FDG uptake was suppressed to 25–45% in cerebral regions and 6–7% in cerebellum. The lactate concentration in the surrounding medium decreased after lactate loading. Hence brain tissue preferentially uses lactate over glucose under normoxic and euglycemic condition.  1998 Published by Elsevier Science Ireland Ltd. All rights reserved

Keywords: Brain slice; Lactate; [18F]2-fluoro-2-deoxy-D-glucose; Hypoxia; Glucose uptake; Positron

The brain lactate concentration has been assumed to be stable unless hypoxic conditions cause a mismatch between glycolysis and respiration [4,14]. Recent in-vivo studies in both humans and animals, however, suggest that lactate production and accumulation occur during cerebral stimulation under normoxic conditions [2,13,20]. The possible mechanism is that astrocytes glycolytically metabolize glucose to lactate which is in turn used by neurons [12,19]. Schurr et al. [16] showed that lactate, as the sole energy substrate, can support neuronal function in hippocampal slice preparations, which finding was later reproduced by others [1,6]. Nevertheless, few studies are available as to which substrate, glucose or lactate, is preferentially utilized by brain tissue under normoxic condition. The in-vitro experimental model using brain slices differs from in-vivo systems in that a circulatory system and * Corresponding author. Tel.: +81 776 613111, ext. 2335; fax: +81 776 618137; e-mail address: [email protected]

blood–brain barrier are absent. Thus, any desired alteration in that environment can be studied in the absence of bloodborne factors with precise control over the extracellular environment of this preparation. Measurement of glucose utilization based on the glucose analogue 2-deoxyglucose (2DG) and the tracer kinetic model has become a fundamental method to assess brain function [18]. Newman et al. [10] first applied of the [14C]2DG method to living brain slices. The amount of radioactivity uptake to the tissue is, in this method, counted either with a liquid scintillation counter or by image analysis after autoradiography, only at a defined time-point for sampling after removing the free radiotracer surrounding the tissue. To extend this approach, we recently established an imaging technique in living brain slices, named ‘dynamic positron autoradiography technique (dPAT)’, utilizing positron emitter-labeled ligands as a probe and an imaging plate as a detector [8,9]. Due to the high specific radioactivity of the radiotracers, high energy of beta particles, and high sensitivity of the imaging plate,

0304-3940/98/$19.00  1998 Published by Elsevier Science Ireland Ltd. All rights reserved PII S0304- 3940(98) 00413- 3

156

T. Murata et al. / Neuroscience Letters 249 (1998) 155–158

serial two-dimensional images of radioactivity can be constructed quantitatively with a short exposure time, while the cells in the slices are still alive in the incubation solution. The purpose of the present study was to investigate the dynamic changes in glucose uptake by lactate loading. Using fresh rat brain slices and dPAT with [18F]2-fluoro2-deoxy-D-glucose([18F]FDG), we quantitatively analyzed the serial changes in the rate of [18F]FDG uptake when lactate was exogenously loaded in the medium and removed. Another experiment with similar settings was performed to estimate the amount of lactate consumption during lactate loading. Male Sprague–Dawley rats weighing 250–300 g were anesthetized with diethyl ether. After decapitation, the brains were quickly removed and immersed in oxygenated and cooled (1–4°C) Krebs–Ringer solution (having the following composition in mM concentrations: NaCl, 124; KCl, 5; MgCl2, 1; CaCl2, 2; KH2PO4, 1.2; NaHCO3, 26; glucose, 10). Brain slices of 300-mm thickness were prepared with a microslicer (DTK-2000; Dosaka EM, Kyoto, Japan). For one trial of measurement, 24 sagittally-sectioned brain slices were taken from two rats. The slices were incubated in double polystyrene chambers (outer and inner ones) as described previously [9]. Briefly, the bottom of the outer and inner chambers were cut to make a rectangular hole, over which a transparent polyvinylidene chloride film (10mm thick) or a fine nylon net (80-mm thick) was tightly stretched and fixed in place by gluing it to the side wall. The outer chamber was filled with 40 ml of Krebs–Ringer solution, in which the inner chamber, having numerous small holes (4 mm in diameter) on the side wall, was immersed. Brain slices were arranged on the nylon net of the inner chamber and were lightly fixed in place by covering them with a fine nylon net, which was stretched and glued to the upper side of a 300-mm thick stainless steel ring (2 cm in inner diameter). The bathing solution was continuously bubbled with 95% O2/5% CO2 gas supplied through fine silicone tubing introduced into the inner chamber at six sites, to promote consistent perfusion within the chamber. L(+)-lactic acid (lactate) was purchased from Sigma (St. Louis, MO) and neutralized with NaOH before use. Osmolarity was kept constant in the lactate containing Krebs–Ringer solution by adjusting the concentration of NaCl. Lactate-loading and its relief were done after transferring brain slices, using the inner chamber used as the container, to another outer chamber containing the same amount of [18F]FDG diluted with prewarmed Krebs–Ringer solution containing or not containing lactate (20 mM). 18 F was produced by 18O(p,n) 18F nuclear reaction, and 18 [ F]FDG was synthesized by the method of Hamacher et al. [3] with an automated [18F]FDG synthesis system (NKK, Tokyo, Japan). The specific radioactivity of [18F]FDG was 1–2 Ci/mmol at the end of synthesis, with the total concentration (labeled plus unlabeled) used in this study ranging from 0.47 to 1.21 mg/ml (2.6–7.1 mM). After 1-h preincubation of brain slices in Krebs–Ringer solution at 36°C (pH

7.3–7.4), the inner chamber, containing the slices, was taken out and put into another outer chamber in which [18F]FDG had been diluted at 3–4 mCi/ml with prewarmed Krebs–Ringer solution. A set of six double chambers, four slices in each of the double chambers, were put on an imaging plate (BAS-MP 2040S, Fuji, Japan), which was replaced with a new plate every 10 min. Exposed imaging plates were scanned with a bio-imaging analyzer (BAS1500, Fuji), and the images were displayed on a Macintosh computer. The pixel size was 100 mm. The linearity of the response of the imaging plates (counts/pixel) to the radioactivity was checked, with all the data presented here within this linear range. As the half-life of 18F is about 110 min, the radioactivity decayed quickly during the experiments. However, the radioactivity of 18F decreases with the same time-course both in the brain slices and in the surrounding medium solution, and it is not necessary to compensate for the radioactive decay when the radioactivity pixel value of a region of interest (ROI) is divided by that of the bathing medium (BM) [9]. Thus the relative increment in the [18F]FDG uptake (ROI − BM) can be expressed in terms of the following ratio Relative Uptake Ratio = (ROI − BM)=BM

(1) 2

where ROI is the radioactivity counts of ROI (per mm ) and BM, the average radioactivity counts measured at four places in the bathing medium surrounding each brain slice (per mm2). To depict dynamic changes of [18F]FDG uptake, the relative uptake ratio was plotted against time. Time zero (t = 0) is when [18F]FDG was introduced into the bathing medium containing brain slices. As shown in Fig. 1, values within the initial part (from t = 0 up to t = 20 min) of each plot were negative, indicating non-steady condition between the medium and the slices. After that, the uptake curves of the unloaded control (open symbols) showed a linear relationship with time (linear regression coefficient r . 0.98, from t = 60 min to t = 350 min), indicating constant FDG uptake by the brain tissue. This linear relationship is usually maintained up to 10 h under normoxic and euglycemic conditions (data not shown). This slope indicates the rate of fixed FDG6-PO4 [5,11], which in turn reflects the metabolic rate of glucose [10,11,18]. To evaluate the dynamic changes in the rate of FDG uptake by lactate loading, the slope was calculated in the pre-loading, loading, and post-loading phases separately using the linear regression analysis. Serial changes in medium lactate concentrations after loading with 20 mM lactate were measured. Eighty slices taken from the cerebella of three rats were newly prepared in a separate double chamber. Then, 60-ml samples were collected from each of eight sites in the medium solution, the lactate concentrations measured, and the mean values calculated. Taking into account evaporative water loss of the solution due to continuous incubation at 36°C and bubbling, in addition to the slice-containing chamber, measure-

157

T. Murata et al. / Neuroscience Letters 249 (1998) 155–158

Fig. 1. Effect of 20 mM lactate on the time-course curve of [18F]FDG uptake in two representative brain regions: frontal cortex (circles) and cerebellum (squares). Brain slices which remained in control solution (open symbols) and brain slices which were exposed to lactate (closed symbols). Addition and washout of lactate are indicated by arrows. Ordinate: relative uptake ratio of [18F]FDG (see text for further explanation). Abscissa: time in min. Time zero is when [18F]FDG was introduced into the bathing medium containing brain slices. Values are means obtained in 12 slices (SEM were omitted).

ments were also made of a chamber with otherwise the same conditions but not containing any slices, as a control. Lactate was measured using an enzymatic kit (TC L-Lactic Acid, Boehringer-Mannheim, Mannheim, Germany), which uses lactate dehydrogenase to convert lactate and NAD+ to pyruvate and NADH respectively. NADH was measured fluorometrically with the excitation at 340 nm and emission at 450 nm using UV-3101PC (Shimadzu, Japan). Lactate loading drastically and reversibly decreased [18F]FDG uptake of the brain with a latency of less than 20–30 min (Fig. 1). Changes in the slope of uptake curves throughout the time-course (before, during and after the application of lactate) are summarized in Table 1. The [18F]FDG uptake rate during lactate loading was markedly lower as compared to the pre-loading and the unloaded control values (25–45% in cerebral regions and 6–7% in cere-

bellum). This suggests that the majority of glucose utilization was transiently inhibited by the lactate loading. The reason why the degree of this inhibition was marked in the cerebellum may be related to structural particularities of the cerebellum (i.e. the density of neurons is higher in cerebellar than in cerebral cortices [7,15]), with further study of this issue awaited. In the present study, after relief of 80 min lactate loading the slope of the uptake curve recovered almost to baseline. This may suggest that brain tissue viability was preserved by the use of lactate in place of glucose as an energy substrate during this period. In electrophysiological studies using rat hippocampal slices it has been reported that synapse potentials did not recover in the case of ≥50 min aglycemia loading even after a 30 min recovery period with adequate oxygen supply [17]. The second experiment was designed to verify the replacement of lactate as an energy substrate during the lactate loading period. Expressing the serial changes in lactate concentrations in the two chambers (slice- and no-slice-containing) as a ratio to the respective values at time zero, at 60 and 80 min values were significantly lower in the slice-containing chamber as compared to that containing no slices (Fig. 2). These results prove that under normoxic and euglycemic conditions brain tissue preferentially uses exogenous lactate. At the start of the experiment, the total amount of lactate present in 40 ml of the medium solution (containing 20 mM lactate) was 800 mmol, approximately 56 mmol of which (amounting to about 7%) was estimated to have decreased during the 80 min period until relief of the lactate loading (based on the ratio to control: 0.93 at 80 min). Assuming that this decreased portion is consumed by all the slices, the amount of lactate estimated to be consumed by the slices was 0.73 mmol/g per min (calculating from the total weight of the 80 cerebellar slices measured after the end of the experiment). Taking as a reference the glucose utilization value of 0.3–0.4 mmol/g/min calculated from the total tissue radioactivity measured by whole-slice homogenization using the [14C]2DG method and living brain slices

Table 1 Changes in the slope of [18F]FDG uptake curves throughout the time-course (before, during and after the application of 20 mM lactate) Brain region

Treatment

Before

Frontal cortex

Control Lactate-load Control Lactate-load Control Lactate-load Control Lactate-load Control Lactate-load

1.23 1.18 1.28 1.44 1.38 1.47 1.28 1.18 1.42 1.48

Striatum Thalamus Hippocampus Cerebellum

± ± ± ± ± ± ± ± ± ±

During 0.18 0.10 0.19 0.20 0.18 0.22 0.14 0.15 0.14 0.13

1.24 0.37 1.09 0.37 1.14 0.42 1.04 0.46 1.28 0.08

± ± ± ± ± ± ± ± ± ±

After 0.16 0.12* 0.14 0.07* 0.16 0.08* 0.21 0.09* 0.12 0.04*

1.28 1.32 1.04 1.33 1.05 1.16 1.10 1.04 1.14 1.18

± ± ± ± ± ± ± ± ± ±

0.13 0.18 0.15 0.19 0.14 0.16 0.18 0.14 0.13 0.15

Slope values (per min × 1000) were obtained from the slope of the regression equation (Y = aX + b) fitted to [18F]FDG uptake curves using the linear regression analysis. Y, relative uptake ratio defined by the Eq. (1) (see text); X, time after introducing [18F]FDG into exposure chambers; a, slope of the line; b, intercept. All values are means ± SEM obtained in 12 slices. *P , 0.01, significantly different from the control values and from those of before and after the application of lactate (Mann–Whitney U test and one-way ANOVA with Scheffe’s post-hoc comparison test, respectively).

158

T. Murata et al. / Neuroscience Letters 249 (1998) 155–158

Fig. 2. Serial changes in medium lactate concentrations after loading with 20 mM lactate. Lactate concentrations of the chambers containing slices (squares), and as a control chambers without slices (circles) were simultaneously measured at each time-point and expressed as a relative ratio to the respective values at time zero (10 min after the start of loading). Values are means ± SEM obtained in four separate experiments. *P , 0.05, significantly different from the control values (Mann–Whitney U test).

[10], and considering glucose 1 mol and lactate 2 mol equivalent as energy substrates, it was thought that almost all the energy obtained from glucose utilization was compensated for by lactate. Many in-vivo studies indicate the occurrence of metabolic uncoupling during physiological activation [2,10], in which an increased glucose utilization is not matched by a commensurate increase in oxygen consumption. These data converge to indicate that neuronal activity is linked to glycolysis leading to lactate production [19]. The set of observations [12,19] provides a cellular localization for the site of lactate production, in which glucose taken up by astrocytes is converted glycolytically into lactate, which is then released into the extracellular space to be utilized by neurons. The present study showed that exogenously-supplied lactate is utilized in preference to glucose by living brain slices in which the microenvironment of the cells is well conserved. This may support the concept that lactate production by glial cells is to provide a more preferable energy substrate for neurons. The behavior of lactate from this aspect deserves much attention, apart from the conventional view of lactate only as a harmful end-product of anaerobic glycolysis during oxygen lack. This study was supported in part by the research Grant (JSPS-RFTF97L00203) for the ‘Research for the Future’ Program from the Japan Society for the Promotion of Science. [1] Bueno, D., Azzolin, I.R. and Perry, M.L.S., Ontogenic study of glucose and lactate utilization by rat cerebellum slices, Med. Sci. Res., 22 (1994) 631–632. [2] Fox, P.T., Raichle, M.E., Mintun, M.A. and Dence, C., Nonoxidative glucose consumption during focal physiologic neural activity, Science, 241 (1988) 462–464. [3] Hamacher, K., Coenen, H.H. and Stocklin, G., Efficient stereospecific synthesis of no-carrier-added 2-[F-18]-fluoro-2-deoxyD-glucose using aminopolyether supported nucleophilic substitution, J. Nucl. Med., 27 (1986) 235–238.

[4] Hossmann, K.-A. and Kleihues, P., Reversibility of ischemic brain damage, Arch. Neurol., 29 (1973) 375–382. [5] Huang, S.C., Phelps, M.E., Hoffman, E.J., Sideris, K., Selin, C.J. and Kuhl, D.C., Noninvasive determination of local cerebral metabolic rate of glucose in man, Am. J. Physiol., 128 (1980) E69–E82. [6] Izumi, Y., Benz, A.M., Zorumski, C.F. and Olney, J.W., Effects of lactate and pyruvate on glucose deprivation in rat hippocampal slices, NeuroReport, 5 (1994) 617–620. [7] Lange, W., Cell number and cell density in the cerebellar cortex of man and some other mammals, Cell Tissue Res., 157 (1975) 115–124. [8] Matsumura, K., Bergstro¨m, M., Onoe, H., Takechi, H., Westerberg, G., Antoni, G., Bjurling, P., Jacobson, G.B., La˚ngstro¨m, B. and Watanabe, Y., In vitro positron emission tomography (PET): use of positron emission tracers in functional imaging in living brain slices, Neurosci. Res., 22 (1995) 219–229. [9] Murata, T., Matsumura, K., Onoe, H., Bergstro¨m, M., Takechi, ¨ gren, M., H., Sihver, S., Sihver, W., Neu, H., Andersson, Y., O Fasth, K.-J., La˚ngstro¨m, B. and Watanabe, Y., Receptor imaging technique with 11C-labeled receptor ligands in living brain slices: its application to time-resolved imaging and saturation analysis of benzodiazepine receptor using [11C]Ro15–1788, Neurosci. Res., 25 (1996) 145–154. [10] Newman, G.C., Hospod, F.E. and Patlak, C.S., Brain slice glucose utilization, J. Neurochem., 51 (1988) 1783–1796. [11] Phelps, M.E., Huang, S.C., Hoffman, E.J., Sokoloff, L. and Kuhl, D.E., Tomographic measurement of local cerebral glucose metabolic rate in humans with (F-18)2-fluoro-2-deoxy-Dglucose: validation of method, Ann. Neurol., 6 (1979) 371– 388. [12] Poitry-Yamate, C.L., Poitry, S. and Tsacopoulos, M., Lactate released by Mu¨ller glial cells is metabolized by photoreceptors from mammalian retina, J. Neurosci., 15 (1995) 5179– 5191. [13] Prichard, J., Rothman, D., Novotny, E., Petroff, O., Kuwabara, T., Avison, M., Howseman, A., Hanstock, C. and Shulman, R., Lactate rise detected by 1H NMR in human visual cortex during physiologic stimulation, Proc. Natl. Acad. Sci. (USA), 88 (1991) 5829–5831. [14] Rehncrona, S., Rosen, I. and Siesjo¨, B.K., Brain lactic acidosis and ischemic cell damage: 1. Biochemistry and neurophysiology, J. Cereb. Blood Flow Metab., 1 (1981) 297– 311. [15] Rockel, A.J., Hiorns, R.W. and Powell, T.R.S., The basic uniformity in structure of the neocortex, Brain, 103 (1980) 221– 244. [16] Schurr, A., West, C.A. and Rigor, B.M., Lactate-supported synaptic function in the rat hippocampal slice preparation, Science, 240 (1988) 1326–1328. [17] Schurr, A., West, C.A. and Rigor, B.M., Electrophysiology of energy metabolism and neuronal function in the hippocampal slice preparation, J. Neurosci. Meth., 28 (1989) 7–13. [18] Sokoloff, L., Reivich, M., Kennedy, C., Des Rosiers, M.H., Patlak, C.S., Pettigrew, K.D., Sakurada, O. and Shinohara, M., The [14C]deoxyglucose method for the measurement of local cerebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat, J. Neurochem., 28 (1977) 897–916. [19] Tsacopoulos, M. and Magistretti, P.J., Metabolic coupling between glia and neurons, J. Neurosci., 16 (1996) 877–885. [20] Ueki, M., Linn, F. and Hossmann, K.A., Functional activation of cerebral blood flow and metabolism before and after global ischemia of rat brain, J. Cereb. Blood Flow Metab., 8 (1988) 486–494.